Vol 17 ¡ Issue 1 ¡ Fall 2018
ps
addiction: an interview with Dr. henry Kransler page 29
ps ps ps
ps
Developing drugs: Women in clinical trials
page 26
ps
ps
ps
elecrochemical oxidization for renewable energy conversion and storage
Drugs
from development to public health
page 33 31 page
PENN SCIENCE FALL 2018 VOL. 17 ISSUE 1
PennScience is a peer-reviewed journal of undergraduate research and related content published by the Science and Technology Wing at the University of Pennsylvania and advised by a board of faculty members. PennScience presents relevant science features, interviews, and research articles from many disciplines, including the biological sciences, chemistry, physics, mathematics, geological science, and computer science. PennScience is funded by the Student Activities Council. For additional information about the journal including submission guidelines, visit www.pennscience.org or email pennscience@gmail.com.
EDITORIAL STAFF editors-in-chief writing managers Kenny Hoang Darsh Shah
editing managers Billy Hasley Kathy Wang
Mia Fatuzzo Jenny (JiCi) Wang
design manager Roshan Benefo
technology manager Rounak Gokhale
communications managers
business managers
Rachel Levinson Aaron Zhang
faculty advisors Dr. M. Krimo Bokreta
Dr. Jorge Santiago-Aviles
Donna Yoo Catherine Ruan
student advisor Grace Ragi
writing
editing
design
business
Emily Lo Andrew Lowrance Michelle Paolicelli Pranshu Suri Hiab Teshome Tamsyn Brann Neelu Paleti Roshni Kailar Rose Nagele Asha Dahiya Amanda Paredes
Abraham Frey Daniel Rodriguez Brian Song Emma O’Neil Ken Jiao Lily Zekavat Sumant Shringari Mimi Lu Brian Zhong Kelly Liang
Farhaanah Mohideen Felicity Qin Julia Davies Amara Okafor Jessica Tan Julia Yan Kaitlyn Thayer Abi Szabo Olivia Meyer
Alexander Massaro Glen Kahan Angela Yang Helen Jiang
communications Jasmine Chen Celia Zhang Arjun Jain Zhiqiao Jiang Charles Rothkrug Joshua Kim Natasha Chity-Guevara David DeVaro Shriya Beesam Lynn Ahrens Eric Teichner
Tab Co le of nte nts
6
the future of immunotherapy
9
hydrogels: the end of daily medication
12
mental illness treatment
15
oxytocin: a hormone of the body, a love from within ourselves
18
the opioid epidemic
20
digital therapeutics: the doctor that fits in your pocket
23
Cannibis Oil: A health trend to weed out?
26
developing drugs: women in clinical trials
29
31
interview with dr. henry kranzler interview with dr. reagan wetherhill
33 electrochemical oxidation of formate on a PdNI/C nanoparticle catalyst for renewable energy conversion and storage
38
ps
modifying reaction diffusion: a numerical model for turing morphogenesis, ben-jacob patterns, and cancer growth
A LETTER FROM THE EDITORS
D
ear Reader, We are thrilled to present the fall issue of the seventeenth volume of the PennScience Journal of Undergraduate Research. We are incredibly grateful for the members of PennScience who worked diligently all semester to assemble this journal, the students who submitted their research findings for publication, and the Penn community who attended our events and engaged in scientific discourse on campus. In this issue, Hiab Teshome examines the future of immunotherapy in treating cancer. Michelle Paolicelli discusses the potential of targeted drug delivery using hydrogels. Roshni Kailar reviews the current treatment options for mental illnesses from multiple perspectives. Andrew Lowrance takes a closer look at oxytocin, a hormone that drives human interconnected behavior. Pranshu Suri investigates the opioid epidemic and its implications for public health. Neelu Paleti explores the emerging intersection between drugs and technology. Finally, Rose Nagele provides a detailed look at the complicated role of women in clinical trials. We are also proud to present the original research of two students – Sai Mamidala and Kai Trepka. We have greatly enjoyed our semester leading PennScience, and we would like to extend a sincere thank you to the many groups and individuals who have made PennScience possible. First, we would like to thank our incredible journal staff – the members of our writing, editing, design, business, and communication committees – for their hard work, dedication, and enthusiasm. Our publication is entirely student-run and relies on the efforts of our scientifically curious undergraduate members. We would also like to recognize the students on campus who shared their experiences with us at our research panel, as well as Dr. Henry Kranzler, who was interviewed for our journal by writing committee member Tamsyn Brann. We owe our funding to the Science and Technology Wing of the King’s Court College House and the Student Activities Council. Thank you also to our faculty mentors, Krimo Bokreta and Jorge Santiago-Aviles, for their constant guidance and support. Finally, we would like to thank you for reading PennScience – enjoy our latest issue! Sincerely, Mia Fatuzzo (C’19) Jenny (JiCi) Wang (C’19) Co-Editors-in-Chief
Looking for a chance to
publish your research? Penn Science is accepting submissions for our spring 2019 issue! Submit your independent study projects, senior design projects, reviews, and other original research articles to share your work with fellow undergraduates at Penn and beyond. Email submissions and any questions to pennscience@gmail.com
Research in any scientific field will be considered, including but not limited to: Biochemistry Biological Sciences Biotechnology Chemistry Computer Science Engineering Geology Mathematics Physics Physiology
Visit the PennScience website www.pennscience.org to see previous issues and for more information.
Spring 2018
Spring 2017
Features
y
f
Im
r e th
a
p
O
U
TUR E
F
THE
muno
By Hiab Teshome Designed by Farhaanah Mohideen
T
raditionally, metastatic melanoma cancer has been targeted with chemotherapy, but new immunotherapies can use the bodies defenses to attack this cancer. These immunotherapies work by developing antibodies in the lab that can be used as a targeted therapy [1]. By tagging these cancer cells, the antibodies signal to the immune system to destroy them, or act as immune checkpoint inhibitors to regulate many of the body’s immune system checkpoints. Scientist have been studying two types of immunotherapy drugs, specifically for metastatic melanoma. First, small-molecule inhibitors target the MAP kinase pathways that regulate many different types of cellular activities. Second, immune-checkpoint inhibitors reactivate suppressed antitumor immune responses by removing blinders on cancer cells that once prevented the immune system from recognizing the cells as cancerous [2]. Both treatments increased the survival rates of patients
6
PENNSCIENCE JOURNAL | fall 2018
with metastatic melanoma. This progress in cancer immunotherapy has shown great promise in activating the immune system against cancer. These drugs target two biochemical pathways, the MAPK and ERK pathway. The MAPK, or mitogen-activated protein kinase, pathway is critical for human cancer cell survival [3]. Errors in this pathway have been shown to alter cell regulation, leading to cancer. MAP kinases work with extracellular signal regulated kinases (ERK), which are a chain of proteins that relay signals from receptors outside of the cell to genes inside of the nucleus. ERKs communicate with a family of Raf Both treatments increased the survival rates of patients with metastatic melanoma.
Tumor Cell
Drug immunotherapies have allowed researchers to target tag tumor cells, using antibody signalling to destroy them. Antigen By destroying pathways that have allowed tumor cells to be T Cell Receptor unrecognizable to T-cells, the growth rate of cancer can be diminished. Immunotherapies target the PD-1/ PD-L1 pathway to target cancer cells. T-Cell
PD-L1 PD-1
proteins which are a series of kinases that are involved in the proliferation, differentiation, and apoptosis of cells [4]. In immune cells, activated Raf is also a component of the innate response in different steps of the inflammatory cascade, increasing the expression of tumor necrosis factor alpha (TNF-ι) and inducible nitric oxide synthase (iNOS), which are both signalling proteins involved in systemic inflammation in order to coordinate an immune response [5]. Because of Raf ’s and ERK’s important role in the immune pathway, studies were carried out to explore if the MAP pathway can be a potential target for anticancer drugs. The most common mutation of the Rafs and ERK involves the BRAF gene, which dictates the proper protein formation of these proteins. Recent studies have involved identifying selective BRAF inhibitors in order to reduce the production of mutated Rafs and ERKs. Selective BRAF inhibitors, such as vemurafenib and dabrafenib, extended the median cancer progression free survival of patients for more than 7 months [6]. Both of these drugs inhibit the protein synthesis of mutated BRAF which results in abnormal growth of cells, and inhibition of the immune system to regulate cancer cells. Although BRAF inhibitors have had clinical benefits, drug resistance is very frequent making these inhibitors limited in their ability to destroy cancer cells In addition, these inhibitors sometimes incorrectly activate MAP kinase pathways in BRAF wild-type cells driving tumor growth [7]. As a result, due to some of the failings of the BRAF inhibitors, researchers began to study immune system checkpoints, and possible drugs to alter and reactivate these systems. MAP kinase pathway immunotherapies inhibit the production of mutated proteins. However, another very commonly used immuno-
Features
Tumor Cell Death
PD-L1 Antigen
Anti PD-L1 Anti PD-1
T Cell Receptor PD-1
T-Cell
fall 2018 | PENNSCIENCE JOURNAL
7
Features
therapy is through using drugs to block immune checkpoints to allow the body’s’ natural immune system to destroy cancer cells Researchers are trying to better understand how tumors can escape immune cell recognition by inhibiting the ability for T cells to respond to antigen and co-stimulatory signals.
without regulation. The major cell involved in mediating immunity against cancers are T cells. The activation of T cells to perform its anti tumoral function requires two signals. Signal one is an antigen-specific signal to the T-cell receptor. These specific signals are expressed on many cells including cancer cells. The second required signal is a co-stimulatory signal provided by B7 and other co-stimulatory molecules that assist in T cell activation [8]. The successful balance of activating and inhibiting T cells determines the surveillance and cancer eliminating function of T cells. Researchers are trying to better understand how tumors can escape immune cell recognition by inhibiting the ability for T cells to respond to antigen and co-stimulatory signals. Under normal conditions, immune checkpoint molecules serve to regulate T-cell responses, which is necessary to avoid uncontrolled destruction. However, tumor cells use these intrinsic ‘brakes’ of the immune system as immune escape mechanisms by inducing functionally exhausted T-cells which can no longer respond to antigens [9]. Anticancer drugs targeting these ‘immune checkpoints’ on T-cells are termed ‘checkpoint inhibitors’[9]. These checkpoint molecules are extremely important, as they help the body to discriminate between ‘foreign’ and ‘self ’. Numerous drugs
have been developed to intercept tumor control of the immune system by specifically targeting these checkpoint molecules on T-cells by blocking inhibitory checkpoints to restore immune system function, however they were not as successful at increasing the lifespan of patients. Recently, two drugs were approved by the FDA to block immune checkpoint molecules, ipilimumab and atezolizumab. These drugs target checkpoint proteins on T cells, such as CRLA-4 and PD-1, that can switch off T cells to keep it from attacking other cells in the body [9]. These drugs can identify and inhibit these targets in order to activate the immune system to destroy cancer cells that it did not once recognize. However, the danger in using this form of immunotherapy is that by blocking inhibiting checkpoints, there is no longer a mechanism for immune cells to recognize foreign cells and self cells. Cancer immunotherapy is a blossoming field in the medical field. Researchers are currently developing drugs that target more inhibitory molecules, and creating drugs that can blockade for more than one inhibitory molecules. However, by globally suppressing the inhibitory molecules, many individuals undergoing these therapies develop serious autoimmune diseases due to the drug’s widespread blocking activity. Researchers therefore are moving to a more “personalized” approach to combination immunotherapy; in that scenario, tumor biopsies would be interrogated for a series of addressable checkpoints, then a personalized checkpoint blockade cocktail administered [10]. An even more promising field in cancer immunotherapy is studying how different strategies for inhibiting the brakes on the immune system can be used in the treatment of cancer. This year’s Nobel Prize winners in Physiology or Medicine, James P. Allison and Tasuku Honjohe, discovered a new protein pathway that when blocked was a promising strategy in the fight against cancer. Results were dramatic, leading to long-term remission and possible cure in several patients with metastatic cancer, a condition that had previously been considered essentially untreatable [11]. For more than 100 years scientists have been attempting to use the immune system in the fight against cancer. Checkpoint therapy has now revolutionized cancer treatment and has fundamentally changed the way we view how cancer can be managed.
References
[1] Cancer.Net. (2018). Understanding Immunotherapy. [online] Available at: https://www.cancer.net/navigating-cancer-care/how-cancer-treated/immunotherapy-and-vaccines/understanding-immunotherapy [Accessed 4 Nov. 2018]. [2] Turajlic, S. and Larkin, J. (2018). Immunotherapy for Melanoma Metastatic to the Brain. New England Journal of Medicine, 379(8), pp.789-790. [3] De Luca A, Maiello MR, D’Alessio A, Pergameno M, Normanno N. The RAS/RAF/MEK/ERK and the PI3K/AKT signalling pathways: role in cancer pathogenesis and implications for therapeutic approaches. Expert Opin Ther Targets. 2012;16(Suppl 2):S17–27. [4] Hagemann, C. and Rapp, U. (1999). Isotype-Specific Functions of Raf Kinases. Experimental Cell Research, 253(1), pp.34-46. [5] Arthur, J. and Ley, S. (2013). Mitogen-activated protein kinases in innate immunity. Nature Reviews Immunology, 13(9), pp.679-692. [6] K.T. Flaherty, I. Puzanov, K.B. Kim, A.Ribas, G.A. McArthur, J.A. Sosman, et al. (2010). Inhibition of mutated, activated BRAF in metastatic melanoma. N Engl J Med, 363, pp. 809-819. [7] Sanchez-Laorden, A. Viros, M.R.Girotti, M. Pedersen, G. Saturno, A.Zambon, et al. (2014). BRAF inhibitors induce metastasis in RAS mutant or inhibitor-resistant melanoma cells by reactivating MEK and ERK signaling. Sci Signal, 7, p. Ra30 [8] Shi, T., Ma, Y., Yu, L., Jiang, J., Shen, S., Hou, Y. and Wang, T. (2018). Cancer Immunotherapy: A Focus on the Regulation of Immune Checkpoints. International Journal of Molecular Sciences, 19(5), p.1389. [9] Greil, R., Hutterer, E., Hartmann, T. and Pleyer, L. (2017). Reactivation of dormant anti-tumor immunity – a clinical perspective of therapeutic immune checkpoint modulation. Cell Communication and Signaling, 15(1). [10]Li, X., Shao, C., Shi, Y. and Han, W. (2018). Lessons learned from the blockade of immune checkpoints in cancer immunotherapy. Journal of Hematology & Oncology, 11(1). [11] Nobelprize.org. (2018). [online] Available at: https://www.nobelprize.org/uploads/2018/10/press-medicine2018.pdf [Accessed 4 Nov. 2018].
8
PENNSCIENCE JOURNAL | fall 2018
Hydrogels:
Features
The End of Daily Medication By Michelle Paolicelli Designed by julia yan
S
cientists today are creating materials with amazing capabilities on a microscopic scale in an effort to improve the human experience. Our ancestors first realized that the objects around us could be manipulated to better suit our needs about two and a half million years ago, and since then the field of materials science has continuously evolved. Today, so-called “smart” materials are being designed to respond to specific and distinct environments. One emerging field is targeted drug delivery using hydrogels. Hydrogels are versatile substances predominately made up of water, and they have received great attention because of their biocompatibility, easy manipulation, and unique physical properties [1]. Hydrogels have a multitude of applications in medicine. From contact lenses to synthetic wound dressings, they serve as the ideal material for use in the human body [2]. Although hydrogels are well established in contact lens technology, their use in pill capsules for targeted drug delivery is still in its formative
stages. During the drug delivery process, the drug and hydrogel are crosslinked before the resulting complex is taken into the body orally [3]. When hydrated, hydrogels become a soft rubbery material that can be easily molded to create an internal drug reservoir, and they serve as a barrier between the drug and the human body. Depending on the properties of the hydrogel packaging, the boundary between drug and body will only break under specific physiological conditions —such as a specific pH [3]— or when the gel reaches a certain state related to swelling
fall 2018 | PENNSCIENCE JOURNAL
9
Features
HYDROGELS
Drug resevoir
hydrogel
are versatile substances predominantly made up of water.
From contact lenses to synthetic wound dressings, hydrogles serve and anatomical conditions. This is beneficial for patients as the ideal material for use suffering from morbidities specific to a particular body system in the human or in need of continuous relief. Traditional medications use body.
Drug resevoir
hydrogel
The progression of hydrogel degradation in the body. Drug diffusion occurs from the core through the hydrogel membrane.
Hydrogels have a multitude of applications in medicine. 10
PENNSCIENCE JOURNAL | fall 2018
rudimentary and slow release techniques that limit their efficacy in providing the patient with satisfactory results because the medication must be frequently re-administered. To encompass numerous delivery parameters, many proposals for the use of hydrogels in drug delivery rely on a diffusion-controlled release mechanism. Hydrogel capsules contain a central drug reservoir surrounded by a hydrogel membrane to create a concentration gradient that allows for a constant release rate of the drug. Some diffusion-controlled applications also incorporate a swelling-controlled aspect. This refers to the swelling response of a hydrogel when it comes in contact with a bio-fluid. Once expanded to the critical point, the drug within begins to diffuse into the body [2]. Currently, there are only a few commercial products that employ hydrogels as a means of drug delivery, but that is likely to change in the coming years. Dr. Kinam Park, a professor of biomedical engineering at the Purdue University School of Pharmacy, has dedicated his decades-long career to the study of the applications of biomaterials in drug delivery. He has proposed that hydrogels be used to create an orally administered drug that can be retained in the stomach for twenty-four hours [2][4]. This allows for a slower release of medication, and grants the patient a greater period of relief for each dosage taken. This method would be applicable for any drugs currently treated with pills. Dr. Park explains his research approach using an analogy of the digestion of food: if a large piece of meat is swallowed, it will sit in the stomach until its volume has been reduced enough to allow passage into the rest of the digestive tract. Similarly, the larger a dose of orally administered medicine, the longer it will take for the
medicine to become small enough to pass into the small intestines. Park also notes that a physical pill capsule must be small enough to be easily swallowed. This underscores the need for a dynamic vessel like a hydrogel capsule that can change size based on the level of hydration. Park’s research employs the use of hydrogels to induce swelling of the pill capsule to an extent that it cannot pass into the small intestine from the stomach. However, such pills need to be eliminated by the body eventually, requiring the hydrogels to not only swell, but also be enzyme-digestable [4]. According to Park’s concept of the hydrogel, a patient would ingest a dried hydrogel-drug complex in the form of a capsule about the size of a Tylenol pill. The pill would then swell in the stomach and be broken down over time by the interaction of pepsin in the stomach and the susceptible hydrogel boundary, releasing the drug at a controlled rate [4]. Park hopes to one day create a targeted drug delivery mechanism specifically for cancer treatment. Certain limitations exist when working with hydrogels due to the limited scientific knowledge of the diffusion of enzymes through gels. This gap poses an obstacle when predicting the diffusion rate of enzymes into hydrogels, and thus the rate of release of the drug itself. More generally, ensuring the complete and natural removal of the hydrogel casing from body has been the primary focus of scientists [5]. Also, further research is necessary to assess the possible side effects of hydrogel use as well as the nature of the gels’ degeneration. Lastly, hydrophobic drugs, including certain antitumor drugs, pose a serious obstacle to to incorporation into the hydrophilic hydrogels [5] [6]. Despite ongoing limitations to the practical applications of hydrogels for targeted drug delivery, early research has shown how hydrogels can be used to encapsulate drugs for slow-release in the body. Scientists like Dr. Kinam Park are focused on improving the reliability of current hydrogel applications within drug delivery as well as expanding the number of compatible medications [7]. The biocompatibility and tunable release mechanisms of hydrogels make them ideal for slow-release drug delivery. With continued developments in hydrogel technology, perhaps the era of daily medication will soon be behind us.
Features
early research suggests that hydrogels can be used to encapsulate drugs for slow release in the body.
References: [1] Kopeček, J. (2007). Hydrogel biomaterials: A smart future?. Biomaterials, 28(34), pp.5185-5192. [2] Caló, E. and Khutoryanskiy, V. (2015). Biomedical applications of hydrogels: A review of patents and commercial products. European Polymer Journal, 65, pp.252-267. [3] Peppas, N. (1997). Hydrogels and drug delivery. Current Opinion in Colloid & Interface Science, 2(5), pp.531-537. [4] Park, K. (1988). Enzyme-digestible swelling hydrogels as platforms for long-term oral drug delivery: synthesis and characterization. Biomaterials, 9(5), pp.435-441. [5] Hoare, T. and Kohane, D. (2008). Hydrogels in drug delivery: Progress and challenges. Polymer, 49(8), pp.1993-2007. [6]Schwendener, R. and Schott, H. (2009). Liposome Formulations of Hydrophobic Drugs. Methods in Molecular Biology, pp.129-138. [7] Pàmies, P. and Stoddart, A. (2013). Materials for drug delivery. Nature Materials, 12(11), pp.957-957.
fall 2018 | PENNSCIENCE JOURNAL
11
Features
Mental
Illness
treatment By Roshni Kailar Designed by Kaitlyn Thayer
Disclaimer: This article discusses treatments of mental illnesses from multiple perspectives but is in no way a comment on what an individual with a mental illness should do. As always, you must consult a professional regarding any form of treatment.
D
epression is one of the leading causes of disability throughout the world estimated to affect 1 in 5 adults [1]. Compared to 1 in 6.6 people in the US who reported smoking or 1 in 37 people in the US who have a stroke, it is clear that mental health is a very pressing public health issue to be considered which is only expected to increase in incidence [2, 3]. In response to this concerning trend, this article hopes to provide a review of current treatments, with a focus on recent advancements, for those with mental illness.
12
PENNSCIENCE JOURNAL | fall 2018
Currently, most drugs used to treat mental illness target neurotransmitters in the brain. Too high or low levels of neurotransmitters are often the cause of symptoms associated with mental illness. For example, low levels of serotonin and norepinephrine are correlated with depression and anxiety. Prescription drugs bring these levels back to normal, thus alleviating symptoms. Common classes of drugs include selective serotonin reuptake inhibitors (SSRIs) and selective norepinephrine reuptake inhibitors (SSNIs), both of which increase the amount of serotonin in the synaptic space. However, SSRIs and SSNIs don’t work for all patients and have several associated side effects. For example, patients can have seizures or other cerebrovascular complications. Another complication is serotonin syndrome which causes autonomic instability, gastrointestinal issues, and neuromuscular hyperactivity [4]. In addition to the side effects of drugs like SSRIs, another problem is that two-thirds of patients
with depression do not respond to the first drug they are prescribed [5]. This implies that a majority of patients diagnosed with depression effectively go without treatment for quite some time. An alternative for these traditional treatments is new pharmaceutical drugs, in development, that hopefully will not have the same side effects. Some new targets for treatment include glutamate and ion channel receptors. Glutamate is an excitatory neurotransmitter that activates receptors on neurons, such as the NMDA receptor. Elevated glutamate levels have been observed in patients with depression and other mood disorders. Novel antidepressant drugs like NMDA receptor antagonists, which block the function of receptors, have shown significant behavioral effects in rats. In a 2008 study, a single dose caused rapid and sustained antidepressant effects that lasted for up to one week in rats. By blocking glutamate from binding to the NMDA receptor, the drug alleviates symptoms associated with high glutamate levels [6]. An effective alternative to drug treatments is exercise. Research conducted by Duke University found that between 40-45% of people with depression can overcome their symptoms through exercise alone [5]. This research suggests that there are natural ways to try to combat mental illnesses. Although it is unclear specifically how exercise is able to alleviate mental illnesses, some hypotheses do exist. One theory is that exercise may be involved in the hypothalamic-pituitary-adrenal (HPA) axis via an increase in blood flow to the brain, which allows for the reduction of stress. Another potential way that exercise can help a patient with depression is by facilitating social interaction and providing a distraction from harmful thoughts. People often exercise in groups and meet new people through exercise; exercise can therefore help patients by allowing them to form new social relationships, which are generally therapeutic in terms of mental illnesses. Thus, potential alternative treatments suggested by this research are to use exercise or a combination of exercise and
Features
SSRIs and SSNIs don’t work for all patients and have several associated side effects.
Presynaptic Cell
SSRI
Seratonin
Seratonin channel
Seratonin receptor Postsynaptic Cell
The addition of SSRI’s block the re-uptake of seratonin by the presynaptic cell.
fall 2018 | PENNSCIENCE JOURNAL
13
Features
medications [7]. Another way to treat mental illnesses is through behavioral therapy. The main form of behavioral therapy is cognitive behavioral therapy (CBT). CBT targets both cognitive and behavioral factors involved in mental illnesses. CBT involves strategizing with a psychologist about how to deal with certain situations by identifying misconceptions or wrong assumptions which lead to harmful thoughts [8]. Patients then monitor their thoughts and substitute any harmful thoughts arising from misconceptions with more realistic thoughts. CBT also develops methods to deal with negative thoughts and behaviors. Current research suggests CBT is just as effective as, if not better, at dealing with depressive symptoms compared to prescription drugs for symptoms associated with bipolar disorder. However, research has suggested that the efficacy of CBT is dependent on the condition. For example, CBT is not very effective at controlling aggression in patients with schizophrenia. Thus, the comparison between drug treatments and CBT still needs to be considered [9]. A similar method of behavioral therapy that is based on CBT is dialectic behavior therapy (DBT) which involves developing acceptance and change strategies. DBT
involves 4 stages, in which patients develop their coping mechanisms. The first stage is stabilization of thoughts, the second stage involves stabilization of behaviors, the third stage promotes happiness, and the fourth stage develops skills to allow for fulfillment. DBT differs from CBT in developing mindfulness skills whereas CBT involves focusing on reasoning and rationality. Also, DBT fosters therapy in a group setting whereas CBT is generally one-on-one. According to data by Bohus et al. patients who underwent DBT improved at a greater rate than those who didn’t, and DBT was effective at helping every factor except anger in mentally ill patients [10]. Overall, there are two main treatments for mental illness: prescription drugs and behavioral therapy. The obvious question based on the two very different types of treatments is that of efficacy. Since every person is different and both forms of treatment involve very different ways of targeting the condition, it is difficult to say if one is generally better than the other. A possible consideration, however, in the comparison of efficacies of drugs and psychotherapy is patient compliance in terms of listening to treatment prescribed by the psychologist or psychiatrist. Another great alternative involves the simultaneous use of both prescription drugs and behavioral therapies together. Current research has already suggested that optimized treatment would involve both methods, but this idea is perhaps one that can be researched further in the future. Since the two methods target different aspects of mental illness, namely behavior and chemistry, the eventual goal is to help all patients and reduce, if not eliminate, mental illnesses.
References [1] Nami.org. (2018). Mental Health By the Numbers | NAMI: National Alliance on Mental Illness. [online] Available at: https://www.nami.org/Learn-More/Mental-Health-By-the-Numbers [Accessed 4 Nov. 2018]. [2] Adaa.org. (2018). Treatment | Anxiety and Depression Association of America, ADAA. [online] Available at: https://adaa.org/understanding-anxiety/depression/treatment [Accessed 4 Nov. 2018]. [3] Benjamin, E., Virani, S., Callaway, C. et al, (2018). Heart Disease and Stroke Statistics—2018 Update: A Report From the American Heart Association. Circulation, 137(12). [4] Haddad, P. and Dursun, S. (2007). Neurological complications of psychiatric drugs: clinical features and management. Human Psychopharmacology: Clinical and Experimental, 23(S1), pp.S15-S26. [5] Adaa.org. (2018). [online] Available at: https://adaa.org/sites/default/files/SomberQuestions_Antidepressants.pdf [Accessed 4 Nov. 2018]. [6] Dang, Y., Ma, X., Zhang, J., Ren, Q., Wu, J., Gao, C. and Hashimoto, K. (2014). Targeting of NMDA Receptors in the Treatment of Major Depression. Current Pharmaceutical Design, 20(32), pp.5151-5159. [7] Sharma, Ashish et al. “Exercise for mental health” Primary care companion to the Journal of clinical psychiatry vol. 8,2 (2006): 106. [8] Gettingunstuck.com. (2018). [online] Available at: http://www.gettingunstuck.com/cpu/PSY412/Overhead/COGNITIVE%20BEHAVIORA L%20THERAPY%20OH.pdf [Accessed 4 Nov. 2018]. [9] Hofmann, S., Asnaani, A., Vonk, I., Sawyer, A. and Fang, A. (2012). The Efficacy of Cognitive Behavioral Therapy: A Review of Meta-analyses. Cognitive Therapy and Research, 36(5), pp.427-440. [10] Dialectical behavior therapy: current indications and unique elements. Psychiatry (Edgmont). 2006;3(9):62-8.
14
PENNSCIENCE JOURNAL | fall 2018
T (1) Lee, H. J., Macbeth, A. H., Pagani, J. H., and Young W. S. (2009). Oxytocin: the great facilitator of life. Progress in Neurobiology 88, 127-151. (2) Insel, T. R., and Fernald, R.D. (2004). How the brain processes social information: searching for the social brain. Annual Review of Neuroscience 27, 697-722. (3) Feldman, R. (2015). Sensitive periods in human social development: new insights from research on oxytocin, synchrony, and high-risk parenting. Development and Psychopathology 27, 369-395. (4) Bartz, J. A., Zaki, J., Bolger, N., and Ochsner, K. N. (2011). Social effects of oxytocin in humans: context and person matter. Trends in Cognitive Sciences 15, 301-309. (5) Kosfeld, M., Heinrichs, M., Zak, P. J., Fischbacher, U., and Fehr, E. (2005). Oxytocin increases trust in humans. Nature 435, 673-676. (6) Cho, M. M., DeVries, A. C., Williams, J. R., and Carter, C. S. (1999). The effects of oxytocin and vasopressin on partner preferences in male and female prairie voles (Microtus ochrogaster). Behavioral Neuroscience 113, 1071-1079. (7) Windle, R. J. (1997). Central oxytocin administration reduces stress-induced corticosterone release and anxiety behavior in rats. Endocrinology 138, 2829-2834.
Features
here is an inevitable crossroad between our subjective desire for interconnectedness and the scientific understanding of that desire — the drive to understand the fundamental basis from which we are drawn to one another. At the heart of our desire? Oxytocin (OT): the neurochemical arguably responsible for feelings ranging from self-actualization to codependency. The understanding of OT is fundamental to the work of neuroscience researchers and enthusiasts alike. From the psychological compulsions of the mind, to the neurobiological explanations of the body, the generality of oxytocin is hardly insignificant. In simple terms, OT is a peptide secreted by the hypothalamus and transported to the posterior lobe of the pituitary gland where neurochemical interactions give rise to its flagship sense of attachment [1]. Though an understanding of the psychological, behavioral, and neurochemical determinants of OT should be conveyed for a thorough picture of its overall effects on the developing mind, an understanding of its biological implications is sufficient for practical deliberation of its importance. In recent years, researchers looked to the amygdala for an understanding of human interconnective behavior. Scientists have demonstrated that the component of fear that underpins attachment is explained by the process of neurochemical activation from the hypothalamus, particularly for infants and their maternal figures [2]. Persistent se paration, within a level of cognit i ve
(8) Baumgartner, T., Heinrichs, M., Vonlanthen, A., Fischbacher, U., and Fehr, E. (2008). Oxytocin shapes the neural circuitry of trust and trust adaptation in humans. Neuron 58, 639-650. (9) Kirsch, P. (2005). Oxytocin modulates neural circuitry for social cognition and fear in humans. Journal of Neuroscience 25, 11489-11493. (10) Kumsta, R., Hummel, E., Chen, F. S., and Heinrichs, M. (2013). Epigenetic regulation of the oxytocin receptor gene: implications for behavioral neuroscience. Frontiers in Neuroscience 7.
X OXYTOCIN A Hormone of the Body, the Love from Within Ourselves
By Andrew Lowrance Designed by Felicity Qin
fall 2018 | PENNSCIENCE JOURNAL
15
Features
growth, impairs the development of basic psychological cues, illustrated by the release of OT. This region of time concerning perpetual psychological neglect, as within a child’s early years, is most commonly aligned with the notion of sensitive periods (SP), where basic physical and psychological phenomena play a greater role in shaping our sense of consciousness and well-being [3]. Studies have demonstrated a statistically significant link between the perpetuation of SP’s and the promulgation of neural plasticity, with OT playing a key role in the reconciliation of this emotional advancement to a state of psychological equilibrium. To effectively delineate the process by which the SP impacts infants, scientists looked to the neurochemical basis by which OT diffuses. Experiments have been conducted to understand the relationship between OT and psychological codependency using both human and non-human subjects. However, attempts by researchers to correlate the human sense of close interaction and the neurochemical basis from which that euphoria is de-
rived using non-human brains will arguably produce less conclusive results; contextualization is a critical factor to be considered when understanding the behavior of humans, as opposed to non-human subjects [4]. However, for ethical and financial reasons, scientific experiments involving human subjects are often not feasible. Non-human testing has therefore been utilized by the scientific community for decades in the pursuit of greater neurobiological understanding. These studies demonstrated linear* relationships between the release of OT throughout the hypothalamus and the corresponding sense of apparent emotional attachment [5]. Although these experiments were conducted on animal models, comparative physiology has demonstrated the remarkably similar composition and function of human and non-human brains. That is, similar patterns of OT diffusion within non-human subjects provide insight into the possibilities of understanding the chemical basis of human behavior. For instance, non-human variables, such as rats, were used for investigating OT’s socio-biological
* (for purposes of practical understanding, linear in this context refers to the direct link between two given factors within a degree of statistical certainty, without significant consideration of outlying factors)
16
PENNSCIENCE JOURNAL | fall 2018
Features
components and how it heightens the nociceptive (pain) thresholds, thereby reducing relative anxiety levels within an organism [6]. An analytical approach was considered in devising a breakdown of the fundamental effects of OT in rats through varying dosages to determine the associative behavioral and neurochemical attributes of the additional peptide, thereby assessing the hypothalamo-pituitary-adrenal (HPA). Measurable changes in neurochemical balances within the amygdala were recorded, with a coexistent shift in characteristic behavior toward a reduced state of anxiety [7]. Similarly, a study concerning the internasal, double-blind administration of OT within human test subjects, subsequently analyzed by an fMRI, found associative trust adaptations linked to the administration of OT. The amygdala, midbrain region, and dorsal striatum were specifically activated in subjects receiving OT [8]. Thus, the link between OT increase and certain behavioral outputs, like trust, signifies a corresponding shift in the diffusion rate o f
O T through targeted neural pathways. Likewise, from the fundamental link between human and non-human testing, OT reveals an association between the complex emotional states derived from subjects and consequent chemical distribution. More particularly, there is a symmetry between the human brain and OT’s connection to attachment, social recognition, and aggression in non-human mammals: strong evidence conveys the perpetuation and modulation of OT within the amygdala, as in the instance of SP’s with younger subjects. A study concerning 15 healthy males after double-blind crossover intranasal application of OT and a placebo illustrated the possible mediation of the amygdala by OT. Thus, despite the non-linear relationship between OT diffusion and consequent behavior, a practical understanding of OT’s implicative characteristics refines the bridge between human and non-human testing [9]. The mind-brain symmetry that concerns psychological behavior unravels a necessary beauty to the universe, a universe that could be sought within the lives of individuals and the interconnectedness of their minds. More particularly, with respect to the continuity of time, human evolution necessitates the precision with which this “love drug” reshapes the human sense of desire and the innate composition of the self.
fall 2018 | PENNSCIENCE JOURNAL 17
Features
the M
By Pranshu Suri
T
P
W
T
F
I O I epidemic
S
D
By Pranshu Suri Designed by Amara Okafor
M
ore people in the U.S. die every day from opioid overdoses than car accidents [1]. This fact may strike some as surprising, but, to those who are close to the rampant opioid epidemic, it is a tragic reality. Deaths resulting from opioid painkillers have been steadily increasing in the U.S. over the past 20 years, so it is imperative that we are informed about this sphere of public health. Despite all of the coverage on the opioid epidemic, many Americans report still not knowing exactly what an opioid is and why they are so uniquely addictive. Opioids are prescription narcotics that act to relieve pain [2]. The most common trigger event for long-term addiction seems to be a prolonged use and/or misuse of prescription painkillers such as Oxycontin, Percocet, and Vicodin [3]. When someone consumes an opioid drug, the drug molecules rapidly enter into their bloodstream; from there, they travel through the body to the brain and bind to individual proteins called m-opioid receptors that exist on the surface of opioid-sensitive neurons [3]. Specifically, these neurons and receptors are interconnected with the reward pathways in the human brain. These reward pathways run through a region called the limbic system, which is located in the center of the brain [3]. The limbic system has many functions ranging from the regulation of emotions and sensory processing to the formation of memories [4]. The aforementioned opioid-sensitive neurons
18
PENNSCIENCE JOURNAL | fall 2018
Illustration of a brain not on drugs versus a brain of a person who regularly uses. Opioids prompt an increase in neurotransmitters like serotonin, dopamine, etc. This creates feelings of pain relief and euphoria. are found in a region of the limbic system called the ventral tegmental area (VTA) [4]. Thus, when a person takes an opioid painkiller, the opioid-sensitive neurons in the VTA release dopamine, the “feel-good� neurotransmitter, into the surrounding regions [3]. Under normal circumstances, the feelings of pleasure caused by this dopamine release serve to counteract any pain one might be feeling at that given time; however, when the person consuming the drug is not feeling any pain to begin with, the limbic system’s essential functions are disrupted [5]. Opioids can make the user feel extremely happy, content, and relaxed; because the limbic system is also responsible for the formation of memories, the brain associates each subse-
quent drug consumption with this skewed perception of reality [4]. Addiction plagues millions across the country, for whom there is currently no discernible solution. For years, lawmakers have debated various policy ideas to ameliorate this pervasive epidemic, to no avail. With tens of thousands of opioid-related deaths each year, the economic toll on the U.S. has accumulated to over $1 trillion and is only expected to increase in coming years [6]. Despite the increased use of registries and alternative or holistic therapies to limit the over-prescription of painkillers, the opioid epidemic in America persists, which arguably indicates deeper underlying issues that cannot be solved simply by limiting the number of prescriptions doctors are allowed to write [7]. As it turns out, the opioid crisis that America faces today cannot be traced to just one origin; it is a complex and multifaceted issue that has many underlying causes. More often than not, addiction typically arises from taking a higher-than-normal dosage in order to control pain. People who overdose on opioids to curb their pain often aren’t aware of other treatment options and rehabilitative services for pain, so they default to opioids [8].
Besides this, drug companies heavily market their opioid products to the general public, which makes people more predisposed to turn to opioids when they are in pain, even if such highly addictive drugs are not necessarily needed [8]. Although it may seem counterintuitive, the government’s attempt to limit the number of prescriptions that physicians are allowed to write for opioid drugs may actually be contributing to the problem rather than solving it. Because of this policy, in recent years, prescription drugs have become harder to acquire and some individuals reliant on prescription opioids turned to illegal drugs such as heroin and fentanyl [7]. Almost half of the deaths linked to opioids in 2016 were due to these illicit drugs [7]. In many cases, those prone to addiction are not aware of other treatment options for their pain and instead select prescription painkillers, which are not always necessary. Learning about and spreading knowledge of the opioid crisis—even down to how these drugs interact with our bodies biochemically—is crucial in order to ensure that we are making informed decisions and is the first step to effectively managing this public health epidemic.
Features
GABA
Morphine Opioids
GABA
GABA Receptors
DOPAMINE
Dopamine
Dopamine Receptors
Opioids and morphine hyper polarize GABA neurons which then suppress the release of GABA. Usually, GABA inhibits the release of Dopmaine– opioids, however, supress GABA which then inceases the release of Dopmaine.
references [1] Durkin, E. (2018). US drug overdose deaths rose to record 72,000 last year, data reveals. The Guardian. [2] (2018). Opiate Addiction/Opioid Addiction. (https://www.therecoveryvillage.com/opiate-addiction/#gref) [3] Kosten, T., and George, T. (2002). The Neurobiology of Opioid Dependence: Implications for Treatment. Science & Practice Perspectives 1, 13-20. [4] Poot, M., Yuska, B., and Tolen, L. (2018). Drugs and the Limbic System. [5] Nabipour, S., Ayu Said, M., and Hussain Habil, M. (2014). Burden and Nutritional Deficiencies in Opiate Addiction- Systematic Review Article. Iranian Journal Of Public Health 43, 1022–1032. [6] Mangan, D. (2018). Economic cost of the opioid crisis: $1 trillion and growing faster. CNBC. [7] Hellmann, J. (2018). What caused the opioid crisis?. The Hill. [8] (2018). The Origin and Causes of the Opioid Epidemic. (http://www.georgetownbehavioral.com/node/2013)
fall 2018 | PENNSCIENCE JOURNAL
19
Features
Digital Therapeautics: The Doctor that Fits in Your Pocket
By Neelu Paleti Designed by Jessica Tan 20
PENNSCIENCE JOURNAL | fall 2018
T
he first picture that comes to mind when one thinks about pharmaceuticals is a doctor prescribing a chalky white pill. Yet, this picture may change quite significantly in the near future, where a touch of a button on your smartphone could replace the need for that doctor. From apps about birth control to opioid addiction, there are a range of digital tools that are beginning to change the way we interact with drugs on a daily basis. This emerging intersection between drugs and technology is shifting the power of medical decision-making from doctors to the apps and their users. Each of these applications, with its new range of accessible, personalized treatments, is poised to have an unprecedented impact on the way patients receive care.
the emerging intersection between drugs and technology is shifting the power of medical decision-making from doctors to the apps and their users.
“
“
Features
administration is diabetics who must track insulin dosage. Whenever it is time for an insulin shot, patients must calculate how much insulin to inject. Apps such as mySugr, PredictBGL, and Diabetes:M fulfill this function through a device that sits under the skin. The connected apps project the predicted blood glucose levels and allow users to visualize these trends. Additionally, the apps serve as daily logs where users can input information about diet and exercise to yield more accurate projections of blood glucose level [4]. By awarding points to anything from logging carbs to making a periodic endocrinologist appointment, PredictBGL formas its interface as a game involving healthy competition between users [5]. Eventually, this can maximize patient independence and efficiency, offering users a seamless
12:00 PM
Touted to be the first FDA-approved digital birth control app, Natural Cycles represents a new form of individualized birth control directly available on a phone. Using daily basal temperature measurements and monthly menstrual cycle data, the application’s algorithm analyzes trends to show whether or not the user is fertile. This algorithm observes an array of metrics such as sperm survival, cycle length, ovulation day, and temperature fluctuations to produce results; it displays a red light to indicate fertility and a green light otherwise [1]. This application mainly serves to collect the necessary data and notify users of when to take further measures to prevent pregnancy. According to a study conducted by the European Journal of Contraception and Reproductive Health Care, this app is shown to be 93% effective, and by adding more data such as luteinizing hormone test results, this app can be more accurately tailored to the user [2]. Despite its high success rate, Natural Cycles depends on consumer compliance in recording basal temperatures five times a day for best results. Additionally, this app’s algorithms depend on previous body patterns; any sudden hormonal imbalances could lead to incorrect results. Yet, amidst such drawbacks, this app serves as an addition to the toolbox of birth control methods, thereby reducing the necessity for doctors in women’s pregnancy and health with a smartphone [3]. Another patient population where technology can aid drug
fall 2018 | PENNSCIENCE JOURNAL
21
incentive to keep their diabetes under control through self-monitoring. Furthermore, Diabetes:M is able to collect data on diet and daily patterns of blood glucose or insulin activity and send them directly to doctors, further bridging the gap between patients and healthcare providers [6]. The blood glucose levels that once required careful examination at periodic checkups can now be sent remotely within moments, yet another advantage of these newer drug technologies. In addition, smartphone apps are now able to utilize the efficient and customizable features of the digital space to treat a population of users coping with drug abuse and addiction recovery. One such app, Triggr, can not only help prevent cravings and relapse but also use data from screen engagement, texting patterns, phone calls, and sleep trends to predict an episode of relapse and provide necessary intervention [7]. By initially inputting a user’s drug preferences, relapse history, trigger words, and other personalized data, the app’s algorithms are able to implement machine learning techniques to search for trends and alert a recovery team member or a user’s specific care team. This offers care providers a user-friendly platform to consistently track high-risk patients and prevent relapse and overdose with 92% accuracy right from their smartphones [7]. Similarly, Reset, the References:
[1] (2018). FDA allows marketing of first direct-to-consumer app for contraceptive use to prevent pregnancy. U.S. Food & Drug Administration. https://www. fda.gov/newsevents/newsroom/pressannouncements/ucm616511.htm [2] Our research. Natural Cycles. https://www.naturalcycles.com/en/science/ research/ [3] Wakeman, J. (2018). FDA approves controversial birth control app. Healthline. https://www.healthline.com/health-news/fda-approves-contro versial-birth-control-app [4] Huckvale, K., Adomaviciute, S., Prieto, J. T., Leow, M. K. S., and Car, J. (2015). Smartphone apps for calculating insulin dose: a systematic analysis. BMC Medicine 13. [5] Ward, J. (2017). PredictBGL is a diabetes management app predicting fluctuation in blood sugar levels. Startup Daily. http://www.startupdaily.net/2017/06/ predictbgl-diabetes-management-app-fluctuations-blood-sugar/ [6] The ultimate way to understand and manage your diabetes. Diabetes-M. https://www.diabetes-m.com/ [7] Byrnes, N. (2017). Treating addiction with an app. MIT Technology Review. https://www.technologyreview.com/s/604085/treating-addiction-withan-app/ [8] (2017). FDA permits marketing of mobile medical application for substance use disorder. U.S. Food & Drug Administration. https://www.fda.gov/NewsEvents/Newsroom/PressAnnouncements/ucm576087.htm [9] Singer, N. (2018). Take this app and call me in the morning. New York Times. https://www.nytimes.com/2018/03/18/technology/take-this-app-andcall-me-in-the-morning.html [10] Loy, J., Ali, E. E., and Yap, K. Y. (2016). Quality assessment of medical apps that target medication-related problems. Journal of Managed Care & Specialty Pharmacy 22, 1124-1140
22
PENNSCIENCE JOURNAL | fall 2018
first FDA-approved health app to target substance abuse, can track triggers including hunger, anger, loneliness, social pressures, and other situations to report results back to the user’s doctor.8 Providers can then track the trends of these triggers and cravings to either reward the user or prescribe them further treatment. A typical 30-day inpatient treatment costs insurers and government agencies, such as Medicaid, an average of $17,000 per person [9]. Furthermore, the cost of treating hepatitis C, a disease highly associated with injection drug abuse, is around $5 billion per year [9]. In addition to preventing such costs, these drug technologies could spare a portion of the expenditure spent towards rehabilitation visits and periodic doctor appointments. Using a customizable and accessible approach, these digital therapeutics could offset the traditionally time-consuming aspects of healthcare required
“
While drug technologies can offer a range of new treatment possibilities, the road to definitive digital treatments is not yet complete.
“
Features
for this population of substance abuse patients. The rise of digital therapeutics offers an accessible, remote way of interacting with drugs without the constant need for doctor-patient interaction. However, it is imperative to consider the downsides to user-based treatment avenues. Many of these apps and the corresponding treatments rely heavily on user compliance and the accurate input of information. A slight mistake while logging blood glucose data or a user’s failure to consistently feed data on substance abuse triggers can yield incorrect results from these machine-learning based apps [10]. In addition, such apps are often used without proper medical supervision, which could cause users to misinterpret the results. For example, insulin injections may be taken at unnecessary times if patients misunderstand the app’s data. This also raises the question of liability between the app, doctors, and patients if treatment steps are taken without proper guidance. Furthermore, treatment is limited to the patient population with access to smartphones. While drug technologies can offer a range of new treatment possibilities, the road to definitive digital treatments is not yet complete. Digital therapeutics has pushed the image of pharmaceuticals far past that of a doctor and a pill, and this path to effective, accessible treatments continues to unfold with the introduction of newer innovative technologies each day.
Features
Cannabis Oil
A Health Trend to Weed Out? By Emily Lo Designed by Abi Szabo
“C
annabis” is a term that perhaps leads you to conjure the image of a popular yet illicit five-leaf plant known for its psychoactive effects on the brain. The characteristic high that marijuana induces can be attributed to tetrahydrocannabinol (THC): the primary active ingredient in cannabis that binds to the human brain’s cannabidiol-1 (CB1) receptor and interferes with neurotransmitter signaling. Yet today, cannabis also garners increasing attention from non-prescriptive drug consumers who believe the substance could hold promising anti-inflammatory properties. They are interested in cannabidiol (CBD), marijuana’s second most prominent cannabinoid (chemical constituent of cannabis), which can be contained in an insoluble liquid extract known as cannabis oil [1]. Unlike THC, CBD does not bind to CB1 and is
fall 2018 | PENNSCIENCE JOURNAL 23
Features
in fact legal and non-intoxicating. People use cannabis oil to treat conditions ranging from acne to colds, some even speculating that it could address cancer and Alzheimer’s Disease [2]. Major companies such as Coca-Cola have also taken note of the substance’s increasing popularity and are exploring the possibility of incorporating CBD into “functional wellness beverages [3].” Yet, despite recognition of cannabis oil’s potential
THese results suggest that cannabis oil has the ability to alleviate the frequency of seizures among Epileptic patients in a dose-dependent manner. medicinal value by researchers, industry leaders, and average consumers alike, the product also attracts skepticism as its benefits have yet to be fully characterized and understood. As the market for cannabis oil continues to expand on an international scale, many are encouraged by existing evidence that the substance could improve the lives of those suffering from specific chronic and debilitating illnesses. Much of this proof is anecdotal--for instance, this past June, a story about Billy Caldwell, a severely epileptic twelve-year-old boy from Britain, made headlines [4]. Cannabis oil was able to alleviate his life-threatening seizures, which intensified when his CBD was confiscated at an airport. Following the controversy spurred by this high-profile case, United Kingdom Home Secretary Sajid David announced on October 11th that medical cannabis oil will be approved for prescription use starting November 1st [5]. In addition, earlier this year on June 25th, the Federal Drug Association (FDA) approved Epidolex, “an oral solution of cannabis oil,” to address severe forms of epilepsy such as Lennox-Gastaut syndrome (LGS) and Dravet syndrome (DS), both of which are challenging to address using other medications [6]. This followed three phase 3 randomized double-blind, placebo-controlled clinical trials that showed Epidolex decreased 24
PENNSCIENCE JOURNAL | fall 2018
the frequency of seizures when taken concurrently with other medications by human patients [7]. The most recent of the three studies, published this past May in the New England Journal of Medicine, examined the efficacy of a CBD solution by collecting data from 225 patients. Non-placebo patients were orally dosed with CBD at either 10 mg/kg of body weight or 20 mg/kg of body weight. The researchers found that patients treated with CBD experienced a greater percent reduction in drop-seizure frequency compared to the placebo group [7]. These results suggest that cannabis oil has the ability to alleviate the frequency of seizures among epileptic patients in a dose-dependent manner. Furthermore, this correlation highlights the substance’s potential to combat symptoms of epilepsy and provide relief. “Lennox-Gastaut is one of the most difficult epilepsy conditions to treat,” notes Dr. Elaine Wirrell, the director of Pediatric Epilepsy at the Mayo Clinic and a contributor to the aforementioned study. She adds, “Having another treatment that holds promise to significantly reduce drop seizures is important [7].” However, in all three Epidolex studies, adverse effects such as sleepiness, elevated liver enzymes, and diarrhea were also noted. Given the benefits and risks of cannabis oil that have been uncovered thus far, Epidiolex researcher Dr. Jacqueline French notes that its use would need to be monitored by a physician [7]. As CBD has yielded promising results when used to address epilepsy and psychological conditions, it continues to attract interest and spark hope among patients, doctors, and researchers seeking effective solutions to various medical conditions. Nonetheless, there remains concern regarding the substance’s novelty and the adverse effects tied to cannabis oil. According to Marcel Bonn-Miller, faculty at the Perelman School of Medicine, “It’s important to know that the research in this area is in its infancy, partly
Features
because we haven’t really understood much about CBD until relatively recently [9].” Scientific evidence regarding the benefits of cannabis remains limited, and many currently question the substance’s inconsistent physiological impact on consumers. Considering that there are many varieties of CBD oil with differing concentrations and chemical compositions, it is imperative for future research to explore the nature of this compound so that consumers are aware of proper dosage and potential side effects. As Martin Lee, director of Project CBD, remarks, “There really is a scientific basis for understanding why CBD can work, but we’re still really a long way of mastering the hows [1].” As studies and success stories fuel the world’s growing fixation on cannabis oil, the substance is likely to retain its popularity for the time being. It may be that the assessment of the safety and efficacy of cannabis oil continues to pose a challenge to the scientific community. The liquid extract also attracts increasing skepticism, and its popularity has created the demand for scientific proof validating its status as a billion-dollar industry and an ever-growing health trend [1]. Yet, one thing is certain: there is
more to be discovered about this substance’s properties and value in the medical world. While the emergence of CBD indeed holds great promise, only time and additional research can truly dictate how and when this compound should be used in our lives. References:
[1] Snyder, C. (2018). What is CBD oild and how did it become a $1 billion industry? Business Insider. https://www.businessinsider.com/what-is-cbd-oil-how-made-marijuana-cannabis-plant-health-2018-8 [2 Johnson, J. (2018) Everything you need to know about CBD oil. Medical News Today. www.medicalnewstoday.com/articles/317221.php [3] (2018). Coca-Cola eyes cannabis oil market. Web MD. www.webmd.com/diet/ news/20180918/coca-cola-eyes-cannabis-oil-market. [4] (2018). Billy Caldwell ‘could die’ unless given cannabis oil, says mum. BBC News. https://www.bbc.com/news/uk-44504142 [5] Tobin, O. (2018). Medicinal cannabis oil will be available on prescription by next month, Sajid Javid says. Evening Standard. https://www.standard.co.uk/news/uk/ medicinal-cannabis-oil-will-be-available-on-prescription-by-next-month-sajid-javidsays-a3959866.html [6] (2018). FDA approves first drug comprised of an active ingredient derived from marijuana to treat rare, severe forms of epilepsy. U.S. Food & Drug Administration. https://www.fda.gov/NewsEvents/Newsroom/PressAnnouncements/ucm611046.htm [7] Ellis, F. J. (2018). FDA approves oral cannabidiol for Dravet and Lennox-Gastaut Syndromes. Neurology Today. https://journals.lww.com/neurotodayonline/blog/ breakingnews/Pages/post.aspx?PostID=740 [8] Stern, C. A. J., and Gazarini, L., Takahashi, R. N., Guimarães, F. S., and Bertoglio, L. J. (2012). On disruption of fear memory by reconsolidation blockade: evidence from cannabidiol treatment. Neuropsychopharmacology 37, 2132-2142. [9] Hickock, K. (2018). What is CBD oil, and does it really work? LiveScience. https:// www.livescience.com/63452-what-is-cannabis-oil.html really work? LiveScience. https://www.livescience.com/63452-what-is-cannabis-oil. html
fall 2018 | PENNSCIENCE JOURNAL
25
Features
Developing Drugs:
Women in Clinical Trials By Rose Nagele Designed by Julia Davies
I
n 1977, Andrea Goldstein filed a lawsuit against drug manufacturer Eli Lilly and Company. The drug in question, diethylstilbestrol, or DES, had been prescribed to Goldstein’s mother, Mrs. Schwartz, while she was pregnant in 1953. [1] At the time, DES was commonly prescribed to pregnant women to prevent miscarriages. During the 1970s, it became clear that the drug was not only ineffective, but also detrimental to the resulting children. The DES prescribed to Mrs. Schwartz caused abnormalities in the development of Goldstein’s uterus, leading to complications with pregnancy, miscarriage, and eventually infertility. Determining accountability for the harms of a drug prescribed and manufactured a generation ago, however, is hardly straightforward. Eli Lilly and Company
26
PENNSCIENCE JOURNAL | fall 2018
battled the lawsuit against Andrea Goldstein for a decade as her symptoms developed. Goldstein was only one of many women, called DES daughters, calling for restitution for the harm caused by DES manufacturers like Lilly. As recently as 2013, Lilly faced yet another DES lawsuit, brought by four sisters, each of whom had developed breast cancer in their 40s. The fallout surrounding DES treatments illustrates the complicated history of the under-representation of women in clinical trials. DES was one of several problematic medications offered to pregnant women in the 1950s and 60s, including thalidomide, an anti-nausea drug that caused severe limb malformations in newborns. These cases brought to light the differential responses of men and women to certain compounds. While much has changed in the conduct of medical research to address these nuances, issues of safety and representation in drug development continue to this day. Every drug approved by the FDA undergoes three phases of clinical trials that investigate both the efficacy of a drug at treating a specific condition and the drugs safety (see figure). Drugs are first tested on animal models before they are moved to human subjects in clinical trials. Phase I is focused on ensuring there are no extreme side effects and involves a small number of patients. With each phase, the sample size becomes larger and the questions more nuanced. This process is intended to minimize risk to human subjects while developing the drug. While healthy subjects are usually used in early trials when the effects of the drug are still hazy, in order for the drug to effectively treat a specific illness, subjects with that condition must be included in developmental studies. A number of factors cause differences in the way individuals respond to drugs. The genetic and societal differences between men and women is just one example. Some molecular, physiological, and behavioral differences between men and women impact their susceptibility to disease their response to treatment. Whether a person has an XX or XY genotype impacts the composition of hormones and chemicals in the bloodstream, organ development, metabolism, and other physiological processes. For example, women (XX) produce more of the hormone estrogen, while men (XY) produce more androgens. These hormones are thought to account for many sex-specific responses to treatments such as quinidine, a drug used to treat irregular heart rhythms. Women who are given quinidine are more likely to have a potentially lethal reaction called “torsades de pointes” than men who are prescribed the same drug. The different physiological responses to quinidine among men and women demonstrate that treatments marketed to both men and women should undergo clinical trials that account for those differences. The DES crisis did not immediately result in efforts to safely include women in the drug development process, however. Instead, the FDA implemented policies that restricted women
Features
from participating in clinical trials on the basis of protecting a vulnerable population. In 1977, the FDA released “General Considerations for the Clinical Evaluation of Drugs,” a guideline that effectively excluded women from early phase clinical trials. In order to protect “the fetus from unanticipated exposure to potentially harmful drugs,” all women “of childbearing potential” were to be excluded from Phase I and II trials, unless they had a life-threatening condition. A woman of childbearing potential” was defined as a “premenopausal female capable of becoming pregnant,” which meant that even women who had no plans to become pregnant were excluded. This policy received criticism for its paternalistic denial of the “autonomy and decision-making capacity” of women. Moreover, it did little to actually improve the quality of medications available to pregnant women. Medical professionals had to assume that the drugs they prescribed female patients would have the same effectiveness and side effects as they had on men. And since the effects on a developing fetus were virtually unknown, many pregnant women were advised to stay off medications altogether. The predominately male composition of the research field in the twentieth century also biased clinical trial design. In 1978, about 20% of medical school graduates were female. This was up from about 6% in 1950. Roughly the same ratios existed for students earning science related doctorates. When men are the ones planning and carrying out research, men are more likely to be the study subjects. Men are innately more likely to assume the male form as a suitable model of human physiology. Quantifying the degree to which women are disadvantaged due to regulatory and social restrictions, however, has not been simple. A 1994 committee on “Ethical and Legal Issues Relating to the Inclusion of Women in Clinical Studies” stated that determining “whether women have been disadvantaged by policies and practices regarding their participation or by a failure to focus on their health interests in the conduct of research, are hindered by a scarcity of reported data.” The lack of data, however, may have only served to reinforce the demand for changing practices. In the 1990s, the FDA began to address both the regulatory and social restrictions of women in clinical trials. The 1977 restrictions were finally loosened in 1993 with the “FDA Guidance Study and Evaluation of Gender Differences in the Clinical Evaluation of Drugs.” A Special Report on the new guidances acknowledged that “it is possible to reduce the risk of fetal exposure through protocol design” without excluding women from early clinical trials. To ensure that those possibilities would be pursued, however,
fall 2018 | PENNSCIENCE JOURNAL 27
Features
the FDA needed to address the bias of the male dominated research field. An advisory group called the Office of Women’s Health formed at the FDA in 1994 to advise officials on women’s health, recommend policy and initiatives, and promote the inclusion of women in clinical trials. A parallel office opened at the National Institutes of Health (NIH), the organization that oversees the distribution of federal funds for health and science research. The Office of Research on Women’s Health at the monitors the inclusion of women in clinical trials and the allocation of federal grants for research on women’s health. Perhaps the most effective way to reduce bias, however, is to increase the diversity of those carrying out the research. An increasing number of women are entering scientific and medical professions, and experimental designs are changing to reflect the populations they come from. In the early 2000s, a number of studies on the presence of women in clinical trials indicated that representation may have improved following these institutional changes. A 2003 review of the inclusion of sex and gender in medical studies noted a shift in perspective among researchers. While historically “the research agenda” for women’s health was limited to their reproductive capacity, research has expanded to include more nuanced concerns for sex-based variation in disease. The author identified sex-based variation in the prevalence, symptoms experienced, and responses to treatment for numerous diseases, including HIV, type-2 diabetes, irritable bowel syndrome, chronic fatigue, autoimmune diseases, and heart disease. Though the literature on women’s health has grown, more research can be conducted, as a review published in Pharmacy Practice in 2016 points out. The biological differences between men and women are complex, with implications on cellular, physiological, and morphological scales. The life experiences and external pressures facing men and women--which are related to but distinct from their genetic makeup--also impact their health. While women are generally seeing more representation in clinical trials, the issues surrounding inclusion of pregnant women are still unresolved. Women with chronic conditions, like multiple sclerosis, diabetes, hypertension, and psychiatric disorders, face the decision of whether to continue taking drugs while pregnant. There is little research on how these drugs may interact with the physiological changes brought on by pregnancy, but stopping treatment can pose significant risk to the mother. In April 2018, the FDA released a draft guidance titled “Pregnant Women: Scientific and Ethical Considerations for Inclusion in Clinical Trials.” The document, which “contains nonbinding recommendations,” acknowledges that “development of accessi28 PENNSCIENCE JOURNAL | fall 2018
ble treatment options for the pregnant population is a significant public health issue.” Whether the guidance, which permits the inclusion of pregnant women in clinical trials provided certain conditions are met, results in an improved body of knowledge regarding the effects of medications in pregnant women is yet to be seen. All clinical trials must balance the risk imposed on the study subjects with the need to produce new and effective treatments. While no drug can ever be 100 percent safe due to the complexity and variation of human biology, careful adjustments to study design can go a long way to improving treatment. The past few decades have seen improved representation of women in clinical trials as a result of policy changes and the recognition of women themselves as capable researchers. This improvement exemplifies how equitable regulations as well as diversity in the medical field results in the development of more effective treatments. In order for women as well as other underrepresented populations, such as ethnic minorities or gender-nonconforming individuals, to have access to effective treatment, they must have a presence in the research process. References:
[1] Silbersweig, S. E. (1980). Payton vs. Abbott Laboratories: an analysis of the Massachusetts DES class action suit. American Journal of Law and Medicine 6, 243-282. [2] Rheingold, P. D. (1977). Litigation involved DES. Women Health 1, 26-27. (2013). Settlement reached in Eli Lilly pregnancy drug linked to breast cancer case. CBS News. https:// www.cbsnews.com/news/settlement-reached-in-eli-lilly-pregnancy-drug-linked-to-breast-cancer-case/ [3] Institute of Medicine (US) Committee on Ethical and Legal Issues Relating to the Inclusion of Women in Clinical Studies. (1994). Women’s participation in clinical studies. In Women and Health Research. Mastroianna, A. C., Faden, R., and Federman, D., ed. (Washington DC: National Academies Press). [4] (2017). Investigational new drugs: FDA has taken steps to improve the expanded access program but should further clarify how adverse events data are used. U.S. Government Accountability Office. https://www.gao.gov/products/GAO-17-564 [5] Benton, R. E., et al. (2000). Greater quinidine-induced QTc interval prolongation in women. Clinical Pharmacology & Therapeutics 67, 413-418. [6] (1977). General considerations for the clinical evaluation of drugs. U.S. Food & Drug Administration. (Rockville, MD: U.S. Department of Health). [7] (2010). Diversity in the physician workforce: facts and figures 2010. Association of American Medical Colleges. https://www.aamc.org/download/432976/data/factsandfigures2010.pdf [8] Staff Care: an AMN Healthcare Company (2015). Women in Medicine:  A Review of Changing Physician Demographics, Female Physicians by Specialty, State and Related Data (Irving, Texas). [9 ] National Center for Science and Engineering Statistics (2018). TABLE A-2. Citizenship status, race/ethnicity, and sex of Ph.D.s, by field of doctorate: 1975–99 total and 5-year cohorts from 1975. [10] Mastroianni, A., Faden, R., and Federman, D. (1994). Women and Health Research: Ethical and Legal Issues of Including Women in Clinical Studies: Volume I. (Washington, D.C.: National Academy Press). [11] McCarthy, L., Milne, E., Waite, N., Cooke, M., Cook, K., Chang, F., and Sproule, B. (2017). Sex and gender-based analysis in pharmacy practice research: A scoping review. Research In Social And Administrative Pharmacy 13, 1045-1054. [12] (2018). Regulations, Guidance, and Reports related to Women’s Health. [13] Merkatz, R., Temple, R., Sobel, S., Feiden, K., and Kessler, D. (1993). Women in Clinical Trials of New Drugs -- A Change in Food and Drug Administration Policy. New England Journal Of Medicine 329, 292-296. [14] (2018). Office of Women’s Health. [15] Miller, M. (2001). Gender-Based Differences in the Toxicity of Pharmaceuticals—The Food and Drug Administration’s Perspective. International Journal Of Toxicology 20, 149-152. [16] (2018). Medical scientists, & life scientists, all other. [17] (2018). Physicians & surgeons. [18] National Science Foundation (2018). Survey of Doctorate Recipients; Survey Year 2015. [19] Pinn, V. (2003). Sex and Gender Factors in Medical Studies. JAMA 289, 397. [20] Liu, K., and DiPietro Mager, N. (2016). Women’s involvement in clinical trials: historical perspective and future implications. Pharmacy Practice 14, 708-708. [21] Wizemann, T., and Pardue, M. (2001). Exploring the biological contributions to human health (Washington, D.C.: National Academy Press). [22] Lyerly, MD, MA, A. (2018). Should pregnant women be included in clinical trials?. [23] McCormack, S., and Best, B. (2014). Obstetric Pharmacokinetic Dosing Studies are Urgently Needed. Frontiers In Pediatrics 2. [24] Office of Communications, Division of Drug Information, Center for Drug Evaluation and Research, Food and Drug Administration (2018). [25] Pregnant Women: Scientific and Ethical Considerations for Inclusion in Clinical Trials Guidance for Industry.
interview
H
ow would you define addiction?
developing a substance use disorder. However, the notion of an “addictive personality” is too simplistic; the origin of substance use disorders is quite complex.
I should preface this by saying that addiction is, to some people, a dirty word. Some people view it as a pejorative term, yet it remains In regard to the opioid Interview by Tamsyn Brann widely used and recognized. In epidemic: do you think the fact, our center is the Center problem can be amelioratHenry Kranzler, M.D. is the Benjamin Rush Profor Studies of Addiction. Thus, ed through further refessor of Psychiatry and Director of the Center for I’ll use the term addiction search, or are other kinds Studies of Addiction at the University of Pennsylbecause it remains useful. My of efforts required? vania’s Perelman School of Medicine. He researchdefinition of addiction is imes the genetics and pharmacology of the treatment paired control over a behavior. As with substance use disorof substance abuse disorders, with a specific focus Substance addiction, such as ders, the opioid epidemic, does on precision medicine (the identification of specific drug and alcohol addiction, is not have a simple solution, as medical treatments that match the characteristics impaired control over the use it is complex and multiply-deof individual patients). His work has appeared in of such substances. I emphatermined. A big contributor over 500 journal articles and other publications size impaired control because to the epidemic has been the and he is the Editor of the journal Alcoholism: simply using a substance is not, sharp increase in prescribing Clinical and Experimental Research. by definition, addiction. It’s of opioid pain relievers by the need to actively seek and physicians, which was promotcontinue to use the substance ed by pharmaceutical compadespite knowing that its use nies and advocates for more has adverse consequences that is the aggressive treatment of pain. Howessence of addiction. ever, there’s been a drop in the rate Are there risk factors in a of prescription of opioids in the person’s genome that predisHow can precision medicine past few years. Physicians have been pose him or her to an addictive made aware of the problem, and add to the research already disorder? Is there such a thing there are state prescription monibeing done about addiction? as an addictive personality? toring programs – such as in PennPrecision medicine aims to match sylvania – that require that, before In addition to conducting treattreatment to the characteristics of I prescribe an opioid to a patient, I the individual and the disease being ment research, I study the genetics look up their opioid treatment hisof substance use disorders, that is, diagnosed and treated. Pharmatory, including whether they’ve been the genetic variation (differences cogenetics is relevant both to the prescribed opioids by different docin DNA sequence) that predisposes therapeutic effects of a drug and tors, which would suggest that they people to develop a substance use the adverse effects of a drug. A may be “doctor shopping.” Soludisorder. It’s very clear, for examgenetic variant, for example, could tions like these are new, and they ple, that alcohol dependence is 50 predict who’s going to respond well have begun to reduce over-prescribpercent heritable, which means that ing of these drugs. A troublesome to a medication in terms of the it’s also 50 percent environmental. treatment response but it can also trend that has contributed to the predict who’s going to suffer adverse The same is true for other drugs continued rise in the opioid overas well, including opioids, cocaine, consequences of that medication. dose death rate is that many people and cannabis. Psychological facIf the potential adverse effects are who first became dependent on presevere, knowing who’s at risk is very tors, which are also both genetic scribed opioids have turned to using and environmental, contribute to important clinically. illicit opioids – heroin and the much
Dr. Henry Kranzler
fall 2018 | PENNSCIENCE JOURNAL 29
interview
more potent synthetic opioid fentanyl. What made you pursue this particular branch of research? I started by doing research on medications to help treat alcohol use disorder. There are three medications that have been approved to treat the disorder, and I participated in some of those development efforts. However, the medications have comparatively modest treatment effects. The reasons for that include the fact that alcohol (unlike opioids or nicotine, for which there are specific binding sites, which multiple medications target) has no specific receptors that can be targeted for treatment. Thus, it is much harder to find a drug that’s likely to exert a large treatment effect in patients with an alcohol use disorder. Around the time that I began conducting treatment research, I began collaborating with a colleague at Yale on the genetics of alcohol use disorder, which then progressed to studies of cocaine, opioid, and cannabis use disorders. Those efforts led us to identify genetic variants that appeared to predict the response not only to drugs of abuse but also to therapeutic drugs. Is pharmacogenetics the same as precision medicine? Pharmacogenetics is a major component of precision medicine. Another approach involves animal models that can be used to inform precision medicine research in humans. However, one problem with animal studies is that the findings may not generalize to humans. Although humans and rodents, for example, are mammals, there are some big species differences. For example, most animals won’t self-administer alcohol, so there are procedures that have been developed to facilitate alcohol self-administration studies in animals. Have you worked with animal models before? Not directly, but I collaborate with animal researchers. I am currently working with Mariella De Biasi, a neuropharmacologist here at Penn. Her laboratory is researching a topic that is of particular interest to me and is related to pharmacogenetics. It is a study of an experimental medication that blocks GluK1-containing kainite receptors, a specific kind of glutamate receptor. Glutamate is the most abundant excitatory neurotransmitter in the brain. We chose to look at this receptor subtype because, in a prior study of the medication topiramate in humans, we found that variation in the 30 PENNSCIENCE JOURNAL | fall 2018
gene encoding the GluK1 subunit implicated it as a key target for reducing heavy drinking. Dr. De Biasi’s lab has followed up on that human finding in a mouse model of alcohol drinking to determine the mechanism of the pharmacogenetic effect. Do you find that between the work you’ve done peripherally with animal models and your own work with humans that there’s a specific advantage with using human models? I have chosen to work with humans because I am interested in improving the human condition, and it’s much more direct to do that by working with humans. That said, a lot of what we know about human genetics and pharmacology has been gained through animal research and it’s a very useful way to identify molecular mechanisms of effects seen in humans. What is the focus of your current research? About 20 years ago, I began looking for an easily measured genetic moderator that can be used to identify people who are likely to respond well to a medication to reduce heavy drinking. This ultimately led to a study of topiramate, which we completed and published, and which enabled us to obtain funding to do a replication study. That study, which were doing at Penn, should be finished in the next year. To date, we have randomized nearly 150 patients to treatment with topiramate or placebo based on their genotype to see whether we can validate our initial findings. I am also conducting two other pharmacogenetic trials, one to treat alcohol use disorder and the other smoking cessation in pregnant women. Finally, I am conducting multiple studies of the genetics of alcohol and drug use disorders. What advice would you have for undergraduates looking to get into the field of research? Start early. Give yourself time to find out what you really like because you have to have a passion for research to enable you to stick with it. It requires a lot of commitment and there will be disappointments along the way, but if it excites you and satisfies a burning curiosity, it’s a wonderful career option. To sustain a research career, you have to get research funding, which is difficult, and if you don’t have a “fire in the belly,” as they say, you’re going to lose interest. It’s best to find out early whether research excites you. If you love it, it’s worth the effort. That said, however, what you find exciting early in your college career may not excite you later, so it’s hard to predict where your interests may take you. My bachelor’s degree waas in anthropology, and I thought I was going to do anthropological field work, but I ended up doing much more biologically-oriented research. The bottom line, though, is that you should pursue your passions and see where they lead you.
Features
W
Dr. reagan wetherill hat is your area of research?
My research focuses on identifying the neural features that predict and/or maintain substance use or relapse, as well as the consequences of alcohol and substance use on the brain. This research involves integrating several research approaches, including genetics, neuroimaging, neuropsychological assessments, and clinical trials. What does your research suggest about the differences between addiction in men and women? Men and women are different, not surprisingly. Males typically initiate substance use at an earlier age, but women escalate use more quickly. Females are also at greater risk of negative
Interview by Asha Dahiya
Dr. Reagan Wetherill is an Assistant Professor and Licensed Clinical Psychologist at the Center for Studies of Addiction in the Department of Psychiatry at the Perelman School of Medicine at the University of Pennsylvania. Dr. Wetherill’s training and research experience encompass both basic and clinical neuroscience, and her program of research integrates genetic and neuroimaging approaches to help elucidate the etiology of alcohol use disorder, cannabis use disorder, and nicotine use disorder and to optimize pharmacological treatments.
health-related consequences. For example, female smokers are at greater risk of developing lung cancer, death from chronic obstructive pulmonary disease, and so on. There are hormones at play there, particularly estrogen. Substance use is harmful to everyone, but for women, the effects of use seem to be more severe.
From a biological perspective, how does substance addiction affect the brain?
From the perspective of a neuroscientist, there are stages of addiction with brain processes that underlie each stage. The limbic area, or reward network, is the key network underlying addictive behaviors, and as such, plays a role in all stages of addiction. The first stage is often described as the binge/intoxication stage. Another side of addiction is when the drug is not present in the brain and we see a heightened state of negative affect and stress. Over time, the changes that go on in the brain because of repeated use leads to tolerance and crav-
fall 2018 | PENNSCIENCE JOURNAL
31
Features
ing. These stages culminate into preoccupation when the person spends a great deal of time thinking about, obtaining, using and recovering from using. You have conducted research on cannabis, nicotine and cocaine. What are the differences between how addiction develops for each of these substances? Each drug works on a different receptor system, but all of them affect dopamine, which is the primary neurotransmitter for the limbic system. Those that directly hit the dopamine system, such as cocaine, have a much quicker and more powerful effect than, say, marijuana, which affects the endocannabinoid system. These drugs still work along the same neural networks, but through different processes. What kind of pharmacological treatments exist to treat substance addiction, and how do these treatments work? Essentially, there are agonists and antagonists that can either mimic or block the drug from causing their usual effects. What we do here [at the Treatment Research Center] is try to find the best fit for the individual. One of the studies Dr. Kranzler, the Center Director, and I are working on now is a prospective examination of whether a genetic polymorphism in a specific SNP [single nucleotide polymorphism] moderates treatment response to Topimarate (which is not yet approved for alcohol use disorder), but has been shown to reduce heavy drinking days in individuals with alcohol use disorder. Foreseeable, we could one day have a person come in, screen them, and then match them to the medication that works best for that individual. In general, research suggests that treatment response to pharmacotherapy for alcohol and substance use disorders is modest, which pushes the drive towards precision
32
PENNSCIENCE JOURNAL | fall 2018
medicine. You have conducted research on drinking among college-aged students and on the link between drinking, sexual activity and perceived risk. What kind of drinking patterns are present in this demographic, and what is their social and behavioral impact? That’s a big area of research that I’ve spent quite a lot of time in. Essentially, the type of drinking behaviors that we see quite often among college-age individuals is the binge-drinking or heavy episodic drinking pattern; drinking large amounts of alcohol over short periods. An area of research I have worked on is alcohol-induced blackouts. That is, memory loss for all or part of a drinking episode. Alcohol-induce blackouts have been reported by 40-50% of college students who consume alcohol. Do you find that these patterns are indicative of addiction later in life? It’s an interesting thought. Research suggests that people mature out; that is that once a person leaves the college environment and takes on more responsibilities, the lifestyle of drinking doesn’t fit well with managing and maintaining adult roles, such as having a career, spouse, or children, so many individuals stop engaging in binge-drinking behavior. However, there is a subset that continues on and maintains the high level of alcohol consumption regardless of life changes. Several factors likely contribute to maintaining that type of behavior. In your opinion, what are the most critical scientific questions surrounding substance addiction that have yet to be answered? One of the big topics that is underway right now is identifying the brain structures and behaviors that influence neurodevelopment and contribute to alcohol and substance use before a person initiates use. It could and will hopefully provide information that we need to improve prevention and treatment efforts. Another big area of interest would be finding a treatment that works for the masses. That is why we do what we do and try to help as many people as we can until we get there.
RESEARCH
Electrochemical Oxidation of Formate on a PdNi/C Nanoparticle Catalyst for Renewable Energy Conversion and Storage Sai Mamidala Center for Automation Technology, Drexel University, Philadelphia, PA, USA Acknowledgments: This research was conducted from June 20 to September 1, 2017 at the Center for Automation Technology at Drexel University under the guidance and mentorship of Dr. Joshua Snyder.
Abstract: Rising global temperatures are increasingly being attributed to excess CO2 in the Earth’s atmosphere, commonly generated by the use of nonrenewable energy sources such as coal and natural gas. Addressing the large disparity between peak energy demand and peak energy production by renewable sources requires the development of reliable methods/devices for the storage and conversion of the generated power. This study examines the utility of formate as an electrochemical fuel through the assessment of the activity and mechanistic progression of the half-cell electrooxidation of formate. We show that a catalyst composed of palladium-nickel alloy nanoparticles supported on high surface area carbon is both active and operationally stable during extended polarization as assessed by metrics including peak current density, rate of decay, and intrinsic activity. Pd-Ni/C exhibits clear performance superiority to more traditional materials including Pd/C and Pt/C. We also assess mechanistic shift in the reaction progression for alloy catalysts, including any changes to a rate determining step, by determining the kinetic isotope effect through the comparison of formate oxidation activity of regular and deuterated formate. The exact mechanisms of formate oxidation on the three catalysts (Pt/C, Pd/C, Pd-Ni/C) are also discussed here; the slightly different properties and compositions of each of the catalysts analyzed in this study prove to have a visible effect on the mechanisms and therefore the performance metrics of the catalysts themselves, as relevant to the formate oxidation reaction.
Intro/Background: While Earth’s current climate change crisis can be traced back to a variety of factors, the abundance of “greenhouse gases” in the atmosphere is cited as contributing significantly.18 The increase in both rate and quantity of CO2 released into the atmosphere is a confluence of multiple events and processes; the burning of fossil fuels is arguably the most detrimental. Nonrenewable energy sources account for about 85.1% of the United States’ annual energy production, and over 5.1 billion metric tons of CO2 was released by the US energy sector in 2016 alone.15 The problem is not limited to the United States; coal, a nonrenewable energy source, is responsible for 41% of the world’s electricity needs, while renewable sources account for just 21%.14, 10 This disparity underscores why research into new energy storage and conversion devices is necessary; the development of technologies to either capture atmospheric CO2 or limit its continued release into the atmosphere.20 The reaction of electrochemical CO2 reduction to small
organic compounds such as methanol and formic acid has recently been a popular area of research as it presents a possible method of renewable energy storage in the form of chemical bonds.22 Renewable sources such as solar or wind power can be used to power these reduction reactions.22 Oxidation of these reduced fuel molecules in a fuel cell device can then extract the stored energy to power devices and processes during off-peak hours. Following this rationale, a combined system if devised in the future may function as a device to both generate and store power. The formate oxidation reaction, a method of power generation, would be coupled with the reverse reaction, CO2 reduction, a potential technique for storing energy. Such a device, for which many further developments are necessary, would have a number of practical applications in addition to increasing the viability of energy production through renewable sources. For example, in a submarine or spacecraft, where hostile outside conditions prevent crews from being able to utilize a large amount of resources, the
fall 2018 | PENNSCIENCE JOURNAL
33
research
CO2 reduction reaction could be used to store energy, remove CO2 from the enclosed atmosphere, and generate O2 gas. The formate oxidation reaction could then be used in situations when emergency power generation capabilities are necessary. Over the last few decades, extensive work on formate/formic acid electrochemical oxidation has outlined the predicted mechanisms of the reaction and explored the process’s properties, advantages, and limitations on various electrocatalytic materials.5 Formic acid oxidation, which is analogous to the formate oxidation reaction, can take place through one of two distinct mechanisms. One pathway goes through an active intermediate, not only producing current but also leaving the active site cleanly and efficiently for the next molecule of reactant to be catalyzed into product (direct pathway). The second pathway goes through an alternate intermediate which tends to adsorb strongly onto the metal surface, severely blocking the number of available active sites on the catalyst (indirect pathway).23, 3 This dual-mechanism proposal is frequently discussed in the literature and generally accepted in the field of electrocatalysis.24, 25 CO was effectively identified as the likely poison intermediate in the second pathway, but finding, definitively, the active intermediate and its subsequent relevant mechanism has proved challenging due to the short lifetime of the intermediate species during the reaction.3 However, it is known that formate oxidation occurs slightly differently on different catalysts because of each catalyst’s unique properties.2 A small change in the composition of a catalyst can have large effects on the resulting maximum current density, rate of decay, etc when tested for the formate oxidation reaction, which is why a study of different metals and their behaviors as catalysts is necessary. A widely accepted and mathematically rigorous hypothesis which helps predict how certain metals may function as catalysts using thermodynamics is Density Functional Theory (DFT), which was employed to create an “activity volcano” predicting binding strengths for metal catalysts. This is especially important for the design of catalysts because it translates the complex compromising of binding strengths of key reaction intermediates (binding too strongly leads to catalyst poisoning, while binding too weakly inhibits the progress of the reaction itself) into a simple one-dimensional optimization problem with an ideal catalyst being near the top of the volcano. DFT and the activity volcano have been historically fairly accurate when making inferences about specific catalysts’ binding strengths, and we predict that our alloy catalyst, Pd-Ni/C, will yield more optimal 34
PENNSCIENCE JOURNAL | fall 2018
results than plain Pd and Pt, which are already at a high position on the volcano.7 This paper will include a thorough discussion of the Pd/C, Pt/C, and Pd-Ni/C catalysts, their properties, and overall performance metrics in various solutions, with accompanying graphs. We will also provide evidence for the difference in the mechanisms for the formate oxidation reaction on the Pd/C and Pd-Ni/C catalysts, which will increase our understanding of how the formate oxidation reaction proceeds on different surfaces. We aim for these conclusions to assist in the future design of catalysts to be used in formic acid/formate fuel cells. Experimental: The Pd/C and Pd-Ni/C nanoparticles were synthesized using a solvothermal process. Catalytic electrodes were made through drop-casting of a catalyst ink. Transmission electron microscopy and scanning TEM were performed to visually characterize the microstructure of the nanoparticles; energy-dispersive X-Ray spectroscopy was used to measure Ni and Pd fractions. Results & Discussion: (1) Material Characterization In this study, commercially-made Pt catalyst and our asmade Pd catalyst serve as respective controls to compare the relative functionality of the Pd-Ni/C catalyst. Pd-Ni/C, a catalyst carrying the properties of an alloy into the electrochemical testing, was chosen to benefit from both the inherent advantage of using Pd as the parent metal over Pt and that of another metal bonded into the lattice structure of the metal. While its potential as an effective catalyst for CO2 reduction has been recently examined, an assessment of its performance for the formate oxidation reaction was necessary.9 (2) Electrochemical Characterization Although the mechanism of the formate oxidation process is not completely certain, some probable conjectures do exist.26 Research done previously concerning formate/ formic acid oxidation points out that certain characteristics of Pd based catalysts give them higher current densities than Pt based catalysts. The high activity of Pd for heterogeneous hydrogenation reactions has been anecdotally tied to its ability to adsorb as well as absorb hydrogen.27 With a H/Pd ratio near 0.7, Pd readily forms a hydride at electrochemically relevant conditions.4 This is in contrast to Pt, which can only form surface adsorbed H and cannot form a hydride. It has been proposed that this hydride behavior partly contributes to the differences in formic acid/formate oxidation between Pd and Pt.1 The formate oxidation reaction on a Pt catalyst proceeds via a dual mechanism (indirect pathway), with one pathway utilizing an active intermediate, likely COO2-, and the other pathway allowing the formation of a strongly adsorbed intermediate species, CO, which most likely CO.4 This CO poisons active sites
of the catalyst, decreasing the electrochemically active area over time and eventually resulting in complete deactivation. It has been purported that formic acid/formate oxidation on Pd progresses through a direct mechanism, avoiding the formation of CO and resulting in faster rates and greater operational stability, lower degree of intermediate poisoning.22
for adsorbed formate and the recorded oxidation current, evidence of the effects of differing Figure 2 contains the format oxidation CVs for Pt/C and Pd/C where the currents have been normalized by the electrochemically active surface area (ECSA) of the electrodes. ECSA is calculated from the charge due to the oxidation of a surface layer of adsorbed hydrogen. In Figure 2, it is evident that Pd/C exhibits both higher oxidation currents as well as a lower potential separation between the anodic and cathodic peaks which is an indication of the superior poisoning tolerance/ avoidance of the Pd/C catalyst. The Pd-Ni/C catalyst was synthesized as a 75/25 at. % alloy of Pd and Ni, respectively. A core-shell structure is formed upon annealing in H2/Ar where the surface is passivated in several atomic layers of Pd with a high concentration of Ni in the underlying atomic layer, as has been shown for other platinum group metal alloys.9 Formation of this Pd-skinned structure results in a change in the surface lattice parameter the valence electronic structure of the surface atoms.3 In addition to their strong adsorption of CO, Pd and Pt bind the active intermediates of formic acid/ formate oxidation too strongly.26 The modification of the valence electronic structure of the surface through alloying effectively weakens the adsorption strength of both poisoning species and active intermediates, improving reaction rates for a range of electrochemical reactions.26 Nørskov et al. have used ‘d-band theory’ to model changes in surface reactivity due to alloying as well as provide a metric for predicting the activity of different materials.12 The chief principle underlying the theory is that the binding energy of an
RESEARCH
adsorbate to a (pure or alloy) metal surface is largely dependent on the electronic structure of the surface itself.11 This model uses the d-band center (εd) shift, with respect to the reference, or Fermi, level as a measure of tracking the variation of electronic structure, binding strengths, and corresponding changes in the catalytic activity.5 εd is controlled and affected by a variety of factors, including shape of the d-band, how full the d-orbital is, and the distances between the atoms of the lattice. In addition, the type of metal that Pd is alloyed with determines the degree and sign of strain that the surface metal atoms experience.1, 28 The position of εd is known to have an approximately linear correlation with the binding energy of adsorbed intermediates, chiefly those composed of hydrogen or oxygen, and consequently the catalytic activity with respect to electrochemical formate oxidation.7 Depending on whether the second metal has a larger or smaller lattice constant, which is the physical parameter of unit atoms in the lattice, than the parent, εd will shift up or down, the d-band will become narrower or broader, and the overlap between the d-orbitals of the parent and second metal decreases or increases, respectively.16 This shifting of the d-band center and charge transfer affects the corresponding binding energy of reaction intermediates, with a higher εd resulting in stronger bonding and a lower one resulting in weaker bonding of organic molecules to a metal surface.5 Pd-Ni alloys fall into the latter case: Ni, the metal being alloyed with its parent, Pd, has a smaller lattice constant than Pd (352.4 pm for Ni > 389.1 pm for Pd). Therefore, this leads to a compressive strain effect that facilitates the overlap of the d-orbital of Pd, lowering εd, which we speculate ultimately leads to an optimal adsoption/desorption balance during formate oxidation. The d-band center may also be shifted down by a phenomenon known as the ligand effect, in which, according to calculations from Density Functional Theory, the electrons from less noble metals such as Ni tend to shift to more precious metals such as Pd.5 Alloying palladium with other non-precious metals lowers εd, by both altering the electronic structure and introducing a degree of irregularity into the palladium lattice, which in turn causes the resulting alloy to bind oxygen or other small organic compounds more weakly than pure palladium would. It is established that this effect lowers the binding energy for adsorbed poison intermediates, most prevalently CO in this case. This, therefore, allows for not only a lower rate of decay, as poison intermediates will not be permanently adsorbed as easily or as quickly, but also higher current densities relating to the formate oxidation reaction because less of
fall 2018 | PENNSCIENCE JOURNAL
35
RESEARCH
the catalyst active sites will be taken up by adsorbed poison intermediate species.16 This is important because it points out the potential advantages of a metal alloy catalyst, more specifically Pd-Ni, over a pure metal catalyst. Ni was chosen here because when alloyed with Pd, it decreases the adsorption strength of intermediates, in addition to producing a homogenous alloy in the specified ratio.6 Verification of the hypothesized comparison of peak current densities can be found in Figure 3: the ECSA normalized peak current densities produced by a Pd-Ni/C catalyst are significantly higher than those produced by a pure palladium or even a pure platinum catalyst. We propose that this difference is due to the weakened interaction between the Pd-Ni/C and the active adsorbed intermediate, increasing the rate of the reaction propagation as a function of applied potential.
As previously considered, a catalyst’s rate of decay is also a useful method of characterization. The rate of decay, when normalized by the initial current for each catalyst to create an accurate standard of comparison, gives a measure of the catalyst’s operational durability and is obtained by potential holding and observing the behavior of the catalyst’s decaying activity (overall/initial speed of decay, time taken for activity to fully diminish, etc.). Figure 4 shows the current transient for the Pd/C and Pd-Ni/C during a constant potential hold at 0.3 V vs RHE. Although some poison species builds up on both the Pd/C and Pd Ni/C catalysts, it is evident that the rate of decay is much slower for Pd Ni/CSince it is clear that some poison species does build up on both the Pd/C and Pd-Ni/C catalysts, they both exhibit some degree of decay in activity over time. However, it is evident that the rate of decay is much slower 36
PENNSCIENCE JOURNAL | fall 2018
for Pd-Ni/C. We purport that this difference is due to the weakened interaction between Pd-Ni/C and intermediates including CO as well as other active, oxidized species. This, again, is reflective of the fact that the Pd-Ni/C catalyst does not bind poison intermediates as strongly as Pd/C; stronger binding of poison intermediates during the reaction would lead to quick buildup of adsorbed poison on the available active sites on the catalyst and consequently a faster initial rate of decay.
To gain more insight into any potential mechanistic shifts, including any changes to the rate determining step of the reaction mechanism as a consequence of alloying, we measured the kinetic isotope effect (KIE) for formate oxidation on Pd/C and Pd-Ni/C. We probed the relative importance of the first dehydrogenation step...Comparing the relative current densities for the oxidation of HCOO- versus DCOO-, we can probe the relative importance of the first dehydrogenation step, HCOO-/ DCOO- —> H+/D+ + COO2-. Replacing H with D deliberately introduces a slow step as D is a heavier isotope and generally reacts with slower rates than the lighter H.8 For Pd/C, a clear KIE exists where significant decrease in oxidative current is observed for DCOO- in comparison to HCOO-, Figure 5. As shown in Figure 6, however, no KIE is observed for the Pd-Ni/C electrocatalysts. The oxidative current densities on Pd-Ni/C when tested in 0.1 M KHCO3 + 0.01 M NaHCOO and 0.1 M KHCO3 + 0.01 M NaDCOO solutions are nearly identical, showing that this catalyst’s mechanism is not as drastically affected by the heavier hydrogen atom. From these observations we are able to draw some conclusions. Since the deuterated formate solution affected the behavior of the Pd/C but not the Pd-Ni/C catalyst, the formate oxidation mechanisms must differ on each catalyst such that the dehydrogenation step is the rate-limiting step on Pd/C catalyst but not Pd-Ni/C. This result gives insight into the properties of the formate oxidation reaction’s mechanism on an alloy such as Pd-Ni/C, especially when compared to pure metals such as Pd. By better understanding the progression of this mechanism, more active and stable catalysts can be designed for improved device power and efficiency as well as reduced cost. Future work will focus on further elucidation of the exact elementary reaction steps that govern formate oxidation of Pd alloy nanomaterials.
RESEARCH
step of the reaction on Pd/C but not on Pd-Ni/C, which is not significantly affected by the heavier H isotope. All these results provide important insight that may be used for the design of next generation formic acid/formate oxidation electrocatalysts for application to efficient, reversible renewable energy storage and conversion. References:
Conclusions: In this study we examined the properties of Pd-Ni/C catalyst with respect to the formate oxidation reaction. These properties were evaluated by comparing to two control catalysts, Pd/C and Pt/C. We consistently see the advantages of using a Pd-Ni/C alloy nanoparticle catalyst. In the cyclic voltammogram depicting a test in a 0.1 M KHCO3 + 0.01 M NaHCOO electrolyte, the Pd-Ni/C catalyst outperforms both Pd/C and Pt/C in peak current density. The potential hold experiment at 0.3 V makes evident the decreased rate of decay for the Pd-Ni/C catalyst in comparison to Pd/C. The reasons for these results can be traced back to the fact that Pd-Ni is an alloy, and its structure brings the properties unique to alloys, such as modified electronic structure, to the overall catalyst. These properties are proven here to have a positive impact on the formate oxidation reaction by optimizing the binding strength of key reaction intermediates while eluding the negative effects of strongly adsorbed poisons. In addition, a further experiment designed to test one part of the formate oxidation mechanism, the dehydrogenation step, revealed that it is the rate determining
1 An, L., and R. Chen. “Direct formate fuel cells: A review.” Journal of Power Sources320 (2016): 127-39. 2 Grozovski, Vitali, Víctor Climent, Enrique Herrero, and Juan M. Feliu. “Intrinsic Activity and Poisoning Rate for HCOOH Oxidation at Pt(100) and Vicinal Surfaces Containing Monoatomic (111) Steps.” ChemPhysChem10, no. 11 (2009): 1922-926. 3 Grozovski, Vitali, Francisco J. Vidal-Iglesias, Enrique Herrero, and Juan M. Feliu. “Adsorption of Formate and Its Role as Intermediate in Formic Acid Oxidation on Platinum Electrodes.” ChemPhysChem12, no. 9 (2011): 1641-644. 4 Haan, John L., and Richard I. Masel. “Recent Progress in Improving the Oxidation of Formic Acid on High Surface Area Platinum and Palladium Catalysts: Surface Alloying and pH Effects.” ECS Transactions, 2008. 5 Jiang, Kun, Han-Xuan Zhang, Shouzhong Zou, and Wen-Bin Cai. “Electrocatalysis of formic acid on palladium and platinum surfaces: from fundamental mechanisms to fuel cell applications.” Phys. Chem. Chem. Phys.16, no. 38 (2014): 20360-0376. 6 Liu, Weifeng, and Michael Pecht. IC component sockets. New York: Wiley, 2004. 7 Kibler, Ludwig A., Ahmed M. El-Aziz, Rüdiger Hoyer, and Dieter M. Kolb. “Tuning Reaction Rates by Lateral Strain in a Palladium Monolayer.” Angewandte Chemie International Edition44, no. 14 (2005): 2080-084. 8 Knowles, Rob. Kinetic Isotope Effects in Organic Chemistry. PPT. Princeton, NJ: Princeton University, September 14, 2005. 9 Kortlever, Ruud, Jing Shen, Klaas Jan P. Schouten, Federico Calle-Vallejo, and Marc T. M. Koper. “Catalysts and Reaction Pathways for the Electrochemical Reduction of Carbon Dioxide.” The Journal of Physical Chemistry Letters6, no. 20 (2015): 4073-082. 10 Organization for Economic Cooperation and Development. “Coal.” International Energy Agency. 2017. 11 Rossmeisl, Jan, Gustav S. Karlberg, Thomas Jaramillo, and Jens K. Nørskov. “Steady state oxygenreduction and cyclic voltammetry.” Faraday Discuss.140 (2009): 337-46. 12 Ruban, A., B. Hammer, P. Stoltze, H.l Skriver, and J.k Nørskov. “Surface electronic structure and reactivity of transition and noble metals1Communication presented at the First Francqui Colloquium, Brussels, 19–20 February 1996.1.” Journal of Molecular Catalysis A: Chemical 115, no. 3 (1997): 421-29. 13 Snyder, Joshua, Ian Mccue, Ken Livi, and Jonah Erlebacher. “Structure/Processing/Properties Relationships in Nanoporous Nanoparticles As Applied to Catalysis of the Cathodic Oxygen Reduction Reaction.” Journal of the American Chemical Society134, no. 20 (2012): 8633-645. 14 US Department of Energy. “How much of world energy consumption and production is from renewable energy? - FAQ - U.S. Energy Information Administration (EIA).” U.S. Energy Information Administration - EIA - Independent Statistics and Analysis. September 15, 2017. 15 US Energy Information Administration. “U.S. Energy Information Administration - EIA - Independent Statistics and Analysis.” What is U.S. electricity generation by energy source? - FAQ - U.S. Energy Information Administration (EIA). April 18, 2017. 16 Witte, Jon. “D-Band Theory - ORR Catalysis with Pt-based CSNPs.” ORR Catalysis with Pt-based CSNPs. December 13, 2011. 17 Leahy, Stephen. “Hidden Costs of Climate Change Running Hundreds of Billions a Year.” National Geographic. September 27, 2017. 18 “Climate change causes: A blanket around the Earth.” NASA. August 10, 2017. 19 Harvey, Chelsea. “Scientists Can Now Blame Individual Natural Disasters on Climate Change.” Scientific American. January 02, 2018. 20 London Battersea Power Station. “Promising technologies to reduce power plant emissions - Horizon 2020 - European Commission.” Horizon 2020: The EU Framework Programme for Research and Innovation. June 01, 2017. 21 “Greenhouse Gases Explained: Where Greenhouse Gases Come From.” US Energy Information Administration. July 7, 2017. 22 Kortlever, Ruud, Collin Balemans, Youngkook Kwon, and Marc T.m. Koper. “Electrochemical CO2 reduction to formic acid on a Pd-based formic acid oxidation catalyst.” Catalysis Today244 (2015): 58-62. 23 Markovic, N. “Structural-effects in electrocatalysis Oxygen reduction on the gold single crystal electrodes with (110) and (111) orientations.” Journal of Electroanalytical Chemistry 165, no. 1-2 (1984): 121-33. 24 Climent, Vı́ctor, Enrique Herrero, and Juan M. Feliu. “Electrocatalysis of formic acid and CO oxidation on antimony-modified Pt(111) electrodes.” Electrochimica Acta 44, no. 8-9 (1998): 1403-414. 25 Vidal-Iglesias, F. J., J. Solla-Gullón, E. Herrero, A. Aldaz, and J. M. Feliu. “Formic acid oxidation on Pd-modified Pt(100) and Pt(111) electrodes: A DEMS study.” Journal of Applied Electrochemistry 36, no. 11 (2006): 1207-214. 26 Min, Xiaoquan, and Matthew W. Kanan. “Pd-Catalyzed Electrohydrogenation of Carbon Dioxide to Formate: High Mass Activity at Low Overpotential and Identification of the Deactivation Pathway.” Journal of the American Chemical Society 137, no. 14 (2015): 4701-708. 27 Nag, Nabin K. “A Study on the Formation of Palladium Hydride in a Carbon-Supported Palladium Catalyst.” The Journal of Physical Chemistry B 105, no. 25 (2001): 5945-949. 28 Xin, Hongliang, Aleksandra Vojvodic, Johannes Voss, Jens K. Nørskov, and Frank Abild-Pedersen. “Effects of d-band shape on the surface reactivity of transition-metal alloys.” Physical Review B 89, no. 11 (2014).
fall 2018 | PENNSCIENCE JOURNAL
37
research
Modifying Reaction Diffusion: A Numerical Model for Turing Morphogenesis, Ben-Jacob Patterns, and Cancer Growth Kai Trepka Department of Chemistry and Chemical Biology, Harvard University, Cambridge, MA, USA
Abstract: Modeling the development of organisms and diseases has been of interest for decades. Often, complex systems of development or signaling pathways can be explained and modeled to a high degree of accuracy with only a few simplifying assumptions. Complex systems such as pattern development, bacterial growth, and tumor formation can be modeled numerically using a reaction diffusion model with relatively few factors and still give accurate results, allowing exploration of equilibrium and non-equilibrium solutions. Here, applications of such an approach to a few model problems are presented – morphogenesis, bacterial growth, and cancer treatment.
Mathematical Background and Model: The Diffusion Equation Given a collection of small particles (for example bacterial nutrients, or oxygen in water) and a heterogeneous concentration profile, over time the particles will be pushed by random thermal fluctuations into a more homogeneous (i.e. uniform) profile. This gives us a flux, j, that scales with the diffusion constant, D, and the concentration gradient.
System Validation: Gaussian Fitting As a basis for this model, I wrote a 2D Laplacian operator in Matlab and verified that the concentration profile of a diffusing point source fits to the appropriate Gaussian (as it should – we see Green’s function gives a Gaussian concentration profile.) The concentration profile after 1000 timesteps is shown in Figure 1.
This tells us that the flux of particles is higher near “pointier” places in the concentration profile (causing us to expect fractal-like growth in certain systems). Plugging this into Fick’s law, we obtain the final diffusion equation of a collection of particle concentrations C over time, i.e. Figure 1: System Validation
38
PENNSCIENCE JOURNAL | fall 2018
Animal Morphology: Model: Small Fluctuations around Equilibrium How do animals get their shapes? At a biochemical level, various complicated signaling pathways are involved (Fgf, Bmp, Hedgehog, and Notch, for example [1]). However, Alan Turing proposed that complex pattern formation can be described by a reaction diffusion system with at least two factors, with minor random or directed fluctuations from a homogeneous equilibrium resulting in a very heterogeneous final distribution, for example leopard spots and zebra stripes [2]. Turing solved this system by linearizing around the steady state, which yields interesting patterns in and of itself, but has important limitations in that we can’t know the behavior at long timeframes. To model this numerically, I adopted the following system using two dummy morphogens (factors that affect animal growth) A and B:
RESEARCH
I then investigated the behavior by varying a-d, as well as the locations of pairs relative to each other (i.e. are the points in the pairs right next to each other, one apart from each other, or in totally random locations entirely). Results and Conclusions Varying the values a-d results in changes in the density of heterogeneous defects/lines in the steady state – higher (absolute) values of coefficients result in more dense patterns (see Figure 2). This means the difference between an animal with very complicated patterns everywhere and one with larger spots/ defects may be varying reaction rates based on local concentrations.
Suppose the diffusion coefficients of A and B are 0.5 and 4.5, respectively, in arbitrary units. Also suppose that the change in concentrations over time are defined by
Initialize a uniform system at a steady state (i.e. with no perturbations, there will be no changes). Consider the case Define the numerical value of a small fluctuation (this represents random perturbations from the environment), Randomly generate pairs of fluctuations (these represent elements being thermally bounced “back and forth”,) where the first element in the pair has concentrations A1, B1, and the second has concentrations A2, B2, where
Figure 2: Steady state patterns of systems with 1000 randomly located point pairs. In the left image, a = 100000, b = -100001, c = 100001, d = -1000002. In the right image a = 5, b = -6, c = 6, and d = 7. Pair location has little effect – if each point is randomly located (i.e. each point is not next to its corresponding point, A1 defect is not next to the A2 defect), it produces the same patterns as if each point in the pair is one apart (i.e. A1-normal-A2 in a row). However, if each pair is right next to each other, i.e. the point with the A1 defect is directly adjacent to the point with the A2 defect, the changes, the defects end up canceling each other out in the long-run, resulting in homogeneity. This is just one example of the failure of the linearization approximation (i.e. long-term stability even with predicted short-term non-equilibrium perturbations). Turing’s approximation is useful for some patterns, but over time some smaller aspects of these fall 2018 | PENNSCIENCE JOURNAL
39
research
patterns disappear – this is why it is useful to numerically model the full timescale of pattern formation so we know the behavior at equilibrium, rather than just the behavior after short-term perturbations (see Figure 3).
Figure 3: Short-term (linear) approximation (left) vs long-term steady-state (right) Future Directions We have shown it is possible to numerically model Turing pattern formation, the effects of tweaking parameters such as number of defect and governing constants, and potential developmental relevance. To improve on this model and understand arbitrary patterns, it would be useful to consider increased numbers of morphogens (rather than just using Turing’s simplified system with two morphogens linearized around the steady state), as well as considering more complicated reactions. Modeling Bacteria in Non-Equilibrium Growth Model: Adding Growth While reaction diffusion on its own leads to interesting patterns and conclusions, it is instructive to incorporate growth into the model, given that cells do not exist in a static environment and are constantly considering whether to synthesize a new copy of their DNA and divide of the conditions are right. Ben-Jacob observed interesting growth regimes when growing bacteria on agar with different peptone concentrations (see Figure 4) [3,4]. Our goal was to see if we could simulate and obtain the same patterns 40
PENNSCIENCE JOURNAL | fall 2018
mathematically, using reasonable physical constants.
Figure 4: Bacterial Patterns from [6] (Empirical). Peptone concentrations of 0.1, 1, 3, and 10 g/l, respectively (a-d). To model bacterial growth, we assumed each pixel represents one bacterium – a bacterial radius is about 5 um, so a pixel is about 10x10 um, and each pixel is one bacteria. Each bacterium has a certain uptake rate of nutrients, and will divide when the nutrient concentration inside exceeds a certain threshold, and each resulting bacterium will have half the nutrients of the initial bacteria. In this model, there is no diffusion of bacteria themselves, and no death if a nutrient threshold is passed. In addition, recall that <Δx^2> =2nDΔt For reasonable physical constants, we have D = 10-7 cm2/s, and delta x is 10 microns, so the timestep we use is dt = 5 seconds. Additionally, let the amount of nutrient a bacterium needs before division be 3*1012 g/bacterium (which is per pixel). Finally, we run through a range of nutrient (peptone) concentrations from 10-6 to 2*10-6 g/cm2. Results and Conclusions
Figure 5: Bacterial Patterns (modeled). Peptone concentrations of 10-6, 1.25*10-6, 1.5*10-6, and 2*10-6 g/cm2, respectively (a-d). Using even this simple model of growth, we can roughly reproduce the four growth regimes observed by Ben-Jacob’s groups – at low peptone concentration, there is a circle in the center, and some small
radial branching. At intermediate peptone concentration, there is radial branching and fractal-like growth on these branches. At high peptone concentration, there is radial “finger” growth, and at very high peptone concentrations, there is just a bacterial blob. In other words, we have shown numerically that bacterial growth is diffusion limited and can be modeled as such, in very close agreement with reported experimental data. One interesting thing to note is these fractal-like patterns emerge because the flux through points is locally higher than the flux through a large surface, resulting in increased growth and branching from branches that already exist rather than from the center. The utility of this result is we can determine what patterns we expect based on certain nutrient concentrations and rates of growth, compare it to empirical patterns, and use this to determine the accuracy of our constants. Additionally, this has applications to other fractal-like processes, such as crystal growth via physical vapor deposition at high temperatures (flux is highest through points). Future Directions To improve the model and give even more accurate patterns, we should consider the doubling time of bacteria (which is 1500 seconds for all 4 stages of the cell cycle, rather than the 5 second timesteps we used), account for bacterial motility (for example swarming or the “run and tumble” model), and add a parameter for cell death if nutrient concentration around the bacteria is too low for cell maintenance. Reaction Diffusion and Cancer Model: Incorporating Multiple Factors One particularly interesting application of reaction diffusion models is diffusion-limited cancer growth [5]. For example, the following tumor morphologies from [5] can be roughly modeled using a reaction diffusion model. Cancer cells are often mutated such that growth factor networks are out of control – they require less growth factor than normal cells to divide, and can often produce more, resulting in a positive feedback loop of growth and risks metastasis [6,7]. Cancer cells also kill surrounding tissue, either by outcompeting it for nutrients or secreting a “death factor” that kill surrounding WT cells [8].
RESEARCH
We will model cancer cells by considering a system in which there are a network of blood vessels that constantly replenish nutrients (especially since the speed of blood, 0.1 m/s, is much faster than the speed of oxygen diffusion through tissue, which is under 10-4 m/s, so we can say the vessel locations constantly replenish all nutrients in each timestep.) Nutrients diffuse from the blood vessel, cells grow if above a certain nutrient threshold and no surrounding cells (if WT) or even through surrounding cells (if a cancer cell), cells die below a certain nutrient threshold, cells produce and consume growth factors at certain rates, and growth factors and nutrients diffuse according to the diffusion equation. Cancer cells have a lower threshold of growth factor to divide in all of our models, as found in [6,7]; however, we also set the model so that cancer cells produce far less growth factor than normal cells, effectively depending on normal cells for their own division. We then investigate a variety of starting conditions (possibility of detachment from basal lamina, effect of different nutrient demands and uptake from cancer cells, effect of cancer cells making more growth factors, and growth along different blood vessel locations.
Figure 6: Simulation Results - Steady State Profiles of Cancer Cells Results and Conclusions Above are the images of cancer cell steady state locations for different initial parameters. Note that in each of them, there were 50 randomly generated cancer cells at different points, and certain ones survived and then grew into the above morphologies. For figures 9a-9d, there are 7 parallel vertical blood vessels; for 9e, there is a grid of such blood vessels instead. For most simulations, cancer cells can diffuse across the nutrient source; for 9a, they cannot (prefall 2018 | PENNSCIENCE JOURNAL
41
research
sumes basal lamina provides some protection against metastasis). From this, we can see that diffusion can limit tumor growth – these models reached a steady state past which the cancer cells could no longer grow. This has relevance for therapies that attempt to prevent formation of blood vessels in tumors in order to starve them of nutrients. Furthermore, differences in rates of growth factor production have a huge influence on whether a tumor stays benign or undergoes a feedback expansion. Future Directions To improve this model, more realistic rather than arbitrary diffusion and reaction coefficients could be used, as well as a 3D generalization, and considering different initial locations of blood vessels and cancer cells. Further questions to be investigated via similar means include comparing a nutrient competition vs death factor secretion model, as well as modeling drugs that inhibit growth factors as a method to treat tumors and what expected morphology changes are. Conclusions and Future Directions: Conclusions Often, complex systems of development or signaling pathways can be explained and modeled to a high degree of accuracy with only a few simplifying assumptions – we don’t need to understand every gene expressed in every pathway to get a good idea of what will happen. Complex systems such as pattern development, bacterial growth, and tumor formation can be modeled numerically and with relatively few factors and still give interesting and roughly accurate results, which obviates the need for a lot of complicated math and lets us find the eventual t=∞ solution. Although not very optimized as presented here, a reaction-diffusion-growth model of cancer is particularly interest. If you can set up an accurate system (i.e. appropriate diffusion coefficients, rates of growth factor production/consumption, blood vessel loca42
PENNSCIENCE JOURNAL | fall 2018
tions), you can predict how you expect cancer to grow normally. Further, you can predict how certain diffusion-limited treatments (chemotherapy) should affect tumor morphology, and compare predicted results to MRI to see if we understand drug mechanism. Outside of biology, this work has interesting applications in determining non-equilibrium or interesting steady-state solutions to physical systems such as fluid flow and the heat equation. Future Model Improvements To improve on the work presented, the model could be expanded to 3D (which is not very challenging conceptually – it’s just setting up a different Laplacian and tracking across a 4D rather than 3D array, but is annoying with edge cases and visualization), and compared to the continuous case (where we use Green’s function convolved with the distribution at each step, rather than just discretizing the Laplacian). Further, models of cells should consider cell motility, as well as allow for arbitrary factors and cell types. The challenge is to determine the appropriate constants to use, as well as which factors are relevant – for any given pathway, there could be dozens of genes involved, but the morphology of the result can be predicted with far fewer components in the model (as we showed in the case of bacterial growth). Acknowledgements Kai Trepka would like to acknowledge Dr. Adam Cohen for his assistance in the planning and execution of this model during Chem 163 at Harvard. References
[1] Cooper GM. The Cell: A Molecular Approach. 2nd edition. Sunderland (MA): Sinauer Associates; 2000. Signaling in Development and Differentiation. [2] Turing, A. M. (1952). The chemical basis of morphogensis, Phil. Trans. Roy. Soc. B, 237, 37–72. [3] Ben-Jacob, Eshel, and Peter Garik. “The formation of patterns in non-equilibrium growth.” Nature 343.6258 (1990): 523-530. [4] Golding, Ido, et al. “Studies of bacterial branching growth using reaction–diffusion models for colonial development.” Physica A: Statistical Mechanics and its Applications 260.3 (1998): 510-554. [5] Ferreira Jr, S. C., M. L. Martins, and M. J. Vilela. “Reaction-diffusion model for the growth of avascular tumor.” Physical Review E 65.2 (2002): 021907. [6] Jechlinger, Martin, et al. “Autocrine PDGFR signaling promotes mammary cancer metastasis.” The Journal of clinical investigation 116.6 (2006): 1561. [7] Demoulin, Jean-Baptiste, and Ahmed Essaghir. “PDGF receptor signaling networks in normal and cancer cells.” Cytokine & growth factor reviews 25.3 (2014): 273-283. [8] Suijkerbuijk, Saskia JE, et al. “Cell competition drives the growth of intestinal adenomas in Drosophila.” Current Biology 26.4 (2016): 428-438.
www.pennscience.org Penn Science is sponsored by the Science and Technology Wing at the University of Pennsylvania