Download PDF Systems biology methods in molecular biology 2745 mariano bizzarri full chapter pdf
Systems Biology Methods in Molecular Biology 2745 Mariano Bizzarri
Visit to download the full and correct content document: https://textbookfull.com/product/systems-biology-methods-in-molecular-biology-2745mariano-bizzarri/
More products digital (pdf, epub, mobi) instant download maybe you interests ...
Systems Biology 1st Edition Mariano Bizzarri (Eds.)
University of Hertfordshire Hatfield, Hertfordshire, UK
For further volumes: http://www.springer.com/series/7651
For over 35 years, biological scientists have come to rely on the research protocols and methodologies in the critically acclaimed Methods in Molecular Biology series. The series was the first to introduce the step-by-step protocols approach that has become the standard in all biomedical protocol publishing. Each protocol is provided in readily-reproducible step-bystep fashion, opening with an introductory overview, a list of the materials and reagents needed to complete the experiment, and followed by a detailed procedure that is supported with a helpful notes section offering tips and tricks of the trade as well as troubleshooting advice. These hallmark features were introduced by series editor Dr. John Walker and constitute the key ingredient in each and every volume of the Methods in Molecular Biology series. Tested and trusted, comprehensive and reliable, all protocols from the series are indexed in PubMed.
SystemsBiology
Second Edition
Edited by
Mariano Bizzarri
Department of Experimental Medicine, Sapienza University, Rome, Italy
Editor Mariano Bizzarri
Department of Experimental Medicine
Sapienza University
Rome, Italy
ISSN 1064-3745ISSN 1940-6029 (electronic)
Methods in Molecular Biology
ISBN 978-1-0716-3576-6ISBN 978-1-0716-3577-3 (eBook)
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This Humana imprint is published by the registered company Springer Science+Business Media, LLC, part of Springer Nature.
The registered company address is: 1 New York Plaza, New York, NY 10004, U.S.A.
Paper in this product is recyclable.
Preface: Challenges and Methodological Advances in Systems Biology
The Meaning of “Systems Biology”
A few set of definitions strive to capture the more appropriate meaning of “Systems Biology.” Systems Biology “is the computational and mathematical analysis and modeling of complex biological systems. It is a biology-based interdisciplinary field of study that focuses on complex interactions within biological systems, using a holistic approach (holism instead of the more traditional reductionism) to biological research” [1]. However, the conceptualization of Systems Biology is still open. Currently, two primary streams can be recognized within Systems Biology: (1) pragmatic Systems Biology, which emphasizes the use of large-scale molecular interactions (“omic” approach), aimed at building huge signaling networks by applying mathematical modeling and thus showing how cells make decisions based on the “information” flowing through their networks. (2) Theoretic Systems Biology which posits that the theoretical (and consequently the methodological) basis of biological study should be deeply modified in antithesis to the reductionist paradigm adopted by molecular biology [2]. According to Sauer et al. [3], “the reductionist approach has successfully identified most of the components and many of the interactions but, unfortunately, offers no convincing concepts or methods to understand how system properties emerge [given that] the pluralism of causes and effects in biological networks is better addressed by observing, through quantitative measures, multiple components simultaneously and by rigorous data integration with mathematical models.”
Therefore, Systems Biology addresses those operational procedures able to integrate (rather than to reduce) different aspects of the intertwined molecular and biophysical factors that cooperate in accomplishing a biological function. This approach strives to substitute single “molecular player” with significant correlations arising among different “actors,” distributed across a multi-level organization. The “causative biological factor” arises as a consequence of collective (instead of “single”) changes in reciprocal relationships, involving a complex network of internal/external controls. The objective is finally to establish a dynamical model capable in capturing the intrinsic complexity of the system, which is frequently submitted to the rules on nonequilibrium thermodynamic [4]. Since the objective is to recapitulate a model of the dynamic interactions, the experimental techniques –including transcriptomics, metabolomics, proteomics, and high-throughput technologies –that most suit systems biology are those that are system-wide, able in recording an impressive body of raw data, which are further analyzed through a specific bioinformatic strategy.
The present volume – an updated re-edition of a previous book published in 2018 –precisely focuses on methodological aspects of Systems Biology in order to provide this new theoretical approach with a robust and tailored experimental support.
As Editor, I have collected an eclectic cluster of articles. This is not a “one view fits all” approach. It is rather one to “let a hundred flowers bloom,” specifically aimed to identify key methodological issues that actually challenge the reliability of “true” Systems Biology studies, namely addressing the following topics: (1) models in Systems Biology and parameters identification; (2) computational methods; (3) critical transition states; (4) specific fields of investigation.
Models in Systems Biology and Parameters Identification
A very basic challenge in performing systems biology studies deals with the appropriate choice of models – wherein both top-and bottom-up approaches are adopted – and critical parameters – not only restricted to molecular factors – “selected” among the huge number of molecules that can actually be recorded with high-throughput techniques.
The first chapters discuss these complex issues by addressing the methodological challenges arising in specific areas of research.
In The Search for System’s Parameters, A. Giuliani argues that the integration of physics and biology is the key for copying with the “deluge” of information collected about an overwhelming body of molecular components that is becoming a real threat for true knowledge advancement. A solution based on new strategies to perform the network formalization of molecular knowledge is proposed, suggesting a perspective based on the association of both flux and dynamic modelization. Some theoretical and applicative cases are debated, highlighting how the two models could help in grasping subtle and hidden differences underlying the observed biological process. Ideally, Kumar Selvarajoo further reinforces this approach in the chapter Complexity of Biochemical and Genetic Responses Reduced Using Simple Theoretical Models. The chapter outlines the bewildering complexity of metabolic data provided usually by metabolomic studies. To extract useful information a combination of both linear and nonlinear ordinary differential equations is mandatory in order to “reconstruct” an instructive, dynamic model. Time-series data of living cells are included into the modelization to recapitulate the overall process, especially by focusing on critical transition points. This approach undergoes an extensive analysis, after having been vindicated through appropriate experimental settings.
However, the appropriate theoretical framework to explain the complexity observed in some physiological phenomena must include the thermodynamic formalism of nonequilibrium systems together with the theory of complex systems, as highlighted in the chapter by Nieto-Villar et al. – Metastasis Models: Thermodynamics and Complexity – dealing with the intricate matter of distant colonization of cancerous cells. Noticeably, this approach allows the establishment of an appropriate conceptual and operational framework to investigate emergent phenomena that cannot be easily grasped by reductionist-based models. Difficulties in managing this kind of problems arise from the fact that we are facing nonlinear systems, temporally and spatially self-organized out of thermodynamic equilibrium. Different mathematical models and techniques are used to investigate the evolution of cancerous process across intertwined states depicted in a Waddington-like landscape. Occurrence of specific conditions – including epithelial-mesenchymal transition and response to chronotherapy – are discussed as exemplary situations.
A ver y critical challenge is represented by the analysis of metabolomic data, as aptly discussed by Syarul Nataqain Baharum in the chapter A Systems Biology Approach in Metabolomics. Metabolomics can provide diagnostic, prognostic, and therapeutic biomarker profiles of individual patients because a large number of metabolites can be simultaneously measured in biological samples in an unbiased manner. As expected when nonlinear processes occur in self-organizing systems, even minor changes/stimuli in the initial conditions can result in substantial alterations of the overall process. This property makes metabolomics a very promising field in assessing the reliability of the system biology approach in investigating complex processes. However, this is not a simple task. Due to the complexity and sensitivity of the metabolome, studies must be devised to maintain consistency, minimize
subject-to-subjectvariation, and maximize information recovery. Technological advances in experimentaldesign,animal models, and instrumentation have recently aided this effort. Protonnuclearmagnetic resonance (1H-NMR) spectroscopy of biofluids – including plasma,urine,andfeces – provides by now the opportunity to identify biomarker change patternsthatreflectthephysiological or pathological status of an individual patient. Metabolomicshastheultimate potential to be useful in a clinical context, where it could be used topredicttreatmentresponse and survival and for early disease diagnosis. Moreover, during drugtreatment,anindividual’s metabolic status can be monitored, and used to predict deleteriouseffects.
A critical improvement in metabolomic studies has been provided by the ever-increasing use of mass spectrometry, as outlined in the chapter by Anna Laura Capriotti and Aldo Lagana ` – Mass Spectrometry in Systems Biology. Namely, the possibility to study a class of underrated small molecules – which support relevant functions albeit being represented in very small amounts in biological fluids – has gained momentum in recent years. Particularly, short-chain peptides have attracted increasing attention in different research fields, including biomarker discovery. The analysis of small peptides – from 2 to 10 amino acid residues –constitutes an analytical challenge in complex matrices due to their low abundance compared to other molecules, which can cause extensive ion suppression during massspectrometric acquisition. Furthermore, there is a lack of analytical workflows for their comprehensive characterization, since ordinary peptidomics strategies cannot identify them. In this context, an enrichment strategy is absolutely required to isolate and clean up short-chain peptides by graphitized carbon black solid phase extraction. The highresolution mass spectrometry protocol discussed in the chapter allowed the detection of the eluting peptides by data-dependent mode using a suspect screening strategy with an inclusion list. To ensure a better coverage of peptide polarity, when urine samples are under scrutiny, the analytical strategy should associate ultrahigh performance liquid chromatography by reversed-phase and hydrophilic interaction liquid chromatography.
Systems biology cannot avoid associating semi-quantitative data obtained by fluorescence and immunofluorescence techniques performed on either histological samples or 3D constructs. In the chapter authored by Michael Levin – Bioelectric Fields in Living Organisms: A Systems Biology Approach – fluorescent lifetime imaging (FLIM) is described as a powerful tool for visualizing physiological parameters in vivo. This technique is especially useful for mapping bioelectric patterns in living organisms (such as Xenopus laevis embryos). The chapter discuss how to disentangle physiological artifacts from true bioelectric signals, a method for dye delivery via transcardial injection, and how to visualize and interpret the fluorescent lifetime of the dyes in vivo. Such data are then associated to molecular analysis to recapitulate an integrated model.
Computational Methods
Mathematical modeling is irreplaceable in understanding and modeling natural phenomena, as extensively discussed by Mael Monte ´ vil in the chapter A Primer on Mathematical Modeling in the Study of Organisms and Their Parts. Such a tool carries its own assumptions and should always be used in a critical manner. The key observables and steps of modeling are presented altogether with their basic biological interpretation. Namely, the relevance of theoretical principles as the very basic bricks underpinning different methodological approaches is investigated, especially when conceptual tools are used in the “construction”
ofplausiblemodels.This discussion entails the use (and the interpretation) of equations, and theoverallchapteraimsat facilitating the interaction between biologists and mathematical modelers.
Cancer is specific case of mathematical modeling in biology, due to the intrinsic nonlinear dynamics of many critical processes. Abdallah Alameddine extensively addresses these hurdles in the chapter Systems Biology Modeling of Cancer Nonlinear Dynamics. The chapter explains how to connect temporal measurements of a nonlinear dynamical and unstable complex system, such as cancer, with well-established engineering methods, which are usually applied in linear dynamical systems. This proof of concept deserves to be included in clinical studies given its relevance in the development of newer treatment options, by either adding an appropriate external “damping” or a “forcing” term, or by a “control” actuator such that its nonlinear dynamic is steered to a spiral stably into zero forever as a sink attractor. The chapter address a question too often underestimated, i.e., how to influence in an efficient way the “attractor” stability in order to trigger a “therapeutic” effect, by looking at cancer as a complex system rather than a “sum” of single molecular reactions.
Critical Transition States
Orchestrated changes that promote an overall phenotypic modification are usually investigated by adopting the framework of critical transition state, borrowed from statistical mechanics [5]. The chapter by Masa Tsuchiya et al., – From Cell States to Cell Fates: Control of Cell State Transitions – specifically discusses the emergence of transition during cell fate commitment. The chapter suggests that the coordinated behavior of thousands of genes in cell fate transitions can be addressed through genome expression as an integrated dynamical system, by using concepts borrowed from self-organized criticality and coherent stochastic behavior. To quantify effects of collective behavior of genes, the authors adopted a flux balance approach to obtain a new conceptual/operational tool (termed expression flux analysis (EFA)). EFA is instrumental in “decoding” specific experimental genome-wide expression data, providing unexpected insights into the dynamics of the cell-fate transitions. In particular, the authors show that in cell fate change, specific stochastic perturbations can spread over the entire system to guide distinct transitions through switching cyclic flux flow in the genome engine. Utilization of EFA enables in discovering a unified genomic mechanism for when and how cell-fate change occurs through critical transitions.
Along this field, a further contribution comes from Andras Paldi – A Systems Biology Approach in Investigating/Understanding Cell Differentiating Processes – highlighting the complexity behind a specific kind of transition processes, such as those involved during differentiation. The cells of a multicellular organism are derived from a single zygote and genetically almost identical. Yet, they are phenotypically very different. This difference is the result of a process commonly known as cell differentiation. How the phenotypic diversity emerges during ontogenesis or regeneration is a central and intensely studied but still unresolved issue in biology. Here, cell biology is facing a true conceptual challenge that is frequently confused with “methodological” difficulties. How to define a cell type? What stability or change means in the context of cell differentiation and how to deal with the ubiquitous molecular variations seen in the living cells? What are the driving forces of the change? The authors propose to start with a “conceptual revolution,” in order to reframe the problem of cell differentiation in a systemic way by incorporating different theoretical approaches. This new conceptual framework is able to capture the insights made at different
levelsofcellularorganization, and – what is even more important – to explain and accommodatewithanumberof experimental finding previously considered at odds with the dominantexplanatoryparadigm. Moreover, the strategy depicted by Paldi helps in establishingaformalstrategyforfurther experimental studies.
Epithelial-mesenchymal transition (EMT) constitutes a further case of clinical transition, especially relevant in morphogenesis and carcinogenesis studies. In the chapter authored by Noemi Monti et al. – Systems Biology Methodology for the Study of Cell Motility and Invasiveness – a Systems Biology protocol for the investigation of EMT is discussed. Epithelial-mesenchymal transition is a trans-differentiating and reversible process that leads to dramatic cell phenotypic changes, enabling epithelial cells in acquiring mesenchymal phenotypes and behaviors. EMT plays a crucial role during embryogenesis, and occurs in several para-physiologic and pathological conditions, as during fibrosis or cancer development. EMT displays some hallmarks of critical transitions, as a sudden change in the overall configuration of a system in correspondence of specific tipping point around which a “catastrophic bifurcation” happens. The transition occurs when external conditions breach specific thresholds. This definition helps in highlighting two main aspects: (1) the change involves the overall system, rather than single, discrete components; (2) cues from the microenvironment play an irreplaceable role in triggering the transition. Definitely, this evidence implies that critical transition should be ascertained focusing the investigation at the system level (rather than investigating only molecular parameters) in a well-defined context, as the transition is strictly dependent on the microenvironment in which it occurs. Therefore, we need a systems biology approach to investigate EMT across the Waddingtonlike epigenetic landscape wherein the participation of both internal and external cues is henceforth studied to follow the extent and the main characteristics of the phenotypic transition. The proposed investigative protocol suggests a set of systems parameters (motility, invasiveness), altogether with specific molecular/histological markers to identify those critical “observables,” which can be integrated into a comprehensive mechanistic model. We are all aware of the limitations of a 2D model of cell culture. The main question therefore is how to extend the Systems Biology approach to a 3D model of cell culture. This question is extensively discussed in the chapter by Carlos Sonnenschein and Ana Soto, Modeling Cancer in 3D: A Systems Biology Approach. It is well known that stromal–epithelial interactions mediate mammary gland development and the formation and progression of breast cancer. To study these interactions in vitro, 3D models are indeed essential. We have successfully developed novel 3D in vitro models that allow the formation of mammary gland structures closely resembling those found in vivo and that respond to the hormonal cues that regulate mammary gland morphogenesis and function. Due to their simplicity when compared to in vivo studies, and to their accessibility to visualization in real time, these models are well suited to conceptual and mathematical modeling.
Specific Fields of Investigation
Besides cancer and metabolism, an increasing number of different fields have attracted the interest of Systems Biology-based study by now. A promising area of investigation is particularly represented by studies performed on metabolism defects due to inborn errors. The chapter authored by Antonio Angeloni et al., Systems Biology and Inborn Error of Metabolism, provides an instructive example about how to set a Systems Biology-based study related to Phenylketonuria.
Inborn errors of metabolism (IEM) are a group of about 500 rare genetic diseases with large diversity and complexity due to number of metabolic pathways involved in. Establishing a correct diagnosis and identifying the specific clinical phenotype is consequently a difficult task. However, an inclusive diagnosis able in capturing the different clinical phenotypes is mandatory for successful treatment. However, in contrast with Garrod’s basic assumption “one-gene one-disease,” no “simple” correlation between genotype–phenotype can be vindicated in IEMs. An illustrative example of IEM is phenylketonuria (PKU), an autosomal recessive inborn error of L-phenylalanine (Phe) metabolism, ascribed to variants of the phenylalanine hydroxylase (PAH) gene encoding for the enzyme complex phenylalanine-hydroxylase. Blood values of Phe allow classifying PKU into different clinical phenotypes, albeit the participation of other genetic/biochemical pathways in the pathogenetic mechanisms remains elusive. Indeed, it has been shown that the most serious complications, such as cognitive impairment, are not only related to the gene dysfunction but also to the patient’s background and the participation of several non-genetic factors.
Therefore, a Systems Biology-based strategy is required in addressing IEM complexity and in identifying the interplay between different pathways in shaping the clinical phenotype. Such an approach should entail the concerted investigation of genomic, transcriptomics, proteomics, and metabolomics profiles altogether with phenylalanine and amino acids metabolism. Noticeably, this “omic” perspective could be instrumental in planning personalized treatment, tailored accordingly to the disease profile and prognosis.
Another specific field of research that has gained momentum in the last years is represented by microRNAs. Mahmood Tavallaei discusses such topic in the chapter MicroRNA Networks in Cancer Cells: A Systems Biology Approach.
Short RNAs called microRNAs (miRNAs) bind to and inhibit target messenger RNAs in gene regulatory networks. Recent study suggests that miRNAs circulate in a stable, cell-free form and that particular miRNAs in plasma or serum may be biomarkers for cancer and other diseases. Circulating miRNAs as biomarkers provide distinct challenges including pre-analytic variance and data standardization. These tasks are not only challenging but require an inclusive strategy based on Systems Biology, which associated a qRT-PCR approach for measuring circulating miRNAs as biomarkers, as well as sample preparation, experimental design, and data processing issues.
A ver y unusual application of Systems Biology is finally represented by studies dealing with the COVID-19 pandemic. Marko Djordjevic aptly analyzes this argument in the chapter Systems Biology Approaches in COVID-19 Studies. In fact, the COVID-19 pandemic can be regarded as a systems biology problem, with the entire world representing “the system,” and the human population as the element transitioning from one state to another with certain transition rates. While capturing all the relevant features of such a complex system is hardly possible, compartmental epidemiological models can be used as an appropriate simplification to model the system’s dynamics and infer its important characteristics, such as basic and effective reproductive numbers of the virus. These measures can later be used as response variables in feature selection methods to uncover the main factors contributing to disease transmissibility. We here demonstrate that a combination of dynamic modeling and machine learning approaches can represent a powerful tool in understanding the spread, not only of COVID-19 but also of any infectious disease of epidemiological proportions.
Contributions included in this updated version of the book Systems Biology – Springer Protocols, taught us that physiological and pathological processes are complex contextdependent entities to which our genes make a necessary but only partial contribution [6]. Systems Biology helps in establishing a new theoretical framework, based upon new conceptual tools. In turn, these conceptual premises imply a profound reconsideration of the methodological framework. We have to rethink how an experiment is planned, what kind of parameters are worth of investigation, and how their mutual relationship should be described by means of a different mathematical modeling spanning through different space and temporal scales. That task require an open-minded attitude and a true multidisciplinary approach. I mean that biologists, physicists, and mathematicians should be working together, in a true cooperative effort for establishing a new experimental method of research.
Taken as a whole, this set of articles not only challenges some of the current methodological paradigms but also lays the groundwork for alternative approaches and in many cases takes those approaches further toward the goal of understanding living systems as complex processes, governed by both local and general control factors operating at different space and temporal scales.
Rome,ItalyMarianoBizzarri
References
1. Tavassoly I, Goldfarb J, Iyengar R (2018) Systems biology primer: the basic methods and approaches. Essays Biochem 62(4):487–500. https://doi.org/10.1042/EBC20180003. PMID: 30287586
2. Bizzarri M, Palombo A, Cucina A (2013) Theoretical aspects of systems biology. Prog Biophys Mol Biol 112(1–2):33–43. https://doi.org/10.1016/j.pbiomolbio.2013.03.019. Epub 2013 Apr 3. PMID: 23562476
3. Sauer U, Heinemann M, Zamboni N (2007) Genetics. Getting closer to the whole picture. Science 316(5824):550–551. https://doi.org/10.1126/science.1142502. PMID: 17463274
4. Kondepudi DK, De Bari B, Dixon JA (2020) Dissipative structures, organisms and evolution. Entropy (Basel) 22(11):1305. https://doi.org/10.3390/e22111305. PMID: 33287069; PMCID: PMC7712552
5. Bizzarri M, Giuliani A (2022) Soft statistical mechanics for biology. Methods Mol Biol 2449:263–280. https://doi.org/10.1007/978-1-0716-2095-3_11. PMID: 35507267
6. Rosslenbroich B (2011) Outline of a concept for organismic systems biology. Semin Cancer Biol 21(3): 156–164. https://doi.org/10.1016/j.semcancer.2011.06.001. Epub 2011 Jun 28. PMID: 21729754
Preface: Challenges and Methodological Advances in Systems Biology
IMODELS IN SYSTEMS BIOLOGY AND PARAMETERS IDENTIFICATION
1 Identifying Key In Silico Knockout for Enhancement of Limonene Yield Through Dynamic Metabolic Modelling
Jasmeet Kaur Khanijou, Yan Ting Hee, and Kumar Selvarajoo
2 The Search for System’s Parameters: Statistical and Dynamical Description from Complex Network Analysis
Alessandro Giuliani
3 Untargeted Analysis of Short-Chain Peptides in Urine Samples Short Peptides Analysis .
SaraElsa Aita, Andrea Cerrato, Aldo Lagana ` , Carmela Maria Montone, Enrico Taglioni, and Anna Laura Capriotti 4 Metastasis Models: Thermodynamics and Complexity
A. Guerra, J. A. Betancourt-Mar, J. A. Llanos-Pe´rez, R. Mansilla, and J. M. Nieto-Villar
5 Metabolomics: Challenges and Opportunities in Systems Biology Studies .
Ahmed Mediani and Syarul Nataqain Baharum
6 Optical Estimation of Bioelectric Patterns in Living Embryos
Patrick McMillen and Michael Levin
PART II COMPUTATIONAL METHODS
7 Mathematical Modeling in the Study of Organisms and Their Parts
Mae¨l Monte´vil 8 Systems Biology Modeling of Cancer Nonlinear Dynamics
Abdallah Alameddine PART III CRITICAL TRANSITION STATES ACROSS WADDINGTON’S LIKE LANDSCAPES
9 From Cell States to Cell Fates: Control of Cell State Transitions
Masa Tsuchiya, Alessandro Giuliani, and Paul Brazhnik
10 Understanding Cell Differentiation Through Single-Cell Approaches: Conceptual Challenges of the Systemic Approach
Lae¨titia Racine and Andras Paldi
11 Modeling Mammary Organogenesis from Biological First Principles: A Systems Biology Approach
Cheryl M. Schaeberle, Victoria A. Bouffard, Carlos Sonnenschein, and Ana M. Soto
PART IV SPECIFIC FIELDS OF INVESTIGATION
12 Systems Biology and Inborn Error of Metabolism: Analytical Strategy in Investigating Different Biochemical/Genetic Parameters
Aurora Piombarolo, Cristiano Ialongo, Mariano Bizzarri, and Antonio Angeloni
13 System Biology Approach in Investigating Epithelial-Mesenchymal Transition (EMT)
Noemi Monti, Alessandro Querqui, Guglielmo Lentini, Marco Tafani, and Mariano Bizzarri
14 An Enhanced Quantitative Reverse Transcription-PCR Approach for Measuring Circulating MicroRNAs
Mahmoud Tavallaie and Mostafa Khafaei
15 Systems Biology Approaches to Understanding COVID-19 Spread in the Population
Sofija Markovic ´ , Igor Salom, and Marko Djordjevic
Index
Contributors
SARAELSA AITA • Dipartimento di Chimica, Universita di Roma La Sapienza, Rome, Italy
ABDALLAH ALAMEDDINE • Heart and Vascular Program, Baystate Health, University of Massachusetts Chan Medical School-Baystate, Springfield, MA, USA
ANTONIO ANGELONI • Department of Experimental Medicine, Sapienza University, Rome, Italy
SYARUL NATAQAIN BAHARUM • Institute of Systems Biology (INBIOSIS), Universiti Kebangsaan Malaysia, Bangi, Selangor, Malaysia
J. A. BETANCOURT-MAR • Mexican Institute of Complex Systems, Tamaulipas, Mexico
MARIANO BIZZARRI • Department of Experimental Medicine, Sapienza University, Rome, Italy
VICTORIA A. BOUFFARD • Tufts University School of Medicine, Boston, MA, USA
PAUL BRAZHNIK • Academy of Integrated Science, Virginia Tech, Blacksburg, VA, USA
ANNA LAURA CAPRIOTTI • Dipartimento di Chimica, Universita di Roma La Sapienza, Rome, Italy
ANDREA CERRATO • Dipartimento di Chimica, Universita di Roma La Sapienza, Rome, Italy
MARKO DJORDJEVIC • Quantitative Biology Group, Faculty of Biology, University of Belgrade, Belgrade, Serbia
ALESSANDRO GIULIANI • Environment and Health Department, Instituto Superiore di Sanita, Rome, Italy
A. GUERRA • Department of Chemical-Physics, A. Alzola Group of Thermodynamics of Complex Systems M.V. Lomonosov Chair, Faculty of Chemistry, University of Havana, Havana, Cuba
YAN TING HEE • Bioinformatics Institute (BII), Agency for Science, Technology and Research (A*STAR), Singapore, Republic of Singapore
CRISTIANO IALONGO • Department of Experimental Medicine, Sapienza University, Rome, Italy
MOSTAFA KHAFAEI • National Institute of Genetic Engineering and Biotechnology (NIGEB), Tehran, Iran
JASMEET KAUR KHANIJOU • Singapore Institute of Food and Biotechnology Innovation (SIFBI), Agency for Science, Technology and Research (A*STAR), Singapore, Republic of Singapore
ALDO LAGANA ` • Dipartimento di Chimica, Universita di Roma La Sapienza, Rome, Italy
GUGLIELMO LENTINI • Department of Experimental Medicine, Sapienza University, Rome, Italy; Systems Biology Group, Sapienza University, Rome, Italy
MICHAEL LEVIN • Department of Biology, Allen Discovery Center, Tufts University, Medford, MA, USA; Wyss Institute, Harvard University, Cambridge, MA, USA
J. A. LLANOS-PE ´ REZ • Mexican Institute of Complex Systems, Tamaulipas, Mexico
R. MANSILLA • Centro Peninsular en Humanidades y Ciencias Sociales (CEPHCIS), National Autonomous University of Mexico (UNAM), Merida, Mexico
SOFIJA MARKOVIC ´ • Quantitative Biology Group, Faculty of Biology, University of Belgrade, Belgrade, Serbia
PATRICK MCMILLEN • Department of Biology, Allen Discovery Center, Tufts University, Medford, MA, USA
AHMED MEDIANI • Institute of Systems Biology (INBIOSIS), Universiti Kebangsaan Malaysia, Bangi, Selangor, Malaysia
MAEL MONTE ´ VIL • Centre Cavaille´s, Re´publique des savoirs UAR, USA 3608, E ´ cole Normale Supe´rieure, Paris, France
NOEMI MONTI • Department of Experimental Medicine, Sapienza University, Rome, Italy; Systems Biology Group, Sapienza University, Rome, Italy
CARMELA MARIA MONTONE • Dipartimento di Chimica, Universita di Roma La Sapienza, Rome, Italy
J. M. NIETO-VILLAR • Department of Chemical-Physics, A. Alzola Group of Thermodynamics of Complex Systems M.V. Lomonosov Chair, Faculty of Chemistry, University of Havana, Havana, Cuba
ANDRAS PALDI • Ecole Pratique des Hautes Etudes, PSL Research University, St-Antoine Research Center, INSERM U938, Paris, France
AURORA PIOMBAROLO • Department of Experimental Medicine, Sapienza University, Rome, Italy
A. QUERQUI • Department of Experimental Medicine, Sapienza University, Rome, Italy; Systems Biology Group, Sapienza University, Rome, Italy
LAETITIA RACINE • Ecole Pratique des Hautes Etudes, PSL Research University, St-Antoine Research Center, INSERM U938, Paris, France
IGOR SALOM • Institute of Physics Belgrade, National Institute of the Republic of Serbia, Belgrade, Serbia
CHERYL M. SCHAEBERLE • Tufts University School of Medicine, Boston, MA, USA
KUMAR SELVARAJOO • Bioinformatics Institute (BII), Agency for Science, Technology and Research (A*STAR), Singapore, Republic of Singapore; Synthetic Biology for Clinical and Technological Innovation (SynCTI), National University of Singapore (NUS), Singapore, Republic of Singapore; School of Biological Sciences, Nanyang Technological University (NTU), Singapore, Republic of Singapore
CARLOS SONNENSCHEIN • Tufts University School of Medicine, Boston, MA, USA
ANA M. SOTO • Tufts University School of Medicine, Boston, MA, USA
M. TAFANI • Department of Experimental Medicine, Sapienza University, Rome, Italy
ENRICO TAGLIONI • Dipartimento di Chimica, Universita di Roma La Sapienza, Rome, Italy
MAHMOUD TAVALLAIE • Kawsar Human Genetic Research Center, Tehran, Iran
MASA TSUCHIYA • SEIKO Life Science Laboratory, SEIKO Research Institute for Education, Osaka, Japan
Part I
Models in Systems Biology and Parameters Identification
Chapter 1
Identifying Key In Silico Knockout for Enhancement of Limonene Yield Through Dynamic Metabolic Modelling
Jasmeet Kaur Khanijou, Yan Ting Hee, and Kumar Selvarajoo
Abstract
Living cells display dynamic and complex behaviors. To understand their response and to infer novel insights not possible with traditional reductionist approaches, over the last few decades various computational modelling methodologies have been developed. In this chapter, we focus on modelling the dynamic metabolic response, using linear and nonlinear ordinary differential equations, of an engineered Escherichia coli MG1655 strain with plasmid pJBEI-6409 that produces limonene. We show the systems biology steps involved from collecting time-series data of living cells, to dynamic model creation and fitting the model with experimental responses using COPASI software.
Key words Dynamic metabolic modelling, COPASI, Parameter estimation, Limonene, Time-series data
1 Introduction
The issue of food security has been brought to light in recent years, primarily driven by climate change affecting crop yields across the world. The recent coronavirus disease 2019 (COVID-19) pandemic, rising fuel prices and various political strife between countries affecting supply chains magnified the importance of food security especially in land-scarce, import-dependent nations [1, 2]. The environmental impact of current agricultural methods, including destruction of natural habitats and emission of greenhouse gases, has become increasingly urgent to address as the Earth’s global temperatures inches towards the 1.5 °C warming limit. In view of the above mentioned, but not limited to, concerns, many countries have been looking to diversify their food sources, one of which is to investigate alternative and cultured food
Jasmeet Kaur Khanijou and Yan Ting Hee contributed equally with all other contributors.
possibilities. These include plant-based meat, lab-grown meat and using microbes as biofactories to produce compounds of interest [3].
(S)-Limonene, hereafter referred to as limonene, is a monoterpene compound with a sweet citrusy fragrance. Holding the generally recognized as safe (GRAS) status issued by the US Food and Drug Administration, there has been a growing application for it, besides its current use as a food, flavoring, and fragrance additive [4]. Traditionally, limonene is produced from waste orange peel with current market price ranging between $7.92 to $20.15 per kilogram [5]. However, this approach can be limited both by crop yields and the difficulty of citrus rind recycling [4, 5]. To address these issues, several groups have engineered E. coli to increase limonene and other terpenes yields [6–11], and consequently to decrease their market price.
The precursor to limonene, geranyl diphosphate (GPP), is produced from the condensation of isopentenyl pyrophosphate (IPP) and its isomer, dimethylallyl pyrophosphate (DMAPP). IPP/DMAPP can be formed from two pathways the deoxyxyluose phosphate (DXP) and the mevalonate (MVA) pathways (see Fig. 1). While the DXP pathway occurs in E. coli natively, it only results in the production of small amounts of terpenes. Engineering of the DXP pathway for higher yields of terpenes were met with limitations possibly due to physiological pathway regulations in E. coli [11]. Therefore, the MVA pathway was engineered to be expressed alongside instead [6, 9, 11] (see Fig. 1). Building on these past works, there have been commendable efforts to further increase limonene yields from E. coli, ranging from transcriptional tuning [12] to genetically engineering for better ribosome binding sites, promoter optimization and utilization of plant, instead of native, ispA gene [13]. However, these methods primarily target the MVA pathway elements alone and do not consider other relevant metabolic pathways, such as upstream glycolysis from a carbon source. Existence of bottlenecks and loss of flux are equally important [14] for consideration to maximize limonene yields.
Analyses of metabolic bottlenecks and flux distribution can be performed by several computational modeling approaches [14]. Kinetic models are dynamic models that are based on linear and nonlinear differential equations governed by the underlying (enzyme) kinetics for each reaction in a pathway. They are popularly used to describe dynamic biological processes, ranging from metabolic [15], to protein signaling [16, 17], to gene regulatory networks [18–20], for systems biology research. For instance, dynamic modelling of TRAIL signaling using time-series data revealed PKC as a target to overcome TRAIL resistance in cancer [21]. The same method also helped identify RIP1 as a target to suppress but not abolish proinflammatory signaling [22]. These examples, thus, highlight the huge potential of dynamic models built from
Fig. 1 Schematic of different pathways involved in limonene production in E. coli engineered with the mevalonate (MVA) pathway. Cofactor consumption is represented by curved arrows. Intermediates: Glcex, glucose extracellular; Glc, glucose; G6P, glucose-6-phosphate; 6PG, 6-phosphogluconate; X5P, xylulose-5phosphate; Ru5P, ribulose-5-phosphate; R5P, ribose-5-phosphate; F6P, fructose-6-phosphate; F16BP,
time-series experimental data to uncover important insights in complex biological networks.
In the context of metabolic network modelling, the rates of change of metabolites in the network are represented by deterministic ordinary differential equations (ODEs). The rate of each reaction step in the network is defined by a rate law that describes the enzyme’s kinetics, for instance the Michaelis-Menten equation (or mass-action kinetics for protein signaling reactions). Defining the ODE of metabolite A in terms of its associated rate laws, we obtain the generalized Eq. 1:
where si represents the stoichiometric coefficient of species A in reaction i and vi denotes the rate law of reaction i. Reactions that produce and consume A will have a positive and negative stoichiometric coefficient respectively. All the species in the network are thus represented by Eq. 1. An example is shown in Fig. 2. From the network topology above (see Fig. 1), BPG is produced and consumed by the GDH and PGK reactions respectively, with a
Fig. 2 ODE of BPG, which is produced from GAP by the GDH enzyme and later converted into PG3 by PGK. 3PG is written here as PG3 instead (see Note 1). Kinetic parameters, Vmax of forward (Vf) and reverse (Vr) reaction, and KM belong to their respective enzyme as denoted by the parentheses as either (GDH) or (PGK)
stoichiometric coefficient of 1 in both. These are thus represented by the “+” and “-” signs in its ODE, respectively. MichaelisMenten kinetics were used for both GDH and PGK.
Here, we describe the experimental methods to collect timeseries data, forming the network topology and kinetic model construction using COmplex PAthway SImulator (COPASI) [23]. Finally, we will also discuss how the time-series data is fed into the model for parameter estimations, further model tuning and model validation.
2 Materials
2.1 Cell Culture
2.2 Sample Preparation
1. MG1655 E. coli strain with plasmid pJBEI-6409 obtained from Taek Soon Lee (Addgene plasmid #47048) [6].
2. Pre-pre-culture Luria-Bertani medium: 10 g/L tryptone, 5 g/ L yeast extract, 10 g/L NaCl, and 30 μg/mL chloramphenicol.
5. Internal standard mixture: 50 μg/mL mevalonic acid-d3 (MVA-d3) and 10 μg/mL thymolphthalein monophosphate (TMP).
6. Centrifugal vacuum concentrator.
7. Methanol: 10 mM ammonium hydroxide (7:3).
1. Interface an Agilent 6230 time of flight-mass spectrometer (TOF-MS) and a Dual Agilent Jet Stream (AJS) ion source with an Agilent ultra-performance liquid chromatography (UPLC) 1290 system and a Waters Acquity UPLC BEH C18 (2.1 × 150 mm, 1.7 μm) column with a VanGuard pre-column (2.1 × 5 mm).
2. Use a dual mobile phase system mobile phase A: 5 mM ammonium formate in water (pH 9.5) and mobile phase B: 5 mM ammonium formate (pH 9.5) in acetonitrile: water (9:1).
3. Calibration mixtures containing different concentrations of intracellular intermediates are used to construct calibration curves: DHAP (DHAP + GAP pool) – 0.04 to 5 μg/mL; DXP – 0.04 to 10 μg/mL; F6P (F6P + G6P pool) – 0.04 to 10 μg/mL; F1, 6BP – 0.04 to 6 μg/mL; MVA – 0.04 to 10 μg/mL; R5P (R5P + Ru5P + X5P pool) – 0.04 to 10 μg/ mL; Pyruvate – 0.04 to 1.5 μg/mL; MVAP – 0.01 to 0.3 μg/ mL; GPP – 0.05 to 1.5 μg/mL; FPP – 0.05 to 1.5 μg/mL. To each 100 μL calibration mixture, add 10 μL of internal standard mixture. Calibration curves are used to determine linearity and the concentration of each compound in the prepared samples.
4. Software for metabolite quantitation (e.g., Agilent Masshunter Workstation Quantitative Analysis for TOF).
1. Interface an Agilent 1200 high performance liquid chromatography (HPLC) system with a Bio-rad Aminex HPX-87H column (300 × 7.8 mm) and 1260 Infinity II Refractive Index Detector (RID).
2.4.3 Secreted Limonene
2. Prepare 0.01 N sulphuric acid as mobile phase.
3. Use calibration mixtures with different concentrations of extracellular metabolites to construct calibration curves: Glucose –0.5 to 80 g/L; Lactic acid – 0.125 to 8 g/L; Acetic acid –0.125 to 80 g/L; Ethanol – 0.5 to 80 g/L. Calibration curves are used to establish linearity and the concentration of each extracellular metabolite in the prepared samples.
4. Software for metabolite quantitation.
1. Equip an Agilent 7890B gas chromatography mass spectrometry (GC-MS) system with a DB-5 ms column.
2. Calibration standards for limonene prepared in ethyl acetate from 0.05 μg/mL to 10 μg/mL. Use calibration curves to determine linearity and the concentration of limonene in the dodecane extracts from samples.
3. Software for limonene quantitation.
3 Methods
3.1 Production of Bacterial Culture
An overview of the experimental protocol is presented in Fig. 3. The protocol has been divided into five parts: production of bacterial culture; harvesting of samples; preparation of samples; instrumental analysis and mass spectrometry; and data analysis.
Limonene bacterial strains are grown as 50 mL cell cultures in 250 mL flasks and induced with IPTG. Each flask is sacrificed in duplicates from 2 h, 3 h, 6 h, and 7 h post-IPTG induction to determine intracellular and extracellular metabolite concentrations.
1. For pre-pre-culture, select a single colony and inoculate in 5 mL Luria-Bertani medium and chloramphenicol overnight at 37 °C and 220 rpm.
2. For pre-culture, wash cell pellet from pre-pre-culture and re-suspend in 50 mL M9 medium, leave overnight at 30 °C and 220 rpm.
3. For culture, add 100 μL of pre-cultures to 50 mL M9 medium in 250 mL flasks. Incubate at 30 °C and 220 rpm.
4. Upon reaching optical density of 1 at 600 nm, add IPTG.
5. Add 5 mL dodecane to cell cultures.
6. Leave at 30 °C and 220 rpm.
7. Sacrifice flasks over time-points at 2 h, 3 h, 6 h, and 7 h postIPTG induction.
Fig. 3 Workflow of generating time-series metabolomics data
3.2 Harvesting of Samples
3.2.1 Intracellular Metabolites
3.2.2 Extracellular Metabolites and Secreted Limonene
1. Subject 10 mL of cell cultures from various time-points to fast filtration.
2. Wash cell pellet left on polyamide membrane with 5 mL wash solution.
3. Place the polyamide membrane onto an aluminium foil and fold into half prior to quenching in liquid nitrogen.
4. Remove from liquid nitrogen and store at -80 °C prior to metabolite extraction.
1. Centrifuge 50 mL cell cultures from various time-points for 10 min at 3000 rpm.
2. Remove dodecane layer containing secreted limonene and store at -80 °C prior to GC-MS analysis.
3. Place collected supernatant samples at -80 °C prior to HPLCRID analysis for extracellular metabolites.
4. Dry cell pellets in oven overnight prior to weighing.
3.3 Preparation of Samples
3.3.1 Intracellular Metabolites
3.3.2 Extracellular Metabolites and Secreted Limonene
3.4 Instrumental Analysis and Mass Spectrometry
3.4.1 Intracellular Metabolites
3.4.2 Extracellular Metabolites
3.4.3 Secreted Limonene
1. Remove membranes from aluminum foil and place into 5 mL extraction solvent consisting of methanol, acetonitrile, and water.
2. Vortex for 1 min.
3. Sonicate for 3 min, 3 times, placing in ice prior to each sonication.
4. Place extracts into glass tubes and spike with 20 μL internal standards mixture.
5. Remove solvents using vacuum concentrator and reconstitute with 200 μL methanol and 10 mM ammonium hydroxide mixture. Dilute samples by 10 times and 100 times, as necessary.
1. Filter 1 mL of supernatant sample using polyamide filter prior to HPLC-RID analysis.
2. Dilute limonene samples using ethyl acetate by 10 or 100 times prior to GC-MS quantitative analysis.
1. For each run, inject 2 μL of sample. Start with 100% mobile phase A from 0–3.5 min with 0.1 mL/min flow rate to 100% mobile phase B at 12 min and hold for 8 min with increased flow rate to 0.5 mL/min. At 20 min, recalibrate system back to 100% mobile phase A for 5 min and hold for another 5 min. Keep column temperature constant at 35 °C.
2. Negative electrospray ionisation is used with the following TOF settings: Gas temperature, 325 °C; Gas flow, 11 L/min; Nebuliser pressure, 35 psi; Sheath gas temperature, 375 °C; Sheath gas flow, 11 L/min; Vcap voltage, 3500 V; Nozzle voltage, 500 V; Skimmer, 65; OctopoleRFPeak, 750; Scan rate, 2 spectra/s. Fragmentor voltage varied is throughout each 35 min sample analysis: 2–7.5 min, 140 V; 7.5–15 min, 100 V, 140 V and 150 V. UPLC flow diversions are as follows: 0–2 min to waste, 2–15 min to TOF-MS, and 15–35 min to waste.
1. Set RID temperature at 30 °C with positive polarity.
2. For each run, inject 5 μL of sample with isocratic gradient and 0.6 mL/min flow rate for 28 min. Set column temperature at 35 °C.
1. Equip an Agilent 7890B GC-MS system with a DB-5 ms column.
2. For each run, use a 10:1 split ratio with 10 mL/min split flow and an injection of 1 μL. Use a GC oven temperature program of 40 °C for 3 min, followed by a 10 °C/min ramp to 100 °C
3.5 Data Analysis
3.6 COPASI Software and Execution
3.6.1 Model Construction
and another 60 °C/min ramp to 220 °C with a hold time of 2 min. Injector and MS transfer line temperatures are 250 °C and 280 °C, respectively.
3. Operate the MS in selected ion-monitoring (SIM) mode using ions of m/z 136, 68 and 93, representing the molecular ion and two abundant fragmental ions of limonene.
1. Execute the quantitation of metabolites with appropriate dilutions as necessary with calibration curves obtained from metabolite standards.
The construction and analysis of kinetic models can be done via the open-source and stand-alone program COPASI [23]. COPASI (https://copasi.org) is available both in a graphical user interface (CopasiUI) and command line version (CopasiSE). Basico is also available as a simplified interface to using COPASI from Python.
1. Begin first by defining the model’s units under the Model tab. Set “Quantity Unit” and “Volume Unit” to be the same as the metabolite concentration unit used in the time-series data. Thus, if μmol/l is used, set “μmol” and “l” respectively. Keep “Time Unit” no longer than the minute scale (see Note 2).
2. Under Model → Biochemical → Compartments, create the compartment(s) to describe where the reactions would be taking place. For model simplicity, we would create just one compartment, using COPASI’s default settings of 3D dimensionality and fixed simulation type.
3. Under Reactions, add the reactions that occur in the metabolic network. Specify the reaction name and its chemical equation. Use equal sign (=) and “->” to indicate reversible and irreversible reaction respectively. Alternatively, use the reversible checkbox to toggle the reaction reversibility. For reactions that involve stoichiometric coefficient greater than 1, for instance AtoB reaction of 2AcCoA = AtAcCoA + CoA, write the reaction equation as shown in Fig. 4a. If “2AcCoA” or “2*AcCoA” (no spaces) were used instead, COPASI will treat it as a separate species from AcCoA.
4. Assign the appropriate rate law to describe the reaction kinetics from the drop-down menu. Note that rate laws that match the reaction characteristics (reversibility and the same number of substrates and/or products) would only appear for selection. COPASI has a few predefined rate laws to choose from, but a user-defined rate law can also be created by clicking on the “+” sign (see Fig. 4a).
5. In the window that appears next, enter the rate law formula in the Formula box. If the equation is syntactically correct, the
Fig. 4 Model construction. (a) Adding reaction to the model. (b) Adding user-defined rate laws. Top and bottom show syntactically correct and incorrect (due to a missing closing bracket) mathematical formula. (c) Performing mapping of variables between the reaction and rate law variables
3.6.2 Parameter
Estimation
button next to the box will become white and clickable to view the equation in a more visually understandable manner (see Fig. 4b, top). The button will appear grey and unclickable otherwise (see Fig. 4b, bottom; also see Note 1). Additionally, indicate if the rate law is for a “reversible,” “irreversible,” or “general” reaction below.
6. While the rate law is being entered, COPASI automatically parse and extract the variables as the default “Parameter” type. From the dropdown under Description, select if a variable is a “Substrate,” “Product,” or “Modifier” (reaction activator or inhibitor), where necessary.
7. When defining the new rate law is complete, click “Commit,” return to the reaction, and select it from the dropdown menu. Perform mapping of variables between the reaction and formula variables (see Fig. 4c).
8. As the reactions are created, species in the network will be automatically populated under the Species tab.
The reaction parameters such as Vmax and KM can be obtained from in vitro studies of the enzyme kinetics. However, such information may not be available for some enzymes. In addition, in vitro conditions may not necessarily reflect the conditions in living cells or systems [24]. In such cases, parameter values would need to be estimated based on time-series experimental data relevant to the metabolic network being built.
1. Save experimental time course results in tab-or commadelimited plain text file format (.csv, .tsv, .txt, etc.), having the column names in the first row.
2. Under Tasks → Parameter Estimation, import the experimental data and correctly inform COPASI the separator that should be used to automatically parse in the table of values (see Fig. 5a).
3. Select “Time Course” as the Experiment Type and indicate how each column of the input data associates with model elements, either “Time,” “independent,” “dependent,” or “ignored” (default). Upon selecting “dependent,” a pop-up window appears prompting you to specify what model element the column maps to. Navigate to Species → Transient Concentrations and select for the correct mapping (see Fig. 5b).
4. Choose either mean or mean square as the “Weight Method” as either ensures that all data columns contribute in the same order of magnitude towards the error of fit regardless of the size of their numerical values (see Fig. 5a)[25].
5. Save the setup of the imported experimental data by clicking “OK.”
Another random document with no related content on Scribd:
impression respectively Oblique rows of prominent cushions wind round the surface of the stem and branches: each cushion is prolonged upwards and downwards in the form of a narrow ridge with sloping sides which connects adjacent cushions by an ogee curve. At the upper limit of the broader kite-shaped portion of the cushion the ligular pit forms a conspicuous feature; immediately below this is the leaf-scar with its three small scars,—the lateral parichnos strands and the central leaf-trace. The two oval areas shown in fig. 185, D, just below the lower edge of the leaf-scars, represent the parichnos arms which impinge on the surface of the cushions on their way to the leaves, as explained on a previous page. It is possible that these areas were visible on the living stem as strands of loose parenchyma comparable with the lenticel-like pits on the stipules of Angiopteris[372] and the leaf-bases of Cyatheaceous ferns, or it may be that their prominence in the specimen before us is the result of the decay of a thin layer of superficial cortex which hid them on the living tree. Fig. 185, B, illustrates the appearance of a stem in a partially decorticated condition (Bergeria state). A further degree of decortication is seen in fig. 185, A, which represents the Knorria condition.
F . 185. Lepidodendron Veltheimianum. From specimens in Dr Kidston’s Collection. (Approximately nat. size.)
Fig. 157 shows a Ulodendron axis of this species; in the lower part the specimen illustrates the partial obliteration of the surface features as the result of the splitting of the outer bark consequent on growth in thickness of the tree. By an extension of the cracks, shown in an early stage in fig. 157, the leaf-cushions would be entirely destroyed and the surface of the bark would be characterised by longitudinal fissures simulating the vertical grooves and ridges of a Sigillarian stem. The large stumps of trees shown in the frontispiece to Volume I. are probably, as Kidston[373] suggests, trunks of L. Veltheimianum in which the leaf-cushions have been replaced by irregular longitudinal fissures. In old stems of Sigillaria the enlarged parichnos areas constitute a characteristic feature (p. 205), but it does not follow that the absence of large parichnos scars is a distinguishing feature of all Lepidodendra.
In this species, as in others, the form of the leaf-cushion exhibits a considerable range of variation dependent on the thickness of the shoot; the contiguous cushions of young branches become stretched
apart as the result of increasing girth of the whole organ, and casts of still older branches may exhibit very different surface-features[374] . The leaves as seen on impressions of slender branches are comparatively short, reaching a length of 1–2 cm. It is important to notice that leafy twigs of this species may bear terminal cones[375] resembling in form those of Picea excelsa and other recent conifers, though differing essentially in their morphological features.
The fossil stumps of trees represented in the frontispiece to Volume I. bear horizontally spreading and dichotomously branched root-like organs having the characters of Stigmaria ficoides[376] . Geinitz has suggested that Stigmaria inaequalis Göpp. may be the underground portion of Lepidodendron Veltheimianum.
It is unfortunately seldom possible to connect petrified Lepidodendron cones with particular species of the genus based on purely vegetative characters, but it is practically certain that we are justified in recognising certain strobili described by Williamson[377] from the Calciferous Sandstone series of Burntisland on the Firth of Forth as those of Lepidodendron Veltheimianum. Williamson believed that the cone which he described belonged to the plant with shoots characterised by the anatomical features of his species Lepidodendron brevifolium (= L. Veltheimianum), a conclusion which is confirmed by Kidston[378] . The cone of L. Veltheimianum, which reached a diameter of at least 1 cm. and a length of 4 cm., agrees in essentials with other species of Lepidostrobus; the axis has a single medullated stele of the same general type as that of the vegetative shoots of Lepidodendron fuliginosum and L. Harcourtii. The sporophylls are described by Williamson as spirally disposed, and Scott notices that in some specimens they are arranged in alternate whorls; as in recent Lycopods both forms of phyllotaxis may occur in the same species. The heterosporous nature of this strobilus, to which Scott first applied the name Lepidostrobus Veltheimianus, is clearly demonstrated by the two longitudinal sections contributed by Mr Carruthers and figured by Williamson in 1893[379] .
Each sporophyll, attached almost at right angles to the cone-axis, bears a radially elongated sporangium seated on the median line of its upper face; its margins are laterally expanded as a thin lamina;
from the middle of the lower face a narrow keel extends downwards between two sporangia belonging to a lower series. From the base of a sporangium a mass of sterile tissue penetrates into the sporeproducing region as in the large sporangia of Isoetes (cf. fig. 191, H, a, and fig. 133, H). The distal and free portion of the sporophylls is bent upwards as a protecting bract. Some of the sporangia in the upper part of the cone produced numerous microspores, while 8–16 megaspores occur in the lower sporangia. The megaspores, having a mean diameter of 0·8 mm. “quite 40 times the size of the microspores[380],” are characterised by tubular capitate appendages, and by a conspicuous three-lobed projection (fig. 191, E)[381] which, as Scott suggests, may represent the outer spore-wall which has split as the result of germination. It is not improbable, as shown in fig. 191, I, that this cap was present before germination. The megaspores represented in fig. 191, I, illustrate their characteristic form as seen in a section of a megasporangium, Sm; the open beaklike portion of the larger spore is probably the apical region which has split along the three-rayed lines. These lines form a characteristic feature of both recent and extinct spores and denote their origin in tetrads. The spore shown in fig. 191, E[382] , illustrates the external features. The apical region of the prothallus of a megaspore of Lepidodendron Veltheimianum described by Mr Gordon[383] consists of smaller cells than those occupying the greater part of the spore-cavity, a differentiation which he compares with that of the prothallus of Selaginella.
F . 186.
A, B. Lepidodendron Veltheimianum. (Botany School, Cambridge.)
C. Lepidodendron macrophyllum. (British Museum. No. 377.)
There can be little doubt that the petrified shoots described by Williamson[384] from the Calciferous Sandstone beds of Burntisland as Lepidodendron brevifolium are identical with specimens possessing the external features of L. Veltheimianum. In 1872 Dawson expressed the opinion that Williamson’s species should be referred to L. Veltheimianum, and evidence subsequently obtained confirms this view. The stele of this species is of the medullated type, differing from that of L. fuliginosum and L. Harcourtii in the absence
of prominent ridges on the external surface of the primary xylem, and from L. vasculare in the possession of a parenchymatous pith. In younger twigs the cortex consists of fairly homogeneous tissue, but in older branches there is a greater distinction between a delicate middle cortex and a stronger outer cortex. Fig. 186, A, represents a stem in which the vascular cylinder is composed of a primary xylem ring, x, 1·5 mm. broad, succeeded by a zone of secondary wood 1·2 cm. in breadth. The junction between the primary and secondary xylem is shown on a larger scale in fig. 186, B. The tissues abutting on the secondary xylem have not been preserved; the outer cortex, which consists chiefly of secondary elements, is divided superficially into unequal ridges corresponding to the leaf-cushions which have been more or less obliterated as the result of growth in thickness of the stem.
9. Lepidodendron Pedroanum (Carruthers).
In 1869 Mr Carruthers described some specimens of vegetative stems and isolated sporangia, collected by Mr Plant in Brazil, as Flemingites Pedroanus[385] . From a more recent account published by Zeiller[386] it is clear that Carruthers’ species is a true Lepidodendron; an examination of the type-specimens in the British Museum confirms this determination. The contiguous leaf-cushions have rounded angles similar in form to those of Lepidodendron Veltheimianum and L. dichotomum, but it is not unlikely that the Brazilian plant is specifically distinct from European species. A figure of one of the specimens on which Carruthers founded the species is given by Arber[387] in his Glossopteris Flora. The Brazilian plant is chiefly interesting as affording proof of the existence of Lepidodendron in the southern hemisphere; the species has also been recognised in South Africa from material collected by Mr Leslie at Vereeniging[388] .
As Zeiller[389] has suggested, it is not improbable that the fossils described by Renault[390] from Brazil as Lycopodiopsis Derbyi may be the petrified stems of Lepidodendron Pedroanum. The structure of the central cylinder of Renault’s species is of the type represented by
L. Harcourtii; the xylem forms a continuous ring and does not consist of separate strands of tracheae as Renault believed.
Specimens described under this name are interesting rather on account of their extended geographical range and geological antiquity than on botanical grounds. The drawings reproduced in fig. 187 illustrate the characteristic appearance of this Lower Carboniferous and Upper Devonian type, as represented by a specimen recently described[391] from the Lower Karroo (Dwyka) series, which is probably of Carboniferous age, near Orange River Station, South Africa. The surface is divided into polygonal or rhomboidal areas (figs. A and B) 8–9 mm. long and 7–8 mm. broad, arranged in regular series and representing leaf-scars, comparable with those of Sigillaria Brardi and other species, or possibly partially decorticated leaf-cushions. A short distance below the apex of each area there is a more or less circular prominence or depression (fig. 187, B) and on a few of the areas there are indications of a groove (fig. A, g) extending from the raised scar to the pointed base, as at g, g.
F . 187. Lepidodendron australe. Fig. A, nat. size.
In examining the graphitic layer on the surface of the South African specimen shown in fig. 187, A, use was made of a method recently described by Professor Nathorst[392] . A few drops of collodion were placed on the surface, and after a short interval the film was removed and mounted on a slide. The addition of a stain facilitated the microscopic examination and the drawing of the collodion film. The cell-outlines (fig. 187, C) on the surface of the polygonal areas may be those of the epidermis, but they were more probably formed by a subepidermal tissue; the scar, which interrupts the continuity of the flat surface, may mark the position of a leaf-base, or, assuming a partial decortication to have occurred prior to fossilisation, it may represent a gap in the cortical tissue caused by the decay of delicate
tissue which surrounded the vascular bundle of each leaf in its course through the cortex of the stem. If the impression were that of the actual surface of a Lepidodendron or a Sigillaria, we should expect to find traces of the parichnos appearing on the leaf-scar as two small scars, one on each side of the leaf-bundle. In specimens from Vereeniging described in 1897[393] as Sigillaria Brardi, which bear a superficial resemblance to that shown in fig. A, the parichnos is clearly shown. On the other hand, an impression of a partially decorticated Lepidodendroid stem need not necessarily show the parichnos as a distinct feature: owing to its close association with the leaf-trace in the outer cortex, before its separation in the form of two diverging arms, it would not appear as a distinct gap apart from that representing the leaf-bundle. The absence of the parichnos may be regarded as a point in favour of the view that the impression is that of a partially decorticated stem. Similarly, the absence of any demarcation between a leaf-cushion and a true leaf-scar such as characterises the stems of Lepidodendra and many Sigillariae is also favourable to the same interpretation.
In 1872 Mr Carruthers[394] described some fossils from Queensland, some of which appear to be identical with that shown in fig. 187 under the name Lepidodendron nothum, Unger[395] , a species founded on Upper Devonian specimens from Thuringia. The Queensland plant is probably identical with Dawson’s Canadian species, Leptophloeum rhombicum[396] . In 1874 M’Coy[397] instituted the name Lepidodendron australe for some Lower Carboniferous specimens from Victoria, Australia: these are in all probability identical with the Queensland fossils referred by Carruthers to Unger’s species, but as the identity of the German and Australian plants is very doubtful[398] it is better to adopt M’Coy’s specific designation.
Krasser[399] has described a similar, but probably not specifically identical, type from China; from Devonian rocks of Spitzbergen Nathorst[400] has figured, under the name Bergeria, an example of this form of stem, and Szajnocha[401] has described other specimens from Lower Carboniferous strata in the Argentine.
Lepidodendron australe has been recorded from several Australian localities[402] from strata below those containing the genus Glossopteris and other members of the Glossopteris, or, as it has recently been re-christened, the Gangamopteris[403] Flora.
viii. Fertile shoots of
Lepidodendron.
A. Lepidostrobus.
The generic name Lepidostrobus was first used by Brongniart[404] for the cones of Lepidodendron, the type-species of the genus being Lepidostrobus ornatus, the designation given by the author of the genus to a Lepidostrobus previously figured by Parkinson[405] in his Organic Remains of a Former World. The generic name Flemingites proposed by Carruthers[406] in 1865, under a misapprehension as to the nature of spores which he identified as sporangia, was applied to specimens of true Lepidostrobi. Brongniart also instituted the generic name Lepidophyllum for detached leaves of Lepidodendron, both vegetative and fertile; the specimen figured by him in 1822 as Filicites (Glossopteris) dubius[407] , and which was afterwards made the type-species of the genus, was recognised as being a portion of the lanceolate limb of a large single-veined sporophyll belonging to a species of Lepidostrobus.
In an unusually large Lepidophyllum, or detached sporophyll of Lepidostrobus, in the Manchester University Museum, the free laminar portion reaches a length of 8 cm.
It is not uncommon to find Lepidodendron preserved in the form of a shell of outer cortex, which has become separated along the phellogen from the rest of the stem; as the result of compression the cylinder of bark may assume the appearance of a flattened stem covered with leaf-cushions. A specimen preserved in this way was described by E. Weiss as a cone of Lomatophloios macrolepidotus Gold., and is quoted by Solms-Laubach and other authors[408] as an example of an unusually large Lepidostrobus. An examination of the type-specimen in the Bergakademie of Berlin convinced me that Weiss had mistaken the partially destroyed leaf-cushions for
sporophylls, and Stigmarian rootlets, which had invaded the empty space, for sporangia[409] .
In external appearance some species of Lepidostrobus bear a superficial resemblance to the cone of a Spruce Fir (Picea excelsa), but the surface of a lycopodiaceous strobilus is usually covered by the overlapping and upturned laminae which terminate the more or less horizontal sporangium-bearing portion of the sporophyll.
Fig. 188 affords a good example of a long and narrow Lepidostrobus. This specimen from the Middle Coal-Measures of Lancashire has a length of 23 cm.; like other Lepidostrobi it is borne at the tip of a slender shoot. The fossil is sufficiently well preserved to show the characteristic radially elongated form of the large sporangia and the long and upturned distal portions of the sporophylls.
We may briefly describe Lepidostrobus as follows:—Cylindrical strobili consisting of an axis containing a single cylindrical stele which agrees generally with that of the vegetative shoots of L. Harcourtii and other species. The amount of parenchymatous pith varies in different forms; in some the primary xylem is almost solid. The middle cortical region, which has usually been destroyed before fossilisation, possesses the loose lacunar structure characteristic of this region in the vegetative branches. The thicker walled outer cortex is continued at the periphery into crowded, usually spirally disposed sporophylls, each of which consists of a more or less horizontal pedicel, which may be characterised by a keel-like median ridge on its lower surface, while to the central region of the upper face is attached a large radially elongated sporangium. One of the chief differences between a Lepidodendron cone and those of the recent genus Lycopodium is the greater radial elongation of the sporangia in the former Some species of Lepidostrobus may have been homosporous; some are known to be heterosporous. In the latter the megasporangia borne on the lower sporophylls usually contain several megaspores as in Isoetes (cf. fig. 133, E). Beyond the distal end of the sporangium the sporophyll becomes broader in a horizontal plane and is bent upwards as a lanceolate limb; it may
also be prolonged a short distance downwards as a bluntly triangular expansion.
F . 188. Lepidostrobus. Middle Coal-Measures, Bardsley, Lancashire. From a specimen in the Manchester Museum. (½ nat. size.)
There can be little doubt that the Palaeozoic Lepidodendra, like Lycopodium cernuum (fig. 123) and other recent Lycopods, usually bore their cones at the tips of slender shoots. The fertile shoot of Lepidophloios scoticus shown in fig. 160, B, affords one of several instances supporting this statement; similar examples are figured by Brongniart[410], Morris[411] , and by more recent writers. The apparently
sessile cone figured by Williamson[412] from a specimen in the Manchester Museum is certainly not in situ, but is accidentally associated with the stem.
The general absence of secondary wood in the steles of Lepidostrobi is, as Dr Kidston[413] points out, consistent with the view that the cones were shed on maturity and that fertilisation probably took place on the ground, or perhaps on the surface of the water where the slender hairs of the megaspores (fig. 191, F, I) may have served to catch the microspores.
F . 189. Lepidostrobus. Section through the apical region of a cone above the axis. (Manchester University Collection.)
Fig. 189 is an accurate representation of a transverse section, 6 mm. in diameter, of what is no doubt the apical portion of a Lepidostrobus from the Coal-Measures of Shore, Lancashire. The section cuts across the upturned free laminae above the level of the
apex of the cone-axis. Each lamina contains a small vascular bundle composed of a few tracheae and some thin-walled cells surrounded by delicate mesophyll tissue. Immediately in front of the distal end of a sporangium a small ligule is borne on the upper face of the sporophyll (fig. 191, A, B, l) occupying the same position as in Selaginella (cf. fig. 131, F). Strands of vascular tissue pass in a steeply ascending course from the xylem to the pedicels of sporophylls, finally curving upwards and ending in the upper limb. Each vascular bundle consists of a strand of xylem, apparently of mesarch structure, accompanied by a few layers of parenchyma on its outer face and by a group of cambiform elements, the whole being enclosed in a sheath of parenchyma continuous with the inner cortex of the cone axis. The vascular bundle is accompanied by a parichnos in the outer cortex and in the sporophyll.
Reference has already been made to the belief on the part of some palaeobotanists that the large scars of Ulodendron represent attachment-surfaces of sessile cones, and reasons have been given against the acceptance of this view.
There is considerable range in the size of Lepidostrobi An incomplete specimen, 33 cm. long and 6 cm. broad, which may have been 50 cm. in length, is described by Renault and Zeiller[414] from the Commentry Coal-field. The larger cones afford a striking demonstration of the enormous spore-output of some species of Lepidodendron.
Among the earliest accounts of the anatomy of Lepidostrobus are those by Hooker[415] and Binney[416] . One of the specimens described by the former author (fig. 190) affords an interesting example of an unusual manner of fossilisation; a hollow stem or Lepidodendron is filled with sedimentary material containing several pieces of Lepidostrobi in an approximately vertical position.
F . 190. Lepidodendron stem with Lepidostrobi. (After Hooker.)
A. Side-view showing leaf-cushions on the left-hand side and the Knorria condition on the right.
B. View of transverse section; s, sections of Lepidostrobi.
The fact that Lepidostrobi usually occur as isolated specimens renders it impossible in most cases to refer them to particular species of Lepidodendron. Neither external features nor anatomical characters afford satisfactory criteria by which to correlate vegetative and fertile shoots; in some measure this is due to the imperfection of our knowledge as regards the range of structure within the limits of species; it is also due to lack of information as to the extent to which
the transition from sterile to fertile portions of a shoot is accompanied by anatomical differences. Prof. Williamson wrote: “I have for many years endeavoured to discover some specific characters by which different Lepidostrobi can be distinguished and identified, but thus far my efforts have been unsuccessful[417].” In a few cases, such as those mentioned in the description of Lepidodendron Veltheimianum and L. Wünschianum, it has been possible to correlate cones and vegetative shoots.
The most complete account we possess of the anatomy of Lepidodendron cones is that by Mr Maslen[418] , who first demonstrated the occurrence of a ligule on the sporophylls, and thus supplied a missing piece of evidence in support of the generally accepted view as to the homology of the sporangium-bearing members and foliage leaves.
i. Lepidostrobus variabilis (Lindley and Hutton).
1811. “Strobilus,” Parkinson, Organic Remains, Vol. . p. 428, Pl. . fig. 1.
1828. Lepidostrobus ornatus, Brongniart, Prodrome, p. 87.
1831. L. variabilis, Lindley and Hutton, Foss. Flora, Pls. . .
1831. L. ornatus, Lindley and Hutton, Foss. Flora, Pl. .
1837. L. ornatus var. didymus, Ibid. Pl. .
1850. Arancarites Cordai, Unger, Genera et Spec. Plant. foss. p. 382.
1875. Lepidostrobus variabilis, Feistmantel, Palaeontographica, Vol. . Pl. .
1886. L. variabilis, Kidston, Cat. Palaeozoic Plants, p. 197.
1890. L. ornatus, Zeiller, Flor. Valenciennes, p. 497, Pl. . figs. 5, 6.
—— L. variabilis, Zeiller, Flor. Valenciennes, p. 499, Pl. . figs. 3, 4.
Under this specific name are included strobili from Upper Carboniferous rocks which, in spite of minor differences, may be considered as one type. The cylindrical cones vary considerably in size, some reaching a length of 50 cm. or more. The sporophylls are
attached by a pedicel, 4–8 mm. long, at right angles to the axis, while the distal portion forms an oval lanceolate limb 10–20 mm. in length. The sporangia are 4–8 mm. long.
The branched example figured by Lindley and Hutton[419] as a variety (L. ornatus var. didymus) illustrates a phenomenon not uncommon in both Palaeozoic and recent lycopodiaceous strobili.
A–D. L. oldhamius.
F . 191. Lepidostrobus.
B, C, D. From sections in the Binney Collection, Cambridge.
E. Megaspore. (After Kidston.)
F. Megaspore (Coal-Measures, Halifax). (After Williamson.)
G. Megaspore of Lepidostrobus foliaceus. (After Mrs Scott.)
H. Tangential section of sporangium. (After Bower.)
I. Part of sporangium wall, Sm, of the cone of Lepidodendron Veltheimianum, enclosing two megaspores. (Cambridge Botany School.)
ii. Lepidostrobus oldhamius Williamson[420]. Fig. 191, A–D.
Williamson[421] instituted this term for strobili previously described by Binney[422] , without adequate evidence, as the cones of Lepidodendron Harcourtii. In shape and in the main morphological features this type resembles L. variabilis, which is however known only in the form of casts and impressions. A cone of L. oldhamius, 2–3 cm. in diameter, possesses a medullated stele consisting of a ring of primary xylem (fig. 191, D, x) with exarch protoxylem and no secondary elements. Maslen found several short tracheae at the periphery of the xylem and states that these led him to compare the cone with the vegetative shoots of Lepidodendron vasculare, but the common occurrence of such elements in different types of shoot renders them of little or no specific value. The inner cortex is like that of vegetative shoots of Lepidodendron and the middle cortex, which was no doubt of the type described in Lepidostrobus Brownii, is represented by a gap in the sections, beyond which is the stronger outer cortex (fig. 191, D) passing into the horizontal pedicels of the sporophylls. The section of the axis reproduced in fig. 191, D, was figured by Binney[423] as Lepidodendron vasculare. The leaf-traces, several of which are seen in the middle cortical region in fig. D, lt, consist of a strand of scalariform tracheae, with a mesarch protoxylem, succeeded by a few parenchymatous cells; beyond these there is usually a small gap which was originally occupied by a strand of thin-walled cells. It is important to note that in one sporophyll-trace figured by Maslen[424] there is a strand of thin-walled elongated elements abutting on the xylem, which he describes as phloem. This tissue is certainly more like true phloem than any which has hitherto been described in the leaf-traces of vegetative shoots. The state of preservation is not, however, sufficiently good to enable us to recognise undoubted phloem features.
In such cones as I have examined no tissue has been seen which shows the histological features characteristic of the secretory zone of vegetative shoots: the “phloem” (Maslen) occupies the position in the sporophyll bundle which in the vascular bundles of foliage leaves is occupied by a dark-celled and partially disorganised tissue in continuity with the secretory zone of the main stele. It may be that in
the strobili this tissue occurred in a modified form, but even assuming that the section figured by Maslen shows true phloem, an assumption based on slender evidence, this is not sufficient justification for the application of the term phloem to a tissue occupying a corresponding position in vegetative shoots and distinguished by well-marked histological features.
The sporophyll-traces, as seen in the outer cortex in fig. 191, D, are partially surrounded by a large crescentic space, p, which was originally occupied by the parichnos. The sporangia are attached along the middle line of the sporophyll and, as in Lepidostrobus Brownii, a cushion of parenchyma projects into the lower part of the sporangial cavity (fig. 191, A, a; C, a).
The diagrammatic sketch of part of a section in the Binney Collection reproduced in fig. 191, B, shows the position of the ligule, l. No megaspores have been discovered in any specimens of this type; the microspores, which occur both singly and in tetrads, have a length of 0·02–0·03 mm.
The drawing shown in fig. 191, A, based on a section in the Binney Collection, illustrates the general arrangement of the parts of a typical Lepidostrobus. I have made use of this sketch instead of that given by Maslen, as his figure conveys the idea that the sporophylls are superposed, whereas, whether they are verticillate or spiral, a radial longitudinal section would not cut successive sporangia in the same plane.
iii. Lepidostrobus Brownii (Brongn.).
In 1843 a specimen of a portion of a petrified cone was purchased by the British Museum, assisted by the Marquis of Northampton and Robert Brown, for £30 from a French dealer. This fossil, from an unknown locality, was briefly described by Brown in 1851[425] and named by him Triplosporites, but in a note added to his paper he expressed the opinion that the generic designation Lepidostrobus would be more appropriate. Brongniart afterwards named the cone Triplosporites Brownii[426] , and Schimper[427] described it in his Traité as Lepidostrobus Brownii. The type-specimen is preserved in the
British Museum and the Paris Museum possesses a piece of the same fossil.
The central axis of the cone has a stele of the type characteristic of Lepidodendron fuliginosum and L. Harcourtii, and the xylem is surrounded by a thin-walled tissue described by Bower[428] as possibly phloem; but in the absence of longitudinal sections it is impossible to say how far the tissue external to the xylem agrees with that in Lepidodendron stems. The sporophylls consist of a horizontal portion, to the upper face of which the radially elongated sporangia are attached, one to each sporophyll; beyond the distal end of the sporangium the sporophyll bends sharply upwards as a fairly stout lamina. The wall of the sporangium is composed of several layers of cells, as shown in a drawing published by Bower[429]; in the interior occur groups of microspores, and from a ridge of tissue which extends along the whole length of the sporangium irregular trabeculae of sterile tissue project into the sporangial cavity, as in Isoetes (fig. 191, H: cf. fig. 133, H).
Further information in regard to Lepidostrobus Brownii has recently been supplied by Prof. Zeiller[430] , who recognises the existence of a ligule, and draws attention to some interesting histological features in the tissue of the sporophylls[431] .
Spores of Palaeozoic Lycopodiales.
The calcareous nodules from the Coal seams of Yorkshire and Lancashire are rich in isolated spores, many of which are undoubtedly those of Lepidostrobi. Examples of spores were figured by Morris[432] in 1840, and their occurrence in coal has been described by several authors, one of the earliest accounts being by Balfour[433] . The drawings of Palaeozoic and recent spores published by Kidston and Bennie[434] demonstrate a striking similarity between the megaspores of existing and extinct Lycopods, the chief difference being the larger size of the fossils.
The general generic name Triletes, originally used by Reinsch[435] , is a convenient term by which to designate Pteridophytic spores which cannot be referred to definite types.
