Proefschrift Ramekers by Nicole Nijhuis

Assessment and Preservation of Auditory Nerve Integrity in the Deafened Guinea Pig Dyan Ramekers 64 ISBN 978-94-6108-814-7

UITNODIGING voor het bijwonen van de openbare verdediging van het proefschrift

Assessment and Preservation of Auditory Nerve Integrity in the Deafened Guinea Pig door Dyan Ramekers

op donderdag 27 november 2014 om 12.45 uur in de Senaatszaal van het Academiegebouw van de Universiteit Utrecht Domplein 29 te Utrecht Receptie na afloop van de promotie in Zaal 1636 in het Academiegebouw

Paranimfen: Renate Buijink (renatebuijink@gmail.com) Sarah Havenith (s.havenith@umcutrecht.nl) Dyan Ramekers Boerhaavelaan 86 3552 CZ Utrecht 06 18031396 d.ramekers@umcutrecht.nl

Assessment and Preservation of Auditory Nerve Integrity in the Deafened Guinea Pig

Dyan Ramekers

Publication of this thesis was sponsored by: MED-EL GmbH, Innsbruck, Austria Foundation of Scientific Research in Otorhinolaryngology Utrecht “ORLU” Department of Otorhinolaryngology and Head & Neck Surgery, University Medical Center Utrecht Dutch Association of Otorhinolaryngology and Head & Neck Surgery Brain Center Rudolf Magnus © 2014 Dyan Ramekers ISBN 978-94-6108-814-7 Layout and printed by Gildeprint, Enschede, the Netherlands Cover image: eCAP and aABR waveforms positioned over spiral ganglion cell photomicrographs Cover design inspired by Stanley Donwood’s London Views

Assessment and Preservation of Auditory Nerve Integrity in the Deafened Guinea Pig Bepaling en behoud van de functionaliteit van de gehoorzenuw in de dove cavia (met een samenvatting in het Nederlands)

Proefschrift ter verkrijging van de graad van doctor aan de Universiteit Utrecht op gezag van de rector magnificus, prof. dr. G.J. van der Zwaan, ingevolge het besluit van het college voor promoties in het openbaar te verdedigen op donderdag 27 november 2014 des middags te 12.45 uur. door

Dyan Ramekers

geboren op 4 februari 1983 te Veldhoven

Promotor: Copromotoren:

Prof. dr. W. Grolman Dr. H. Versnel Dr. S.F.L. Klis

Voor mijn vader

Table of contents List of abbreviations Chapter 1. Chapter 2. Chapter 3. Chapter 4. Chapter 5. Chapter 6. Chapter 7. Chapter 8. Chapter 9.

General introduction

Spiral ganglion cell morphology in guinea pigs after deafening and neurotrophic treatment

Auditory-nerve responses to varied inter-phase gap and phase duration of the electric pulse stimulus as predictors for neural degeneration

109

Temporary neurotrophic treatment prevents deafness-induced auditory nerve degeneration and preserves functionality

169

General discussion

225

Neurotrophins and their role in the cochlea

The peripheral processes of spiral ganglion cells after intracochlear application of brain-derived neurotrophic factor in deafened guinea pigs

Recovery characteristics of the electrically stimulated auditory nerve in deafened guinea pigs: relation to neuronal status

139

Spiral ganglion cell loss following ototoxically induced hair cell loss: simultaneous rather than retrograde degeneration

207

Nederlandse samenvatting

235

Dankwoord

241

Curriculum vitae

247

List of publications

251

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

List of abbreviations

List of abbreviations 1WD, one week deaf 2WD, two weeks deaf 6WD, six weeks deaf 14WD, fourteen weeks deaf aABR, acoustically evoked auditory brainstem response ABR, auditory brainstem response aFGF, acidic fibroblast growth factor BDNF, brain-derived neurotrophic factor CAP, compound action potential CI, cochlear implant CP, central process (of an SGC) CNTF, ciliary neurotrophic factor eABR, electrically evoked auditory brainstem response eCAP, electrically evoked compound action potential FGF, fibroblast growth factor GDNF, glial cell line-derived neurotrophic factor HP, habenula perforata IHC, inner hair cell IPG, inter-phase gap IPI, inter-pulse interval MPI, masker-probe interval NGF, nerve growth factor NH, normal hearing NT-3, neurotrophin-3 NT-4/5, neurotrophin-4/5 NTF, neurotrophic factor NTR, neurotrophin receptor OHC, outer hair cell OSL, osseous spiral lamina p75NTR, p75 neurotrophin receptor PBS, phosphate-buffered saline PD, phase duration peSPL, peak equivalent sound pressure level PP, peripheral process (of an SGC) SGC, spiral ganglion cell SNHL, sensorineural hearing loss TrkA-C, tropomyosin-related kinase A-C

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

CHAPTER 1 General introduction

We do not rely solely upon science and reason, because these are necessary rather than sufficient factors, but we distrust anything that contradicts science or outrages reason. – Christopher Hitchens, “God is not great”

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 1

1. The ear: more than meets the eye

The ear is a peripheral sensory organ that transforms sound pressure waves into electrical information and feeds it to the brain where the electrical signal is processed and sound is perceived. The ear comprises much more than the visible external part, the pinna, which together with the auditory canal forms the outer ear; it includes the tympanic membrane (or ear drum) and the three middle ear ossicles, together forming the middle ear, and finally the snail-shell-like inner ear, or cochlea, which together with the vestibular system characterized by its three semicircular canals forms the osseous labyrinth (Fig. 1). vestibular system auditory ossicles auditory canal

auditory nerve

cochlea tympanic membrane pinna

Fig. 1. Schematic overview of the anatomy of the ear. Image courtesy of MED-EL, Innsbruck, Austria. Fig. 1. Schematic overview of the anatomy of the ear.

Image courtesy of MED-EL, Innsbruck, Austria. The main function of the outer ear is conducting sound to the tympanic membrane, thereby allowing the vulnerable inner ear to be protected in the petrous part of the temporal bone, but its specific shape gives it an additional important function. By amplifying or attenuating specific frequencies, the shape of the pinna modulates the reflection of sound waves toward the tympanic membrane. Since this modulation depends on the direction of the incoming sound wave and thus on the location of the sound source, the outer ear enables us to localize sound sources in the vertical plane. The tympanic membrane vibrates in response to sound pressure waves and passes on these vibrations to the middle ear ossicles: malleus, incus, and stapes, respectively. In turn, the stapes conveys the vibrations to the oval window membrane of the inner ear. Because the tympanic membrane is much larger than the oval window membrane and because of the lever arm difference between malleus and incus, this mechanical

General introduction

transduction is accompanied by a more than twentyfold increase in pressure. This amplification is necessary since at the oval window membrane sound pressure waves are transferred from air to fluid, which would otherwise result in reflection of a substantial part of the sound energy. The ossicles in the middle ear therefore fulfill an important role; however, in the absence of (functional) ossicles hearing is impaired but not completely obliterated. In the osseous labyrinth two separate functions are carried out by distinct but interconnected organs: the vestibular system, making use of gravity and inertia, controls balance and spatial orientation; the cochlea is the essential part of the peripheral hearing organ. The function of the cochlea is to transform sound pressure waves into an electrical signal that is subsequently transmitted to the brain. The process by which this is realized can be divided into a mechanical and an electrochemical phase. The transition between these two phases may be considered to be the essential step in sound processing in the ear and it is carried out by the cochlear hair cells. Since the cochlea is a complex structure, both functionally and anatomically, and because it is the central anatomical structure in this thesis, it requires a more comprehensive account, as outlined below.

2. The cochlea The complexity of the cochlea is not only a question of function; its three-dimensional structure adds to this complexity and a detailed description is essential to fully understand the function of the cochlea. As mentioned above, the entry of a sound wave into the cochlea is the oval window membrane, set in motion by the stapes footplate. Behind this membrane is a fluid-filled compartment called the scala vestibuli, which spirals upward like a spiral staircase (“scala” is Latin for ladder) from the base of the cochlea to the apex. At the most apical part of the cochlea the scala vestibuli has an open connection with a second compartment, the scala tympani, which spirals downward the same two and a half turns and ends at the round window membrane in the outer wall of the cochlea, close to the oval window. Vibrations of the oval window membrane thus cause a pressure wave that travels up and down the cochlea and finally causes the round window membrane to vibrate in antiphase. Figure 2A shows a longitudinal (midmodiolar) section of a guinea pig cochlea, in which eight transections of the scalae are visible from base to apex. The transverse section in Fig. 2B gives a more detailed depiction of the mutually continuous scalae, which are both filled with perilymph – a high Na+ and low K+ fluid, resembling the extracellular cerebrospinal fluid found in the central nervous system. In between, a third scala – the scala media – is located, filled with 13

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 1

intracellular fluid-resembling low Na+ and high K+ endolymph. The scalae are separated by flexible membranes – Reissner’s membrane and the basilar membrane – so that pressure waves travelling through the perilymph cause these membranes to vibrate. Whereas Reissner’s membrane merely has a barrier function, the basilar membrane supports the organ of Corti; a structure that contains the four rows of several thousands of cochlear hair cells (Fig. 3A). From base to apex the basilar membrane becomes wider and more flexible, so that it is set in motion by low-frequency vibrations at the base and by high frequencies at the apex. This separation of frequencies along a spatial axis is referred to as tonotopy.

A SM

tectorial membrane

organ of Corti

basilar membrane

Reissner’s membrane SV

M AN B

PPs

ST 200 µm

20 µm

SGCs

Fig. 2. Histological sections of a guinea pig cochlea. A Midmodiolar section in which eight Fig. 2. Histological of are a guinea cochlea. A Midmodiolar which transections throughsections the scalae visible pig from the cochlear base (B) tosection its apexin(A). The eight axons through the are (M) visible cochlear nerve base (AN). (B) toThe its black apex rectangle (A). The oftransections the SGCs pass through thescalae modiolus andfrom form the the auditory indicates thatthrough is shown inmodiolus (B). B Transection of the scalae at the lower axons of the the region SGCs pass the (M) and form thethree auditory nerve (AN). Themiddle black (third) turn.indicates The organ Corti isthat situated on theinbasilar thescalae scala tympani rectangle theof region is shown (B). Bmembrane, Transectioninofbetween the three at the (ST) andmiddle the scala media (SM); scalaofmedia scala vestibuli (SV) aremembrane, separated by lower (third) turn. Thethe organ Corti and is situated on the basilar in Reissner’s between membrane. The IHCs in the organ of Corti synapse with the peripheral processes (PPs) of the the scala tympani (ST) and the scala media (SM); the scala media and scala vestibuli (SV) are SGCs, of which the cell bodies are located in Rosenthal’s canal. The tectorial membrane is slightly separated bydue Reissner's membrane. IHCs in the organ of Corti synapse with the peripheral bent upward to the tissue fixatingThe process. processes (PPs) of the SGCs, of which the cell bodies are located in Rosenthal's canal. The tectorial membrane is slightly bent upward due to the tissue fixating process.

The hair cells owe their name to the bundle of hair-like stereocilia, which reach up from the cell soma toward the tectorial membrane – a structure that is fixed at the medial side of the scala media, and protrudes just far enough to cover the most laterally positioned hair cells. Two types of hair cells are distinguished: there is one row of inner (IHCs) and three rows of outer hair cells (OHCs). The IHCs are the actual (passive) sensory receptors of sound pressure waves, while OHCs possess the ability to 14

General introduction

actively influence the vibration that is perceived by the IHCs by being able to elongate and contract along their basoapical axis. The IHCs convert the vibration caused by a sound wave into an electrochemical signal. As the basilar membrane vibrates up and down, the endolymph in between the reticular lamina, covering the organ of Corti, and the tectorial membrane is moved along the mediolateral axis (i.e., in the horizontal direction in Fig. 3A) by shear force and the stereocilia are displaced along with the endolymph. This displacement causes nonspecific cation channels to be forced open, through which K+ ions enter the hair cell. The change in membrane potential caused by this ion influx causes the opening of voltagegated Ca2+ channels and the influx of Ca2+ mediates the release of the IHC’s chemical signal – probably the neurotransmitter glutamate – onto the peripheral processes of the spiral ganglion cells (SGCs; see Fig. 2B). The large potential between the endolymph at the apex of the hair cell and the perilymph at its base causes rapid depolarization of the cell and rapid repolarization once the neurotransmitter has been released. The cell bodies of the SGCs are located in Rosenthal’s canal, which spirals around the center of the cochlea – the modiolus – medial to the scalae and the organ of Corti. From this canal their peripheral processes (PPs) extend outward to the organ of Corti, while their central processes (or axons) extend through the modiolus to form the majority of the auditory nerve which projects to the brain. The majority of SGCs are socalled type-I SGCs, which receive sensory input from the inner hair cells. Approximately 5-10% are type-II SGCs, which are associated with the active mechanical responses of OHCs. Neurotransmitter released into the synaptic cleft between IHC and the PP of an SGC binds to ligand-gated ion channels, which subsequently mediate the initiation of an action potential that propagates toward the SGC cell body, and subsequently along the axon toward the brain. The auditory nerve exits the cochlea at its base; together with the vestibular nerve it forms the vestibulocochlear or VIIIth cranial nerve.

3. Hearing: the perception of sound The information that is encoded in the cochlea and which is subsequently conveyed to the brain distinguishes three aspects that are crucial for understanding sound. First, temporal information is encoded by the time patterns of action potentials. Second, loudness is encoded by the frequency with which action potentials are conveyed to the brain. Additionally, if the amplitude of the vibration of the basilar membrane becomes larger, IHCs over an increased portion of the basilar membrane are activated; bandwidth therefore correlates with loudness as well. Third, the gradient in width and stiffness of the basilar membrane allows for the separation of frequencies and thus of pitch. Since 15

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 1

type-I SGCs convey information from a single IHC to the brain, the aforementioned tonotopy is preserved not only in the auditory nerve, but all the way through to the primary auditory cortex (Moerel et al., 2014). Several nuclei in the brainstem process the auditory information before it reaches the primary auditory cortex. These nuclei perform functions such as binaural integration (combining information from both ears), multisensory integration and sound localization in the horizontal plane. Since the tonotopy is preserved in these brainstem nuclei, the more complex processing of, for example, speech and music is believed to be performed solely at the cortical level, where the actual perception of sound is realized.

10 Âľm

PPs

Fig. 3. Histological sections of the organ of Corti in the guinea pig cochlea. A A healthy organ of Corti guinea pig cochleaofasCorti in Fig.in 2A,the B). guinea Arrows indicate the nuclei Fig.in3.a normal-hearing Histological sections of(same the organ pig cochlea. A Aof the hair cells, arrow heads indicate their stereocilia (gray: OHCs; black: IHC). B Organ of Corti of an healthy organ of Corti in a normal-hearing guinea pig (same cochlea as in Fig. 2A, ototoxically deafened guinea pig. After the destruction of hair cells the organ of Corti is collapsed Arrows indicatedegeneration the nucleiofofSGCs thecan hair cells, arrow heads indicate their asB). well. The secondary be inferred from the absence of peripheral processes. Both A andOHCs; B represent transections of the of lower middle turn of the cochlea. stereocilia (gray: black: IHC). B Organ Corti of an ototoxically deafened

guinea pig. After the destruction of hair cells the organ of Corti is collapsed as well. The secondary degeneration of SGCs can be inferred from the absence of processes. Both A loss and Band represent transections of the lower middle 4.peripheral Sensorineural hearing cochlear implantation turn of the cochlea.

A 2012 report by the World Health Organization estimates that 360 million people worldwide (5.3% of the world population) suffer from disabling hearing loss (WHO, 2012). Defects of any of the components in the auditory pathway, from pinna to cortex, may lead to hearing loss in one form or another, as can be logically deduced from their function in the normal situation. However, by far the most common cause of severe to complete hearing loss is the loss, absence, or dysfunction of hair cells; this condition is referred to as sensorineural hearing loss (SNHL; Fig. 3B). As SNHL is essentially a 16

General introduction

description of a symptom, it has a variety of genetic or environmental causes, including meningitis, antibiotic ototoxicity and noise exposure. Whereas conventional hearing aids can alleviate the discomfort for most cases of moderate losses essentially by amplification, they are rendered useless in case of severe SNHL. For those who do not benefit from hearing aids, cochlear implantation is currently the only available therapy to restore the sense of hearing. Cochlear implants (CIs) can be seen as a replacement for the mechanical part of the ear up to, and including, the IHCs. In short, a CI converts sound to electric pulses with which it stimulated the SGCs directly. The external part of the device consists of a microphone, an audio processor and a magnetic coil which transmits the processed signal to the subcutaneously placed receiver/stimulator (Fig. 4). The stimulator generates electric pulses that are conducted through an electrode lead that is inserted into the scala tympani of the cochlea, either through the round window membrane or via a cochleostomy as depicted in Fig. 4. As can be seen in Fig. 2B the electrode array inserted into the scala tympani is in close proximity to its target tissue – the SGCs in Rosenthal’s canal. With electrical stimulation the representation of timing and loudness (the first two aspects of sound encoding as discussed in the previous section) can easily be achieved. In addition, since the electrode array can be inserted all the way up toward the apex, SGCs in specific regions along the spiral axis can be stimulated by selectively using the contacts on the array, so that the third aspect of sound encoding – pitch discrimination by means of tonotopy – can be restored to a certain degree as well. The number of people worldwide who have received a CI has recently been estimated to be around 300,000 (Géléoc and Holt, 2014). While the first experimental CIs implanted in the 1950s and 60s provided little benefit for the recipients (Eshraghi et al., 2012), their present-day successors often allow CI users to have a telephone conversation (Géléoc and Holt, 2014). Unfortunately, performance may vary considerably among CI users (Peterson et al., 2010). Several factors are known to contribute to this variability, such as etiology of hearing loss and age at onset of hearing loss (relative to both language development and implantation), but these do not have accurate explanatory value. Technological and surgical developments in for example speech encoding strategies or atraumatic electrode insertion are crucial for improving CI performance in general; these developments are aided by a better understanding of the biology of hearing, of hearing loss, and of neurostimulation. In addition, studying the biological implications of cochlear implantation might elucidate the variability in performance among CI users.

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 1

transmission coil receiver/ stimulator

audio processor

electrode array electrode lead

Fig. 4. Schematic depiction of the components of a cochlear implant. Sound picked up by a microphone built into the processordepiction is encoded. signal through the skin Fig.audio 4. Schematic of The the encoded components ofisasent cochlear to the subcutaneously placed receiver, the stimulator sends the electric to the contacts implant. Soundand picked up by a microphone builtpulses into the in the cochlea via the electrode lead. Image courtesy of MED-EL, Innsbruck, Austria.

audio processor is encoded. The encoded signal is sent through the skin to the subcutaneously placed receiver, and the stimulator sends the electric pulses to the contacts in the cochlea via the electrode lead. Image courtesy of MEDDegenerationEL,of spiralAustria. ganglion cells following sensorineural Innsbruck,

5. hearing loss

In the healthy cochlea, the organ of Corti provides neurotrophic support for the SGCs by secreting growth factors. The two most predominant growth factors are the structurally homologous neurotrophins brain-derived neurotrophic factor (BDNF) and neurotrophin-3 (NT-3). These neurotrophins bind to specific receptors in the SGC membrane (tropomyosin-related kinase [Trk] B and TrkC, for BDNF and NT-3, respectively) and thereby mediate a multitude of intracellular pathways, including proliferation. Severe loss of hair cells results in the collapse and degeneration of the organ of Corti and as a consequence the neurotrophic support is discontinued (Ernfors et al., 1995; Fritzsch et al., 1999; Zilberstein et al., 2012). Accordingly, in animal studies induction of SNHL consistently shows significant degeneration of SGCs (Ylikoski et al., 1974; Spoendlin, 1975; Webster and Webster, 1981; Versnel et al., 2007; Fig. 3B). In humans a direct relation between SGC degeneration and loss of hair cells is difficult to confirm, but post-mortem histological analysis of CI usersâ&#x20AC;&#x2122; cochleas has shown that SNHL is often accompanied by substantial SGC loss (Fayad and Linthicum, 2006). A recent study has furthermore shown that speech perception is correlated with the number of surviving SGCs (Seyyedi et al., 2014). 18

General introduction

As these findings illustrate, it is important to protect SGCs from secondary degeneration as a result of hair cell loss. In the past two decades numerous studies have shown that, when exogenously applied, a variety of neurotrophic factors (including BDNF and NT-3) has the potential to prevent SGC degeneration in animal models of SNHL (e.g., Shah et al., 1995; Ylikoski et al., 1998; Gillespie et al., 2004). A comprehensive literature review on the role of neurotrophins in the developing and the mature cochlea, as well as on application of exogenous neurotrophins in animal models of SNHL, can be found in chapter 2 of this thesis (Ramekers et al., 2012).

6. Electrophysiology in CI users The electrical activity of neurons causes small voltage changes in the neuronal tissue that can be measured from relatively large distances; a well-known example of this principle is electroencephalography (EEG). Using similar techniques the activity of the brainstem (auditory brainstem responses; ABR) or the auditory cortex (cortical auditory evoked potentials) can be recorded in a brief time window immediately following an auditory stimulus. These recordings can be used to study the neurobiology of auditory processing in the brain, but they are specifically useful in clinical diagnostics, such as neonatal hearing screening. In CI users these electrophysiological recordings provide a useful objective contribution to implant fitting and to the assessment of CI performance in general, in addition to the more subjective psychophysical and cognitive tests (e.g., Smoorenburg et al., 2002; Willeboer and Smoorenburg, 2006; Kim et al., 2009; RungeSamuelson et al., 2009). Besides measurements of electrical activity in the central nervous system, the activity of the SGCs in the auditory periphery (compound action potentials; CAPs) can be recorded as well. This is however not possible without invasive recording techniques, which renders it generally impractical for clinical purposes. CI users form an exception, since CIs possess a telemetry function, which means that in addition to using the intracochlear electrodes for stimulation of the SGCs they can also be used to record the neural response. Since the CAPs recorded from CIs are without exception evoked by electrical instead of acoustic stimuli, these are referred to as the electricallyevoked compound action potential (eCAP). The eCAP is a valuable addition to evoked potentials recorded from the brain, as it records the direct unprocessed responses from the electrically stimulated SGCs. Therefore, it can be used to assess stimulation efficacy separately for each of the electrode contacts throughout the cochlea, which may reflect the functionality of the contacts, the electrical resistance of the tissue in between the electrode and the target SGCs, and, moreover, it may reflect the condition of the SGC 19

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 1

population. It is however difficult â&#x20AC;&#x201C; if not impossible â&#x20AC;&#x201C; to distinguish between these underlying factors. A recent study for instance showed that the steepness of the slope of the eCAP growth curve correlates well with speech perception (Kim et al., 2010). While it might be logical to induce that a steeper slope implies a larger maximum eCAP amplitude and therefore a larger SGC population, the evidence for this kind of assumptions must come from animal research.

7. Electrophysiology in animals In animal research a wider range of electrophysiological techniques is applied to study the auditory system, including in vivo single-fiber recordings form the auditory nerve and in vitro patch clamping of cultured SGCs. However, the focus in CI-related research is mostly on eCAPs and electrically evoked auditory brainstem responses (eABRs), since these can be related to similar recordings in humans. For instance, several studies reported that after induction of SNHL in guinea pigs eABR thresholds increase (Shepherd and Javel, 1997; Maruyama et al., 2008; Fransson et al., 2010), although Agterberg et al. (2009) reported unchanged thresholds after deafening. eABR amplitudes are consistently found to be lower after deafening (Hall, 1990; Shepherd and Javel, 1997; Agterberg et al., 2009; Havenith et al., 2011). Treatment with neurotrophic factors has often been found to reduce the eABR threshold (e.g., Yamagata et al., 2004; Maruyama et al., 2008; Fransson et al., 2010; Leake et al., 2011), while again unchanged thresholds have been reported as well (Agterberg et al., 2009; Havenith et al., 2011). The eABR amplitude is higher in animals subjected to neurotrophic treatment than in deafened controls (Maruyama et al., 2008; Agterberg et al., 2009), which most likely reflects the enhanced survival of SGCs. One of the main advantages of electrophysiological studies in animals is that functional outcome measures can be directly compared to the condition of the neural substrate by means of histology. This possibility makes animal research crucial in finding characteristics of the electrical activity of SGCs or other auditory cells that reliably reflect their physical condition. For example, Hall (1990) showed that the amplitude of the first peak of the eABR complex, which like the eCAP corresponds to activity of the auditory nerve, correlates with the amount of surviving SGCs in deafened rats. Correlations such as these may be crucial in order to be able to understand and interpret differences in electrophysiological outcome measures among CI users and to relate these to neurobiological processes such as degeneration or adaptation on the one hand, and differences in speech perception on the other. 20

General introduction

While the eCAP or eABR amplitude may in theory accurately reflect the number of present (or, rather, contributing) SGCs, factors such as the distance and impedance between the stimulation electrode and the SGCs and between the source of activity (SGCs or brainstem) and the recording electrode (Grill et al., 2009) will complicate comparison between subjects, or within subjects over time. For this reason, electrophysiological measures that are not subject to these factors, and that correlate with the condition of the neuronal population, would provide a valuable tool for the assessment of neuronal status in CI users and of its relation to CI performance. An example of such measures is the effect size of a time delay between the two phases in a biphasic stimulation pulse. This time delay, called the inter-phase gap (IPG), generally results in a more efficacious stimulus, objectified by a decrease in excitation threshold and an overall shift of the input-output curve toward lower stimulus levels (van den Honert and Mortimer, 1979; Shepherd and Javel, 1999; Cappaert et al., 2013). An increase in IPG in CI stimulation furthermore results in an increase in loudness perception (McKay and Henshall, 2003) as well as a decrease in behavioral thresholds (Carlyon et al., 2005). In a study with deafened guinea pigs, Prado-Guitierrez et al. (2006) showed that the magnitude of this IPG effect on both the eCAP and eABR level correlates with the number of surviving SGCs. The importance of this finding is that this IPG effect is considered to be largely â&#x20AC;&#x201C; if not completely â&#x20AC;&#x201C; independent of the aforementioned factors, such as electrode impedance. So far, a relation between speech perception and this IPG effect has not been found (Kim et al., 2010), which only illustrates that a more thorough examination of the origin of this effect (both at the cellular and at the population level) and of its implications for SGC functionality and subsequent higherlevel processes is in order.

8. Aim of this thesis The main objective of this thesis is to contribute to the knowledge that is necessary for the preservation or recuperation of the functional integrity of the auditory nerve in CI recipients. This contribution largely consists of three elements. First, a detailed morphological description of SGCs in the healthy and the deafened cochlea is provided. Second, a detailed characterization of SGC responsiveness to electrical stimulation is given, both in the normal-hearing and in the deafened cochlea, and functional correlates for SGC degeneration are described. Third, it is assessed whether exogenously administered BDNF is capable of restoring or at least preventing further changes associated with SGC degeneration, both functionally and morphologically. 21

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 1

9. Thesis outline

It is often assumed that a healthy nerve is one that is made up by a large number of SGCs, but this account completely disregards the morphological and electrophysiological changes the individual SGC may experience in the neurotrophin-deprived cochlea. Therefore, chapters 2-4 of this thesis are dedicated to a close examination of neurotrophic regulation in the healthy cochlea, and of the effects of deafening and subsequent treatment with exogenous BDNF on morphology and functionality of SGCs. In chapter 2 the current literature on the role of neurotrophic factors in the developing and the mature cochlea is reviewed. Furthermore, a comprehensive overview is provided of the effects of treatment of the deafened cochlea with exogenous neurotrophic factors. Chapter 3 deals with SGC morphology in the healthy, the deafened, and the BDNFtreated cochlea. It provides a detailed description of the number and appearance of the SGCs and takes into account the basoapical gradients in the cochlea with respect to neurotrophin presence and dependence. Chapter 4 is analogous to chapter 3, as it describes the morphology of SGC in the normal, deafened, and BDNF-treated cochlea, but its focus is entirely on the presence and morphology of the peripheral processes of the SGCs. These processes are of specific interest since although their function is essentially lost in the absence of hair cells, their presence may influence the electrical excitation by CIs because of their closer proximity to the electrode array in the scala tympani. Chapters 5 and 6 describe extensive electrophysiological characterizations of SGCs in the normal-hearing and deafened cochlea of the guinea pig using intracochlear eCAP recordings. Chapter 5 is dedicated to the identification of objective electrophysiological measures â&#x20AC;&#x201C; using variation of phase duration and IPG â&#x20AC;&#x201C; that predict the degree of degeneration of the SGC population. Chapter 6 describes the temporal response properties of electrically stimulated SGCs in normal-hearing and deafened guinea pigs and relates these properties to quantified measures of SGC degeneration. The experiments described in chapter 7 were designed to evaluate the protective potential of BDNF both on SGC morphology and functionality in chronically implanted guinea pigs. Additionally, in order to assess the practicability of clinical application of treatment with neurotrophic factors in the future, the long-term effects of temporary treatment were evaluated. The data presented in chapter 8 mainly comprise histological findings combined from several studies (chapters 3, 5 and 7). The time course of SGC degeneration after deafening is compared with the time course of degeneration of their peripheral and central processes. The course of SGC degeneration is relevant for CI performance and may be important for the development of strategies to preserve or regrow neural tissue in the cochlea. 22

General introduction

The main findings from this thesis are summarized and briefly discussed in chapter 9. In this concluding chapter both the scientific and clinical implications as well as future directions are outlined.

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 1

References

Agterberg, M.J.H., Versnel, H., Van Dijk, L.M., De Groot, J.C.M.J., Klis, S.F.L., 2009. Enhanced survival of spiral ganglion cells after cessation of treatment with brain-derived neurotrophic factor in deafened guinea pigs. J. Assoc. Res. Otolaryngol. 10, 355-367. Cappaert, N.L.M., Ramekers, D., Martens, H.C.F., Wadman, W.J., 2013. Efficacy of a new chargebalanced biphasic electrical stimulus in the isolated sciatic nerve and the hippocampal slice. Int. J. Neural Syst. 23, 1250031. Carlyon, R.P., van Wieringen, A., Deeks, J.M., Long, C.J., Lyzenga, J., Wouters, J., 2005. Effect of interphase gap on the sensitivity of cochlear implant users to electrical stimulation. Hear. Res. 205, 210–224. Ernfors, P., Van De Water, T., Loring, J., Jaenisch, R., 1995. Complementary roles of BDNF and NT-3 in vestibular and auditory development. Neuron 14, 1153-1164. Eshraghi, A.A., Nazarian, R., Telischi, F.F., Rajguru, S.M., Truy, E., Gupta, C., 2012. The cochlear implant: historical aspects and future prospects. Anat. Rec. (Hoboken) 295, 1967-1980. Fayad, J.N., Linthicum, F.H. Jr., 2006. Multichannel cochlear implants: relation of histopathology to performance. Laryngoscope 116, 1310-1320. Fritzsch, B., Pirvola, U., Ylikoski, J., 1999. Making and breaking the innervation of the ear: neurotrophic support during ear development and its clinical implications. Cell Tissue Res. 295, 369-382. Fransson, A., Maruyama, J., Miller, J.M., Ulfendahl, M., 2010. Post-treatment effects of local GDNF administration to the inner ears of deafened guinea pigs. J. Neurotrauma 27, 1745-1751. Géléoc, G.S., Holt, J.R., 2014. Sound strategies for hearing restoration. Science 344, 1241062. Gillespie, L.N., Clark, G.M., Marzella, P.L., 2004. Delayed neurotrophin treatment supports auditory neuron survival in deaf guinea pigs. Neuroreport 15, 1121-1125. Grill, W.M., Norman, S.E., Bellamkonda, R.V., 2009. Implanted neural interfaces: biochallenges and engineered solutions. Annu. Rev. Biomed. Eng. 11, 1-24. Hall, R.D., 1990. Estimation of surviving spiral ganglion cells in the deaf rat using the electrically evoked auditory brainstem response. Hear. Res. 49, 155-168. Havenith, S., Versnel, H., Agterberg, M.J.H., de Groot, J.C.M.J., Sedee, R.J., Grolman, W., Klis, S.F.L., 2011. Spiral ganglion cell survival after round window membrane application of brainderived neurotrophic factor using gelfoam as carrier. Hear. Res. 272, 168-177. Kim, J.R., Abbas, P.J., Brown, C.J., Etler, C.P., O’Brien, S., Kim, L.S., 2010. The relationship between electrically evoked compound action potential and speech perception: a study in cochlear implant users with short electrode array. Otol. Neurotol. 31, 1041-1048. Kim, J.R., Brown, C.J., Abbas, P.J., Etler, C.P., O’Brien, S., 2009. The effect of changes in stimulus level on electrically evoked cortical auditory potentials. Ear Hear. 30, 320-329. Leake, P.A., Hradek, G.T., Hetherington, A.M., Stakhovskaya, O., 2011. Brain-derived neurotrophic factor promotes cochlear spiral ganglion cell survival and function in deafened, developing cats. J. Comp. Neurol. 519, 1526-1545. Maruyama, J., Miller, J.M., Ulfendahl, M., 2008. Glial cell line-derived neurotrophic factor and antioxidants preserve the electrical responsiveness of the spiral ganglion neurons after experimentally induced deafness. Neurobiol. Dis. 29, 14-21. McKay, C.M., Henshall, K.R., 2003. The perceptual effects of interphase gap duration in cochlear implant stimulation. Hear. Res. 181, 94–99. Moerel, M., De Martino, F., Formisano, E., 2014. An anatomical and functional topography of human auditory cortical areas. Front. Neurosci. 8, 225. 24

General introduction

Peterson, N.R., Pisoni, D.B., Miyamoto, R.T., 2010. Cochlear implants and spoken language processing abilities: review and assessment of the literature. Restor. Neurol. Neurosci. 28, 237-250. Prado-Guitierrez, P., Fewster, L.M., Heasman, J.M., McKay, C.M., Shepherd, R.K., 2006. Effect of interphase gap and pulse duration on electrically evoked potentials is correlated with auditory nerve survival. Hear. Res. 215, 47-55. Ramekers, D., Versnel, H., Grolman, W., Klis, S.F.L., 2012. Neurotrophins and their role in the cochlea. Hear. Res. 288, 19-33. Runge-Samuelson, C., Firszt, J.B., Gaggl, W., Wackym, P.A., 2009. Electrically evoked auditory brainstem responses in adults and children: effects of lateral to medial placement of the nucleus 24 contour electrode array. Otol. Neurotol. 30, 464-470. Seyyedi, M., Viana, L.M., Nadol, J.B. Jr., 2014. Within-Subject Comparison of Word Recognition and Spiral Ganglion Cell Count in Bilateral Cochlear Implant Recipients. Otol. Neurotol. 35, 14461450. Shah, S.B., Gladstone, H.B., Williams, H., Hradek, G.T., Schindler, R.A., 1995. An extended study: protective effects of nerve growth factor in neomycin-induced auditory neural degeneration. Am. J. Otol. 16, 310-314. Shepherd, R.K., Javel, E., 1997. Electrical stimulation of the auditory nerve. I. Correlation of physiological responses with cochlear status. Hear. Res. 108, 112-144. Shepherd, R.K., Javel, E., 1999. Electrical stimulation of the auditory nerve: II. Effect of stimulus waveshape on single fibre response properties. Hear. Res. 130, 171-188. Smoorenburg, G.F., Willeboer, C., van Dijk, J.E., 2002. Speech perception in nucleus CI24M cochlear implant users with processor settings based on electrically evoked compound action potential thresholds. Audiol. Neurootol. 7, 335-347. Spoendlin, H., 1975. Retrograde degeneration of the cochlear nerve. Acta Otolaryngol. 79, 266– 275. Van den Honert, C., Mortimer, J.T., 1979. The response of the myelinated nerve fiber to short duration biphasic stimulating currents. Ann. Biomed. Eng. 7, 117-125. Versnel, H., Agterberg, M.J.H., De Groot, J.C.M.J., Smoorenburg, G.F., Klis, S.F.L., 2007. Time course of cochlear electrophysiology and morphology after combined administration of kanamycin and furosemide. Hear. Res. 231, 1-12. Webster, M., Webster, D.B., 1981. Spiral ganglion neuron loss following organ of Corti loss: a quantitative study. Brain Res. 212, 17-30. WHO, 2012. WHO global estimates on prevalence of hearing loss. Programme of the Prevention of Blindness and Deafness. Geneva, Switzerland: World Health Organization (WHO). (http:// www.who.int/pbd/deafness/WHO_GE_HL.pdf?ua=1). Willeboer, C., Smoorenburg, G.F., 2006. Comparing cochlear implant users’ speech performance with processor fittings based on conventionally determined T and C levels or on compound action potential thresholds and live-voice speech in a prospective balanced crossover study. Ear Hear. 27, 789-798. Yamagata, T., Miller, J.M., Ulfendahl, M., Olivius, N.P., Altschuler, R.A., Pyykkö, I., Bredberg, G., 2004. Delayed neurotrophic treatment preserves nerve survival and electrophysiological responsiveness in neomycin-deafened guinea pigs. J. Neurosci. Res. 78, 75-86. Ylikoski, J., Pirvola, U., Virkkala, J., Suvanto, P., Liang, X.Q., Magal, E., Altschuler, R., Miller, J.M., Saarma, M., 1998. Guinea pig auditory neurons are protected by glial cell line-derived growth factor from degeneration after noise trauma. Hear. Res. 124, 17-26. Ylikoski, J., Wersall, J., Bjorkroth, B., 1974. Degeneration of neural elements in the cochlea of the guinea pig after damage to the organ of corti by ototoxic antibiotics. Acta Otolaryngol. Suppl. 326, 23-41. 25

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 1

Zilberstein, Y., Liberman, M.C., Corfas, G., 2012. Inner hair cells are not required for survival of spiral ganglion neurons in the adult cochlea. J. Neurosci. 32, 405-410.

CHAPTER 2 Neurotrophins and their role in the cochlea

Dyan Ramekers, Huib Versnel, Wilko Grolman, Sjaak F.L. Klis Hearing Research 288 (2012) 19-33.

Cold silence has/ A tendency to/ Atrophy any/ Sense of compassion – Maynard James Keenan, “Schism”

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 2

Abstract

Spiral ganglion cell (SGC) degeneration following hair cell loss can be prevented by administration of exogenous neurotrophic factors. Many of these neurotrophic factors, in particular the neurotrophins brain-derived neurotrophic factor (BDNF) and neurotrophin-3 (NT-3), have been described to be involved in the development of the rodent cochlea. While expression of most of the neurotrophins has decreased to below detectable levels during adulthood (only NT-3 remains highly expressed), their respective receptors remain present in SGCs. Indeed much less is known about the function of neurotrophins in the mature cochlea. Such knowledge is crucial in the search for tools to improve SGC survival following cochlear implantation. In this review, we will critically regard the current experimental findings of neurotrophic treatment of the SGCs in the perspective of fundamental cellular mechanisms underlying neurotrophin signaling. We conclude that, in order to fully apprehend the effects of neurotrophic treatment of degenerating SGCs and in order to consider clinical application of neurotrophins, future research should focus (a) on characterizing the expression pattern of neurotrophins in the cochlea after deafening, (b) on more detailed characterization of functional and morphological changes of SGCs associated with both deafening and neurotrophic treatment and (c) on the possible self-supporting state of SGCs after cessation of shortterm neurotrophic treatment.

Neurotrophins and their role in the cochlea

1. Neurotrophins 1.1. Neurotrophins and their receptors Neurotrophic factors (NTFs) are proteins that are produced and secreted by neurons, glial cells, sensory cells and muscle fibers. The structurally homologous neurotrophins form a subclass of NTFs (e.g., Reichardt, 2006). Neurotrophins are the most ubiquitously expressed of all NTFs and are therefore the most extensively studied. In mammals four types of neurotrophins have been identified: nerve growth factor (NGF), brain-derived neurotrophic factor (BDNF), neurotrophin-3 (NT-3) and neurotrophin-4/5 (NT-4/5; also called neurotrophin-4 or NT-4). The remainder of the NTF family consists of at least a dozen more members, including glial cell line-derived neurotrophic factor (GDNF), fibroblast growth factor (FGF) and ciliary neurotrophic factor (CNTF). Although these factors are not as universally distributed within the nervous system as are the four neurotrophins, their actions are often of similar importance (e.g., Ylikoski et al., 1998; Chao, 2000; Reichardt, 2006). All four neurotrophins bind to p75 neurotrophin receptor (p75NTR), but have a second ligand-specific receptor of the tropomyosin-related kinase (Trk) family: NGF binds to TrkA, BDNF and NT-4 bind to TrkB and NT-3 binds to TrkC (Barbacid, 1994; see Fig. 1). All three Trk receptors have been shown to be homologous at the amino acid level for 87% (Dawbarn and Allen, 2003). In this review we focus on the influence of neurotrophins in the cochlea. However, since the role of NGF and NT-4/5 is not as large as that of BDNF and NT-3, emphasis will be on the latter two. In addition, since especially the non-neurotrophin GDNF is often experimentally applied in the inner ear alongside BDNF and NT-3, we will occasionally digress from the strict boundaries of NTF categorization. A

sensory cell (target cell)

innervation

neurotrophic signaling

NGF BDNF NT-4/5

TrkA

NT-3

TrkC

TrkB

p75NTR ganglion cell (receptor cell)

to CNS

GDNF target cell

RET GFRÎą1 receptor cell

Fig. 1. Neurotrophin signaling. (A) Target cells such as sensory cells or muscle cells are innervated by the more centrally located receptor cells such as ganglion cells. Receptor cell survival depends on the presence of a target cell, which is conveyed by the release of neurotrophins. CNS, central nervous system. (B) Neurotrophins and their receptors. Neurotrophins are released by the target cell by means of exocytosis. Solid arrows indicate high affinity between ligand and receptor; dashed arrows indicate low affinity. Note that the non-neurotrophin GDNF binds to a heterodimer receptor complex formed by RET and GFRÎą1. 29

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 2

1.2. Neurotrophin function Neurotrophins are secreted by the targets of neuronal innervation (target cell) and have their effect on more centrally located neurons (see Fig. 1). Activation of any of the four neurotrophin receptors will trigger extensive intracellular signaling cascades, leading to complex physiological changes in the receptor cell (Thoenen, 1995; Huang and Reichardt, 2003; Reichardt, 2006). In short, these signaling pathways comprise cell proliferation, cell differentiation, cell maturation, neuronal plasticity, axonal outgrowth or repulsion, and even apoptosis (Huang and Reichardt, 2001; Greenberg et al., 2009; Teng et al., 2010). To explain this wide variety of cellular fates with as few as only four neurotrophins, the key can be found in combined activation and protein complex formation. For example, it has been found that p75NTR activation by any of the four neurotrophins leads to apoptosis, but only when there is no accompanying activation of a member of the Trk family (Frade et al., 1996). However, if there is coactivation of p75NTR and Trk, the outcome can be proliferation, actual physical growth, or differentiation – depending on the specific neurotrophin, on the cell type or its environment and depending on co-activation of even more (but structurally unrelated) transmembrane receptors (Bothwell, 1995; Teng et al., 2010). 1.3. Target cell regulation of neurotrophin signaling The widespread consequences and therefore the great importance of neurotrophin signaling are reflected in the large amount of controllable steps in this pathway from gene transcription to protein synthesis, enzyme activation and ligand-receptor interaction. Already at the posttranscriptional stage, different neurotrophin mRNA sequences can be formed by the process of RNA splicing. By alternative splicing, one or more of the coding exons are removed along with the intermediate non-coding introns, giving rise to a variety of different transcripts, all coding for the same protein. The NT-3 gene has been reported to consist of three exons including two 5’ short untranslated exons and a 3’ long exon encoding the entire protein, and to give rise to two classes of transcripts by alternative splicing of the 5’ exons to the 3’ coding exon (Sekimoto et al., 1998). Little information is available on the differential usage of alternative promoters of different exons of NT-3 although two different promoters have been described (Leingärtner and Lindholm, 1994). This is different for BDNF, which comprises of 34 possible transcripts, and although the most abundant BDNF transcripts in humans all contain exons VI and IXabcd, most of the 34 possible transcripts can be found in human tissue (Pruunsild et al., 2007; Baj and Tongiorgi, 2009). Each 5′ exon of BDNF is alternatively spliced to a downstream exon that contains the coding region of BDNF with a 3′ untranslated region containing two potential polyadenylation signals (Aid et al., 2007). However, the functional consequences of multiple transcripts that encode the same protein are not understood. 30

Neurotrophins and their role in the cochlea

Posttranslational modification occurs by proteolytic cleavage of proneurotrophins in the cell soma, but is has also been reported to occur extracellularly (Lee et al., 2001). Pro-forms of neurotrophic factors are traditionally thought to be inactive proteins, silent until activation by cleaving enzymes. However, the pro-forms are now thought to be biologically active (see below), and the nature of their function depends on the activity of proteolytic enzymes. The rate by which pro-forms of neurotrophic factors are proteolytically cleaved into mature forms sets the balance between pro-function and mature function (Edwards et al., 1988; Lee et al., 2001; Reichardt, 2006). Consequently, it is not just the amount but also the structure of secreted neurotrophins that determines the strength and the nature of the neurotrophic message.

1.4. Receptor cell regulation of neurotrophin signaling The amplification of the neurotrophin signal is proportional to the density of neurotrophin receptors (NTRs). The physiological effect of this signal on the receptor cell is dependent on the relative expression of the various NTRs. In addition, the level of neurotrophin maturation is influential at this point as well. Various experimental findings suggest that the primary ligands for p75NTR are not the mature neurotrophins, but their precursors – the proneurotrophins (Dechant and Barde, 1997; Lee et al., 2001; Reichardt, 2006; Ivanicevic and Saragovi, 2006; Arévalo and Wu, 2006). This means that the relative NTR expression influences both the physiological effect of binding of mature neurotrophins and the strength of the effect of proneurotrophins. And as mentioned in the previous section, for the receptor cell the difference between p75NTR and Trk activation can literally denote the distinction between life and death. In addition, high concentrations of neurotrophins – when Trk receptors are near saturation – will have a stronger effect on the low affinity p75NTR than low concentrations. As a consequence, extreme concentrations of neurotrophins can be lethal to the receptor cell as well (Lee et al., 2001). Kinase activity of all Trk receptors is regulated by ten intracellular tyrosines that can be phosphorylated, resulting in enhanced tyrosine kinase activity upon binding of a neurotrophin (Patapoutian and Reichardt, 2001; Reichardt, 2006). Without any tyrosine phosphorylation, neurotrophin binding has virtually no consequence, whereas a fully phosphorylated Trk receptor is highly active following ligand binding. It could thus be said that a neuron can decide for itself whether it wants to be susceptible to external neurotrophic signals. However, phosphorylation of the individual tyrosines of Trk receptors is strictly regulated following local cell-type-specific standards. The intracellular pathways activated by NTRs are numerous and well-documented (e.g., Huang and Reichardt, 2003; Bramham and Messaoudi, 2005; Reichardt, 2006; Numakawa et al., 2010). Among others, the activation of Trk receptors leads to 31

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 2

activation of the Ras – Raf – MEK – ERK and Phospholipase-C – Ca2+ pathways. In general terms, these pathways lead to activation of the cellular metabolic machinery, the former mainly by initiating gene transcription and the latter by enhancing protein synthesis. What is important to note is that this so-called machinery contains a vast amount of entities, which all need to be supervised strongly. Faults in this metabolic system frequently lead to unrestrained cell growth. More and more, neurotrophins and their receptors are seen as causes for all kinds of neural and nonneural forms of cancer (e.g., Thiele et al., 2009). Interestingly in this context is that patients whose tumors that have elevated TrkA and TrkC expression levels have a better prognosis than those with elevated levels of TrkB and BDNF (Thiele et al., 2009; Tauszig-Delamasure and BouzasRodriguez, 2011). Unfortunately, however, it is extremely difficult to intervene in these extensive signaling cascades, because of the vast amounts of agents which cause amplification and attenuation as well as divergence and convergence of numerous pathways simultaneously.

1.5. The role of neurotrophins in neurodegenerative disorders Apart from cancer, many other diseases of the nervous system have been associated with abnormal expression or functioning of neurotrophins and their receptors (e.g., Huang and Reichardt, 2003; Allan and Dawbarn, 2006; Hennigan et al., 2007). Below the possible use of neurotrophins as a pharmacological treatment is discussed for a selection of neurodegenerative diseases: Alzheimer’s disease, Parkinson’s disease and amyotrophic lateral sclerosis. Brains of patients with Alzheimer’s disease (AD) show region-specific alterations in distribution of neurotrophins; especially NGF levels are reduced. Numerous studies have established a relation between lower levels of NGF and cholinergic cell degeneration, causing cognitive deficits (reviewed by Dawbarn and Allen, 2003). Transgenic studies have revealed that mice with antibodies to NGF develop an AD-like phenotype during late adulthood, which is abolished by administration of exogenous NGF (Capsoni et al., 2000). In addition, a clinical trial study has shown that AD symptoms were overall slightly diminished during continuous intracerebroventricular administration of mice NGF (Eriksdotter Jonhagen et al., 1998). However, side effects, including pain, were prevailing. Parkinson’s disease (PD) is an example of a disease not caused by – or in any way related to – a change in neurotrophin concentrations, but for which neurotrophins can be used to develop a fundamental treatment nonetheless. A first possible neurotrophin treatment for PD is fighting the dopamine-releasing cell apoptosis in the substantia nigra (SN). Because apoptosis in these cells is mediated by pro-NGF and p75NTR, and although it has not been shown that this mechanism is hyperactive in PD patients, 32

Neurotrophins and their role in the cochlea

inhibiting proNGF-p75NTR coupling might restore the balance and preserve SN neurons (Chen et al., 2008). Second, because the administration of dopamine substitutes like L-dopa have only a transient effect and thus do not cure PD, it is now tried to implant dopamine-producing cells into the midbrain in rat models of PD. Survival rate of these cells is dramatically low, but trials using BDNF to stimulate dopamine release of these cells in culture (Hoglinger, et al., 1998), and NT-4/5 and GDNF for cell survival (Haque et al., 1996; Lin et al., 1993) have proven successful. Amyotrophic lateral sclerosis (ALS) is caused by a mutation in the SOD1 gene, which causes defects in the mitochondrial respiratory system. All four neurotrophins have been found to be upregulated in muscle tissue of ALS patients (Küst et al., 2002), but an in vitro study showed that the administration of exogenous NT-4/5 seems to yield a therapeutic effect, while there is no neuroprotective effect of BDNF or NT-3 (Van Westerlaak et al., 2001). It can be hypothesized that, at least in the case of ALS, upregulation of neurotrophins may not be a goal-directed response, but rather is a last resort – a crude means for trying to reduce motoneuronal apoptosis. It should then be considered to investigate whether almost aimlessly elevated endogenous neurotrophin levels could contribute to the symptoms of the disease. Alternatively, the mere occurrence of an increase of all neurotrophin levels – whether a crude hailshot approach to tissue-preservation or not – indicates that it is possible and ostensibly safe to clinically intervene in the neurotrophin homeostasis, without immediately causing either mass apoptosis or unrestrained cell growth. In conclusion, although it cannot be excluded that the prolonged elevation of neurotrophin levels is ultimately harmful, clinical neurotrophic treatment can be considered for neurodegenerative disorders.

1.6. Experimental applications of neurotrophins Since the late 1980s the application of neurotrophins has been used in animal experiments (e.g., Schinstine et al., 1991). The first experiments concerned the infusion of NGF in the ventricular system of rats (Hefti, 1986; Gage et al., 1988) or macaques (Tuszynski et al., 1990; Koliatsos et al., 1990) in order to prevent atrophy of cholinergic neurons in the forebrain after a lesion of the septo-hippocampal projections. In the two macaque studies, NGF levels in the infused ventricles was 20-1500 times higher than in controls, yet there were no indications that these increased levels had any toxic effects. The administration of neurotrophins via the ventricular system obviously has one major drawback. Because the infusion is directly into the cerebrospinal fluid, the neurotrophin will quickly be distributed throughout the entire brain, despite the fact that the opening of the infusion cannula was in close proximity to the target area. Furthermore, a 20-to-1500-fold increase of NGF within the ventricular boundaries is an irrelevant measure for the concentration of NGF in the septal nucleus. Nonetheless, 33

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 2

Tuszynski et al. (1990) found a survival rate of 80% of the cholinergic neurons treated with NGF, compared to the unlesioned and untreated contralateral septal nucleus; in lesioned untreated macaques survival was only 40%. Comparable results were found by Koliatsos et al. (90% vs. 60%) and in the rat study performed by Hefti (88% vs. 50%). An elegantly designed experiment by Altar et al. (1992) showed that infusion of neurotrophins can be performed much more localized: both specifically into the striatum and right above the SN of the rat. They first showed that around 80% of the radioactively labeled BDNF infused into the striatum is transported to the SN. Next, they showed that BDNF infused directly above the SN elicits behavior comparable to that when the activity of the SN is enhanced by an injection of amphetamine, signifying that BDNF succeeded in enhancing the dopamine production in the SN. In addition, their study included a more elaborate monitoring of the bodily functions and the behavior of the animals in response to elevated BDNF levels. They did not find any indication that the exogenous BDNF caused any side effects, apart from a slight and transient loss of body weight. Lucidi-Phillipi et al. (1995) repeated both the septohippocampal lesion study (discussed in the previous paragraph) and the SN study in rats, but instead of infusing the BDNF with a cannula they used BDNF-producing grafts consisting of transduced rat fibroblast cells (see Schinstine et al., 1991, for a review on this technique). Not surprisingly, their SN findings were consistent with those of Altar et al. (1992). The amount of preserved dopaminergic neurons was much lower, which was attributed to lower rate of BDNF secretion by grafts than that by infusion. However, with respect to the septo-hippocampal lesion study, BDNF producing grafts did not protect septal cholinergic neurons. The inability of BDNF to prevent septal nucleus neurons from degeneration, as opposed to the protective function of NGF as shown by Hefti (1986) and Tuszynski et al. (1990), indicate that the four neurotrophins are not simply interchangeable and that each has its own target.

2. Neurotrophin function and distribution in the cochlea 2.1. Neurotrophic support and innervation of the cochlea The main neural structures in the adult cochlea are (1) the sensory hair cells positioned on the basilar membrane, (2) the spiral ganglion cells (SGCs) in the modiolus and (3) the auditory nerve (see Fig. 2). In short, inner hair cells depolarize in response to vibrations of the basilar membrane and consequently excite the SGCs, whose axons compose the auditory or cochlear nerve. Together with the vestibular nerve, the auditory nerve forms the vestibulocochlear nerve (n. VIII), projecting to the central nervous system. 34

Neurotrophins and their role in the cochlea

As with all peripheral sensory cells, it is primarily the hair cells that determine the fate of the SGCs during development as well as during adulthood. There is evidence implying only a small role for central targets in trophic support for primary neurons in general (Ard et al., 1985; Johnson and Yip, 1985). In addition, survival by autocrine neurotrophic support, independent from peripheral or central targets has only been suggested for dorsal root ganglion neurons (Acheson et al., 1995).

OHCs

IHC

peripheral processes

SGCs AN

Fig. 2. The organ of Corti and spiral ganglion shown in a midmodiolar section of a guinea pig cochlea. Inner hair cells synapse onto the peripheral processes of SGCs, which in turn, via the auditory nerve, synapse onto central neurons in the cochlear nucleus. Histological preparation was done as described previously by Van Ruijven et al. (2004). AN, auditory nerve; IHC, OHCs, inner and outer hair cells; SGCs, spiral ganglion cells.

2.2. Distribution of neurotrophins and neurotrophin receptors in the cochlea The predominant neurotrophins in the inner ear are BDNF and NT-3 (Wheeler et al., 1994; Ylikoski et al., 1993; Fritzsch et al., 1997a, 1999). A clear distinction should be made between developmental expression patterns and neurotrophin expression during adulthood (see Table 1, Fig. 3). In early development of the rat, BDNF mRNA has been found in both inner and outer hair cells, but not in the adult organ of Corti (insignificant levels at post-natal day P4; Wheeler et al., 1994). NT-3 is continuously expressed by inner hair cells during development and throughout adulthood, while outer hair cell expression of NT-3 is present in development but ceases at maturation. The high affinity receptors TrkB and TrkC, for BDNF and NT-3 respectively, are both strongly expressed in spiral ganglia around birth and during adulthood (Ylikoski et al., 1993).

Figure 2 35

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

NT-3 mRNA expressed near CVG in otic vesicle (E101 and E11.54) and strongly in the developing cochlea (E15)1; NT-3 transcripts found in E17 IHCs and OHCs, but decreased towards perinatal period3 Age 1 week: apical high mRNA level in OHCs, moderate levels in IHCs; both weaker in basal turn2

Low levels of BDNF mRNA around E11 â&#x20AC;&#x201C; stronger expression at E13 in cochlea and vestibular system1; From E11.5 to P1, mRNA found in vestibular sensory epithelium in saccule, utricle and ampullae4; BDNF mRNA in IHCs and OHCs decreases from E17 to insignificant levels at P43; Age 1 week: mRNA expression in both IHCs and OHCs2

Weak mRNA expression in non-sensory epithelium1; NGF transcripts found in membranes of ampullae and otolithic organs2; No expression in the OC3 or entire cochlea5; No expression detected from E11.5 to P14

during adulthood

No expression2

TrkB

TrkA

receptor

p75NTR

High NT-3 mRNA levels in IHCs, TrkC none in OHCs; in IHCs higher levels in apical than in basal regions2; After birth, NT-3 levels in IHCs (but not in OHCs) increase rapidly to reach embryonic levels3

No BDNF mRNA found in the OC 2,3

NGF transcripts found in membranes of ampullae and otolithic organs2; No expression in the OC3

presence in the inner ear

during development

NT-4/5 No expression1,2,4; Mice lacking NT-4/5 show normal OC fenotype6

NT-3

BDNF

NGF

ligand

From E11.5 to P1, moderate mRNA expression in the CVG4; At E21 p75NTR expression is found in pillar cells and SGCs, but not in hair cells3; Age 1 week: moderate p75NTR mRNA expression in spiral and vestibular ganglia2

Exogenous NT-3 induces neurite outgrowth from in vitro SAGs at E11-E12 (however, less effective than BDNF)1; Age 1 week: TrkC mRNA expression (weaker than TrkB2) in spiral and vestibular ganglia2,5

Moderate p75NTR mRNA expression in spiral and vestibular ganglia2

TrkC mRNA expression (weaker than TrkB) in spiral and vestibular ganglia2

Exogenous BDNF induces neurite Strong expression of TrkB outgrowth from in vitro SAGs at E11-E121; mRNA in spiral and vestibular From E11.5 to P1 slowly decreasing mRNA ganglia2 expression4; Age 1 week: strong expression of TrkB mRNA in spiral and vestibular ganglia2,5

Exogenous NGF does not induce neurite No TrkA mRNA expression2; outgrowth from in vitro SAGs at E11-E121; TrkA expression in adult rats From E11.5 moderate mRNA expression, and mice17 but absent from E14.54; Age 1 week: no TrkA mRNA expression2; No expression in rat embryonic cochlea5

presence in the inner ear

during development

Table 1. Expression patterns of neurotrophic factors and their receptors during development and in the healthy adult inner ear. during adulthood

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39 Chapter 2

Strong/moderate GFRα1 expression in SGCs, not in OC 7

No RET expression in guinea pig cochlea7;

PDGFR

FGFR2 FGFR3 FGFR4

FGFR1

P3: Both α and β expressed in OC and SGCs14; E12-P1: high but decreasing PDGFRα, low but stable PDGFRβ15

E13.5: OHCs, pillar and Deiters’ cells, weaker towards birth, nearly absent postnatal13; Only non-sensory, continuous presence13; Pillar and Deiters’ cells (from E15.5)13; In mesenchyme surrounding the scalae13

CNTFRα At P3: expression in OC, stria vascularis and in SGCs14

From P12 and throughout adulthood moderate expression RET in IHCs7,8 and SGCs8; Cochlear protein levels doubled GFRα1 8h after noise9

No expression in guinea pig cochlea7, low expression in rat modiolus and hair cells8; high expression in SGCs16; Weak GFRα1 expression in SGCs, not in OC 7

BDNF, brain-derived neurotrophic factor; CVG, cochleovestibular ganglion; E#, embryonic day #; FGF, fibroblast growth factor; FGFR1-4, FGF receptor 1-4; GDNF, glial cell line-derived neurotrophic factor; IHC, inner hair cell; NGF, nerve growth factor; NT-3, neurotrophin-3; NT-4/5, neurotrophin-4/5; OC, organ of Corti; OHC, outer hair cell; P#, post-natal day #; p75NTR,p75 neurotrophin receptor; PDGF, platelet-derived growth factor; SAG, statoacoustic ganglion; TrkA-C, tropomyosinerelated kinase A-C. 1 Pirvola et al., 1992 (rat); 2Ylikoski et al., 1993 (rat); 3Wheeler et al., 1994 (rat); 4Schecterson and Bothwell, 1994 (mouse); 5Ernfors et al., 1992 (rat); 6Bianchi et al., 1996 (mouse); 7Ylikoski et al., 1998 (guinea pig); 8Stöver et al., 2000 (rat); 9Nam et al., 2000 (rat); 10Lefebvre et al., 1991 (rat, in vitro); 11Hossain and Morest, 2000 (mouse, in vitro); 12Wright et al., 2003 (mouse); 13Hayashi et al., 2010 (mouse); 14Malgrange et al., 1998 (rat); 15Lee et al., 2004 (rat); 16Kuang et al., 1999 (rat) ; 17 Dai et al., 2004 (rat/mouse).

At P3: PDGFA,B expressed in the OC14; E12-P1: high PDGFA, but no PDGFB expression15

PDGF

FGF1 stimulates in vitro SGC survival, migration and neuritogenesis10,11; At E8.5 FGF4 expression in otic placode12; At E9 FGF16 expression in otic cup/otic vesicle12

FGF

At P3: CNTF expression in OC and stria vascularis, but not in SGCs14

CNTF

At P7 (but not at E16,18,21 or P1,3,5) GDNF mRNA is expressed in the IHCs and OHCs of the basal turn, at P9 GDNF expression extends into apex7; At P3: not present in OC or SGCs14

GDNF

Neurotrophins and their role in the cochlea

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 2

TrkA is only transiently expressed during embryonic days 11.5-14.5, but is abundantly expressed after birth in inner and outer hair cells and SGCs in both rats and mice (Dai et al., 2004). However, its ligand, NGF, is only weakly expressed and its expression is not found in the organ of Corti (Pirvola et al., 1992; Ylikoski et al., 1993; Wheeler et al., 1994). The presence of TrkA and NGF has been associated with the autonomic efferent innervation of the inner ear, rather than with its sensory function (Fagan et al., 1996; Fritzsch et al., 1997a). This is in accordance with the idea that sensory neurons derived from neural crest – but not from placodal origin, such as the inner ear – are supported by NGF (Pirvola et al., 1992). GDNF expression starts around the end of the first postnatal week in rats, while it is absent throughout the entire embryonic stage (Qun et al., 1999). RET, the transducing receptor of GDNF, is not expressed during development in either guinea pigs or rats. The expression of RET during adulthood is not clear, as in three studies it is respectively marked as nonexistent, low and high (Ylikoski et al., 1998; Stöver et al., 2000; Kuang et al., 1999).

development

IHC

SGC

NGF BDNF NT-3 NT4/5 GDNF CNTF FGF PDGF

adulthood

IHC

SGC

TrkA TrkB TrkC p75NTR

NGF BDNF NT-3 NT4/5

TrkA TrkB TrkC p75NTR

RET GFRα1 CNTFRα FGFR1-4 PDGFRα,β

GDNF

RET GFRα1

to CNS

Fig. 3. Summary of the data presented in Table 1. Expression of neurotrophins, other neurotrophic factors and their respective receptors in the developing (A) and in the adult rodent cochlea (B). Bold font indicates high expression, regular font indicates moderate expression and grey font indicates that the ligand or receptor is absent. The term development encompasses the period from embryonic to early postnatal development (see Table 1). Expression levels of CNTF, FGF, PDGF and their receptors in the adult cochlea have not been studied. IHC, inner hair cell; SGC, spiral ganglion cell; CNS, central nervous system.

Neurotrophins and their role in the cochlea

2.3. Interplay between BDNF and NT-3 in inner ear innervation Null-mutation studies have shown that two weeks old mice lacking BDNF show only a minor loss of mostly type-II SGCs, which innervate the outer hair cells (Bianchi et al., 1996). In addition, this loss of cells was found in the middle and particularly in the apical turns of the cochlea. These findings suggest a base-to-apex gradient in BDNF dependence along the spiral axis of the developing cochlea. Interestingly, NT-3 dependence has been shown to be the exact opposite: NT-3 null-mutant mice show a lack of both inner and outer hair cell innervation at the basal turn around embryonic day E15 (Fritzsch et al., 1997a; Fariñas et al., 2001). This pattern is in agreement with findings of Fritzsch et al. (1995), who found that TrkB knock-out mice have reduced innervation of the middle and apical turns, whereas TrkC knock-outs show reduced innervation of the basal part of the cochlea. Not surprisingly, double mutant mice – lacking both TrkB and TrkC – display complete absence of inner ear innervation (both cochlear and vestibular). Although BDNF and NT-3 provide trophic support in distinct areas in the inner ear, they might be interchangeable to some extent – or even functionally completely redundant, since the effect of null-mutations could be purely related to a general decrease in neurotrophic signaling rather than to a specific loss of neurotrophic function (Fariñas et al., 2001; Tessarollo et al., 2004). This idea is supported by the fact that TrkB and TrkC are found alongside each other in spiral ganglion cells of rats (Ernfors et al., 1992; Ylikoski et al., 1993; Fariñas et al., 2001) and by the fact that NT-3 has been known to also bind to TrkB, albeit with significantly lower affinity (Bothwell, 1995; although disputed by Stenqvist et al., 2005). To assess their interchangeability, Coppola et al. (2001) replaced the gene coding for NT-3 with the BDNF gene in mice, thus yielding NT-3BDNF/BDNF mice. The results were striking, as the basal turn of the cochlea – in developing wild-type mice supported by NT-3 – was fully innervated. Conversely, replacement of the coding region of the BDNF gene with that of NT-3 (BDNFNT-3/NT-3) has been shown to result in a much more innervated cochlea than that in BDNF knock-out mice (Agerman et al., 2003). Consequently, the virtually complete loss of hearing in BDNF knock-out mice was rescued in these mice. These results might indeed suggest interchangeability, however, data from Tessarollo et al. (2004) yielded a different interpretation: indisputably, the basal turn in NT-3BDNF/ BDNF mice is completely innervated in the absence of NT-3, but this is due to redirection of vestibular fibers into the cochlea. Indeed, Agerman et al. (2003) indicated a role of BDNF in fiber guidance by showing that the vestibular system in their BDNFNT-3/NT-3 mice is only sparsely innervated, while survival of vestibular neurons is substantial. Taken together, these findings have lead to the supposition that the expression of BDNF and NT-3 is carefully orchestrated both spatially and temporally during inner ear embryonic development, and that BDNF (and possibly NT-3), apart from their role in survival, function as neurite guidance agents.

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 2

Notwithstanding these results from developmental research, there has unfortunately been a lack of reports on the co-operation between BDNF and NT-3 in the adult cochlea. The gradient in NT-3 expression along the cochlear spiral axis is shown to reverse from embryonic development towards adulthood (from highest in basal turns to highest in apex; Pirvola et al., 1992; Ylikoski et al., 1993; Fritzsch et al., 1997b), but it is telling that the accompanying directional change in the BDNF expression gradient over time is still only a hypothesis (Davis, 2003). If the aforementioned interplay between NT-3 and BDNF is purely relevant during fiber guidance, it is to be expected that in the adult inner ear these neurotrophins become interchangeable. If, in the normal functioning adult cochlea, neurotrophic support is solely realized in terms of survival, BDNF and NT-3 would be interchangeable due to the uniform distribution pattern of their receptors TrkB and TrkC in the cochlea and their similar intracellular signaling pathways in general (Wheeler et al., 1994; Fritzsch et al., 2004; Reichardt, 2006). While this has not yet been investigated in healthy cochleas, the effect of exogenous neurotrophins in the damaged inner ear has been studied extensively.

3. Neurotrophic treatment of the damaged cochlea 3.1. Sensorineural hearing loss and cochlear implantation Profound hearing loss can often be characterized as sensorineural hearing loss (SNHL), a condition in which a substantial or virtually complete loss of hair cells leads to the inability to transfer basilar membrane vibrations into neuronal signaling. Cochlear implants (CIs) help partially retrieve hearing by direct electrical stimulation of the SGCs. However, as described in section 2.2 (Table 1, Fig. 3), the hair cells in a healthy cochlea provide NTFs to the SGCs (NT-3 and GDNF in the mature cochlea), and therefore, a secondary problem that emerges from the loss of hair cells is that the SGCs lose their neurotrophic support and, as a consequence, gradually degenerate. This secondary degeneration has been demonstrated in various animal models (Ylikoski et al., 1974; Webster and Webster, 1981; Xu et al., 1993; Dodson, 1997; Leake and Rebscher, 2004; McFadden et al., 2004; Versnel et al., 2007) and in humans (Linthicum, et al., 1991; Khan et al., 2005; Fayad and Linthicum, 2006; Pfingst et al., 2011). The speed of degeneration depends on species and cause of hearing loss. For instance in guinea pigs with a (near-) complete deprivation of hair cells, a loss of 50% of SGCs may occur within a month, and in humans such loss may be found after several years of deafness. Under certain conditions (observed after noise trauma), however, loss of SGCs or their peripheral processes may occur without hair cell loss (Kujawa and Liberman, 2006, 2009). The presumed underlying mechanism is postsynaptic excitotoxicity, caused by highly augmented presynaptic hair cell activity. 40

Neurotrophins and their role in the cochlea

Although a clear correlation between CI user performance and SGC count has not been found (Linthicum, et al., 1991; Khan et al., 2005; Fayad and Linthicum, 2006), recent advances in CI technology more and more depend on a well functioning auditory nerve (Roehm and Hansen, 2005; Green et al., 2008). Therefore, in order to maintain the hearing ability for CI users over time and, in addition, to keep pace with technological progress now and in the future it is thought to be crucial to prevent SGC degeneration.

3.2. SGC preservation with exogenous neurotrophic factors Ideally, treatment of SNHL should focus on the integrity of the hair cell population. Indeed, hair cells can be successfully protected from both noise and aminoglycosideinduced ototoxicity (e.g., Duan et al., 2000), but in clinical practice hair cells are already lost in patients suffering from SNHL, and therapeutical focus is on the protection of SGCs from secondary degeneration. Hence, the application of exogenous neurotrophic factors has been studied extensively in animal models of SNHL. An overview of these studies, performed during the last two decades, is presented in Table 2. It has been noted that the two primary neurotrophins in the adult cochlea are BDNF and NT-3, and that the two may to some extent be functionally interchangeable (Wheeler et al., 1994; Fritzsch et al., 2004; Reichardt, 2006; see section 2.3). Hence, these two are the most common neurotrophins used in deafened animals in the literature (Ernfors et al., 1996; Staecker et al., 1996; Miller et al., 1997, 2007; Shinohara et al., 2002; Gillespie et al., 2003, 2004; Yamagata et al., 2004; Endo et al., 2005; McGuinness and Shepherd, 2005; Noushi et al., 2005; Richardson et al., 2005; Shepherd et al., 2005, 2008; Wise et al., 2005; Rejali et al., 2007; Agterberg et al., 2008, 2009; Chikar et al., 2008; Glueckert et al., 2008; Havenith et al., 2011; Leake et al., 2011; Pettingill et al., 2011). The third, and most often explored alternative, is GDNF (Ylikoski et al., 1998; Kuang et al., 1999; Suzuki et al., 2000; Yagi et al., 2000; Kanzaki et al., 2002; Liu et al., 2008; Maruyama et al., 2008; Scheper et al., 2009; Fransson et al., 2010). The neurotrophin NGF is applied in a few studies (Shah et al., 1995, Schindler et al., 1995, Gillespie et al., 2004), and NT4/5 in only one study (Gillespie et al., 2004). Perhaps the most striking result is that, despite the fact that BDNF and NT-3 are the predominant trophic factors in the unaffected adult organ of Corti, other neurotrophic factors give a similar positive effect (Gillespie et al., 2004; Table 2). On the one hand this is remarkable because neurotrophic factors other than BDNF and NT-3 are traditionally not thought to play important roles in the cochlea; on the other hand it is not surprising given the presence of their respective receptors in SGCs. Unfortunately, the studies with CNTF (Shinohara et al., 2002; Yamagata et al., 2004) and FGF (Miller et al., 2007; Glueckert et al., 2008) did not include an experimental group which received only this particular neurotrophic factor. Because both factors were co-administered with BDNF it is not possible to discuss their individual contribution to the observed protective effect.

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

osmotic pump with cannula into scala tympani

osmotic pump with cannula into scala tympani encapsulated BDNF overexpressing Schwann cells injected into scala tympani

variable delay 5 days

2 weeks

BDNF (100 μg /ml) + CNTF 2 weeks (100 ng/ml) (at 0.5 μl/h) 6 weeks for 4 weeks

BDNF (100 μg/ml) + CNTF no delay (100 ng/ml) (at 0.5 μl/h) for 26 days

BDNF (100 μg/ml at 0.25 μl/h) for 4 weeks BDNF (6 μg) for 2 weeks or 4 weeks BDNF (94 μg/ml at 0.25 μl/h) for 10 weeks BDNF (>100 times more than controls) for 2 weeks Or 4 weeks

2.9 1.9

2.2

1.3 1.4

1.3 1.3 1.3

3.0

2.0

1.6 1.5

5 days

BDNF (4-7-fold higher level 1 week than controls) for 41 days BDNF (100 μg/ml at 0.25 2 weeks μl/h) for 4 weeks BDNF (80 days) 4 days

3.2

1.6 0.9 (2 weeks cessation)

1.3

18 days

1 week

0.6 0.6 0.8

1.0

0.7

1.0

0.9

0.9 0.3

1.0 1.0 1.3

1.4

1.6

1.2

treated/ treated/ normal untreated*

1.0 1.0 1.1

0.9

1.2

1.0

treated/ normal*

1.8

treated/ untreated*

delay

SGCs

packing density (ratio) perikaryal area (ratio) peripheral processes

treatment

2 weeks

BDNF (5.4 μg/ml at 0.25 μl/h) for 4 weeks BDNF (10 μg) for 4 weeks

BDNF (84 µg) for 1 week

BDNF (62.5 μg/ml at 0.25 μl/h) for 4 weeks

osmotic pump with cannula into scala tympani

biodegradable hydrogel on RWM osmotic pump with cannula into scala tympani osmotic pump with cannula into scala tympani transfected cell coated electrode in scala tympani osmotic pump with cannula into scala tympani adenoviral transfection into scala tympani osmotic pump with cannula into scala tympani Gelfoam on RWM

treatment

-5 dB -2 dB

-4 dB

McGuinness and Shepherd, 2005 Shepherd et al., 2005 Rejali et al., 2007

Endo et al., 2005

Gillespie et al., 2003

Yamagata et al., 2004

Shinohara et al., 2002

Agterberg et al., 2008 -2 dB monopolar Chikar et al., -9 dB bipolar 2008 0 dB Agterberg et al., 2009 0 dB Havenith et al., 0 dB 2011 -6 dB Leake et al., 2011 Pettingill et al., 2011

-6 dB

-2 dB

treated/ untreated

eABR threshold references

means

Table 2. Various in vivo strategies of neurotrophic treatment of the cochlea after deafening.

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39 Chapter 2

osmotic pump with cannula into scala tympani osmotic pump with cannula into scala tympani adenoviral transfection into scala tympani

osmotic pump with cannula into scala tympani

osmotic pump with cannula into scala tympani osmotic pump with cannula into scala tympani osmotic pump with cannula into scala tympani

osmotic pump with cannula into scala tympani

single infusion/

osmotic pump with cannula into scala tympani

osmotic pump with cannula into scala tympani osmotic pump with cannula into scala tympani alginate beads on RWM/ into scala tympani osmotic pump with cannula into scala tympani

osmotic pump with cannula into scala tympani

NGF (200 μg/ml at 0.5 μL/h) for 2 weeks GDNF (72 ng/h for 7 days, then 50 ng/h for 14 days) GDNF (24 or 21 days)

NGF (200 μg/ml at 0.5 μL/h) for 2 weeks

4 days 1 week

no delay 2 weeks 4 days

no delay

2 weeks

5 days 33 days 2 weeks

1.7 1.5

1.1 1.4 3.0

1.7 1.6 1.6 1.5

1.4 1.6 1.8

1.0 1.2 2.1

NT-3 (100 ng) for 1 week 4 weeks NT-3 (140 ng) for 1 week NT-3 + BDNF (50 μg/ml at 0.25 μl/h) for 4 weeks

BDNF + NT-3 (50 μg/ml at 0.25 μl/h) for 4 weeks BDNF + NT-3 (30 μg/ml at 0.25 μl/h) for 4 weeks BDNF NT-3 NT-4/5 or NGF (62.5 μg/ml at 0.25 μl/h) for 4 weeks

1.5 1.2

1.4 1.3 3.9 3.3 3.9

1.3 2.3 6.6

2.4 3.2 1.7

1 week

BDNF (50ng/ml) or NT-3 (50ng/ml) for 2 weeks

5 days BDNF + NT-3 BDNF NT-3 (1 mg/ml at 2.5 µl/h) for 8 weeks

1 week

NT-3 (4 weeks)

BDNF (100 μg/ml) + 3 days FGF (50 ng/ml) for 26 days 3 weeks NT-3 (5 µl/h) for 2 weeks no delay

4 day 3 weeks 6 weeks

BDNF (100 μg/ml) + FGF (50 ng/ml) at 0.5 μl/h for 26 days

0.7 0.7

1.0 0.2 1.0

0.9 0.9 0.9 0.8

1.0 0.7 0.9

0.5 0.8 0.9

0.9 0.8

0.9 0.8 0.9

0.8 1.1 1.0

1.2 1.2 0.6

1.1 1.3 1.9

1.2 1.4 1.7

1.1 1.3 1.9

0.9 1.1 2.0

1.9

1.1 1.4

2.2 4.6 2.2

-3 dB

Ylikoski et al., 1998 Yagi et al., 2000

Shah et al., 1995

Schindler et al., 1995

Landry et al., 2011 Gillespie et al., 2004

Wise et al., 2005

Richardson et al., 2005

Miller et al., 1997

Glueckert et al., 2008 Ernfors et al., 1996 Noushi et al., 2005 Staecker et al., 1996

Miller et al., 2007

Neurotrophins and their role in the cochlea

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

44 no delay

GDNF (100 ng/ml at 0.5 3 weeks μl/h) for 4 weeks GDNF (1 μg/ml at 0.5 μl/h) 4 weeks for 4 weeks cocktail of neurotrophins 5 weeks for 8 months

GDNF (10μl/ml) +/- AO (at 0.5 μl/h) for 4 weeks

no delay

4 days

1.1

2.6

1.4

3.2

2.1

1.7

0.7

0.9

1.0

treated/ treated/ normal untreated*

1.1

treated/ normal*

1.8

treated/ untreated*

delay

SGCs packing density (ratio) perikaryal area (ratio) peripheral processes

treatment

-15 dB

-6 dB

treated/ untreated

Scheper et al., 2009 Fransson et al., 2010 Wise et al., 2011

Maruyama et al., 2008

Kanzaki et al., 2002 Liu et al., 2008

eABR threshold references

All studies in guinea pigs with exception of McGuinness and Shepherd (2005) and Liu et al. (2008): rat; Leake et al. (2011) and Wise et al. (2011): cat. All animals deafened with aminoglycosides except Ylikoski et al. (1998): noise-induced hearing loss. Values averaged over all available cochlear locations. AO, anti-oxidants; RWM, round window membrane. *: non-significant values in italics.

osmotic pump with cannula into scala tympani osmotic pump with cannula into scala tympani encapsulated pig choroid plexus cells injected into scala tympani

osmotic pump with cannula into scala tympani

GDNF (39 days)

adenoviral transfection into scala tympani adenoviral transfection through round window membrane

GDNF (>10 times more than in control ears) for 1 month

treatment

means

Table 2. Continued

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39 Chapter 2

Neurotrophins and their role in the cochlea

It is important to realize that, besides the differences in biochemical and systemic effectiveness of the neurotrophic factors that are used, the method of administration of these compounds can determine the magnitude of their neurotrophic efficacy. Global administration by intraventricular infusion, as applied in the pioneering work of Hefti (1986) and Gage et al. (1988), evidently is an inefficient way to deliver neurotrophins to the inner ear. Furthermore, as the experimental intracochlear delivery of neurotrophins is aimed to eventually facilitate clinical applications, multiple techniques have been developed trying to minimize the surgical or post-surgical discomfort. Direct intracochlear delivery is more efficacious than delivery onto the round window (Table 2), but the surgical procedure to enter the cochlea bears a higher risk of both infection and further hearing loss.

3.3. Clinical feasibility In terms of SGC preservation, the most potent means of administration of neurotrophic factors seems to be intracochlear infusion by use of an osmotic pump. Unfortunately, when it comes to clinical applicability there are various objections against this method, such as its invasive nature â&#x20AC;&#x201C; raising the risk of both infection and loss of possible residual hearing â&#x20AC;&#x201C; and the finite supply of neurotrophic factors (for a review on the feasibility of clinical application for the various administration techniques see e.g., Pettingill et al., 2007; Bowe and Jacob, 2010). On the other hand, the intracochlear administration of neurotrophic factors can be combined with cochlear implant surgery, which means that there is no risk of infection or damage in addition to that of the cochlear implantation itself (e.g., Staecker et al., 2010). The development of less invasive techniques includes injection through the tympanic membrane (Kuang et al., 1999) and the placement of neurotrophic factor containing alginate beads (Noushi et al., 2005), hydrogel (Endo et al., 2005) and gelfoam (Havenith et al., 2011) onto the round window membrane. These techniques yield only moderately positive results, probably because their preservative effect tends to be limited to the basal part of the cochlea (Noushi et al., 2005; Havenith et al., 2011). Moreover, it is limited over time by the fact that immediately upon placement release of the neurotrophic factor starts to decrease.

3.4. Cell-based therapy for permanent neurotrophic supply Cell-based therapy is a promising method to provide a permanent supply of neurotrophic factors to the cochlea. Although introducing alien cells into the cochlea bears risk of immune response or migration, several techniques have been developed to considerably minimize these risks (Rejali et al., 2007; Pettingill et al., 2011; Wise et al., 2011). Rejali et al. (2007) showed that BDNF-releasing fibroblasts attached to a 45

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 2

cochlear implant electrode have a positive effect on SGC survival in the basal half of the cochlea after 48 hours. However, the life span of these fibroblasts was limited to several weeks, probably due to fixation of the cells to the electrode. The idea of restricting the movement of NTF-producing cells has been further improved by encapsulating them in 500-600 µm alginate particles that are permeable to NTFs, nutrients and waste products and are injected into the scala tympani (Pettingill et al., 2011; Wise et al., 2011). Thus encapsulated cells survived and kept producing NTFs for at least eight months in cats (Wise et al., 2011). Cell-based techniques provide the means to give continuous neurotrophic support to SGCs. This makes them in principle good candidates for future clinical applications, but the long-term effects of elevated neurotrophin concentrations must be thoroughly assessed (e.g., risk of tumorigenesis) since cessation of cell-based treatment would be difficult to achieve. 3.5. Functional changes after neurotrophic treatment Overall, all neurotrophic factors that have been used to counteract SGC degeneration after hair cell loss proved successful to some extent (see Table 2). However, the diversity of criteria that are used to assess neurotrophic efficacy after ototoxic treatment complicates a closer examination. Most studies define neurotrophic efficacy by the potential to enhance SGC survival compared to non-treated SGC populations. This is undeniably one of the main effects of treatment with neurotrophic factors, but in addition to the fact that in humans the mere number of surviving SGCs is not predictive for CI performance (Linthicum et al., 1991; Khan et al., 2005; Fayad and Linthicum, 2006), the functional condition of the surviving SGC population is neglected by more than half of the studies reviewed here. Admittedly, the aim of several of these studies is to develop a novel delivery method – less invasive or with longer-lasting effect, but especially when a therapy is being fine-tuned for possible clinical application its effects should be assessed at every level, including the functional level. Now that it has been sufficiently demonstrated that the administration of all investigated neurotrophic factors will enhance SGC survival, the next step is to characterize the differences between the various factors – both functionally and morphologically. For example, BDNF or NT-3 administration appears to increase the SGC soma in addition to its positive effect on survival (Ernfors et al., 1996; Richardson et al., 2005; Shepherd et al., 2005, 2008; Agterberg et al., 2008; Glueckert et al., 2008; Leake et al., 2011). Indeed, it is known that activation of Trk receptors induces cell growth via the Akt pathway (e.g. Reichardt, 2006), although Akt activated by GDNFRET interaction (Takahashi, 2001) does not seem to induce SGC growth (Scheper et al., 2009; Fransson et al., 2010). For both cellular physiology and electrical excitability it is 46

Neurotrophins and their role in the cochlea

arguably best for SGCs to retain their original size. Hence, based on effect on size GDNF would be preferable above the neurotrophins. The investigation into the effect of neurotrophic treatment of SGCs on their functional properties is mostly limited to the assessment of threshold shifts of the electrically evoked auditory brainstem response (eABR). While this measure is certainly informative of the overall excitability of the auditory nerve, much more electrophysiological measures can be derived from eABR or eCAP (electrically evoked compound action potential) recordings, which can be used to quantify the functional effect of neurotrophic treatment and can ultimately be predictive for human CI user performance. For example, Prado-Guitierrez et al. (2006) showed that – for both eABR and eCAP recordings – the beneficial effect of the introduction of an inter-phase gap in biphasic stimuli is smaller in deafened animals than in normal hearing controls. The auditory nerve of deafened animals treated with neurotrophic factors might regain this sensitivity, but only if the neurotrophic support adequately mimics the physiologically healthy situation as found in normal hearing animals. In human CI users, Kim et al. (2010) found that the eCAP slope (of eCAP amplitude as function of current) strongly correlates with speech perception. Furthermore, for speech perception in CI users carrier pulse trains are used, and it has been shown that high-rate pulse trains (4 kHz) increase the dynamic range (Pfingst et al., 2007). However, the same study also showed that high pulse rates increase psychophysical modulation detection thresholds, which is likely caused by the auditory nerve’s limited ability to follow high stimulation rates because of refractoriness. So in order to facilitate speech processing by CIs, it is crucial to maintain the auditory nerve’s responsiveness, or, more specifically, to maintain SGC temporal electrophysiological properties. For example, lower psychophysical thresholds have been found in guinea pigs treated with BDNF compared to untreated controls (Chikar et al., 2008). And although in this particular study a correlation between SGC survival and psychophysical thresholds has not been found, these thresholds might correlate with other functional or even morphological changes of SGCs in response to the BDNF treatment. Mechanisms underlying functional changes may be found in altered firing characteristics. Adamson et al. (2002) showed in cultured SGCs that the natural gradient in these characteristics along the tonotopic axis disappears with administration of exogenous neurotrophins: treatment with BDNF gives all SGCs the properties of basally located neurons with respect to latency, adaptation and number of action potentials during a single period of depolarization, while NT-3 gives SGCs the response characteristics of apical neurons. While for SGC functioning in the intact cochlea this gradient is important for proper transduction of a wide range of frequencies, this need might arguable be different in cochlear implant patients. Whether the fast basal 47

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 2

properties or the slower apical characteristics would be more appropriate for electrical stimulation is yet to be determined. Given these findings, the choice for a specific neurotrophic factor for clinical application should at least partly depend on its effect on the electrophysiological behavior of SGCs. For mere SGC survival the neurotrophins might be interchangeable, but it is clear that, at least to some extent, the functional properties of SGCs depend on the neurotrophin that is applied (Adamson et al., 2002; Davis, 2003).

3.6. Peripheral process outgrowth induced by neurotrophic treatment The first symptom of SGC degeneration after hair cell loss is the regression of the afferent peripheral processes that formerly synapsed with the inner hair cells (Leake and Hradek, 1988; Altschuler et al., 1999, Glueckert et al., 2008). Whereas SGC loss is irreversible since there is no evidence for neurogenesis in the adult cochlea (Landry et al., 2011), these peripheral processes can regrow towards a new source of neurotrophic factors which act as chemoattractants (Altschuler et al., 1999; Wise et al., 2005, 2011; Miller et al., 2007; Glueckert et al., 2008; Shibata et al., 2010). While this regrowth is definitively informative with regard to the health of the SGC population (as argued by Wise et al., 2005) and thereby informative of the trophic potential of the neurotrophic factor in question, the functional consequences of afferent resprouting for CI users is unclear (Shibata et al., 2011). Because resprouting of peripheral processes decreases the distance between electrode and target tissue the stimulation current can be lowered, which is good in terms of power consumption of the implant and in terms of reducing risk of stimulation induced tissue damage (Merrill et al., 2005). However, the site of excitation initiation is not necessarily located at the smallest distance to the stimulation electrode, but may depend on diameter of the neural element, the number of available ion channels in this compartment and the excitability of neighboring compartments (Rattay et al., 2001). A homogeneous population of excitable tissue without the complexity of peripheral processes might be a more efficient interface than a heterogeneous population with various degrees of resprouted processes. Recent developments of gene transfer technology may overcome this heterogeneity and provide an optimal interface, as it allows guided regrowth of peripheral processes (Shibata et al., 2010, 2011). In addition to altering the interface for CIs, the potential of a neurotrophic factor to induce peripheral process outgrowth can be beneficial for hair cell reinnervation as well. This can be relevant with respect to possible hair cell replacement by stem cell therapy or after deafferentation â&#x20AC;&#x201C; but not hair cell loss â&#x20AC;&#x201C; caused by excitotoxicity as decribed by, for example, Puel et al. (1998) and Kujawa and Liberman (2009). 48

Neurotrophins and their role in the cochlea

4. Exogenous neurotrophins and endogenous receptors In this section we relate the previous two sections to each other. When relating the expression patterns of the various NTFs that are present in the inner ear during development or adulthood with the findings from the literature search on SGC survival after treatment with NTFs (sections 2 and 3, respectively), several points of discussion arise. The most remarkable observation is that all NTFs that have been experimentally applied increased SGC survival rate in deafened adult animals, even though only NT-3 and GDNF are found to be expressed in the healthy adult organ of Corti. Furthermore, the complex and highly dynamic temporal and spatial variation in expression pattern for at least eight neurotrophic factors that are at some point present in the inner ear suggests specialized signaling leading to specific functions for receptor cells. The application of a physiologically high dose of only one type of neurotrophic factor – although it apparently rescues SGCs from degeneration – may well have other significant morphological and functional consequences. 4.1. TrkA and TrkB in the adult cochlea – silent receptors? Only NT-3 and GDNF are found in the adult cochlea (Table 1, Fig. 3). The first question to address is why SGCs in the adult cochlea express receptors – notably, TrkA and TrkB – while their respective ligands – NGF and BDNF or NT-4/5 – are not present. A substantial number of studies, listed in Table 2, inform us that these receptors are functional and able to prevent SGC degeneration when activated with exogenous neurotrophins. Three reasons for their presence are possible: (1) the presence of TrkB is a remnant, reflecting the importance of BDNF during development. This, however, would still leave us with the question why TrkA is expressed, since NGF is not found in the cochlea in any developmental stage, nor is it found during adulthood. Another possibility is that (2) NGF and BDNF are present in the cochlea but in such low concentrations that they are undetectable. Finally, (3) both receptors might be present in anticipation of a possible NGF or BDNF signal. In this case, there should be a specific (and not too uncommon) change in cochlear physiology that triggers a sudden supply of BDNF or NGF. Direct evidence to support this hypothesis is lacking, but it has been shown that within eight hours after noise trauma GDNF levels are doubled (Nam et al., 2000). A similar mechanism may exist in which BDNF or NGF levels are upregulated in response to cochlear stress due to acoustic overstimulation (noise trauma) or exposure to ototoxins. Excitotoxicity after noise trauma might trigger activity-dependent neurotrophin expression that provides a built-in protective mechanism to the peripheral processes of SGCs, and their synapses with hair cells (Hong et al., 2008).

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 2

4.2. Long-term effect of temporary neurotrophic treatment As discussed above, neurotrophic treatment of SGCs is very effective; survival after neurotrophic treatment can be as high as 100% in the basal cochlear turns, even when treatment is several weeks delayed (Ylikoski et al., 1998; Shepherd et al., 2005; Miller et al., 2007; Agterberg et al., 2009). In addition to a delayed onset of neurotrophic treatment, several studies have looked into the survival of SGCs several weeks after cessation of the treatment as well (Gillespie et al., 2003; Maruyama et al., 2008; Agterberg et al., 2009; Fransson et al., 2010). Another study examined survival after cessation of neurotrophic treatment in animals that were additionally treated with chronic electrical stimulation during the entire experiment (Shepherd et al., 2008). This assessment is of great relevance because a sustained preservative effect of temporary treatment would simplify the development of clinical application to a great extent. Unfortunately, the findings so far do not lead to solid conclusions since fundamentally different observations have been made. According to Gillespie et al. (2003) cessation of neurotrophic treatment leads to accelerated SGC death, which leaves treated cochleas indistinguishable from untreated controls as early as two weeks later. Shepherd et al. (2008) showed that degeneration was significantly reduced, in particular in the basal part of the cochlea, which they attributed to chronic electrical stimulation that was applied during and after neurotrophic treatment. However, in sharp contrast to the Gillespie et al. study, the other three studies (Maruyama et al. 2008; Agterberg et al., 2009; Fransson et al., 2010) showed a sustained pattern of preservation that does not change upon treatment cessation. The discrepancy is not caused by a difference in neurotrophic factors (BDNF vs. GDNF) as previously suggested by Maruyama et al. (2008), since Agterberg et al. (2009) used the same neurotrophic factor as Gillespie et al. (2003), namely BDNF. Also, it appears that the delay between deafening and treatment should not be relevant given the variety of treatment onset delays. Data on cessation of treatment with NTFs in other neural systems are rare and caution should be exercised when comparing those with the findings referred to here. For example, Montero and Hefti (1988) found a decline in neuronal preservation after cessation of NGF treatment in lesioned rats. These results may appear homologous to those of Gillespie et al. (2003), but there are crucial methodological differences â&#x20AC;&#x201C; the main difference being the stepwise intraventricular administration of high dose NGF versus continued local infusion. Moreover, different tissues may act differently in response to temporary neurotrophic treatment and subsequent cessation. The difference between a failed prolonged preservation in one and successful preservation in the other three studies might be attributed to the weekly eABR recordings in the latter, as was suggested in Agterberg et al. (2009). However, although the positive effect of electrical stimulation on SGC survival has been demonstrated 50

Neurotrophins and their role in the cochlea

(Shepherd et al., 2005, 2008; Scheper et al., 2009), there is evidence that suggests that this effect only exists in the presence of neurotrophic factors (Shepherd et al., 2005; Agterberg et al., 2010), which is discussed in terms of activity-dependent Trk receptor trafficking in Wise et al. (2011). Therefore, although there is evidence to suppose that electrical stimulation has a preservative effect on SGCs after cessation of the neurotrophic treatment (Shepherd et al., 2008), and that even as little as 30 minutes of eABR recording weekly may still be beneficial (Miller and Altschuler, 1995; Mitchell et al., 1997), there should be a source of neurotrophic factors to exploit the electrical stimulation-induced Trk recruitment weeks after cessation of treatment with exogenous neurotrophic factors. It is unlikely that this source is the NTF that was infused weeks before, given the short in vivo half-life of BDNF (approximate 3 hours; Kishino et al., 2001) and GDNF (approximately 37 hours; Ejstrup et al., 2010). 4.3. Autocrine neurotrophic support It has been hypothesized that the continued SGC preservation after cessation of neurotrophic treatment might reflect a re-established endogenous neurotrophic supply (Maruyama et al., 2008), but with a virtually complete loss of hair cells the source for this support must be found elsewhere. Zha et al. (2001) showed that cultured rat SGCs produce BDNF (in vivo data not available). Although these cells were harvested during early postnatal development – a period in which hair cells too are still releasing BDNF – the idea of a self-sustaining mechanism for SGCs is appealing. Indeed, there is evidence that autocrine BDNF signaling plays an important role in survival and growth of developing and mature dorsal root ganglion cells (Davies and Wright, 1995), developing hippocampal neurons (Cheng et al., 2011) and mature pituitary melanotrope cells (Kuribara et al., 2011a). Besides induction of survival or growth, the activation of TrkB by BDNF leads to an increased release of BDNF via cyclic AMP and protein kinase A activity, TrkB trafficking via PI-3 kinase activation (Cheng et al., 2011) and BDNF transcription via ERK (Kuribara et al., 2011b) – all of which acting to sustain or amplify the autocrine loop. Whether it is possible that this autocrine survival mechanism is activated in SGCs in deafened animals remains to be investigated, as well as the then apparent role that temporary neurotrophic treatment has in inducing it. More importantly, it should be investigated whether this self-sustaining state is of permanent nature. A first step would be to study expression of neurotrophic factors and their receptors in cochleas of deafened animals of which the SGC population has survived several weeks after cessation of neurotrophic treatment.

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 2

5. Concluding remarks and future directions

5.1 Concluding remarks Several points relevant for the role of neurotrophins in cochlear physiology can be concluded on the basis of this review. 1) The important neurotrophins in the developing cochlea are NT-3 and BDNF. They have distinct functions in the development and maintenance of the SGCs, and interchangeability is low. In the healthy adult cochlea the only neurotrophin observed is NT-3, while remarkably, the receptors for the neurotrophins that have not been observed, TrkA and TrkB are present alongside the receptor of NT-3, TrkC. The interchangeability of the neurotrophins in the adult cochlea may be high. 2) In cochleas deprived of hair cells (as in severe sensorineural hearing loss) SGCs degenerate, and this degeneration can be prevented very successfully by exogenous neurotrophins which replace the lost endogenous neurotrophic support. All four neurotrophins and GDNF work similarly well in preserving the SGCs, not only NT-3 and GDNF that have been shown to be present in the adult cochlea. This is not surprising given the presence of all receptors in the adult SGCs, and it supports the hypothesis that the neurotrophins are highly interchangeable in the adult cochlea. 3) Withdrawal of exogenous neurotrophic treatment does not lead to degeneration of the SGCs (at least it is slower than the degeneration induced by a sudden loss of hair cells). An attractive mechanism explaining the continued preservative effect is autocrine neurotrophic support which is triggered by the temporary exogenous neurotrophic treatment. Recent evidence for such an autocrine mechanism is provided by research in other neural systems. 4) Neurotrophic treatment may be considered for various types of neurodegenerative disorders, including treatment of the auditory nerve in case of severe sensorineural hearing loss. A clinical trial has thus far only been performed in AD patients, which demonstrated negative side-effects outweighing small positive effects.

5.2 Future perspectives In this review various fundamental questions (regarding, e.g., interchangeability of neurotrophins, and an autocrine mechanism) are raised, which need to be addressed before considering the possible clinical applications of neurotrophins in the inner ear. In the following we will formulate these questions, and we will provide some directions on how to tackle these questions. 1) What is the function of Trk receptors without the presence of neurotrophins? Although most of the neurotrophins are not present in the healthy adult cochlea, their respective receptors are all expressed in SGCs. Their function could be related to stress 52

Neurotrophins and their role in the cochlea

situations, in which a change in neurotrophic signaling calls for different receptors. The expression pattern of most neurotrophins and their receptors in the inner ear during development and during adulthood has been studied extensively. But if we want to understand and mimic repair and preservation of tissue in distress, we should first focus on factors that are involved in the repair mechanisms of the cochlea itself. In addition, we should learn from research into other neurodegenerative disorders which are accompanied by changes in neurotrophin expression. 2) Does treatment with exogenous neurotrophins cause morphological or physiological changes to SGCs which affect their ability to transduce electrically presented information by cochlear implants (and even by remaining hair cells)? More detailed functional and morphological parameters should be measured after neurotrophic treatment. Mere SGC survival does not necessarily relate to SGC health nor to hearing preservation. In addition to the threshold and amplitude of the compound action potential or the auditory brainstem response, electrophysiological properties such as excitability and refractoriness should be included, as well as psychophysical measures. 3) Is there a risk of tumorigenesis attached to permanent cell-based treatment in the long run? Since neurotrophins and Trk receptors play important roles in tumor growth, the risks of permanently enhanced concentrations in the cochlea should be seriously considered. To minimize the risk of tumorigenesis, the release of neurotrophins by these implanted cells should be equal to the release by hair cells in the healthy cochlea. To this end, choroid plexus cells, which by nature release a variety of neurotrophins (Wise et al., 2011), might be preferable over transfected cells that produce and release relatively large amounts of neurotrophins. 4) Is the effect of temporary neurotrophic treatment transient in nature or do SGCs have the potential to support themselves in an autocrine fashion in vivo? The prolonged preservative effect of neurotrophic treatment described in section 4 is an important finding both for clinical application purposes and fundamental understanding of the underlying mechanisms. An obvious but crucial question is whether the prolonged preservative effect holds for longer than a mere 4 weeks after cessation of the treatment. Moreover, and irrespective of its duration, understanding the cellular and subcellular mechanisms underlying this prolonged effect can be crucial in optimizing the neurotrophic treatment. 5) Finally, the above questions can be addressed for each of the neurotrophins separately: do the neurotrophins have the same effect for the different aspects? An intriguing question is whether BDNF, present in the developing but not in the mature cochlea, has a different effect than NT-3, present in both developing and mature cochlea. 53

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 2

Acknowledgements We are grateful to Marlies Knipper for helpful comments on the manuscript.

Neurotrophins and their role in the cochlea

References Acheson, A., Conover, J.C., Fandl, J.P., DeChiara, T.M., Russell, M., Thadani, A., Squinto, S.P., Yancopoulos, G.D., Lindsay, R.M., 1995. A BDNF autocrine loop in adult sensory neurons prevents cell death. Nature 374, 450-453. Adamson, C.L., Reid, M.A., Davis, R.L., 2002. Opposite actions of brain-derived neurotrophic factor and neurotrophin-3 on firing features and ion channel composition of murine spiral ganglion neurons. J. Neurosci. 22, 1385-1396. Agerman, K., Hjerling-Leffler, J., Blanchard, M.P., Scarfone, E., Canlon, B., Nosrat, C., Ernfors, P., 2003. BDNF gene replacement reveals multiple mechanisms for establishing neurotrophin specificity during sensory nervous system development. Development 130, 1479-1491. Agterberg, M.J.H., Versnel, H., De Groot, J.C.M.J., Smoorenburg, G.F., Albers, F.W.J., Klis, S.F.L., 2008. Morphological changes in spiral ganglion cells after intracochlear application of brainderived neurotrophic factor in deafened guinea pigs. Hear. Res. 244, 25-34. Agterberg, M.J.H., Versnel, H., Van Dijk, L.M., De Groot, J.C.M.J., Klis, S.F.L., 2009. Enhanced survival of spiral ganglion cells after cessation of treatment with brain-derived neurotrophic factor in deafened guinea pigs. J. Assoc. Res. Otolaryngol. 10, 355-367. Agterberg, M.J.H., Versnel, H., de Groot, J.C.M.J., van den Broek, M., Klis, S.F.L., 2010. Chronic electrical stimulation does not prevent spiral ganglion cell degeneration in deafened guinea pigs. Hear. Res. 269, 169-79. Aid, T., Kazantseva, A., Piirsoo, M., Palm, K., Timmusk, T., 2007. Mouse and rat BDNF gene structure and expression revisited. J. Neurosci. Res. 85, 525-535. Allen, S.J., Dawbarn, D., 2006. Clinical relevance of the neurotrophins and their receptors. Clin. Sci. 110, 175-191. Altar, C.A., Boylan, C.B., Jackson, C., Hershenson, S., Miller, J., Wiegand, S.J., Lindsay, R.M., Hyman, C., 1992. Brain-derived neurotrophic factor augments rotational behavior and nigrostriatal dopamine turnover in vivo. Proc. Natl. Acad. Sci. USA 89, 11347-11351. Altschuler, R.A., Cho, Y., Ylikoski, J., Pirvola, U., Magal, E., Miller, J.M., 1999. Rescue and regrowth of sensory nerves following deafferentation by neurotrophic factors. Ann. NY Acad. Sci. 884, 305-311. Ard, M.D., Morest, D.K., Hauger, S.H., 1985. Trophic interactions between the cochleovestibular ganglion of the chick embryo and its synaptic targets in culture. Neurosci. 16, 1151-1170. ArĂŠvalo, J.C., Wu, S.H., 2006. Neurotrophin signaling: many exciting surprises! Cell. Mol. Life Sci. 63, 1523-1537. Baj, G., Tongiorgi, E., 2009. BDNF splice variants from the second promoter cluster support cell survival of differentiated neuroblastoma upon cytotoxic stress. J. Cell. Sci. 122, 36-43. Barbacid, M., 1994. The Trk family of neurotrophin receptors. J. Neurobiol. 25, 1386-1403. Bianchi, L.M., Conover, J.C., Fritzsch, B., DeChiara, T., Lindsay, R.M., Yancopoulos, G.D., 1996. Degeneration of vestibular neurons in late embryogenesis of both heterozygous and homozygous BDNF null mutant mice. Development 122, 1965-1973. Bothwel, M., 1995. Functional interactions of neurotrophins and neurotrophin receptors. Ann. Rev. Neurosci. 18, 223-253. Bowe, S.N., Jacob, A., 2010. Round window perfusion dynamics: implications for intracochlear therapy. Curr. Opin. Otolaryngol. Head Neck Surg. 18, 377-385. Bramham, C.R., Messaoudi, E., 2005. BDNF function in adult synaptic plasticity: The synaptic consolidation hypothesis. Prog. Neurobiol. 76, 99â&#x20AC;&#x201C;125. Capsoni, S., Ugolini, G., Comparini, A., Ruberti, F., Berardi, N., Cattaneo, A., 2000. Alzheimer-like neurodegeneration in aged anti-nerve growth factor transgenic mice. Proc. Natl. Acad. Sci. USA 97, 6826-6831. 55

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 2

Chao, M.V., 2000. Trophic factors: An evolutionary cul-de-sac or door into higher neuronal function? J. Neurosci. Res. 59, 353-355. Chen, L.W., Yung, K.K.L., Chan, Y.S., Shum, D.K.Y., Bolam, J.P., 2008. The proNGF-p75NTR-Sortilin Signalling Complex as New Target for the Therapeutic Treatment of Parkinson’s Disease. CNS Neurol. Disord. Drug Targets 7, 512-523. Cheng, P.L., Song, A.H., Wong, Y.H., Wang, S., Zhang, X., Poo, M.M., 2011. Self-amplifying autocrine actions of BDNF in axon development. Proc. Natl. Acad. Sci. USA 108, 18430-18435. Chikar, J.A., Colesa, D.J., Swiderski, D.L., Di Polo, A., Raphael, Y., Pfingst, B.E., 2008. Over-expression of BDNF by adenovirus with concurrent electrical stimulation improves cochlear implant thresholds and survival of auditory neurons. Hear. Res. 245, 24-34. Coppola, V., Kucera, J., Palko, M.E., Martinez-De Velasco, J., Lyons, W.E., Fritzsch, B., Tessarollo, L., 2001. Dissection of NT3 functions in vivo by gene replacement strategy. Development 128, 4315-4327. Dai, C.F., Steyger, P.S., Wang, Z.M., Vass, Z., Nuttall, A.L., 2004. Expression of Trk A receptors in the mammalian inner ear. Hear. Res. 187, 1-11. Davies, A.M., Wright, E.M., 1995. Neurotrophic Factors: Neurotrophin autocrine loops. Curr. Biol. 5, 723-726. Davis, R.L., 2003. Gradients of neurotrophins, ion channels, and tuning in the cochlea. Neuroscientist 9, 311-316. Dawbarn, D., Allen, S.J., 2003. Neurotrophins and neurodegeneration. Neuropathol. Appl. Neurobiol. 29, 211-230. Dechant, G., Barde, Y.A., 1997. Signalling through the neurotrophin receptor p75NTR. Curr. Opin. Neurobiol. 7, 413-418. Dodson, H.C., 1997. Loss and survival of spiral ganglion neurons in the guinea pig after intracochlear perfusion with aminoglycosides. J. Neurocytol. 26, 541–556. Duan, M., Agerman, K., Ernfors, P., Canlon, B., 2000. Complementary roles of neurotrophin 3 and a N-methyl-D-aspartate antagonist in the protection of noise and aminoglycoside-induced ototoxicity. Proc. Natl. Acad. Sci. USA 97, 7597-7602. Edwards, R.H., Selby, M.J., Garcia, P.D., Rutter, W.J., 1988. Processing of the native nerve growth factor precursor to form biologically active nerve growth factor. J. Biol. Chem. 263, 68106815. Ejstrup, R., Kiilgaard, J.F., Tucker, B.A., Klassen, H.J., Young, M.J., La Cour, M., 2010. Pharmacokinetics of intravitreal glial cell line-derived neurotrophic factor: experimental studies in pigs. Exp. Eye Res. 91, 890-895. Endo, T., Nakagawa, T., Kita, T., Iguchi, F., Kim, T.S., Tamura, T., Iwai, K., Tabata, Y., Ito, J., 2005. Novel strategy for treatment of inner ears using a biodegradable gel. Laryngoscope 115, 20162020. Eriksdotter Jonhagen, M.E., Nordberg, A., Amberla, K., Backman, L., Ebendal, T., Meyerson, B., Olson, L., Shigeta, M., Theodorsson, E., Viitamen, M., Winblab, B., Wahlund, L.O., 1998. Intracerebroventricular infusion of nerve growth factor in three patients with Alzheimer’s disease. Dement. Geriatr. Cogn. Disord. 9, 246-257. Ernfors, P., Merlio, J.P., Persson, H., 1992. Cells expressing mRNA for neurotrophins and their receptors during embryonic rat development. Eur. J. Neurosci. 4, 1140-1158. Ernfors, P., Duan, M.L., ElShamy, W.M., Canlon, B., 1996. Protection of auditory neurons from aminoglycoside toxicity by neurotrophin-3. Nat. Med. 2, 463-467. Fagan, A.M., Zhang, H., Landis, S., Smeyne, R.J., Silos-Santiago, I., Barbacid, M., 1996. TrkA, but not TrkC, receptors are essential for survival of sympathetic neurons in vivo. J. Neurosci. 16, 6208-6218. 56

Neurotrophins and their role in the cochlea

Fariñas, I., Jones, K.R., Tessarollo, L., Vigers, A.J., Huang, E., Kirstein, M., de Caprona, D.C., Coppola, V., Backus, C., Reichardt, L.F., Fritzsch, B., 2001. Spatial shaping of cochlear innervation by temporally regulated neurotrophin expression. J. Neurosci. 21, 6170-6180. Fayad, J.N., Linthicum, F.H. Jr., 2006. Multichannel cochlear implants: relation of histopathology to performance. Laryngoscope 116, 1310-1320. Frade, J.M., Rodríguez-Tébar, A., Barde, Y., 1996. Induction of cell death by endogenous nerve growth factor through its p75 receptor. Nature 383, 166-168. Fransson, A., Maruyama, J., Miller, J.M., Ulfendahl, M., 2010. Post-treatment effects of local GDNF administration to the inner ears of deafened guinea pigs. J. Neurotrauma 27, 1745-1751. Fritzsch, B., Fariñas, I., Reichardt, L.F., 1997b. Lack of neurotrophin 3 causes losses of both classes of spiral ganglion neurons in the cochlea in a region-specific fashion. J. Neurosci. 17, 62136225. Fritzsch, B., Pirvola, U., Ylikoski, J., 1999. Making and breaking the innervation of the ear: neurotrophic support during ear development and its clinical implications. Cell Tissue Res. 295, 369-382. Fritzsch, B., Silos-Santiago, I., Bianchi, L.M., Fariñas, I., 1997a. The role of neurotrophic factors in regulating the development of inner ear innervation. Trends Neurosci. 20, 159-164. Fritzsch, B., Silos-Santiago, I., Smeyne, R., Fagan, A.M., Barbacid, M., 1995. Reduction and loss of inner ear innervation in TrkB and TrkC receptor knockout mice: a whole mount DiI and scanning electron microscopic analysis. Aud. Neurosci. 1, 401-417. Fritzsch, B., Tessarollo, L., Coppola, E., Reichardt, L.F., 2004. Neurotrophins in the ear: their roles in sensory neuron survival and fiber guidance. Prog. Brain Res. 146, 265-278. Gage, F.H., Armstrong, D.M., Williams, L.R., Varon, S., 1988. Morphological response of axotomized septal neurons to nerve growth factor. J. Comp. Neurol. 269, 147-155. Gillespie, L.N., Clark, G.M., Bartlett, P.F., Marzella, P.L., 2003. BDNF-induced survival of auditory neurons in vivo: Cessation of treatment leads to accelerated loss of survival effects. J. Neurosci. Res. 71, 785-790. Gillespie, L.N., Clark, G.M., Marzella, P.L., 2004. Delayed neurotrophin treatment supports auditory neuron survival in deaf guinea pigs. Neuroreport 15, 1121-1125. Glueckert, R., Bitsche, M., Miller, J.M., Zhu, Y., Prieskorn, D.M., Altschuler, R.A., Schrott-Fischer, A., 2008. Deafferentation-associated changes in afferent and efferent processes in the guinea pig cochlea and afferent regeneration with chronic intrascalar brain-derived neurotrophic factor and acidic fibroblast growth factor. J. Comp. Neurol. 507, 1602-1621. Green, S.H., Altschuler, R.A., Miller, J.M., 2008. Cell death and cochlear protection. In: Schacht, J., Popper, A.N., Fay, R.R. (Eds.), Auditory Trauma, Protection, and Repair. New York, Springer, pp. 275–319. Greenberg, M.E., Xu, B., Lu, B., Hempstead, B.L., 2009. New insights in the biology of BDNF synthesis and release: implications in CNS function. J. Neurosci. 29, 12764-12767. Haque, N.S.K., Hlavin, M., Fawcett, J.W., Dunnett, S.B., 1996. The neurotrophin NT4/5, but not NT3, enhances the efficacy of nigral grafts in a rat model of Parkinson’s disease. Brain Res. 712, 45-52. Havenith, S., Versnel, H., Agterberg, M.J.H., De Groot, J.C.M.J., Sedee, R., Grolman, W., Klis, S.F.L., 2011. Spiral ganglion cell survival after round window membrane application of brainderived neurotrophic factor using gelfoam as carrier. Hear. Res. 272, 168-177. Hayashi, T., Ray, C.A., Younkins, C., Bermingham-McDonogh, O., 2010. Expression patterns of FGF receptors in the developing mammalian cochlea. Dev. Dyn. 239, 1019-1026. Hefti, F., 1986. Nerve Growth Factor promotes survival of septal cholinergic neurons after fimbrial transections. J. Neurosci. 6, 2155-2162. 57

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 2

Hennigan, A., O’Callaghan, R.M., Kelly, Á.M., 2007. Neurotrophins and their receptors: roles in plasticity, neurodegeneration and neuroprotection. Biochem. Soc. Trans. 35, 424-426. Hoglinger, G.U., Sautter, J., Meyer, M., Spenger, C., Seiler, R.W., Oertel, W.H., Widmer, H.R., 1998. Rat ventral mesencephalon grown as solid tissue cultures: Influence of culture time and BDNF treatment on dopamine neuron survival and function. Brain Res. 813, 313-322. Hong, E.J., McCord, A.E., Greenberg, M.E., 2008. A biological function for the neuronal activitydependent component of Bdnf transcription in the development of cortical inhibition. Neuron 60, 610-624. Hossain, W.A., Morest, D.K., 2000. Fibroblast growth factors (FGF-1, FGF-2) promote migration and neurite growth of mouse cochlear ganglion cells in vitro: immunohistochemistry and antibody perturbation. J. Neurosci. Res. 62, 40-55. Huang, E.J., Reichardt, L.F., 2001. Neurotrophins: roles in neuronal development and function. Annu. Rev. Neurosci. 24, 677-736. Huang, E.J., Reichardt, L.F., 2003. Trk receptors: Roles in neuronal signal transduction. Ann. Rev. Biochem. 72, 609-642. Ivanicevic, L., Saragovi, H.U., 2006. The neurotrophins. In: Kastin, A.J. (Ed.), Handbook of biologically active peptides, Burlington, Academic Press, pp. 1409-1411. Johnson, E.M., Yip, H.K., 1985. Central nervous system and peripheral nerve growth factor provide trophic support critical to mature sensory neuronal survival. Nature 314, 751-752. Kanzaki, S., Stöver, T., Kawamoto, K., Prieskorn, D.M., Altschuler, R.A., Miller, J.M., Raphael, Y., 2002. Glial cell line-derived neurotrophic factor and chronic electrical stimulation prevent VIII cranial nerve degeneration following denervation. J. Comp. Neurol. 454, 350-360. Khan, A.M., Handzel, O., Burgess, B.J., Damian, D., Eddington, D.K., Nadol, J.B. Jr., 2005. Is word recognition correlated with the number of surviving spiral ganglion cells and electrode insertion depth in human subjects with cochlear implants? Laryngoscope 115, 672-677. Kim, J.R., Abbas, P.J., Brown, C.J., Etler, C.P., O’Brien, S., Kim, L.S., 2010. The relationship between electrically evoked compound action potential and speech perception: a study in cochlear implant users with short electrode array. Otol. Neurotol. 31, 1041-1048. Kishino, A., Katayama, N., Ishige, Y., Yamamoto, Y., Ogo, H., Tatsuno, T., Mine, T., Noguchi, H., Nakayama, C., 2001. Analysis of effects and pharmacokinetics of subcutaneously administered BDNF. Neuroreport 12, 1067-1072. Koliatsos, V.E., Nauta, H.J., Clatterbuck, R.E., Holtzman, D.M., Mobley, W.C., Price, D.L., 1990. Mouse nerve growth factor prevents degeneration of axotomized basal forebrain cholinergic neurons in the monkey. J. Neurosci. 10, 3801-3813. Kujawa, S.G., Liberman, M.C., 2006. Acceleration of age-related hearing loss by early noise exposure: evidence of a misspent youth. J. Neurosci. 26, 2115-2123. Kujawa, S.G., Liberman, M.C., 2009. Adding insult to injury: cochlear nerve degeneration after “temporary” noise-induced hearing loss. J. Neurosci. 29, 14077-14085. Kuribara, M., Hess, M.W., Cazorla, M., Roubos, E.W., Scheenen, W.J.J.M., Jenks, B.G., 2011a. Brainderived neurotrophic factor stimulates growth of pituitary melanotrope cells in an autocrine way. Gen. Comp. Endocrinol. 170, 156-161. Kuribara, M., Kidane, A.H., Vos, G.A., de Gouw, D., Roubos, E.W., Scheenen, W.J., Jenks, B.G., 2011b. Extracellular-signal regulated kinase regulates production of pro-opiomelanocortin in pituitary melanotroph cells. J. Neuroendocrinol. 23, 261-268. Küst, B.M., Copray, J.C., Brouwer, N., Troost, D., Boddeke, H.W., 2002. Elevated levels of neurotrophins in human biceps brachii tissue of amyotrophic lateral sclerosis. Exp. Neurol. 177, 419-427. Landry, T.G., Wise, A.K., Fallon, J.B., Shepherd, R.K., 2011. Spiral ganglion neuron survival and function in the deafened cochlea following chronic neurotrophic treatment. Hear. Res. 282, 303-313. 58

Neurotrophins and their role in the cochlea

Leake, P.A., Hradek, G.T., 1988. Cochlear pathology of long term neomycin induced deafness in cats. Hear. Res. 33, 11-33. Leake, P.A., Hradek, G.T., Hetherington, A.M., Stakhovskaya, O., 2011. Brain-derived neurotrophic factor promotes cochlear spiral ganglion cell survival and function in deafened, developing cats. J. Comp. Neurol. 519, 1526-1545. Leake, P.A., Rebscher, S.J., 2004. Anatomical considerations and long-term effects of electrical stimulation. In: Zeng, F.G., Popper, A.N., Fay, R.R. (Eds.), Cochlear Implants: Auditory prostheses and electric hearing, New York, pp. 101–148. Leingärtner, A., Lindholm, D., 1994. Two promoters direct transcription of the mouse NT-3 gene. Eur. J. Neurosci. 6, 1149-1159. Lee, R., Kermani, P., Teng, K.K., Hempstead, B.L., 2001. Regulation of cell survival by secreted proneurotrophins. Science 294, 1945-1948. Lee, Y.-W., Ozeki, M., Juhn, S.K., Lin, J., 2004. Expression of platelet-derived growth factor in the developing cochlea of rats. Acta Otolaryngol. 124, 558-562. Lefebvre, P.P., Van de Water, T.R., Weber, T., Rogister, B., Moonen, G., 1991. Growth factor interactions in cultures of dissociated adult acoustic ganglia: neuronotrophic effects. Brain Res. 567, 306-312. Lin, L.F., Doherty, D.H., Lile, J.D., Bektesh, S., Collins, F., 1993. A glial cell line-derived neurotrophic factor for midbrain dopaminergic neurones. Science 260, 1130-1132 Linthicum, F.H. Jr., Fayad, J., Otto, S.R., Galey, F.R., House, W.F., 1991. Cochlear implant histopathology. Am. J. Otol. 12, 245-311. Liu, Y., Okada, T., Shimazaki, K., Sheykholeslami, K., Nomoto, T., Muramatsu, S., Mizukami, H., Kume, A., Xiao, S., Ichimura, K., Ozawa, K., 2008. Protection against aminoglycoside-induced ototoxicity by regulated AAV vector-mediated GDNF gene transfer into the cochlea. Mol. Ther. 16, 474-480. Lucidi-Phillipi, C.A., Gage, F.H., Shults, C.W., Jones, K.R., Reichardt, L.F., Kang, U.J., 1995. Brainderived neurotrophic factor-transduced fibroblasts: production of BDNF and effects of grafting to the adult rat brain. J. Comp. Neurol. 354, 361-376. Malgrange, B., Rogister, B., Lefebvre, P.P., Mazy-Servais, C., Welcher, A.A., Bonnet, C., Hsu, R.-Y., Rigo, J.-M., Van de Water, T.R., Moonen, G., 1998. Expression of growth factors and their receptors in the postnatal rat cochlea. Neurochem. Res. 23, 1133-1138. Maruyama, J., Miller, J.M., Ulfendahl, M., 2008. Glial cell line-derived neurotrophic factor and antioxidants preserve the electrical responsiveness of the spiral ganglion neurons after experimentally induced deafness. Neurobiol. Dis. 29, 14-21. McFadden, S.L., Ding, D., Jiang, H., Salvi, R.J., 2004. Time course of efferent fiber and spiral ganglion cell degeneration following complete hair cell loss in the chinchilla. Brain Res. 997, 40-51. McGuinness, S.L., Shepherd, R.K., 2005. Exogenous BDNF rescues rat spiral ganglion neurons in vivo. Otol. Neurotol. 26, 1064-1072. Merrill, D.R., Bikson, M., Jefferys, J.G., 2005. Electrical stimulation of excitable tissue: design of efficacious and safe protocols. J. Neurosci. Methods 141, 171-198. Miller, J.M., Altschuler, R.A., 1995. Effectiveness of different electrical stimulation conditions in preservation of spiral ganglion cells following deafness. Ann. Otol. Rhinol. Laryngol. Suppl. 166, 57-60. Miller, J.M., Chi, D.H., O’Keeffe, L.J., Kruszka, P., Raphael, Y., Altschuler, R.A., 1997. Neurotrophins can enhance spiral ganglion cell survival after inner hair cell loss. Int. J. Dev. Neurosci. 15, 631-643. Miller, J.M., Le Prell, C.G., Prieskorn, D.M., Wys, N.L., Altschuler, R.A., 2007. Delayed neurotrophin treatment following deafness rescues spiral ganglion cells from death and promotes regrowth of auditory nerve peripheral processes: effects of brain-derived neurotrophic factor and fibroblast growth factor. J. Neurosci. Res. 85, 1959-1969. 59

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 2

Mitchell, A., Miller, J.M., Finger, P.A., Heller, J.W., Raphael, Y., Altschuler, R.A., 1997. Effects of chronic high-rate electrical stimulation on the cochlea and eighth nerve in the deafened guinea pig. Hear. Res. 105, 30-43. Montero, C.N., Hefti, F., 1988. Rescue of lesioned septal cholinergic neurons by nerve growth factor: specificity and requirement for chronic treatment. J. Neurosci. 8, 2986-2999. Nam, Y.-J., Stöver, T., Hartman, S.S., Altschuler, R.A., 2000. Upregulation of glial cell line-derived neurotrophic factor (GDNF) in the rat cochlea following noise. Hear. Res. 146, 1-6. Noushi, F., Richardson, R.T., Hardman, J., Clark, G., O’Leary, S., 2005. Delivery of neurotrophin-3 to the cochlea using alginate beads. Otol. Neurotol. 26, 528-533. Numakawa, T., Suzuki, S., Kumamaru, E., Adachi, N., Richards, M., Kunugi, H., 2010. BDNF function and intracellular signaling in neurons. Histol. Histopathol. 25, 237-258. Patapoutian, A., Reichardt, L.F., 2001. Trk receptors: mediators of neurotrophin action. Curr. Opin. Neurobiol. 11, 272-280. Pettingill, L.N., Richardson, R.T., Wise, A.K., Shepherd, R.K., O’Leary, S., 2007. Neurotrophic factors and neural prostheses: potential clinical applications based upon findings in the auditory system. IEEE Trans. Biomed. Eng. 54, 1138-1148. Pettingill, L.N., Wise, A.K., Geaney, M.S., Shepherd, R.K., 2011. Enhanced auditory neuron survival following cell-based BDNF treatment in the deaf guinea pig. PLoS One 6, e18733. Pfingst, B.E., Bowling, S.A., Colesa, D.J., Garadat, S.N., Raphael, Y., Shibata, S.B., Strahl, S.B., Su, G.L., Zhou, N., 2011. Cochlear infrastructure for electrical hearing. Hear. Res. 281, 65-73. Pfingst, B.E., Xu, L., Thompson, C.S., 2007. Effects of carrier pulse rate and stimulation site on modulation detection by subjects with cochlear implants. J. Acoust. Soc. Am. 121, 2236-2246. Pirvola, U., Ylikoski, J., Palgi, J., Lehtonen, E. Arumäe, U., Saarma, M., 1992. Brain-derived neurotrophic factor and neurotrophin 3 mRNAs in the peripheral target fields of developing inner ear ganglia. Proc. Natl. Acad. Sci. USA 89, 9915-9919. Prado-Guitierrez, P., Fewster, L.M., Heasman, J.M., McKay, C.M., Shepherd, R.K., 2006. Effect of interphase gap and pulse duration on electrically evoked potentials is correlated with auditory nerve survival. Hear. Res. 215, 47-55. Pruunsild, P., Kazantseva, A., Aid, T., Palm, K., Timmusk, T., 2007. Dissecting the human BDNF locus: bidirectional transcription, complex splicing, and multiple promoters. Genomics 90, 397-406. Puel, J.L., Ruel, J., Gervais d’Aldin, C., Pujol, R., 1998. Excitotoxicity and repair of cochlear synapses after noise-trauma induced hearing loss. Neuroreport 9, 2109-2114. Qun, L.X., Pirvola, U., Saarma, M., Ylikoski, J., 1999. Neurotrophic factors in the auditory periphery. Ann. NY Acad. Sci. 884, 292-304. Rattay, F., Lutter, P., Felix, H., 2001. A model of the electrically excited human cochlear neuron. I. Contribution of neural substructures to the generation and propagation of spikes. Hear. Res. 153, 43-63. Rejali, D., Lee, V.A., Abrashkin, K.A., Humayun, N., Swiderski, D.L., Raphael, Y., 2007. Cochlear implants and ex vivo BDNF gene therapy protect spiral ganglion neurons. Hear. Res. 228, 180-187. Reichardt, L.F., 2006. Neurotrophin-regulated signalling pathways. Phil. Trans. R. Soc. 361, 15451564. Richardson, R.T., O’Leary, S., Wise, A., Hardman, J., Clark, G., 2005. A single dose of neurotrophin-3 to the cochlea surrounds spiral ganglion neurons and provides trophic support. Hear. Res. 204, 37–47. Roehm, P.C., Hansen, M.R., 2005. Strategies to preserve or regenerate spiral ganglion neurons. Curr. Opin. Otolaryngol. Head Neck Surg. 13, 294-300. 60

Neurotrophins and their role in the cochlea

Schecterson, L.C., Bothwell. M., 1994. Neurotrophin and neurotrophin receptor mRNA expression in developing inner ear. Hear. Res. 73, 92-100. Scheper, V., Paasche, G., Miller, J.M., Warnecke, A., Berkingali, N., Lenarz, T., Stöver, T., 2009. Effects of delayed treatment with combined GDNF and continuous electrical stimulation on spiral ganglion cell survival in deafened guinea pigs. J. Neurosci. Res. 87, 1389-1399. Schindler, R.A., Gladstone, H.B., Scott, N., Hradek, G.T., Williams, H., Shah, S.B., 1995. Enhanced preservation of the auditory nerve following cochlear perfusion with nerve growth factors. Am. J. Otol. 16, 304-309. Schinstine, M., Kawaja, M.D., Gage, F.H., 1991. Intracerebral delivery of growth factors: potential application of genetically modified fibroblasts. Prog. Growth Factor Res. 3, 57-66. Sekimoto, M., Fukamachi, K., Nemoto, F., Miyata, S., Degawa, M., Senba, E., Ueyama, T., Nemoto, K., 1998. Novel alternative splicing in the 5’ exon of the neurotrophin-3 gene. Neuroreport 9, 3675-3679. Shah, S.B., Gladstone, H.B., Williams, H., Hradek, G.T., Schindler, R.A., 1995. An extended study: protective effects of nerve growth factor in neomycin-induced auditory neural degeneration. Am. J. Otol. 16, 310-314. Shepherd, R.K., Coco, A., Epp, S.B., Crook, J.M., 2005. Chronic depolarization enhances the trophic effects of brain-derived neurotrophic factor in rescuing auditory neurons following a sensorineural hearing loss. J. Comp. Neurol. 486, 145-158. Shepherd, R.K., Coco, A., Epp, S.B., 2008. Neurotrophins and electrical stimulation for protection and repair of spiral ganglion neurons following sensorineural hearing loss. Hear. Res. 242, 100-109. Shibata, S.B., Cortez, S.R., Beyer, L.A., Wiler, J.A., Di Polo, A., Pfingst, B.E., Raphael, Y., 2010. Transgenic BDNF induces nerve fiber regrowth into the auditory epithelium in deaf cochleae. Exp. Neurol. 223, 464-472. Shibata, S.B., Budenz, C.L., Bowling, S.A., Pfingst, B.E., Raphael, Y., 2011. Nerve maintenance and regeneration in the damaged cochlea. Hear. Res. 281, 56-64. Shinohara, T., Bredberg, G., Ulfendahl, M., Pyykkö, I., Olivius, N.P., Kaksonen, R., Lindström, B., Altschuler, R., Miller, J.M., 2002. Neurotrophic factor intervention restores auditory function in deafened animals. Proc. Natl. Acad. Sci. USA 99, 1657-1660. Staecker, H., Jolly, C., Garnham, C., 2010. Cochlear implantation: an opportunity for drug development. Drug Discov. Today 15, 314-321. Staecker, H., Kopke, R., Malgrange, B., Lefebvre, P., Van de Water, T.R., 1996. NT-3 and/or BDNF therapy prevents loss of auditory neurons following loss of hair cells. Neuroreport. 7, 889894. Stenqvist, A., Agerman, K., Marmigère, F., Minichiello, L., Ernfors, P., 2005. Genetic evidence for selective neurotrophin 3 signalling through TrkC but not TrkB in vivo. EMBO Rep. 6, 973-978. Stöver, T., Gong, T.L., Cho, Y., Altschuler, R.A., Lomax, M.I., 2000. Expression of the GDNF family members and their receptors in the mature rat cochlea. Brain Res. Mol. Brain Res. 76, 25-35. Tauszig-Delamasure, S., Bouzas-Rodriguez, J., 2011. Targeting neurotrophin-3 and its dependence receptor tyrosine kinase receptor C: a new antitumoral strategy. Expert Opin. Ther. Targets 15, 847-858. Takahashi, M., 2001. The GDNF/RET signaling pathway and human diseases. Cytokine Growth Factor Rev. 12, 361-373. Teng, K.K., Felice, S., Kim, T., Hempstead, B.L., 2010. Understanding proneurotrophin actions: Recent advances and challenges. Dev. Neurobiol. 70, 350-359. Tessarollo, L., Coppola, V., Fritzsch, B., 2004. NT-3 replacement with brain-derived neurotrophic factor redirects vestibular nerve fibers to the cochlea. J. Neurosci. 24, 2575-2584. 61

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 2

Thiele, C.J., Li, Z., McKee, A.E., 2009. On Trk – the TrkB signal transduction pathway is an increasingly important target in cancer biology. Clin. Cancer Res. 15, 5962-5967. Thoenen, H., 1995. Neurotrophins and neuronal plasticity. Science 270, 593-598. Tuszynski, M.H., U, H.S., Amaral, D.G., Gage, F.H., 1990. Nerve Growth Factor Infusion in the Primate Brain Reduces Lesion-Induced Cholinergic Neuronal Degeneration. J. Neurosci. 10, 3604-3614. Van Ruijven, M.W.M., de Groot, J.C.M.J., Smoorenburg, G.F., 2004. Time sequence of degeneration pattern in the guinea pig cochlea during cisplatin administration. A quantitative histological study. Hear. Res. 197, 44-54. Van Westerlaak, M.G.H., Bär, P.R., Cools, A.R., Joosten, E.A.J., 2001. Malonate-induced corticomotoneuron death is attenuated by NT-4, but not by BDNF or NT-3. Neuroreport 12, 13551358. Versnel, H., Agterberg, M.J.H., de Groot, J.C.M.J., Smoorenburg, G.F., Klis, S.F.L., 2007. Time course of cochlear electrophysiology and morphology after combined administration of kanamycin and furosemide. Hear. Res. 231, 1-12. Webster, M., Webster, D.B., 1981. Spiral ganglion neuron loss following organ of Corti loss: a quantitative study. Brain Res. 212, 17–30. Wheeler, E.F., Bothwell, M., Schecterson, L.C., Von Bartheld, C.S., 1994. Expression of BDNF and NT-3 mRNA in hair cells of the organ of Corti: Quantitative analysis in developing rats. Hear. Res. 73, 46-56. Wise, A.K., Fallon, J.B., Neil, A.J., Pettingill, L.N., Geaney, M.S., Skinner, S.J., Shepherd, R.K., 2011. Combining cell-based therapies and neural prostheses to promote neural survival. Neurotherapeutics 8, 774-787. Wise, A.K., Richardson, R., Hardman, J., Clark, G., O’leary, S., 2005. Resprouting and survival of guinea pig cochlear neurons in response to the administration of the neurotrophins brainderived neurotrophic factor and neurotrophin-3. J. Comp. Neurol. 487, 147-165. Wright, T.J., Hatch, E.P., Karabagli, H., Karabagli, P., Schoenwolf, G.C., Mansour, S.L., 2003. Expression of mouse fibroblast growth factor and fibroblast growth factor receptor genes during early inner ear development. Dev. Dyn. 228, 267-272. Xu, S.A., Shepherd, R.K., Chen, Y., Clark, G.M., 1993. Profound hearing loss in the cat following the single co-administration of kanamycin and ethacrynic acid. Hear. Res. 70, 205-215. Yagi, M., Kanzaki, S., Kawamoto, K., Shin, B., Shah, P.P., Magal, E., Sheng, J., Raphael, Y., 2000. Spiral ganglion neurons are protected from degeneration by GDNF gene therapy. J. Assoc. Res. Otolaryngol. 1, 315-325. Yamagata, T., Miller, J.M., Ulfendahl, M., Olivius, N.P., Altschuler, R.A., Pyykkö, I., Bredberg, G., 2004. Delayed neurotrophic treatment preserves nerve survival and electrophysiological responsiveness in neomycin-deafened guinea pigs. J. Neurosci. Res. 78, 75-86. Ylikoski, J., Pirvola, U., Moshnyakov, M., Palgi, J., Arumäe, U., Saarma, M., 1993. Expression patterns of neurotrophin and their receptor mRNAs in the rat inner ear. Hear. Res. 65, 69-78. Ylikoski, J., Pirvola, U., Virkkala, J., Suvanto, P., Liang, X.Q., Magal, E., Altschuler, R., Miller, J.M., Saarma, M., 1998. Guinea pig auditory neurons are protected by glial cell line-derived growth factor from degeneration after noise trauma. Hear. Res. 124, 17-26. Ylikoski, J., Wersall, J., Bjorkroth, B., 1974. Degeneration of neural elements in the cochlea of the guinea pig after damage to the organ of corti by ototoxic antibiotics. Acta Otolaryngol. Suppl. 326, 23–41. Zha, X.M., Bishop, J.F., Hansen, M.R., Victoria, L., Abbas, P.J., Mouradian, M.M., Green, S.H., 2001. BDNF synthesis in spiral ganglion neurons is constitutive and CREB-dependent. Hear. Res. 156, 53-68. 62

CHAPTER 3 Spiral ganglion cell morphology in guinea pigs after deafening and neurotrophic treatment

Maarten C. van Loon, Dyan Ramekers, Martijn J.H. Agterberg, John C.M.J. de Groot, Wilko Grolman, Sjaak F.L. Klis, Huib Versnel Hearing Research 298 (2013) 17-26.

There are seven species of great ape on the planet. Six of them live in nature. One cannot survive without artificial aid. – Timothy Taylor, “The artificial ape”

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 3

Abstract

It is well known that spiral ganglion cells (SGCs) degenerate in hair-cell-depleted cochleas and that treatment with exogenous neurotrophins can prevent this degeneration. Several studies reported that, in addition, SGC size decreases after deafening and increases after neurotrophic treatment. The dynamics of these cell size changes are not well known. In a first experiment we measured size, shape (circularity) and intracellular density of SGCs in guinea pigs at various moments after deafening (1, 2, 4, 6, and 8 weeks) and at various cochlear locations. In a second experiment, the effect of treatment with brain-derived neurotrophic factor (BDNF) on SGC morphology was investigated at various cochlear locations in deafened guinea pigs. We found that SGC size gradually decreased after deafening in the basal and middle cochlear turns. Already after one week a decrease in size was observed, which was well before the number of SGCs started to decrease. After BDNF treatment SGCs became noticeably larger than normal throughout the cochlea, including the middle and apical turns, whereas an effect on survival of SGCs was primarily observed in the basal turn. Thus, both after deafening and after neurotrophic treatment a change in size occurs before survival is affected. Morphological changes were not restricted to a subpopulation of SGCs. We argue that although changes in cell size and changes in survival might be manifestations of two separate mechanisms, morphological measures such as size, circularity and intracellular density are indicative for survival and degeneration.

Spiral ganglion cell morphology

1. Introduction Degeneration and protection of the auditory nerve have been studied extensively over the past decades. An important impetus for these studies is the cochlear implant, which relies on a viable nerve in order to be effective in providing electrical hearing in patients with profound sensorineural hearing loss. It is well known that spiral ganglion cells (SGCs) degenerate after loss of cochlear hair cells (Ylikoski et al., 1974; Spoendlin, 1975; Webster and Webster, 1981; Koitchev et al., 1982; Leake and Hradek, 1988; McFadden et al., 2004; Shepherd et al., 2004; Versnel et al., 2007). There is a gradual loss of SGCs, and the cells that survive are smaller and less ovoid-shaped than SGCs in normal cochleas (Staecker et al., 1996; Leake et al., 1999; Richardson et al., 2005; Shepherd et al., 2005; Glueckert et al., 2008; Agterberg et al., 2008). This degeneration of SGCs is thought to be a consequence of loss of neurotrophic support by the hair cells, as the absence of endogenous neurotrophins, such as brain-derived neurotrophic factor (BDNF) or neurotrophin-3 (NT-3), leads to a loss of SGCs (Ernfors et al., 1995; Fritzsch et al., 1999; Schimmang et al., 2003). Administration of exogenous neurotrophins can prevent SGC loss (Ernfors et al., 1996; Staecker et al., 1996; Miller et al., 1997; Shinohara et al., 2002; Gillespie et al., 2004; Shepherd et al., 2005; Glueckert et al., 2008; Agterberg et al., 2008; Song et al., 2009; Leake et al., 2011). An additional effect of such treatment is that cells become larger than normal (McGuinness and Shepherd, 2005; Shepherd et al., 2005; Glueckert et al., 2008; Agterberg et al., 2008; Leake et al., 2011). A significant reduction of SGC perikaryal area (by more than 20%) was observed as early as two weeks after deafening, while the number of SGCs remained near-normal (Agterberg et al., 2008, their Fig. 6). Neurotrophic treatment of deafened animals might lead to larger surviving SGCs without enhancement of survival (Richardson et al., 2005). Considering these observations one might say that a change in cell size occurs prior to cell death or cell rescue. In this paper, we address this notion in two experiments in deafened guinea pigs. First, the morphological changes of SGCs after deafening were examined as a function of both time after deafening and cochlear location. Specifically, it was examined whether SGC size or shape changed after one week, when the number of cells is not yet decreased (Versnel et al., 2007). Second, the morphological changes of the SGCs after deafening and subsequent neurotrophic treatment were examined as a function of cochlear location. Dependence on cochlear location is expected for two reasons. First, gradients along the cochlea are found with respect to, among others, expression levels of BDNF and NT-3, and with respect to discharge properties of SGCs (Adamson et al., 2002; Schimmang et al., 2003; Davis and Liu, 2011). Second, delivery of drugs through the round window membrane or through a cochleostomy in the basal turn will allow the drugs to reach the apical turn, albeit in lower concentrations than 65

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 3

in the basal turn (Salt and Plontke, 2009; Hahn et al., 2012). Also, the effect of BDNF treatment is often more pronounced in the basal turn (Shepherd et al., 2005; Glueckert et al., 2008; Agterberg et al., 2008); therefore, we examined here whether the effect of BDNF on SGC morphology depends on location in a similar fashion.

2. Methods

2.1. Animals Thirty-seven albino female guinea pigs (strain: Dunkin Hartley; weighing 250-600 g) were obtained from Harlan Laboratories (Horst, the Netherlands) and housed in the animal care facility of the Rudolf Magnus Institute of Neuroscience. All animals had free access to both food and water and were kept under standard laboratory conditions. Lights were on between 7:00 am and 7:00 pm. Temperature and humidity were kept constant at 21 ºC and 60%, respectively. All experimental procedures were approved by the University’s Committee on Animal Research (DEC-UMC #03.04.036).

2.2. Experimental design To describe the morphological changes of SGCs after deafening and the effect of neurotrophic treatment we performed histological analyses of cochleas used in two previous studies of our laboratory (Versnel et al., 2007; Agterberg et al., 2008). Note that in the Versnel et al. study, which described effects of deafening, SGC morphology was not addressed, and in the Agterberg et al. study, which described effects of BDNF treatment, SGC morphology was addressed using electron microscopy but only for one particular cochlear location (upper basal turn). The treatment schedules of the two experiments are shown in Fig. 1. In the first experiment (Fig. 1A), we quantified the size and shape of SGCs after deafening. Animals were sacrificed at different survival times following the deafening procedure: 1, 2, 4, 6, and 8 weeks. A control group consisting of five guinea pigs was sacrificed at eight weeks after a sham treatment. Except for the 6-week group, all animals were equipped with an electrode consisting of a stainless steel wire with a gold-ball tip positioned in the round window niche of the right cochlea in order to record compound action potentials (CAPs) to tone pips and clicks (for details, see Versnel et al., 2007), which enabled us to measure hearing thresholds. Guinea pigs showing insufficient threshold shifts for click responses (< 50 dB), one or two weeks after the deafening procedure, were excluded (N = 3). After exclusion, 23 experimental animals remained (for details, see Fig. 1A). In the second experiment (Fig. 1B) we studied the effect of neurotrophic treatment on the size and shape of SGCs. The right cochleas of six animals were implanted with 66

Spiral ganglion cell morphology

a cannula two weeks after deafening and were treated with BDNF for four consecutive weeks. The left cochleas remained untreated and were used in the analyses of the first experiment as well. Stainless steel screws were inserted into the skull to record auditory brainstem responses (ABRs) to click stimuli (for details, see Agterberg et al., 2008). Since this study addresses cell morphology given a positive effect of BDNF, one animal was excluded because it did not respond to BDNF treatment (both SGC packing density and SGC size in the treated cochlea were similar to those in the untreated ear; see Fig. 6 in Agterberg et al., 2008). The five remaining animals demonstrated large threshold shifts (> 50 dB). They were sacrificed for histology immediately after cessation of BDNF treatment, which was six weeks after deafening (Fig. 1B).

N=5 Df H

3 H

4 H

6 H

Sham

pump

5 H

(Ntotal = 23)

(N = 5) Df: deafening H : histology BDNF

Right

H (N = 5) Left 0

Time relative to deafening (weeks)

Fig. 1. Treatment schedules of the experiments. A Animals were bilaterally deafened and at various times after deafening the animals were sacrificed for histology: 1, 2, 4, 6, and 8 weeks. As control, five animals were sham-treated and examined after eight weeks. In all but the 6-weeks animals, both the right and left cochlea were examined; in the 6-weeks animals only the left cochlea was examined. B Five animals were implanted with a mini-osmotic pump and cannula two weeks after deafening. The cannula was inserted in the right cochlea, and BDNF was delivered during four weeks. Immediately thereafter, six weeks after deafening, the animals were sacrificed for histology. Note that the left cochleas were used in the deafening experiment as well.

2.3. Deafening procedure Except for a control group in experiment 1, all animals were deafened. The animals were anaesthetized with ketamine (Ketanest®, 40 mg/kg, i.m.) and xylazine (Sedamun®, 10 mg/kg, i.m.). Kanamycin sulphate in isotonic saline (400 mg/kg) was administered subcutaneously, and the external jugular vein was exposed and cannulated. Furosemide Figure was 1 slowly infused, 15-60 minutes after (100 mg/kg, Centrafarm, the Netherlands) kanamycin injection. The dose of kanamycin is similar as used in numerous other studies applying kanamycin in combination with ethacrynic acid or furosemide for 67

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 3

deafening of guinea pigs (e.g., Staecker et al., 1996; Miller et al., 1997; Nourski et al., 2004; Gillespie et al., 2004; Shepherd et al., 2005; Wise et al., 2005). This procedure, introduced by West et al. (1973), has been shown to eliminate almost all cochlear hair cells when furosemide is infused within 4 hours after kanamycin injection (Brummett et al., 1979). The control animals in experiment 1 received isotonic saline (subcutaneously and intravenously), instead of kanamycin and furosemide. 2.4. Cannula implantation and BDNF treatment The animals in experiment 2 (Fig. 1B) were treated with BDNF as follows. Two weeks after the deafening procedure the animals were implanted with a cannula in the right cochlea. The animals were anaesthetized with ketamine and xylazine, and the right bulla was exposed retro-auricularly. A small hole was drilled to visualize the cochlea. The round window membrane was perforated and the tip of the cannula was inserted into the scala tympani (Prieskorn and Miller, 2000). The cannula, filled with BDNF solution, was fixed onto the bulla with dental cement (Ketac-Cem Aplicap; ESPE Dental Supplies, Utrecht, the Netherlands). Mini-osmotic pumps were attached to the cannula and inserted into a subcutaneous pocket. The cannula was connected to the skull with a screw (Brown et al., 1993) and fixed with dental cement, and the wound was closed with two layers of Vicryl®. Mini-osmotic pumps were filled with BDNF (PeproTech®) solution (100 µg/ml), and then incubated in sterile saline for 48 h at 37 °C in order to guarantee a constant flow rate at the time of implantation. The BDNF concentration of 100 µg/ml was chosen, because it is in the range (50 to 100 µg/ml) that has been proven to be effective (Gillespie et al., 2004; Shepherd et al., 2005; Wise et al., 2005; Miller et al., 2007). The cochleas were treated with BDNF for four consecutive weeks. 2.5. Histology After the animals were sacrificed for histological preparation, the left and right cochlea were fixed by intralabyrinthine perfusion with a tri-aldehyde fixative and processed as described by De Groot et al. (1987). Cochleas were embedded in toto in Spurr’s lowviscosity resin, divided into two halves along a standardized midmodiolar plane, using the attachment point of the cochlea on the medial wall of the bulla and both the oval and round windows as anatomical landmarks, and were then re-embedded in fresh resin. From each cochlea, five semithin (1 µm) sections were cut at 30-µm intervals, which is twice the average diameter of SGCs (and greater than large BDNF-treated SGCs, Agterberg et al., 2008; Glueckert et al., 2008), to ensure that consecutive sections would not contain the same cell. The sections were then stained with methylene blue and azur II in sodium tetraborate and examined with a Leica DMRA light microscope. All sections were used to determine SGC packing densities; the first section, which was 68

Spiral ganglion cell morphology

closest to the midmodiolar plane of sectioning, was used for determination of perikaryal area, cell circularity and intracellular density. Digitized light-microscopical images of the spiral ganglia were merged into photomontages and analyzed using the imageprocessing program ImageJ (version 1.39t; http://rsbweb.nih.gov/ij). We ensured that the overlap between adjacent images was at least 25% on either side, resulting in a final photomontage without any obvious marginal distortion. To determine SGC packing densities, all profiles of both type-I and type-II perikarya in transections of Rosenthal’s canal were counted (see Slepecky, 1996, for general characteristics of type-I and typeII cells). For quantitative analysis of the morphological features (perikaryal area, cell circularity, and intracellular density), only type-I (myelinated) SGCs with an obvious nucleus were included. The cochleas were analyzed at four or five locations: lower and upper basal locations B1 and B2, lower and upper middle locations M1 and M2, and (in experiment 2 only) the lower apical location A1 (Fig. 2). Location

Distance from Characteristic round window (%) frequency (kHz)

1.6

0.8

0.4

0.2

Fig. 2. Midmodiolar section (1 µm) of a guinea pig cochlea from the normal-hearing control group showing the different cochlear locations at which the SGCs were examined. The table gives the distance from the round window (in %) of locations B1 through A3 and corresponding characteristic frequencies (computed according to Tsuji and Liberman, 1997). N. VIII, cochlear nerve; B1,2, lower and upper basal locations; M1,2, lower and upper middle locations; A1-3, lower, middle, and upper apical locations; H, helicotrema.

2.5.1. Determination of SGC packing density To determine SGC packing densities, the bony boundaries of Rosenthal’s canal were Figure 2 outlined using a pressure-sensitive stylus on a Wacom Bamboo digitizer tablet. The area of Rosenthal’s canal was calculated by converting the number of pixels to mm2. Type-I and type-II SGC perikarya, with or without a nucleus, were counted in all five cochlear sections. SGC packing densities were calculated by dividing the number of 69

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 3

counted perikarya by the cross-sectional area of Rosenthal’s canal. For each cochlear location the packing density was then averaged over all (ideally five) sections. In some cases reliable counts could only be made in less than five sections. When SGCs have different sizes, an error is introduced in calculating the SGC packing densities (Coggeshall and Lekan, 1996; Leake et al., 1999). Therefore, the packing density was corrected for size using the equation: bcor = b·√(An/A) where bcor is the corrected packing density, b is the actually counted packing density, An is the average perikaryal area of the SGCs in the normal group, and A is the average perikaryal area of the SGCs at the corresponding cochlear location. This equation assumes that slice thickness (1 µm) is much smaller than the average cell diameter (~15 µm) and it assumes a spherical cell shape.

2.5.2. Determination of perikaryal area Perikaryal area (in µm2) was measured by manually delineating the innermost layer of the myelin sheath of the type-I SGCs. In case of cell shrinkage, the plasma membrane surrounding the perikaryon was outlined, and the area within was measured.

2.5.3. Determination of cell circularity Cell circularity was measured directly in ImageJ after delineating the perimeter of the cell. It is calculated according to 4πA/L2, where A is the area and L is the perimeter. Circularity is dimensionless, varying from 0 to 1, and, therefore, it is independent of size. Circularity is 1 for a perfect circle and less than 1 for other shapes (e.g., 0.78 for a square).

2.5.4. Determination of intracellular density Intracellular density of the perikaryon of the SGC, including both the nucleus and the cytoplasm was measured directly in ImageJ. A gray level between 0 (white) and 255 (black) was assigned to each image pixel, and then the mean gray value for each individual SGC was calculated. The same program settings for image contrast, brightness and resolution were used for all images analyzed. Because of differences in exposure and staining between each section, the data were normalized. The mean gray value of four sampling fields in the modiolar bone was measured in each photomontage and the gray value of the SGCs was then divided by the average gray value of the bone. 2.6. Statistical analyses In case of the morphological properties perikaryal area, circularity and intracellular density, the data of the individual type-I SGCs were first averaged per location. For analysis of the effect of deafening (first experiment), the data of the left and right cochlea

Spiral ganglion cell morphology

within one animal were averaged (except for the 6-week data, which were obtained from the left cochlea only). Subsequently, repeated-measures analysis of variance (rm ANOVA) was performed with cochlear location as within factor and survival time as between factor. For analysis of the effect of BDNF treatment (second experiment), where a linear relationship with cochlear location was expected (a decrease of treatment effect from base to apex), a linear mixed model analysis was applied under the assumption of compound symmetry, with cochlear location as covariate and treatment (right vs. left cochlea) as factor. In this analysis location was expressed in relative distance from location B1 (see Fig. 2). To test correlations between the four SGC variables Pearson’s correlation coefficients were computed. All statistical analyses were performed with SPSS for Windows (version 20.0).

3. Results 3.1 Hair cell loss and threshold shifts after deafening Midmodiolar counts in the deafening experiment (Fig. 1A) showed a complete OHC loss in the basal regions of the cochlea (locations B1, B2, and M1) in each animal. In 16 out of 22 animals, no remaining IHCs could be observed. In 6 animals, remaining IHCs were found, both basally and apically. The remaining IHCs were evenly distributed across the groups (1 week deaf: N = 2 animals; 2 weeks deaf: N = 1; 4 weeks deaf: N = 2; 6 weeks deaf: N = 1; 8 weeks deaf: N = 0). The average threshold shift in the deafening experiment was 75 dB (range: 59-90 dB), as assessed by CAPs to clicks. In the BDNF experiment (Fig. 1B) only 1 out of 5 animals showed remaining hair cells, and the treated ear did not show more remaining hair cells than the untreated ear. The average threshold shift was 82 dB (range: 55-95 dB), as assessed by ABRs to clicks.

3.2. Examples of morphological features of SGCs The effect of deafening on SGCs with and without subsequent BDNF treatment is shown in Fig. 3 at two different cochlear locations, the lower basal (B1) and the upper middle turn (M2). In a normal cochlea (Fig. 3A, B), Rosenthal’s canal was packed with SGCs embedded in a matrix of vascularized connective tissue. SGC packing density, and cell size and shape appeared very similar at both cochlear locations. One week after deafening, there were no apparent SGC changes in either turn (Fig. 3C, D). Six weeks after deafening, however, the number of SGCs had declined substantially, particularly in the basal turn, and surviving SGCs were smaller after deafening (Fig. 3E, F). The cells had lost their typical round to ovoid shape and demonstrated a shrunken appearance with only a small quantity of cytoplasm. 71

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 3

Fig. 3. Light micrographs of Rosenthalâ&#x20AC;&#x2122;s canal at cochlear locations B1 (A, C, E, G) and M2 (B, D, F, H) in different conditions. A, B spiral ganglia from a normal-hearing control animal; C, D spiral ganglia one week after deafening; E, F spiral ganglia of the left untreated cochlea from an animal in which the right cochlea was treated with BDNF, six weeks after deafening; G, H spiral ganglia of the right BDNF-treated cochlea of the same animal as represented in E and F. 72

Spiral ganglion cell morphology

Treatment with BDNF prevented loss of SGCs in the basal turn; after six weeks the distribution of SGCs in Rosenthal’s canal appeared near-normal (Fig. 3G). The surviving SGCs had retained their ovoid shape and seemed to be larger than those in the normal group. The middle turn of the same cochlea demonstrated an evident loss of SGCs (Fig. 3H) and when compared to the contralateral untreated cochlea (Fig. 3F) there appeared to be little protective effect of BDNF. The remaining SGCs, however, were larger and had a rounder appearance than their counterparts in the contralateral untreated cochlea. 3.3. Morphological features of SGCs after deafening

3.3.1. Packing density and size Figure 4 shows the time courses of SGC packing density and SGC size after deafening at four different cochlear locations. In accordance with previous studies, there was a gradual decrease with time in packing density showing a loss of 50-60% after 6 to 8 weeks. Cell size decreased noticeably faster; the average perikaryal area was already reduced after one week, eventually leading to a decrease of 10-25% of its original size after 8 weeks. The reductions with time are statistically significant as demonstrated by rm ANOVA [within factor: cochlear location; between factor: survival time; packing density: F(5,22) = 21.1, P < 0.001; cell size: F(5,21) = 5.5, P < 0.01]. Both variables have been plotted relative to normal in order to compare the time course of cell size with that of packing density (Fig. 4A-D). The main point emerging from this comparison is that the decrease in SGC size occurred prior to a decrease in number of cells. While the packing densities in animals examined one week after deafening were not different from those in normal-hearing controls, the cell size was reduced by 10% at locations B2, M1 and M2. The loss of SGCs first occurred in the second week after deafening and strongly progressed in the following weeks. The time course of the SGC packing density was similar across the different cochlear locations, which was confirmed by a lack of interaction between location and survival time [rm ANOVA: F(15,66) = 0.9, P > 0.5]. A significant interaction was found for SGC size [rm ANOVA: F(5,21) = 5.5, P < 0.01), caused by a deviant time course at location M2. A secondary finding in this analysis over all animals was a significant difference in packing density and size across cochlear locations [F(3,63) > 7, P < 0.001]. The packing density was lower at B1 than at locations B2 through M2 (Fig. 4E) and SGCs were larger at B1 than at B2 through M2 (Fig. 4F).

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

120%

location B1

100%

80%

60%

40%

40% packing density

20%

perikaryal area

0% 0

120% 100%

location B2

120%

100% packing density and perikaryal area relative to normal

location M1

120%

80%

60%

40%

20%

location M2

100%

0% 0 2 4 6 8 Time after deafening (weeks) 2000 1500 1000 500 0

B1 B2 M1 M2

Cochlear location

Perikaryal area (µm2)

Packing density (cells/mm2)

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 3

2 4 6 8 Time after deafening (weeks)

260 240 220 200 180

B1 B2 M1 M2

Cochlear location

Fig. 4. Effect of deafening on SGC packing density and perikaryal area at four different cochlear locations examined (B1, B2, M1, and M2). Data shown represent means with standard error of mean. Data are averaged across the right and left cochlea before averaging across animals. The packing density is corrected for SGC size. A, B, C, D Packing density and perikaryal area relative to normal as a function of time after deafening for SGCs in locations B1 (A), B2 (B), M1 (C), and M2 (D). E Packing density in normal-hearing controls. F Perikaryal area in normal-hearing controls.

3.3.2. Circularity and intracellular density Figure 4 The circularity of the SGCs did not vary significantly with time (Fig. 5A; rm ANOVA or linear regression, P > 0.3). There was a significant trend with turn [rm ANOVA, F(3,63) = 6.9, P < 0.001]: cells at location B1 were more circular than at B2 through M2, which is illustrated in Fig. 5A for locations B1 and M2. The intracellular density of the SGCs gradually increased with time after deafening (Fig. 5B). For each cochlear location (B1 through M2) this cellular change was significant [linear regression: R2 > 0.16, P < 0.05]. 74

Spiral ganglion cell morphology

1.6

0.86

Intracellular density

Perikaryal circularity

0.88

1.4

0.84 0.82

1.2

0.8 0.78

1.0

0.76 0.74

0.8 0

2 4 6 8 Time after deafening (weeks)

2 4 6 Time after deafening (weeks)

Fig. 5. Effect of deafening on SGC perikaryal circularity and intracellular density at two cochlear locations (lower basal turn, B1; upper middle turn, M2). Data shown represent means with standard error of mean. Data are averaged across the right and left cochlea before averaging across animals. A Circularity as a function of time after deafening. B Intracellular density as a function of time after deafening.

3.4. SGC properties after BDNF treatment

3.4.1. Packing density and size Figurewith 5 the contralateral untreated cochleas Figure 6 compares the BDNF-treated cochleas with respect to packing density and SGC size. The data are presented as percentage relative to normal. The packing densities in the treated cochleas were significantly higher than those in the untreated cochleas [linear mixed model: F(1,44) = 24.0, P < 0.001]. Loss of SGCs after deafening was prevented by BDNF treatment, particularly in the basal turn where the packing density was similar to normal, and about twice as large as in the untreated ear. The protective effect of BDNF disappeared towards the apex with packing densities being similar in treated and untreated cochleas at cochlear locations M2 and A1. The interaction between location and treatment was significant [linear mixed model: F(1,42.6) = 8.0, P < 0.01]. At the middle and apical locations (M1, M2, A1), the packing densities of the BDNF-treated cochleas were also significantly smaller than those in the normal-hearing control group [independent samples t test: P < 0.05]. Figure 6B demonstrates that SGCs were significantly larger in BDNF-treated cochleas than in untreated cochleas [linear mixed model: F(1,43) = 113, PÂ <Â 0.001]. The data of SGC size showed a notably different trend than the packing density data: at each location, including the more apical region (M2, A1), BDNF had a substantial effect. For locations B1, B2, M1 and A1 the differences were significant [paired t test: P < 0.01]. Also, the BDNF-treated SGCs were larger than in normal controls at each location, which was significant in B1 and B2 [independent samples t test: P < 0.01]. The size data were comparable to the packing density data, in that the differences decreased from base to 75

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Packing density relative to normal

apex. The interaction between location and treatment was confirmed by linear mixed model analysis [F(1,43) = 15.1, P < 0.001].

untreated (left)

150%

BDNF-treated (right)

100%

50% B1 0% 0%

B2 20%

M1 40%

60%

80%

100%

Distance from round window

Perikaryal area relative to normal

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 3

150%

100%

50%

B1 0% 0%

B2 20%

M1 40%

M2 60%

A1 80%

100%

Distance from round window

Fig. 6. Effect of BDNF treatment on SGC packing density and perikaryal area as a function of distance from the round window (in %). The five different cochlear locations examined (B1, B2, M1, M2, and A1) are indicated at the corresponding distance. Data shown represent means with standard error of mean. The packing density is corrected for SGC size. A Packing density relative to normal. B Perikaryal area relative to normal.

3.4.2. Circularity and intracellular density The circularity of the SGCs followed the trend of the size: at each location the BDNFtreated SGCs were more circular than the untreated Figure 6SGCs in the contralateral cochleas (Fig. 7A). The differences, while less pronounced, were statistically significant [linear mixed model: F(1,41) = 9.7, P < 0.01]. Figure 7B shows that the intracellular density was higher in untreated ears than in BDNF-treated ears at the more basal locations (B1, B2, M1) but not at the more apical 76

Spiral ganglion cell morphology

locations (M2, A1). The treatment effect and the interaction between location and treatment were statistically significant [linear mixed model – treatment: F(1,43) = 12.3, P < 0.01; interaction: F(1,43) = 7.2, P < 0.05]. Circularity relative to normal

140%

untreated (left)

BDNF-treated (right) 120%

100%

80% B1 60% 0%

B2 20%

M1 40%

60%

80%

100%

Intracellular density relative to normal

Distance from round window 140%

120%

100%

80% B1 60% 0%

B2 20%

M1 40%

M2 60%

A1 80%

100%

Distance from round window

Fig. 7. Effect of BDNF treatment on SGC circularity and intracellular density as a function of distance from the round window (in %). The five different cochlear locations examined (B1, B2, M1, M2, and A1) are indicated at the corresponding distance. Data shown represent means with standard error of mean. A Circularity relative to normal. B Intracellular density relative to normal.

3.5. Distributions of size, circularity and intracellular density in the SGC population So far, the morphological features of the SGCs were described by values averaged across an individual transection of Rosenthal’s canal. Figure 7 It is possible that a change in an averaged feature, for instance a decrease in size, is caused by a subpopulation of SGCs. This subpopulation may shrink and degenerate, while surviving cells retain their normal size. A bimodal distribution, then, would indicate for an individual cell 77

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

that a decrease in size causes cell death. Alternatively, a homogeneous population of shrinking cells would imply that shrinkage does not cause cell death, but instead merely indicates an increased probability of cell death. To address this issue, we examined the distributions of the morphological features of the individual cells at each cochlear location. Figure 8 shows the distribution of SGC size at one location (B2) for various conditions. The distributions were unimodal, demonstrating that the average change in cell size (decrease over time after deafening, or increase after BDNF treatment) was a result of a change in size for the entire SGC population. The unimodality of the distribution was confirmed for each condition in each cochlear turn with a KolmogorovSmirnov test for normal distribution [P > 0.1]. It should be noted that 6 to 8 weeks after deafening almost all cells were smaller than the average size in normal-hearing animals (Fig. 8, dashed line). Circularity and intracellular density for individual cells were unimodally distributed for all conditions and cochlear locations as well (data not shown) [Kolmogorov-Smirnov test for normal distribution: P > 0.1]. 20% 10% 0%

BDNF 8 weeks 6 weeks

percentage

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 3

4 weeks 2 weeks 1 week normal hearing

0 80 160 240 320 400 SGC perikaryal area (µm²) Fig. 8. Distribution of SGCs according to their perikaryal area in 20 µm2 bins at cochlear location B2. Cells are pooled across animals and the number of cells per bin is plotted as a percentage of the total cell count. Vertical dashed line indicates median for normal hearing animals. BDNF, BDNF-treated cells; 1-8 weeks, 1-8 weeks after deafening.

Spiral ganglion cell morphology

3.6. Correlations between packing density and cell properties The four SGC variables examined might be related to one another in their changes after deafening and/or subsequent neurotrophic treatment. For instance, cell circularity seemed to follow the time course of cell size, i.e., cells became less circular as they shrank (compare Fig. 5A to Fig. 4), and vice versa, they were rounder as they were larger after neurotrophic treatment (compare Fig. 7A to Fig. 6B). Intracellular density rather seemed to have the same course as survival, i.e., the cell content started to become more dense when survival started to decline (compare Fig. 5B to Fig. 4) and only at cochlear locations where BDNF resulted in maintained survival the cells were less dense (compare Fig. 7B to Fig. 6A). Correlation analyses performed on the deafening data (N = 28) for each cochlear location confirmed these observations. Circularity was positively correlated with cell size, which was significant for locations B2, M1, and M2 [R > 0.43, P < 0.05]. Packing density was negatively correlated with intracellular density, which was significant for each location [R < -0.46, P < 0.05]. Also, size and circularity were negatively correlated to intracellular density [P < 0.05 for each location]. Multiple regression analysis, with packing density as dependent variable and the three SGC features as independent variables, demonstrated that intracellular density was the only significant feature explaining the variance for 18%-38% [P < 0.01 at B1, B2, and M2].

4. Discussion In two experiments, one on the degeneration of spiral ganglion cells after severe cochlear hair cell loss and the other on preventing such degeneration, we observed a similar pattern. In both experimental conditions a change in cell size was observed prior to a change in survival. Specifically, already one week after deafening a decrease in size was observed throughout the cochlea, which was well before the number of SGCs started to decrease (Fig. 4). After BDNF treatment, SGCs had become noticeably larger than normal throughout the cochlea, including the middle and apical turns, whereas a positive effect on survival of SGCs was mainly observed in the basal turn (Fig. 6). 4.1. Packing density and SGC size The progressive decrease we found in both SGC size and packing density after deafening (Fig. 4) is consistent with previous findings (Staecker et al., 1996; McGuinness and Shepherd, 2005; Shepherd et al., 2005; Glueckert et al., 2008; Agterberg et al., 2009; Havenith et al., 2011; Leake et al., 1999, 2011). One week after deafening, the mean SGC size in three out of four cochlear locations examined was found to be smaller as compared to normal SGCs, while SGC packing densities did not decrease until one week 79

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 3

later (Figs. 4C-F). In contrast with these findings, Glueckert et al. (2008) reported a decrease of SGC packing densities, only seven days after deafening, across all cochlear turns along with a decrease in cell size. However, this rapid initial decrease might be an overestimation as SGC packing densities had not reduced further after another 14 days in their experiments. Furthermore, the packing density presented here was corrected for soma size (according to Coggeshall and Lekan, 1996), while their data were not, thereby causing an augmented SGC packing density decrease when cell size is decreasing. The latter explains 10-50% of the decrease in packing density, depending on the cochlear turn. It should be noted that histological data from other studies earlier than two weeks after deafening are not available (McGuinness and Shepherd, 2005; Shepherd et al., 2005; Agterberg et al., 2008; Leake et al., 2011). A virtually instantaneous decrease in average cell size followed by a delayed decrease in SGC packing density suggests that a decrease in SGC size is the first step in the degeneration process. However, this idea is complicated by the shift of the entire distribution of cell size as shown in Fig. 8, which means that not just the 60% of SGCs that have eventually degenerated after 8 weeks have been subject to a decrease in size. Taken together, these observations suggest that size reduction is an indicator for degeneration without a direct causal relationship. In other words, a reduction in cell size does not automatically lead to SGC death. SGC packing densities in BDNF-treated cochleas were higher than those in untreated cochleas (Fig. 6A) and BDNF-treated SGCs were larger than untreated ones and even larger than healthy SGCs (Fig. 6B), which is comparable with previous findings in guinea pigs (Glueckert et al., 2008; Richardson et al., 2005; Shepherd et al., 2005). Furthermore, it has been demonstrated that BDNF induces cell growth in cultured melanotrope cells from Xenopus laevis, specifically via tropomyosin-related kinase B (TrkB) (Kuribara et al., 2011). Possibly, infusion of BDNF results in a larger concentration of BDNF than in the normal cochlea. Cochlear concentrations of BDNF in normal-hearing guinea pigs are in the order of 10 pg/ml (Ito et al., 2005); arguably, continuous infusion with 100 Îźg/ml BDNF at an infusion rate of 0.25 Îźl/h leads to much higher cochlear concentrations, which might account for the difference in size between treated and normal SGCs. In contrast to findings in guinea pigs, BDNF-treated SGCs in neonatally deafened cats are found not to be larger than SGCs in normal-hearing cats (Leake et al., 2011). It is unclear whether this reflects a difference in age at the time of deafening or a difference across species. Interestingly, treatment of SGCs in deafened guinea pigs with glial cell line-derived neurotrophic factor (GDNF) does not lead to an increase in cell size compared to deafened controls (Scheper et al., 2009). For both packing density and SGC size, the effect of BDNF decreased from base to apex, which might be caused by a base-to-apex concentration gradient of BDNF due to basal infusion. 80

Spiral ganglion cell morphology

This effect may further be explained by a possible gradient in BDNF expression in the healthy adult cochlea and, therefore, a gradient in BDNF dependence decreasing from base to apex, although evidence for this gradient is circumstantial (Davis, 2003; Davis and Liu, 2011; Ramekers et al., 2012). Note that, although SGC loss is smaller in apical than in basal regions (40% vs. 60%), the absence of an effect of BDNF treatment apically cannot be explained by a lack of an effect of deafening in the more apical regions. The combined observation that SGCs in the more apical regions (M2, A1) of BDNF-treated ears show an increased size, but not enhanced survival, corresponds to findings of Richardson et al. (2005). One week after a single infusion of NT-3, they found larger-than-normal cells but without a significant effect on survival. A single dose may instantly initiate cell growth while continuous support, at least for a certain period (say days or weeks), is needed for survival. Likewise, in the present study, a relatively low concentration of BDNF towards the apex of the cochlea may account for the positive effect on cell size combined with a lack of effect on SGC preservation. This supports the idea that a change in SGC size and SGC survival or loss â&#x20AC;&#x201C; although both consequences of alterations in neurotrophin availability â&#x20AC;&#x201C; might be manifestations of separate mechanisms. For example, it is known that activation of the two types of neurotrophin receptors (the high-affinity TrkB, and the low-affinity p75NTR) each induce the activation of a distinct intracellular pathway, while activation of both receptors simultaneously has a synergistic effect (e.g., Reichardt, 2006). It might then be suggested that, upon basal BDNF infusion, apical concentrations are high enough to activate the high-affinity TrkB and induce SGC growth, while higher concentrations still are necessary for the additional co-activation of p75NTR for actual SGC survival. Interestingly, Havenith et al. (2011) reported enhanced survival two weeks and four weeks after placement of BDNF-soaked gelfoam onto the round window, but they did not see a corresponding change in cell size. Agterberg et al. (2009) showed that two weeks after cessation of BDNF infusion into the cochlea SGC size is comparable to normal again, while SGC preservation is still ongoing. Both studies show that the positive effect on cell size is annihilated after two weeks, suggesting a dynamic and adaptive mechanism, while the processes underlying cell survival appear to be much more rigid. These observations support the above-mentioned suggestion that a change in size and a change in survival might be manifestations of separate mechanisms.

4.2. Circularity and intracellular density Cell circularity can easily be seen in relation to cell growth. A growing cell will stretch its plasma membrane and will thereby become more spherical, while a shrinking cell will display the opposite effect. Accordingly, we have shown that a decrease in cell size after deafening is accompanied by a decrease in cell circularity and that, conversely, the typically larger BDNF-treated SGCs are more circular. 81

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 3

Intracellular density reflects the relative amount of organelles such as the nucleus, endoplasmic reticulum and the Golgi apparatus. The gradual increase in intracellular density after deafening could reflect one of two possible processes. Either the amount of the intracellular content is actually growing over time, while the cell body volume does not increase, or the cell body shrinks while the amount of organelles remains constant. When comparing Fig. 4C-F with Fig. 5B, the decrease in cell size over time is indeed accompanied by an increase in intracellular density. The effect of BDNF treatment on intracellular density and cell size is different from the degeneration data; the larger-than-normal cell size after neurotrophic treatment is only partially reflected by an intracellular density that is lower than in the untreated ear, but not different from that in normal-hearing controls. If the change in intracellular density is a direct effect of size change, the explanation could be that the SGC size after initial growth is stabilized early during the four-week neurotrophic treatment, after which the intracellular content has time to adjust itself proportionally â&#x20AC;&#x201C; a steady state, which is not reached during the continuously ongoing process of degeneration. Intracellular density is an uncommon morphological measure and direct comparison with previous findings in literature is therefore not possible. However, the intracellular density data presented here are consistent with electron-density measures found in an electron microscopic study by Agterberg et al. (2008), in that both show an increase after deafening and a normal value after BDNF treatment. Cell survival is significantly negatively correlated with intracellular density, but whether this reflects a causal role for the density of cellular content or a result of an intermediate stage eventually leading to cell death is unclear. In conclusion, although the underlying cellular mechanisms remain unclear, both the intracellular density and cell circularity might be used, along with cell size, as indicators for the stability of the cell and, therefore, as a measure for the stability of the SGC population as a whole.

4.3 Methodological considerations The change in SGC appearance (i.e., size and shape) over time or in response to cochlear damage has been shown to introduce counting errors (Richter et al., 2011). We have therefore chosen to correct packing density for the change in size (cf., Agterberg et al., 2008) assuming cell sphericity and assuming slice thickness (1 Âľm) being much smaller than cell diameter (~15 Âľm). This method is appropriate for assessments of counts relative to normal as done in the current study (Figs. 4 and 6A). Other studies have used more complex and arguably more accurate methods, such as the optical fractionator method (Camarero et al., 2001) and a physical disector stereological method (Leake et al., 2011). For assessing absolute numbers, those methods would indeed be preferable 82

Spiral ganglion cell morphology

to ours. However, for quantifying changes of packing density over time our method is appropriate, since it is not sensitive to the effect of cell size and shape because changes in size and shape are small as compared to changes in number of cells (Figs. 4 and 5). Cell circularity is on average relatively high, which indicates that the assumption of sphericity is reasonable. In addition, cell circularity does not vary much between groups (~0.78-0.82). The changes in packing density in response to deafening and BDNF treatment, as presented in this study, are similar to those found by Leake et al. (2011).

4.4. Concluding remarks It is well-documented that upon hair cell loss the number of SGCs decreases, and that this process of degeneration ceases with BDNF treatment. In addition, both hair cell loss and successive treatment with BDNF have distinct effects on SGC morphology as well (Ylikoski et al., 1974). The morphological changes described in this study are caused by a change in the availability of neurotrophic support. After hair cell loss the lack of neurotrophic support leads to smaller SGCs that are also less circular and denser. Treatment with exogenous BDNF restores their original morphology, but enhances cell size and circularity beyond normal proportions â&#x20AC;&#x201C; probably due to the higher-than-normal physiological concentration of the neurotrophin. The functional consequences of these morphological changes might influence hearing performance of cochlear implant users. Reduced populations of small SGCs in deafened animals still yield fairly good thresholds and amplitude of electrically evoked auditory brainstem responses (Shepherd et al., 2005; Agterberg et al., 2009; Havenith et al., 2011; Leake et al., 2011), which indicates normal cell function, at least in simple stimulus conditions. Nevertheless, considering that functional properties that are relevant in more complex stimulus conditions, such as refractoriness, might be affected, the restoration of their original size and shape would arguably lead to a better functioning neural interface for cochlear implants.

Acknowledgements This study was supported by the Heinsius-Houbolt Foundation and CochlearÂŽ. We are grateful to Ferry Hendriksen for assisting with histology, and Kelly Maijoor for assisting with surgery. We thank RenĂŠ Eijkemans (Biostatistics & Research Support, Julius Center) for his advice on statistical analysis.

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 3

References

Adamson, C.L., Reid, M.A., Davis, R.L., 2002. Opposite actions of brain-derived neurotrophic factor and neurotrophin-3 on firing features and ion channel composition of murine spiral ganglion neurons. J. Neurosci. 22, 1385-1396. Agterberg, M.J.H., Versnel, H., de Groot, J.C.M.J., Smoorenburg, G.F., Albers, F.W.J., Klis, S.F.L., 2008. Morphological changes in spiral ganglion cells after intracochlear application of brainderived neurotrophic factor in deafened guinea pigs. Hear. Res. 244, 25-34. Agterberg, M.J.H., Versnel, H., Van Dijk, L.M., De Groot, J.C.M.J., Klis, S.F.L., 2009. Enhanced survival of spiral ganglion cells after cessation of treatment with brain-derived neurotrophic factor in deafened guinea pigs. J. Assoc. Res. Otolaryngol. 10, 355-367. Brown, J.N., Miller, J.M., Altschuler, R.A., Nuttall, A.L., 1993. Osmotic pump implant for chronic infusion of drugs into the inner ear. Hear. Res. 70, 167-172. Brummett, R.E., Brown, R.T., Himes, D.L., 1979. Quantitative relationship of the ototoxic interaction of kanamycin and ethacrynic acid. Arch. Otolaryngol. 105, 240-246. Camarero, G., AvendaĂąo, C., FernĂĄndez-Moreno, C., Villar, A., Contreras, J., de Pablo, F., Pichel, J.G., Varela-Nieto, I., 2001. Delayed inner ear maturation and neuronal loss in postnatal Igf-1deficient mice. J. Neurosci. 21, 7630-7641. Coggeshall, R.E., Lekan, H.A., 1996. Methods for determining numbers of cells and synapses: a case for more uniform standards of review. J. Comp. Neurol. 364, 6-15. Davis, R.L., 2003. Gradients of neurotrophins, ion channels, and tuning in the cochlea. Neuroscientist 9, 311-316. Davis, R.L., Liu, Q., 2011. Complex primary afferents: What the distribution of electrophysiologicallyrelevant phenotypes within the spiral ganglion tells us about peripheral neural coding. Hear. Res. 276, 34-43. De Groot, J.C.M.J., Veldman, J.E., Huizing, E.H., 1987. An improved fixation method for guinea pig cochlear tissues. Acta Otolaryngol. 104, 234-242. Ernfors, P., Van De Water, T., Loring, J., Jaenisch, R., 1995. Complementary roles of BDNF and NT-3 in vestibular and auditory development. Neuron 14, 1153-1164. Ernfors, P., Duan, M.L., ElShamy, W.M., Canlon, B., 1996. Protection of auditory neurons from aminoglycoside toxicity by neurotrophin-3. Nat. Med. 2, 463-467. Fritzsch, B., Pirvola, U., Ylikoski, J., 1999. Making and breaking the innervation of the ear: neurotrophic support during ear development and its clinical implications. Cell Tissue Res. 295, 369-382. Gillespie, L.N., Clark, G.M., Marzella, P.L., 2004. Delayed neurotrophin treatment supports auditory neuron survival in deaf guinea pigs. Neuroreport 15, 1121-1125. Glueckert, R., Bitsche, M., Miller, J.M., Zhu, Y., Prieskorn, D.M., Altschuler, R.A., Schrott-Fischer, A., 2008. Deafferentation-associated changes in afferent and efferent processes in the guinea pig cochlea and afferent regeneration with chronic intrascalar brain-derived neurotrophic factor and acidic fibroblast growth factor. J. Comp. Neurol. 507, 1602-1621. Hahn, H., Salt, A.N., Biegner, T., Kammerer, B., Delabar, U., Hartsock, J.J., Plontke, S.K., 2012. Dexamethasone levels and base-to-apex concentration gradients in the scala tympani perilymph after intracochlear delivery in the guinea pig. Otol. Neurotol. 33, 660-665. Havenith, S., Versnel, H., Agterberg, M.J.H., De Groot, J.C.M.J., Sedee, R.J., Grolman, W., Klis, S.F.L., 2011. Spiral ganglion cell survival after round window membrane application of brainderived neurotrophic factor using gelfoam as carrier. Hear. Res. 272, 168-177. Ito, J., Endo, T., Nakagawa, T., Kita, T., Kim, T.S., Iguchi, F. 2005. A new method for drug application to the inner ear. ORL J. Otorhinolaryngol. Relat. Spec. 67, 272-275. 84

Spiral ganglion cell morphology

Koitchev, K., Guilhaume, A., Cazals, Y., Aran, J.M., 1982. Spiral ganglion changes after massive aminoglycoside treatment in the guinea pig. Counts and ultrastructure. Acta Otolaryngol. 94, 431-438. Kuribara, M., Hess, M.W., Cazorla, M., Roubos, E.W., Scheenen, W.J.J.M., Jenks, B.G., 2011. Brainderived neurotrophic factor stimulates growth of pituitary melanotrope cells in an autocrine way. Gen. Comp. Endocrinol. 170, 156-161. Leake, P.A., Hradek, G.T., 1988. Cochlear pathology of long term neomycin induced deafness in cats. Hear. Res. 33, 11-33. Leake, P.A., Hradek, G.T., Snyder, R.L., 1999. Chronic electrical stimulation by a cochlear implant promotes survival of spiral ganglion neurons after neonatal deafness. J. Comp. Neurol. 412, 543-562. Leake, P.A., Hradek, G.T., Hetherington, A.M., Stakhovskaya, O., 2011. Brain-derived neurotrophic factor promotes cochlear spiral ganglion cell survival and function in deafened, developing cats. J. Comp. Neurol. 519, 1526-1545. McFadden, S.L., Ding, D., Jiang, H., Salvi, R.J., 2004. Time course of efferent fiber and spiral ganglion cell degeneration following complete hair cell loss in the chinchilla. Brain Res. 997, 40-51. McGuinness, S.L., Shepherd, R.K., 2005. Exogenous BDNF rescues rat spiral ganglion neurons in vivo. Otol. Neurotol. 26, 1064-1072. Miller, J.M., Chi, D.H., O’Keeffe, L.J., Kruszka, P., Raphael, Y., Altschuler, R.A., 1997. Neurotrophins can enhance spiral ganglion cell survival after inner hair cell loss. Int. J. Dev. Neurosci. 15, 631-643. Miller, J.M., Le Prell, C.G., Prieskorn, D.M., Wys, N.L., Altschuler, R.A., 2007. Delayed neurotrophin treatment following deafness rescues spiral ganglion cells from death and promotes regrowth of auditory nerve peripheral processes: Effects of brain-derived neurotrophic factor and fibroblast growth factor. J. Neurosci. Res. 85, 1959-1969. Nourski, K.V., Miller, C.A., Hu, N., Abbas, P.J., 2004. Co-administration of kanamycin and ethacrynic acid as a deafening method for acute animal experiments. Hear. Res. 187, 131-133. Prieskorn, D.M., Miller, J.M., 2000. Technical report: chronic and acute intracochlear infusion in rodents. Hear. Res. 140, 212-215. Ramekers, D., Versnel, H., Grolman, W., Klis, S.F.L., 2012. Neurotrophins and their role in the cochlea. Hear. Res. 288, 19-33. Reichardt, L.F., 2006. Neurotrophin-regulated signalling pathways. Phil. Trans. R. Soc. B 361, 1545-1564. Richardson, R.T., O’Leary, S., Wise, A., Hardman, J., Clark, G., 2005. A single dose of neurotrophin-3 to the cochlea surrounds spiral ganglion neurons and provides trophic support. Hear. Res. 204, 37-47. Richter, C.P., Kumar, G., Webster, E., Banas, S.K., Whitlon, D.S. 2011. Unbiased counting of neurons in the cochlea of developing gerbils. Hear. Res. 278, 43-51. Salt, A.N., Plontke, S.K., 2009. Principles of local drug delivery to the inner ear. Audiol. Neurootol. 14, 350-360. Scheper, V., Paasche, G., Miller, J.M., Warnecke, A., Berkingali, N., Lenarz, T., Stöver, T., 2009. Effects of delayed treatment with combined GDNF and continuous electrical stimulation on spiral ganglion cell survival in deafened guinea pigs. J. Neurosci. Res. 87, 1389-99. Schimmang, T., Tan, J., Muller, M., Zimmermann, U., Rohbock, K., Kopschall, I., Limberger, A., Minichiello, L., Knipper, M., 2003. Lack of Bdnf and TrkB signalling in the postnatal cochlea leads to a spatial reshaping of innervation along the tonotopic axis and hearing loss. Development 130, 4741-4750. Shepherd, R.K., Roberts, L.A., Paolini, A.G., 2004. Long-term sensorineural hearing loss induces functional changes in the rat auditory nerve. Eur. J. Neurosci. 20, 3131-3140. 85

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 3

Shepherd, R.K., Coco, A., Epp, S.B., Crook, J.M., 2005. Chronic depolarization enhances the trophic effects of brain-derived neurotrophic factor in rescuing auditory neurons following a sensorineural hearing loss. J. Comp. Neurol. 486, 145-158. Shinohara, T., Bredberg, G., Ulfendahl, M., Pyykko, I., Olivius, N.P., Kaksonen, R., Lindstrom, B., Altschuler, R., Miller, J.M., 2002. Neurotrophic factor intervention restores auditory function in deafened animals. Proc. Natl. Acad. Sci. U.S.A 99, 1657-1660. Slepecky, N., 1996. Structure of the mammalian cochlea. In: Dallos, P., Popper, A.N., Fay, R.R. (Eds.), The Cochlea. Springer-Verlag, New York, pp. 44-129. Song, B.N., Li, Y.X., Han, D.M., 2009. Delayed electrical stimulation and BDNF application following induced deafness in rats. Acta Otolaryngol. 129, 142-154. Spoendlin, H., 1975. Retrograde degeneration of the cochlear nerve. Acta Otolaryngol. 79, 266275. Staecker, H., Kopke, R., Malgrange, B., Lefebvre, P., Van De Water, T.R., 1996. NT-3 and/or BDNF therapy prevents loss of auditory neurons following loss of hair cells. Neuroreport 7, 889894. Tsuji, J., Liberman, M.C., 1997. Intracellular labeling of auditory nerve fibers in guinea pig: central and peripheral projections. J. Comp. Neurol. 381, 188-202. Versnel, H., Agterberg, M.J.H., de Groot, J.C.M.J., Smoorenburg, G.F., Klis, S.F.L., 2007. Time course of cochlear electrophysiology and morphology after combined administration of kanamycin and furosemide. Hear. Res. 231, 1-12. Webster, M., Webster, D.B., 1981. Spiral ganglion neuron loss following organ of Corti loss: A quantitative study. Brain Res. 212, 17-30. West, B.A., Brummett, R.E., Himes, D.L., 1973. Interaction of kanamycin and ethacrynic acid. Severe cochlear damage in guinea pigs. Arch. Otolaryngol. 98, 32-37. Wise, A.K., Richardson, R., Hardman, J., Clark, G., Oâ&#x20AC;&#x2122;Leary, S., 2005. Resprouting and survival of guinea pig cochlear neurons in response to the administration of the neurotrophins brainderived neurotrophic factor and neurotrophin-3. J. Comp. Neurol. 487, 147-165. Ylikoski, J., Wersall, J., Bjorkroth, B., 1974. Degeneration of neural elements in the cochlea of the guinea-pig after damage to the organ of corti by ototoxic antibiotics. Acta Otolaryngol Suppl. 326, 23-41.

CHAPTER 4 The peripheral processes of spiral ganglion cells after intracochlear application of brain-derived neurotrophic factor in deafened guinea pigs

Laurien Waaijer, Sjaak F.L. Klis, Dyan Ramekers, Martinus H.W. van Deurzen, Ferry G.J. Hendriksen, Wilko Grolman Otology & Neurotology 34 (2013) 570-578.

Consider replacing a tiny portion of my brain with its neuromorphic equivalent. Okay, I’m still here: the operation was successful (...) We know people like this already, such as those with cochlear implants, implants for Parkinson’s disease, and others. Now replace another portion of my brain: okay, I’m still here... and again... At the end of the process, I’m still myself. – Ray Kurzweil, “The singularity is near”

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 4

Abstract Objective: To characterize the effects of deafening and subsequent treatment with brain-derived neurotrophic factor (BDNF) on the peripheral processes (PPs) of spiral ganglion cells (SGCs) in guinea pigs.

Background: BDNF may prevent degeneration of neural structures after loss of hair cells with possible relevance for cochlear implant candidates.

Methods: Guinea pigs were deafened with a combination of kanamycin and furosemide. Two weeks after deafening, intracochlear BDNF treatment was started with osmotic pumps for four weeks. Two weeks after cessation of BDNF treatment the cochleas were analyzed. PPs were counted and morphologically characterized with respect to myelination, size and shape. Results: Deafening reduced the number of PPs. We found that BDNF treatment, started 2 weeks after deafening, significantly reduced this degenerative effect. The remaining processes showed an altered morphology; compared to normal the size was reduced in deafened untreated animals and increased in BDNF treated animals. The myelin sheath appeared thinner in BDNF treated animals. Conclusion: We conclude that BDNF has potential as an agent that can improve the interface between cochlear implants and the auditory periphery.

BDNF-mediated survival of peripheral processes

1. Introduction Cochlear implants (CIs) are successful in giving patients with severe sensorineural hearing loss (SNHL) the ability to hear. For their function, these prostheses depend on direct electrical stimulation of the auditory neurons. One of the factors explaining the variability in perceptual performance of CI users might be the amount of surviving neural supply by spiral ganglion cells (SGCs) and their peripheral processes (PPs; Fayad et al., 1991). Loss of neural supply secondary to SNHL can compromise the efficacy of the implants. The spiral ganglion is a cranial sensory ganglion located in the cochlea. The SGCs are connected to the sensory hair cells in the organ of Corti by the SGCs’ PPs and connected to the brainstem by the central processes. About 95% of the SGCs are associated with the inner hair cells (IHCs) and only about 5% with the outer hair cells (OHCs) (Spoendlin, 1972; Romand and Romand, 1987). These two types of neurons have distinctive features (Romand and Romand, 1987; Spoendlin, 1985). Type-I SGCs, innervating the IHCs, are large (with diameters ranging from 10 to 15 µm) and surrounded by a myelin sheath. Type-II SGCs, innervating the OHCs, are smaller (diameter 6 to 12 µm) and mostly unmyelinated. To reach the IHCs, the type-I SGCs’ PPs project radially through the osseous spiral lamina (OSL). The peripheral portion of the process is myelinated by Schwann cells. Near the organ of Corti the processes become unmyelinated and pass through the habenula perforata (HP). The density of the afferent PPs is highest in the basal turn with a gradual decrease towards the apex (Spoendlin, 1972). SNHL is usually associated with damage to the cochlear hair cells. When hair cells are degenerated, the PPs and the SGCs progressively disappear (Webster and Webster, 1981; Spoendlin, 1984; Wise et al., 2005;Agterberg et al., 2008; Agterberg et al., 2009) as a result of retrograde degeneration (Spoendlin, 1975, 1984). The remaining PPs show morphological differences, including swelling, demyelination and degeneration of the PPs (Terayama et al., 1977, 1979; Leake and Hradek, 1988; Wise et al., 2005). The degeneration of neurons and PPs is traditionally thought to be due to the absence of neurotrophic stimulation by the hair cells in the organ of Corti (Lefebvre et al., 1992; Pirvola et al., 1992), although there is increasing evidence that the loss of supporting cells is also important (Stankovic et al., 2004; Sugawara et al., 2005; Zilberstein et al., 2012). Several studies have proven that (delayed) intracochlear application of exogenous neurotrophic factors, like brain-derived neurotrophic factor (BDNF) or neurotrophin-3 (NT-3), is effective in preventing SGCs from degeneration after induced deafness (Staecker et al., 1996; Miller et al., 1997; Gillespie and Shepherd, 2005; Pettingill et al., 2007; Agterberg et al., 2008; Agterberg et al., 2009). However, less 89

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 4

is known about the influence of BDNF on the PPs. Wise et al. (2005) found a significant PP survival in light microscopic analysis of histological sections of a long-term (33 days) deafened group of guinea pigs after a subsequent 4 weeks of combined neurotrophic treatment with BDNF and NT-3. In their short-term (5 days) deafened guinea pigs they found no significant neurotrophic influence. Glueckert et al. (2008) described an increased survival of the PPs after both immediate and delayed application of BDNF and acidic fibroblast growth factor (aFGF). Although each of the studies described above confirms that neurotrophic factors have an initial positive effect on SGC and PP survival, this effect must be permanent to be successful as a treatment. There are three studies that found a prolonged preservation of SGCs after cessation of the neurotrophic treatment (Murayama et al., 2008; Agterberg et al., 2009; Fransson et al., 2010). Thus, from previous experimental findings it is evident that BDNF alone or in a combination treatment promotes SGC survival in deafened animals. There is no quantitative evidence of the influence of BDNF alone on the PPs. The aim of the present study was to characterize and quantify morphological changes in SGC PPs after deafening and subsequent BDNF treatment. We used electron microscopy for determination of the packing densities of the PPs and for determination of their morphological parameters: size, myelin sheath thickness and shape.

2. Methods 2.1. Animals and experimental design In the present study we used the cochleas of guinea pigs studied previously by Agterberg et al. (2009). Briefly, ten healthy albino female guinea pigs (strain: Dunkin Hartley; weighing 250-350g) were purchased from Harlan Laboratories (Horst, the Netherlands) and housed in the animal care facility of the Rudolf Magnus Institute of Neuroscience (Utrecht, the Netherlands). All animals had free access to both food and water and were kept under standard laboratory conditions. Lights were on between 7:00 am and 7:00 pm. Temperature and humidity were kept constant at 21 째C and 60%, respectively. Six guinea pigs were bilaterally deafened and two weeks thereafter implanted in the right cochlea with an electrode array and cannula. Consecutively, BDNF was administered to the implanted cochleas during a 4-week period via a mini-osmotic pump. After completion of BDNF treatment, the osmotic pumps were removed. Two weeks after cessation of BDNF treatment (i.e., eight weeks after deafening) the animals 90

BDNF-mediated survival of peripheral processes

were killed and processed for histology. For histological analysis, a comparison was made between BDNF-treated and untreated ears in the same animals. For details on deafening procedure, cannula implantation and cochlear infusions see Agterberg et al. (2009). The histological data of the deafened ears were compared to data from normal cochleas presented in an earlier study (Agterberg et al., 2008). Two weeks after a sham deafening procedure a cannula was implanted through the round window membrane into the right cochlea of these animals. Through this cannula the cochleas were infused with phosphate-buffered saline (PBS) during a 4-week period. Immediately after finishing PBS infusion, the animals were sacrificed for histological preparation. Summarizing, three cohorts were studied. The deafened group consisted of six deafened animals. Both left and right cochleas of these animals were analyzed. The left cochlea remained untreated (Untreated, N = 6). The right cochleas were treated with BDNF (BDNF, N = 6). The control group consisted of four animals (Normal, N = 4). Only the right cochleas were studied. All experimental procedures were performed with the approval of the University’s Committee on Animal Research (DEC-UMC #03.04.036). 2.2. Deafening procedure and ABRs Before and after deafening acoustically evoked auditory brainstem responses (aABRs) were measured to assess the extent of hearing loss. All deafened animals included in this study demonstrated a threshold shift of ≥ 50 dB, measured 14 days after deafening. After cochlear implantation electrically evoked ABRs (eABRs) were recorded twice per week. For further details on the used method of recording the ABRs and their results we refer to Agterberg et al. (2009). The animals were deafened with a subcutaneous injection of kanamycin (400mg/ kg) (aminoglycoside antibiotic) followed by a slow intravenous infusion of furosemide (100mg/kg) (loop diuretic). This procedure has been shown to cause severe damage to the organ of Corti, including loss of hair cells and supporting cells (Gillespie et al., 2003; Versnel et al., 2007). For the normal animals a subcutaneous and intravenous injection of isotonic saline was used instead (Agterberg et al., 2008). 2.3. Cannula implantation and BDNF treatment In the deafened group, two weeks after the deafening procedure a six-electrode array with drug-delivery cannula was inserted through a cochleostomy in the basal turn near the round window of the right cochlea (Agterberg et al., 2009). This cannula was connected to an Alzet® mini-osmotic pump (model 2004; flow rate 0.25 μl/h; reservoir

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 4

200 μl). The cannula and the mini-osmotic pump were filled with BDNF (PeproTech Inc., Rocky Hill, NJ, USA) solution at 100 μg/ml. We have chosen this concentration because several studies (Gillespie et al., 2003; Yamagata et al., 2004; Wise et al., 2005; Miller et al, 2007; Agterberg et al., 2008; Shepherd et al., 2008) have proven that concentrations in the range of 50-100 µg/ml are effective. The osmotic pump generated a constant flow rate of BDNF during four weeks. In the control group of normal hearing animals the right cochleas were inserted with a cannula through the round window membrane, the osmotic pumps were loaded with PBS.

2.4. Fixation Immediately after sacrificing the animals, the cochleas were fixated by intralabyrinthine perfusion with a fixative consisting of 3% glutaraldehyde, 2% formaldehyde, 1% acrolein and 2.5% DMSO in 0.08 M sodium cacodylate buffer (pH 7.4), followed by immersion in the same fixative for 3 h at room temperature and subsequent histological processing for light and electron microscopy. For further details, see de Groot et al. (1987). 2.5. Cochlear sections Figure 1 demonstrates the used sectioning method. After fixation and preparation, cochleas were divided into two halves along a standardized midmodiolar plane. One half of each cochlea was divided in three transverse sections resulting in a basal, medial and apical turn. These three different half turns were re-embedded separately. After reembedding, the half turns were cut in semithin (1 µm) sections from the outer line of the cochlea in the direction of the modiolus. The tapering shape of the OSL, the radial course and the degeneration pattern of the PPs might result in different PP packing densities at different locations in the OSL. Therefore we used a standardized sectioning method, using the junction of the inner sulcus and limbus as a landmark. When this level was reached, we made four extra semithin coupes. From this level we made ultrathin (60-90 nm) sections of the reembedded half turns of the cochleas, which were contrast-stained with 7% methanolic uranyl acetate and Reynold’s lead citrate.

BDNF-mediated survival of peripheral processes

Fig. 1. A Overview of the performed cochlear preparation and subsequent sectioning. B Midmodiolar view of the OSL and the Organ of Corti in a basal turn. The line projects the position in the OSL at which the SGCs’ peripheral processes were analyzed.

2.6. Electron microscopy The characteristics of the PPs in the mid 100 µm of the OSL were examined with a JEOL 1200EX transmission electron microscope (Fig. 2). For morphological analysis it is important to analyse the cross-sections of the PPs in the middle of the OSL, since these processes are then cut perpendicularly. Mean axoplasm area, myelin sheath area and roundness in Figures 6, 7 and 8 were obtained by averaging all individual measurements of the PPs within one cochlear location, and consecutively averaging all data of that location across animals within a treatment group. The number of analyzed individual PPs cross-sections in the Normal group was 2607, in the Untreated 1722 and in the BDNF treated 2881. Measurements on the digitalized electron micrographs were performed by using image analysis program NIH ImageJ (version 1.63; US National Institutes of Health, Bethesda, MD, USA).

2.6.1. Determination of packing density The bony boundaries of the OSL were outlined and the cross sectional area (in 1000 µm2) without adjacent blood vessels was calculated. The number of cross-sections of PPs was counted. Packing densities of the PPs were calculated by dividing the number of PPs by the cross sectional area of the OSL. 2.6.2. Determination of axoplasm area The axoplasm area delineated by the most inner layer of the myelin sheath was outlined. We chose this measurement since the myelin is the most clearly distinctive feature of the PPs, whereas the axoplasm was not always clearly distinguishable in case of axoplasm shrinking.

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 4

Fig. 2. Electron micrograph of the mid 100 µm in the OSL, used for analysis of the morphological characteristics of the SGCs’ peripheral processes.

2.6.3. Determination of myelin sheath area The morphological features of the myelin sheath were quantified by measuring the absolute area of myelin per peripheral process. We outlined the outer and inner layer of myelin and then measured the donut-shaped area between the outer and inner circle in µm2.

2.6.4. Determination of roundness Roundness is a feature that can be measured directly in NIH ImageJ after delineating the perimeter of the process. It is calculated as: 4A / (π * major axis2), where A is the area. We used the most outer layer of the myelin sheath to determine the roundness.

2.7. Statistical analysis The statistical analyses were carried out with SPSS, version 15.0 for Windows. We used a two-way repeated-measures analysis of variance (rm ANOVA) to compare differences between the Untreated and BDNF group, with cochlear location (basal, middle or apical) and treatment (BDNF treated or untreated) as within-factors. To compare the data of the Untreated and BDNF cochleas with the Normal cochleas we used a rm ANOVA with the cochlear location (basal, middle, apical) as within and group (Normal, Untreated, BDNF) as between factor. We performed a two-sided posthoc Dunnett test, comparing Untreated and BDNF with the Normal cochleas.

BDNF-mediated survival of peripheral processes

3. Results One basal turn of a BDNF-treated cochlea could not be analyzed due to a mistaken sectioning plane resulting in transversally sectioned PPs. For statistical analysis these missing values were substituted per treatment group per location in the cochlea.

3.1. Packing density of the peripheral processes Figure 3 shows electron micrographs of the PPs in the basal turn, providing typical examples of the PPs in a normal (3A), deafened untreated (3B) and deafened BDNF treated (3C) cochlea. In the Normal, the OSL is occupied by round cross-sections of the PPs. Eight weeks after deafening a major decrease in number is evident in the Untreated group. In cochleas treated with BDNF the amount of PPs appears comparable to Normal. In this example the processes appear less round as compared with the Normal, but rounder as compared with the Untreated. Figure 4 compares the mean packing densities of the PPs in the Normal (N = 4), Untreated (N = 6) and BDNF (N = 6) groups. The densities are separately illustrated for the different cochlear locations. The data confirm the result shown in Fig 3A-C. The mean PP packing density in the Untreated group was reduced by about 60% as compared to the mean PP packing density of Normal hearing guinea pigs. The PP packing densities in the BDNF-treated cochleas (BDNF) were near normal at all locations and by a factor 2 larger than those in Untreated cochleas. The PP packing densities in the BDNF-treated cochleas (BDNF) were significantly larger than those in the Untreated [rm ANOVA â&#x20AC;&#x201C; main effect of BDNF: F(1,2) = 39.7, P < 0.01). There was no significant influence of location on the packing density, nor did location have an (interaction) effect on BDNF. Comparison of the PP packing densities with the normal situation revealed a significant difference between the Normal and Untreated group [rm ANOVA: F(1,2) = 42.9, P < 0.001; post-hoc Dunnett Untreated vs. Normal: P < 0.001; BDNF vs. Normal: P > 0.1].

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 4

Fig. 3. Electron micrographs of the peripheral processes in the OSL of the basal turn of the cochleas of the three different groups. A Normal B Untreated C BDNF.

BDNF-mediated survival of peripheral processes

Packing density (PPs/1000 µm2)

90 80 70 60 50 Untreated BDNF Normal

40 30 20 10 0 basal

middle

apical

Cochlear location

Fig. 4. Mean packing densities of the peripheral processes in the Normal (N = 4), Untreated (N = 6) and BDNF (N = 6) groups, separately illustrated for the different cochlear locations. Dashed lines represent the packing densities in the normal cochleas, with their SEM in grey. **: P < 0.01. Statistical analyses between the Normal and deafened groups are not shown but are described in the text. Error bars: SEM.

3.2. Morphology of the peripheral processes The Normal PPs demonstrated a prominent myelin sheath and had a near-round shape (Fig. 3A). After deafening not only the number of PPs in the Untreated cochleas dramatically decreased; also their appearance changed. In the deafened Untreated cochleas the PPs appeared elongated and collapsed (Fig. 3B). In the BDNF treated cochleas the PPs had an appearance more similar to the normal processes (Fig. 3C). Most often they had a size similar to those seen in normal cochleas. Some processes in the BDNF group however, seemed increased far beyond the size seen in the Normal or Untreated group. These large processes appeared as ‘empty’ and ‘swollen’ and were lined by a thin myelin sheath (Fig. 5A, D). Nevertheless, most processes had a size and shape similar to normal. Further, the BDNF-treated processes showed a wide variation in myelin sheath appearance and thickness (Fig. 5C). In the BDNF group we saw numerous un- and thinly myelinated processes. The frequency of unmyelinated processes was higher than in the other groups. Some of the myelin sheaths in the BDNF group were different from the other groups (Fig. 5B); there were damaged sheaths, some had vacuoles within the myelin, and others showed delamination. Also processes with thin non-continuous myelin linings were seen. Some processes had a myelin sheath within their myelin sheath (Fig. 5E). 97

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 4

Additionally, we observed an increased number of Schwann cell nuclei in the BDNF group compared to the contralateral Untreated cochlea, comparable with the increase in number of PPs, a finding that has been described previously (Wise et al., 2005).

Fig. 5. Detailed electron micrographs of some interesting features of the peripheral processes in the BDNF group. Magnification 6000x. A Swollen processes and non-continuous myelin lining. B Different conditions of myelin sheaths. C High prevalence of unmyelinated processes. D Swollen process. E Myelin sheath within myelin sheath.

BDNF-mediated survival of peripheral processes

3.2.1. Axoplasm area Figure 6 shows a histogram of the mean values of the axoplasm area of the PPs at the three different cochlear locations for each group. Compared to Normal, the axoplasm in the Untreated group was decreased at all cochlear locations, with a reduction by about 55% as compared to the mean area of normal PPs, indicating a shrinking of the PPs in case of deafening. When treated with BDNF, however, the axoplasm area was similar to the Normal situation. This effect was most profound in the basal turn. The rm ANOVA showed a significant effect of BDNF [rm ANOVA â&#x20AC;&#x201C; main effect of BDNF: F(1,2) = 18.0; P < 0.01]. Comparison of the PP axoplasm area with the normal situation revealed a significant difference between the Untreated and Normal group [rm ANOVA: F(1,2) = 22.2, P < 0.001; post-hoc Dunnett Untreated vs. Normal: P < 0.001; BDNF vs. Normal: P > 0.1].

3.5

Axoplasm (Âľm2)

3.0 2.5 Untreated

2.0

BDNF Normal

1.5 1.0 0.5 0.0 basal

middle

apical

Cochlear location

Fig. 6. Mean axoplasm area of the cross-sections of the peripheral processes at different cochlear locations. Dashed lines represent the Normal cochleas, with their SEM in grey. **: P < 0.01. Statistical analyses between the Normal and deafened groups are not shown but are described in the text. Error bars: SEM.

3.2.2. Myelin sheath area Figure 7 shows a histogram of the mean values of the absolute myelin sheath area of the PPs at three cochlear locations. In the Untreated animals the myelin did not significantly differ from the Normal situation. In the processes treated with BDNF however, we observed a decrease in the mean myelin sheath area of about 40% as compared to the mean myelin sheath area of normal PPs. 99

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Compared to the contralateral Untreated PPs, this decrease was significant [rm ANOVA – main effect of BDNF: F(1,2) = 11.2, P < 0.05]. In all groups we saw an increase in myelin sheath area from basal to apical. This was a significant main effect of cochlear location [rm ANOVA – main effect of location: F(1,2) = 34.0, P < 0.01]. There was no interaction between treatment and location. Comparison of the myelin sheath area with the normal situation revealed a significant difference between the BDNF and Normal group [rm ANOVA: F(1,2) = 11.4, P < 0.01; post-hoc Dunnett BDNF vs. Normal: P < 0.01; Untreated vs. Normal: P > 0.4]. 4.0 3.5

Myelin sheath (µm2)

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 4

3.0 2.5

Untreated

2.0

Normal

BDNF

1.5 1.0 basal

middle

apical

Cochlear location

Fig. 7. Mean myelin sheath area of the cross-sections of the peripheral processes at different cochlear locations. Dashed lines represent the Normal cochleas with their SEM in grey. *: P < 0.05. Statistical analyses between the Normal and deafened groups are not shown but are described in the text. Error bars: SEM.

3.2.3. Roundness Figure 8 shows a histogram of the mean values of the roundness of the PPs at the three different cochlear locations. Untreated PPs are less round compared to Normal. Again we saw an influence of treatment; BDNF-treated cochleas appeared to have significantly rounder PPs compared with the Untreated cochleas. This effect was significant [rm ANOVA – main effect of BDNF: F(1,2) = 24.3, P < 0.01]. The difference between the Untreated and the BDNF group was most profound in the basal turn. This is reflected in a small but significant main effect of cochlear location [rm ANOVA – main effect of location: F(1,2) = 5.3, P < 0.05). There was no interaction between the factors treatment and location. 100

BDNF-mediated survival of peripheral processes

Comparison of the roundness with the normal situation revealed a significant difference between the Untreated and Normal group [rm ANOVA: F(1,2) = 13.4, P < 0.01; post-hoc Dunnett Untreated vs. Normal: P < 0.01; BDNF vs. Normal: P > 0.3].

0.7

Roundess

0.6 0.5

Untreated BDNF Normal

0.4 0.3

0.2 basal

middle

apical

Cochlear location

Fig. 8. Mean roundness of the cross-sections of the peripheral at different cochlear locations. Dashed lines represent the Normal cochleas, with their SEM in grey. **: P < 0.01. Statistical analyses between the Normal and deafened groups are not shown but are described in the text. Error bars: SEM.

4. Discussion 4.1. Packing density Our finding that the packing density in deafened BDNF-treated cochleas is comparable with the normal situation indicates that BDNF is successful in maintaining the PPs of the SGCs. Our findings are in accordance with two previous studies with various neurotrophic agents. Wise et al. (2005) found the same effect of neurotrophic treatment on the number of PPs in a long-term (33 days) deafened group that received a delayed neurotrophic treatment with a combination of BDNF and Neurotrophin-3 (NT-3). In their short-term (5 days) deafened group, however, they found no significant treatment effect, probably because there was not much effect of the deafening after such a short interval. In contrast with our findings, both of their long-term neurotrophin treated groups showed a significant decline in number of PPs compared to normal indicating that the protective effect in their experiment was less than reported here. Possibly, this 101

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 4

discrepancy is caused by a difference in the administered neurotrophin – 50 µg/ml of BDNF and NT-3 in their study, whereas we used 100 µg/ml of BDNF. Interestingly, this would imply that BDNF is a more potent agent for the protection or resprouting of PPs than NT-3. Glueckert et al. (2008) qualitatively analyzed the afferent PPs of SGCs in an immediate (3 days post-deafening) and a delayed (3 weeks post-deafening) group treated with a combination of BDNF and aFGF. They also found a significant prevention of degeneration of the number of peripheral processes in both groups, but did not include a quantitative comparison with normal hearing animals. 4.2. Morphology of the peripheral processes

4.2.1. Axoplasm area Axoplasm area decreased significantly after deafening. This result is in accordance with previous findings; neuronal shrinkage is a well-known feature of degenerating auditory neurons (Ylikoski et al., 1974; Terayama et al., 1977, 1979; Leake and Hradek, 1988; Shepherd and Javel, 1997; Araki et al., 1998; Agterberg et al., 2009). Neurotrophic treatment apparently resulted in an increase in axoplasm area, adding up to a situation comparable to that of healthy controls. This may indicate that these PPs are as healthy as those in normal hearing animals. 4.2.2. Myelin sheath Deafening does not cause changes in the extent of myelination. In the deafened untreated group we found no significant change in the mean area of the myelin sheath compared with the normal situation. In the BDNF-treated group, however, we found a significant decrease in mean cross-sectional myelin sheath area. This may represent early signs of degeneration, but we did not see this in untreated degenerating cochleas. Another possibility may be that there is sprouting of new PPs which are in the process of developing a myelin sheath. Several studies report a resprouting of PPs following disruption of the organ of Corti. Some found that Schwann cells enter the damaged organ of Corti along the surviving PPs, and form myelin lamellae around them (Webster and Webster, 1981; Terayama et al., 1977; Bohne and Harding, 1992; Strominger et al., 1995; Lawner et al., 1997). Resprouting of PPs after neurotrophic treatment has been shown by Wise et al. (2005). They also observed an increase in number of Schwann cells, which are responsible for myelination. We observed similar Schwann cell aggregation (e.g., Fig. 3). Larger numbers of Schwann cells may indicate PP regrowth – rather than preservation of existing PPs – since it is known that myelin internodes are shorter for newly developed neurites than for old ones (e.g., Friede et al., 1985). Hence, in order to cover the entire length of the PPs, a larger number of Schwann cells is needed. 102

BDNF-mediated survival of peripheral processes

4.2.3. Roundness Roundness of the PPs was decreased after deafening; this is in accordance with previous studies that reported a collapsing myelin sheath with a shrunken axoplasm in case of deafening (Wise et al., 2005; Leake and Hradek, 1988). Neurotrophic treatment results in rounder processes that do not significantly differ from normal processes. Again, this feature may indicate that not only are there more PPs in BDNF-treated animals, but that they appear healthier as well.

4.3. Comparison with ganglion cell analyses The cochleas we analyzed for effects of deafening and neurotrophic treatment on PPs were used previously in a study in which the features of spiral ganglion cell somata were determined (Agterberg et al., 2009). Briefly, they reported similar effects on SGC packing density. Somatic area was larger than in normal animals, which is a feature we do not see in PPs. A point which was made in the previous study of Agterberg et al. (2009) was that the effect of BDNF persisted for two weeks after cessation of treatment. This was in contrast with the results of Gillespie et al. (2003). These authors reported accelerated loss of SGCs after cessation of BDNF treatment. However, several other authors also reported a persisting effect of neurotrophic treatment (Maruyama et al., 2008; Fransson et al., 2010), which we have now shown for PPs as well.

4.4. Functional implications BDNF-mediated regrowth of PPs arguably has functional consequences. When the CI electrode is closer to excitable tissue, excitation thresholds may be lower. Lower current levels would in turn lead to less spread of excitation, which may improve frequency discrimination in CI users. The thresholds that were obtained from eABR recordings in the animals from the present study have been presented before (Agterberg et al., 2009). Interestingly, the eABR threshold is similar for BDNF-treated animals and untreated controls. Possibly eABR recordings are not suitable to assess the excitation threshold of the peripheral SGCs. This may result from augmented sensitivity of brainstem nuclei to reduced sensory auditory input – this kind of normalizing mechanism has been reported in a variety of neural tissues (Turrigiano, 1999). Alternatively, it could simply be that the most sensitive SGCs – including their PPs – survive, so that the whole-nerve excitation threshold remains unchanged. In contrast to the unchanged eABR threshold after deafening, the amplitude of wave I decreased after deafening and was preserved with BDNF treatment (Agterberg et al., 2009). This, however, might just as well have resulted from the degeneration and preservation of SGC somata rather than from that of their PPs. 103

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 4

NT-3 is expressed in an apex-to-base gradient in the organ of Corti, whereas there is an opposite base-to-apex gradient of cochlear BDNF expression (Schimmang et al., 2003; Tan and Shepherd, 2006; Flores-Otero and Davis, 2011). These gradients are correlated with apex-to-base differences in SGC membrane electrical properties and expression of ion channels and synaptic proteins (Green et al., 2012). The spatial gradient of neurotrophic expression plays an important role in specifying positiondependent features of SGCs. For example, exposure to NT-3 in vitro causes SGNs to adopt an apical phenotype (Adamson et al., 2002) and basal portions of the organ of Corti co-cultured with SGCs induce a basal SGC phenotype in a BDNF-dependent manner (Flores-Otero et al., 2007). Since BDNF and NT-3 appear to have distinct actions in regulation of the electrophysiological and synaptic phenotype of the SGCs there may be a need for delivery of different neurotrophins to different areas.

4.5. Conclusion / Clinical implications Our finding that BDNF treatment can enhance the amount and apparent quality of peripheral processes has clinical relevance for the use of future cochlear implants. Regrowth or preservation of PPs leads to closer proximity between the cochlear implant electrode array and its neuronal target, enhancing the efficacy of the cochlear implant by lowering of the excitation thresholds, thereby decreasing power consumption and lowering the risk of electrochemically-induced tissue damage. It might furthermore increase the range of stimulation and improve the tonotopic specificity of electrical stimulation. Our findings also have relevance in relation to the international effort to apply a regenerative medicine approach for curing deafness. Even when, in the future, we may find a way to regenerate the organ of Corti, we still need a population of auditory neurons and their processes to restore the functional link between regenerated hair cells and the brain. BDNF is a promising treatment option for maintaining this population.

104

BDNF-mediated survival of peripheral processes

References Adamson, C.L., Reid, M.A., Davis, R.L., 2002. Opposite actions of brain-derived neurotrophic factor and neurotrophin-3 on firing features and ion channel composition of murine spiral ganglion neurons. J. Neurosci. 22, 1385–1396. Agterberg, M.J.H., Versnel, H., De Groot, J.C.M.J., Smoorenburg, G.F., Albers, F.W.J., Klis, S.F.L., 2008. Morphological changes in spiral ganglion cells after intracochlear application of brainderived neurotrophic factor in deafened guinea pigs. Hear. Res. 244, 25-34. Agterberg, M.J.H., Versnel, H., Van Dijk, L.M., De Groot, J.C.M.J., Klis, S.F.L., 2009. Enhanced survival of spiral ganglion cells after cessation of treatment with brain-derived neurotrophic factor in deafened guinea pigs. J. Assoc. Res. Otolaryngol. 10, 355-367. Araki, S., Kawano, A., Seldon, L., Shepherd, R.K., Funasaka, S., Clark, G.M., 1998. Effects of chronic electrical stimulation on spiral ganglion neuron survival and size in deafened kittens. Laryngoscope 108, 687-695. Bohne, B.A., Harding, G.W., 1992. Neural regeneration in the noise-damaged chinchilla cochlea. Laryngoscope 102, 693-703. de Groot, J.C.M.J., Veldman, J.E., Huizing, E.H., 1987. An improved fixation method for guinea pig cochlear tissues. Acta Otolaryngol. 104, 234-242. Fayad, J., Linthicum, F.H., Jr, Otto, S.R., Galey, F.R., House, W.F., 1991. Cochlear implants: histopathologic findings related to performance in 16 human temporal bones. Ann. Otol. Rhinol. Laryngol. 100, 807-811. Flores-Otero, J., Davis, R.L., 2011. Synaptic proteins are tonotopically graded in postnatal and adult type I and type II spiral ganglion neurons. J. Comp. Neurol. 519, 1455–1475. Flores-Otero, J., Xue, H.Z., Davis, R.L., 2007. Reciprocal regulation of presynaptic and postsynaptic proteins in bipolar spiral ganglion neurons by neurotrophins. J. Neurosci. 27, 14023–14034. Fransson, A., Maruyama, J., Miller, J.M., Ulfendahl, M., 2010. Post-treatment effects of local GDNF administration to the inner ears of deafened guinea pigs. J. Neurotrauma 27, 1745-1751. Friede, R.L., Brzoska, J., Hartmann, U., 1985. Changes in myelin sheath thickness and internode geometry in the rabbit phrenic nerve during growth. J. Anat. 143, 103-113. Gillespie, L.N., Clark, G.M., Bartlett, P.F., Marzella, P.L., 2003. BDNF-induced survival of auditory neurons in vivo: Cessation of treatment leads to accelerated loss of survival effects. J. Neurosci. Res. 71, 785-790. Gillespie, L.N., Shepherd, R.K., 2005. Clinical application of neurotrophic factors: the potential for primary auditory neuron protection. Eur. J. Neurosci. 22, 2123-2133. Glueckert, R., Bitsche, M., Miller, J.M., Zhu, Y., Prieskorn, D.M., Altschuler, R.A., Schrott-Fischer, A., 2008. Deafferentation-associated changes in afferent and efferent processes in the guinea pig cochlea and afferent regeneration with chronic intrascalar brain-derived neurotrophic factor and acidic fibroblast growth factor. J. Comp. Neurol. 507, 1602-1621. Green, S.H., Bailey, E., Wang, Q., Davis, R.L., 2012. The Trk A, B, C’s of neurotrophins in the cochlea. Anat. Rec. 295, 1877-1895. Lawner, B.E., Harding, G.W., Bohne, B.A., 1997. Time course of nerve-fiber regeneration in the noise-damaged mammalian cochlea. Int. J. Dev. Neurosci. 15, 601-617. Leake, P.A., Hradek, G.T., 1988. Cochlear pathology of long term neomycin induced deafness in cats. Hear. Res. 33, 11-33. Lefebvre, P.P., Weber, T., Rigo, J.M., Staecker, H., Moonen, G., Van De Water, T.R., 1992. Peripheral and central target-derived trophic factor(s) effects on auditory neurons. Hear. Res. 58:185192. 105

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 4

Maruyama, J., Miller, J.M., Ulfendahl, M., 2008. Glial cell line-derived neurotrophic factor and antioxidants preserve the electrical responsiveness of the spiral ganglion neurons after experimentally induced deafness. Neurobiol. Dis. 29, 14-21. Miller, J.M., Chi, D.H., O’Keeffe, L.J., Kruszka, P., Raphael, Y., Altschuler, R.A., 1997. Neurotrophins can enhance spiral ganglion cell survival after inner hair cell loss. Int. J. Dev. Neurosci. 15, 631-643. Miller, J.M., Le Prell, C.G., Prieskorn, D.M., Wys, N.L., Altschuler, R.A., 2007. Delayed neurotrophin treatment following deafness rescues spiral ganglion cells from death and promotes regrowth of auditory nerve peripheral processes: effects of brain-derived neurotrophic factor and fibroblast growth factor. J. Neurosci. Res. 85, 1959-1969. Pettingill, L.N., Richardson, R.T., Wise, A.K., Shepherd, R.K., O’Leary, S., 2007. Neurotrophic factors and neural prostheses: potential clinical applications based upon findings in the auditory system. IEEE Trans. Biomed. Eng. 54, 1138-1148. Pirvola, U., Ylikoski, J., Palgi, J., Lehtonen, E. Arumäe, U., Saarma, M., 1992. Brain-derived neurotrophic factor and neurotrophin 3 mRNAs in the peripheral target fields of developing inner ear ganglia. Proc. Natl. Acad. Sci. USA 89, 9915-9919. Romand, M.R., Romand, R., 1987. The ultrastructure of spiral ganglion cells in the mouse. Acta Otolaryngol. 104, 29-39. Schimmang, T., Tan, J., Müller, M., Zimmermann, U., Rohbock, K., Kôpschall, I., Limberger, A., Minichiello, L., Knipper, M., 2003. Lack of BDNF and TrkB signalling in the postnatal cochlea leads to a spatial reshaping of innervation along the tonotopic axis and hearing loss. Development 130, 4741-4750. Shepherd, R.K., Coco, A., Epp, S.B., 2008. Neurotrophins and electrical stimulation for protection and repair of spiral ganglion neurons following sensorineural hearing loss. Hear. Res. 242, 100-109. Shepherd, R.K., Javel. E., 1997. Electrical stimulation of the auditory nerve. I. Correlation of physiological responses with cochlear status. Hear. Res. 108, 112-144. Spoendlin, H., 1972. Innervation densities of the cochlea. Acta Otolaryngol. 73, 235-248. Spoendlin, H., 1975. Retrograde degeneration of the cochlear nerve. Acta Otolaryngol. 79, 266275. Spoendlin, H., 1984. Factors inducing retrograde degeneration of the cochlear nerve. Ann. Otol. Rhinol. Laryngol. 93 (Suppl. 112), 76-82. Spoendlin, H., 1985. Anatomy of cochlear innervation. Am. J. Otolaryngol. 6, 453-467. Staecker, H., Kopke, R., Malgrange, B., Lefebvre, P., Van de Water, T.R., 1996. NT-3 and/or BDNF therapy prevents loss of auditory neurons following loss of hair cells. Neuroreport. 7, 889894. Stankovic, K., Rio, C., Xia, A., Sugawara, M., Adams, J.C., Liberman, M.C., Corfas, G., 2004. Survival of adult spiral ganglion neurons requires erbB receptor signaling in the inner ear. J. Neurosci. 24, 8651-8661. Strominger, R.N., Bohne, B.A., Harding, G.W., 1995. Regenerated nerve fibers in the noise-damaged chinchilla cochlea are not efferent. Hear. Res. 92, 52-62. Sugawara, M., Corfas, G., Liberman, M.C., 2005. Influence of supporting cells on neuronal degeneration after hair cell loss. J. Assoc. Res. Otolaryngol. 6, 136-147. Tan, J., Shepherd, R.K., 2006. Aminoglycoside-induced degeneration of adult spiral ganglion neurons involves differential modulation of Tyrosine Kinase B and p75 neurotrophin receptor signaling. Am. J. Pathol. 169, 528–543. Terayama, Y., Kaneko, Y., Kawamoto, K., Sakai, N., 1977. Ultrastructural changes of the nerve elements following disruption of the organ of Corti. I. Nerve elements in the organ of Corti. Acta Otolaryngol. 83, 291-302. 106

BDNF-mediated survival of peripheral processes

Terayama, Y., Kaneko, K., Tanaka, K., Kawamoto, K., 1979. Ultrastructural changes of the nerve elements following disruption of the organ of Corti. II. Nerve elements outside the organ of Corti. Acta Otolaryngol. 88, 27-36. Turrigiano, G.G., 1999. Homeostatic plasticity in neuronal networks: the more things change, the more they stay the same. Trends Neurosci. 22, 221-227. Versnel, H., Agterberg, M.J.H., de Groot, J.C.M.J., Smoorenburg, G.F., Klis, S.F.L., 2007. Time course of cochlear electrophysiology and morphology after combined administration of kanamycin and furosemide. Hear. Res. 231, 1-12. Webster, M., Webster, D.B., 1981. Spiral ganglion neuron loss following organ of Corti loss: a quantitative study. Brain Res. 212, 17–30. Wise, A.K., Richardson, R., Hardman, J., Clark, G., O’leary, S., 2005. Resprouting and survival of guinea pig cochlear neurons in response to the administration of the neurotrophins brainderived neurotrophic factor and neurotrophin-3. J. Comp. Neurol. 487, 147-165. Yamagata, T., Miller, J.M., Ulfendahl, M., Olivius, N.P., Altschuler, R.A., Pyykkö, I., Bredberg, G., 2004. Delayed neurotrophic treatment preserves nerve survival and electrophysiological responsiveness in neomycin-deafened guinea pigs. J. Neurosci. Res. 78, 75-86. Ylikoski, J., Wersall, J., Bjorkroth, B., 1974. Degeneration of neural elements in the cochlea of the guinea pig after damage to the organ of corti by ototoxic antibiotics. Acta Otolaryngol. Suppl. 326, 23–41. Zilberstein, Y., Liberman, M.C., Corfas, G., 2012. Inner hair cells are not required for survival of spiral ganglion neurons in the adult cochlea. J. Neurosci. 32, 405-410.

107

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

CHAPTER 5 Auditory-nerve responses to varied inter-phase gap and phase duration of the electric pulse stimulus as predictors for neuronal degeneration

Dyan Ramekers, Huib Versnel, Stefan B. Strahl, Emma M. Smeets, Sjaak F.L. Klis, Wilko Grolman Journal of the Association for Research in Otolaryngology 15 (2014) 187-202.

There’s a gap in between/ There’s a gap where we meet/ Where I end and you begin – Thom Yorke, “Where I end and you begin”

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 5

Abstract

After severe hair cell loss, secondary degeneration of spiral ganglion cells (SGCs) is observed â&#x20AC;&#x201C; a gradual process that spans years in humans but only takes weeks in guinea pigs. Being the target for cochlear implants (CIs), the physiological state of the SGCs is important for the effectiveness of a CI. For assessment of the nerveâ&#x20AC;&#x2122;s state, focus has generally been on its response threshold. Our goal was to add a more detailed characterization of SGC functionality. To this end, the electrically evoked compound action potential (eCAP) was recorded in normal-hearing guinea pigs and guinea pigs that were deafened 2 or 6 weeks prior to the experiments. We evaluated changes in eCAP characteristics when the phase duration (PD) and inter-phase gap (IPG) of a biphasic current pulse were varied. We correlated the magnitude of these changes to quantified histological measures of neurodegeneration (SGC packing density and SGC size). The maximum eCAP amplitude, derived from the input-output function, decreased after deafening, and increased with both PD and IPG. The eCAP threshold did not change after deafening, and decreased with increasing PD and IPG. The dynamic range was wider for the six-weeks-deaf animals than for the other two groups. Excitability increased with IPG (steeper slope of the input-output function and lower stimulation level at the halfmaximum eCAP amplitude), but to a lesser extent for the deafened animals than for normal-hearing controls. The latency was shorter for the six-weeks-deaf animals than for the other two groups. For several of these eCAP characteristics the effect size of IPG correlated well with histological measures of degeneration, whereas effect size of PD did not. These correlations depend on the use of high current levels, which could limit clinical application. Nevertheless, the potential of these correlations towards assessment of the condition of the auditory nerve may be of great benefit to clinical diagnostics and prognosis in cochlear implant recipients.

110

eCAP can predict neural health

1. Introduction Loss of cochlear hair cells leads to sensorineural hearing loss (SNHL). Bypassing the lost hair cells, cochlear implants (CIs) electrically stimulate the spiral ganglion cells (SGCs), of which the axons form the auditory nerve. Proper functioning of CIs thus depends on the presence of a healthy and sufficiently large population of SGCs that is able to transduce the encoded auditory information to the brainstem. However, animal studies have shown that SGCs become smaller and degenerate as a consequence of hair cell loss (Ylikoski et al., 1974; Spoendlin, 1975; Webster and Webster, 1981; Versnel et al., 2007). This degeneration has been associated with discontinued neurotrophic support from the organ of Corti (Ernfors et al., 1995; Fritzsch et al., 1999; Zilberstein et al., 2012). Administration of exogenous neurotrophic factors has prevented SGC degeneration in multiple animal models of SNHL (for a review, see Ramekers et al., 2012); SGCs are larger and more numerous after neurotrophic treatment than in untreated controls (Ernfors et al., 1996; Richardson et al., 2005; Shepherd et al., 2005; Glueckert et al., 2008; Agterberg et al., 2008, 2009; Leake et al., 2011; van Loon et al., 2013). In human CI users, a relation between the number of surviving SGCs and CI performance has not been found (Fayad and Linthicum, 2006; Xu et al., 2012), suggesting that criteria for a well-functioning interface for CIs comprise more than mere numerical quantification. So far, attempts to assess the functional consequence of SGC degeneration mainly involved characterization of electrophysiological activity at single unit or population level in animal models of SNHL. Inconsistent findings have been reported from changes in excitation thresholds after deafening. In single neuron studies excitation thresholds become as often lower as they become higher (reviewed in Sly et al., 2007). Excitation thresholds for electrically evoked auditory brainstem responses (eABRs) have been reported to become elevated after deafening (Shepherd and Javel, 1997; Maruyama et al., 2008; Fransson et al., 2010), but also to remain unchanged (Agterberg et al., 2009). For electrically evoked compound action potentials (eCAPs) excitation thresholds are found to be higher after deafening (Stypulkowski and van den Honert, 1984; Prado-Guitierrez et al., 2006). Being consistently smaller after deafening, response amplitudes seem to reflect function better than threshold (Hall, 1990; Shepherd and Javel, 1997; Agterberg et al., 2009). However, the amplitude largely depends on factors such as the distance between the stimulation electrode and the excitable tissue, the impedance between the two, and, likewise, on the distance and impedance between the excited neural tissue and the recording electrode (Grill et al., 2009). Overcoming these unpredictable and often confounding factors, Prado-Guitierrez et al. (2006) employed a method with which the difference in excitability between 111

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 5

two stimulus pulse shapes can be utilized as a quantitative functional measure for SGC health. The magnitude of the change in excitability brought about by either a prolonged inter-phase gap (IPG) or phase duration (PD) correlated with the amount of surviving SGCs. Increasing PD leads to higher excitability because the amount of charge that is displaced toward or away from the electrode increases proportionally when phase area (PD × current level) is increased (Shepherd and Javel, 1999). The beneficial effect of an IPG increase was recognized soon after the introduction of biphasic stimulation, since it partly abolishes the increase in threshold that arises when adding a hyperpolarizing phase to a depolarizing monophasic pulse (van den Honert and Mortimer, 1979). Besides the effect on electrophysiological thresholds, an IPG increase in cochlear implant stimulation results in an increase in loudness perception (McKay and Henshall, 2003) as well as a decrease in behavioral thresholds (Carlyon et al., 2005). The aim of this study was (1) to provide a comprehensive characterization of eCAP input-output properties in normal-hearing and deafened guinea pigs, (2) to evaluate the effect of changing both IPG and PD on these properties, and (3) to correlate the magnitude of this effect to quantified histological measures of neurodegeneration. Finding electrophysiological correlates for the amount of degeneration of the SGC population may provide an important tool for clinical diagnostics and prognosis in cochlear implant recipients.

2. Methods 2.1. Animals and experimental design Eighteen female albino guinea pigs (Dunkin Hartley; 250–350 g) were obtained from Harlan Laboratories (Horst, the Netherlands) and kept under standard laboratory conditions (food and water ad libitum; lights on between 7:00 am and 7:00 pm; temperature 21 °C; humidity 60%). One group served as normal-hearing controls (NH; N = 6), while the remaining guinea pigs were exposed to ototoxic treatment either two (2WD; N = 6) or six (6WD; N = 6) weeks prior to acute electrophysiological measurements. Immediately after these experiments the animals were sacrificed and their cochleas processed for histological analysis. All surgical and experimental procedures were approved by the Animal Care and Use Committee of Utrecht University (DEC 2010.I.08.103).

112

eCAP can predict neural health

2.2. Deafening procedure Animals were anesthetized by intramuscular injection of dexmedetomidine (Dexdomitor®; 0.25 mg/kg) and ketamine (Narketan®; 40 mg/kg), and acoustically evoked auditory brainstem responses (aABRs) were recorded to measure hearing thresholds. Ototoxic treatment involved subcutaneous injection of kanamycin (SigmaAldrich, St. Louis, MO, USA; 400 mg/kg) and subsequent infusion of furosemide (Centrafarm, Etten-Leur, the Netherlands; 100 mg/kg) into the external jugular vein, which has been shown to eliminate the great majority of both inner and outer hair cells (Versnel et al., 2007; West et al., 1973). Post-operatively, the animals were injected subcutaneously with the non-ototoxic antibiotic enrofloxacin (Baytril®, 5 mg/kg) and the anti-inflammatory drug carprofen (Rimadyl®, 5 mg/kg).

2.3. Implantation For the acute experiments anesthesia was initiated with Hypnorm® (Vetapharma; 0.5 ml/kg i.m.), followed by a gas mixture of 2% isoflurane evaporated in O2 and N2O (1:2), delivered by a mouth cap. The skull was exposed and four transcranial screws were placed for ABR and eCAP recordings (one 1 cm posterior to bregma, one 2 cm anterior to bregma, and two 1 cm bilateral to bregma). The animals were subsequently tracheostomized and artificially ventilated (Amsterdam infant ventilator mk3, Hoekloos, Schiedam, the Netherlands) with a gas mixture of O2 and N2O (1:2) and 1–1.5% isoflurane (45-50 cycles/min respiration rate, 1.8-2.0 kPa) throughout the experiment. The animals were placed on a 37 °C heating pad. Their heart rate, O2 consumption, and expiratory CO2 were monitored, and, based on those readings, anesthetic settings were adjusted if needed. Every 2 hours 2 ml of lactated Ringer’s solution was injected subcutaneously to avoid dehydration. Via a retro-auricular approach the bulla was opened to expose the cochlea. A 0.5 mm cochleostomy was then drilled in the basal turn, within 1 mm from the round window, through which the electrode array was inserted into the scala tympani (see below). 2.4. Auditory brainstem responses Prior to ototoxic treatment aABRs were recorded using subcutaneously positioned needle electrodes (active electrode behind the right pinna; reference electrode on the skull, rostral to the brain; ground electrode in hind limb). During the acute experiments the posterior transcranial screw was used as active electrode, the anterior screw as reference electrode, and a subcutaneous needle electrode in the hind limb as ground. Broadband acoustic clicks (20 μs monophasic rectangular pulses; inter-stimulus interval 99 ms) were synthesized and attenuated using a TDT3 system (modules RP2, PA5 (2x) and HB7; Tucker-Davis Technologies, Alachua, FL, USA), and presented in free 113

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 5

field using a Blaupunkt speaker (PCxb352; 4 Ω; 30 W). The responses were differentially amplified using a Princeton Applied Research (Oak Ridge, TN, USA) 5113 pre-amplifier (amplification ×5,000; band pass filter 0.1–10 kHz), digitized by the TDT3 system (100 kHz sampling rate, 24-bit sigma-delta converter) and stored on a PC for off-line analysis. Hearing thresholds were obtained by starting at approximately 110 dB peak equivalent SPL, and decreasing the sound level in steps of 10 dB until the response had disappeared. The threshold was then defined as the interpolated sound level at which the aABR N1-P2 peak was 0.3 µV.

2.5. Compound action potentials Both stimulation and recording of eCAPs was done with monopolar configuration, using the transcranial screw right from bregma as return electrode for stimulation and the one left from bregma as reference electrode for recording. eCAPs were recorded with a custom-made four-contact electrode array connected to a MED-EL PULSARci100 cochlear implant (MED-EL GmbH, Innsbruck, Austria). The electrode array was inserted up to 4 mm into the cochleostomy, so that all 4 contacts – evenly spaced over the first 3 mm of the array – were in the scala tympani. For all data presented here, the most apical contact (5 mm from the round window) was used for stimulation and the most basal one (2 mm from the round window) for recording. Electrode impedance ranged from 3.4 to 4.6 kΩ (4.1 kΩ on average), so that in all cases, given the maximum implant output voltage of 6.1 V in our set-up, a 1200-µA current was within the implant’s current range. The implant was controlled by a PC via a Research Interface Box 2 (RIB2; Department of Ion Physics and Applied Physics, University of Innsbruck, Innsbruck, Austria) and a National Instruments data acquisition card (PCI-6533, National Instruments, Austin, TX, USA). eCAP data were sampled using adaptive sigma-delta modulation at 1.2 MHz (Neustetter et al., 2012). Stimulation and recording paradigms were created in MATLAB (version 7.11.0; Mathworks, Natick, MA, USA). Biphasic current pulses were presented with alternating polarity to reduce stimulation artifact (see Fig. 1A), and the responses to 50 pairs of these stimuli were averaged. We chose alternating stimulation rather than a masker-probe paradigm, since it is thought that by doing so a larger portion of the SGC population is recruited (Westen et al., 2011). The resulting waveforms did not seem to contain significant remaining stimulation artifact (e.g., see sub-threshold traces in Figs. 1B, C). In addition, the absence of significant stimulation artifact can be deduced from the absence of linear increases in the N1-P2 amplitude sub-threshold or supra-saturation in the input-output curves. Therefore, further techniques such as scaled template subtraction (Miller et al., 1998) were not applied. The phase duration (PD) and the inter-phase gap (IPG) of the current pulses were varied (PD: 20-50 µs in steps of 5 µs; IPG: 2.1-10-20-30 µs; see Fig. 114

eCAP can predict neural health

1A). Each of the 7 Ă&#x2014; 4 = 28 possible pulse shapes was presented at 10 current levels, which were adjusted for PD, such that for each PD the charge ranged from 2.4 to 24 nC: e.g., pulses with 25 Âľs PD had a maximum stimulation level of 960 current units (cu; 1 cu â&#x2030;&#x2C6; 1 ÂľA; linearity of this relationship was verified via control measurements), while 50 Âľs/phase pulses had a maximum stimulation level of 480 cu. The recording to the lowest, sub-threshold stimulation level was subtracted from the other recordings in order to reduce measurement onset artifacts. Resulting eCAPs are exemplified in Fig. 1B, C. The eCAP amplitude was defined as the difference in voltage between the N1 and the P2 peak (first positive peak following N1), and was plotted against current level. Input-output functions thus obtained were fitted with a Boltzmann sigmoid, using the Levenberg-Marquardt algorithm (Marquardt, 1963; R2 was between 0.95 and 1, and Eq. 1 voor chapter 5, Eq. 3 voor chapter 6, Eq. 1 voor chapter 7: 0.996 on average):

đ?&#x2018;&#x2030;đ?&#x2018;&#x2030;đ?&#x2018;&#x2030;đ?&#x2018;&#x2030;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; = đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; +

đ??ľđ??ľđ??ľđ??ľ

đ??źđ??źđ??źđ??źâ&#x2C6;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; đ??ˇđ??ˇđ??ˇđ??ˇ

1 â&#x2C6;&#x2019; đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2019;

(Eq. 1)

where V is amplitude in ÂľV, I is stimulus current in ÂľA and A-D are fitting parameters. parameters from the Eq. 1 voorThe chapter 6, Eq. 2derived voor chapter 7: fitting are the outcome variables (see also Fig. 1D) that were used to assess differences between pulse shapes, between animal groups, đ??źđ??źđ??źđ??źđ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ??źđ??źđ??źđ??ź â&#x2C6;&#x2019;đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ą 0 and to relate to SGC packing density and perikaryalâ&#x2C6;&#x2019; area. đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; = đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2122; ďż˝1 â&#x2C6;&#x2019; đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; đ?&#x153;?đ?&#x153;?đ?&#x153;?đ?&#x153;? ďż˝These outcome variables are maximum N1-P2 amplitude (defined by B), current level to achieve 50% of the maximum N1-P2 amplitude (defined by C), slope at C (defined by B/4D), threshold (defined by C-2D, the current level at which the tangent to the curve at C crosses A), and dynamic range 4D).3 The 1 peak latency, averaged over the three highest current levels, Eq. 2 voor(defined chapter by 6, Eq. voorNchapter 7: was analyzed in addition to these input-output characteristics. Note that the slope đ?&#x2018;&#x20AC;đ?&#x2018;&#x20AC;đ?&#x2018;&#x20AC;đ?&#x2018;&#x20AC;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2019;đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ą 0range 4D; therefore đ?&#x2018;&#x20AC;đ?&#x2018;&#x20AC;đ?&#x2018;&#x20AC;đ?&#x2018;&#x20AC;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2019;đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ą 0 B/4D equals the ratio of the amplitude B and the dynamic these â&#x2C6;&#x2019; â&#x2C6;&#x2019; đ?&#x153;?đ?&#x153;?đ?&#x153;?đ?&#x153;? đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; đ?&#x153;?đ?&#x153;?đ?&#x153;?đ?&#x153;? đ??ľđ??ľđ??ľđ??ľ â&#x2C6;&#x2122; â&#x2C6;&#x2019; đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; = đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2122; + đ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;? â&#x2C6;&#x2122; đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; ďż˝ ďż˝1 ďż˝ as an ďż˝1 three variables are interrelated. The noise level (defined by A) has not been used outcome variable. All analysis was performed using custom-made software in MATLAB.

Eq. 1 voor chapter 8: đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ą

đ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;? = đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2019;đ?&#x153;?đ?&#x153;?đ?&#x153;?đ?&#x153;?

115

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

PD IPG PD

↔↔↔

↓

dynamic range

threshold

1.5 1.0

→

amplitude

slop e

←

2.0

336 µA

level50%

100

200

noise level

300

288 µA 240 µA

192 µA

2 1

384 µA

0.5 0

signal amplitude (mV)

anodic first

480 µA 432 µA

cathodic first

N −P amplitude (mV)

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 5

400

stimulation level (µA)

500

↑ N1

144 µA 96 µA

0.4 0.8 1.2 1.6

time since stimulus onset (ms)

Fig. 1. A Schematic of the electric pulse stimuli used in this study. To reduce stimulation artifact alternating polarity stimuli were applied. The biphasic pulses were applied with IPGs of 2.1, 10, 20 and 30 µs, for each PD (20 to 50 µs in 5-µs steps). B, C Examples of eCAPs at various current levels recorded in a normal-hearing animal (B) and in a 6-weeks deaf animal (C). Stimulation conditions were identical in both cases (PD 50 µs; IPG 30 µs) and stimulation amplitude is given for each trace. D Input-output function derived from eCAP N1-P2 amplitude in (B); the solid line represents the fitted Boltzmann curve (see Eq. 1). The dashed lines indicate the eCAP characteristics that were derived from the Boltzmann equation (see Methods section). IPG, interphase gap; PD, phase duration.

2.6. Histology After completion of the experiment, all animals were sacrificed and their right cochleas were harvested. Processing and analysis was performed as previously described by van Loon et al. (2013). Intra-labyrinthine cochlear fixation was achieved with a fixative of 3% glutaraldehyde, 2% formaldehyde, 1% acrolein and 2.5% DMSO in a 0.08 M sodium cacodylate buffer. The cochleas were then decalcified, post-fixated and embedded in Spurr’s low-viscosity resin. After dividing the cochleas into two halves along a standardized midmodiolar plane, they were re-embedded in fresh resin. From each cochlea five semi-thin (1 µm) sections were cut at 30-µm intervals – to ensure that consecutive sections could not contain the same cell – which were subsequently stained with 1% methylene blue and 1% azur B in 1% borax. 116

eCAP can predict neural health

Using a Leica DC300F digital camera mounted on a Leica DMRA light microscope and a 40x oil immersion lens (Leica Microsystems GmbH, Wetzlar, Germany), pictures of each transection of Rosenthal’s canal (2 basal, 2 middle, and 3 apical transections), as well as the organ of Corti were obtained. Within each transection of Rosenthal’s canal, the number of type-I SGCs was counted and packing density was averaged across all five sections. In one of the five sections, the average perikaryal area was determined for SGCs with a visible nucleus using ImageJ (Version 1.42q, National Institutes of Health, Bethesda, MA, USA). Subsequently both packing density and perikaryal area were averaged over cochlear turns. Since the likelihood of detecting an individual SGC depends on its perikaryal size, the average packing density was corrected for perikaryal size as previously described (Coggeshall and Lekan, 1996; van Loon et al., 2013). Hair cell counts included hair cells with a nucleus, a cilia bundle, or a clear cochlear hair-celllike outline.

2.7. Statistical analysis Differences in aABR threshold shifts between the 2WD and 6WD groups were evaluated with an independent samples t test. Changes in SGC packing density and perikaryal area after ototoxic treatment were assessed with a one-way analysis of variance (ANOVA) and post-hoc Bonferroni. Pearson’s correlation coefficient was used to assess the correlation between maximum eCAP amplitude and SGC packing density. The effects of PD and IPG on eCAP characteristics (amplitude, slope, threshold, dynamic range, level50% and latency) were assessed with two-way repeated measures ANOVA; Greenhouse-Geisser correction was applied when the assumption of sphericity was violated. Correlations between differences in eCAP characteristics in response to PD or IPG changes and histological findings (SGC packing density and SGC perikaryal area) were evaluated using multiple linear regression. All statistical tests were performed with SPSS 20.0 for Windows (IBM, Armonk, NY, USA), except for R2 values in Figs. 8 and 9, which were obtained using the Statistics toolbox in MATLAB.

3. Results 3.1. Deafening aABR N1-P2 peak threshold shifts after deafening ranged from 59 to 82 dB with the exception of one animal in the 2WD group (27 dB shift). This animal was therefore excluded from analysis with respect to group averages, but not from the multiple regression analyses. Average threshold shifts per group were 61 dB for 2WD and 74 dB for 6WD, and did not significantly differ [t(9) = 2.0; P = 0.079]. 117

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 5

Ototoxic treatment resulted in significant reduction of both inner (IHCs) and outer hair cells (OHCs). 100% of IHCs were present in the NH group, and this was reduced to 45% and 13% in the 2WD and 6WD group respectively. Average OHC presence decreased from 96% in NH to 37% in 2WD and 18% in 6WD animals. OHC loss was more pronounced in the basal half of the cochlea (17% remaining in 2WD and 10% remaining in 6WD) than in the apical half (48% and 25%, respectively). For IHCs this gradient was smaller: 38% for 2WD and 8% for 6WD in the basal half, and 50% and 17%, respectively, in the apical half. For each group a typical transection of Rosenthalâ&#x20AC;&#x2122;s canal is depicted in Fig. 2. The packing density of SGCs was lower in the 2-weeks deafened group (Fig. 2B) than in the normal hearing group (Fig. 2A), and lower still in the 6-weeks deaf group (Fig. 2C). Figure 3A shows the quantification of numerical SGC packing density averaged across animals. SGC packing density significantly decreased after deafening [one-way ANOVA: F(2,14) = 86.3; P < 0.001]. Post-hoc Bonferroni showed that SGC packing density was significantly different between groups (P < 0.001 for each pair). In normal-hearing animals packing density was similar across turns. The decrease in packing density after deafening was most pronounced in the basal turn, and smallest in the apical turn. The average perikaryal area (shown in Fig. 3B) also significantly decreased after deafening [one-way ANOVA: F(2,14) = 14.1; P < 0.001; post-hoc Bonferroni: NH-2WD: P = 0.013; NH6WD: P < 0.001; 2WD-6WD: P = 0.41]. For normal-hearing animals the perikaryal area was similar across turns. The decrease in cell size after deafening was slightly larger toward the apex. In Fig. 3C SGC perikaryal area is plotted against packing density for each animal separately. Although SGC degeneration was clearly accompanied by a decrease in cell size, a large variation can be observed within experimental groups. Note that the 2WD animal with a small threshold shift has a normal packing density but a SGC size between normal and the average size found in 2WD animals.

3.2. eCAP waveforms The examples of eCAP recordings in Fig. 1B (normal hearing) and 1C (6-weeks deaf) show that the N1-P2 peak amplitude is smaller for the deafened animal, while the eCAP threshold is roughly similar (see first obvious response at the same stimulation level). The P2 peak was not as pronounced in deafened animals as it was in normal-hearing animals, but its approximate location was evident in all instances.

118

eCAP can predict neural health

Fig. 2. Examples of transections of Rosenthalâ&#x20AC;&#x2122;s canal (lower middle turn) containing the SGCs in a normal-hearing cochlea (A), a cochlea 2 weeks after deafening (B), and a cochlea 6 weeks after deafening (C). 119

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

*** ***

1600

***

1200

250

SGC perikaryl area (µm2)

2000

SGC packing density (cells/mm2)

SGC perikaryal area (µm2)

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 5

200

***

150 100

800 400 0

NH 2WD 6WD

250

50 0

NH 2WD 6WD

225 200 175 150

500

1000

1500

2000

SGC packing density (cells/mm2)

Fig. 3. Quantification of histological SGC analysis. Group averages are shown for SGC numerical packing density (A) and SGC perikaryal area (B), both first averaged across all cochlear turns per animal. C SGC perikaryal area as function of SGC packing density for each animal separately. The blue-crossed red symbol indicates the 2WD animal excluded from the group averages. NH, normal-hearing; 2WD, 2 weeks deaf; 6WD, 6 weeks deaf; SGC, spiral ganglion cell. * P < 0.05, *** P < 0.001; N = 6 for NH and 6WD, N = 5 for 2WD; error bars represent SEM.

3.3. Effect of pulse shape alterations on eCAP characteristics In Fig. 4, N1-P2 peak amplitudes for a single (normal-hearing) animal are plotted as a function of current (left column) and charge (right column) for different pulse shapes. The beneficial effect of increasing IPG on stimulation efficacy (i.e., larger amplitude and lower threshold) becomes apparent when comparing Figs. 4A and 4B (2.1 μs IPG) with 4C and 4D (30 μs IPG), respectively. Increasing PD also enhanced stimulation efficacy (see 4A, C), but only partly so when eCAP amplitude is plotted against charge (i.e., phase area) instead of current level (i.e., phase amplitude; compare 4A and 4C with 4B and 4D, respectively). Specifically, the effect of PD is abolished for 30 μs IPG (4D), but not for 2.1 μs IPG (4B). In other words, the effect of pulse shape (the combination of phase duration and phase amplitude) does not influence the response amplitude for 30 μs IPG 120

eCAP can predict neural health

as long as the phase area (charge level) is the same, while for 2.1 μs IPG pulse shape is relevant since short phases elicit lower maximum eCAP amplitudes. IPG 2.1 µs

IPG 2.1 µs

IPG 30 µs

1.5 1.0 0.5

N −P amplitude (mV)

A 2.0

1.5

PD 20 µs PD 25 µs PD 30 µs PD 35 µs PD 40 µs PD 45 µs PD 50 µs

1.0 0.5

N −P amplitude (mV)

C 2.0

0 5 10 15 20 current (dB re 100 µA)

16 20 24 28 charge (dB re 1 nC)

Fig. 4. Input-output curves from a single normal-hearing animal. Separate lines represent responses to pulses with different PD. The upper two plots show the same data for pulses with a 2.1-µs IPG, but plotted against stimulation phase amplitude in µA in (A) and against phase area in nC in (B). The lower two plots show the same input-output curves for 30-µs IPG pulses either plotted as function of phase amplitude in µA (C) or as function of phase area in nC (D). IPG, interphase gap; PD, phase duration.

For each animal the input-output functions as shown in Fig. 4 are fitted according to Eq. 1 (see Methods section). The five input-output characteristics that are derived from these fits (amplitude, slope, threshold, dynamic range and level50%) and N1 latency are averaged per group and plotted as function of IPG in Fig. 5 for 20-μs (first and third column) and 50-μs PD (second and fourth column). In Fig. 6 these eCAP characteristics are plotted as function of PD, for either 2.1 μs IPG (first and third column) or 30 μs IPG (second and fourth column). F and P values resulting from two-way repeated measures ANOVA, with IPG and PD as within factors and group as between factor, are given in Table 1. The maximum eCAP amplitude (Fig. 5A, B and 6A, B) and the slope (related to the amplitude; Fig. 5C, D and 6C, D) decreased markedly with time after deafening. Accordingly, for all 28 pulse shapes there was a significant correlation between maximum amplitude and SGC packing density [R2 between 0.49 and 0.71; P < 0.01 for all 121

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 5

pulse shapes], and between slope and SGC packing density [R2 between 0.48 and 0.67; P < 0.01 for all pulse shapes]. For all groups the maximum amplitude increased similarly with increasing IPG (Fig 5A, B) and with increasing PD (Fig. 6A, B). For the slope, dependence on IPG varied among groups as was indicated by a significant interaction. The slope increased markedly with IPG for the NH group, while this increase was less pronounced for the 2WD and smaller still for the 6WD group (Fig. 5C, D). Differences in slope with increasing PD were not different among groups (Fig. 6C, D). Note that the overall charge was kept constant as described in the previous section. eCAP threshold decreased with an increase in both IPG (Fig. 5E, F) and PD (Fig. 6E), while neither overall threshold levels nor dependence on IPG or PD were significantly different between groups. The dynamic range was similar for NH and 2WD animals, but noticeably wider for the 6WD animals (Figs. 5G, H and 6G, H). In addition, dependence on IPG was significantly different among groups: whereas it hardly changed with IPG for the NH and 2WD groups, there was an apparent increase in dynamic range for the 6WD group (Fig. 5G, H). Level50% â&#x20AC;&#x201C; at which the eCAP amplitude is half of the maximum amplitude â&#x20AC;&#x201C; decreased significantly with increasing IPG (Fig. 5I, J), but did not significantly vary with PD (Fig. 6I, J). Although there was no significant overall difference in level50% among groups, a significant interaction effect with IPG indicates that dependence on IPG varied: the decrease with increasing IPG is smaller for 6WD than for the other groups. When comparing the effect of IPG on level50% with that on threshold (Figs. 5E, F and 5I, J), it is consistently larger at threshold; this means that the amount of charge needed to equalize eCAP amplitude is larger at threshold level than at the half-maximum eCAP amplitude. Latency of the N1 peak was shorter for the 6-weeks deaf animals than for the other two groups (Figs. 5K, L and 6K, L). Dependence on PD was similar for all groups (Fig. 6K, L), while N1 latency increased with increasing IPG for both deafened groups, but not for the NH group (Fig. 5K, L). Finally, for each of the six eCAP characteristics we found a significant interaction of IPG and PD (Table 1), a feature that is well illustrated by comparison of Figs. 4B and 4D. In summary, virtually all eCAP characteristics were dependent on both IPG and PD, and most of these characteristics changed after deafening. The change with PD was similar for all groups, whereas with increasing IPG the slope, dynamic range, level50% and latency of the eCAP were affected differently after deafening (demonstrated by significant interactions between IPG and group; see Table 1). In other words, it was possible to distinguish between groups by varying IPG, but not by varying PD.

122

eCAP can predict neural health

Table 1. Results from the two-way repeated measures ANOVAs, applied separately for each eCAP characteristic. Degrees of freedom are given in brackets. amplitude

slope

threshold

main effects PDa group

IPGa

IPG*PDa

F 56.9 (1.2,17) 28.9 (1.4,20) 11.6 (2,14) P < 0.001 < 0.001 0.0011

13.5 (4.7,66) 1.0 (2.4,17) < 0.001 0.42

F 40.1 (1.1,15) 11.0 (2.2,30) 11.3 (2,14) P < 0.001 < 0.001 0.0011

F 131 (1.5,21) P < 0.001

6.4 (2.3,32) 0.0034

F 242 (1.9,27) P < 0.001

2.9 (1.4,20) 0.091

interaction effects IPG*groupa PD*groupa

2.3 (2,14) 0.13

4.8 (4.8,68) < 0.001

5.3 (2.2,15) 0.016

5.5 (2.6,18) 0.0089

13.8 (2.9,40) 2.1 (3.0,21) < 0.001 0.13

1.0 (2.8,20) 0.42

0.84 (4.3,30) 0.52

1.6 (4.6,32) 0.21

dynamic range

F 15.8 (1.3,18) 3.7 (2.3,32) P < 0.001 0.031

9.4 (2,14) 0.0026

3.3 (3.4,48) 0.024

latency

F 28.2 (1.9,26) 247 (2.6,37) P < 0.001 < 0.001

4.9 (2,14) 0.024

4.1 (7.1,100) 13.6 (3.7,26) 0.74 (5.3,37) < 0.001 < 0.001 0.60b

level50%

3 (2,14) 0.31

28.6 (4.5,62) 6.1 (3.9,27) < 0.001 0.0013

Degrees of freedom given in brackets. aGreenhouse-Geisser corrected, PD*IPG*group was significant (F(14,100) = 1.8; P = 0.049).

PD 20 µs

PD 50 µs

slope (mV/nC)

amplitude (mV)

2.0 1.5 1.0

level50% (dB re 1 nC) threshold (dB re 1 nC)

0 26

dynamic range (dB)

0.5

22 18

PD 20 µs

interaction of

PD 50 µs

0.2 0.1 0 12 10 8 6 4 2

450

22 18

0.55 (2.9,20) 0.65

latency (µs)

0.3

1.6 (4.6,32) 0.20

400

NH 2WD 6WD

2.1 10 20 IPG (µs)

350

30 2.1 10 20 IPG (µs)

300

2.1 10 20 IPG (µs)

30 2.1 10 20 IPG (µs)

Fig. 5. eCAP characteristics averaged per group and plotted as function of IPG. PD is 20 µs in the first and third column, and 50 µs in the second and fourth. A, B amplitude; C, D slope; E, F threshold; G, H dynamic range; I, J level50%; K, L latency. IPG, inter-phase gap; PD, phase duration; NH, normal-hearing; 2WD, 2 weeks deaf; 6WD, 6 weeks deaf. For each plot N = 6 for NH and 6WD, N = 5 for 2WD; error bars represent SEM. 123

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

IPG 2.1 µs

IPG 30 µs

slope (mV/nC)

amplitude (mV)

2.0 1.5 1.0 0 26

dynamic range (dB)

0.5

level50% (dB re 1 nC) threshold (dB re 1 nC)

22 18 14 26

IPG 2.1 µs

IPG 30 µs

0.1 0 12 10 8 6 4 2

400

NH 2WD 6WD

0.2

450

0.3

latency (µs)

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 5

30 40 PD (µs)

350

50 20

30 40 PD (µs)

300

30 40 PD (µs)

50 20

30 40 PD (µs)

Fig. 6. eCAP characteristics averaged per group and plotted as function of PD. IPG is 2.1 µs in the first and third column, and 30 µs in the second and fourth. A, B amplitude; C, D slope; E, F threshold; G, H dynamic range; I, J level50%; K, L latency. IPG, inter-phase gap; PD, phase duration; NH, normal-hearing; 2WD, 2 weeks deaf; 6WD, 6 weeks deaf. For each plot N = 6 for NH and 6WD, N = 5 for 2WD; error bars represent SEM.

3.4. Correlation between changes in response to pulse shape alterations and histology The differences in eCAP characteristics (amplitude, slope, threshold, dynamic range, level50% and latency) that emerge when comparing responses to pulses with the shortest (2.1 µs) and the longest IPG (30 µs), as well as responses to the shortest (20 µs) and longest PD (50 µs), were subsequently put in a regression model with SGC packing density and perikaryal area (averaged over all cochlear turns) as predictor variables. Since interaction effects with group were exclusively found for changes in IPG (see Table 1), our emphasis is on changes in eCAP characteristics in response to IPG variation. Upon PD variation, SGC packing density and perikaryal area only significantly predicted a change in slope [R2 = 0.50, P = 0.008], a change in threshold [R2 = 0.40, P = 0.021] and a change in dynamic range [R2 = 0.43, P = 0.015] – all three only for an IPG of 2.1 µs. In Table 2 the R2 and P values derived from multiple regression analyses are given for changes in eCAP characteristics in response to increased IPG for the shortest and largest PD. The observation that either histological measure can show a significant partial contribution demonstrates that both play a considerable functional role. The 124

eCAP can predict neural health

corresponding scatter plots are shown in Fig. 7; changes in response to IPG increase are plotted against SGC packing density (on the horizontal axis), while the marker size is proportional to the perikaryal area. The four eCAP measures that showed a significant interaction effect with group (slope, dynamic range, level50% and latency) also showed significant correlations with SGC packing density and perikaryal area (Fig. 7C, D, G, I-L). In addition, the regression model significantly predicts threshold shifts for 20-μs PD (Fig. 7E). The absence of a significant interaction effect between IPG and group for eCAP amplitude (see Table 1) is paralleled by a lack of predictive value for SGC packing density or perikaryal area with respect to amplitude change with increasing IPG (Table 2; Fig. 7A,B). The latency shift after IPG increase significantly correlated with SGC degeneration for both PDs (Table 2, Fig. 7K, L). Interestingly, the latency shift we found in the 6WD animals (22 µs on average) roughly matched the increase in IPG duration (30.0 - 2.1 = 27.9 µs). Moreover, the extrapolated trend line for packing density predicts a latency shift of approximately 30 µs for both PDs for a hypothetical packing density of 0, while the latency shift is close to zero for high packing densities. Table 2. Multiple regression analyses. For each eCAP characteristic separately, and for short and long PD, the difference as a result of IPG increase from 2.1 to 30 µs is put in a multiple regression model with SGC perikaryal area and packing density as predictors. maximum amplitudea

slopeb

thresholda dynamic rangea level50%a latencya a

N = 18; bN = 17

partial contribution packing density perikaryal area P R2 P R2

total P

20 50

0.15 0.13

0.37 0.40

0.15 0.10

0.37 0.50

0.30 0.23

0.071 0.14

20 50

0.14 0.00

0.41 0.96

0.21 0.00

0.24 0.95

0.35 0.00

0.041 1.0

20 50 20 50 20 50 20 50

0.43 0.48 0.19 0.22 0.22 0.49 0.19 0.40

0.038 0.027 0.28 0.17

0.24 0.019 0.27 0.012

0.18 -0.03 0.30 -0.02 0.26 -0.04 0.38 0.41

0.29 0.85 0.12 0.90 0.17 0.74

0.052 0.011

0.61 0.46 0.49 0.21 0.47 0.45 0.57 0.80

0.0014 0.014 0.0066 0.17 0.008 0.012

0.0017 < 0.001

125

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

PD 50 µs

120

0.6 0.4 0.2 0 0 −2

R2= 0.35*

−4 −6

0 R2= 0.47** −1

−2 −3 −4

180

∆ slope (µV/nC)

PD 20 µs

R2= 0.45*

0 500 1500 0 500 1500 2 SGC packing density (cells/mm )

∆ dynamic range (dB)

0.8

PD 20 µs **

R = 0.61

PD 50 µs *

R = 0.46

60 0 8 6

NH 2WD 6WD

2 0 R2= 0.49**

−2

∆ latency (µs)

∆ amplitude (re max) ∆ threshold (dB)

1.0

∆ level50% (dB)

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 5

20 0

−20 R2= 0.57** −40

R2= 0.80***

0 500 1500 0 500 1500 2 SGC packing density (cells/mm )

Fig. 7. Differences in eCAP characteristics when IPG is increased from 2.1 to 30 µs plotted as function of numerical SGC packing density for 20-µs (first and third column) and 50-µs PD (second and fourth column). Symbols represent individual animals; symbol size is proportional to the average SGC perikaryal area for that animal. Both histological predictors are averages from the basal, middle and apical turns. Solid lines are regression lines with SGC packing density as single predictor, shown here for purely visual purposes. R2 values represent the amount of variation that can be explained by both SGC packing density and perikaryal area. The blue-crossed red symbol indicates the 2WD animal excluded from the group averages. PD, phase duration; SGC, spiral ganglion cell; *, P < 0.05; **, P < 0.01; ***, P < 0.001. N = 18 for A-B and E-L; N = 17 for C and D.

In Fig. 8 R2 values for the same multiple linear regression analyses are plotted for all seven PDs. Whereas the predictors in the above-mentioned regression models are histological measures obtained by averaging across all cochlear turns, in this figure the ability to predict changes in eCAP characteristics with increased IPG was assessed per cochlear turn (basal, middle, and apical). In general, differences among turns did not appear to vary substantially, suggesting that both histological measures correlated considerably over cochlear turns, and/or that neural elements from all turns contributed to the eCAP response. (Note that the electrode array was placed up to 5 mm from the round window, so that both the stimulation and the recording electrode were in the basal turn.) The results in Fig. 8 are similar to those shown in Fig. 7. The predictive power of SGC packing density and perikaryal area for the change in maximum amplitude with increased IPG remained mostly insignificant (Fig. 8A; filled circles indicate statistically

126

eCAP can predict neural health

significant values). For slope the regression model significantly predicted changes with IPG, although the R2 decreases slightly toward longer PD (Fig. 8B). The same gradient was observed for dynamic range (Fig. 8D) and, more distinctly, for threshold (Fig. 8C). The R2 values for the change in level50% (Fig. 8E) provide a different perspective on the data shown in Fig. 7I, J: the predictive power of the regression model is statistically significant only for the extreme PDs. For all PDs the latency shift brought about by increased IPG is for at least 50% explained by both histological predictors (Fig. 8F).

A 1.0

slope

threshold

dynamic range

level50%

latency

amplitude

basal middle apical

0.8 0.6 0.4 0.2 0 C 1.0 0.8 0.6 0.4 0.2 0 E 1.0 0.8 0.6 0.4 0.2 0

30 40 PD (Âľs)

Fig. 8. R2 values derived from multiple linear regression analyses with SGC packing density and perikaryal area as predictor variables (determined separately for each cochlear turn) and change with IPG increase as dependent variable. Filled circles represent statistically significant R2 values (P < 0.05). PD, phase duration.

Figure 9 shows that the high predictability for latency as shown in Fig. 8F is restricted to the three highest charge levels used, as the R2 rapidly drops to insignificant levels below 20 nC. This effect, illustrated here for 20, 35 and 50 Âľs, is present for all PDs applied.

127

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

1.0

PD 20 µs PD 35 µs PD 50 µs

0.8 R2

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 5

0.6 0.4 0.2 0

16 20 charge (nC)

Fig. 9. R2 values derived from multiple linear regression analyses with SGC packing density and perikaryal area as predictor variables and change of latency with IPG increase as dependent variable. R2 values are statistically significant (filled circles; P < 0.05) only for the higher stimulation levels. PD, phase duration.

4. Discussion

In this study we have evaluated the effect of PD and IPG on eCAP characteristics in normal-hearing and deafened guinea pigs. The effect of increasing PD in most cases roughly resembled increasing IPG: the maximum amplitude, slope, dynamic range and latency increased, while threshold and level50% decreased. A strong interaction effect for all six characteristics furthermore suggests that both act via a similar mechanism, since, in general, the effect of increasing IPG was smaller when PD was long, and vice versa – the effect of increasing PD was smaller when IPG was long (Figs. 5 and 6). Probably, the beneficial effect of an increase in either lies in the temporal separation of action potential initiation and hyperpolarization (van den Honert and Mortimer, 1979). However, in contrast to PD, increasing the IPG has a differential effect on various eCAP characteristics with respect to experimental groups, and the effect size often correlates well with histological measures for SGC degeneration. Based on these findings, variation of IPG, and not PD, in a clinical setting possibly makes a suitable diagnostic tool for the evaluation of the auditory nerve in CI users. 4.1. The effect of deafness on eCAPs The gradual loss of SGCs after ototoxic treatment was reflected by a gradual decrease in eCAP amplitude (Figs. 5A, B; 6A, B). With a decrease in amplitude, the slope of the input-output function became – fairly proportionally – lower as well. We did not observe significant changes in excitation threshold, although there was a small but 128

eCAP can predict neural health

consistent increase for the 2WD group followed by a decrease for the 6WD group for virtually all applied pulse shapes (Figs. 5E, F; 6E, F). A similar pattern was observed for the N1 latency, in which case group differences were statistically significant. Together, these results are in agreement with the theory that the site of excitation shifts to a more central location, and that loss of tissue may lead to lower impedances of the electrodetissue interface (Stypulkowski and van den Honert, 1984; Frijns et al., 1996; Shepherd and Javel, 1997). The stimulation level at which the half-maximum eCAP amplitude is reached (level50%) did not significantly change (Figs. 5I, J; 6I, J), although again there was a small but consistent increase after deafening. A wider dynamic range in deafened animals probably reflects the more heterogeneous disposition of the SGC population with respect to excitation thresholds, given that the dynamic range of single neurons appear to remain unaltered (Sly et al., 2007).

4.2. Phase duration With increasing PD, firing thresholds for single fibers (Shepherd and Javel, 1999), eABRs (Miller et al., 1995) and eCAPs (Prado-Guitierrez et al., 2006; this study) and psychophysical detection level thresholds (Miller et al., 1999) become lower when current level is kept constant. The main reason for this is that the amount of displaced charge (PD times current level) increases, which is why we chose to express stimulation level in term of charge rather than current. Accordingly, for 30-µs IPG the threshold expressed in charge was equal for all PDs (Figs. 4C, D; 6F). For short-IPG pulses the threshold became lower for longer PDs (Figs. 4A, B; 6E), suggesting more efficient recruitment with longer pulses. This latter finding seems to contradict single-fiber data from Shepherd and Javel (1999) and psychophysical data from Moon et al. (1993), who both found an increase in threshold with PD for PDs longer than 50 µs. Two different mechanisms with opposite effects but operating at different PD ranges may explain how our data are complementary to the literature data. On the one hand, in a pulse with short (<50 µs) PD the second – hyperpolarizing – phase may be so close to the onset of the first phase that it abolishes excitation initiation at low levels, thereby increasing threshold. Increasing PD to 50 µs (Fig. 6E) then reduces threshold by moving the onset of the hyperpolarizing phase further away from that of the depolarizing phase, as does increasing IPG (compare Figs. 6E and 6F). Indeed, by separating both phases with a 40-µs IPG, Moon et al. (1993) observed similar psychophysical detection thresholds for PDs shorter than 50 µs. On the other hand, a longer PD (>50 µs) may introduce significant influence of the leaky integrator, which means that more charge is needed to compensate membrane leakage over the length of the phase duration (Moon et al., 1993). In that respect a stronger “leaky” effect for the deafened groups might be expected, since the demyelination associated with SGC degeneration (Agterberg et al., 129

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 5

2008) causes leakier cell membranes (Koles and Rasminsky, 1972; Spoendlin, 1984). Instead, we did not find differences between healthy and degenerating SGC populations. This might be explained by the relatively short PD we used, which may have been too short for significant charge leakage to occur. Note that longer PDs would require an increase in blanking time, which would interfere with the N1 peak. Alternatively, the effect of demyelination on membrane leakage might not be that substantial.

4.3. Inter-phase gap The potential of an IPG to decrease excitation threshold has been demonstrated in various neural tissues and at various levels of neural processing (van den Honert and Mortimer, 1979; Shepherd and Javel, 1999; McKay and Henshall, 2003; Carlyon et al., 2005; Prado-Guitierrez et al., 2006; Cappaert et al., 2013). In the present study this potentiating effect of an IPG is confirmed unambiguously for threshold (Fig. 5E, F) and level50% (Fig. 5I, J). Interestingly, this effect is more pronounced for threshold than for level50% (Figs. 5E, F and 5I, J), which is consistent with findings on loudness perception in CI users: McKay and Henshall (2003) reported a larger effect of IPG on threshold than on the loudest comfortable level. In addition, as is shown in Table 1, the effect of IPG on amplitude, slope, dynamic range, and latency is highly significant as well. When comparing Figs. 5 and 6, it appears that IPG and PD have roughly similar effects on all six eCAP characteristics. However, the key difference between the two is that the effect size with a change in IPG, but not PD, highly depends on the physiological state of the cochlea. This is demonstrated by the significant interaction between IPG and group for slope, dynamic range, level50%, and latency and the lack of interactions between PD and group (Table 1). A smaller increase in slope (Fig. 5C, D) and a smaller decrease in level50% (Fig. 5I, J) indicate that the increased excitability brought about by an increased IPG is lower for the 6WD animals. More strikingly, the dynamic range remains mostly unchanged for the NH and 2WD groups, while it increases significantly for the 6WD animals (Fig. 5G, H). Probably the SGCs that are additionally recruited by the increased IPG have higher excitation thresholds, thereby enhancing the dynamic range. Given that this increase in dynamic range only occurs for deafened animals, these high individual thresholds may correspond to degenerating SGCs, or SGCs with a degenerated myelin sheath (Koles and Rasminsky, 1972; Spoendlin, 1984). Both cathodic and anodic currents can initiate action potentials in extracellular auditory nerve stimulation (reviewed in Macherey et al., 2008). Shepherd and Javel (1999) showed that when the second phase of a biphasic pulse is the excitatory one, the fiber response is delayed by the IPG duration as expected (see Fig. 10C, D). In the present study, since alternating stimulation was applied for artifact reduction, we therefore expected the latency to increase with half the increase in IPG duration (14 Âľs; 130

eCAP can predict neural health

see Fig. 10). Instead, for 50-µs PD, the increase was negligible for NH animals, 14 µs for 2WD, and 22 µs for 6WD animals (Fig. 7L). The pre-processed eCAPs in response to the cathodic-first and anodic-first pulses were often too much obscured by the stimulation artifact to reliably discern which pulse polarity contributed most to the averaged eCAP. However, based on several less-obscured examples from normal-hearing animals (i.e., with large eCAP amplitudes), we gathered that, in the normal-hearing group, cathodicfirst pulses contributed to a substantially greater extent. The observation that in this group the latency was not delayed in case of IPG increase implies that it was the first (i.e., cathodic) phase that most potently excited the SGCs (Fig. 10A, B). We hypothesize that the latency shift we observed with IPG increase for deafened animals, but not for the normal-hearing controls, may therefore have resulted from a shift in excitation phase preference from cathodic-first in NH to cathodic-second in 6WD animals: an increased IPG will not influence the eCAP latency when the majority of all responses is evoked by the cathodic phase of the cathodic-first pulse (Fig. 10A, B), while the latency shift will be more than half the IPG increase when the majority is evoked by the cathodic phase of the anodic-first pulse (Fig. 10C, D). Note that because we applied alternating polarity stimulation, this reasoning would pertain if the anodic rather than the cathodic phase (as is assumed here) were the most effective one. In a similar fashion, for the sake of simplicity, we here assume that also for deafened animals, as for the normalhearing animals, the cathodic phase is the most effective one. This assumption may be incorrect, as pointed out by a modeling study of Rattay et al. (2001) showing that SGC degeneration caused a change from lower thresholds for cathodic monophasic pulses to lower thresholds for anodic pulses. While the effect for biphasic pulses was much weaker, it would imply a shift from cathodic excitation by the cathodic-first pulse in NH animals (Fig. 10A, B) to anodic excitation by the same pulse in 6WD animals. The expected latency shifts with increasing IPG in this case would remain as observed. In case of the 20-µs PD (Fig. 7K) the hypothesis concerning the shift in excitation phase holds as well, with the exception that the latency decreased with increased IPG for NH animals (see negative latency shifts, as opposed to no shift, in Fig. 7K). Probably, a 20-µs cathodic phase almost immediately followed by the anodic hyperpolarizing phase is too short for the majority of SGCs to initiate a successful action potential, so that action potential initiation gradually shifts from the cathodic phase of the anodicfirst pulse to the cathodic phase of the cathodic-first pulse as IPG increases, resulting in a shorter latency. In addition to explaining the latency shift, this hypothesis can explain the limited potential of an increased IPG to enhance excitability (increase in slope and decrease in level50%; Figs. 5C, D and 5I, J) in the 6WD animals relative to the NH controls. If indeed the majority of SGCs in NH animals responds to the cathodic phase of the cathodic-first pulse 131

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 5

rather than to that of the anodic-first pulse, these SGCs may benefit substantially from the delay of the second (anodic) phase, which otherwise can annihilate the successful initiation of action potentials (van den Honert and Mortimer, 1979). If, then, in deafened animals the cathodic phase of the anodic-first pulse is the preferred exciter, the function of the IPG is fundamentally different as it separates the excitatory phase from the preceding hyperpolarizing phase rather than from the subsequent hyperpolarization. In the latter case, the IPG is arguably of less importance to action potential initiation. This hypothesized shift in excitation phase is partly based on visual inspection of recordings to individual polarities, and largely inferred from the objective analysis of averaged data. For a more direct approach to test this hypothesis a masker-probe paradigm for artifact reduction might have been more suited, since artifact-free recordings to both individual stimulus polarities would then have been available. We chose alternating stimulation, since a larger portion of the SGC population is thought to be recruited than with a masker-probe paradigm (Westen et al., 2011). excitating phase first

IPG 2.1 µs

IPG 30 µs

excited by first phase: no shift (NH)

exciting phase second C

excited by second phase: ~28 µs shift (6WD)

Fig. 10. The effect of IPG increase in case of alternating stimulation. For cathodic-first stimulus polarity (A and B) a latency shift is not expected, since the excitatory phase (shaded area) remains unchanged. For anodic-first pulses (C and D) the latency shift relative to the stimulus onset is proportional to the IPG duration. In case of equal contribution of both pulse polarities a latency shift of 14 µs is expected. The black bar indicates the time between stimulus onset and termination of the excitatory phase.

4.4. Correlations with histology Both the maximum eCAP amplitude and the slope of the input-output function correlated significantly with SGC packing density. This is in agreement with findings by Hall (1990), who found a correlation between the amplitude of the eABR P1 peak and the number of spiral ganglion cells in rats. Although maximum amplitude and slope may be reliable and straightforward measures for the estimation of the number of SGCs in animals, they have restrictions when translating to clinical practice. First, 132

eCAP can predict neural health

factors such as the position of the electrode array (angle, depth, location of insertion), possible fibrosis or ossification, and a wide range of different available electrode arrays affect the eCAP amplitude and thereby also the slope. Second, the current range in CI users is much more limited than in (anesthetized) laboratory animals, so that the maximum eCAP amplitude often cannot be attained. The eCAP slope as we have defined it is the steepest slope in the input-output curve (at level50%, see Fig. 1D), and can only be reasonably estimated when current levels above level50% have been applied, which depends on maximum current level. The alternative in clinical settings is to apply linear fits through all supra-threshold data points (Kim et al., 2010), which, taking into account the non-linear nature of the input-output function, makes the slope dependent of the maximum current level. Hence, we have explored the possibility to use more indirect within-subject differential eCAP measures that therefore inherently suffer less from the above-mentioned limitations to predict the neuronal status. However, most of the relative measures still depend on the use of high current levels (second limitation), and it cannot be ruled out that any of the factors mentioned under the first limitation to some extent still influence the effect size of pulse shape variation. We did not find any significant interaction effects between PD and group (Table 1), and accordingly we found hardly any predictive power for either SGC packing density or perikaryal area for changes in eCAP characteristics with changed PD. Specifically, in contrast to Prado-Guitierrez et al. (2006), we did not find a significant correlation between a change in level50% with increasing PD and SGC packing density. The discrepancy with their findings most likely resulted from a difference in PD range (20-50 Âľs versus 104-208 Âľs), but could also be due to eCAP versus eABR recordings, or to monopolar versus bipolar stimulation. For PDs beyond the range we used, the necessary blanking time would obscure the N1 peak. This limitation can be circumvented by recording eABRs instead of eCAPs, like Prado-Guitierrez et al. did, but for eCAP recordings the PD range cannot be widened substantially, which so much as disqualifies PD as a suitable predictor for neural health using currently available CI telemetry. In contrast, the increase in IPG (from 2.1 to 30 Âľs) did result in a series of significant correlations (Fig. 7). Specifically, the change in level50% (Fig. 7I, J) is highly similar to the results Prado-Guitierrez et al. obtained with IPG variation: the slopes of the trend lines are 0.8 and 1.0 dB per 1000 cells/mm2, respectively. Whereas Prado-Guitierrez et al. (2006) used only SGC packing density as histological measure to relate to their electrophysiological measures, we added perikaryal area as second histological measure to cover both the degree of SGC loss and the state of the surviving SGCs. The two measures are obviously related, although, as is shown in Fig. 3C, variability within groups is substantial. Moreover, it is hypothesized that SGC loss and shrinkage may result from separate subcellular mechanisms (van Loon et al., 2013). 133

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 5

SGC soma size affects functionality directly since it is highly (positively) correlated with membrane capacitance (Limón et al., 2005). Furthermore, a decrease in cell size may result from intracellular changes due to lost neurotrophic support (van Loon et al., 2013), which could cause a change in electrical activity as well. The possible effect of SGC packing density on SGC functionality is less straightforward. Less densely packed cells could result in a change in the current path from electrode to the neural substrate, resulting in lower thresholds and shorter latency as discussed above.

4.5. Considerations for clinical application The present study has demonstrated that, by increasing the stimulus’s IPG, it is possible to predict the magnitude of a change in eCAP response for up to 80% with histological measures for SGC degeneration (Fig. 8). Thus, the magnitude of an IPG-induced change in eCAP morphology is a fairly accurate predictor for neural health. By means of the CI’s telemetry function in human CI users, eCAPs can be evoked and recorded using the same parameters as used in this study. Although there are some critical differences between animal and human cochleas – e.g., with respect to myelination, anatomy and excitation (Frijns et al., 2001; Abbas and Miller, 2004) – this method may prove to be suitable for clinical purposes. During cochlear implantation, intra-operatively acquired knowledge of the state of the auditory nerve may provide a choice to apply personalized supplementary neuroprotective treatment. Furthermore, these measures may have instant prognostic value for CI performance, and the state of the auditory nerve can be monitored over time. Since this method is independent of the highly variable absolute amplitude, it can easily be used for comparison between subjects. In this light, it should be noted that in human CI users a correlation between a change in eCAP sensitivity (level50%) and speech perception has not been found (Kim et al., 2010). However, we have shown that this measure only marginally correlates with neural health, and that the predictive power furthermore depends on the PD that is used (Fig. 8E). Based on the present findings, most likely candidates would be slope, dynamic range and latency. Of these candidates, latency not only has the most predictive power (Fig. 8F), it is also the only characteristic for which it is sufficient to use only one current level, instead of having to construct an input-output curve. This will considerably reduce the duration of the measurement session, thereby making it more suitable for clinical purposes than the other characteristics. However, it is important to note that the predictive power of latency exists only for near-saturation stimulus levels (above 20 nC; see Fig. 9). This is not a result of less accurate N1 peak assessment – which might be expected at lower stimulus levels – since the within-animal variance of N1 peaks is similar for all stimulus levels above 12 nC (data not shown). Possibly, at medium stimulus levels the preference for pulse polarity or even for excitation phase 134

eCAP can predict neural health

polarity is less pronounced than at near-saturation levels. If this is the case, the IPG effect on latency obviously becomes less straightforward. If this restriction for higher-level stimulation would be the case in human CI users as well, for all eCAP characteristics used in the present study high-level stimulation would be required for the prediction of neural status, which could be a considerable limitation to clinical applicability. Instead of using a single eCAP characteristic, employing a combination of all three identified candidates may therefore be crucial to develop a reliable objective measure for auditory nerve status in CI users.

Acknowledgments The authors would like to thank Ferry Hendriksen for histological processing, RenĂŠ van de Vosse for technical support, and Roland Hessler at MED-EL, Innsbruck, for the electrode arrays. This work was supported by MED-EL GmbH, Innsbruck, Austria.

135

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 5

References

Abbas, P.J., Miller, C.A., 2004. Biophysics and Physiology. In: Zeng F-G, Popper AN and Fay RR (eds) Cochlear Implants: Auditory Prostheses and Electric Hearing. Springer Handbook of Auditory Research. Springer-Verlag, New York, pp 149-212. Agterberg, M.J.H., Versnel, H., De Groot, J.C.M.J., Smoorenburg, G.F., Albers, F.W.J., Klis, S.F.L., 2008. Morphological changes in spiral ganglion cells after intracochlear application of brainderived neurotrophic factor in deafened guinea pigs. Hear. Res. 244, 25-34. Agterberg, M.J.H., Versnel, H., Van Dijk, L.M., De Groot, J.C.M.J., Klis, S.F.L., 2009. Enhanced survival of spiral ganglion cells after cessation of treatment with brain-derived neurotrophic factor in deafened guinea pigs. J. Assoc. Res. Otolaryngol. 10, 355-367. Cappaert, N.L.M., Ramekers, D., Martens, H.C.F., Wadman, W.J., 2013. Efficacy of a new chargebalanced biphasic electrical stimulus in the isolated sciatic nerve and the hippocampal slice. Int. J. Neural Syst. 23, 1250031. Carlyon, R.P., van Wieringen, A., Deeks, J.M., Long, C.J., Lyzenga, J., Wouters, J., 2005. Effect of interphase gap on the sensitivity of cochlear implant users to electrical stimulation. Hear. Res. 205, 210-224. Coggeshall, R.E., Lekan, H.A., 1996. Methods for determining numbers of cells and synapses: a case for more uniform standards of review. J. Comp. Neurol. 364, 6-15. Ernfors, P., Duan, M.L., ElShamy, W.M., Canlon, B., 1996. Protection of auditory neurons from aminoglycoside toxicity by neurotrophin-3. Nat. Med. 2, 463-467. Ernfors, P., Van De Water, T., Loring, J., Jaenisch, R., 1995. Complementary roles of BDNF and NT-3 in vestibular and auditory development. Neuron 14, 1153-1164. Fayad, J.N., Linthicum, Jr., F.H., 2006. Multichannel cochlear implants: relation of histopathology to performance. Laryngoscope 116, 1310-1320. Fransson, A., Maruyama, J., Miller, J.M., Ulfendahl, M., 2010. Post-treatment effects of local GDNF administration to the inner ears of deafened guinea pigs. J. Neurotrauma 27, 1745-1751. Frijns, J.H.M., Briaire, J.J., Grote, J.J., 2001. The importance of human cochlear anatomy for the results of modiolus-hugging multichannel cochlear implants. Otol. Neurotol. 22, 340-349. Frijns, J.H.M., de Snoo, S.L., ten Kate, J.H., 1996. Spatial selectivity in a rotationally symmetrical model of the electrically stimulated cochlea. Hear. Res. 95, 33-48. Fritzsch, B., Pirvola, U., Ylikoski, J., 1999. Making and breaking the innervation of the ear: neurotrophic support during ear development and its clinical implications. Cell Tissue Res. 295, 369-382. Glueckert, R., Bitsche, M., Miller, J.M., Zhu, Y., Prieskorn, D.M., Altschuler, R.A., Schrott-Fischer, A., 2008. Deafferentation-associated changes in afferent and efferent processes in the guinea pig cochlea and afferent regeneration with chronic intrascalar brain-derived neurotrophic factor and acidic fibroblast growth factor. J. Comp. Neurol. 507, 1602-1621. Grill, W.M., Norman, S.E., Bellamkonda, R.V., 2009. Implanted neural interfaces: biochallenges and engineered solutions. Annu. Rev. Biomed. Eng. 11, 1-24. Hall, R.D., 1990. Estimation of surviving spiral ganglion cells in the deaf rat using the electrically evoked auditory brainstem response. Hear. Res. 49, 155-168. Kim, J.R., Abbas, P.J., Brown, C.J., Etler, C.P., Oâ&#x20AC;&#x2122;Brien, S., Kim, L.S., 2010. The relationship between electrically evoked compound action potential and speech perception: a study in cochlear implant users with short electrode array. Otol. Neurotol. 31, 1041-1048. Koles, Z.J., Rasminsky, M., 1972. A computer simulation of conduction in demyelinated nerve fibres. J. Physiol. 227, 351-364. 136

eCAP can predict neural health

Leake, P.A., Hradek, G.T., Hetherington, A.M., Stakhovskaya, O., 2011. Brain-derived neurotrophic factor promotes cochlear spiral ganglion cell survival and function in deafened, developing cats. J. Comp. Neurol. 519, 1526-1545. Limón, A., Pérez, C., Vega, R., Soto, E., 2005. Ca2+-activated K+-current density is correlated with soma size in rat vestibular-afferent neurons in culture. J. Neurophysiol. 94, 3751-3761. Macherey, O., Carlyon, R.P., van Wieringen, A., Deeks, J.M., Wouters, J., 2008. Higher sensitivity of human auditory nerve fibers to positive electrical currents. J. Assoc. Res. Otolaryngol. 9, 241-251. Marquardt, D.W., 1963. An Algorithm for Least-Squares Estimation of Nonlinear Parameters. J. Soc. Ind. Appl. Math. 11, 431-441. Maruyama, J., Miller, J.M., Ulfendahl, M., 2008. Glial cell line-derived neurotrophic factor and antioxidants preserve the electrical responsiveness of the spiral ganglion neurons after experimentally induced deafness. Neurobiol. Dis. 29, 14-21. McKay, C.M., Henshall, K.R., 2003. The perceptual effects of interphase gap duration in cochlear implant stimulation. Hear. Res. 181, 94-99. Miller, A.L., Smith, D.W., Pfingst, B.E., 1999. Across-species comparisons of psychophysical detection thresholds for electrical stimulation of the cochlea: II. Strength-duration functions for single, biphasic pulses. Hear. Res. 135, 47-55. Miller, C.A., Abbas, P.J., Rubinstein, J.T., Robinson, B.K., Matsuoka, A.J., Woodworth, G., 1998. Electrically evoked compound action potentials of guinea pig and cat: responses to monopolar, monophasic stimulation. Hear. Res. 119, 142-154. Miller, C.A., Woodruff, K.E., Pfingst, B.E., 1995. Functional responses from guinea pigs with cochlear implants. I. Electrophysiological and psychophysical measures. Hear. Res. 92, 85-99. Moon, A.K., Zwolan, T.A., Pfingst, B.E., 1993. Effects of phase duration on detection of electrical stimulation of the human cochlea. Hear. Res. 67, 166-178. Neustetter, C., Zangerl, M., Spitzer, P., Zierhofer, C., 2012. In-vitro characterization of a cochlear implant system for recording of evoked compound action potentials. Biomed. Eng. Online 11,22. Prado-Guitierrez, P., Fewster, L.M., Heasman, J.M., McKay, C.M., Shepherd, R.K., 2006. Effect of interphase gap and pulse duration on electrically evoked potentials is correlated with auditory nerve survival. Hear. Res. 215, 47-55. Ramekers, D., Versnel, H., Grolman, W., Klis, S.F.L., 2012. Neurotrophins and their role in the cochlea. Hear. Res. 288, 19-33. Rattay, F., Lutter, P., Felix, H., 2001. A model of the electrically excited human cochlear neuron. I. Contribution of neural substructures to the generation and propagation of spikes. Hear. Res. 153, 43-63. Richardson, R.T., O’Leary, S., Wise, A., Hardman, J., Clark, G., 2005. A single dose of neurotrophin-3 to the cochlea surrounds spiral ganglion neurons and provides trophic support. Hear. Res. 204, 37-47. Shepherd, R.K., Coco, A., Epp, S.B., Crook, J.M., 2005. Chronic depolarization enhances the trophic effects of brain-derived neurotrophic factor in rescuing auditory neurons following a sensorineural hearing loss. J. Comp. Neurol. 486, 145-158. Shepherd, R.K., Javel, E., 1997. Electrical stimulation of the auditory nerve. I. Correlation of physiological responses with cochlear status. Hear. Res. 108, 112-144. Shepherd, R.K., Javel, E., 1999. Electrical stimulation of the auditory nerve: II. Effect of stimulus waveshape on single fibre response properties. Hear. Res. 130, 171-188. Sly, D.J., Heffer, L.F., White, M.W., Shepherd, R.K., Birch, M.G., Minter, R.L., Nelson, N.E., Wise, A.K., O’Leary, S.J., 2007. Deafness alters auditory nerve fibre responses to cochlear implant stimulation. Eur. J. Neurosci. 26, 510-522. 137

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 5

Spoendlin, H., 1975. Retrograde degeneration of the cochlear nerve. Acta Otolaryngol. 79, 266â&#x20AC;&#x201C; 275. Spoendlin, H., 1984. Factors inducing retrograde degeneration of the cochlear nerve. Ann. Otol. Rhinol. Laryngol. 93 (Suppl. 112), 76-82. Stypulkowski, P.H., van den Honert, C., 1984. Physiological properties of the electrically stimulated auditory nerve. I. Compound action potential recordings. Hear. Res. 14, 205-223. Van den Honert, C., Mortimer, J.T., 1979. The response of the myelinated nerve fiber to short duration biphasic stimulating currents. Ann. Biomed. Eng. 7, 117-125. Van Loon, M.C., Ramekers, D., Agterberg, M.J.H., De Groot, J.C.M.J., Grolman, W., Klis, S.F.L., Versnel, H., 2013. Spiral ganglion cell morphology in guinea pigs after deafening and neurotrophic treatment. Hear. Res. 298, 17-26. Versnel, H., Agterberg, M.J.H., De Groot, J.C.M.J., Smoorenburg, G.F., Klis, S.F.L., 2007. Time course of cochlear electrophysiology and morphology after combined administration of kanamycin and furosemide. Hear. Res. 231, 1-12. Webster, M., Webster, D.B., 1981. Spiral ganglion neuron loss following organ of Corti loss: a quantitative study. Brain Res. 212, 17-30. West, B.A., Brummett, R.E., Himes, D.L., 1973. Interaction of kanamycin and ethacrynic acid. Severe cochlear damage in guinea pigs. Arch. Otolaryngol. 98, 32â&#x20AC;&#x201C;37. Westen, A.A., Dekker, D.M., Briaire, J.J., Frijns, J.H.M., 2011. Stimulus level effects on neural excitation and eCAP amplitude. Hear. Res. 280, 166-176. Xu, H.X., Kim, G.H., Snissarenko, E.P., Cureoglu, S., Paparella, M.M., 2012. Multi-channel cochlear implant histopathology: are fewer spiral ganglion cells really related to better clinical performance? Acta Otolaryngol. 132, 482-490. Ylikoski, J., Wersall, J., Bjorkroth, B., 1974. Degeneration of neural elements in the cochlea of the guinea pig after damage to the organ of corti by ototoxic antibiotics. Acta Otolaryngol. Suppl. 326, 23-41. Zilberstein, Y., Liberman, M.C., Corfas, G., 2012. Inner hair cells are not required for survival of spiral ganglion neurons in the adult cochlea. J. Neurosci. 32, 405-410.

138

CHAPTER 6 Recovery characteristics of the electrically stimulated auditory nerve in deafened guinea pigs: relation to neuronal status

Dyan Ramekers, Huib Versnel, Stefan B. Strahl, Sjaak F.L. Klis, Wilko Grolman

My ear barely caught signals coming in regular succession which could not have been produced on earth. â&#x20AC;&#x201C; Nikola Tesla, 1919

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 6

Abstract

Successful cochlear implant performance requires adequate responsiveness of the auditory nerve to prolonged pulsatile electrical stimulation. Degeneration of the auditory nerve as a result of severe hair cell loss could considerably compromise this ability. The main objective of this study was to characterize the recovery of the electrically stimulated auditory nerve, as well as to evaluate possible changes caused by deafnessinduced degeneration. To this end we studied the electrically evoked compound action potential (eCAP) using two-pulse masker-probe and pulse train paradigms in a guinea pig model of sensorineural hearing loss. Following ototoxically induced hair cell loss, spiral ganglion cells (SGCs) substantially and progressively degenerated as observed after two and six weeks; SGC peripheral processes degenerated at a highly comparable rate. Masker-probe interval and pulse train inter-pulse interval was varied from 0.4 to 16 ms. Whereas recovery assessed with masker-probe was roughly similar for normalhearing and deafened animals, it was considerably faster for six weeks deaf animals (τ ≈ 1.2 ms) than for two weeks deaf or normal-hearing animals (τ ≈ 3-4 ms) when 100-ms pulse trains were applied. With pulse train stimulation eCAP amplitudes were modulated with pulse number. Latency increased with decreasing inter-pulse intervals, and this was more pronounced with pulse trains than with masker-probe stimulation. The relative refractory period (τ) and amplitude modulation depth for pulse trains, as well as the latency increase for both paradigms significantly correlated with quantified measures of auditory nerve degeneration (size and packing density of SGCs). In addition to these findings, separate masker-probe recovery functions for the eCAP N1 and N2 peaks displayed a robust non-monotonic course for both peaks in all animals. The time interval between the N1 and N2 correlated with neuronal refractoriness, suggesting that the N2 peak reflects a second firing of part of the SGC population. We conclude that – compared to the commonly used masker-probe recovery functions – recovery functions obtained with pulse train stimulation may provide a means to augment differences and, by doing so, to more potently discriminate between auditory nerve conditions.

140

Recovery characteristics of the auditory nerve

1. Introduction Sensorineural hearing loss (SNHL) is commonly characterized by loss of cochlear hair cells. Severe SNHL can be treated with cochlear implantation. Cochlear implants (CIs) convert sound into electrical pulses that are conveyed to the spiral ganglion cells (SGCs), which constitute the auditory nerve. Animal models consistently show that these SGCs degenerate shortly after SNHL induction (Ylikoski et al., 1974; Spoendlin, 1975; Webster and Webster, 1981; Versnel et al., 2007; Ramekers et al., 2014). This secondary SGC degeneration has been shown to occur in human CI users as well (Fayad and Linthicum, 2006), and it has recently been reported that word recognition is positively correlated to SGC count (Seyyedi et al., 2014). An important tool to study the functional condition of the auditory nerve is the CIâ&#x20AC;&#x2122;s electrically evoked compound action potential (eCAP) recording function, with which the nerveâ&#x20AC;&#x2122;s input to the central auditory system can be recorded. Comparison of eCAP characteristics with psychophysical measures in humans (e.g., Kirby et al., 2012; Carlyon and Deeks, 2013) or measures for speech perception (e.g., Kim et al., 2010) can be used to assess the relation between eCAP characteristics and CI performance. Animal research is complementary to these clinical techniques; the comparison of eCAP characteristics with SGC histology in animals provides an opportunity to indirectly relate human capability for speech understanding with a CI to detailed histology of the auditory nerve. For instance, introducing an inter-phase gap into a biphasic current pulse leads to a decrease in both animal eCAP thresholds (Prado-Guitierrez et al., 2006; Ramekers et al., 2014) as well as human behavioral thresholds (Carlyon et al., 2005). Since on the eCAP level the effect of inter-phase gap varies with the degree of SGC degeneration (Prado-Guitierrez et al., 2006; Ramekers et al., 2014), an inter-phase gap effect on behavioral thresholds in human CI users might show similar variation and would then provide a link between the condition of the auditory nerve and sound perception. In the present study we aimed to provide similar objective tools to relate eCAP recovery characteristics to quantified histological measures for SGC degeneration. Since it is crucial that the SGC population is able to follow high pulse rate stimulation for proper CI use, focus was on the temporal response properties of the electrically stimulated auditory nerve, and, in particular, on changes in these properties that are associated with SGC degeneration as a result of hair cell loss. To this end we first recorded eCAPs using masker-probe paradigms to characterize in detail the refractory properties of healthy and degenerating SGCs, which we compared to similar studies in both animals (e.g., Stypulkowski and van den Honert, 1984) and humans (e.g., Botros and Psarros, 2010). Second, to mimic more realistic situations we examined neural 141

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 6

recovery using pulse train paradigms, in which firing synchrony and neural fatigue play a prominent role.

2. Methods

2.1. Animals and surgery The data presented in this report form a subset of data obtained from a series of acute experiments. During these experiments a second set of eCAP data was acquired, which has been published previously (Ramekers et al., 2014). Hence, for details about animal care, surgical procedures and recording techniques we refer to the methods section of that paper. In short, eighteen female albino guinea pigs (Dunkin Hartley; 250–350 g; Harlan Laboratories, Horst, the Netherlands) were divided into three groups. One group of six animals functioned as normal-hearing controls (NH), while the other two were deafened either two (2WD) or six weeks (6WD) before the eCAP recordings, by means of coadministration of kanamycin (Sigma-Aldrich, St. Louis, MO, USA; 400 mg/kg s.c.) and the loop diuretic furosemide (Centrafarm, Etten-Leur, the Netherlands; 100 mg/ kg i.v.). Before the deafening procedure, and during the acute experiments, the hearing threshold was determined with recording of click-evoked auditory brainstem responses (ABRs), so that threshold shifts caused by the ototoxic treatment could be established. During the acute experiments the animals were anesthetized with a gas mixture of O2 and N2O and 1-2% isoflurane. The right cochlea was exposed via a retro-auricular approach, and a 0.5 mm cochleostomy was drilled in the basal turn, through which a four-contact electrode array (MED-EL GmbH, Innsbruck, Austria) was inserted into the scala tympani. Transcranial screws were placed on the skull for stimulation and recording reference purposes. All surgical and experimental procedures were approved by the Animal Care and Use Committee of Utrecht University (DEC 2010.I.08.103).

2.2. Electrically evoked compound action potentials eCAPs were recorded using a MED-EL PULSARci100 stimulator (MED-EL GmbH, Innsbruck, Austria). The electrode array was inserted up to 4 mm into the cochleostomy, so that all 4 contacts – evenly spaced over the first 3 mm of the array – were in the scala tympani. For all data presented here, the most apical contact was used for stimulation and the most basal one for recording. The implant was controlled by a PC via a Research Interface Box 2 (RIB2; Department of Ion Physics and Applied Physics, University of Innsbruck, Innsbruck, Austria) and a National Instruments data acquisition card (PCI6533, National Instruments, Austin, TX, USA). Stimulation and recording paradigms were created in MATLAB (version 7.11.0; Mathworks, Natick, MA, USA). 142

Recovery characteristics of the auditory nerve

2.3. Recording and stimulation paradigm Two different stimulus paradigms were used: masker-probe and pulse train. The paradigms are illustrated in Fig. 1. Both pulse train and masker-probe stimulation consisted of monopolar biphasic current pulses with 30 µs phase duration and 30 µs inter-phase gap. Current level was varied from 0 to 800 current units (1 current unit ≈ 1 µA) in ten steps. Alternating polarity stimulation was applied for stimulation artifact reduction (see Fig. 1A, B), while recordings to a subthreshold stimulus were subtracted to eliminate measurement onset artifacts. Every stimulation step was repeated 50 times, and the N1-P2 and P2-N2-P3 peak amplitudes were calculated from these averages (see Fig. 1A for definition of peaks). The N1 latency was determined by averaging the latencies of the N1 peaks obtained with the three highest stimulation levels (640, 720 and 800 µA). The masker-probe interval (MPI) was varied from 0.3 to 16 ms in 18 steps, while the inter-pulse interval (IPI) ranged from 0.4 to 16 ms in 10 steps for pulse trains. Pulse train duration was as close to 100 ms as possible, taking into account variable IPI duration. For both stimulation paradigms, MPIs/IPIs were presented in decreasing order, while the order of current levels was permuted. As the implant did not allow continuous recording, the stimulation paradigm was designed such that the eCAPs in response to both the masker and the probe (maskerprobe stimulation) were recorded in separate (consecutive) runs. Similarly, the responses to the last ten pulses in the pulse train were recorded individually (see Fig. 1C). Furthermore, in case of small IPIs/MPIs (≤1.2 ms) the response to the previous pulse was present in the recording, which was therefore recorded separately (if not already part of the paradigm) and subtracted from the initial recording. In Fig. 2 examples of eCAP recordings from a masker-probe paradigm (A, B) and from a pulse train paradigm (C-E) are shown.

143

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 6

MPI

A cathodic first

MPI

cathodic first anodic first anodic first

↓↓ P2 P3 ↑ ↓N↓ 2 N ↑ 1

IPI

B cathodic first

IPI

↓

anodic first

IPI N

↓

cathodic first

IPI

N N−1 N−1 N−2 N−2

6 N−9 N−9

Fig. 1. Stimulation paradigm for masker-probe (A) and pulse train stimulation (B). Stimulus polarity was alternated for artefact reduction purposes. C In order to record the responses to the last ten pulses of a pulse train, a series of recordings (N through N-9) was used. MPI, maskerprobe interval; IPI, inter-pulse interval.

144

Recovery characteristics of the auditory nerve

masker

probe

MPI (ms) 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.2 1.4 1.6 1.8 2 3 4 8 12 16

signal amplitude (mV)

5 4 3 2 1 0 −1 −2

0.4

0.8

1.2

1.6

IPI 8 ms

0.4

0.8

1.2

signal amplitude (mV)

1.6

IPI 0.4 ms

IPI (ms)

0.4

N−9

0.6

N−8

0.8

N−7

1.0

N−6

1.2

N−5

1.6

N−4

N−3

N−2

N−1

0 0.4

0.8

1.2

1.6

0.4

0.8

1.2

1.6

time since stimulus onset (ms)

0.4

0.8

1.2

1.6

Fig. 2. Examples of eCAP recordings at 800 µA current level. Masker and probe eCAPs were recorded in separate runs. A The 18 traces to the masker are effectively responses to singlepulse stimuli – their similarity demonstrates the reproducibility of the eCAP. These responses were used for normalization of the probe-evoked eCAPs. B eCAPs recorded after the probe pulse gradually decrease in amplitude as the MPI becomes shorter, and are absent for the shortest MPI. C eCAPs recorded after the last ten pulses of a 100-ms pulse train consisting of N pulses with 8 ms IPI, and D 0.4 ms IPI. E eCAPs recorded after the last (Nth) pulse of pulse trains with different IPIs. MPI, masker-probe interval; IPI, inter-pulse interval. 145

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

2.4. Data analysis 2.4.1. Amplitude modulation The eCAP N1 peak amplitude was defined by the voltage difference between the N1 and P2 peaks (see Fig. 1). Figure 3 shows examples of N1 amplitudes for the last ten pulses of a pulse train for different IPIs. The oscillating pattern that emerges with higher pulse rate (e.g., for 0.6 ms IPI in Fig. 3C) was evaluated with Fourier analysis, yielding the mean amplitude averaged over all ten pulses (DC component; used for eCAP recovery analysis), and a measure for the eCAP amplitude modulation depth (the amplitude of the most dominant frequency). 1.2

A − NH

IPI (ms)

1.0

16 8.0 4.0 2.0 1.6 0.8 0.6 0.4

0.8 0.6 0.4 0.2 0

normalized eCAP amplitude

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 6

1.2

B − 2WD

1.0 0.8 0.6 0.4 0.2 0 1.2

C − 6WD

1.0 0.8 0.6 0.4 0.2 0

N−9 N−8 N−7 N−6 N−5 N−4 N−3 N−2 N−1

pulse number

Fig. 3. eCAP N1-P2 amplitudes plotted for the last 10 pulses of pulse trains with different IPIs and 800 µA current level for (A) a normal-hearing (NH) animal, (B) a two-weeks deaf (2WD) animal, and (C) a six-weeks deaf animal (6WD). IPI, inter-pulse interval. 146

Recovery characteristics of the auditory nerve

2.4.2. N6, assessed pter 5, Eq. 3 voor chapter Eq. 1recovery voor chapter 7: with masker-probe and pulse train paradigms 1 peak

All amplitude data were normalized to the amplitude obtained with a single pulse đ??ľđ??ľđ??ľđ??ľ (i.e., the masker paradigm) at 800 ÂľA, for each animal separately. đ?&#x2018;&#x2030;đ?&#x2018;&#x2030;đ?&#x2018;&#x2030;đ?&#x2018;&#x2030;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; = đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; in + the masker-probe đ??źđ??źđ??źđ??źâ&#x2C6;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2019; Figure 3 shows examples 1 â&#x2C6;&#x2019; đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; ofđ??ˇđ??ˇđ??ˇđ??ˇnormalized amplitudes for the last ten pulses of the pulse pter 5, Eq. 3 voor chapter 6, Eq.removal 1 voor chapter 7: train. After of subthreshold amplitudes, recovery functions were fitted to the normalized data. đ??ľđ??ľđ??ľđ??ľ eCAP described with an exponential function (e.g., đ?&#x2018;&#x2030;đ?&#x2018;&#x2030;đ?&#x2018;&#x2030;đ?&#x2018;&#x2030;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; recovery = đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; + is commonly đ??źđ??źđ??źđ??źâ&#x2C6;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2019; pter 6, Eq. 2 voor chapter 7: đ??ˇđ??ˇđ??ˇđ??ˇ Morsnowski et al., 2006; 1 â&#x2C6;&#x2019; đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;Botros and Psarros, 2010), as depicted here: đ??źđ??źđ??źđ??źđ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ??źđ??źđ??źđ??ź â&#x2C6;&#x2019;đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ą 0 đ?&#x153;?đ?&#x153;?đ?&#x153;?đ?&#x153;?

đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; = đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2122; ďż˝1 â&#x2C6;&#x2019; đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2019;

pter 6, Eq. 2 voor chapter 7:

ďż˝ ,

(Eq. 1)

where eCAP is the normalized eCAP amplitude for a given inter-pulse interval IPI, A đ??źđ??źđ??źđ??źđ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ??źđ??źđ??źđ??ź â&#x2C6;&#x2019;đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ą 0 is the maximum eCAP amplitude evoked by a probe pulse after a sufficiently long IPI â&#x2C6;&#x2019; đ?&#x153;?đ?&#x153;?đ?&#x153;?đ?&#x153;? đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; = đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2122; â&#x2C6;&#x2019; đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; ďż˝1 ďż˝ pter 6, Eq. 3 voor chapter 7: (comparable to a single-pulse-evoked eCAP), Ď&#x201E; is the recovery time constant, and t0 is the absolute refractory period. đ?&#x2018;&#x20AC;đ?&#x2018;&#x20AC;đ?&#x2018;&#x20AC;đ?&#x2018;&#x20AC;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2019;đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ą 0 đ?&#x2018;&#x20AC;đ?&#x2018;&#x20AC;đ?&#x2018;&#x20AC;đ?&#x2018;&#x20AC;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2019;đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ą 0 â&#x2C6;&#x2019; â&#x2C6;&#x2019; Since from this conventional đ?&#x153;?đ?&#x153;?đ?&#x153;?đ?&#x153;? đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2019; đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; đ?&#x153;?đ?&#x153;?đ?&#x153;?đ?&#x153;? đ??ľđ??ľđ??ľđ??ľ deviated đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;our = đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;masker-probe â&#x2C6;&#x2122; ďż˝1 + đ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;? â&#x2C6;&#x2122; đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; recovery ďż˝ â&#x2C6;&#x2122; ďż˝1 functions ďż˝ recovery function as they were non-monotonic (Fig. 5A, B), we used a two-component pter 6, Eq. 3 voor chapter 7: function adapted from Prijs et al. (1993) to describe masker-probe recovery: exponential

pter 8:

đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; = đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2122; ďż˝1 + đ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;? â&#x2C6;&#x2122; đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;

đ?&#x2018;&#x20AC;đ?&#x2018;&#x20AC;đ?&#x2018;&#x20AC;đ?&#x2018;&#x20AC;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2019;đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ą 0 đ?&#x153;?đ?&#x153;?đ?&#x153;?đ?&#x153;? đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;

â&#x2C6;&#x2019;

ďż˝ â&#x2C6;&#x2122; ďż˝1 â&#x2C6;&#x2019; đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;

â&#x2C6;&#x2019;

ďż˝ ,

(Eq. 2)

đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ą where eCAPâ&#x2C6;&#x2019;is the normalized eCAP amplitude evoked by the probe for a given maskerđ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;? = đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; đ?&#x153;?đ?&#x153;?đ?&#x153;?đ?&#x153;? probe interval MPI, A is the maximum eCAP amplitude evoked by a probe pulse after a sufficiently long MPI, c is a dimensionless constant defining the ratio between the two exponential components, Ď&#x201E;A is the recovery time constant of the first exponential, Ď&#x201E;B is đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ą the đ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;? recovery = đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2019;đ?&#x153;?đ?&#x153;?đ?&#x153;?đ?&#x153;?time constant of the second exponential, and t0 is the absolute refractory period. Fitting examples for a normal-hearing and a six weeks deaf animal are given in Fig. 5A, B. Pulse train recovery functions were constructed using the DC component obtained with the aforementioned Fourier analysis â&#x20AC;&#x201C; i.e., the average N1-P2 amplitude of the eCAPs evoked by the last ten pulses of the pulse train. Since these data (Fig. 5C, D) did not display the kind of non-monotonic recovery as was observed with the masker-probe data, they were fitted with a single exponential recovery function (Eq. 1).

147

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 6

R1 The maximum amplitude evoked by a single pulse (masker) was derived from a apter 6, Eq.sigmoid 1 voor chapter R2 5, Eq. 3 voor chapter Boltzmann fit to the7:input-output function (Ramekers et al., 2014): R3 đ??ľđ??ľđ??ľđ??ľ đ?&#x2018;&#x2030;đ?&#x2018;&#x2030;đ?&#x2018;&#x2030;đ?&#x2018;&#x2030;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; = đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; + R4 (Eq. 3) đ??źđ??źđ??źđ??źâ&#x2C6;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; , 1 â&#x2C6;&#x2019; đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2019; đ??ˇđ??ˇđ??ˇđ??ˇ R5 R6 where eCAP is the eCAP N1-P2 amplitude evoked with current level I, A is the noise level, R7 B is the maximum eCAP amplitude, C is the current level to achieve 50% of the maximum R8 6, Eq. 2 voor chapter eCAP amplitude, and 4D is the dynamic range. apter 7: R9 đ??źđ??źđ??źđ??źđ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ??źđ??źđ??źđ??ź â&#x2C6;&#x2019;đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ą 0 R10 2.4.3. đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; N2 peak =recovery đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2122; ďż˝1 â&#x2C6;&#x2019; assessed đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2019; đ?&#x153;?đ?&#x153;?đ?&#x153;?đ?&#x153;? ďż˝with masker-probe paradigm R11 The N2 peak amplitude (see Fig. 1A) was defined by the distance between the N2 peak R12 and a straight line from the P2 to the P3 peak. This peak was only analyzed for 800-ÂľA R13 stimuli, because of its low amplitude and higher threshold than the N1 peak. The N2 R14 apter 6, Eq. 3 voor chapter 7: were normalized to the amplitude obtained with the masker pulse, for each amplitudes R15 animal separately. đ?&#x2018;&#x20AC;đ?&#x2018;&#x20AC;đ?&#x2018;&#x20AC;đ?&#x2018;&#x20AC;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2019;đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ą 0 đ?&#x2018;&#x20AC;đ?&#x2018;&#x20AC;đ?&#x2018;&#x20AC;đ?&#x2018;&#x20AC;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2019;đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ą 0 â&#x2C6;&#x2019; â&#x2C6;&#x2019; R16 đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; = đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2122; ďż˝1 + đ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;? â&#x2C6;&#x2122; đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; đ?&#x153;?đ?&#x153;?đ?&#x153;?đ?&#x153;? đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; ďż˝ â&#x2C6;&#x2122; ďż˝1 â&#x2C6;&#x2019; đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; đ?&#x153;?đ?&#x153;?đ?&#x153;?đ?&#x153;? đ??ľđ??ľđ??ľđ??ľ ďż˝ R17 2.5. Histology R18 After completion of the experiment, all animals were sacrificed and their right cochleas R19 were harvested for histological analysis. For a detailed description we refer to Ramekers R20 apter 8: et al. (2014). In short, cochleas were fixated with 3% glutaraldehyde, 2% formaldehyde, R21 1% acrolein and 2.5% DMSO in a 0.08 M sodium cacodylate buffer, decalcified, postđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ą R22 đ?&#x153;?đ?&#x153;?đ?&#x153;?đ?&#x153;? fixated, and in Spurrâ&#x20AC;&#x2122;s low-viscosity resin. From each cochlea five semi-thin đ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;? = đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2019;embedded R23 (1 Âľm) midmodiolar sections were cut at 30-Âľm intervals which were subsequently R24 stained with 1% methylene blue and 1% azur B in 1% borax. R25 Within each transection of Rosenthalâ&#x20AC;&#x2122;s canal, the number of type-I SGCs was R26 counted and packing density was averaged across all five sections. In one of the five R27 sections, the average perikaryal area was determined for SGCs with a visible nucleus R28 using ImageJ (Version 1.42q, National Institutes of Health, Bethesda, MA, USA). Since R29 the likelihood of detecting an individual SGC depends on its perikaryal size, the average R30 packing density was corrected for perikaryal size as previously described (Coggeshall R31 and Lekan, 1996; van Loon et al., 2013). R32 Packing densities of peripheral processes (PPs) were determined to a large extent R33 as previously described (Waaijer et al., 2013). In short, semi-thin (1 Âľm) sections were R34 cut parallel to the midmodiolar plane at the level of the osseous spiral lamina (OSL). R35 Sections were examined using a Leica DC300F digital camera mounted on a Leica DMRA R36 light microscope, and a 63Ă&#x2014; oil immersion lens (Leica Microsystems GmbH, Wetzlar, R37 Germany). Packing densities were determined by delineating the bony boundaries of R38 the OSL and subsequently dividing the number of PPs by the thus obtained transection R39 area of the OSL.

148

Recovery characteristics of the auditory nerve

2.6. Statistical analysis Repeated measures (rm) ANOVA was applied to evaluate differences among groups and for the various stimulation protocols for maximum eCAP N1-P2 peak amplitude; among groups and for various current levels for amplitude modulation depth, masker-probe and pulse train recovery parameters; and among groups and for various MPIs or IPIs for N1 peak latency. For rm ANOVAs in which current level was the within-subject factor, only the highest four current levels were used, since current levels below 560 µA did not lead to eCAP responses above noise floor in all animals. In case the assumption of sphericity was violated Greenhouse-Geisser correction was applied. All rm ANOVAs were performed with SPSS 20.0 for Windows (IBM, Armonk, NY, USA). The correlation between SGC and PP packing densities, as well as the correlation between N2 latency and eCAP recovery measures, was assessed using Pearson’s correlation coefficient in MATLAB. Correlations between eCAP parameters and quantified histological measures were assessed with multiple linear regression analysis, using the Statistics toolbox in MATLAB.

3. Results 3.1. Animal inclusion, confirmation of deafening and histological quantification Reliable eCAP recordings could be achieved in all but two animals. In one 2WD animal eCAPs could not be obtained with pulse train stimulation, and in one 6WD animal no eCAPs could be recorded with either stimulation paradigm; recordings in these animals contained strong stimulation artifacts masking any eCAP responses. Successful ototoxic treatment was confirmed with click-evoked ABR recordings and hair cell counts. ABR threshold shifts for all deafened animals ranged from 59 to 82 dB, with the exception of one animal in the 2WD group that had a 27 dB threshold shift. This animal is excluded from group averages. Hair cell counts were significantly lower two or six weeks after ototoxic treatment (see Ramekers et al. [2014] for details). Progressive SGC degeneration was confirmed by histological quantification (see also Ramekers et al., 2014). SGC packing density decreased from 1671 ± 43 in NH animals to 1238 ± 56 in 2WD and 648 ± 68 cells per mm2 in 6WD (average ± SEM), while mean perikaryal area decreased from 213 ± 4 in NH to 186 ± 3 in 2WD and 172 ± 6 µm2 in 6WD animals (average ± SEM). Similarly, PP packing density decreased from 80 ± 2 in NH animals to 62 ± 2 in 2WD and 30 ± 3 processes per 1000 µm2 in 6WD [F(2,14) = 85.6; P < 0.001]. In addition, a strong correlation between SGC and PP packing density was observed [R2 = 0.93; P < 0.001]. 149

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 6

3.2. eCAP waveform Examples of eCAP recordings from both masker-probe and pulse train stimulation are shown in Fig. 2. Recorded before each masker-probe pair, eCAPs in response to the masker pulse (Fig. 2A) exemplify the reproducibility of single-pulse eCAP recordings in a NH animal. The N1-P2 peak amplitude in response to the probe decreased gradually as the MPI decreased (Fig. 2B; same NH animal). In addition, the eCAP N1 and N2 peak latency became gradually longer. In Fig. 2C the eCAPs evoked by the last ten pulses of an 8-ms IPI pulse train are plotted. The similar appearance of these ten eCAPs illustrates that the SGC population from this particular 2WD animal was perfectly able to follow a 125 Hz pulse train. In case of a 0.4-ms IPI pulse train in the same animal (Fig. 2D), the last ten eCAPs exhibit an oscillating pattern with large and early waveforms for odd-numbered pulses (N-9, N-7, etc.) and with small and late waveforms for even-numbered pulses. In a different 2WD animal, the gradual decrease in amplitude with decreasing IPI after the last (Nth) pulse is depicted (Fig. 2E; compare with masker-probe recovery in Fig. 2B). The N2 peak disappeared for IPIs smaller than 4 ms, whereas for the masker-probe example (Fig. 2B) it was still present for 0.6 ms MPI.

3.3. eCAP amplitude modulation Figure 3 shows for three individual animals (one from each group) the N1-P2 peak amplitude for the last ten eCAPs and for several IPIs, for 800 µA stimulation level, normalized to a single-pulse eCAP. The amplitude gradually decreased with higher pulse rate (as also shown in Fig. 2E), with no indication of a decrease in amplitude over pulse number – suggesting that the 100-ms pulse train duration was long enough for the SGC population to stabilize its response after an initial decrease. However, for IPIs below 1 ms the amplitude displayed a robust oscillating pattern, in particular for the deafened animals (see also Fig. 2D). Note that these amplitudes are not derived from single recordings but from averages of 50 iterations. Fourier analysis of the amplitude modulation yielded the average amplitude (AC component) of the oscillation. Examples of amplitude modulation for 800 µA stimulation level as function of IPI for a NH, a 2WD and a 6WD animal are shown in Fig. 4A. In these examples the amplitude modulation depth is present mainly for the two deafened animals and it peaks around 0.6 ms IPI. Therefore, the across-animal average of the modulation amplitude in Fig. 4B is shown for 0.6 ms IPI. Amplitude modulation was stronger for the 6WD group than for the other two [rm ANOVA over 4 highest current levels – main effect of group, F(2,14) = 7.3; P = 0.007; post hoc Bonferroni – NH-2WD, P = 1.0; NH-6WD, P = 0.008; 2WD-6WD, P = 0.044]. Current level did not significantly influence amplitude modulation [F(1.4,20) = 1.4; P = 0.26]. 150

amplitude modulation (re max amp)

Recovery characteristics of the auditory nerve

30% A

30% B

NH 2WD 6WD

20%

IPI: 0.6 ms

20%

10%

0% 0.3

10%

1.0 3.0 inter−pulse interval (ms)

400

480

560 640 720 current level (µA)

800

Fig. 4. A Amplitude modulation depth plotted relative to the maximum eCAP amplitude for three individual animals, as function of IPI (current level is 800 µA). B Group averages of the amplitude modulation for 0.6-ms IPI plotted as function of current level. NH, normal hearing; 2WD, 2 weeks deaf; 6WD, 6 weeks deaf; IPI, inter-pulse interval. N = 6 for NH, N = 5 for 2WD and 6WD; error bars represent SEM.

3.4. Recovery

3.4.1. Examples Examples of recovery function fitting are depicted in Fig. 5. For all 17 animals and current levels the masker-probe recovery was non-monotonic to some extent, as is exemplified for a NH animal (Fig. 5A) and a 6WD animal (Fig. 5B) around 1-2 ms IPI. Accordingly, the R2 when fitting with a single exponential (Eq. 1) was 0.87, whereas it increased to 0.98 when fitting the double exponential (Eq. 2; averaged over all animals and current levels). Recovery functions constructed with pulse train data almost invariably followed a monotonic course (e.g., the NH animal in Fig. 5C). Sporadic instances resembled the recovery as seen with masker-probe stimulation (e.g., the 800 µA trace of 6WD animal in Fig. 5D plateaus around 1.2 ms IPI). Fitting of these data with the single exponential function (Eq. 1) gave an average R2 of 0.92. The parameters derived from both exponential fits were first analyzed as group averages (Figs. 8 and 9) and pulse train data were subsequently individually correlated with quantified histological measures (Fig. 11).

151

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

eCAP amplitude (probe/masker) eCAP amplitude (re single pulse)

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 6

1.0

A − NH

B − 6WD

800 µA 560 µA 480 µA 400 µA 320 µA

0.8 0.6 0.4 0.2 0 0.3 1.0

C − NH

10 0.3 MPI (ms)

10 0.3 IPI (ms)

D − 6WD

0.8 0.6 0.4 0.2 0 0.3

Fig. 5. Examples of recovery function fitting. Recovery functions obtained with masker-probe paradigms were fitted with a double exponential function (Eq. 2) as exemplified for several current levels for a normal-hearing (A) and a 6 weeks deaf animal (B). Recovery functions obtained with pulse train stimulation were fitted with a single exponential function (Eq. 1) as exemplified for several current levels for a normal-hearing (C) and a 6 weeks deaf animal (D). All amplitudes were normalized to an 800-µA masker response. For clarity, recovery functions for current levels 720 and 640 µA are not shown. Note that the logarithmic scale of the time axes slightly distorts the fitting curves (400 µA trace in C). MPI, masker-probe interval; IPI, inter-pulse interval. Symbols indicate actual eCAP amplitude; dotted lines represent fitted curves.

3.4.2. N1 and N2 recovery functions In Fig. 6 averaged N1 and N2 recovery functions are depicted for NH (Fig. 6A) and 2WD animals (Fig. 6B) for masker-probe stimulation at 800 µA current level. The averaged N1 recovery functions display the same course as the 800-µA traces in Fig. 5A, B. Noticeably, the N2 peak recovery shows similar characteristics – an initial increase in amplitude resulting in a peak around 1 ms, followed by a decrease and a final increase for MPIs beyond 2 ms. Apparent differences are the consistently higher amplitude than the masker-evoked N2 for long MPIs, and the more pronounced peak around 1 ms. N2 152

Recovery characteristics of the auditory nerve

eCAP amplitude (probe/masker)

recovery functions for 6WD animals are not shown since for this group the N2 peak was often too small for meaningful analysis – on average it decreased after deafening from 93 ± 16 (NH) to 87 ± 17 (2WD) to 19 ± 6 µV (6WD) for masker pulses (average ± SEM). A − NH

B − 2WD

N peak 1 N peak 2

0.3

MPI (ms)

Fig. 6. Averaged recovery functions for N1 and N2 peaks for NH animals (A) and 2WD animals (B) for masker-probe stimulation at 800 µA current level. Amplitudes for probe N1 and N2 peaks are separately normalized to N1 and N2 peaks evoked by the masker (dotted lines). Note that the solid lines are not fitted curves (in contrast to the dotted lines in Fig. 5). NH, normal hearing; 2WD, 2 weeks deaf; N = 6 for NH, N = 5 for 2WD; error bars represent SEM.

3.4.3. Maximum amplitude The maximum eCAP N1-P2 peak amplitude decreased after deafening, as shown for various stimulation protocols in Fig. 7 [rm ANOVA – F(2,12) = 8.6; P = 0.005; post hoc Bonferroni – NH-2WD, P = 1.0; NH-6WD, P = 0.005; 2WD-6WD, P = 0.041]. While theoretically the maximum eCAP amplitude should be similar regardless of the stimulation protocol, it did significantly differ among protocols [rm ANOVA – F(1.4,17) = 15.9; P < 0.001; simple contrast – pulse train (not masker or probe) lower than single pulse, F(1,12) = 62.6; P < 0.001]. The lower maximum amplitude for pulse train protocols probably reflects a temporary adaptation mechanism.

153

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

2.0

eCAP N1−P2 amplitude (mV)

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 6

NH 2WD 6WD

1.5

1.0

0.5

single

masker

probe

train

stimulation protocol

Fig. 7. Maximum eCAP N1-P2 amplitude for various stimulation protocols, calculated with a Boltzmann sigmoid fit (Eq. 3) for single pulse (Ramekers et al., 2014) and masker, with a double exponential fit (Eq. 2) for probe, and with a single exponential fit (Eq. 1) for pulse train stimulation. Note that the single pulse amplitudes were obtained with recordings published previously (Ramekers et al., 2014) and shown here for comparison only. NH, normal hearing; 2WD, 2 weeks deaf; 6WD, 6 weeks deaf; N = 6 for NH, N = 5 for 2WD and 6WD; error bars represent SEM.

3.4.4. Recovery assessed with masker-probe paradigm The double exponential function fitted to the masker-probe data implies conventional recovery of the SGC population (described by the absolute refractory period t0, recovery time constant τB and maximum amplitude A; Eq. 2), interrupted by a relatively fast facilitatory mechanism – described by a facilitation amplitude constant c and a facilitation time constant τA. Group averages for these four fitting parameters are shown in Fig. 8. The absolute refractory period (Fig. 8A) significantly decreased with current level [rm ANOVA over 4 highest current levels – F(1.5,22) = 16.7, P < 0.001]; possible differences among groups were not statistically significant [rm ANOVA – F(2,14) = 1.9, P = 0.18]. The relatively slow recovery time constant τB (Fig. 8B) also became significantly shorter with increasing current level [rm ANOVA over 4 highest current levels – F(1.5,21) = 9.2, P = 0.003]; a significant interaction effect with group furthermore indicated that the effect size of current level varied among groups [rm ANOVA – F(3,21) = 6.4, P = 0.003]. A significant main effect of group was not observed [rm ANOVA – F(2,14) = 0.20, P = 0.84]. The facilitation amplitude constant c (Fig. 8C) increased with current level [rm ANOVA over 4 highest current levels – F(1.4,18) = 9.7, P < 0.01]; differences among groups 154

Recovery characteristics of the auditory nerve

were not statistically significant [rm ANOVA – F(2,14) = 1.0, P = 0.39]. Finally, the facilitation time constant τA (Fig. 8D) did not change with current level [rm ANOVA over 4 highest current levels – F(1.5,21) = 0.95, P = 0.38], nor were there differences among groups [rm ANOVA – F(2,14) = 1.1, P = 0.35]. In summary, the results show that recovery from refraction (t0 and τB) was dependent on current level – but not similarly so for all groups. In contrast, the magnitude of the facilitation (c) and its timing relative to the preceding eCAP (τA) were more robust with respect to different current levels or groups. 2.5 B

recovery time constant τB (ms)

1.0 A

0.6

1.5

0.4

1.0

0.2

0.5 0 0.8 D

facilitation time constant τA (ms)

facilitation constant c

0 6 C

0.6

4 3

0.4

0.2

1 0

NH 2WD 6WD

2.0

t (ms)

0.8

400

480

560 640 720 current level (µA)

800

400

480

560 640 720 current level (µA)

800

Fig. 8. Masker-probe recovery parameters derived from fitting with double exponential (Eq. 2) plotted as function of current level. A t0; B recovery time constant τB; C facilitation constant c; D facilitation time constant τA. NH, normal hearing; 2WD, 2 weeks deaf; 6WD, 6 weeks deaf; N = 6 for NH, N = 5 for 2WD and 6WD; error bars represent SEM.

3.4.5. Recovery assessed with pulse train paradigm The two parameters derived from the single exponential function fitted to the pulse train data are shown in Fig. 9. t0 was much longer for pulse trains (1-2 ms) than for maskerprobe stimulation (0.4 – 0.6 ms; Fig. 8A). No significant differences were observed in t0 (Fig. 9A) either with current level [rm ANOVA over 4 highest current levels – F(1.7,22) = 0.71, P = 0.48] or among groups [rm ANOVA – F(2,13) = 1.7, P = 0.23]. The recovery time constant τ (Fig. 9B) significantly decreased with current level [rm ANOVA over 4 highest 155

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

current levels – F(3,39) = 21.6, P < 0.001] and varied among groups [rm ANOVA – F(2,13) = 23.7, P < 0.001; post hoc Dunnett: NH-2WD, P = 0.027; NH-6WD, P = 0.0016]. For the 6WD group τ was similar to the values found for masker-probe stimulation (1-1.5 ms; Fig. 8B) whereas for NH and 2WD τ was much longer (3-5 ms).

t0 (ms)

3 A

recovery time constant τ (ms)

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 6

NH 2WD 6WD

0 6 B 5 4 3 2 1 0

400

480

560 640 720 current level (µA)

800

Fig. 9. Pulse train recovery parameters derived from fitting with single exponential (Eq. 1) plotted as function of current level. A t0; B recovery time constant τ. NH, normal hearing; 2WD, 2 weeks deaf; 6WD, 6 weeks deaf; N = 6 for NH, N = 5 for 2WD and 6WD; error bars represent SEM.

3.5. N1 latency For single pulse stimuli, the N1 latency was significantly shorter six – but not two – weeks after deafening (Ramekers et al., 2014). Accordingly, the latency for the masker pulses – plotted at the right-hand side of both Figs. 10A and 10B – displayed similar differences among groups. Overall, latency differences among groups were statistically significant for both masker-probe [rm ANOVA – F(2,14) = 6.3, P = 0.011; post hoc Dunnett: NH-2WD, P = 0.67; NH-6WD, P = 0.0070] and pulse train stimulation [rm ANOVA – F(2,9) = 5.4, P = 0.029; post hoc Dunnett: NH-2WD, P = 0.046; NH-6WD, P = 0.010]. The latency increased with decreasing MPI for masker-probe (Fig. 10A; rm ANOVA – F(2.5,35) = 28.6, P < 0.001) and with decreasing IPI for pulse train stimulation (Fig. 10B; rm ANOVA – F(1.6,14) = 20.1, P < 0.001). A subsequent latency decrease for MPIs shorter than 0.4-0.5 ms and for IPIs shorter than 1.2-1.4 ms was observed. For all MPIs the relative 156

Recovery characteristics of the auditory nerve

differences between groups remained largely similar (NH and 2WD animals overlap, shorter latency for 6WD animals), whereas with pulse train stimulation differences among all three groups became more pronounced with decreasing IPI. 0.7 A

masker−probe

N1 latency (ms)

0.6

NH 2WD 6WD

0.5

↓

0.4 0.3

0.3

0.7 B

1.0 3.0 10 masker−probe interval (ms)

pulse train

N1 latency (ms)

0.6 0.5

↓

0.4 0.3

0.3

1.0 3.0 10 inter−pulse interval (ms)

Fig. 10. A N1 peak latency of eCAPs from masker-probe stimulation, relative to probe stimulus onset, and plotted as function of MPI. B N1 peak latency of eCAPs from pulse train stimulation, averaged for the last ten pulses, relative to the onset of the respective pulse, and plotted as function of IPI. For both (A) and (B), the latency of the eCAP evoked by the masker pulse is plotted separately at the right hand side (arrow) for comparison. MPI, masker-probe interval; IPI, interpulse interval; NH, normal hearing; 2WD, 2 weeks deaf; 6WD, 6 weeks deaf; N = 6 for NH, N = 5 for 2WD and 6WD; error bars represent SEM.

3.6. Correlations with histology 3.6.1. Correlation between pulse train recovery characteristics and SGC degeneration In Fig. 11 several eCAP characteristics for pulse train stimulation at fixed current levels are plotted for each animal individually as function of their respective SGC packing density. Symbol size corresponds to the mean SGC perikaryal area. t0 was not correlated with either histological measure for SGC degeneration (Fig. 11A, B), whereas the recovery constant τ correlated well with SGC size and packing density. Although differences between NH and 6WD animals were larger at lower 157

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

current levels (Fig. 11C), this prediction was only statistically significant at 800 µA (Fig. 11D). The amplitude modulation that occurred with short IPIs was clearly larger after deafening (Figs. 3, 4). For both 640 and 800 µA, more than 50% of the variance could be explained with the combination of SGC size and packing density (Fig. 11E, F).

640 µA

800 µA

1.2 0.8

t (ms)

1.6

0.4 R2 = 0.16

P = 0.316

R = 0.14 P = 0.381

τ (ms)

0 6 C D 5 R2 = 0.17 R2 = 0.42 P = 0.030 4 P = 0.302 3 2 NH 2WD 1 6WD 0 30 E F 2 2 R = 0.54 R = 0.57 25 P = 0.004 P = 0.003 20 15 10 5 0 0 500 1500 0 500 1500 SGC packing density (cells/mm2) Fig. 11. Several pulse train response parameters plotted as function of numerical SGC packing density. Each symbol represents an individual animal; symbol size is proportional to the mean SGC perikaryal area for that animal. Solid lines are regression lines with SGC packing density as single predictor, shown here for purely visual purposes; R2 values represent the amount of variation that can be explained by both SGC packing density and mean perikaryal area, assessed with multiple linear regression analysis. A, B t0; C, D recovery time constant τ; E, F amplitude modulation depth for 0.6 ms IPI, relative to maximum amplitude. The crossed symbol indicates the 2WD animal excluded from the group averages. NH, normal hearing; 2WD, 2 weeks deaf; 6WD, 6 weeks deaf; SGC, spiral ganglion cell; IPI, inter-pulse interval.

amplitude modulation (%)

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 6

3.6.2. Correlation between N1 latency and SGC degeneration The increase in N1 latency brought about by short MPIs/IPIs relative to latencies of masker-evoked N1 peaks (Fig. 10) was averaged over pulse intervals from 0.4 to 2 ms, and plotted as function of SGC packing density in Fig. 12. Symbol size corresponds to the mean SGC perikaryal area. With masker-probe stimulation, 60% of the increase in 158

Recovery characteristics of the auditory nerve

latency could be explained with SGC histology (Fig. 12A); with pulse train stimulation this was 47% (Fig. 12B).

latency difference (µs)

100

masker−probe

75 50 25

R2 = 0.60 P = 0.002

latency difference (µs)

0 300

pulse train NH 2WD 6WD

200

100

R2 = 0.47 P = 0.017

500 1000 1500 2000 SGC packing density (cells/mm2)

Fig. 12. Difference in N1 peak latency between single pulse (masker) and the average of 0.42.0 ms MPI (A) or IPI (B) plotted as function of numerical SGC packing density. Each symbol represents an individual animal; symbol size is proportional to the mean SGC perikaryal area for that animal. Solid lines are regression lines with SGC packing density as single predictor, shown here for purely visual purposes; R2 values represent the amount of variation that can be explained by both SGC packing density and mean perikaryal area, assessed with multiple linear regression analysis. The crossed symbol indicates the 2WD animal excluded from the group averages. NH, normal hearing; 2WD, 2 weeks deaf; 6WD, 6 weeks deaf; SGC, spiral ganglion cell; MPI, maskerprobe interval; IPI, inter-pulse interval.

3.7. N1-N2 peak interval and the origin of the N2 peak In order to assess the origin of the N2 peak visible in the eCAP waveform, we compared the timing of this second peak (relative to the first) with a combined measure of recovery, t0+τB (Fig. 13). The correlation between the two measures was statistically significant, suggesting that the N1 and N2 peak are essentially separate eCAPs each composed of action potentials from the same SGC population, rather than responses of two separate neuronal populations. 159

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

0.70 N1−N2 interval (ms)

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 6

0.65

NH 2WD 6WD

0.60 0.55 0.50 0.45 0.8

R = 0.35 P = 0.013

1.2

1.6 2.0 t +τ (ms) 0

2.4

Fig. 13. Correlation between the N1-N2 peak interval and recovery measure t0+τB obtained with masker-probe stimulation. A significant correlation was confirmed using Pearson’s correlation coefficient. NH, normal hearing; 2WD, 2 weeks deaf; 6WD, 6 weeks deaf.

4. Discussion

The main objective in this study was to characterize the recovery of the electrically stimulated auditory nerve, as well as to evaluate possible changes caused by deafnessinduced SGC degeneration. To this end, we not only assessed recovery characteristics using the classical masker-probe paradigm, but in addition employed more demanding – and more realistic – 100-ms long pulse train stimuli. Our data show that, by assessing neuronal recovery with pulse trains, differences between deafened guinea pigs and normal-hearing controls are enhanced compared to masker-probe stimulation, which could prove useful in clinical diagnostics. Furthermore, the amplitude-modulated responses observed at high pulse rates (and for high current levels) may be indicative of the size of the neural population.

4.1. Non-monotonicity in masker-probe recovery The non-monotonic course of the recovery function for masker-probe stimulation was unmistakable in all animals for both N1 and N2. This non-monotonicity has been observed before in eCAP recovery functions in animals (Stypulkowski and van den Honert, 1984), but it is also present when the probe is an acoustic stimulus (Stronks et al., 2010). It has furthermore been observed in eABR wave-I recovery functions (Miller et al., 1993) and in psychophysical recovery functions in CI users (Nelson and Donaldson, 2001). Stypulkowski and van den Honert (1984) showed a non-monotonic course in eCAP N2 peak (our nomenclature) recovery functions of both normal-hearing and 160

Recovery characteristics of the auditory nerve

acutely ototoxically deafened cats for MPIs around 1 ms. The N2 peak was successfully (and consistently) abolished after cochlear laminectomy, implying that it originated from stimulation of PPs. Furthermore, in their laminectomy preparation the nonmonotonicity was absent in the N1 peak (our nomenclature) recovery function, suggesting that the non-monotonicity, too, resulted from PP stimulation. Based on these observations, Stypulkowski and van den Honert (1984) proposed two mechanisms that might contribute to the non-monotonic recovery behavior in N2 peaks. The first is temporal summation – after subthreshold stimulation of the PP by the masker pulse, the cell membrane remains transiently depolarized, thereby temporarily increasing the likelihood of a successful PP-initiated action potential to the probe. The second hypothesized mechanism entails a shift in excitation site for part of the population from axon for masker (adding to the N1 peak) to PP for probe stimulation (adding to the N2 peak). The explanatory potential of both hypothetical mechanisms is strictly limited to N2 recovery, and, moreover, cannot hold in case the N1 peak displays the same nonmonotonic pattern, as is the case in our findings. Indeed, since the non-monotonic behavior over IPIs is conspicuously similar for N1 and N2 (Fig. 6), the underlying mechanism is more likely one that unites peripheral and axonal components instead of discriminating between them. An alternative explanation for this non-monotonic behavior comes from the examination of spontaneous activity of single auditory nerve fibers. Prijs et al. (1993) found that almost half of the fibers they recorded from had a preference for inter-spike intervals of 1.1 ms on average (range: 0.9-1.5 ms), which is remarkably similar to the 1-ms MPI at which the facilitation for N1 and N2 recovery reached its peak (Fig. 6). If indeed the observed facilitation is a manifestation of a preferred inter-spike interval inherent to (at least some) SGCs, it furthermore explains the unchanged facilitation time constant τA over current levels. The underlying mechanisms causing the relatively short preferred inter-spike interval might be hyperpolarization-activated cation current (Ih), for which the molecular correlate in SGCs – the hyperpolarization-activated cation nonselective (HCN) channel – has recently been characterized in detail (Kim and Holt, 2013). During the hyperpolarization phase following an action potential, these depolarizing currents are known to be able to initiate rebound action potentials. While the existence of a preferred inter-spike interval for SGCs provides a better explanation for the observed non-monotonicity (in both N1 and N2 peak recovery) than temporal summation or shift in excitation site, the question remains why the nonmonotonicity was absent in the laminectomy preparation of Stypulkowski and van den Honert (1984). It should first be noted that they did not present N1 recovery functions from the intact cochlea, which leaves open the possibility that these too would not have 161

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 6

shown non-monotonicities. It is furthermore possible that the laminectomy damaged the SGCs such that their inherent firing characteristics were to some extent altered.

4.2. Recovery assessed with masker-probe and pulse train paradigms Masker-probe recovery characteristics t0 and τB were both dependent on current level (Fig. 8). Shorter t0 with higher current level probably indicates that with increasing current level the t0 approaches the absolute refractory period of the SGC population. While the absolute refractory period of individual SGCs has been shown to increase after deafening in rats from 0.7 to 0.9 ms (Shepherd et al., 2004), we observed comparable t0 (around 0.5 ms) for all groups. Although relative refractoriness (τB) decreased with current level for NH and 2WD animals, it did not appear to vary for 6WD animals (Fig. 8B). In addition, the effect of current level is smaller for 2WD than for NH animals. Faster recovery with higher current level might reflect various mechanisms. First, high current levels elicit faster and more synchronized responses, resulting in earlier onset and completion of recovery. Second, a probe with high current level may recruit more SGCs that are in their relative refractory period than a probe with lower current level. Assuming that the two aforementioned mechanisms apply universally, another (third) mechanism might be operating. In a computational model Botros and Psarros (2010) showed that – with all other parameters kept constant – a population of neurons of a certain size recovers more slowly than a population half the size. Possibly, the disadvantage that the larger population (NH animals) has compared to the smaller population (6WD animals) is annihilated for higher current levels. Pulse train recovery functions revealed much more pronounced differences among animal groups. Recovery was significantly faster for 6WD animals than for NH animals, whereas it was slower for 2WD than for NH animals (Fig. 9B). Faster recovery with greater neural loss is in line with previous findings, showing that faster recovery is correlated with increasing duration of hearing loss in CI users (Botros and Psarros, 2010). Similarly, it is in line with the aforementioned computational model by the same authors, although their model assumed similar thresholds and refractory properties of the individual neurons. Single fiber studies have shown, however, that after deafening latency becomes shorter and threshold decreases in guinea pigs (Sly et al., 2007), and that the absolute refractory period increases in rats (Shepherd et al., 2004). Nonetheless, we observed a fairly good correlation between the recovery time constant τ and histological measures for SGC degeneration (Fig. 11C, D), which is in agreement with the model’s prediction. The observation that the 2WD animals showed slower recovery than NH, while recovery was faster for the 6WD animals, is reminiscent of the single-pulse threshold 162

Recovery characteristics of the auditory nerve

and latency of the same animals (Ramekers et al., 2014). It is conceivable that upon ototoxic treatment the tissue debris as a result of the initial degeneration of the organ of Corti and subsequent degeneration of SGCs negatively affects the electrode-tissue interface. This is consistent with Prijs et al. (1993), who did not find differences in temporal response characteristics of spontaneous discharges between SGCs in normal and noise-damaged cochleas. Since their data were obtained without stimulation, any degenerating process influencing the electrode-tissue interface became irrelevant. Additionally, since the process of SGC degeneration is arguably at its peak around two weeks after ototoxic treatment (slight but significant loss of SGCs, while about half of the population will die within four weeks), a substantial portion of the SGC population that responds to the electric stimulus might be subject to degeneration.

4.3. Neural fatigue The long t0 observed for pulse train recovery (compared to masker-probe recovery) can be interpreted as neural fatigue rather than a representation of the absolute refractory period (Fig. 9A). Since with 1-ms MPI only 80-90% of the SGC population was able to respond (Fig. 5A, B and Fig. 6), a 100-ms pulse train with 1-ms IPI undoubtedly has the potential to exhaust SGCs. Mechanisms underlying this process â&#x20AC;&#x201C; for example utilized to inhibit specific brain regions with deep brain stimulation treatment â&#x20AC;&#x201C; include energy depletion and depolarization blockade (McIntyre et al., 2004). These mechanisms may be expected to arise sooner in a degenerating population (and with higher current level), and it is therefore remarkable that we did not find differences in t0 among groups nor with current level (Figs. 9A; 11A, B). However, it is in agreement with a previous study in guinea pigs, which showed that fatigue was independent of the presence of hair cells (Killian et al., 1994). 4.4. eCAP amplitude modulation For IPIs below t0 an oscillatory pattern arose (Figs. 4A; 5C, D). It has been demonstrated before that failure of a neuronal population to respond to each pulse in a pulse train not only results in an overall decrease in amplitude, but also gives rise to an amplitude modulation over pulse numbers in the auditory nerve (Wilson et al., 1997; Matsuoka et al., 2000a; Hu et al., 2003; Hughes et al., 2012) and in other neural tissue (Cappaert et al., 2013). This pattern of decreasing and subsequent modulated amplitude for neuronal populations is reflected in firing probability for single neurons (Sly et al., 2007; Campbell et al., 2012). Since every data point in Fig. 3 is an average of 50 recordings (and taking into account that for 0.6-ms IPI more than 150 pulses preceded the recorded ten eCAPs) the observed amplitude modulation implies highly consistent response patterns of 163

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 6

individual neurons over trials. Not surprisingly, this pattern is easily disrupted by introducing noise (Matsuoka et al., 2000b). This observation might explain the scarce presence of amplitude modulation for NH animals (Fig. 4), since spontaneous activity of hair cells probably provided sufficient noise to disrupt SGC firing synchronicity. While hair cell activity might explain the absence of amplitude-modulated responses in NH animals, the high correlation between modulation depth and SGC degeneration (Fig. 11E, F) implies an additional mechanism. It is conceivable that in an ever smaller and thus less densely packed population of SGCs the current path to individual cells across Rosenthalâ&#x20AC;&#x2122;s canal becomes ever more similar (i.e., less variability among cells), resulting in a negative relation between the degree of synchronous activity and SGC population size.

4.5. N1 Latency N1 latency increased with decreasing MPI/IPI (Fig. 10) in a manner similar to latency increases in single neurons (Shepherd et al., 2004). For both masker-probe and pulse train stimulation, the latency continued to increase with ever smaller intervals until t0 was reached. In case of masker-probe stimulation the subsequent decrease (from 0.4 to 0.3 ms MPI) resulted from automated peak picking in the absence of detectable eCAPs, while for pulse train stimulation actual latency decrease was observed once the response became amplitude-modulated. This observation can be explained by the longer inter-spike interval for individual SGCs as they respond to odd or even pulses only. The smaller latency increase for deafened animals may be a result of degeneration of the organ of Corti, causing a more direct current path. Degeneration of part of the SGC population may add to this effect. Furthermore, the initial shorter latency for deafened animals may be related to smaller SGC soma size in these animals (Ramekers et al., 2014), which might influence the (lack of) latency increase as well. In accordance with these proposed mechanisms, the increase in latency from single pulse to the average of 0.4-2.0 ms MPI/IPI could be explained for up to 60% with SGC size and packing density as predictors (Fig. 12). 4.6. N2 latency and recovery The significant correlation between the N1-N2 interval and recovery of the N1 peak (Fig. 13) implies that both peaks represent responses from the same population. Recalling that the high correlation between SGC and PP packing density in the present study indicates that as long as an SGC exists, its PP will too (see section 3.1), two situations are then possible:

164

Recovery characteristics of the auditory nerve

(i) Intrascalar stimulation depolarizes both the peripheral and axonal processes, but whereas the axonal depolarization in the majority of instances leads to a successful action potential initiation (and propagation), the peripherally evoked response is generally cancelled as the axon is refractory. Some cases of peripheral depolarization may result in secondary (additional) action potentials of which the latency depends on the duration of the refractory period of the axon. Note that while this situation resembles the hypotheses postulated by Stypulkowski and van den Honert (1984) described above, it can only account for the emergence of the N2 peak â&#x20AC;&#x201C; not for the emergence of recovery non-monotonicities, which were present for both N1 and N2 peaks. (ii) The SGC soma and its peripheral and axonal processes are excited simultaneously, resulting in a single action potential. The emergence of a higher-threshold second N2 peak is the result of a second firing of some SGCs due to prolonged depolarization of the tissue or rebound activity of the membrane (as discussed in section 4.1). As in (i), the latency of this second eCAP is related to the time constant of recovery.

4.7. Clinical implications In human CI users, masker-probe recovery functions can often be easily recorded using the manufacturerâ&#x20AC;&#x2122;s clinical software. However, recovery characteristics thus obtained appear not to be overly sensitive measures to work with. For instance, although Botros and Psarros (2010) suggested that SGC population size is the dominant factor in temporal responsiveness, their results obtained from adult CI users demonstrated only a small and marginally significant correlation between masker-probe recovery time constants and duration of hearing loss. In another study the average slope of the eCAP recovery function in two groups of implanted children with different etiology of deafness was remarkably similar (Kim et al., 2011). While it is very well possible that the observed recovery characteristics in these studies accurately reflect temporal responsiveness, our findings suggest that recovery functions obtained with pulse train stimulation instead of masker-probe stimulation provides a means to augment differences and, by doing so, to more potently discriminate between (groups of) CI users. Amplitude-modulated responses to high pulse rate stimulation have been shown to occur in CI users for IPIs smaller than 2.5-5 ms (Wilson et al., 1997; Hughes et al., 2012; Carlyon and Deeks, 2013); both the onset of occurrence and modulation depth vary among CI users. Functional implications of this modulation have not yet been reported, although Carlyon and Deeks (2013) demonstrated that the amplitude-modulated eCAP is not related to temporal pitch perception. Alternatively, whether or not this response pattern will prove to have functional consequences, the high correlation between modulation depth and SGC degeneration reported here (Fig. 11E, F) suggests that this measure might be tried as an objective measure to assess the condition of the auditory nerve in CI users. 165

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 6

Acknowledgments The authors would like to thank Ferry Hendriksen for histological processing, Emma Smeets for assistance with histological analysis, RenĂŠ van de Vosse for technical support, and Roland Hessler at MED-EL, Innsbruck, for manufacturing the electrode arrays. This work was supported by MED-EL GmbH, Innsbruck, Austria.

166

Recovery characteristics of the auditory nerve

References Botros, A., Psarros, C., 2010. Neural Response Telemetry Reconsidered: II. The Influence of Neural Population on the ECAP Recovery Function and Refractoriness. Ear. Hear. 31, 380-391. Campbell, L.J., Sly, D.J., O’Leary, S.J., 2012. Prediction and control of neural responses to pulsatile electrical stimulation. J. Neural. Eng. 9, 026023. Cappaert, N.L.M., Ramekers, D., Martens, H.C.F., Wadman, W.J., 2013. Efficacy of a new chargebalanced biphasic electrical stimulus in the isolated sciatic nerve and the hippocampal slice. Int. J. Neural. Syst. 23, 1250031. Carlyon, R.P., Deeks, J.M., 2013. Relationships between auditory nerve activity and temporal pitch perception in cochlear implant users. Adv. Exp. Med. Biol. 787, 363-371. Carlyon, R.P., van Wieringen, A., Deeks, J.M., Long, C.J., Lyzenga, J., Wouters, J., 2005. Effect of interphase gap on the sensitivity of cochlear implant users to electrical stimulation. Hear. Res. 205, 210-224. Coggeshall, R.E., Lekan, H.A., 1996. Methods for determining numbers of cells and synapses: a case for more uniform standards of review. J. Comp. Neurol. 364, 6-15. Fayad, J.N., Linthicum, F.H. Jr, 2006. Multichannel cochlear implants: relation of histopathology to performance. Laryngoscope 116, 1310-1320. Hu, N., Abbas, P.J., Miller, C.A., Robinson, B.K., Nourski, K.V., Jeng, F.C., Abkes, B.A., Nichols, J.M., 2003. Auditory response to intracochlear electric stimuli following furosemide treatment. Hear. Res. 185, 77-89. Hughes, M.L., Castioni, E.E., Goehring, J.L., Baudhuin, J.L., 2012. Temporal response properties of the auditory nerve: data from human cochlear-implant recipients. Hear. Res. 285, 46-57. Killian, M.J.P., Klis, S.F.L., Smoorenburg, G.F., 1994. Adaptation in the compound action potential response of the guinea pig VIIIth nerve to electric stimulation. Hear. Res. 81, 66-82. Kim, J.R., Abbas, P.J., Brown, C.J., Etler, C.P., O’Brien, S., Kim, L.S., 2010. The relationship between electrically evoked compound action potential and speech perception: a study in cochlear implant users with short electrode array. Otol. Neurotol. 31, 1041-1048. Kim, J.R., Kim, L.S., Jeong, S.W., Kim, J.S., Chung, S.H., 2011. Recovery function of electrically evoked compound action potential in implanted children with auditory neuropathy: preliminary results. Acta Otolaryngol. 131, 796-801. Kim, Y.H., Holt, J.R., 2013. Functional contributions of HCN channels in the primary auditory neurons of the mouse inner ear. J. Gen. Physiol. 142, 207-223. Kirby, B., Brown, C.J., Abbas, P.J., Etler, C.P., O’Brien, S., 2012. Relationships between electrically evoked potentials and loudness growth in bilateral cochlear implant users. Ear. Hear. 33, 389-398. Matsuoka, A.J., Abbas, P.J., Rubinstein, J.T., Miller, C.A., 2000a. The neuronal response to electrical constant-amplitude pulse train stimulation: evoked compound action potential recordings. Hear. Res. 149, 115-128. Matsuoka, A.J., Abbas, P.J., Rubinstein, J.T., Miller, C.A., 2000b. The neuronal response to electrical constant-amplitude pulse train stimulation: additive Gaussian noise. Hear. Res. 149, 129137. McIntyre, C.C., Savasta, M., Walter, B.L., Vitek, J.L., 2004. How does deep brain stimulation work? Present understanding and future questions. J. Clin. Neurophysiol. 21, 40-50. Miller, C.A., Abbas, P.J., Robinson, B.K., 1993. Characterization of wave I of the electrically evoked auditory brainstem response in the guinea pig. Hear. Res. 69, 35-44. Morsnowski, A., Charasse, B., Collet, L., Killian, M., Müller-Deile, J., 2006. Measuring the refractoriness of the electrically stimulated auditory nerve. Audiol. Neurootol. 11, 389-402. 167

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 6

Nelson, D.A., Donaldson, G.S., 2001. Psychophysical recovery from single-pulse forward masking in electric hearing. J. Acoust. Soc. Am. 109, 2921-2933. Prado-Guitierrez, P., Fewster, L.M., Heasman, J.M., McKay, C.M., Shepherd, R.K., 2006. Effect of interphase gap and pulse duration on electrically evoked potentials is correlated with auditory nerve survival. Hear. Res. 215, 47-55. Prijs, V.F., Keijzer, J., Versnel, H., Schoonhoven, R., 1993. Recovery characteristics of auditory nerve fibres in the normal and noise-damaged guinea pig cochlea. Hear. Res. 71, 190-201. Ramekers, D., Versnel, H., Strahl, S.B., Smeets, E.M., Klis, S., Grolman, W., 2014. Auditory-nerve responses to varied inter-phase gap and phase duration of the electric pulse stimulus as predictors for neuronal degeneration. J. Assoc. Res. Otolaryngol. 15, 187-202. Seyyedi, M., Viana, L.M., Nadol, J.B. Jr., 2014. Within-Subject Comparison of Word Recognition and Spiral Ganglion Cell Count in Bilateral Cochlear Implant Recipients. Otol. Neurotol. 35, 14461450. Shepherd, R.K., Roberts, L.A., Paolini, A.G., 2004. Long-term sensorineural hearing loss induces functional changes in the rat auditory nerve. Eur. J. Neurosci. 20, 3131-3140. Sly, D.J., Heffer, L.F., White, M.W., Shepherd, R.K., Birch, M.G., Minter, R.L., Nelson, N.E., Wise, A.K., O’Leary, S.J., 2007. Deafness alters auditory nerve fibre responses to cochlear implant stimulation. Eur. J. Neurosci. 26, 510-522. Spoendlin, H., 1975. Retrograde degeneration of the cochlear nerve. Acta Otolaryngol. 79, 266– 275. Stronks, H.C., Versnel, H., Prijs, V.F., Klis, S.F.L., 2010. Suppression of the acoustically evoked auditory-nerve response by electrical stimulation in the cochlea of the guinea pig. Hear. Res. 259, 64–74. Stypulkowski, P.H., van den Honert, C., 1984. Physiological properties of the electrically stimulated auditory nerve. I. Compound action potential recordings. Hear. Res. 14, 205-223. van Loon, M.C., Ramekers, D., Agterberg, M.J.H., de Groot, J.C.M.J., Grolman, W., Klis, S.F.L., Versnel, H., 2013. Spiral ganglion cell morphology in guinea pigs after deafening and neurotrophic treatment. Hear. Res. 298, 17-26. Versnel, H., Agterberg, M.J.H., De Groot, J.C.M.J., Smoorenburg, G.F., Klis, S.F.L., 2007. Time course of cochlear electrophysiology and morphology after combined administration of kanamycin and furosemide. Hear. Res. 231, 1-12. Waaijer, L., Klis, S.F.L., Ramekers, D., Van Deurzen, M.H.W., Hendriksen, F.G.J., Grolman, W., 2013. The Peripheral Processes of Spiral Ganglion Cells After Intracochlear Application of BrainDerived Neurotrophic Factor in Deafened Guinea Pigs. Otol. Neurotol. 34, 570-578. Webster, M., Webster, D.B., 1981. Spiral ganglion neuron loss following organ of Corti loss: a quantitative study. Brain Res. 212, 17-30. Wilson, B.S., Finley, C.C., Lawson, D.T., Zerbi, M., 1997. Temporal representations with cochlear implants. Am. J. Otol. 18, S30-S34. Ylikoski, J., Wersall, J., Bjorkroth, B., 1974. Degeneration of neural elements in the cochlea of the guinea pig after damage to the organ of corti by ototoxic antibiotics. Acta Otolaryngol. Suppl. 326, 23-41.

168

CHAPTER 7 Temporary neurotrophic treatment prevents deafness-induced auditory nerve degeneration and preserves functionality

Dyan Ramekers, Huib Versnel, Stefan B. Strahl, Sjaak F.L. Klis, Wilko Grolman

Let’s see if I’m hearing this right/ You suggest I should take/ A never-ending supply – Joshua Homme, “Better Living Through Chemistry”

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 7

Abstract

After substantial loss of cochlear hair cells, a condition known as sensorineural hearing loss (SNHL), exogenous neurotrophic factors prevent degeneration of the auditory nerve. Since cochlear implantation – currently the only therapy for severe SNHL – depends on a functional auditory nerve, possible clinical applications of neurotrophic factors are being investigated. Here we addressed two questions which are both important for fundamental insight in effect of exogenous neurotrophins on a degenerating neural system, and crucial for translation to the clinic: first, whether temporary treatment with brain-derived neurotrophic factor (BDNF) has the potential to prevent nerve degeneration on the long term; and second, whether a BDNFtreated nerve is electrophysiologically healthy. Two weeks after deafening guinea pigs received a cochlear implant and their cochleas were infused with BDNF for four weeks. Either directly afterwards, or four or eight weeks later (when without treatment the guinea pig auditory nerve drastically degenerates), their cochleas were histologically analyzed. Electrically evoked compound action potentials (eCAPs) were recorded at several time points after implantation. BDNF treatment prevented spiral ganglion cell (SGC) degeneration during the treatment and, importantly, up to eight weeks after treatment cessation. The potentiating effect of an inter-phase gap in the biphasic current pulse on stimulation efficacy – previously reported to reflect neural survival – was different between normal-hearing and deaf controls for amplitude, latency, and current level to reach the half-maximum eCAP amplitude (level50%). Importantly, for BDNF-treated groups these eCAP characteristics were indistinguishable from normalhearing responses, suggesting healthy responsiveness in BDNF-treated SGCs. Temporal response properties slightly changed with BDNF treatment, but recovered after treatment cessation. We conclude that clinically practicable short-term neurotrophic treatment is sufficient for long-term survival of SGCs, and that it can restore or preserve SGC functionality well beyond the treatment period as well.

170

Neurotrophic treatment prevents SGC degeneration

1. Introduction Profound hearing loss is often caused by hair cell loss, a condition which is referred to as sensorineural hearing loss (SNHL). Cochlear implants (CIs) essentially replace hair cells by encoding sound and conveying the signal by means of pulsatile electrical stimulation to the spiral ganglion cells (SGCs) which form the auditory nerve. Being essential for CI therapy, the effect of hair cell loss on these cells has been investigated thoroughly. It has been demonstrated in numerous animal studies that SGCs degenerate in response to induced hair cell loss (Ylikoski et al., 1974; Spoendlin, 1975; Webster and Webster, 1981; Versnel et al., 2007). The underlying mechanism causing this secondary degeneration is thought to be the discontinuation of neurotrophic support from hair cells and their supporting cells (Ernfors et al., 1995; Fritzsch et al., 1999; Zilberstein et al., 2012) â&#x20AC;&#x201C; a notion that is supported by the fact that treatment with exogenous neurotrophic factors, such as brain-derived neurotrophic factor (BDNF), neurotrophin-3 (NT-3) and glial cell line-derived neurotrophic factor (GDNF), prevents SGC degeneration (reviewed in Ramekers et al., 2012). Histological data of CI usersâ&#x20AC;&#x2122; cochleas is sparse, and until recently the necessity of preserving a large SGC population in order to ensure good CI performance in human subjects was subject to debate, as findings implied that CI performance was negatively correlated with the number of surviving SGCs in a population of mostly unilaterally implanted patients (Fayad and Linthicum, 2006). Recently however, using a withinsubject design, Seyyedi et al. (2014) found a strong positive correlation based on SGC counts and word recognition scores from both ears in bilaterally implanted patients. For treatment with neurotrophic factors to become suitable for clinical application, the means of delivery is an important issue (Staecker and Garnham, 2010). The requirements for an appropriate delivery method are directly related to the intended duration of the treatment. Cell-based or viral-based delivery of neurotrophic factors may be appropriate if continuous support is necessary for SGC survival (e.g., Wise et al., 2011; Atkinson et al., 2014), although there are practical and ethical objections attached to these methods. Administration by means of osmotic pumps (e.g., Agterberg et al., 2009), or single-dose administration onto the round window membrane (e.g., Havenith et al., 2011) are then considered to be impractical because of the limited lifespan of the treatment. However, evidence of prolonged effect (2-4 weeks) of four-week treatment with neurotrophic factors is accumulating (Maruyama et al., 2008; Agterberg et al., 2009; Fransson et al., 2010). However brief this prolonged effect may seem, it is arguably long enough to reflect a specific survival mechanism rather than a lingering effect of the administered neurotrophic factors. It has been suggested that short-term treatment might be sufficient to trigger a self-supporting mechanism in SGCs (Ramekers 171

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 7

et al., 2012). In this light, the first objective of the present study was to assess SGC survival after an extended time period after cessation of BDNF treatment. Before considering neurotrophic treatment for clinical application, the fundamental question on the effect of the treatment on SGC responsiveness to electrical stimulation needs to be addressed. So far, assessment of the effect of neurotrophic treatment on SGC functionality has consisted exclusively of the evaluation of the electrically evoked auditory brainstem response (eABR). Predominantly eABR thresholds have been reported, which are mostly found to be lower in animals treated with neurotrophic factors than in deaf controls (Shinohara et al., 2002; Endo et al., 2005; Shepherd et al., 2005; Chikar et al., 2008; Maruyama et al., 2008; Agterberg et al., 2009; Leake et al., 2011). Given that the eABR threshold is generally found to increase with SGC loss after deafening, its decrease in response to neurotrophic treatment may well reflect SGC population size rather than actual SGC functionality. A more crucial limitation of the use of eABRs to assess SGC responsiveness to electrical stimulation is that it is an indirect measure of the cells that are the target of the treatment; central processes of degeneration and homeostatic plasticity may play an important role (e.g., Schaette and Kempter, 2008). The second objective was therefore to more directly examine the effect of neurotrophic treatment on the functionality of the SGC population, by means of electrically evoked compound action potential (eCAP) recordings. The first part of this functional assessment consisted of a detailed evaluation of response properties in singlepulse stimulation paradigms. Specifically, the effect of an inter-phase gap (IPG) in the biphasic current pulse was examined, since the magnitude of this effect has previously been shown to correlate with SGC survival (Prado-Guitierrez et al., 2006; Ramekers et al., 2014b). The second part consisted of an evaluation of temporal response properties, assessed with masker-probe and pulse train stimulation paradigms (Ramekers et al., 2014a). Together, the two objectives of this study were aimed at determining whether brief treatment with BDNF has long-term protective effects on the SGC population, which may additionally be beneficial for responsiveness to electrical stimulation. From a clinical perspective, this study contributes to the assessment of clinical applicability of neurotrophic factors as a treatment for SGC degeneration in CI users. In more general terms, long-lasting protection of neural tissue with short-term BDNF treatment may be an important tool against other neurodegenerative disorders as well.

172

Neurotrophic treatment prevents SGC degeneration

2. Methods 2.1. Animals and experimental design Forty female albino guinea pigs (Dunkin Hartley; 250–350 g) were obtained from Harlan Laboratories (Horst, the Netherlands) and kept under standard laboratory conditions (food and water ad libitum; lights on between 7:00 am and 7:00 pm; temperature 21 °C; humidity 60%). All animals had bilateral normal hearing prior to any experimental treatment, as assessed with click-evoked auditory brainstem responses (ABRs). One group served as normal-hearing controls (NH, N = 6) and was implanted directly after confirmation of normal hearing. The remaining groups (14 weeks deaf, 14WD, N = 6; BDNF-treated, BDNF0, N = 9; BDNF-treated with an additional four-week period after cessation of the treatment, BDNF4, N = 10; and BDNF-treated with an additional eightweek period after cessation of the treatment, BDNF8, N = 9) were exposed to ototoxic treatment two weeks before implantation (Fig. 1). Animals were implanted with an intracochlear electrode array combined with a cannula which was connected to a subcutaneously placed osmotic pump. This procedure was identical for all animals, with the only exception being that the pump reservoir was filled with phosphate-buffered saline (PBS) either with or without BDNF. Four weeks after implantation the osmotic pumps were surgically removed to end the treatment. Depending on the treatment schedule animals were sacrificed four, eight or twelve weeks after implantation (Fig. 1). Electrophysiological measurements were performed immediately after each surgical procedure, and immediately prior to termination and histological processing of the cochleas. After implantation and pump removal in deafened animals, the measurements that involved electrical stimulation (eCAP recordings) were performed in a brief session of 2-3 minutes, so as not to allow possible neurotrophic effects of electrical stimulation to occur, even though such effects were assumed to be negligible (Agterberg et al., 2010). In the normal-hearing animals these recordings were somewhat extended. All surgical and experimental procedures were approved by the Animal Care and Use Committee of Utrecht University (DEC 2010.I.08.103). 2.2. Surgical procedures

2.2.1. Deafening procedure Animals were anesthetized by intramuscular injection of dexmedetomidine (Dexdomitor®; 0.25 mg/kg) and ketamine (Narketan®; 40 mg/kg), and click-evoked ABRs were recorded in order to confirm normal hearing. Deafening was done by subcutaneous injection of kanamycin (Sigma-Aldrich, St. Louis, MO, USA; 400 mg/kg) and subsequent infusion of furosemide (Centrafarm, Etten-Leur, the Netherlands; 100 173

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 7

mg/kg) into the external jugular vein, which has been shown to eliminate the majority of both inner and outer hair cells (West et al. 1973; Versnel et al., 2007. Post-operatively, the animals were injected subcutaneously with the non-ototoxic antibiotic enrofloxacin (Baytril®; 5 mg/kg) and the anti-inflammatory drug carprofen (Rimadyl®; 5 mg/kg). treatment implantation cessation ↓ treatment ↓ deafening period ↓ NH PBS

14WD

PBS

BDNF0

BDNF

BDNF4

BDNF

BDNF8

BDNF 0

6 time in weeks

Fig. 1. Treatment schedule for all five groups. At time point t = 0 all groups except the NH controls were ototoxically deafened after confirmation of normal hearing. At t = 2 all animals were implanted with an electrode array and infusion cannula. The subcutaneously positioned osmotic pump contained PBS either with (BDNF0, BDNF4 and BDNF8 groups) or without (NH and 14WD groups) BDNF. At t = 6 the osmotic pumps were surgically removed for all groups except for the BDNF0 group. At the end of the experimental procedure (t = 6 for BDNF0, t = 10 for BDNF4 and t = 14 for all other groups) extensive eCAP recordings were performed, and the animals were sacrificed for histological analysis of both cochleas immediately afterwards. BDNF, brain-derived neurotrophic factor; PBS, phosphate-buffered saline.

2.2.2. Osmotic pump preparation and cochlear implantation We followed procedures for BDNF delivery that were comparable to those performed previously in our laboratory (Agterberg et al., 2008, 2009). Approximately 40 h before cochlear implant surgery the Alzet® osmotic pumps (model 2004; reservoir volume: 200 µl; flow rate 0.25 μl/h; Durect, Cupertino, CA, USA) were filled with PBS containing 1% guinea pig serum (Sigma-Aldrich, St. Louis, MO, USA) for NH and 14WD groups. For the BDNF-treated groups 100 µg/ml BDNF (PeproTech, Rocky Hill, NJ, USA) was added. Pumps were then incubated in sterile PBS at 37 °C until implantation. Anesthesia was induced as described for deafening above; in addition 0.05 mg/ kg atropine was given by intramuscular injection to reduce bronchial secretion. Clickevoked ABRs were recorded in order to confirm a successful deafening procedure (i.e., threshold shift > 50 dB) before initiation of the implantation procedure. 174

Neurotrophic treatment prevents SGC degeneration

The skull was exposed and five transcranial screws (1.2 mm diameter) were placed for eCAP stimulation and eCAP and ABR recording purposes: one 2 cm anterior to bregma (ABR reference electrode), one 1 cm posterior to bregma (ABR active electrode), one 1 cm left of bregma (eCAP recording reference electrode), and two 1 cm right of bregma (ABR ground electrode; eCAP stimulation reference electrode); an additional screw was positioned just anterior to bregma for anchoring of the implant connector onto the skull. Via a retro-auricular approach the right bulla was opened to gain access to the basal turn of the cochlea. A 0.4 mm cochleostomy was drilled within 1 mm from the round window through which the array/cannula was inserted into the scala tympani for approximately 4 mm. The array had a diameter of 0.3 mm at the tip and slightly tapered so that it sealed off the cochleostomy; it consisted of two electrodes, located approximately 0.5 and 2.1 mm from the tip, and a 75 µm diameter fluid outlet located in between. To ensure proper functioning of the electrodes and correct insertion of the array, electrode impedances were measured and eCAP recordings were tested. The opening in the bulla was closed with glass ionomer cement (GC Fuji PLUS; GC corporation, Tokyo, Japan), and the connector of the electrode array was fixed onto the skull with polymer dental cement (ProBase Cold; Ivoclar Vivadent AG, Schaan, Liechtenstein). The eCAP reference electrodes were soldered to the designated transcranial screws, which were then covered with dental cement. The cannula, which passed through the connector, was filled with sterile PBS (either with or without BDNF) before insertion into the cochlea; the flow moderator of the osmotic pump was attached and the pump was subsequently placed subcutaneously anterior to the left shoulder. 2.2.3. Removal of the osmotic pump In order to ensure treatment cessation the osmotic pumps were surgically removed four weeks after implantation (Agterberg et al., 2009). Anesthesia was induced as described for deafening above. After carefully removing the encapsulating tissue, the osmotic pump was examined for signs of treatment failure (e.g., detachment of the cannula; opaque content of the cannula), and subsequently removed. The open cannula was tied off using nonabsorbable nylon sutures (Ethilon™; Ethicon, Somerville, NJ, USA) and fixed to the subcutis. After the surgical procedure, brief impedance measurements and eCAP recordings were performed.

2.2.4. Surgical procedure for eCAP experiment At the end of the treatment period (see Fig. 1) extensive eCAP recordings were done. Preparations for these recording sessions were done as has been described previously (Ramekers et al., 2014b). In brief, anesthesia was initiated with Hypnorm® (Vetapharma, Leeds, UK; 0.5 ml/kg i.m.), followed by a gas mixture of 2% isoflurane evaporated in 175

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 7

O2 and N2O (1:2). Atropine (0.05 mg/kg) was given to reduce bronchial secretion. The animals were then tracheostomized, and were artificially ventilated throughout the recording session. 2.3. Electrophysiology

2.3.1. Auditory brainstem responses After induction of anesthesia and prior to ototoxic treatment (2.2.1) and implantation (2.2.2) click-evoked ABRs were recorded using subcutaneously positioned needle electrodes (active electrode behind the right pinna; reference electrode on the skull, rostral to the brain; ground electrode in hind limb). Broadband acoustic clicks (20 μs monophasic rectangular pulses; inter-stimulus interval 99 ms) were synthesized and attenuated using a TDT3 system (modules RP2, PA5 [2x] and HB7; Tucker-Davis Technologies, Alachua, FL, USA), and presented in free field using a Blaupunkt speaker (PCxb352; 4 Ω; 30 W; Blaupunkt GmbH, Hildesheim, Germany). The responses were differentially amplified using a Princeton Applied Research (Oak Ridge, TN, USA) 5113 pre-amplifier (amplification ×5,000; band pass filter 0.1–10 kHz), digitized by the TDT3 system (100 kHz sampling rate, 24-bit sigma-delta converter) and stored on a PC for off-line analysis. Hearing thresholds were obtained by starting at 110 dB peak equivalent (pe) SPL (calibrated using Brüel & Kjær sound level meter and condenser microphone), and decreasing the sound level in steps of 10 dB until the response had disappeared. This was done for each ear separately by plugging the contralateral ear. The threshold was then defined as the interpolated sound level at which the ABR was 0.3 µV. Thresholds ≤ 40 dB peSPL were considered to indicate normal hearing.

2.3.2. Compound action potentials For a detailed description of eCAP recording settings and stimulation paradigms we refer to Ramekers et al. (2014b) for single-pulse and to Ramekers et al. (2014a) for multiple-pulse stimulation. A shielded cable was used to connect the electrode array and reference electrodes to a MED-EL PULSARci100 cochlear implant (MED-EL GmbH, Innsbruck, Austria). The implant was controlled by a PC via a Research Interface Box 2 (RIB2; Department of Ion Physics and Applied Physics, University of Innsbruck, Innsbruck, Austria) and a National Instruments data acquisition card (PCI-6533, National Instruments, Austin, TX, USA). Stimulation and recording paradigms, as well as data analysis scripts, were created in MATLAB (version 7.11.0; Mathworks, Natick, MA, USA). Both stimulation and recording of eCAPs was done with monopolar configuration; the most apical of the two intracochlear electrodes was used for stimulation and the

176

Neurotrophic treatment prevents SGC degeneration

most basal one for recording. Biphasic current pulses were presented with alternating polarity to reduce stimulation artifact, while recordings to a subthreshold stimulus were subtracted to eliminate measurement onset artifacts.

2.3.2.1. Single pulse paradigm The phase duration (PD) of the current pulses was 50 Âľs; the IPG was varied from 2.1 to 30 Âľs (Fig. 2A). Maximum current levels were determined for each animal and each recording session individually, based on saturation of the input-output curves, or in some cases based on the maximum output level of the implant (6.1 V in our set-up), which could be the limiting factor in case of high impedances. The maximum charge per phase was 23 nC on average (range among animals: 14-30 nC; mean range among groups: 2124 nC). Ten current levels were applied in 10%-steps of the thus determined maximum current level. The N1 amplitude was defined as the difference in voltage between the N1 and the P2 peak (first positive peak following N1; see Fig. 2B-D). Input-output functions pter 5, Eq. 3 voor chapter 6, Eq. 1 voor chapter 7: were fitted with a Boltzmann sigmoid, following Ramekers et al. (2014b): đ??ľđ??ľđ??ľđ??ľ đ?&#x2018;&#x2030;đ?&#x2018;&#x2030;đ?&#x2018;&#x2030;đ?&#x2018;&#x2030;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; = đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; + (Eq. 1) đ??źđ??źđ??źđ??źâ&#x2C6;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;

1 â&#x2C6;&#x2019; đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2019;

đ??ˇđ??ˇđ??ˇđ??ˇ

where V is amplitude in ÂľV, I is stimulus current in ÂľA and A-D are fitting parameters. R2 was between 0.91 and 1; 0.99 on average. From these parameters several variables were derived (see Fig. 2E): the maximum N1 amplitude (defined by B); the current level pter 6, Eq. 2 voor chapter 7: to attain 50% of the maximum N1 amplitude (defined by C); the slope at C (defined by đ??źđ??źđ??źđ??źđ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ??źđ??źđ??źđ??ź â&#x2C6;&#x2019;đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ą 0 B/4D); the threshold (defined by C-2D, the current level at which the tangent to the đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; = đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2122; ďż˝1 â&#x2C6;&#x2019; đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2019; đ?&#x153;?đ?&#x153;?đ?&#x153;?đ?&#x153;? ďż˝ curve at C crosses A); and the dynamic range (defined by 4D). The N1 peak latency, averaged over the three highest current levels, was analyzed in addition to these inputoutput characteristics.

pter 6, Eq. 3 voor chapter 7:

đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; = đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2122; ďż˝1 + đ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;? â&#x2C6;&#x2122; đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;

â&#x2C6;&#x2019;

ďż˝ â&#x2C6;&#x2122; ďż˝1 â&#x2C6;&#x2019; đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;

â&#x2C6;&#x2019;

ďż˝

pter 8: đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ą

177

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

A PD IPG PD

0.5

↓

100%

amplitude

signal amplitude (mV)

→

80% 70%

60%

50%

threshold

←

slop

1.0

dynamic range

level50%

1.5

↓

anodic first

2.0

90%

cathodic first

↔↔↔

N1−P2 amplitude (mV)

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 7

noise level

0 0% 20% 40% 60% 80% 100%

stimulation level (re max)

↑N1

40% 30% 20%

0.4 0.8 1.2 1.6 0.4 0.8 1.2 1.6 0.4 0.8 1.2 1.6 time since stimulus onset (ms)

Fig. 2. A Schematic representation of the pulses used for eCAP recordings. In single-pulse paradigms phase duration (PD) was 50 µs and inter-phase gap (IPG) varied between 2.1 and 30 µs; in multiple-pulse paradigms both PD and IPG were 30 µs. Pulse polarity was alternated to reduce the stimulus artifact in the recording. B-D eCAP examples for different current levels (percentage of maximum current level) for an NH animal (B), a BDNF0 animal (C) and a 14WD animal (D). E Example of an input–output function derived from the eCAP N1–P2 amplitude (filled circles); the solid line represents the fitted Boltzmann curve (see Eq. 1). The dashed lines indicate the eCAP characteristics that were derived from the Boltzmann equation (see section 2.3.2.1.).

2.3.2.2. Multiple-pulse paradigms Current levels for masker-probe and pulse train stimulation were 10% (for artifact reduction) and 100% of the previously defined maximum current level. Biphasic current pulses (30 µs PD and 30 µs IPG) were presented either in pulse pairs (maskerprobe stimulation) with variable masker-probe interval (MPI; 0.4-16 ms in 18 steps), or in 100-ms pulse trains with variable inter-pulse interval (IPI; 0.4-16 ms in 10 steps). The implant did not allow continuous recording, so that the stimulation paradigm had to be designed such that the eCAPs in response to both the masker and the probe (masker-probe stimulation) were recorded in separate (consecutive) runs. Similarly, the responses to the last ten pulses of the pulse train were recorded individually; first the response after N pulses in a 100-ms pulse train was recorded, then the responses after N-1 through N-9 pulses were separately recorded. In case of small IPIs/MPIs (≤ 1.2 ms) the response to the previous pulse was present in the recording, which was therefore recorded separately (if not already part of the paradigm) and subtracted from the initial 178

Neurotrophic treatment prevents SGC degeneration

recording. In Fig. 3 examples of eCAP recordings from a masker-probe paradigm (A) and from a pulse train paradigm (B) are shown for all used pulse intervals. A

maskerâ&#x2C6;&#x2019;probe

0.3 0.4 0.5 0.6 0.7 0.8

signal amplitude (mV)

MPI (ms)

0.9 1.0

1.2 1.4

1.6 1.8 2

3 4

8 12 16

0 0.4

0.8

1.2

1.6

pulse train

IPI (ms)

0.4 0.6

0.8 1.0

1.2 1.6 2

4 8

0 0.4

0.8

1.2

1.6

time since stimulus onset (ms) Fig. 3. Examples of eCAPs recorded with multiple-pulse paradigms. A Responses to the probe pulse in a masker-probe paradigm at maximum current level for various masker-probe intervals (MPIs) in a NH animal. B Responses to the last (Nth) pulse of a 100-ms pulse train at maximum current level for various inter-pulse intervals (IPIs) in a BDNF0 animal.

Thus obtained N1 amplitudes were normalized to the amplitude obtained with a single pulse at maximum current level, for each animal and each recording session separately. Figure 4 shows examples of normalized amplitudes for the last ten pulses of the pulse train. Fourier analysis of these ten amplitudes yielded the average eCAP amplitude (DC component; used for recovery function fitting) and the amplitude of the oscillation around this average (AC component â&#x20AC;&#x201C; i.e., amplitude of the most dominant frequency; used for amplitude modulation analysis).

179

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 7

pter 8:

normalized eCAP amplitude

A â&#x2C6;&#x2019; NH R1 1.0 IPI (ms) R2 0.8 16 R3 8.0 0.6 4.0 R4 2.0 R5 1.6 0.4 0.8 R6 0.6 0.2 0.4 R7 0 R8 B â&#x2C6;&#x2019; BDNF0 R9 1.0 R10 0.8 R11 0.6 R12 R13 0.4 R14 0.2 R15 0 R16 C â&#x2C6;&#x2019; 14WD R17 1.0 R18 0.8 R19 0.6 R20 R21 0.4 R22 0.2 R23 0 R24 Nâ&#x2C6;&#x2019;9 Nâ&#x2C6;&#x2019;8 Nâ&#x2C6;&#x2019;7 Nâ&#x2C6;&#x2019;6 Nâ&#x2C6;&#x2019;5 Nâ&#x2C6;&#x2019;4 Nâ&#x2C6;&#x2019;3 Nâ&#x2C6;&#x2019;2 Nâ&#x2C6;&#x2019;1 N pulse number R25 Fig. 4. eCAP amplitudes plotted for the last 10 pulses of pulse trains with different inter-pulse R26 pter 5, Eq. 3 voor chapter 6, Eq. 1 voor chapter 7: intervals (IPIs) at maximum current level for (A) an NH animal, (B) a BDNF0 animal, and (C) a R27 14WD animal. đ??ľđ??ľđ??ľđ??ľ R28 đ?&#x2018;&#x2030;đ?&#x2018;&#x2030;đ?&#x2018;&#x2030;đ?&#x2018;&#x2030;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; = đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; + đ??źđ??źđ??źđ??źâ&#x2C6;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; R29 After removal of the data for pulse intervals shorter than the longest pulse interval 1 â&#x2C6;&#x2019; đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2019; đ??ˇđ??ˇđ??ˇđ??ˇ R30 that resulted in sub-threshold N1 amplitude, recovery functions were fitted to the R31 remaining N1 amplitudes. Pulse train data were fitted with a commonly used exponential R32 function (e.g., Morsnowski et al., 2006; Botros and Psarros, 2010; Ramekers et al., pter 6, Eq. 2 voor chapter 7: R33 2014a): R34 đ??źđ??źđ??źđ??źđ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ??źđ??źđ??źđ??ź â&#x2C6;&#x2019;đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ą 0 R35 (Eq. 2) đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; = đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2122; ďż˝1 â&#x2C6;&#x2019; đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2019; đ?&#x153;?đ?&#x153;?đ?&#x153;?đ?&#x153;? ďż˝ , R36 R37 where eCAP is the normalized eCAP amplitude for a given inter-pulse interval IPI, A R38 is the maximum eCAP amplitude evoked by a probe pulse after a sufficiently long IPI R39 pter 6, Eq. 3 voor chapter 7:

180

đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; = đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2122; ďż˝1 + đ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;? â&#x2C6;&#x2122; đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;

â&#x2C6;&#x2019;

ďż˝ â&#x2C6;&#x2122; ďż˝1 â&#x2C6;&#x2019; đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;

â&#x2C6;&#x2019;

ďż˝

đ??ľđ??ľđ??ľđ??ľ

đ??źđ??źđ??źđ??źâ&#x2C6;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; đ??ˇđ??ˇđ??ˇđ??ˇ

1 â&#x2C6;&#x2019; đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2019;

Neurotrophic treatment prevents SGC degeneration

pter 6, Eq. 2 voor chapter 7:

đ??źđ??źđ??źđ??źđ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ??źđ??źđ??źđ??ź â&#x2C6;&#x2019;đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ą 0 (comparable to a single-pulse-evoked eCAP), Ď&#x201E; is the recovery time constant, and t0 is â&#x2C6;&#x2019; đ?&#x153;?đ?&#x153;?đ?&#x153;?đ?&#x153;? đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; = đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2122; â&#x2C6;&#x2019; đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; ďż˝1 ďż˝ the absolute refractory period. In accordance with previous findings (Ramekers et al., 2014a), masker-probe recovery functions deviated from this conventional recovery function as they displayed a non-monotonic course (Fig. 5A). We therefore used a two-component exponential pter 6, Eq. 3 voor chapter 7: adapted from Prijs et al. (1993) to describe masker-probe recovery: function đ?&#x2018;&#x20AC;đ?&#x2018;&#x20AC;đ?&#x2018;&#x20AC;đ?&#x2018;&#x20AC;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2019;đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ą 0 đ?&#x153;?đ?&#x153;?đ?&#x153;?đ?&#x153;? đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;

â&#x2C6;&#x2019;

ďż˝ â&#x2C6;&#x2122; ďż˝1 â&#x2C6;&#x2019; đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;

â&#x2C6;&#x2019;

ďż˝ ,

(Eq. 3)

where eCAP is the normalized eCAP amplitude evoked by the probe for a given maskerprobe interval MPI, A is the maximum eCAP amplitude evoked by a probe pulse after a sufficiently long MPI, c is a dimensionless constant defining the ratio between the two exponential components, Ď&#x201E;A is the recovery time constant of the first exponential, Ď&#x201E;B is đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ą the đ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;? recovery = đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2019;đ?&#x153;?đ?&#x153;?đ?&#x153;?đ?&#x153;?time constant of the second exponential, and t0 is the absolute refractory period. eCAP amplitude (re single pulse)

pter 8:

đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; = đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2122; ďż˝1 + đ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;? â&#x2C6;&#x2122; đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;

1.0

A â&#x2C6;&#x2019; maskerâ&#x2C6;&#x2019;probe

B â&#x2C6;&#x2019; pulse train

0.8 0.6

NH BDNF0 14WD fit NH fit BDNF0 fit 14WD

0.4 0.2 0 0.3

3 MPI (ms)

0.3

3 IPI (ms)

Fig. 5. Examples of recovery function fitting. A Recovery functions obtained with masker-probe paradigms were fitted with a double exponential function (Eq. 3) as exemplified for an NH animal, a BDNF0 animal, and a 14WD animal. B Recovery functions obtained with pulse train stimulation were fitted with a single exponential function (Eq. 2) as exemplified for the same three animals. All amplitudes were normalized to a single-pulse eCAP at maximum current level. Symbols indicate actual eCAP amplitude; solid lines represent fitted curves. MPI, masker-probe interval; IPI, inter-pulse interval.

2.4. Histology After completion of the final eCAP recording session, all animals were sacrificed and their right cochleas were harvested. Processing and analysis was largely performed as previously described by van Loon et al. (2013). In short, intra-labyrinthine cochlear fixation was achieved with a fixative of 3% glutaraldehyde, 2% formaldehyde, 1% 181

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 7

acrolein and 2.5% DMSO in a 0.08 M sodium cacodylate buffer. The cochleas were then decalcified, post-fixated and embedded in Spurr’s low-viscosity resin. After dividing the cochleas into two halves along a standardized midmodiolar plane, they were reembedded in fresh resin. From each cochlea three semi-thin (1 µm) sections were cut at 60-µm intervals which were stained with 1% methylene blue, 1% azur B, and 1% borax in distilled water. Using a Leica DC300F digital camera mounted on a Leica DMRA light microscope and a 40× oil immersion objective (Leica Microsystems GmbH, Wetzlar, Germany), micrographs of each transection of Rosenthal’s canal (2 basal, 2 middle, and 3 apical transections), as well as the organ of Corti were obtained. Within each transection of Rosenthal’s canal, the number of type-I and type-II SGCs was counted and packing density was averaged across all three sections. In two of the three sections, the average perikaryal area was determined for type-I SGCs with a visible nucleus using ImageJ (Version 1.42q; National Institutes of Health, Bethesda, MA, USA). Subsequently both packing density and perikaryal area were first averaged in order to obtain one value per cochlear location (base, middle and apex); these were then averaged to obtain a single value per cochlea. Since the likelihood of detecting an individual SGC depends on its perikaryal size, the average packing density was corrected for perikaryal size as previously described (Coggeshall and Lekan, 1996; van Loon et al., 2013). Hair cell presence was sampled in midmodiolar sections (Agterberg et al., 2008). Hair cell counts included hair cells with a nucleus, a cilia bundle, or a clear cochlear hair-cell-like outline.

2.5. Statistical analysis Differences in ABR threshold shifts were evaluated with a one-way analysis of variance (ANOVA; for differences among animal groups) or a paired-samples t test (within NH animals after implantation). Changes in SGC packing density and perikaryal area after implantation, ototoxic treatment, or BDNF treatment were assessed with oneway ANOVAs or t tests; Bonferroni correction was applied in case of post-hoc multiple comparison. The differences in SGC packing density in BDNF-treated animals for the three cochlear locations separately were assessed using linear mixed model analysis with cochlear location as covariate and time after deafening as factor, under the assumption of compound symmetry. For this analysis cochlear location was defined as the distance from the round window, relative to the length of the basilar membrane (e.g., first basal turn: 0.21; helicotrema: 1.0). Differences in absolute values of single-pulse eCAP characteristics (input-output characteristics and latency) were assessed with repeated measures (rm) ANOVA, using IPG as repeats, and for time points t = 6 (end of BDNF treatment) and t = 14 (end point for groups BDNF8, NH, and 14WD) separately. The differential effect of IPG increase 182

Neurotrophic treatment prevents SGC degeneration

on eCAP characteristics among groups was evaluated with one-way ANOVA for time points t = 6 and t = 14 separately. To assess the longitudinal effect of implantation on eCAP characteristics a linear mixed model was employed with IPG as covariate and time (immediately upon implantation and after osmotic pump removal) as factor; for this analysis only the NH eCAPs were used, so that neither deafening nor BDNF-treatment confounded the data. Differences in multiple-pulse eCAP characteristics (recovery and eCAP amplitude modulation) among groups were assessed with one-way ANOVA and post-hoc multiple comparison with Bonferroni correction. All statistical tests were performed with SPSS 20.0 for Windows (IBM, Armonk, NY, USA).

3. Results 3.1. Animal inclusion For several reasons outlined below one BDNF0 animal and two BDNF8 animals were excluded from the present study. Unless stated otherwise, all data presented here are obtained from the remaining 37 animals (NH, N = 6, 14WD, N = 6; BDNF0, N = 8; BDNF4, N = 10; BDNF8, N = 7).

3.1.1. ABR threshold shifts after deafening and implantation Click-evoked ABR threshold shifts two weeks after the ototoxic treatment were 81 dB on average (range: 59-89 dB) and did not significantly differ among the deafened groups [one-way ANOVA; F(3,29) = 0.64, P = 0.59]. Additional ABR recordings during the final eCAP recording session (6-14 weeks after deafening) did not show any further changes of the ABR threshold. A significant 10 dB ABR threshold shift (from 28 to 38 dB peSPL) was observed for the NH animalsâ&#x20AC;&#x2122; right implanted ears [paired-samples t test; t(5) = 2.8, P = 0.02], which was most likely caused by direct insertion-induced trauma to the organ of Corti or by secondary fibrous tissue growth around the electrode array in the scala tympani.

3.1.2. Hair cell counts A paired-samples t test showed that there were no differences in hair cell counts between BDNF-treated and contralateral untreated ears [inner hair cells (IHCs): t(23) = 1.0; P = 0.33; outer hair cells (OHCs): t(23) = 0.28; P = 0.78]. Hair cell counts from both ears were therefore averaged. In the NH group 100% IHCs and 98% OHCs were present on average. For the deafened animal groups IHC presence was between 0% and 2.2%; OHC presence was between 0% and 13%. Differences in OHC presence

183

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 7

among the deafened groups were statistically significant [one-way ANOVA; F(3,26) = 6.3; P = 0.002; post-hoc multiple comparison with Bonferroni correction showed that differences between BDNF0 and BDNF8 (P = 0.002) and between BDNF0 and 14WD (P = 0.023) were statistically significant]. Note that since there were no differences in OHC presence between BDNF8 and 14WD, OHC survival was most likely related to time after deafening.

3.1.3. Assessment of successful BDNF treatment Four weeks after cochlear implantation and onset of the treatment (termination of the BDNF0 group; pump removal for the other groups), the connection of the cannula to the osmotic pump was examined. In one BDNF8 animal the cannula was disconnected from the pump at this time point, and failure of BDNF treatment was confirmed with histological analysis after termination eight weeks later: the SGC packing density was similar to that of 14WD animals. In a second BDNF8 animal the cannula was partly ruptured at the flow moderator of the osmotic pump. As with the previous case, the SGC packing density of this animal was indistinguishable from that of 14WD animals. SGC packing densities from both animals were identified as statistical outliers in the BDNF8 group by SPSS, and both animals were hence excluded from further analysis.

3.1.4. Successful eCAP recordings Of the 38 animals that received treatment as planned, useful eCAPs could be recorded in all except for one BDNF0 animal, since the eCAP threshold for this animal was unusually high (i.e., near the maximum output voltage of the stimulator). Single-pulse and maskerprobe data presented here are obtained from the remaining 37 animals. Finally, proper eCAPs in response to pulse train could not be obtained in two BDNF0 animals and one BDNF4 animal, hence pulse-train data are presented for 34 animals. 3.2 Spiral ganglion cell histology

3.2.1. Light micrographs of Rosenthalâ&#x20AC;&#x2122;s canal In Fig. 6 representative examples of transections of Rosenthalâ&#x20AC;&#x2122;s canal in the upper basal turn are shown. SGCs in an NH cochlea are typically densely packed (Fig. 6A); six weeks after deafening substantial SGC loss was observed and the remaining cells appear smaller, as exemplified for the left untreated ear of a BDNF0 animal (Fig. 6B). Four-week treatment with BDNF of the contralateral ear of the same BDNF0 animal showed enhanced survival of SGCs and an increase in cell size (Fig. 6C). Fourteen weeks after deafening, SGCs in an untreated cochlea (14WD) had degenerated further (Fig. 6D) relative to the 6WD condition (Fig. 6B). The effect of BDNF eight weeks after cessation 184

Neurotrophic treatment prevents SGC degeneration

of the treatment is still evident (Fig. 6E), showing many more cells than in the untreated ear. Compared to the condition immediately after the end of the treatment period (Fig. 6C), the cell size was clearly reduced.

A – NH

B – 6WD

C – BDNF0

D – 14WD

E – BDNF8

50 µm

Fig. 6. Examples of transections of Rosenthal’s canal (upper basal turn) containing the SGCs in (A) an NH animal (right implanted cochlea), (B) a BDNF0 animal (left untreated cochlea), (C) the same BDNF0 animal (right implanted/BDNF-treated cochlea), (D) a 14WD animal (right implanted cochlea), and (E) a BDNF8 animal (right implanted/BDNF-treated cochlea). 185

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 7

3.2.2. Spiral ganglion cell packing density In Fig. 7 quantified SGC packing densities across all cochlear sections are shown in group averages as function of time after deafening. In addition to the implanted/treated right and contralateral left ears from the present study, data of the right ears of three groups from a previous study (Ramekers et al., 2014b) are shown in this figure. The guinea pigs in these groups were acutely implanted and did not receive any neurotrophic treatment. The first of these three groups consists of six normal-hearing animals represented by the blue circle at time point t = 0. The second consists of five two-weeks deaf (2WD) animals denoted by the black triangle at t = 2 â&#x20AC;&#x201C; the time point at which the animals in the present study were implanted. The third group contains six six-weeks deaf animals (6WD) and is denoted by the black triangle at t = 6. The remaining data points are the implanted right ears (solid lines; filled symbols) and untreated contralateral ears (dashed lines; open symbols) from the present study. There were no statistically significant differences among the three groups of normal-hearing ears [one-way ANOVA; F(2,15) = 0.75; P = 0.49], indicating that neither cochlear implantation nor infusion of PBS into the cochlea had any effect on the number of SGCs. SGC degeneration (black lines) was previously shown to be significant in the three acutely implanted groups (Ramekers et al., 2014b); here, it persisted beyond these six weeks as demonstrated by the 14WD group [t test 6WD-14WD (right ears only); t(10) = 2.6; P = 0.014]. Progressive SGC degeneration was successfully abolished during the four-week treatment with BDNF, and remarkably did not recommence four or eight weeks after cessation of the BDNF treatment. For a more detailed examination, the SGC packing density for the BDNF-treated ears is shown for the three cochlear locations separately in the inset in Fig. 7. There appeared to be location-dependent survival of SGCs after BDNF treatment: whereas the packing density in the basal region did not decrease after treatment cessation, it did slightly decrease from week 10 to 14 in the middle region, and apical SGCs appeared to be largely unaffected by the treatment even during the four-week treatment period itself. Linear mixed model analysis using cochlear location as covariate and time after deafening as factor (groups 2WD, BDNF0, BDNF4 and BDNF8) showed that the course of SGC packing density over time (i.e., among groups) was indeed dependent on cochlear location [interaction effect time Ă&#x2014; location; F(3,54) = 6.9; P < 0.001], while neither time after deafening nor location showed a clear significant main effect [location: F(1,54) = 2.6; P = 0.12; time: F(3,37) = 2.8; P = 0.052]. As expected, the SGC packing density of the unimplanted left ears of the deafened animals (BDNF0, BDNF4, BDNF8, 14WD; dashed lines) was similar to the implanted right ears of the deafened untreated animals (black solid lines). The somewhat larger difference between the left ears of the BDNF8 animals and both ears of the 14WD animals at t = 14 were not statistically significant [one-way ANOVA; F(2,16) = 2.7; P = 186

Neurotrophic treatment prevents SGC degeneration

0.09; post-hoc multiple comparison with Bonferroni correction did not yield significant differences]. SGC packing density (cells/mm2)

2000

treatment period

1500

NH BDNF0 BDNF4 BDNF8 untreated

1000

BDNFâ&#x2C6;&#x2019;treated animals

500

B M A

6 10 time after deafening (weeks)

Fig. 7. Group averages of spiral ganglion cell numerical packing density as function of time after deafening. Filled symbols connected with solid lines represent implanted right ears; open symbols connected with dashed lines represent unimplanted left ears. Blue symbols denote normal-hearing animals: at time point t = 0 NH animals (N = 6) from a previous study (Ramekers et al., 2014b); at t = 12 the NH animals from the present study. Degeneration of SGCs after deafening is shown in black triangles: the animal groups at time points t = 2 and t = 6 are from the aforementioned previous study (N = 5 and N = 6 respectively); the 14WD animals are shown at t = 14. The three BDNF-treated animal groups demonstrate prolonged protection from SGC degeneration after cessation of the BDNF treatment; their left unimplanted ears show SGC packing densities similar to those in deafened untreated ears. Inset: SGC packing densities in BDNF-treated ears shown for each cochlear location separately. The efficacy of BDNF treatment was location-dependent. (Note that connecting lines are shown purely for visualization purposes; connected data points represent separate groups of animals.) Error bars represent SEM. B, basal region; M, middle region; A, apical region.

3.2.3. Spiral ganglion cell mean perikaryal area Group averages of mean SGC perikaryal area are shown as function of time after deafening in Fig. 8. PBS infusion and chronic cochlear implantation did not affect SGC size in the NH animals [one-way ANOVA; F(2,15) = 0.75; P = 0.49]. As for SGC packing density in Fig. 7, previously obtained data on the initial decrease in cell size two and six weeks after deafening (2WD and 6WD) are shown in Fig. 8. Compared to 6WD animals and the untreated left ears of the BDNF8 animals, mean perikaryal area appeared to be larger for both ears of the 14WD animals â&#x20AC;&#x201C; differences were however not statistically significant [one-way ANOVA with groups 6WD, 14WD and BDNF8 left ears; F(2,21) = 2.8; P = 0.083; post-hoc multiple comparison with Bonferroni correction did not yield significant differences]. 187

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

BDNF treatment and subsequent treatment cessation did influence mean perikaryal area: immediately after treatment cells were much larger (by about 18% relative to the start of treatment), and 4 and 8 weeks later, the cells had a similar size again as at the start of treatment. These effects were significant [one-way ANOVA with groups 2WD, BDNF0, BDNF4 and BDNF8; F(3,26) = 5.3; P = 0.005; post-hoc multiple comparison with Bonferroni correction yielded significant differences between BDNF0 and BDNF4 (P = 0.018) and between BDNF0 and BDNF8 (P = 0.009)]. No significant differences among treated and untreated ears were found after treatment cessation [one-way ANOVA; F(3,31) = 1.4; P = 0.27; post-hoc multiple comparison with Bonferroni correction did not yield significant differences]. 250

SGC perikaryal area (Âľm )

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 7

225

NH BDNF0 BDNF4 BDNF8 untreated

200

175

150

treatment period

6 10 time after deafening (weeks)

Fig. 8. Group averages of spiral ganglion cell numerical mean perikaryal area as function of time after deafening â&#x20AC;&#x201C; analogous to Fig. 7. Filled symbols connected with solid lines represent implanted right ears; open symbols connected with dashed lines represent unimplanted left ears. Blue symbols denote normal-hearing animals: at time point t = 0 NH animals (N = 6) from a previous study (Ramekers et al., 2014b); at t = 12 the NH animals from the present study. Changes in mean SGC perikaryal area after deafening is shown in black triangles: the animal groups at time points t = 2 and t = 6 are from the aforementioned previous study (N = 5 and N = 6 respectively); the 14WD animals are shown at t = 14. BDNF treatment resulted in an increase in cell size; upon cessation of the treatment mean perikaryal area became indistinguishable from that of untreated cells. (Note that connecting lines are shown purely for visualization purposes; connected data points represent separate groups of animals.) Error bars represent SEM.

3.2.4. Distributions of spiral ganglion cell perikaryal area In order to assess the effects of deafening, BDNF treatment, and subsequent treatment cessation on cell size in more detail, the perikaryal area of individual SGCs are shown in histograms in Fig. 9. The main observation is that the apparent normal distribution 188

Neurotrophic treatment prevents SGC degeneration

of cell sizes in NH animals persists after deafening and is shifted toward the left. This notion implies that after deafening all SGCs become smaller (probably as a result of loss of neurotrophic support), rather than that a divergence emerges between the ~25% cells that survive fourteen weeks after deafening and those that degenerate during this period. BDNF-treated cells were noticeably larger than untreated deaf or NH animals, but mainly in the basal part of the cochlea. Although the distribution was somewhat skewed, there was no sign of bimodality of the distribution. Finally, after cessation of the BDNF treatment the mean cell size gradually decreased, approximating that of 14WD animals, and a normal distribution was preserved in these animals.

3.3. Single-pulse eCAP recordings Examples of eCAP recordings to various current levels are shown for an NH animal, a BDNF0 animal and a 14WD animal in Fig. 2B-D. Resulting input-output functions were fitted with a Boltzmann sigmoidal curve (Eq. 1, Fig. 2E). Group averages of the inputoutput characteristics thus obtained are shown in Fig. 10. Since single-pulse eCAPs were recorded at three time points during the experimental period (after implantation, after pump removal, and, depending on the treatment schedule, four or eight weeks later), the data shown in Fig. 10 are partly pooled across groups. First, eCAP recordings from all deafened animals are grouped together for the recording session immediately following implantation; second, eCAPs from all BDNF-treated animals at the end of the treatment period (final recording session for BDNF0; pump removal for BDNF4 and BDNF8 animals) are pooled. In addition, for various reasons eCAP recordings were not always possible for all animals for recording sessions immediately postoperatively, so that the number of animals per group for these sessions deviates from those in the final recording session. Early effects of chronic implantation were assessed for NH animals by comparing eCAPs recorded immediately after implantation with those obtained after removal of the osmotic pump four weeks later. Linear mixed model analysis, treating time as factor and IPG as covariate, showed statistically significant decreases in threshold [Fig. 10G, H; F(6.8,1) = 9.8; P = 0.017], level50% [Fig. 10M, N; F(8.1,1) = 11; P = 0.010] and latency [Fig. 10P, Q; F(7.3,1) = 45; P < 0.001]; amplitude, slope and dynamic range did not change significantly (for all three: P > 0.5).

189

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

BDNF8

basal

20% 10%

middle

apical

N = 1124 cells N = 7 animals

N = 409 cells N = 7 animals

N = 392 cells N = 7 animals

N = 1080 cells N = 10 animals

N = 801 cells N = 10 animals

N = 976 cells N = 10 animals

N = 1076 cells N = 8 animals

N = 624 cells N = 8 animals

N = 904 cells N = 8 animals

N = 1121 cells N = 19 animals

N = 827 cells N = 19 animals

N = 1034 cells N = 19 animals

N = 663 cells N = 10 animals

N = 417 cells N = 10 animals

N = 511 cells N = 10 animals

N = 789 cells N = 14 animals

N = 508 cells N = 14 animals

N = 725 cells N = 14 animals

N = 615 cells N = 5 animals

N = 307 cells N = 5 animals

N = 322 cells N = 5 animals

N = 2189 cells N = 12 animals

N = 1674 cells N = 12 animals

N = 1928 cells N = 12 animals

BDNF4

0% 20% 10%

BDNF0

0% 20% 10%

14WD

0% 20% 10%

10WD

0% 20% 10%

6WD

0% 20% 10%

2WD

0% 20% 10% 0% 20% NH

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 7

10% 0%

100

200

300

400

500

SGC perikaryal area (µm2)

100

200

300

400

500

SGC perikaryal area (µm2)

100

200

300

400

500

SGC perikaryal area (µm2)

Fig. 9. Distributions of SGCs according to their perikaryal area in 20-µm2 bins pooled across animals and plotted for three cochlear locations (columns) and for all animal groups (rows). Each histogram is plotted such that its total area represents 100% of all constituting cells. All cells that were included in the averages in Fig. 8 are used in these histograms: NH histograms contain cells from both ears from the NH group from the present study, and additionally from the right ears of 6 NH animals from a previous study (Ramekers et al., 2014b); the 6WD histograms contain cells from the left untreated ears from BDNF0 animals and from 6 6WD animals from the previous study; the 10WD data are formed by left untreated ears of BDNF4 animals; the 14WD data represent a combined pool from both ears from 14WD animals and the untreated left ears from BDNF8 animals. The number of cells and the number of animals from which these derived are shown in each respective plot. The dashed line indicates the mean SGC perikaryal area in NH cochleas, for each cochlear location separately.

190

Neurotrophic treatment prevents SGC degeneration

Differences in absolute eCAP characteristics among groups were evaluated for time points t = 6 and t = 14 separately, treating IPG as repeats in rm ANOVAs (see Table 1). Significant differences among groups were found for all characteristics except for threshold. In general, the differences were largest between NH and untreated deafened animals (Fig. 10). Amplitude and slope were larger for NH animals than for the deafened animals, dynamic range and level50% were larger for deafened than for NH animals, and latencies were longer for NH than for deafened animals For time point t = 14 the differences were often more pronounced than for t = 6 (Table 1). A highly significant main effect of IPG was found for all eCAP characteristics for both time points, among others increasing amplitude and slope, and lowering threshold and level50%. Table 1. Results from the repeated measures ANOVAs, applied separately for each eCAP characteristic, and for time points t = 6 and t = 14. amplitude slope

threshold

dynamic range

level50%

latency

time (weeks) 6 14 6 14

F 50 36

9.8 38

6 14

113 78

6 14

57 71

6 14

15 15

14 79

IPG

P < 0.001 < 0.001

F 0.57 14

< 0.001 < 0.001

1.3 3.3

0.005 < 0.001

2.4 12

< 0.001 0.001

4.8 0.40

< 0.001 < 0.001

6.2 9.8

< 0.001 < 0.001

3.0 4.4

group

P 0.58 < 0.001

0.11 < 0.001 0.30 0.064 0.019 0.68

0.068 0.030

0.007 0.003

Degrees of freedom for IPG (between): 1; for group (between): 2; degrees of freedom (within) at t = 6: 23; at t = 14: 16.

For all six eCAP characteristics the change as a result of IPG increase from 2.1 to 30 Âľs is shown in the right-hand column of Fig. 10. Differences in the magnitude of this change among groups (similar to the interaction effect IPG Ă&#x2014; group in the aforementioned rm ANOVAs) were assessed with one-way ANOVAs, the results of which are shown in Table 2. Differences among groups were found for amplitude (Fig. 10C) and latency (Fig. 10R) at t = 14, and for level50% (Fig. 10O) for both time points. Examining these four particular differences, one observes that there were no differences in the IPG effect between NH and BDNF-treated animals, while these two groups significantly deviated from untreated deaf animals (post-hoc multiple comparison with Bonferroni correction, Table 2). For none of the eCAP characteristics did the IPG effect change after cessation of the BDNF treatment [one-way ANOVAs with groups BDNF0, BDNF4 and BDNF8; P > 0.12]. 191

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

IPG 30 µs

∆ amplitude (re max)

1.0 0.5 0

0.1 0

22 20 18 16 10

8 6 4 2

22 20

192

20 10 0 0

−1 −2 −3 3

2 1 0

0.4 0 2

6 10 14 0 2 6 10 time after deafening (weeks)

−1 −2 30

∆ latency (µs)

0.5

0.3

∆ level

0.6

difference

50%

∆ threshold (dB)

∆ slope (µV/nC)

0.2

0.6 0.5 0.4 0.3 0.2 0.1 0

∆ dynamic range (dB)

IPG 2.1 µs

(dB)

amplitude (mV)

1.5

level50% (dB re 1 nC) dynamic range (dB) threshold (dB re 1 nC)

slope (mV/nC)

2.0

latency (ms)

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 7

20 10 0

−10

0 2 6 10 14 time after deafening (weeks)

NH BDNF0 BDNF4 BDNF8 untreated

Neurotrophic treatment prevents SGC degeneration

 Fig. 10. eCAP characteristics derived from fitting of the eCAP input-output functions with Boltzmann sigmoid curves plotted as function of time after deafening. Phase duration was 50 µs for all pulses; inter-phase gap (IPG) was 2.1 µs for the first column, and 30 µs for the second column. In the third column, the difference between the first two columns (the IPG effect; IPG 30 minus IPG 2.1 µs) is shown. eCAP characteristics are: maximum amplitude (A-C), slope at inflection point C (D-F), threshold (G-I), dynamic range (J-L), level50% (M-O), and latency (P-R). eCAPs were recorded directly after implantation (t = 2), after removal of the osmotic pump (t = 6), and at the end of the experimental procedure (t = 6 for BDNF0; t = 10 for BDNF4; t = 14 for NH, BDNF8 and 14WD). For visualization purposes the NH data at implantation is plotted at t = 0 (black cross) as a theoretical starting point at time of deafening for the deafened groups, and in addition plotted as a blue dotted line for reference. At time point t = 2 all deafened animals (all except NH) are grouped together (black triangle); these groups subsequently diverge into BDNFtreated (BDNF0, BDNF4 and BDNF8) and untreated animals (14WD) at t = 6. Data points at time points after t = 6 represent data from separate (unmixed) groups. Summaries of the statistical analysis of these data are shown in Tables 1 and 2. The gray bars indicate the BDNF or control treatment period. Error bars represent SEM. NH: N = 5 at t = 2 and t = 6, N = 6 at t = 14; untreated: N = 19 at t = 2, N = 5 at t = 6, N = 6 at t = 14; BDNF-treated: N = 16 at t = 6, N = 10 at t = 10, N = 7 at t = 14. Table 2. Results from one-way ANOVAs testing eCAP differences with IPG among groups, applied separately for each eCAP characteristic, and for time points t = 6 and t = 14. amplitude

slope

threshold

time (weeks) 6 14 6 14

6 14

F 0.51 15

0.83 0.69

0.56 0.64

dynamic range

6 14

0.75 2.8

latency

6 14

2.3 9.1

level50%

6 14

group

5.1 12

P 0.61 < 0.001 0.45 0.52

multiple comparison (P) NH-BDNF NH-deaf BDNF-deaf 1.0 0.97 1.0 0.42 < 0.001 0.003 1.0 1.0

0.58 0.54

1.0 0.92

0.014 < 0.001

1.0 1.0

0.49 0.091

0.13 0.003

1.0 1.0

0.60 1.0

1.0 1.0

0.64 0.84

0.72 0.12

1.0 0.28

1.0 1.0

0.044 0.005

0.13 0.005

1.0 1.0

0.017 < 0.001 0.62 0.006

Degrees of freedom for group (between): 2; degrees of freedom (within) at t = 6: 23; at t = 14: 16.

3.4. Masker-probe recovery functions

3.4.1. Summary of examples presented previously eCAP waveforms in response to a probe stimulus in a masker-probe paradigm are exemplified for an NH animal in Fig. 3A. The eCAP amplitude gradually decreased with increasingly shorter MPI, until it disappeared altogether for the shortest MPI used (0.3 ms). Recovery functions were constructed from these eCAP amplitudes. 193

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 7

For three individual animals masker-probe recovery functions are shown in Fig. 5A. For the NH and BDNF0 animals a peak is present for MPIs around 1 ms, giving the recovery function a non-monotonic appearance, and the curve of the 14WD animal shows a shoulder around 1 ms. Fitting of these data was therefore done with a twocomponent exponential function (Eq. 3), of which the first exponential described the peak around 1-ms MPI, and the second described the (slower) recovery, reaching saturation for an MPI of approximately 8 ms. Corresponding fits are plotted as solid lines in the figure.

3.4.2. Recovery function characteristics The average R2 for masker-probe recovery function fitting was 0.96. Group averages of the four parameters derived from the fitted exponential function are depicted in Fig. 11. The absolute refractory period t0 was ~0.5 ms for NH animals, which was significantly shorter than that for the 14WD and BDNF4 animals (~0.6 ms) [one-way ANOVA; F(4,31) = 3.4; P = 0.022; post-hoc multiple comparison with Bonferroni correction: NH-14WD, P = 0.048; NH-BDNF4, P = 0.036]. Recovery time constant τB was ~0.5 ms faster for BDNF0 animals than for NH or 14WD animals, and faster for BDNF4 animals than for NH animals [one-way ANOVA; F(4,31) = 6.1; P < 0.001; post-hoc multiple comparison with Bonferroni correction: NH-BDNF0, P = 0.003; NH-BDNF4, P = 0.014; 14WD-BDNF0, P = 0.024]. The facilitation component of Eq. 3, describing the peak around 1-ms MPI, was defined by facilitation constant c (defining the amplitude) and time constant τA. There were no statistically significant differences in c among groups [one-way ANOVA; F(4,31) = 1.0; P = 0.42]. Time constant τA was shorter for 14WD, BDNF0 and BDNF4 animals than for NH animals by about 0.2 ms [one-way ANOVA; F(4,31) = 5.0; P = 0.003; post-hoc multiple comparison with Bonferroni correction: NH-14WD, P = 0.007; NH-BDNF0, P = 0.023; NH-BDNF4, P = 0.004]. 3.5. Pulse trains

3.5.1. Summary of examples presented previously Figure 3B shows examples of eCAP recordings to the last (Nth) pulse of 100-ms pulse trains with different IPIs. As for the masker-probe example in Fig. 3A, the eCAP amplitude decreased with shorter pulse intervals, but this decrease was typically much faster for pulse trains than for masker-probe stimulation. The eCAP amplitudes for the last ten pulses of pulse trains with different IPIs are plotted in Fig. 4 for an NH, a BDNF0 and a 14WD animal. Although the amplitude decreased with higher pulse rates, the amplitudes of the last ten pulses were relatively stable over pulse number. For the 14WD animal an oscillating pattern emerged at high pulse rates (Fig. 4C). 194

Neurotrophic treatment prevents SGC degeneration

0.2 8 C 6 4 2 0

BD NH N BD F0 N BD F4 N F 14 8 W D

facilitation constant c

2.0 B 1.6 1.2

** *

0.8 0.4 0 0.6 D 0.5 0.4 0.3 0.2 0.1 0

* **

BD NH N BD F0 N BD F4 N F 14 8 W D

0.4

t0 (ms)

0.6

facilitation τA (ms)

0.8 A

recovery τ (ms)

Pulse train recovery functions for three individual animals are shown in Fig. 5B. The fitted exponential function (Eq. 2) is shown in solid lines.

Fig. 11. Group averages of masker-probe recovery characteristics. A Absolute refractory period t0; B recovery time constant τB; C facilitation time constant c; D facilitation time constant τA. NH, N = 6; BDNF0, N = 8; BDNF4, N = 10; BDNF8, N = 7; 14WD, N = 6. *, P < 0.05; **, P < 0.01. Error bars represent SEM.

3.5.2. Recovery function characteristics Pulse train recovery function fitting yielded recovery characteristics t0 (absolute refractory period) and τ (recovery time constant). Group averages of these characteristics are plotted in Fig. 12, showing noticeably longer time constants compared to those obtained with the masker-probe paradigm (t0 and τB). The BDNF-treated animals had longer t0 than the other groups by 0.5-0.8 ms, which was statistically significant for BDNF0 and BDNF4 animals versus 14WD animals [one-way ANOVA; F(4,29) = 5.0; P = 0.004; post-hoc multiple comparison with Bonferroni correction: 14WD-BDNF0, P = 0.006; 14WD-BDNF4, P = 0.014]. No significant differences were found for τ [one-way ANOVA; F(4,29) = 1.9; P = 0.14; post-hoc multiple comparison with Bonferroni correction did not yield significant differences]. 3.5.3. eCAP amplitude modulation For the 14WD animal in Fig. 4C the amplitude displayed an alternating pattern for responses to high-rate pulse trains. In Fig. 12C group averages of this eCAP amplitude modulation are depicted as percentage of the maximum eCAP amplitude. Amplitude modulation was significantly larger for 14WD than for NH animals, although overall 195

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

8 B

0.8

6 4

BD NH N BD F0 N BD F4 N F 14 8 W D

0.4

30%

20% 10%

BD NH N BD F0 N BD F4 N F 14 8 W D

1.2

Ď&#x201E; (ms)

1.6

amplitude modulation

differences among groups were not statistically significant [one-way ANOVA; F(4,29) = 2.4; P = 0.07; post-hoc multiple comparison with Bonferroni correction: NH-14WD, P = 0.02]. t0 (ms)

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 7

Fig. 12. Group averages of eCAP characteristics obtained with pulse train stimulation. A Absolute refractory period t0; B recovery time constant Ď&#x201E;; C eCAP amplitude modulation. NH, N = 6; BDNF0, N = 6; BDNF4, N = 9; BDNF8, N = 7; 14WD, N = 6. *, P < 0.05; **, P < 0.01. Error bars represent SEM.

3.6 eCAP characteristics in relation to SGC packing density Single-pulse eCAP characteristics for which the IPG effect was significantly different among groups (amplitude, level50% and latency; see Fig. 10 and Table 2) were evaluated with respect to their relation to SGC packing density. In addition, the relation between eCAP amplitude modulation and SGC packing density was assessed (Fig. 13). There was a significant correlation between these functional measures and SGC packing density for the NH and 14WD (i.e., untreated) groups. Except for amplitude (Fig. 13A), these correlations were highly similar to those found for acutely implanted animals (Ramekers et al., 2014b; indicated with dashed line). Overall, the BDNF-treated groups are located in close proximity to the regression lines. Importantly, however, for latency (Fig. 13C), and more noticeably level50% (Fig. 13B), these groups are below the line, suggesting that the SGC functionality of BDNF-treated SGCs is more similar to that of healthy SGCs in NH animals than their respective SGC packing densities would imply.

196

∆ level50% (dB)

0.6 0.4 0.2

0 30 C

amp modulation

∆ latency (µs)

∆ amplitude (re max)

Neurotrophic treatment prevents SGC degeneration

20 10 0

0 500

1500

NH BDNF0 BDNF4 BDNF8 14WD

−1 −2

30%

20% 10% 0%

0 500

SGC packing density (cells/mm )

1500

Fig. 13. Group averages of eCAP characteristics plotted as function of their average SGC packing density. A The change in amplitude as a result of IPG increase from 2.1 to 30 µs (see also Fig. 10); B the change in level50% as a result of IPG increase; C the change in latency as a result of IPG increase; D eCAP amplitude modulation obtained with pulse train stimulation. Solid lines are regression lines for NH and 14WD groups only; R2s for these regressions ranged from 0.60 to 0.71 (P ≤ 0.003). Dashed lines are regression lines from data previously obtained in similar experiments (Ramekers et al., 2014a, 2014b); these regressions showed statistically significant relationships for level50%, latency and amplitude modulation.

4. Discussion 4.1. Sustained SGC survival after treatment cessation Histological analysis of BDNF-treated cochleas first of all showed that SGC degeneration was averted during the BDNF treatment period (Fig. 7). This finding is in accordance with a multitude of previous studies, which showed similar results with a variety of neurotrophic factors (reviewed in Ramekers et al., 2012). More strikingly, no significant loss of SGCs was observed up to eight weeks after cessation of the treatment, which extends and strengthens previous findings in guinea pigs showing sustained preservation of SGCs two or four weeks after cessation of treatment with BDNF or GDNF (Maruyama et al., 2008; Agterberg et al., 2009; Fransson et al., 2010). It has been suggested that electrical stimulation of SGCs during weekly eABR recordings (as performed in all three aforementioned studies) may provide sufficient neurotrophic support once the treatment with neurotrophic factors is terminated (Agterberg et al., 2009), since Gillespie et al. (2003) reported rapid SGC degeneration after treatment cessation in the absence of eABR recordings. The present findings 197

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 7

however do not support this notion, since there was no electrical stimulation at all during the period after cessation of BDNF treatment and electrical stimulation during the treatment period was limited to a mere five minutes in total (note that extensive eCAP recordings were performed only at the end). Maruyama et al. (2008) suggested that the four-week treatment period might protect SGCs during a critical period following cochlear trauma, after which endogenous survival factors are able to provide sufficient support. Since the present study showed continued SGC degeneration beyond six weeks after deafening in untreated controls (in agreement with other long-term degeneration studies â&#x20AC;&#x201C; Webster and Webster, 1981; Dodson and Mohuiddin, 2000), yet none during the same period after deafening following BDNF treatment, the role of exogenous neurotrophic factors is arguably more than merely deferring degeneration of the SGC population until it is endogenously sustained. Rather, the results imply that the BDNF treatment causes a specific self-sustaining mechanism to arise. As we have argued elsewhere, it is conceivable that the continued survival of SGCs after treatment cessation involves the establishment of an autocrine neurotrophic mechanism (Ramekers et al., 2012). Autocrine BDNF signaling has previously been shown to mediate growth and survival of mature neurons (Davies and Wright, 1995; Kuribara et al., 2011).

4.2 A gradient in SGC survival along the spiral axis Although on average SGC packing density did not differ among the three BDNF-treated groups, a gradient in SGC survival emerged with time after treatment cessation (see inset in Fig. 7). This gradient in BDNF-mediated survival has been reported repeatedly, both immediately after four-week treatment (e.g., van Loon et al., 2013) and after treatment cessation (Agterberg et al., 2009). Several reasons for this gradient have been proposed, the most obvious of which is that BDNF infusion is generally done from a basal location in the cochlea, and that a perilymphatic concentration gradient of BDNF from base to apex leads to lower survival rates in the apical regions. This explanation however is unlikely since endogenous BDNF levels in normal-hearing guinea pigs are in the order of 10 pg/ml (Ito et al., 2005). Additionally, infusion with BDNF concentrations more than a thousand fold lower than used in the present study has been shown to be sufficient to induce SGC survival (Miller et al., 1997). A more likely explanation may follow from basoapical gradients in neurotrophic signaling in the mature cochlea. In the healthy mature cochlea expression levels of NT-3 are highest in the apex, while the gradient for BDNF is thought to be reversed, with highest expression in the cochlear base (Davis, 2003). Loss of hair cells leads to discontinuation of supply of both neurotrophins, and therefore administration of exogenous BDNF is likely to have a more pronounced effect in the basal region than in the apex. This notion is supported by the absence of a clear survival gradient when BDNF is co-administered with NT-3 (Wise et al., 2005) or with 198

Neurotrophic treatment prevents SGC degeneration

fibroblast growth factor (FGF; Glueckert et al., 2008), and by the finding that in the basal turn BDNF more potently protects SGCs than NT-3 does (Miller et al., 1997).

4.3 Spiral ganglion cell size The increase in SGC perikaryal area beyond that of SGCs in NH animals we observed in BDNF-treated ears is an inherent effect of the treatment commonly described in the literature (e.g., Agterberg et al., 2008). In contrast to the stabilized SGC packing density, the mean perikaryal area rather drastically decreased to become indistinguishable from untreated controls after treatment cessation. Agterberg et al. (2009) found that mean perikaryal area of BDNF-treated SGCs was similar to normal-hearing cells two weeks after cessation of BDNF treatment, which corresponds to an interpolated cell size two weeks after treatment cessation in the present study (Fig. 8), which indicates a gradual process over the course of several weeks. While both the increase in cell size and the protection from degeneration are unmistakably caused directly by the administered BDNF, the opposing response to treatment discontinuation strongly implies two separate mechanisms of action. Possibly, the changes in size are activity-dependent (mediated by neurotrophic signaling), and are as such separate from survival. This notion is supported by studies using GDNF, which have shown SGC preservation while the cell size was not different from untreated controls (Scheper et al., 2009; Fransson et al., 2010). An important finding additional to differences in mean perikaryal area is that in all stages after deafening and subsequent BDNF treatment these differences reflect a change in cell size for the entire population (Fig. 9). As previously argued by van Loon et al. (2013), the fact that the entire population of SGCs becomes smaller, while only a minority is lost with each time point after deafening, favors the existence of two separate mechanisms. Additionally, whereas after treatment cessation the entire population again becomes smaller â&#x20AC;&#x201C; seemingly reinstating the initial degeneration process â&#x20AC;&#x201C; SGC loss is not observed. 4.4 Effects of chronic implantation Quantitative histological analysis of SGCs in NH animals did not reveal any effects of the twelve weeks of implantation compared to the unimplanted contralateral ear. In all implanted ears (regardless of treatment) fibrosis or ossification was regularly observed in the scala tympani around the electrode array. Changes in electrophysiological measures in NH animals over time after implantation were the likely result of this tissue response to the array. ABR thresholds increased 10 dB on average. Implantation-related changes in eCAP measures were found with respect to threshold, level50% and latency, all of which were remarkably stabilized within four weeks after implantation. In human CI

199

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 7

recipients a decreased eCAP threshold has been reported at time of initial stimulation, relative to intra-operatively obtained thresholds (Hughes et al., 2001; Spivak et al., 2010), which is comparable to the present situation. Decreases in threshold after implantation have also been observed in our laboratory for the early wave complex of eABRs – which probably reflects auditory nerve activity – in normal-hearing and deafened guinea pigs (Agterberg et al., 2009, 2010; Havenith et al., 2011).

4.5 The effect of BDNF treatment on eCAP measures For all eCAP characteristics except threshold differences among groups were observed (Fig. 10; Table 1). For all characteristics there was a highly significant effect of IPG, and, as we have previously shown (Ramekers et al., 2014b), for some eCAP characteristics the magnitude of this effect correlated with neural survival in deafened animals. In the present study we first looked at the effect size of IPG for the various eCAP characteristics, which significantly varied among groups for amplitude, level50% and latency. More importantly, in case of significant differences in the IPG effect among groups, BDNFtreated animals were consistently similar to NH animals, while both groups were different from 14WD animals (Table 2). These results show that, even eight weeks after treatment cessation, BDNF-treated SGCs were – at least in part – functionally indistinguishable from healthy SGCs. Since the IPG effect size was previously shown to correlate with neural survival only in normal-hearing and deafened animals, the next question was therefore whether BDNF-treated ears would fit in with this correlation. The first step was to reproduce these correlations in NH and 14WD animals, which resulted in reasonably accurate replications (Fig. 13A-C). While the three BDNF-treated groups fitted well with the regression line for amplitude, they are positioned slightly below the regression lines for level50% and latency – toward the normal-hearing levels. These deviations from the predicted values based on SGC packing density are subtle, but they are consistent among the three BDNF-treated groups. Even though functional differences may in part be caused directly by the size of the SGC population, these results suggest that BDNF treatment has an effect additional to mere mediation of survival on SGC functionality that persists after cessation of the treatment. A possible mechanism by which BDNF might cause changes in SGC functionality is modulation of voltage-gated ion channel expression, distribution, and activity. In in vitro experiments Adamson et al. (2002) have shown that redistribution of ion channels constitutes the underlying mechanism causing basal SGCs to exhibit apical firing properties in case of treatment with NT-3, and causing apical SGCs to display basal firing properties after BDNF treatment. It is therefore possible that the changes we observed in SGC responsiveness after BDNF treatment reflected a more “basal” composition of 200

Neurotrophic treatment prevents SGC degeneration

the SGC population. It remains to be investigated whether for electrical stimulation of SGCs more basal-like or more apex-like properties are more appropriate, and therefore whether BDNF or NT-3 is the most suited neurotrophin to apply. Alternatively, as argued by Needham et al. (2012), it might be crucial to preserve the natural base-to-apex gradient in firing properties by co-administration of BDNF and NT-3.

4.6 The effect of BDNF treatment on temporal response properties Since sound encoding with CIs depends on high-frequency stimulation, a vital prerequisite for BDNF treatment to be suitable for clinical application is that it does not considerably alter temporal response properties of the SGC population. Using two separate stimulation protocols (masker-probe and pulse train stimulation) we have shown that differences between BDNF-treated ears and untreated ears (normalhearing or deafened) are not substantial (Figs. 11 and 12). Moreover, eight weeks after cessation of the treatment both t0 and the recovery time constant τ for both stimulation paradigms were not different from either untreated group. It is possible that the transient increase in SGC size caused by the BDNF treatment initially affected temporal responsiveness (differences between the untreated groups and the BDNF0 and BDNF4 groups were found for t0, τA and τB [masker-probe] and for t0 [pulse train]), since membrane capacitance is linearly related to soma diameter (Limón et al., 2005). Additionally, since BDNF treatment causes apical neurons to change from continuously firing to rapidly adapting under constant depolarization (Adamson et al., 2002), slower recovery may be arguably be a logical consequence of BDNF treatment. Consistent with previous findings in normal-hearing and deafened animals (Ramekers et al., 2014a), recovery functions for masker-probe stimulation displayed a non-monotonic course for all animals. As we have reasoned previously (Ramekers et al., 2014a), a possible mechanism underlying the increased eCAP amplitude for MPIs around 1 ms is a temporary increase in firing probability following recovery from refraction due to hyperpolarization-activated currents (Ih). The magnitude of this effect was similar between groups (Fig. 11C), which was consistent with our previous findings. The time constant τA of this facilitation decreased both after deafening and with BDNF treatment. In contrast, treatment of SGCs in vitro with both BDNF and NT-3 has been shown to slow down Ih activation kinetics (Needham et al., 2012). This discrepancy with our findings may be caused by the addition of NT-3 in their study. eCAP amplitude modulation for pulse trains with 0.6 ms IPI was previously found to increase with duration of deafness, and to correlate highly with SGC loss (Ramekers et al., 2014a). Accordingly, we observed a significant increase in amplitude modulation in 14WD animals compared to NH animals (Fig. 12C). When plotted against SGC packing density, amplitude modulation appears to be an accurate predictor for SGC 201

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 7

loss, regardless of BDNF treatment. As we have hypothesized previously (Ramekers et al., 2014a), with lower packing densities of SGCs in Rosenthal’s canal the current path to each individual cell will become more similar (i.e., more direct), thereby facilitating synchronization of SGCs to odd or even pulses.

4.7 Conclusion We have shown that brief treatment with BDNF leads to protection of SGCs for a period of time that well exceeds the treatment period itself. From this we argue that a specific self-sustaining mechanism must be triggered by the exogenous BDNF, and hypothesize that an autocrine neurotrophic loop is a likely candidate. Functional properties of the SGC population are largely preserved as well, without any indication of the BDNF effect wearing off eight weeks after treatment cessation. Although it remains to be investigated whether neurotrophic treatment – and specifically associated changes in SGC functionality – is in fact beneficial for CI performance in humans, at any rate the present findings have shown it to be clinically practicable since temporary treatment may be sufficient for long-lasting effects. Since there is no reason to assume that the long-lasting effect of temporary neurotrophic treatment on both survival and functionality is restricted to cochlear neurons, the present findings may be valuable for treatment strategies in relation to other neurodegenerative diseases which involve spatially restricted neuronal loss.

Acknowledgments

The authors would like to thank Ferry Hendriksen for histological processing, Koen van Riessen for histological analysis, René van de Vosse for technical support, and Roland Hessler at MED-EL, Innsbruck, for the electrode arrays. This work was supported by MED-EL GmbH, Innsbruck, Austria.

202

Neurotrophic treatment prevents SGC degeneration

References Adamson, C.L., Reid, M.A., Davis, R.L., 2002. Opposite actions of brain-derived neurotrophic factor and neurotrophin-3 on firing features and ion channel composition of murine spiral ganglion neurons. J. Neurosci. 22, 1385-1396. Agterberg, M.J.H., Versnel, H., De Groot, J.C.M.J., Smoorenburg, G.F., Albers, F.W.J., Klis, S.F.L., 2008. Morphological changes in spiral ganglion cells after intracochlear application of brainderived neurotrophic factor in deafened guinea pigs. Hear. Res. 244, 25-34. Agterberg, M.J.H., Versnel, H., de Groot, J.C.M.J., van den Broek, M., Klis, S.F.L., 2010. Chronic electrical stimulation does not prevent spiral ganglion cell degeneration in deafened guinea pigs. Hear. Res. 269, 169-79. Agterberg, M.J.H., Versnel, H., Van Dijk, L.M., De Groot, J.C.M.J., Klis, S.F.L., 2009. Enhanced survival of spiral ganglion cells after cessation of treatment with brain-derived neurotrophic factor in deafened guinea pigs. J. Assoc. Res. Otolaryngol. 10, 355-367. Atkinson, P.J., Wise, A.K., Flynn, B.O., Nayagam, B.A., Richardson, R.T., 2014. Viability of long-term gene therapy in the cochlea. Sci. Rep. 4, 4733. Botros, A., Psarros, C., 2010. Neural Response Telemetry Reconsidered: II. The Influence of Neural Population on the ECAP Recovery Function and Refractoriness. Ear. Hear. 31, 380-391. Chikar, J.A., Colesa, D.J., Swiderski, D.L., Di Polo, A., Raphael, Y., Pfingst, B.E., 2008. Over-expression of BDNF by adenovirus with concurrent electrical stimulation improves cochlear implant thresholds and survival of auditory neurons. Hear. Res. 245, 24-34. Coggeshall, R.E., Lekan, H.A., 1996. Methods for determining numbers of cells and synapses: a case for more uniform standards of review. J. Comp. Neurol. 364, 6-15. Davies, A.M., Wright, E.M., 1995. Neurotrophic Factors: Neurotrophin autocrine loops. Curr. Biol. 5, 723-726. Davis, R.L., 2003. Gradients of neurotrophins, ion channels, and tuning in the cochlea. Neuroscientist 9, 311-316. Dodson, H.C., Mohuiddin, A., 2000. Response of spiral ganglion neurones to cochlear hair cell destruction in the guinea pig. J. Neurocytol. 29, 525-537. Endo, T., Nakagawa, T., Kita, T., Iguchi, F., Kim, T.S., Tamura, T., Iwai, K., Tabata, Y., Ito, J., 2005. Novel strategy for treatment of inner ears using a biodegradable gel. Laryngoscope 115, 20162020. Ernfors, P., Van De Water, T., Loring, J., Jaenisch, R., 1995. Complementary roles of BDNF and NT-3 in vestibular and auditory development. Neuron 14, 1153-1164. Fayad, J.N., Linthicum Jr., F.H., 2006. Multichannel cochlear implants: relation of histopathology to performance. Laryngoscope 116, 1310-1320. Fransson, A., Maruyama, J., Miller, J.M., Ulfendahl, M., 2010. Post-treatment effects of local GDNF administration to the inner ears of deafened guinea pigs. J. Neurotrauma 27, 1745-1751. Fritzsch, B., Pirvola, U., Ylikoski, J., 1999. Making and breaking the innervation of the ear: neurotrophic support during ear development and its clinical implications. Cell Tissue Res. 295, 369-382. Gillespie, L.N., Clark, G.M., Bartlett, P.F., Marzella, P.L., 2003. BDNF-induced survival of auditory neurons in vivo: cessation of treatment leads to accelerated loss of survival effects. J. Neurosci. Res. 71, 785-790. Glueckert, R., Bitsche, M., Miller, J.M., Zhu, Y., Prieskorn, D.M., Altschuler, R.A., Schrott-Fischer, A., 2008. Deafferentation-associated changes in afferent and efferent processes in the guinea pig cochlea and afferent regeneration with chronic intrascalar brain-derived neurotrophic factor and acidic fibroblast growth factor. J. Comp. Neurol. 507, 1602-1621. 203

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 7

Havenith, S., Versnel, H., Agterberg, M.J.H., de Groot, J.C.M.J., Sedee, R.J., Grolman, W., Klis, S.F.L., 2011. Spiral ganglion cell survival after round window membrane application of brainderived neurotrophic factor using gelfoam as carrier. Hear. Res. 272, 168-177. Hughes, M.L., Vander Werff, K.R., Brown, C.J., Abbas, P.J., Kelsay, D.M., Teagle, H.F., Lowder, M.W., 2001. A longitudinal study of electrode impedance, the electrically evoked compound action potential, and behavioral measures in nucleus 24 cochlear implant users. Ear Hear. 22, 471486. Ito, J., Endo, T., Nakagawa, T., Kita, T., Kim, T.S., Iguchi, F. 2005. A new method for drug application to the inner ear. ORL J. Otorhinolaryngol. Relat. Spec. 67, 272-275. Kuribara, M., Hess, M.W., Cazorla, M., Roubos, E.W., Scheenen, W.J.J.M., Jenks, B.G., 2011. Brainderived neurotrophic factor stimulates growth of pituitary melanotrope cells in an autocrine way. Gen. Comp. Endocrinol. 170, 156-161. Leake, P.A., Hradek, G.T., Hetherington, A.M., Stakhovskaya, O., 2011. Brain-derived neurotrophic factor promotes cochlear spiral ganglion cell survival and function in deafened, developing cats. J. Comp. Neurol. 519, 1526-1545. Limón, A., Pérez, C., Vega, R., Soto, E., 2005. Ca2+-activated K+-current density is correlated with soma size in rat vestibular-afferent neurons in culture. J. Neurophysiol. 94, 3751–3761. Maruyama, J., Miller, J.M., Ulfendahl, M., 2008. Glial cell line-derived neurotrophic factor and antioxidants preserve the electrical responsiveness of the spiral ganglion neurons after experimentally induced deafness. Neurobiol. Dis. 29, 14-21. Miller, J.M., Chi, D.H., O’Keeffe, L.J., Kruszka, P., Raphael, Y., Altschuler, R.A., 1997. Neurotrophins can enhance spiral ganglion cell survival after inner hair cell loss. Int. J. Dev. Neurosci. 15, 631-643. Morsnowski, A., Charasse, B., Collet, L., Killian, M., Müller-Deile, J., 2006. Measuring the refractoriness of the electrically stimulated auditory nerve. Audiol. Neurootol. 11, 389-402. Needham, K., Nayagam, B.A., Minter, R.L., O’Leary, S.J., 2012. Combined application of brainderived neurotrophic factor and neurotrophin-3 and its impact on spiral ganglion neuron firing properties and hyperpolarization-activated currents. Hear. Res. 291, 1-14. Prado-Guitierrez, P., Fewster, L.M., Heasman, J.M., McKay, C.M., Shepherd, R.K., 2006. Effect of interphase gap and pulse duration on electrically evoked potentials is correlated with auditory nerve survival. Hear. Res. 215, 47-55. Prijs, V.F., Keijzer, J., Versnel, H., Schoonhoven, R., 1993. Recovery characteristics of auditory nerve fibres in the normal and noise-damaged guinea pig cochlea. Hear. Res. 71, 190-201. Ramekers, D., Versnel, H., Grolman, W., Klis, S.F.L., 2012. Neurotrophins and their role in the cochlea. Hear. Res. 288, 19-33. Ramekers, D., Versnel, H., Strahl, S.B., Klis, S.F.L., Grolman, W., 2014a. Recovery characteristics of the electrically stimulated auditory nerve in deafened guinea pigs: relation to neuronal status. Chapter 6 of this thesis. Ramekers, D., Versnel, H., Strahl, S.B., Smeets, E.M., Klis, S.F.L., Grolman, W., 2014b. Auditory-nerve responses to varied inter-phase gap and phase duration of the electric pulse stimulus as predictors for neuronal degeneration. J. Assoc. Res. Otolaryngol. 15, 187-202. Schaette, R., Kempter, R., 2008. Development of hyperactivity after hearing loss in a computational model of the dorsal cochlear nucleus depends on neuron response type. Hear. Res. 240, 5772. Scheper, V., Paasche, G., Miller, J.M., Warnecke, A., Berkingali, N., Lenarz, T., Stöver, T., 2009. Effects of delayed treatment with combined GDNF and continuous electrical stimulation on spiral ganglion cell survival in deafened guinea pigs. J. Neurosci. Res. 87, 1389-1399. Seyyedi, M., Viana, L.M., Nadol, J.B. Jr., 2014. Within-Subject Comparison of Word Recognition and Spiral Ganglion Cell Count in Bilateral Cochlear Implant Recipients. Otol. Neurotol. 35, 14461450. 204

Neurotrophic treatment prevents SGC degeneration

Shepherd, R.K., Coco, A., Epp, S.B., Crook, J.M., 2005. Chronic depolarization enhances the trophic effects of brain-derived neurotrophic factor in rescuing auditory neurons following a sensorineural hearing loss. J. Comp. Neurol. 486, 145-158. Shinohara, T., Bredberg, G., Ulfendahl, M., Pyykkö, I., Olivius, N.P., Kaksonen, R., Lindström, B., Altschuler, R., Miller, J.M., 2002. Neurotrophic factor intervention restores auditory function in deafened animals. Proc. Natl. Acad. Sci. USA 99, 1657-1660. Spivak, L., Auerbach, C., Vambutas, A., Geshkovich, S., Wexler, L., Popecki, B., 2011. Electrical compound action potentials recorded with automated neural response telemetry: threshold changes as a function of time and electrode position. Ear Hear. 32, 104-113. Spoendlin, H., 1975. Retrograde degeneration of the cochlear nerve. Acta Otolaryngol. 79, 266– 275. Staecker, H., Garnham, C., 2010. Neurotrophin therapy and cochlear implantation: translating animal models to human therapy. Exp. Neurol. 226, 1-5. van Loon, M.C., Ramekers, D., Agterberg, M.J.H., de Groot, J.C.M.J., Grolman, W., Klis, S.F.L., Versnel, H., 2013. Spiral ganglion cell morphology in guinea pigs after deafening and neurotrophic treatment. Hear. Res. 298, 17-26. Versnel, H., Agterberg, M.J.H., De Groot, J.C.M.J., Smoorenburg, G.F., Klis, S.F.L., 2007. Time course of cochlear electrophysiology and morphology after combined administration of kanamycin and furosemide. Hear. Res. 231, 1-12. Webster, M., Webster, D.B., 1981. Spiral ganglion neuron loss following organ of Corti loss: a quantitative study. Brain Res. 212, 17-30. West, B.A., Brummett, R.E., Himes, D.L., 1973. Interaction of kanamycin and ethacrynic acid. Severe cochlear damage in guinea pigs. Arch. Otolaryngol. 98, 32–37. Wise, A.K., Fallon, J.B., Neil, A.J., Pettingill, L.N., Geaney, M.S., Skinner, S.J., Shepherd, R.K., 2011. Combining cell-based therapies and neural prostheses to promote neural survival. Neurotherapeutics 8, 774-787. Wise, A.K., Richardson, R., Hardman, J., Clark, G., O’leary, S., 2005. Resprouting and survival of guinea pig cochlear neurons in response to the administration of the neurotrophins brainderived neurotrophic factor and neurotrophin-3. J. Comp. Neurol. 487, 147-165. Ylikoski, J., Wersall, J., Bjorkroth, B., 1974. Degeneration of neural elements in the cochlea of the guinea pig after damage to the organ of corti by ototoxic antibiotics. Acta Otolaryngol. Suppl. 326, 23-41. Zilberstein, Y., Liberman, M.C., Corfas, G., 2012. Inner hair cells are not required for survival of spiral ganglion neurons in the adult cochlea. J. Neurosci. 32, 405-410.

205

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

CHAPTER 8 Spiral ganglion cell loss following ototoxically induced hair cell loss: simultaneous rather than retrograde degeneration

Dyan Ramekers, Huib Versnel, Emma M. Smeets, Sjaak F.L. Klis, Wilko Grolman

Try your hardest to combat atrophy and routine. To question The Obvious and the given is an essential element of the maxim de omnibus dubitandum.

– Christopher Hitchens, “Letters to a young contrarian”

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 8

Abstract Severe damage to the organ of Corti leads to degeneration of the spiral ganglion cells (SGCs) which form the auditory nerve. This degeneration starts at the level of synaptic connection of the peripheral processes (PPs) of SGCs with the cochlear hair cells. It is thought that from this point SGC degeneration progresses in a retrograde fashion: PPs degenerate first, followed by the SGCs and their axons with a delay of several weeks to many months. Since evidence for this course of events is not unambiguous, we here aimed to provide a comprehensive account of the course of SGC degeneration in the guinea pig cochlea after ototoxic treatment. Histological analysis of six healthy and twenty-three deafened cochleas, used in previous studies, showed that the degeneration of SGCs and their peripheral and central processes was simultaneous rather than sequential. Correlation with the amplitude of the electrically evoked compound action potential (eCAP) showed that roughly all surviving SGCs in the deafened cochlea are functional, and that their contribution to the eCAP is larger than that of SGCs in large populations. The course of the degeneration process may vary among species, and may depend on the cause of deafness, but the present findings at least indicate that gradual retrograde degeneration is not an elemental process following damage to the organ of Corti.

208

Degeneration of peripheral and central processes

1. Introduction Severe trauma to the organ of Corti, including loss of inner hair cells and supporting cells, leads to progressive loss of spiral ganglion cells (SGCs; Ylikoski et al., 1974; Spoendlin, 1975; Webster and Webster, 1981; Versnel et al., 2007; Zilberstein et al., 2012). The rate of this degeneration varies among species and with type and severity of the trauma (Spoendlin, 1975). In several studies the degeneration is characterized as being retrograde, affecting the peripheral processes (PPs) before the SGC somata (Spoendlin, 1975, 1984; Leake and Hradek, 1988; Shepherd and Javel, 1997). Spoendlin (1975) identified damage to the unmyelinated portion of the PP as the initiation for retrograde degeneration in cats, after which it takes weeks for its myelinated part to degenerate, followed by the entire SGC within months. However, there have been reports of simultaneous rather than sequential retrograde degeneration (Hinojosa and Marion, 1983; Spoendlin and Schrott, 1989). Both human and animal studies reporting on the course of degeneration mainly involve case reports without accurate quantification. Our main goal was therefore to provide a detailed characterization of the course of SGC degeneration, by means of histological quantification of SGCs and their PPs in the healthy and ototoxically deafened guinea pig cochlea. In addition to analysis of the degeneration process of SGCs and their peripheral processes, we additionally examined their axons, or central processes (CPs). Histological studies of CP degeneration are scarce, although their role is relevant in understanding for example auditory nerve function to electrical stimulation through a cochlear implant. Since, in general, axons are not able to survive autonomously without their cell bodies, the SGC and its CP are thought to degenerate virtually simultaneously. Nonetheless, studies of the auditory nerve mainly consist of characterization of the surviving CPs, and do not involve quantification of the survival rate (Spoendlin and Schrott, 1989; Nadol, 1990; Shepherd and Javel, 1997). By determining the packing densities of SGC somata, and both their PPs and CPs between normal-hearing guinea pigs and guinea pigs 1, 2, 6, and 14 weeks after deafening, a detailed time course of the degeneration process of all three elements is presented. In addition, the functional contribution of the surviving elements was assessed by comparing their packing densities to the amplitude of the electrically evoked compound action potential (eCAP). On the one hand this comparison can be informative of the electrophysiological condition of the degenerating SGC population; on the other it may clarify which neural elements are essential for excitation by electrical stimulation by cochlear implants.

209

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 8

2. Methods 2.1. Animals and experimental design The histological and functional data presented in this study were obtained from several previous studies and have partly been presented before (van Loon et al., 2013; Ramekers et al., 2014a, 2014b). For detailed descriptions of the methods regarding surgical treatment and electrophysiological procedures we refer to those studies (see below). In total twenty-nine female albino guinea pigs were included in this study (strain: Dunkin Hartley; supplier: Harlan Laboratories, Horst, the Netherlands). The animals were treated according to one of five experimental procedures, as shown in Fig. 1 and described below. All experimental procedures were approved by the Animal Care and Use Committee of Utrecht University (DEC-UMC 03.04.036; DEC 2010.I.08.103). The normal hearing (NH; N = 6), two-weeks deaf (2WD; N = 6), and six-weeks deaf (6WD; N = 6) animals were implanted with an intracochlear electrode array shortly before termination and histological processing (Ramekers et al., 2014b). From these animals eCAP recordings are presented in this study. The one-week deaf animals (1WD; N = 5) did not receive any kind of surgery other than the deafening treatment one week before termination (van Loon et al., 2013); eCAP recordings were therefore not available for these animals. The remaining group of animals (14WD; N = 6) was chronically implanted with an intracochlear electrode array two weeks after deafening (Ramekers et al., 2014a). By means of an osmotic pump phosphate-buffered saline (PBS) was infused into the scala tympani for four weeks. This treatment served as a negative control treatment that did not lead to histological differences between the treated right and the untreated left ear (Ramekers et al., 2014a). These animals were euthanized eight weeks after this four-week treatment, once eCAP recordings were performed.

2.2. Deafening procedure After induction of anesthesia click-evoked auditory brainstem responses (ABRs) were recorded in order to confirm normal hearing. Deafening was executed by subcutaneous injection of kanamycin (Sigma-Aldrich, St. Louis, MO, USA; 400 mg/kg) and subsequent infusion of furosemide (Centrafarm, Etten-Leur, the Netherlands; 100 mg/kg) into the external jugular vein, which has been shown to eliminate the majority of both inner and outer hair cells (West et al., 1973; Versnel et al., 2007).

210

Degeneration of peripheral and central processes

NH 1WD 2WD 6WD 14WD

i/e d

(n = 6) (n = 5)

i/e

(n = 6)

i/t

i/e

PBS

(n = 6)

6 time in weeks

d = deafening i = implantation t = treatment onset c = treatment cessation e = eCAP recordings

(n = 6)

Fig. 1. Animal treatment schedule per experimental group. Except for the NH group, time is relative to deafening; the end of the bars represent termination of the animals. All animals except for the 1WD group were subjected to eCAP recordings before they were euthanized. The number of animals per group are shown in brackets.

2.3. Cochlear implantation and eCAP recordings The animals were anesthetized and their right bulla was exposed via a retro-auricular approach. The bulla was subsequently opened to expose the cochlea, in which a 1.window. Chronic implantation (14WD) was cochleostomy was drilled near Figure the round done with a two-contact electrode array and a cannula connected to the subcutaneously positioned osmotic pump. The electrode connector was fixed onto the skull with dental cement for eCAP recordings. Acute implantation (NH, 2WD, 6WD) was done with a four-contact electrode array. In both situations the most apical contact was used for stimulation and the most basal one for recording. eCAP recordings were performed with a MED-EL PULSARci100 cochlear implant (MED-EL GmbH, Innsbruck, Austria), controlled by a PC via a Research Interface Box 2 (RIB2; Department of Ion Physics and Applied Physics, University of Innsbruck, Innsbruck, Austria) and a National Instruments data acquisition card (PCI-6533, National Instruments, Austin, TX, USA). Stimuli consisted of biphasic current pulses with 50 Âľs phase duration and 30 Âľs inter-phase gap, which were presented with alternating polarity to reduce the stimulation artifact. Stimuli were presented at ten current levels, and the input-output curves were fitted with a Boltzmann sigmoid in order to obtain the maximum eCAP amplitude (Ramekers et al., 2014b). 2.4. Histological analysis For each animal only the right cochleas were used for the present study. Histological processing and analysis of SGC packing densities were largely performed as previously described by van Loon et al. (2013). In short, intra-labyrinthine cochlear fixation was achieved with a fixative of 3% glutaraldehyde, 2% formaldehyde, 1% acrolein and

211

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 8

2.5% DMSO in a 0.08 M sodium cacodylate buffer. The cochleas were then decalcified, post-fixated and embedded in Spurr’s low-viscosity resin. Staining was done with 1% methylene blue, 1% azur B, and 1% borax in distilled water. Processing of the cochleas in order to obtain sections suitable for determination of the packing density of the three components of the SGCs is described below. Note that the SGC somata are referred to as “SGCs”, while strictly this term includes both their peripheral and central processes.

2.4.1. Central processes Analysis of CPs was performed for ten of the twenty-nine animals: four NH, three 2WD, and three 6WD animals. Transverse semi-thin (1 µm) sections of the auditory nerve were consecutively cut by starting at the base of the internal acoustic meatus and moving in 20-µm steps in the direction of the cochlea until Scarpa’s ganglion was no longer visible and the auditory nerve could be visually separated from the vestibular nerve (Fig. 2A, B). Using a Leica DC300F digital camera mounted on a Leica DMRA light microscope and a 63× oil immersion objective (Leica Microsystems GmbH, Wetzlar, Germany), approximately 60 frames of the auditory nerve cross-section were obtained. These frames were reassembled and the resulting image was rotated such that the top corresponded with the dorsal and the right with the rostral side. Subsequently, a radial wheel-like grid was superimposed on the reconstructed image (Fig. 2B). The grid was positioned with its center on the center of mass of the nerve under investigation (as calculated by ImageJ [Version 1.42q; National Institutes of Health, Bethesda, MA, USA]). Its eight arms were comprised of frames measuring 30 by 30 μm running from the center in an outward direction. Every second frame was selected for analysis, resulting in a sample size of ~5% of the total auditory nerve. This method was adapted from the fractitioner method (Mayhew and Sharma, 1984) to take into account the tonotopical organization of the nerve (Friede, 1984). Using this sampling method it was therefore possible to approximate fiber origin (basal, medial or apical). Within each grid, the number of intact myelinated CPs was counted using ImageJ and the packing density (fibers/1000 μm²) was determined. Since it was conceivable that in case of substantial degeneration the nerve would collapse into itself, thereby increasing CP density but decreasing the total area of the nerve transection, CP packing density was corrected by multiplying it with the ratio of the area of each respective nerve transection and the average area for NH animals.

2.4.2. Spiral ganglion cells After dividing the cochleas into two halves along a standardized midmodiolar plane, they were re-embedded in fresh resin. From each cochlea five semi-thin sections (only 212

Degeneration of peripheral and central processes

three sections for 14WD animals) were cut at 30-Âľm intervals (Fig. 2A). Using a 40Ă&#x2014; oil immersion objective, micrographs of each transection of Rosenthalâ&#x20AC;&#x2122;s canal were obtained (2 basal, 2 middle, and 3 apical transections for NH, 2WD, 6WD and 14WD animals; 2 basal and 2 middle transections for 1WD animals). The number of type-I and typeII SGCs was determined and the packing density was averaged across sections. It was subsequently averaged in order to obtain one value per cochlear location (base, middle and apex); these averages were then averaged to obtain a single value per cochlea. Since the likelihood of detecting an individual SGC depends on its perikaryal size, the average packing density was corrected for perikaryal size as previously described (Coggeshall and Lekan, 1996; van Loon et al., 2013). 2.4.3. Peripheral processes

hapter 5, Eq. 3 voor chapter 6, Eq. 1 voor chapter 7: Processing and analysis with respect to PP packing densities have previously been

described by Waaijer etđ??ľđ??ľđ??ľđ??ľal. (2013). Semi-thin sections were cut parallel to the đ?&#x2018;&#x2030;đ?&#x2018;&#x2030;đ?&#x2018;&#x2030;đ?&#x2018;&#x2030;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; = đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; +at the level midmodiolar plane đ??źđ??źđ??źđ??źâ&#x2C6;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; of the osseous spiral lamina (OSL; Fig. 2C), for the â&#x2C6;&#x2019; 1 â&#x2C6;&#x2019; đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; basal, middle, and apical turnđ??ˇđ??ˇđ??ˇđ??ˇ separately. Micrographs were obtained using a 63Ă&#x2014; oil immersion objective, yielding an approximately 130-Âľm wide transection of the OSL containing the PPs (Fig. 2D). Packing densities were determined by delineating the bony boundaries of the OSL and subsequently dividing the number of PPs by the thus hapter 6, Eq. 2 voor chapter 7: obtained transection area of the OSL. đ??źđ??źđ??źđ??źđ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ??źđ??źđ??źđ??ź â&#x2C6;&#x2019;đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ą 0 đ?&#x153;?đ?&#x153;?đ?&#x153;?đ?&#x153;?

ďż˝

2.5. Data analysis and statistics All packing densities and eCAP amplitudes were normalized to the average values for the NH animals. Pearsonâ&#x20AC;&#x2122;s correlation coefficient was calculated to characterize the relation between PP and SGC packing densities, between CP and SGC packing hapter 6, Eq. 3 voor chapter 7: densities, and between eCAP amplitudes and SGC, PP or CP packing densities. Since the normalization causes all regression lines to go â&#x2C6;&#x2019;đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąthrough point (1,1), a slope of 1 đ?&#x2018;&#x20AC;đ?&#x2018;&#x20AC;đ?&#x2018;&#x20AC;đ?&#x2018;&#x20AC;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2019;đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ą 0 đ?&#x2018;&#x20AC;đ?&#x2018;&#x20AC;đ?&#x2018;&#x20AC;đ?&#x2018;&#x20AC;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; 0 â&#x2C6;&#x2019; â&#x2C6;&#x2019; đ?&#x153;?đ?&#x153;?đ?&#x153;?đ?&#x153;? đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; đ?&#x153;?đ?&#x153;?đ?&#x153;?đ?&#x153;? đ??ľđ??ľđ??ľđ??ľ â&#x2C6;&#x2122; â&#x2C6;&#x2019; đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; = đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; â&#x2C6;&#x2122; + đ?&#x2018;?đ?&#x2018;?đ?&#x2018;?đ?&#x2018;? â&#x2C6;&#x2122; đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019;đ?&#x2018;&#x2019; ďż˝ ďż˝1 ďż˝ ďż˝1 would mean a proportional decrease toward 0 for both variables; a slope steeper or shallower than 1 would imply that the decrease was disproportional â&#x20AC;&#x201C; for instance faster degeneration of PPs than of SGCs. The time course of degeneration for all three components of SGCs was determined hapter 8: by fitting of the data with an exponential function (see Versnel et al., 2007): đ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ąđ?&#x2018;Ą

(Eq. 1)

where pd is the packing density normalized to the average for NH animals, t is time in weeks, and Ď&#x201E; is the degeneration time constant. 213

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 8

All statistical analyses were performed using the Statistics Toolbox in MATLAB (version 7.11.0; Mathworks, Natick, MA, USA).

200 µm 100 µm

20 µm

50 µm

Fig. 2. A Midmodiolar section used for determination of SGC packing density in Rosenthal’s canal (indicated by arrows). The black line indicates the cutting plane through the auditory nerve used for analysis of CPs; the black square indicated the organ of Corti depicted in C. B Transection through the auditory nerve, overlaid by a radial grid used for CP packing density analysis (see section 2.4.1.). Inset: overview of the transection obtained by cutting along the black line in A; the black circle indicates the auditory nerve. C Transection through the organ of Corti, including the OSL through which the PPs protrude. The black line indicates the cutting plane through the OSL used for PP analysis, as shown in D.

3. Results 3.1. Examples of SGC degeneration Figure 2. Examples of micrographs showing PPs, SGCs and CPs are depicted in Fig. 3 for an NH (A-C), a 2WD (D-F), and a 6WD animal (G-I). The loss of PPs and SGCs two or six weeks after deafening is clearly visible from the absence of tissue within the bony boundaries. 214

Degeneration of peripheral and central processes

Although the loss of CPs is visible as well (Fig. 3C, F, I), it is much less evident because the spaces in between the myelinated processes are occupied – possibly tissue debris and macrophages. Another possible reason for the absence of empty space is that the auditory nerve may collapse into itself – a possibility which led us to correct all CP packing densities for the total nerve cross-section area.

PPs

SGCs

CPs

100 µm

20 µm

2WD

6WD 20 µm

Fig. 3. Examples of histological sections for an NH (A-C), a 2WD (D-F), and a 6WD animal (G-I). The left column shows the OSL containing the PPs in the basal turn; the middle column shows the SGCs in Rosenthal’s canal in the upper basal turn; the right column shows the CPs in a single frame of the radial grid (Fig. 2D).

3.2 Quantification and comparison of packing densities For NH animals SGC packing density was 1717, 1742, and 1554 cells/mm2 for basal, middle, and apical turns, respectively; the overall average was 1671 cells/mm2. These values were used to normalize all SGC packing densities. Likewise, normalization values for PP packing densities were 80 (basal), 71 (middle), 61 (apical), and 71 (average) fibers/1000 µm2; and 51 (basal), 56 (middle), 62 (apical), and 56 (average) fibers/1000 µm2 for CPs. Strong and highly significant correlations were observed between PP and SGC packing densities for all cochlear turns (R2 ≥ 0.85; P < 0.001; Fig. 4A, C, E, G). Moreover, the slope of the regression lines was close to 1, indicating a similar rate in degeneration for SGCs and their PPs. The correlation between CP and SGC packing densities was strong for the basal and middle turn as well (Fig. 4B, D), but weak and statistically non-

Figure 3.

215

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

significant for the apical SGCs (Fig. 4F). The slope of the regression lines was shallower than 1 (~0.5-0.75), which suggested faster degeneration of SGCs than of their CPs. 1.2 A

basal

1.2 B

1.0

0.8

0.6

0.4 R = 0.85 P < 0.001

NH 1WD 2WD 6WD 14WD linear fit y=x

1.2 D 1.0

0.8

0.6

0.4 2

R = 0.89 P < 0.001

1.2 E

apical

1.0 0.8 0.6

CP packing density (re normal)

1.0

0.2

R = 0.63 P = 0.006

middle

0.4

middle

0.4 0.2

R = 0.70 P = 0.002

1.2 F

apical

1.0 0.8 0.6 0.4

0.2

R = 0.89 P < 0.001

1.2 G

total

0.2

R = 0.21 P = 0.180

1.2 H

1.0

0.8

0.6

0.4

total

0.4

0.2

1.2 C

basal

0.4

0.2

PP packing density (re normal)

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 8

R = 0.94 P < 0.001

0.2 0.4 0.6 0.8 1.0 1.2

0.2 0

R2 = 0.64 P = 0.006

0.2 0.4 0.6 0.8 1.0 1.2

SGC packing density (re normal)

Fig. 4. Correlations between PP packing density (left column) or CP packing density (right column) and SGC packing density for the three cochlear locations or origins (basal [A, B], middle Figure 4. (G, H). Packing densities are normalized to [C, D], and apical [E, F]) and for the cochlear average the average values for NH animals, for each location separately. Dashed lines are regression lines â&#x20AC;&#x201C; R2 and P values for the correlations are shown in each plot. Solid lines are defined by y = x, and thereby visualize the relation between PPs/CPs and SGCs in case of simultaneous degeneration.

216

Degeneration of peripheral and central processes

The time course of degeneration for all three components of the SGC is shown in Fig. 5. Degeneration was found to be faster for SGCs (τ = 7.6 weeks) than for PPs (τ = 8.9 weeks) or CPs (τ = 10.8 weeks). This means that 63% (1-e-1) of SGCs was lost after 7.6 weeks, while it took 9 more days for PPs to reach this level, and another 2 weeks for CPs. 2

packing density (re normal)

1.2

SGC (τ = 7.6; R = 0.91) 2 PP (τ = 8.9; R = 0.93) 2 CP (τ = 10.8; R = 0.61)

1.0 0.8 0.6 0.4 0.2 0

0 1 2

time after deafening (weeks)

Fig. 5. Time course of degeneration of SGCs, PPs, and CPs. Fitting of these data with Eq. 1 yielded the τ (in weeks) and R2 values shown in the legend.

3.3 eCAP amplitude The maximum eCAP amplitude decreased after deafening, as exemplified in Fig. 6. The average eCAP amplitude for NH animals was 1.8 mV, and this value was used to normalize all eCAP amplitudes for further analysis.

Figure 5. P2

signal amplitude (mV)

0.5

NH 2WD 6WD 14WD 1.0

1.5

time since stimulus onset (ms)

2.0

Fig. 6. Examples of eCAP recordings in response to a 480-µA stimulus for four individual animals. The amplitude of the N1-P2 peak defined the eCAP amplitude. 217

Figure 6.

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Strong correlations between maximum eCAP amplitude and SGC or PP packing density were observed (Fig. 7A, B), whereas the correlation between eCAP amplitude and CP packing density was low and statistically non-significant (Fig. 7C). A remarkable trend for all three correlations is that for low packing densities the eCAP amplitude is higher than would be predicted if the contribution of each SGC would add linearly to the eCAP (compare regression lines with y = x in Fig. 7). eCAP amplitude (re normal)

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 8

1.2

SGC

NH 2WD 6WD 14WD linear fit y=x

1.0 0.8 0.6 0.4 0.2 0

R = 0.78 P < 0.001

0.2 0.4 0.6 0.8 1.0

R = 0.74 P < 0.001

0.2 0.4 0.6 0.8 1.0

R = 0.29 P = 0.108

packing density (re normal)

0.2 0.4 0.6 0.8 1.0 1.2

Fig. 7. Maximum eCAP amplitude plotted as function of across-cochlear averages of packing density of SGCs (A), PPs (B), and CPs (C). Packing densities and eCAP amplitudes are normalized to the average values for NH animals. Dashed lines are regression lines â&#x20AC;&#x201C; R2 and P values for the correlations are shown in each plot; solid lines represent y = x.

4. Discussion

4.1. Quantification of degenerationFigure 7. The use of packing density in an area confined within a bony structure as a means to sample the total population of SGCs and PPs is common practice. Subsequent comparison of packing densities between normal-hearing and deafened cochleas is then a straightforward method for the estimation of the extent of degeneration. Quantification of CPs in the auditory nerve by means of sampling faces the problem that there are no solid boundaries, except for the internal acoustic meatus through which the vestibular and facial nerve pass as well. The standardized grid which was overlaid onto the transection allowed for an objective sampling of the nerve, but as the examples in Fig. 3F, I suggested, the empty spaces in the nerve resulting from axonal loss appeared to be filled with surrounding CPs, causing the nerve to collapse and the packing density to be enhanced. By compensating for this shrinkage of the nerve transection area the CP packing density became slightly lower in the 2WD and 6WD animals, although the regression lines in Fig. 4B, D, F, H suggest CP packing density may still have been overestimated. This compensatory method also corrects the packing density for changes in CP diameter, which has been shown to be larger in deaf cats than 218

Degeneration of peripheral and central processes

in normal-hearing cats (Shepherd and Javel, 1997), but smaller in deaf humans than in normal-hearing controls (Spoendlin and Schrott, 1989). A possible shortcoming of this method of correcting the CP packing density is that it is applied identically for all frames of the grid, regardless of its location on the radial axis, whereas a more pronounced effect of the collapse may be expected at the center of the nerve.

4.2. Simultaneous degeneration It was expected that the SGCs would either degenerate at a slower or at a similar rate, based on the fact that neurites normally die within 1-2 days after the connection with the soma is lost (Mack et al., 2001). The regression lines resulting from the correlation between PP and SGC packing density approximated y = x (Fig. 4), which implies that degeneration of an SGC and its PP occurs simultaneously. The degeneration time course was slightly faster for SGCs than for both their peripheral and central processes (Fig. 5), which may be in the margins of error, and at any rate does not provide convincing evidence for prolonged survival of PPs or CPs in the absence of an SGC soma. Possibly, the longer time constants of degeneration for PPs and CPs partly reflect a time delay of the Wallerian (as opposed to retrograde) degeneration from the soma toward the level of the cutting planes in the OSL and the internal acoustic meatus.

4.3. Simultaneous versus retrograde degeneration The present findings do not support the idea that degeneration of SGCs after hair cell loss proceeds in a gradual retrograde fashion as a clear simultaneous loss of SGC and PPs was observed. This finding contrasts several previous findings. In deafened cats PPs have been reported to degenerate prior to SGC loss (Liberman and Kiang, 1978; Spoendlin, 1984; Leake and Hradek, 1988; Xu et al., 1993; Shepherd and Javel, 1997). However, some of these concern case reports, or do not provide quantification, and differences in survival rate are not always convincing. In deafened chinchillas, McFadden et al. (2004) found a gradual decrease in SGC packing density similar to that in the present study, but PPs were almost absent as soon as two weeks after deafening. Because the means of deafening varied widely among (and within) the aforementioned studies (noise exposure, intramuscular injections of neomycin, single co-administration of kanamycin and ethacrynic acid, repeated administration of kanamycin, intracochlear direct current stimulation, or co-administration of gentamycin and ethacrynic acid), the difference between the present findings and theirs are more likely to be species-related rather than dependent on the cause of deafness. In a study of the effect of neurotrophic treatment of ototoxically deafened cochleas, Wise et al. (2005) presented SGC packing densities and PP counts from normal-hearing, and 7 or 11 weeks deaf guinea pigs. The relation between PPs and SGCs approximated 219

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 8

y = x, showing that degeneration was simultaneous. As this finding is in accordance with the present study, it adds to the likelihood of a species-related difference in the degeneration process. It is generally believed that SGC degeneration in humans follows a retrograde course, but evidence for this is not unambiguous. Histopathological studies mainly present case reports and often do not quantify presence of PPs (e.g., Otte et al., 1978; Hinojosa and Marion, 1983; Hinojosa et al., 1987; Nadol et al., 1989; Lu and Schuknecht, 1994). While Hinojosa and Marion (1983) reported severe loss of PPs in four patients with SGC presence of up to 50%, this was not the case for all patients, and one even appeared to have more PPs than SGCs. Moreover, based on detailed counts in cochleas from deaf patients and normal-hearing controls, Spoendlin and Schrott (1989) concluded that a PP only disappears when the SGC degenerates. These reports based on histological quantifications are in line with the simultaneous degeneration reported in the present study.

4.4. eCAP amplitude as a predictor for SGC survival The maximum eCAP amplitude correlated well with SGC packing density. This finding is in agreement with a previously reported correlation between the eABR P1 peak and SGC counts (Hall, 1990). Maximum eCAP amplitude correlated equally well with PP packing density, so that these data cannot be used to determine the site of action potential initiation, which may be the SGC soma or its PP. An interesting observation is that for both SGCs and PPs (Fig. 7A, B) the amplitude is higher at low packing densities than would be expected based on a linear relation, which is often assumed in CAP convolution models (e.g., Earl and Chertoff, 2010). The deviation from this assumption at low packing densities may reflect one of two mechanisms. First, it is possible that the remaining SGCs are stimulated more uniformly, as the population becomes smaller. Their responses may therefore be more synchronized, which results in a sharper eCAP with slightly larger amplitude. Second, demyelination of the remaining SGCs (Agterberg et al., 2008) may lead to a larger amplitude for action potentials of individual SGCs as measured extracellularly.

4.5 Functional implications The presence of PPs is believed to be beneficial for cochlear implant stimulation, since their closer proximity to the electrode array may lead to lower thresholds and improved tonotopic specificity (argued by e.g., Wise et al., 2005; Waaijer et al., 2013). It may therefore be important to preserve PPs in cochlear implant users. In numerous animal studies SGCs are shown to be successfully protected from degeneration by treatment with neurotrophic factors (for a review, see Ramekers et al., 2012), and several studies 220

Degeneration of peripheral and central processes

have shown that this protection includes that of PPs (Wise et al., 2005; Glueckert et al., 2008; Waaijer et al., 2013). Apart from protection from degeneration, neurotrophic treatment has also been suggested as a means to induce regrowth of PPs toward the electrode array. Complete PP regrowth is obviously only of interest in case of slow retrograde degeneration, when SGCs are still abundant but their PPs have already degenerated. As the present findings show, this type of degeneration may not be as self-evident as it is often claimed.

Acknowledgments The authors would like to thank Ferry Hendriksen for histological processing, and Koen van Riessen for assistance with histological analysis. This work was supported by MEDEL GmbH, Innsbruck, Austria.

221

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 8

References

Agterberg, M.J.H., Versnel, H., De Groot, J.C.M.J., Smoorenburg, G.F., Albers, F.W.J., Klis, S.F.L., 2008. Morphological changes in spiral ganglion cells after intracochlear application of brainderived neurotrophic factor in deafened guinea pigs. Hear. Res. 244, 25-34. Coggeshall, R.E., Lekan, H.A., 1996. Methods for determining numbers of cells and synapses: a case for more uniform standards of review. J. Comp. Neurol. 364, 6-15. Earl, B.R., Chertoff, M.E., 2010. Predicting auditory nerve survival using the compound action potential. Ear. Hear. 31, 7-21. Friede, R.L., 1984. Cochlear axon calibres are adjusted to characteristic frequencies. J. Neurol. Sci. 66, 193-200. Glueckert, R., Bitsche, M., Miller, J.M., Zhu, Y., Prieskorn, D.M., Altschuler, R.A., Schrott-Fischer, A., 2008. Deafferentation-associated changes in afferent and efferent processes in the guinea pig cochlea and afferent regeneration with chronic intrascalar brain-derived neurotrophic factor and acidic fibroblast growth factor. J. Comp. Neurol. 507, 1602-1621. Leake, P.A., Hradek, G.T., 1988. Cochlear pathology of long term neomycin induced deafness in cats. Hear. Res. 33, 11-33. Hall, R.D., 1990. Estimation of surviving spiral ganglion cells in the deaf rat using the electrically evoked auditory brainstem response. Hear. Res. 49, 155-168. Hinojosa, R., Blough, R.R., Mhoon, E.E., 1987. Profound sensorineural deafness: a histopathological study. Ann. Otol. Rhinol. Laryngol. 96 (Suppl. 128), 43-46. Hinojosa, R., Marion, M., 1983. Histopathology of profound sensorineural deafness. Ann. N. Y. Acad. Sci. 405, 459-484. Liberman, M.C., Kiang, N.Y., 1978. Acoustic trauma in cats. Cochlear pathology and auditory-nerve activity. Acta Otolaryngol. Suppl. 358, 1-63. Lu, C.B., Schuknecht, H.F., 1994. Pathology of prelingual profound deafness: magnitude of labyrinthitis fibro-ossificans. Am. J. Otol. 15, 74-85. Mack, T.G.A., Reiner, M., Beirowski, B., Mi, W., Emanuelli, M., Wagner, D., Thomson, D., Gillingwater, T., Court, F., Conforti, L., Fernando, F.S., Tarlton, A., Andressen, C., Addicks, K., Magni, G., Ribchester, R.R., Perry, V.H., Coleman, M.P., 2001. Wallerian degeneration of injured axons and synapses is delayed by a Ube4b/Nmnat chimeric gene. Nat. Neurosci. 4, 1199-1206. Mayhew, T.M., Sharma, A.K., 1984. Sampling schemes for estimating nerve fibre size. I. Methods for nerve trunks of mixed fascicularity. J. Anat. 139, 45-58. McFadden, S.L., Ding, D., Jiang, H., Salvi R.J., 2004. Time course of efferent fiber and spiral ganglion cell degeneration following complete hair cell loss in the chinchilla. Brain Res. 997, 40-51. Otte, J., Schuknecht, H.F., Kerr, A.G., 1978. Ganglion cell populations in normal and pathological human cochleae. Implications for cochlear implantation. Laryngoscope 88, 1231-1246. Nadol, J.B., Jr., 1990. Degeneration of cochlear neurons as seen in the spiral ganglion of man. Hear. Res. 49, 141-154. Nadol, J.B., Jr., Young, Y.S., Glynn, R.J., 1989. Survival of spiral ganglion cells in profound sensorineural hearing loss: implications for cochlear implantation. Ann. Otol. Rhinol. Laryngol. 98, 411-416. Ramekers, D., Versnel, H., Grolman, W., Klis, S.F.L., 2012. Neurotrophins and their role in the cochlea. Hear. Res. 288, 19-33. Ramekers, D., Versnel, H., Strahl, S.B., Klis, S.F.L., Grolman, W., 2014a. Temporary neurotrophic treatment prevents deafness-induced auditory nerve degeneration and preserves functionality. Chapter 7 of this thesis. 222

Degeneration of peripheral and central processes

Ramekers, D., Versnel, H., Strahl, S.B., Smeets, E.M., Klis, S.F.L., Grolman, W., 2014b. Auditory-nerve responses to varied inter-phase gap and phase duration of the electric pulse stimulus as predictors for neuronal degeneration. J. Assoc. Res. Otolaryngol. 15, 187-202. Shepherd, R.K., Javel. E., 1997. Electrical stimulation of the auditory nerve. I. Correlation of physiological responses with cochlear status. Hear. Res. 108, 112-144. Spoendlin, H., 1975. Retrograde degeneration of the cochlear nerve. Acta Otolaryngol. 79, 266– 275. Spoendlin, H., 1984. Factors inducing retrograde degeneration of the cochlear nerve. Ann. Otol. Rhinol. Laryngol. 93 (Suppl. 112), 76-82. Spoendlin, H., Schrott, A., 1989. Analysis of the human auditory nerve. Hear. Res. 43, 25-38. van Loon, M.C., Ramekers, D., Agterberg, M.J.H., de Groot, J.C.M.J., Grolman, W., Klis, S.F.L., Versnel, H., 2013. Spiral ganglion cell morphology in guinea pigs after deafening and neurotrophic treatment. Hear. Res. 298, 17-26. Versnel, H., Agterberg, M.J.H., De Groot, J.C.M.J., Smoorenburg, G.F., Klis, S.F.L., 2007. Time course of cochlear electrophysiology and morphology after combined administration of kanamycin and furosemide. Hear. Res. 231, 1-12. Waaijer, L., Klis, S.F.L., Ramekers, D., Van Deurzen, M.H.W., Hendriksen, F.G.J., Grolman, W., 2013. The Peripheral Processes of Spiral Ganglion Cells After Intracochlear Application of BrainDerived Neurotrophic Factor in Deafened Guinea Pigs. Otol. Neurotol. 34, 570-578. Webster, M., Webster, D.B., 1981. Spiral ganglion neuron loss following organ of Corti loss: a quantitative study. Brain Res. 212, 17-30. West, B.A., Brummett, R.E., Himes, D.L., 1973. Interaction of kanamycin and ethacrynic acid. Severe cochlear damage in guinea pigs. Arch. Otolaryngol. 98, 32–37. Wise, A.K., Richardson, R., Hardman, J., Clark, G., O’leary, S., 2005. Resprouting and survival of guinea pig cochlear neurons in response to the administration of the neurotrophins brainderived neurotrophic factor and neurotrophin-3. J. Comp. Neurol. 487, 147-165. Ylikoski, J., Wersall, J., Bjorkroth, B., 1974. Degeneration of neural elements in the cochlea of the guinea pig after damage to the organ of corti by ototoxic antibiotics. Acta Otolaryngol. Suppl. 326, 23-41. Xu, S.A., Shepherd, R.K., Chen, Y., Clark, G.M., 1993. Profound hearing loss in the cat following the single co-administration of kanamycin and ethacrynic acid. Hear. Res. 70, 205-215. Zilberstein, Y., Liberman, M.C., Corfas, G., 2012. Inner hair cells are not required for survival of spiral ganglion neurons in the adult cochlea. J. Neurosci. 32, 405-410.

223

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

CHAPTER 9 General discussion

In short, he so bewildered himself in this kind of study, that he passed the nights in reading, from sunset to sunrise, and the days, from sunrise to sunset; and thus, through little sleep and much reading, his brain was dried up in such a manner, that he came at last to lose his wits. – Miguel de Cervantes, “Don Quixote”

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 9

The development of the cochlear implant (CI) has become an interdisciplinary success story. Close co-operation between otologists, neurosurgeons, electrical engineers, and human and animal physiologists in the 1950s-1970s resulted in the first successful partial restoration of hearing in deaf patients, thereby overcoming a widespread skepticism regarding the feasibility of electrical hearing (Eshraghi et al., 2012). At present, hearing performance of many CI users is better than could ever have been expected, owing to a continued effort of physicians, engineers and scientists to improve surgical techniques, to develop the implant, and to broaden our understanding of the biology of hearing loss and neurostimulation. As the expected outcome becomes ever more promising, the criteria for cochlear implantation are broadening, and nowadays include patients with significant residual low-frequency hearing (Stronks et al., 2011; Huarte and Roland, 2014). However, despite the major improvements there are still too many CI users that hardly benefit from their implant, and the neurobiological conditions underlying the huge differences in outcome are not well understood (Peterson et al., 2010). The studies described in this thesis contribute to this understanding. In particular, their focus is on the functional consequences of auditory nerve degeneration that is inherently related to severe sensorineural hearing loss. On the neurobiological level, this involved a detailed histological and electrophysiological characterization of the degeneration process, followed by a pharmacological intervention meant to prevent both numerical and functional decline of the spiral ganglion cells (SGCs). On a more clinically relevant note, objective functional measures predictive of neural loss have been identified, and the clinical applicability of neurotrophic treatment of the auditory nerve was assessed.

1. Histological assessment of spiral ganglion cell degeneration

The histological analyses of SGC degeneration presented in this thesis generally served two purposes. The first was to obtain quantifications of the degenerated neural substrate to correlate with electrophysiological characteristics in order to find predictive functional measures for SGC degeneration. This objective will be discussed in section 5 below. The second purpose was to extend the current knowledge on the degeneration process, which in itself is important in order to improve our understanding of CI functioning. Chapters 3, 4 and 8 (and chapter 7 to some extent) were dedicated to the latter purpose. The degeneration of SGCs following severe hair cell loss has been characterized extensively both in humans (e.g., Otte et al., 1978; Hinojosa and Marion, 1983; Hinojosa 226

General discussion

et al., 1987; Nadol et al., 1989; Spoendlin and Schrott, 1989, Nadol, 1990) and in animals (e.g., Ylikoski et al., 1974; Spoendlin, 1975, 1984; Liberman and Kiang, 1978; Webster and Webster, 1981; Leake and Hradek, 1988; Shepherd and Javel, 1997; Versnel et al., 2007). The results presented in chapter 3 add to this vast amount of data a detailed characterization of surviving SGCs in a population that is gradually becoming smaller. The main finding is that although the majority of the SGCs does not degenerate during the first four weeks after deafening, all SGCs become smaller almost immediately. On the one hand this finding underlines that changes in SGC functionality should not be attributed merely to changes in population size, but that morphological changes of the remaining SGCs should be taken into account as well. On the other hand it shows that a reduction in cell size is indicative of degeneration, without a direct causal relationship. The latter idea is supported by the finding described in chapter 7 that beyond six weeks after deafening the mean SGC size does not decrease further, whereas significant degeneration is still observed. It is suggested that both the rapid decrease in SGC size and the slower degeneration, although both probably directly caused by discontinued neurotrophic support, are manifestations of separate cellular mechanisms. Chapters 4 and 8 focus on the presence of peripheral processes (PPs) of SGCs. Using electron microscopy, chapter 4 provides a detailed characterization of these fibers. It is shown that the PP degeneration associated with severe hair cell loss is abolished by means of treatment with brain-derived neurotrophic factor (BDNF). Additionally, the cross-sectional area drastically decreases after deafening, but is kept at near-normal levels two weeks after cessation of the BDNF treatment. As it is thought that the closer proximity of PPs to the CI electrode array may be beneficial (Wise et al., 2005), these finding indicate that functional consequences of BDNF treatment extend beyond effects on the SGC soma alone. Since the effect of both deafening and subsequent BDNF treatment on survival and appearance seemed to be remarkably similar for SGCs and their PPs (compare chapter 3 with chapter 4), it was decided to study their relation in detail in several experimental groups from previous studies. In chapter 8 the degeneration of SGCs (i.e., the SGC somata) and their PPs was studied in four groups of deafened guinea pigs at several time intervals after deafening. Whereas it is widely believed that this degeneration progresses in a relatively slow retrograde fashion â&#x20AC;&#x201C; loss of PPs well before SGC loss â&#x20AC;&#x201C; this was undoubtedly not the case in our study: the results imply a virtually simultaneous degeneration. Critical evaluation of the literature on this topic furthermore led us to conclude that the evidence for retrograde degeneration of SGCs in both humans and in animal models is not overly convincing. These findings may have important consequences, since PPs are believed to have a functional role in electrical excitability by CI stimulation. 227

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 9

2. Histological effects of BDNF treatment

The literature review in chapter 2 provides an overview of the role of neurotrophic factors in the cochlea, which ranges from guidance of innervating fibers during embryonic development to maintenance of SGCs in the adult state. Its specific relevance to the experimental work described in this thesis is the effect of application of exogenous neurotrophic factors to the deafened cochlea. One of the main findings is that each of the wide variety of neurotrophic factors that have been applied in animal studies has a potent protective effect against SGC degeneration. Furthermore, we proposed that the sustained protective effect observed after cessation of treatment with either glial cell line-derived neurotrophic factor (GDNF; Maruyama et al., 2008; Fransson et al., 2010) or BDNF (Agterberg et al., 2009) might be the result of an autocrine support mechanism triggered by the treatment. In chapter 7 the effect of BDNF on SGC survival and functionality was assessed at three time points after the four-week treatment: directly afterwards, and four and eight weeks after cessation of the treatment. The histological findings presented in this chapter are similar to those in the three aforementioned studies, showing no significant SGC degeneration at any of these time points relative to the start of the treatment. Whereas the previous studies assessed SGC survival two weeks (Agterberg et al., 2009) or four weeks after treatment cessation (Maruyama et al., 2008; Fransson et al., 2010), our findings show that SGCs survive for at least eight weeks after treatment cessation, which greatly strengthens the hypothesis that the BDNF treatment activates a selfsustaining mechanism. Because the previous authors applied electrical stimulation by means of weekly electrically evoked auditory brainstem response (eABR) recordings, Agterberg et al. (2009) proposed that the continued protection of SGCs after treatment cessation might be a result of a synergetic effect between BDNF and electrical stimulation. However, in our study electrical stimulation of the SGCs during the treatment and afterwards was negligible, implying that BDNF treatment alone is sufficient to cause the sustained survival of SGCs. The effect of BDNF treatment on SGC survival and morphology was studied in chapters 3, 4 and 7. A consistent finding throughout these three studies is that the effect of BDNF is most prominent in the basal part of the cochlea, and gradually becomes less pronounced toward the apex, where in some instances the effect is completely absent. This gradient is observed for SGC survival (chapters 3 and 7), increase in both soma size and PP diameter (chapters 3, 4 and 7), intracellular density of the SGC soma (chapter 3), and PP myelination and shape (chapter 4). This gradient in BDNF effectiveness along the basoapical axis is consistent with previous findings with respect to survival (Shepherd et al., 2005; Rejali et al., 2007; Agterberg et al., 2008; Havenith et al., 2011; 228

General discussion

Leake et al., 2011), and with respect to SGC size (Shepherd et al., 2005). The most obvious explanation for this gradient would be that since BDNF is consistently applied in the basal turn, concentrations are highest in the base, and gradually decrease toward the apex. However, as argued by Agterberg et al. (2008), this explanation is unlikely since infusion of BDNF at thousand-fold lower concentrations than is often used has a potent protective effect (Miller et al., 1997). As we argue in chapter 7, an alternative explanation may follow from basoapical gradients in neurotrophic signaling in the mature cochlea. In the healthy mature cochlea expression levels of NT-3 are highest in the apex, while the gradient for BDNF is thought to be reversed, with highest expression in the cochlear base (Davis, 2003). Loss of hair cells leads to discontinuation of supply of both neurotrophins, and therefore administration of exogenous BDNF is likely to have a more pronounced effect in the basal region than in the apex.

3. Functional consequences of spiral ganglion cell degeneration In chapters 5, 6 and 7 the effect of degeneration on SGC functionality was assessed with electrically evoked compound action potentials (eCAPs). Subsequently, changes in SGC functionality were compared to quantified histological measures of SGC degeneration in order to find objective electrophysiological measures that can be used to estimate SGC degeneration. With longer duration of deafness the maximum eCAP amplitude and â&#x20AC;&#x201C; accordingly â&#x20AC;&#x201C; the slope of the eCAP growth function significantly decreased, which most probably reflects the smaller number of contributing SGCs. Neither in the acutely implanted guinea pigs (chapter 5) nor in the chronically implanted animals (chapter 7) did we observe differences in eCAP threshold between normal-hearing and deafened groups. This finding is at odds with several studies reporting higher excitation thresholds for eABR wave III in deafened animals (Shinohara et al., 2002; Yamagata et al., 2004; Shepherd et al., 2005; Chikar et al., 2008; Maruyama et al., 2008). It is, however, similar to findings by Agterberg et al. (2009), who observed similar thresholds for normalhearing and deafened guinea pigs of the eABR wave I, which is generally believed to correspond with the eCAP. For proper CI functioning a crucial requirement for SGCs is the ability to follow high pulse rate stimulation (250-2500 Hz; Middlebrooks, 2008). In chapters 6 and 7 we therefore studied the temporal response properties of the electrically stimulated SGCs, and, in particular, on changes in these properties that are associated with SGC degeneration. Using a two-pulse masker-probe stimulation paradigm, we observed only a slight increase in absolute refractory period of the eCAP fourteen weeks after 229

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 9

deafening. The time constant of the recovery (relative refractory period) was similar for normal-hearing and deafened guinea pigs. In contrast to these findings, recovery from 100-ms pulse trains was significantly faster for two of the three deafened groups than for normal-hearing controls. In chapter 6 we conclude that the use of pulse trains as opposed to masker-probe stimulation may yield a more useful and realistic measure for the temporal responsiveness of the SGC population.

4. Functional consequences of BDNF treatment Before neurotrophic factors can be considered for clinical treatment of the auditory nerve, their effects on SGC functionality must be thoroughly investigated. Whereas the effect on SGC survival has been confirmed repeatedly (discussed in chapter 2), the evaluation of functional consequences in vivo is limited to comparison of eABR thresholds, with only few exceptions (Maruyama et al., 2008; Agterberg et al., 2009; Havenith et al., 2011). This imbalance is surprising since the eventual goal of neurotrophic treatment is to ensure a properly functioning SGC population for CI stimulation, essentially regardless of the population size. In chapter 7 eCAP characteristics in a single-pulse paradigm and temporal response properties in multiple-pulse paradigms were compared among normal-hearing, deaf untreated, and deaf BDNF-treated guinea pigs. The maximum eCAP amplitude, slope, and latency for BDNF-treated groups were overall in between the respective values for normal-hearing and deaf controls; the threshold was similar for all groups. eCAP recovery assessed with masker-probe stimulation showed that recovery was faster for the BDNF-treated animals than for normal-hearing or deaf control groups, whereas recovery assessed with pulse train stimulation showed the opposite effect. Overall, SGC functionality was not greatly affected by the BDNF treatment. Wherever changes had occurred, these mainly meant that functionality was closer to that of healthy than of deaf animals.

5. Predictive value of eCAP measures for SGC degeneration

Although the maximum amplitude (see chapter 8) and slope of the eCAP growth curve correlated well with SGC survival, we argue in chapter 5 that these measures are not practicable for the estimation of neural survival in clinical settings. First, in order to obtain these eCAP measures, current levels are required that often exceed the maximum comfortable level for CI users. Second, these absolute measures are influenced by 230

General discussion

uncontrollable factors such as electrode impedance and the distance between the electrode and the SGC population, and by the design of the electrode array. We therefore used a method that is independent of these unpredictable factors. Introducing a time delay in between the two phases of the biphasic current pulse (inter-phase gap; IPG) results in a more efficacious stimulus, and the magnitude of this effect has been shown to correlate with SGC survival (Prado-Guitierrez et al., 2006). In chapter 5 we showed that this IPG significantly influences eCAP amplitude, slope of the growth curve, threshold, dynamic range, level50% (the current level needed to the half-maximum eCAP amplitude), and latency. By using SGC packing density and perikaryal area as predictors in a multiple regression model, we obtained moderate to high correlations between the IPG effect size and these quantified measures for SGC degeneration for, among others, slope, level50%, and latency. Based on these findings, we concluded that the magnitude of an IPG-induced change in eCAP morphology can be used to predict neural health. In chapter 7 we first reproduced the significant effect of IPG on all six eCAP characteristics, and subsequently showed that for chronically implanted normalhearing and deaf animals the correlation between the magnitude of the IPG effect and SGC packing density was highly similar to that presented in chapter 5 for latency and level50%. Moreover, the IPG effect for the BDNF-treated groups was more similar to normal-hearing animals than would be expected based on their respective SGC survival. Responses to pulse train stimulation were examined in chapters 6 and 7. Apart from recovery measures mentioned above, these responses displayed an amplitude modulated pattern for odd and even pulses, typically at inter-pulse intervals below 1 ms. The amplitude of this modulation, relative to the maximum amplitude, correlated highly with SGC packing density in both studies. Since this measure is an amplitude ratio, it is not subject to factors such as electrode impedance mentioned above, and may therefore be suitable for the estimation of the SGC population size.

6. Relevance to CI research and diagnostics The studies described in this thesis provide novel neurobiological insights concerning SGC degeneration and its prevention, and their respective effects on the functionality of the SGCs with regard to CI stimulation. These studies were designed to have their focus on clinical relevance and applicability. The next step will be to translate these insights into relevant and feasible tests to apply in human CI users. The first steps toward this goal are presently being taken in our department. 231

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Chapter 9

References

Agterberg, M.J.H., Versnel, H., De Groot, J.C.M.J., Smoorenburg, G.F., Albers, F.W.J., Klis, S.F.L., 2008. Morphological changes in spiral ganglion cells after intracochlear application of brainderived neurotrophic factor in deafened guinea pigs. Hear. Res. 244, 25-34. Agterberg, M.J.H., Versnel, H., Van Dijk, L.M., De Groot, J.C.M.J., Klis, S.F.L., 2009. Enhanced survival of spiral ganglion cells after cessation of treatment with brain-derived neurotrophic factor in deafened guinea pigs. J. Assoc. Res. Otolaryngol. 10, 355-367. Chikar, J.A., Colesa, D.J., Swiderski, D.L., Di Polo, A., Raphael, Y., Pfingst, B.E., 2008. Over-expression of BDNF by adenovirus with concurrent electrical stimulation improves cochlear implant thresholds and survival of auditory neurons. Hear. Res. 245, 24-34. Davis, R.L., 2003. Gradients of neurotrophins, ion channels, and tuning in the cochlea. Neuroscientist 9, 311-316. Eshraghi, A.A., Nazarian, R., Telischi, F.F., Rajguru, S.M., Truy, E., Gupta, C., 2012. The cochlear implant: historical aspects and future prospects. Anat. Rec. (Hoboken) 295, 1967-1980. Fransson, A., Maruyama, J., Miller, J.M., Ulfendahl, M., 2010. Post-treatment effects of local GDNF administration to the inner ears of deafened guinea pigs. J. Neurotrauma 27, 1745-1751. Havenith, S., Versnel, H., Agterberg, M.J.H., De Groot, J.C.M.J., Sedee, R., Grolman, W., Klis, S.F.L., 2011. Spiral ganglion cell survival after round window membrane application of brainderived neurotrophic factor using gelfoam as carrier. Hear. Res. 272, 168-177. Hinojosa, R., Blough, R.R., Mhoon, E.E., 1987. Profound sensorineural deafness: a histopathological study. Ann. Otol. Rhinol. Laryngol. 96 (Suppl. 128), 43-46. Hinojosa, R., Marion, M., 1983. Histopathology of profound sensorineural deafness. Ann. N. Y. Acad. Sci. 405, 459-484. Huarte, R.M., Roland, J.T., Jr., 2014. Toward hearing preservation in cochlear implant surgery. Curr. Opin. Otolaryngol. Head Neck Surg. 22, 349-352. Leake, P.A., Hradek, G.T., 1988. Cochlear pathology of long term neomycin induced deafness in cats. Hear. Res. 33, 11-33. Leake, P.A., Hradek, G.T., Hetherington, A.M., Stakhovskaya, O., 2011. Brain-derived neurotrophic factor promotes cochlear spiral ganglion cell survival and function in deafened, developing cats. J. Comp. Neurol. 519, 1526-1545. Liberman, M.C., Kiang, N.Y., 1978. Acoustic trauma in cats. Cochlear pathology and auditory-nerve activity. Acta Otolaryngol. Suppl. 358, 1-63. Maruyama, J., Miller, J.M., Ulfendahl, M., 2008. Glial cell line-derived neurotrophic factor and antioxidants preserve the electrical responsiveness of the spiral ganglion neurons after experimentally induced deafness. Neurobiol. Dis. 29, 14-21. Middlebrooks, J.C., 2008. Cochlear-implant high pulse rate and narrow electrode configuration impair transmission of temporal information to the auditory cortex. J. Neurophysiol. 100, 92-107. Miller, J.M., Chi, D.H., Oâ&#x20AC;&#x2122;Keeffe, L.J., Kruszka, P., Raphael, Y., Altschuler, R.A., 1997. Neurotrophins can enhance spiral ganglion cell survival after inner hair cell loss. Int. J. Dev. Neurosci. 15, 631-643. Nadol, J.B., Jr., 1990. Degeneration of cochlear neurons as seen in the spiral ganglion of man. Hear. Res. 49, 141-154. Nadol, J.B., Jr., Young, Y.S., Glynn, R.J., 1989. Survival of spiral ganglion cells in profound sensorineural hearing loss: implications for cochlear implantation. Ann. Otol. Rhinol. Laryngol. 98, 411-416. 232

General discussion

Otte, J., Schuknecht, H.F., Kerr, A.G., 1978. Ganglion cell populations in normal and pathological human cochleae. Implications for cochlear implantation. Laryngoscope 88, 1231-1246. Peterson, N.R., Pisoni, D.B., Miyamoto, R.T., 2010. Cochlear implants and spoken language processing abilities: review and assessment of the literature. Restor. Neurol. Neurosci. 28, 237-250. Prado-Guitierrez, P., Fewster, L.M., Heasman, J.M., McKay, C.M., Shepherd, R.K., 2006. Effect of interphase gap and pulse duration on electrically evoked potentials is correlated with auditory nerve survival. Hear. Res. 215, 47-55. Rejali, D., Lee, V.A., Abrashkin, K.A., Humayun, N., Swiderski, D.L., Raphael, Y., 2007. Cochlear implants and ex vivo BDNF gene therapy protect spiral ganglion neurons. Hear. Res. 228, 180-187. Shepherd, R.K., Coco, A., Epp, S.B., Crook, J.M., 2005. Chronic depolarization enhances the trophic effects of brain-derived neurotrophic factor in rescuing auditory neurons following a sensorineural hearing loss. J. Comp. Neurol. 486, 145-158. Shepherd, R.K., Javel, E., 1997. Electrical stimulation of the auditory nerve. I. Correlation of physiological responses with cochlear status. Hear. Res. 108, 112-144. Shinohara, T., Bredberg, G., Ulfendahl, M., Pyykkö, I., Olivius, N.P., Kaksonen, R., Lindström, B., Altschuler, R., Miller, J.M., 2002. Neurotrophic factor intervention restores auditory function in deafened animals. Proc. Natl. Acad. Sci. USA 99, 1657-1660. Spoendlin, H., 1975. Retrograde degeneration of the cochlear nerve. Acta Otolaryngol. 79, 266– 275. Spoendlin, H., 1984. Factors inducing retrograde degeneration of the cochlear nerve. Ann. Otol. Rhinol. Laryngol. 93 (Suppl. 112), 76-82. Spoendlin, H., Schrott, A., 1989. Analysis of the human auditory nerve. Hear. Res. 43, 25-38. Stronks, H.C., Versnel, H., Prijs, V.F., Grolman, W., Klis, S.F.L., 2011. Effects of electrical stimulation on the acoustically evoked auditory-nerve response in guinea pigs with a high-frequency hearing loss. Hear. Res. 272, 95-107. Versnel, H., Agterberg, M.J.H., De Groot, J.C.M.J., Smoorenburg, G.F., Klis, S.F.L., 2007. Time course of cochlear electrophysiology and morphology after combined administration of kanamycin and furosemide. Hear. Res. 231, 1-12. Webster, M., Webster, D.B., 1981. Spiral ganglion neuron loss following organ of Corti loss: a quantitative study. Brain Res. 212, 17-30. Wise, A.K., Richardson, R., Hardman, J., Clark, G., O’leary, S., 2005. Resprouting and survival of guinea pig cochlear neurons in response to the administration of the neurotrophins brainderived neurotrophic factor and neurotrophin-3. J. Comp. Neurol. 487, 147-165. Yamagata, T., Miller, J.M., Ulfendahl, M., Olivius, N.P., Altschuler, R.A., Pyykkö, I., Bredberg, G., 2004. Delayed neurotrophic treatment preserves nerve survival and electrophysiological responsiveness in neomycin-deafened guinea pigs. J. Neurosci. Res. 78, 75-86. Ylikoski, J., Wersall, J., Bjorkroth, B., 1974. Degeneration of neural elements in the cochlea of the guinea pig after damage to the organ of corti by ototoxic antibiotics. Acta Otolaryngol. Suppl. 326, 23-41.

233

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Nederlandse samenvatting

Maar stel je nou eens voor dat het niet hypothetisch zou zijn. – Hans Teeuwen, “Dat dan weer wel”

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Nederlandse samenvatting

De meest voorkomende oorzaak van ernstig gehoorverlies is dat geluidstrillingen die de cochlea (het binnenoor) bereiken niet omgezet worden in elektrische signalen. Normaal gesproken wordt deze omzetting bewerkstelligd door de cochleaire binnenste haarcellen; aangeboren dysfunctie van deze haarcellen, of het afsterven door lawaaischade, infectie of ototoxische geneesmiddelen, kan leiden tot volledige doofheid. Deze vorm van gehoorverlies wordt perceptief gehoorverlies genoemd. Een cochleair implantaat (CI) vervangt de functie van deze haarcellen, door geluid om te zetten in elektrische signalen die rechtstreeks aan de spirale ganglioncellen (SGC’s) gepresenteerd worden. De axonen van deze SGC’s vormen de gehoorzenuw, die de auditieve signalen aan de hersenstam doorgeven. Waar in de gezonde humane cochlea elk van de ongeveer 3500 binnenste haarcellen geluidstrillingen met een specifieke frequentie doorgeeft aan een enkele SGC, is die specificiteit veel lager met tot 24 elektroden op een CI. Desondanks zijn mensen met een CI veelal goed in staat om spraak te verstaan. In de gezonde cochlea produceren de haarcellen en omliggende steuncellen in het orgaan van Corti neurotrofe factoren; deze neurale groeifactoren kunnen gezien worden als het essentiële signaal voor de SGC’s om in leven te blijven, zolang er haarcellen zijn die de aanwezigheid van de SGC’s nut geven. Wanneer de haarcellen in groten getale vernietigd of anderszins afwezig zijn, ontbreekt ook dit neurotrofe signaal, en als gevolg hiervan sterven de SGC’s geleidelijk af. Dit natuurlijke proces is in principe nuttig omdat de SGC’s in afwezigheid van haarcellen geen functie hebben; het voortbestaan van SGC’s is daardoor een verspilling van de energie en bouwstoffen nodig voor het onderhoud van deze cellen. Echter, dit proces is uiteraard nadelig bij cochleaire implantatie, waarvoor de SGC’s van essentieel belang zijn. Het aantal overlevende SGC’s is dan ook gecorreleerd met spraakverstaan in CI-gebruikers. Het doel van het onderzoek beschreven in dit proefschrift is tweeledig. Enerzijds is onderzocht wat de effecten zijn van het toedienen van de neurotrofe factor brainderived neurotrophic factor (BDNF) op de degeneratie van de SGC’s in doofgemaakte cavia’s. Daarbij is de praktische haalbaarheid van mogelijke klinische toepassing bij CIgebruikers in ogenschouw genomen. Anderzijds zijn de veranderingen in functionaliteit van de SGC’s als gevolg van degeneratie en als gevolg van behandeling met BDNF onderzocht. Cavia’s werden doofgemaakt door middel van een behandeling met een hoge dosis kanamycine – een antibioticum dat specifiek de cochleaire haarcellen doodt. Functionaliteit van de SGC’s werd gemeten met behulp van de elektrisch opgewekte samengestelde actiepotentiaal (electrically evoked compound action potential; eCAP). De eCAP is de optelsom van potentiaalverschillen die door de elektrische activiteit van individuele SGC’s worden veroorzaakt, als gevolg van elektrische stimulatie met een CI, en gemeten op een korte afstand van de SGC’s met een CI elektrode. Veranderingen in de eCAP als gevolg van het aanpassen van diverse variabelen van de elektrische pulsen werden vervolgens gerelateerd aan met histologie bepaalde maten voor degeneratie. 236

Nederlandse samenvatting

Hoofdstuk 2 betreft een literatuurstudie naar de functie van neurotrofe factoren in de gezonde cochlea, naar de effecten van het verlies van neurotrofe factoren bij schade aan het orgaan van Corti, en naar de toediening van neurotrofe factoren in proefdiermodellen van perceptief gehoorverlies. Een van de belangrijkste conclusies van deze studie is dat het beschermende effect van neurotrofe factoren herhaaldelijk is aangetoond met betrekking tot het aantal overlevende SGC’s, maar dat de functionele consequenties ernstig onderbelicht blijven. Een tweede conclusie betreft de duur van de neurotrofe behandeling. Meerdere studies hebben aangetoond dat het beschermende effect van een vier weken durende behandeling met neurotrofe factoren aanhoudt tot minstens vier weken na het stopzetten van de behandeling. Enerzijds heeft deze bevinding belangrijke consequenties voor mogelijke toepassingen bij CI-gebruikers, aangezien continue toediening van neurotrofe factoren klinisch onpraktisch en mogelijk op de lange termijn zelfs schadelijk kan zijn. Anderzijds is dit een wetenschappelijk interessante bevinding, omdat het impliceert dat de SGC’s autonoom overleven na het beëindigen van de neurotrofe behandeling. In hoofdstuk 3 wordt de morfologie (vorm en uiterlijk) van SGC’s beschreven. Het is bekend dat SGC’s kleiner worden in de afwezigheid van haarcellen, en ook dat deze zelfs groter dan normaal worden na behandeling met BDNF. In dit hoofdstuk geven we een gedetailleerde beschrijving van de morfologische verandering van SGC’s na het vernietigen van de cochleaire haarcellen, en vervolgens na het behandelen met de neurotrofe factor BDNF. Er werd gevonden dat SGC’s één week na het doofmaken al gemiddeld 10% kleiner waren geworden, terwijl de eerste tekenen van degeneratie pas na twee weken optraden. Daarnaast werd gevonden dat het beschermende effect van de BDNF-behandeling sterker was aan het begin van de cochlea (basaal) dan bovenin de cochlea (apicaal). In apicale locaties waar geen significante reductie van SGCdegeneratie plaatsgevonden had, zagen we niettemin dat de overlevende SGC’s duidelijk groter waren dan onbehandelde SGC’s. Op basis van deze bevindingen betogen we dat hoewel celgrootte gerelateerd is aan overleving – en hoewel beide afhankelijk zijn van de beschikbaarheid van neurotrofe factoren – deze waarschijnlijk het gevolg zijn van twee verschillende intracellulaire processen.

In hoofdstuk 4 worden de perifere uitlopers (peripheral processes; PP’s) van de SGC’s bestudeerd. Deze PP’s zijn te zien als smalle axonen, die in de intacte cochlea synaptisch contact maken met de haarcellen. Na verlies van haarcellen is dit het deel van de SGC’s waarvan gedacht wordt dat het als eerste afsterft. Voor CI-onderzoek zijn deze uitlopers interessant omdat ze dichter bij de CI-elektroden liggen dan de cellichamen van de SGC’s; hierdoor is stimulatiedrempel voor een SGC met PP wellicht lager dan voor 237

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Nederlandse samenvatting

een SGC zonder PP. In deze studie hebben we met behulp van elektronenmicroscopie aangetoond dat behandeling met BDNF de PP’s beschermt tegen degeneratie. Tevens zijn de met BDNF behandelde PP’s morfologisch vergelijkbaar met die in gezonde cochlea’s. De isolerende myelineschede was significant dunner na de BDNF-behandeling, wat functionele gevolgen kan hebben voor elektrische stimulatie.

In hoofdstukken 5 en 6 worden experimenten beschreven waarin met behulp van eCAP-metingen bij normaalhorende en doofgemaakte cavia’s met een CI de functionele consequenties van SGC-degeneratie zijn bestudeerd. In de experimenten beschreven in hoofdstuk 5 werd de eCAP opgewekt met een enkele bifasische stroompuls vanuit het CI, en werden verscheidene karakteristieken van de eCAP, zoals de maximale amplitude, de drempel, en de latentie, bepaald. Door het introduceren van een korte pauze tussen de twee fasen van de stroompuls, werd deze significant efficiënter (onder andere lagere drempel en hogere amplitude). De grootte van dit effect op onder meer de latentie verschilde tussen normaalhorende en dove cavia’s, en bleek een hoge correlatie te hebben met het aantal overlevende SGC’s en met de gemiddelde celgrootte. Deze bevindingen kunnen van groot nut zijn bij de bepaling van de mate van SGC-degeneratie bij CI-gebruikers. In hoofdstuk 6 worden resultaten beschreven van dezelfde groep cavia’s. In plaats van enkele pulsen bestonden de stimuli die de eCAP opwekten uit meerdere stroompulsen, zodat verschillen in temporele verwerking van elektrische stimulatie onderzocht kon worden. In klinische studies wordt voor dergelijke doeleinden vaak gebruik gemaakt van een zogeheten masker-probe-paradigma: twee pulsen die met een groot tijdsinterval na elkaar gegeven worden, zullen een eCAP met gelijke amplitude opwekken; maar naarmate de tweede puls sneller op de eerste volgt, zal de amplitude van de tweede eCAP eerst lager worden, en tenslotte volledig verdwijnen. Door dit paradigma met meerdere tijdsintervallen uit te voeren kan een eCAP-herstelcurve gereconstrueerd worden, waarmee de absolute en relatieve refractaire periode van de SGC-populatie bepaald kunnen worden. Zoals bij meerdere klinische studies, vonden we geen verschillen in deze waarden tussen normaalhorende en dove cavia’s. In het tweede deel van deze studie bestonden de stimuli uit 100-milliseconde-lange pulstreinen, en was het paradigma verder vergelijkbaar. Met deze stimuli vonden we grote verschillen tussen de caviagroepen: het herstel was tot tweemaal sneller voor de dove cavia’s dan voor de normaalhorende controles. Op basis van deze bevindingen concluderen we dat degeneratie een duidelijk effect heeft op temporele eigenschappen van de SGC’s, maar dat het veelgebruikte masker-probe-paradigma niet het onderscheidende vermogen heeft dat nodig is om deze effecten bloot te leggen. 238

Nederlandse samenvatting

Het experiment beschreven in hoofdstuk 7 combineert de methoden gebruikt in hoofdstukken 5 en 6 met de toediening van de neurotrofe factor BDNF. Het eerste doel was om te onderzoeken of het beschermende effect van de BDNF-behandeling over een langere periode dan vier weken na het beëindigen van de behandeling in stand bleef. Hiertoe werden de SGC’s van één groep cavia’s direct na de behandeling geanalyseerd, en van een tweede groep vier weken en van een derde groep acht weken na het beëindigen van de behandeling. Het tweede doel was om de functionele eigenschappen van de SGC’s van elk van deze groepen uitgebreid te bestuderen. De aantallen SGC’s waren gelijk voor elk van de drie groepen, wat aantoont dat SGC-degeneratie zelfs tot minstens acht weken na het stopzetten van de neurotrofe behandeling uitblijft. Deze bevinding versterkt de eerdergenoemde theorie aangaande een mogelijk autonoom overlevingsmechanisme van SGC’s na een tijdelijke neurotrofe behandeling. Aangezien een tijdelijke behandeling op de lange termijn effectief blijft, wordt een mogelijke klinische toepassing van BDNF bij CI-gebruikers vele malen haalbaarder. De functionele eigenschappen van de met BDNF behandelde SGC’s – bepaald met eCAP-metingen – waren over het algemeen weinig verschillend van die van SGC’s van normaalhorende cavia’s. Bovendien veranderde de functionele eigenschappen niet na het beëindigen van de BDNF-behandeling. Deze resultaten vormen een belangrijke toevoeging op de huidige literatuur, gezien het feit dat eerder onderzoek naar behandeling met neurotrofe factoren weinig aandacht heeft getoond voor de functionele consequenties.

Tot slot is in hoofdstuk 8 een uitgebreide analyse van de histologische data van SGC’s uit hoofdstukken 3, 5, en 7 beschreven. Er wordt traditioneel gedacht dat de degeneratie van SGC’s begint bij de PP’s en pas later de cellichamen en het axon bereikt (retrograde degeneratie), maar bewijs hiervoor is niet bijzonder sterk. Van één groep normaalhorende en vier groepen dove cavia’s zijn de aantallen PP’s en cellichamen bepaald. Van drie van deze vijf groepen zijn ook tellingen van de axonen in de auditieve zenuw gedaan. De resultaten laten ontegenzeggelijk zien dat in deze caviagroepen de degeneratie van SGC’s simultaan verloopt: zolang als het cellichaam van een SGC bestaat, heeft het een PP en een axon; degeneratie treft alle drie de componenten tegelijkertijd. Deze resultaten zijn relevant voor CI-onderzoek, aangezien veelal wordt gedacht dat SGC’s van CI-gebruikers weinig tot geen PP’s bezitten, en het van functioneel belang kan zijn om deze weer te regenereren. Onze bevindingen geven aan dat retrograde degeneratie geen universeel proces is, en dat het systematischer in andere diersoorten en in mensen onderzocht dient te worden.

239

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Dankwoord

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Dankwoord

Terugkijkend op de afgelopen vier en een half jaar ben ik velen dank verschuldigd voor hun advies, hulp, steun, of afleiding.

Professor Grolman, als eerste wil ik u graag bedanken. Tijdens mijn sollicitatiegesprek vroeg u mij waarom ik eigenlijk zo nodig de wetenschap in wilde. Die vraag had eigenlijk het eenvoudigst van alle te beantwoorden moeten zijn geweest, maar ik herinner mij niet dat mijn antwoord op dat moment van enig inzicht betuigde. Ik wil u bedanken dat ik desondanks het voordeel van de twijfel heb gekregen. Ook wil ik u bedanken voor het geduld en vertrouwen toen er na mijn eerste jaar nog weinig van mijn onderzoek terecht was gekomen. Door een constructie met twee copromotoren heeft u geen directe begeleidende rol gehad, maar de voortgangsgesprekken hebben veelal verhelderend gewerkt â&#x20AC;&#x201C; mede dankzij uw aansporingen om toe te werken naar klinisch haalbare toepassingen. En na ruim vier jaar bedenktijd bij dezen dan eindelijk een antwoord op uw vraag: omdat ik dit alles voor geen goud had willen missen. Vervolgens mijn copromotoren: Huib Versnel en Sjaak Klis. In vele opzichten zijn jullie compleet verschillend, wat vaak tot nuttige discussies over de uit te voeren experimenten en over de behaalde resultaten heeft geleid. Wanneer jullie het ergens over eens zijn, weet ik dat dat bijna een absolute waarheid moet zijn. Zonder al te veel te overdrijven zou ik me geen betere begeleiding tijdens mijn promotietraject hebben kunnen voorstellen. Huib, ik schat dat ik de afgelopen jaren gemiddeld drie keer per dag je kamer binnengelopen ben om even kort wat te vragen, nieuwe bevindingen voor te leggen, of uitgebreid te discussiĂŤren over weer een nieuwe theorie. Het lijkt dan alsof je me initieel probeert te negeren, maar uiteindelijk neem je altijd de tijd om me te helpen. Het inzicht van een fysicus is erg nuttig gebleken bij de hoeveelheid en multidimensionaliteit van de verkregen data. Deze talloze korte bijeenkomsten zijn voor mij heel belangrijk geweest. Ik zal proberen me de komende tijd meer als een echte post-doc te gedragen. Sjaak, als positief ingestelde pessimist bezit je het vermogen om elke situatie een interessante wending te geven. Zo kan je enerzijds enthousiaste maar ongegronde verwachtingen doeltreffend temperen, terwijl je anderzijds met een verrassend enthousiasme probleemsituaties benadert. Je kennis van fundamentele biologische en elektrofysiologische principes hebben vaak geholpen bij het formuleren en theoretisch toetsen van hypotheses en conclusies.

Sarah, om te beginnen wil ik je bedanken omdat je me aan het begin van mijn promotietijd hebt geleerd om te opereren. Maar je hebt me in die tijd nog iets anders geleerd waar ik uiteindelijk misschien wel meer aan heb gehad, en dat is om niet altijd 242

Dankwoord

af te wachten totdat iets volledig uitgedacht is, maar in plaats daarvan soms gewoon te proberen en zien wat er gebeurt. Ik denk dat die strategie me op cruciale momenten in de voorbereiding van experimenten een hoop tijd en moeite heeft bespaard. Tot slot wil ik je bedanken voor je gezelschap in de afgelopen vier en een half jaar – als kamergenoot, en bij alle formele en iets minder formele bijeenkomsten in binnen- en buitenland!

Het uiteindelijke doel van dit proefschrift is om mensen – en niet cavia’s – te helpen. Dat einddoel lijkt soms wat ver weg, en sommigen vragen zich wel eens af wat dierexperimenteel onderzoek op de afdeling KNO te zoeken heeft, maar klinische relevantie bleek vaak dichterbij te zijn dan gedacht. Marc en Jan-Willem, onze gesprekken over onze respectievelijke bevindingen bij cavia’s en patiënten hebben vaak geleid tot interessante ideeën voor vervolgonderzoek. Het besef direct iets te kunnen bijdragen aan klinisch onderzoek werkte vaak enthousiasmerend.

Dan wil ik graag alle studenten bedanken die me de afgelopen jaren geholpen hebben bij mijn onderzoek. Emma, ik hoop dat je evenveel geleerd hebt van jouw stage als ik. Je was mijn eerste student, en ik heb je eindeloos axonen laten tellen. Je enthousiasme was enigszins weggeëbd na enkele maanden, maar karakter en doorzettingsvermogen compenseerden hiervoor. Het zal een schrale troost voor je zijn dat voor vervolgonderzoek het automatisch axonentellen dan nu eindelijk onderweg lijkt te zijn. Ik ben hoe dan ook blij dat je werk uiteindelijk in dit proefschrift terecht is gekomen. Ook wil ik jou en Thijs bedanken voor de avonden waarop jullie mij gezelschap hebben gehouden op het lab tijdens de langdurende experimenten. Koen, jouw bijdrage aan de BDNF-studie is cruciaal geweest. Tijdens je stage heb je bij talloze operaties geassisteerd, en uiteindelijk heb je er een aantal zelf uitgevoerd. Daarnaast heb je het grootste deel van de histologische analyse op je genomen. En alsof dat nog niet genoeg was, heb je na het afronden van je studie nog een aantal maanden als vrijwilliger geteld en geopereerd. Bedankt voor alle hulp! Imme, Rizgar en Wouter, jullie bijdragen zijn niet in dit proefschrift terechtgekomen, maar we gaan hoogstwaarschijnlijk de komende tijd nog aan de slag met de eABRs en met de eCAP-deconvoluties. Bedankt voor jullie hulp en voor de fijne samenwerking op het lab! Van de gebruikte technieken die tot de bevindingen van dit proefschrift hebben geleid is er eigenlijk maar een die ik niet (uiteindelijk) zelf heb uitgevoerd. Ferry, zonder jouw kennis van en vaardigheid in de histologische verwerking van de cochlea’s was er van een groot deel van dit proefschrift weinig terecht gekomen. Daarnaast heb je altijd

243

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Dankwoord

geduldig al die eigenwijze studenten geïnstrueerd over de histologische analyse – niet tot ze dachten dat ze het begrepen, maar totdat ze het uiteindelijk ook echt begrepen. Ik wil René van de Vosse bedanken voor de technische ondersteuning bij het meten van ABRs. Het feit dat er een tweede meetopstelling bij kwam om eCAPs te meten, betekende niet dat ik je minder nodig heb gehad dan mijn voorgangers. Om eABRs te kunnen meten heb je ervoor moeten zorgen dat de twee systemen met elkaar samen konden werken, en die samenwerking ging al snel even soepel als de samenwerking tussen ons tweeën.

Ook de coauteurs van verschillende publicaties die als hoofdstukken in dit proefschrift zijn opgenomen wil ik graag bedanken voor hun bijdrage: Ferry, John, Martijn, Ties, maar vooral Laurien en Maarten. Bedankt voor de prettige samenwerking.

Bij de werkbesprekingen heb ik altijd uitgekeken naar het commentaar van diegenen met wie ik in het dagelijks werk weinig te maken heb gehad. Met name wil ik hier de audiologen Bert, Jan-Willem, Piet en Vera bedanken voor alle kritische vragen en suggesties. Jullie kennis van de fysica gecombineerd met de dagelijkse praktijk in het functiecentrum is voor mij altijd erg nuttig geweest.

Daarnaast wil ik de overige collega’s op H.02 en G.02 bedanken die de afgelopen jaren voor een fijne werkomgeving hebben gezorgd: Alice, Frans, Frits, Hanneke, Hendrik, Inge, Ingrid, Jeroen, Joost, Juliette, Magdalena, Marlien, Nina, Ruben, Stephanie en Yvette. In totaal heb ik 189 operaties op het GDL uitgevoerd, waarvan ik er maar enkele alleen gedaan heb. In alfabetische volgorde wil ik de volgende operatieassistenten bedanken: Anne-Marie, Emma, Feike, Huib, Januska, Jeroen, Koen, Lieke, Luuk, Magdalena, Marieke, Marjolein, Monique, Nina, Pauline, Renate, Robbert, Sarah, Thijs, Varinder, Wouter. Sommigen hebben een enkele keer meegeholpen en durfden daarbij niet eens te kijken, anderen hebben maandenlang zo ongeveer dagelijks meegeholpen – allen hebben een belangrijke bijdrage geleverd aan dit proefschrift. De medewerkers van het GDL wil ik graag bedanken voor de dagelijkse verzorging en controle van de cavia’s. Met name wil ik Jeroen en Ruben bedanken, die na de verhuizing van Geel naar Groot het welzijn van de chronisch geïmplanteerde cavia’s bewaakt hebben.

244

Dankwoord

I would like to thank everyone at MED-EL who has helped me with my project one way or the other. I’m writing these words while in my hotel room in Tokyo during the Hearing and Structure Preservation Workshop, and this meeting is once again exemplary of our good and friendly cooperation during the past four years. In particular I would like to thank Stefan Strahl, Carolyn Garnham, and Roland Hessler. Stefan, your assistance has been crucial at the beginning of my project. You helped me design and write the eCAP recording paradigms with great patience, and when the first results were finally there your skills in data analysis as well as your knowledge of the existing literature have been invaluable. I really appreciated our scientific late-night conversations by e-mail and/or Google hangout, in which we discussed the data almost in real-time as I was collecting them. Carolyn, besides your personal support for – and interest in – this project, the several interesting and useful meetings (both physical and virtual) you initiated have also greatly helped me to perform my research and to interpret the acquired data. Roland, when I first listed all the requirements for the electrodes for the BDNF study, I realized it probably was too much to ask of you. It was a pleasant surprise that it all worked out so well and in so short a time. I am pleased that we have already explored the possibilities for our future project for chronic electrical stimulation – I believe that short meeting left us both equally enthusiastic. De afgelopen jaren had ik niet zo hard kunnen werken zonder de nodige muzikale afleiding. Van wekelijks spelen kwam het vaak niet meer, maar het heeft toch nog twee albums en een aantal optredens opgeleverd. Bedankt Labasheeda – Arne, Saskia en Thanos – en bedankt Unexposed – Bas, Jorn, Simon! Belangrijker nog is de naamloze band die in de afgelopen vijftien jaar ongeveer zevenentachtig albums, één optreden en één beker heeft opgeleverd. Lucas en Tim, de hokavonden zijn nog altijd een perfecte mix van muziek, flauwekul en quasi-intellectueel gepraat. En na een pauze van een paar maanden gaan we binnenkort echt weer verder!

Naast wetenschap en muziek is er niet altijd tijd geweest voor anderen. Ik wil hier met name Justus en Marcel bedanken voor het blijven proberen om zo nu en dan toch eens wat af te spreken. De door jullie georganiseerde barbecue in mijn achtertuin de afgelopen zomer terwijl ik eigenlijk de hele avond wilde werken is waarschijnlijk het beste voorbeeld hiervan. Jasper en Marlies, we zijn tegelijk aan ons eerste jaar psychobiologie begonnen en zijn binnenkort dan ook eindelijk alledrie volwaardig neurowetenschapper. We zien elkaar nog regelmatig, maar dat zegt helaas niets over de frequentie. Sorry. Nu dit alles achter de rug is kom ik graag weer eens naar Amsterdam of Edinburgh! 245

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Dankwoord

Renate, ik beperk me hier graag tot wat je voor me hebt betekent als het gaat om mijn promotie. Als belangrijkste zijn het onze talloze en eindeloze gesprekken over wetenschap in het algemeen, over mijn project, mijn experimenten, over hypothesen, resultaten, analyses, theorieën en conclusies. Daarnaast heb je vooral de afgelopen maanden alle tekst in dit proefschrift minstens één keer gelezen, en elke keer weer probeerde ik je commentaar fanatiek te weerleggen – vaak heb ik het dan uiteindelijk stiekem toch nog wel verwerkt. Laten we ook die keer niet vergeten dat je me hebt geassisteerd bij een implantatie terwijl de rest van Nederland naar de troonswisseling zat te kijken. Vooral de afgelopen maanden, maar eigenlijk de gehele periode, heb je geduldig gewacht, en ervoor gewaakt dat ik ook nog wel aan de nodige ontspanning toekwam. En ja, we gaan binnenkort echt een keer op vakantie. Ties, Wieke, Stéphanie en Taco: joeoeoe, merci!

Tot slot mijn ouders: bedankt voor alle steun, hulp, tijd, aanmoedigingen, interesse en vertrouwen; altijd en onvoorwaardelijk.

246

Curriculum Vitae

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Curriculum Vitae

248

Curriculum Vitae

Dyan Ramekers was born in Veldhoven, the Netherlands on February 4th 1983. After graduation from high school in 2001 (Van Maerlantlyceum, Eindhoven) he studied Life Science & Technology (Leiden University/Delft University of Technology) for two years. Realizing his interests lay with neuroscience, he decided to switch studies. He studied Psychobiology at the University of Amsterdam and obtained the degree of Bachelor of Science in 2007, followed by the degree of Master of Science in neuroscience in 2009. As a postgraduate research assistant he studied the functional regeneration of the sciatic nerve after crush injury at the department of Genome Analysis at the Academic Medical Center in Amsterdam for six months. He started his PhD at the department of Otorhinolaryngology and Head & Neck Surgery at the University Medical Center in Utrecht in 2010, the results of which are described in this thesis. He currently works as a postdoctoral researcher at the department of Otorhinolaryngology and Head & Neck Surgery at the University Medical Center in Utrecht, studying the combined effects of chronic electrical stimulation and clinically practicable treatment with brain-derived neurotrophic factor in guinea pigs.

249

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

List of publications

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

List of publications

252

List of publications

Ramekers, D., Versnel, H., Strahl, S.B., Smeets, E.M., Klis, S.F.L., Grolman, W., 2014. Auditory-nerve responses to varied inter-phase gap and phase duration of the electric pulse stimulus as predictors for neuronal degeneration. J. Assoc. Res. Otolaryngol. 15, 187-202. Sta, M., Cappaert, N.L.M., Ramekers, D., Baas, F., Wadman W.J., 2014. The functional and morphological characteristics of sciatic nerve degeneration and regeneration after crush injury in rats. J. Neurosci. Methods 222, 189-198. Sta, M., Cappaert, N.L.M., Ramekers, D., Ramaglia, V., Wadman, W.J., Baas, F., 2014. C6 deficiency does not alter intrinsic regeneration speed after peripheral nerve crush injury. Neurosci. Res. doi: 10.1016/j.neures.2014.06.008. Cappaert, N.L.M., Ramekers, D., Martens, H.C.F., Wadman, W.J., 2013. Efficacy of a new charge-balanced biphasic electrical stimulus in the isolated sciatic nerve and the hippocampal slice. Int. J. Neural. Syst. 23, 1250031. van Loon, M.C., Ramekers, D., Agterberg, M.J.H., de Groot, J.C.M.J., Grolman, W., Klis, S.F.L., Versnel, H., 2013. Spiral ganglion cell morphology in guinea pigs after deafening and neurotrophic treatment. Hear. Res. 298, 17-26. Waaijer, L., Klis, S.F.L., Ramekers, D., Van Deurzen, M.H.W., Hendriksen, F.G.J., Grolman, W., 2013. The peripheral processes of spiral ganglion cells after intracochlear application of brain-derived neurotrophic factor in deafened guinea pigs. Otol. Neurotol. 34, 570-578. Kuribara, M., van Bakel, N.H.M., Ramekers, D., de Gouw, D., Neijts, R., Roubos, E.W., Scheenen, W.J.J.M., Martens, G.J.M., Jenks, B.G., 2012. Gene expression profiling of pituitary melanotrope cells during their physiological activation. J. Cell. Physiol. 227, 288-296. Ramekers, D., Versnel, H., Grolman, W., Klis, S.F.L., 2012. Neurotrophins and their role in the cochlea. Hear. Res. 288, 19-33.

253

R1 R2 R3 R4 R5 R6 R7 R8 R9 R10 R11 R12 R13 R14 R15 R16 R17 R18 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28 R29 R30 R31 R32 R33 R34 R35 R36 R37 R38 R39

Assessment and Preservation of Auditory Nerve Integrity in the Deafened Guinea Pig Dyan Ramekers 64 ISBN 978-94-6108-814-7

UITNODIGING voor het bijwonen van de openbare verdediging van het proefschrift

Assessment and Preservation of Auditory Nerve Integrity in the Deafened Guinea Pig door Dyan Ramekers

Paranimfen: Renate Buijink (renatebuijink@gmail.com) Sarah Havenith (s.havenith@umcutrecht.nl) Dyan Ramekers Boerhaavelaan 86 3552 CZ Utrecht 06 18031396 d.ramekers@umcutrecht.nl