Investigating the genetic basis of altered activity profiles in the
blind Mexican cavefish, Astyanax mexicanus
A dissertation submitted to the
Graduate School
of the University of Cincinnati
in partial fulfillment of the requirements for the degree of
Doctor of Philosophy
in the Department of Biological Sciences
of the McMicken College of Arts and Sciences
by
by Brian M. Carlson
B.S. Biology, Xavier University, May 2010
Committee Chair: Dr. Joshua B. Gross
June 2015 ABSTRACT
Organisms that have evolved to exploit extreme ecological niches may alter or abandon survival strategies that no longer provide a benefit, or may even impose a cost, in the environment to which they have adapted. Cave environments are characterized by perpetual darkness, isolation and relatively constant temperature and humidity. Accordingly, cave-adapted species tend to converge on a suite of regressive and constructive morphological, physiological and behavioral alterations, including loss or reduction of eyes and pigmentation, increased locomotor activity and reduction or alteration of behavioral rhythmicity. The cave environment and the associated changes in locomotor behavior make species of cavefish prime natural models in which to examine the complex genetic architecture underlying these behavioral phenotypes.
The principal goal of this dissertation was to investigate the genetic basis of altered locomotor activity patterns in the blind Mexican tetra, Astyanax mexicanus. Initially, a custom locomotor assay rig and experimental protocols were developed to assess, characterize and compare activity patterns in surface and Pachón cavefish. The results of these assays clarified differences between the morphotypes, provided evidence that Pachón cavefish retain a weakly-entrainable circadian oscillator with limited capacity to self-sustain entrained rhythms and suggested that patterns in spatial “tank usage” data may be the result of a positive masking effect in response to light stimulus in both morphotypes. In order to facilitate downstream genetic analysis, a high-density genotyping-by-sequencing (GBS) based linkage map was constructed. This map provided much finer mapping ability than any previously published linkage map in this species. Further, the large number of markers enabled the anchoring of ~80% of the total length of the unassembled
Astyanax mexicanus draft genome to the newly developed linkage map, which in turn assisted with downstream comparison of linkage groups and quantitative trait loci (QTL) between maps.
! ii! This greatly facilitated the process of screening for potential candidate genes. Employing the new linkage map and the assay protocols developed for this work, a large F2 surface x Pachón mapping pedigree was assayed and the resulting data was subjected to QTL analysis. QTL results suggest that loci on at least six different linkage groups are associated with patterns in either velocity or tank usage; however, the regions of the genome mediating each of these components are distinct from one another. Further, comparison of linkage groups between maps indicated that the QTL for locomotor activity described here are different than those published in previous studies. Finally, critical genomic intervals underlying activity QTL were screened for candidate genes. A total of 36 potential candidates were identified and genomic and transcriptomic resources were leveraged to highlight several prime candidates for further analysis, as well as to demonstrate that it is unlikely that either catastrophic changes to components of the core molecular clockwork or truncation of members of well-characterized photoreceptor families are behind the altered activity profiles observed in the Pachón cave population.
! iii!
! iv! ACKNOWLEDGEMENTS
As my doctoral training draws to a close and I ready myself to begin what I hope will be a long, successful and satisfying career as an educator, a scientist and a mentor, I would like to take this opportunity to express my sincerest gratitude and heartfelt appreciation to the following people, all of whom have contributed in some significant way to my success in this endeavor:
Drs. Dorothy Engle, George Farnsworth and the rest of the biology faculty at Xavier University,
who provided me with my first opportunities to explore what it’s like to be a “real
biologist,” ensured that I was extremely well-prepared to excel in a doctoral program, and
provided an example of effective teaching and genuinely interested mentoring that
continues to inspire me to this day. It remains my sincere hope that I might some day
have the opportunity to return to my alma mater and help provide future students with the
same excellent learning environment and educational opportunities that were given to me.
Dr. Irene Luken and Mrs. Hedy Fye for giving me the knowledge and skills necessary to
correctly pronounce words like “zeitgeber” and work my way through old journal articles
with titles like, “Untersuchungen über den Lichtsinn und dessen Lokalisation bei dem
Höhlenfisch Anoptichtys jordani Hubbs and Innes (Characidae).”
Drs. Daniel Buchholz, John Layne, Herman Mays and Kenneth Petren for taking time out of
your busy schedules to provide me with the guidance and support necessary to ensure that
I was successful in my coursework and research efforts over the past five years.
Dr. Joshua Gross for being the most intelligent, kind, supportive and inspiring doctoral advisor
that I could have ever asked for. I cannot imagine any better environment in which to
! v! grow as a student, a scientist and a person than working with you against the backdrop of
the warm and inviting lab culture that you so carefully cultivate. You’ve somehow
managed to find the perfect balance between lab boss and lab dad and I am confident that
any students fortunate enough to spend time in your lab will be forever changed for the
better because of it; no one could be a stronger advocate for his or her students than
you’ve been for us.
Drs. Kathryn Rafferty and Kathleen Koenig for providing me the tools, training and
opportunities that have allowed me to develop and grow as an educator throughout the
course of my graduate career.
Mandy Powers for being my confidant, lunch buddy and partner in crime for the last several
years. I wish you success and many happy Pandora jam sessions as you work toward a
red robe and funny hat of your own.
Bethany Stahl for keeping me on my toes. Slacking off was never an option when I knew I’d
constantly have to compare my own drive, work ethic and scientific output to the bar you
somehow always managed to set just the tiniest bit higher than mine. Keep it up!
Tyler Bussian, Allison Furterer, Ian Klinger, Wendy Lu, Brad Meyer and Sam Onusko for
dedicated assistance in pushing my projects forward, no matter how tedious the task. I
hope that the time and effort that you contributed toward making my projects a success
comes back to you three fold!
The rest of the Gross Lab family, both past and present. I wish you all nothing but success and
happiness in whatever awaits you.
! vi!
Dr. Merritt Oleski for letting me take over your conference room while I worked furiously to
finish my dissertation and for distracting me with conversations about our shared belief in
the value of basic science research when you could tell I needed a break. Please consider
the contents of this dissertation as a few humble data points for inclusion in the Library of
Alexandria, Version 2.0.
Mr. L. for being the person most genuinely excited to talk with me about my research.
My friends, family and the Brethren of J.B. Covert Lodge #437 and Alembic Lodge #793 (“Ole
793”) for supporting me, keeping me grounded and providing me with much needed
distractions as I’ve worked hard to achieve my goals and make something of myself.
Most importantly, however, I would like to dedicate this dissertation and all of the time and labor
that went into writing it to the following, without any of whom it is absolutely
inconceivable that I would be where I am today:
To my wife, Julie Carlson, who has never waivered in her love and support for me and whose
quiet strength and encouragement were all that kept me going through the roughest, most
stressful portions of this experience. After well over a hundred pages of tedious
proofreading and weeks of late nights, cranky moods, neglected chores and general
unpleasantness on my part, I am positive that someone is traveling to the Vatican to
advance her cause for living sainthood even as we speak.
To my parents, Mary and Harry Brownfield and Daniel and Kathy Carlson, who have invested
untold amounts of time, energy and resources into making sure that I always had the very
! vii! best education that they could afford, that I never truly wanted for anything and that there
was never any doubt in my mind that I was loved and that they are proud of me. As a
small expression of my gratitude for everything that you’ve done for me:
Dad: Please consider the results of all of my analyses as the solution to the
biggest damn “fish problem” I could come up with.
Mom: Please consider the figures and tables contained within this tome as a poor
substitute for all of the “artist’s renderings” that you’ve missed since I went away
to college.
Finally, to the memory of my late grandfather, Rudolph Flis, who taught me through word and
deed the importance of family, faith and service to others… and that it’s nearly
impossible to stay stressed or unhappy when you’re listening to a good polka. Thanks
especially for that last one, Grandpa; it’s come in quite handy these past few weeks.
! viii!
PREFACE
Once we were blobs in the sea, and then fishes, and then lizards and rats, and then monkeys, and hundreds of things in between. This hand was once a fin, this hand once had claws! In my human mouth I have the pointy teeth of a wolf and the chisel teeth of a rabbit and the grinding teeth of a cow! Our blood is as salty as the sea we used to live in! When we're frightened, the hair on our skin stands up, just like it did when we had fur. We ARE history! Everything we've ever been on the way to becoming us, we still are. Would you like the rest of the story? I'm made up of the memories of my parents and my grandparents, all my ancestors. They're in the way I look, in the color of my hair. And I'm made up of everyone I've ever met who's changed the way I think.
- Terry Pratchett, A Hat Full of Sky
! ix! TABLE OF CONTENTS
Abstract……………………………………………………………………………………….…..ii
Acknowledgements……………………………………………………………………………….v
Preface…………………………………………………………………………………………....ix
List of Tables and Figures……………………………………………………………………….xii
Chapter 1: Introduction and Background……………………………………………………….....1
References…………………………………………………………………………………6
Chapter 2: Characterization and comparison of activity profiles exhibited by the cave and surface morphotypes of the blind Mexican tetra, Astyanax mexicanus……………………………………8
Abstract……………………………………………………………………………………9
Introduction………………………………………………………………………………10
Materials and Methods…………………………………………………………………...13
Results……………………………………………………………………………………18
Discussion………………………………………………………………………………..25
Conclusion……………………………………………………………………...………..37
Acknowledgements………………………………………………………………………38
References………………………………………………………………………………..38
Chapter 3: A high-density linkage map for Astyanax mexicanus using genotyping-by-sequencing technology……………………………………………...……………….………………………..61
Abstract…………………………………………………………………………………..62
Introduction………………………………………………………………………………63
Materials and Methods…………………………………………………………………...65
Results and Discussion……………………………………...………………...…………72
! x! Conclusion…………………………………………………………...…………………..79
Acknowledgements………………………………………………………………………80
References………………………………………………………………………………..80
Chapter 4: Genetic analysis reveals candidate genes potentially underlying altered activity profiles in the blind Mexican tetra, Astyanax mexicanus…………….………………………….98
Abstract…………………………………………………………………………………..99
Introduction……………………………………………………………………………..100
Materials and Methods………………………………………………………………….102
Results and Discussion………………………………………………………...….……107
Conclusion……………………………………………………………………………...118
Acknowledgements…………………………………………………………………..…118
References…………………………………………………………………………..…..119
Chapter 5: General Conclusions………..…………………………….………………………...152
! xi! LIST OF TABLES AND FIGURES
Chapter 2
Figure 2.1: Assay rig and video tracking.
Figure 2.2: 72hr assays reveal altered activity profiles in the Astyanax mexicanus cave morph.
Figure 2.3: Absence of light during development may affect activity profiles.
Figure 2.4: 72hr assays reveal altered spatial usage patterns in the Astyanax mexicanus cave
morph.
Figure 2.5: Absence of light during development may affect spatial usage patterns.
Table 2.1: Mean velocity for surface and cavefish activity assays.
Table 2.2: Results of cosinor rhythmometry for surface and cavefish velocity data.
Table 2.3: Mean top zone usage for surface and cavefish activity assays.
Table 2.4: Mean middle zone usage for surface and cavefish activity assays.
Table 2.5: Mean bottom zone usage for surface and cavefish activity assays.
Table 2.6: Results of cosinor rhythmometry for surface and cavefish top zone usage data.
Table 2.7: Results of cosinor rhythmometry for surface and cavefish bottom zone usage data.
Chapter 3
Figure 3.1: A GBS-based linkage map in the Mexican cave tetra, Astyanax mexicanus.
Figure 3.2: Short GBS sequences identify syntenic stretches between two Ostariophysian
freshwater fish species.
Figure 3.3: Whole-genome synteny between Astyanax and Danio and a proof-of-concept analysis
of albinism.
Figure 3.4: Colinearity between Astyanax linkage groups and genome scaffolds.
! xii! Table 3.1: Summary of BLAST results and identification of markers used in Astyanax-to-Danio
syntenic analysis.
Table 3.2: Comparison of Astyanax linkage maps and syntenic studies with Danio rerio.
Table 3.3: Comparison of syntenic analyses between Astyanax linkage maps and their
association with the Danio rerio genome across multiple studies.
Table 3.4: Representative analysis of linkage group equivalence and quality based on highly
syntenic chromosomes in Danio rerio and linkage groups in Astyanax mexicanus.
Chapter 4
Figure 4.1: F1 surface x Pachón hybrids show surface-like activity patterns with cave-like
influences.
Figure 4.2: F2 surface x Pachón hybrids show a wide array of parental and non-parental activity
patterns.
Figure 4.3: Locomotor activity QTL examined.
Figure 4.4: Non-synonymous mutation causes amino acid substitution at highly conserved
position in cavefish ATOH7.
Table 4.1: Results of QTL analysis for metrics of locomotor activity.
Table 4.2: Summary of candidate gene assessment.
Table 4.S1: Activity data for Asty66 F2 surface x Pachón hybrid pedigree assayed for 24hr under
12:12hr light/dark conditions.
Table 4.S2: Sex, albinism and eye data for Asty66 F2 surface x Pachón hybrid pedigree.
! xiii! Table 4.S3: Distribution of commonly-anchored genomic scaffolds between linkage maps
published by Carlson et al. (2015) and O’Quin et al. (2013).
Table 4.S4: Results of gene ontology-based candidate gene screen.
Table 4.S5: Results of gene expression analysis.
Table 4.S6: Results of sequence variation analysis.
! xiv! ! ! ! ! ! ! ! !
CHAPTER ONE
Introduction and Background
Brian M. Carlson
Department of Biological Sciences University of Cincinnati Cincinnati, OH 45221-0006
! 1! The ability to modify and coordinate behaviors and physiological processes in response to external stimuli is essential for the success and survival of a wide array of taxa (Bell-Pedersen et al. 2005). Cycles in sunlight and, to a lesser extent, temperature are considered the strongest and most ubiquitous environmental timing cues, or “zeitgebers” (lit. “time givers”) in the natural world. Therefore, not only is it advantageous for organisms to be able to rhythmically alter their behavior and physiology throughout the day to ensure the best alignment of endogenous rhythms with environmental cues, but an argument can be made that the development of such “circadian rhythmicity” is an evolutionary necessity (Mehta and Cheng 2013; Ki et al. 2015). As a result, strong selection for the development and maintenance of circadian rhythms has resulted in a high degree of conservation of the critical gene networks that mediate rhythmic, coordinated patterns in behavior and physiology. This is a double-edged sword from a human standpoint: on one hand, perturbations or aberrations within these particular genes or molecular pathways can result in deleterious departures from normal behavioral or physiological rhythms (Liu and Chu 2013).
On the other hand, however, the high level of conservation of these elements means that the more that we understand about the genetic and physiological aspects of circadian rhythmicity in other systems, the better we also understand our own biological rhythms and, potentially, how to treat or even cure associated diseases or disorders in a clinical setting (Takahashi et al. 2008).
Investigation of genetic anomalies underlying phenotypes of evolutionary, developmental and clinical relevance is often conducted in one of a growing number of “model” organisms.
Such organisms often prove to be particularly useful due to their simplicity, their tractability in a lab setting, known genetic background and/or the level of knowledge and number of optimized protocols and resources that can be easily leveraged to design and implement well-conceived experiments that produce consistent, reproducible results (Lowrey and Takahashi 2011).
! 2! However, while exposure to mutagens or artificial selection in a lab setting may result in an organism that reliably phenocopies a trait or syndrome of interest, there is often no guarantee that the specific locus and/or biological process that is responsible for generating the target phenotype in a model organism is one that would conceivably occur in a natural setting.
Therefore, researchers have made convincing arguments for the utilization of a wide array of so- called “natural model organisms” which, while not necessarily ideal to work with, represent the effects of naturally-occurring mutations or the results of positive or negative selective forces acting upon standing genetic variation within a population (Albertson et al. 2009).
When looking for natural models of changes in circadian rhythmicity, cave-adapted species have long been a focus of research attention (Friedrich 2013). While some of this is the result of a largely-erroneous, “surface-centered” notion that circadian rhythmicity will disappear entirely over hundreds or thousands of generations in constant darkness, cave environments do tend to harbor species in which behaviors and physiology that would otherwise rely on light are either altered, diminished or, in a few cases, entirely lost (Friedrich 2013; Menna-Barreto and
Trajano 2015). Given that caves are not only constantly dark, but also fairly isolated and constant with respect to temperature and humidity, each cave essentially functions as a controlled evolutionary experiment (Poulson and White 1969). Interestingly, these natural experiments produce fairly consistent results: more often than not, obligate cave-dwelling organisms that have inhabited their respective caves for thousands of generations are either arthropods or fish species that converge on the reduction or loss of eyes and pigment, along with a suite of other constructive and regressive cave-associated traits, including alterations in circadian rhythmicity and/or light-responsive behaviors (Romero and Green 2005; Protas et al. 2006; Wilkens 2011;
Friedrich 2013).
! 3! The blind Mexican tetra, Astyanax mexicanus, was originally described by Hubbs and
Innes (1936) based on the population found in the Chica cave in NE Mexico. In the decades following their discovery, another 28 cave populations have been identified (Mitchell et al.
1977). In addition to its tractability in the lab setting, the presence of multiple cave populations resulting from at least two waves of colonization by different ancestral surface populations over the last several million years makes Astyanax an ideal model for the study of convergent evolution. This convergence is reflected in the constellation of cave-associated traits that appear to develop again and again as established cave populations age (Hinaux et al. 2011; Bradic et al.
2012; Gross 2012; Gross et al. 2015). However, in addition to these features, it is the existence of an extant surface form with which members of the various cave populations can be bred to reliably produce viable offspring, thereby enabling both classical and molecular genetic studies, that has led Astyanax mexicanus to become the “white rat” of cave evolution (Jeffery 2009).
The primary goal of the research presented in this dissertation was to investigate the genetic basis of alterations in locomotor activity and circadian rhythmicity previously reported in
Astyanax mexicanus (see Chapter 2 for discussion of previous literature) and to better understand the role that lighting cues play in eliciting, modifying and/or entraining rhythmic patterns in behavior. To that end, I began by designing and conducting behavioral assays that allowed me to quantify, characterize and compare patterns in locomotor behavior in the Astyanax surface and cave (Pachón) morphotypes. Relationships between the behaviors observed and the presence or absence of cyclic lighting cues both during assays and over the course of development were then assessed (Chapter 2). In order to facilitate genetic studies, I constructed a high-density GBS- based linkage map in this species and worked to define the relationships between my map, previous maps and the physical genomes of both Astyanax mexicanus and the zebrafish model
! 4! system, thereby both improving the likelihood of identifying loci mediating locomotor activity and discovery of candidate genes associated with the phenotypes observed (Chapter 3). I then employed the knowledge and tools developed by my earlier work to assay locomotor activity in a large F2 surface x cave hybrid mapping pedigree. Subsequent QTL analysis demonstrated that 1) activity differences have a detectable genetic basis, 2) highlighted potential relationships between cave-associated traits and some elements of activity, and 3) suggested that the spatial and velocity components of locomotor activity are controlled independently. Finally, I was able to leverage available genomic and transcriptomic resources to identify potential candidate genes residing on genomic scaffolds anchored to my linkage map, highlight several well-supported candidates for further examination and support the argument that the genetic basis of observed differences of locomotor behavior is unlikely to be found in the core molecular clockwork or one of the well-characterized families of photoreceptors (Chapter 4). Together, the results presented in this dissertation serve to expand our knowledge of the complex genetic architecture likely underlying behavioral phenotypes in this species. This work also provides insights and tools that can inspire and facilitate future research, not only in Astyanax, but potentially a variety of other natural model systems as well (Chapter 5).
! 5! REFERENCES
Albertson, R. C., W. Cresko, H. W. Detrich, 3rd and J. H. Postlethwait, 2009 Evolutionary
mutant models for human disease. Trends Genet 25: 74-81.
Bell-Pedersen, D., V. M. Cassone, D. J. Earnest, S. S. Golden, P. E. Hardin et al., 2005
Circadian rhythms from multiple oscillators: Lessons from diverse organisms. Nat Rev
Genet 6: 544-556.
Bradic, M., P. Beerli, F. J. Garcia-de Leon, S. Esquivel-Bobadilla and R. L. Borowsky, 2012
Gene flow and population structure in the Mexican blind cavefish complex (Astyanax
mexicanus). BMC Evol Biol 12: 9.
Friedrich, M., 2013 Biological clocks and visual systems in cave-adapted animals at the dawn of
speleogenomics. Integr Comp Biol 53: 50-67.
Gross, J. B., 2012 The complex origin of Astyanax cavefish. BMC Evol Biol 12: 105.
Gross, J. B., B. Meyer and M. Perkins, 2015 The rise of Astyanax cavefish. Dev Dyn: Early
online. DOI: 10.1002/dvdy.24253.
Hinaux, H., K. Pottin, H. Chalhoub, S. Pere, Y. Elipot et al., 2011 A developmental staging table
for Astyanax mexicanus surface fish and Pachón cavefish. Zebrafish 8: 155-165.
Hubbs, C. L., and W. T. Innes, 1936 The first known blind fish of the family Characidae: a new
genus from Mexico. Occas Pap Mus Zool Univ Mich 342: 1-7.
Jeffery, W. R., 2009 Regressive evolution in Astyanax cavefish. Annu Rev Genet 43: 25-47.
Ki, Y., H. Ri, H. Lee, E. Yoo, J. Choe et al., 2015 Warming up your tick-tock: Temperature-
dependent regulation of circadian clocks. Neuroscientist: 1-16.
Liu, Z., and G. Chu, 2013 Chronobiology in mammalian health. Mol Biol Rep 40: 2491-2501.
! 6! Lowrey, P. L., and J. S. Takahashi, 2011 Genetics of circadian rhythms in mammalian model
organisms. Adv Genet 74: 175-230.
Mehta, N., and H. Y. Cheng, 2013 Micro-managing the circadian clock: The role of microRNAs
in biological timekeeping. J Mol Biol 425: 3609-3624.
Menna-Barreto, L., and E. Trajano, 2015 Biological rhythmicity in subterranean animals: A
function risking extinction?, pp. 55-68 in Mechanisms of Circadian Systems in Animals
and Their Clinical Relevance, edited by R. Aguilar-Roblero, M. Díaz-Muñoz and M. L.
Fanjul-Moles. Springer International Publishing.
Mitchell, R. W., W. H. Russell and W. R. Elliot, 1977 Mexican Eyeless Characin Fishes, Genus
Astyanax: Environment, Distribution and Evolution. Texas Tech Press, Lubbock, Texas.
Poulson, T. L., and W. B. White, 1969 The cave environment. Science 165: 971-981.
Protas, M. E., C. Hersey, D. Kochanek, Y. Zhou, H. Wilkens et al., 2006 Genetic analysis of
cavefish reveals molecular convergence in the evolution of albinism. Nat Genet 38: 107-
111.
Romero, A., and S. M. Green, 2005 The end of regressive evolution: Examining and interpreting
the evidence from cave fishes. J Fish Biol 67: 3-32.
Takahashi, J. S., H. K. Hong, C. H. Ko and E. L. McDearmon, 2008 The genetics of mammalian
circadian order and disorder: Implications for physiology and disease. Nat Rev Genet 9:
764-775.
Wilkens, H., 2011 Variability and loss of functionless traits in cave animals. Reply to Jeffery
(2010). Heredity (Edinb) 106: 707-708.
! 7!
CHAPTER TWO
Characterization and comparison of activity profiles exhibited by the cave and surface morphotypes of the blind Mexican tetra, Astyanax mexicanus
Brian M. Carlson and Joshua B. Gross*
Department of Biological Sciences University of Cincinnati Cincinnati, OH 45221-0006
*It is anticipated that the information contained in this chapter will be published elsewhere with Dr. Joshua B. Gross serving as a co-author. However, the narrative, analysis and representations of results included here have been developed to their current state solely by Brian M. Carlson.
8!! ABSTRACT
Adaptation to extreme environments has led many species to alter or even entirely abandon their reliance upon cyclic environmental inputs, principally daily cycles of light and darkness. The extreme darkness, stability and isolation of cave ecosystems has made cave- adapted species particularly attractive systems in which to study the consequences of life without light and the strategies that allow species to survive and even thrive in such environments. In order to further explore these questions, we have assessed the rhythmicity of locomotor rhythms in the blind Mexican tetra, Astyanax mexicanus, under controlled laboratory conditions. Using high-resolution video tracking assays, we characterized patterns in locomotor activity and spatial tank usage for members of the surface and Pachón cave populations. Here we demonstrate that cavefish have a higher overall level of activity and use the space within the trial tank differently than surface fish. Further, Pachón cavefish show circadian rhythmicity in both activity and spatial tank usage under a 12:12 light/dark cycle. We provide further evidence that these cavefish retain a weakly light-entrainable, endogenous circadian oscillator with limited capability to sustain rhythms in activity, but not spatial tank usage, in the absence of photic cues. Finally, we demonstrate a putative behavioral “masking effect” contributing to behavioral rhythms and provide evidence that exposure to constant darkness during development may alter behavioral patterns later in life.
9!! INTRODUCTION
The ability to coordinate biological processes and behaviors with changes in the environment is crucial for health and survival and has been demonstrated in a wide array of taxa from cyanobacteria to humans (Bell-Pedersen et al. 2005). Daily cycles in light and temperature are powerful environmental cues that have a prominent impact on a wide range of behaviors and physiological processes essential for life. Circadian rhythmicity is coordinated by one or more endogenous clocks and enables organisms to predict daily changes in their environment. This allows behaviors and physiological processes to occur at the most advantageous point of each
24hr cycle and can therefore be viewed as an evolutionary necessity (Ki et al. 2015). Further, it is becoming increasingly clear that the gene network facilitating rhythmic biological control in animals has deep evolutionary roots, suggesting that there has been strong selective pressure for the development and maintenance of circadian rhythmicity throughout evolutionary time
(Friedrich 2013).
Despite the widespread reliance upon daily cycles of light, many species have adapted to exploit new ecological niches in extreme environments where these critical cues are unreliable or even entirely absent. The ways in which these organisms cope with these pressures are as diverse as the organisms themselves. For example, there is evidence that the deep-sea eel
Synaphobranchus kaupii exhibits rhythms in melatonin production that may be entrained by cyclic changes in underwater current flow influenced by the moon (Wagner et al. 2007). In the subterranean mole rat Spalax ehrenbergi, which is visually blind but entrainable to light cycles, certain individuals have shown a shift from the normally observed “diurnal” activity rhythm to a
“nocturnal” rhythm; a shift accompanied by an apparent “uncoupling” of circadian rhythms from photic input (Oster et al. 2002). Strikingly, there is evidence that arctic reindeer appear to have
10! ! abandoned circadian rhythmicity entirely, relying instead on a “circannual” clock entrained by photic cues around the equinoxes (Lu et al. 2010). Examining the strategies that allow organisms to thrive in extreme environments and the manner in which their biological rhythms and molecular clockwork have changed over time has proven extremely useful as the scientific community works to expand and deepen our understanding of circadian rhythms and their evolutionary, ecological and clinical relevance.
Cave-adapted organisms have proven to be valuable study systems for the examination of circadian rhythmicity since cave environments are not only devoid of cyclic lighting cues, but also feature relatively constant temperature and humidity (Yerushalmi and Green 2009).
Cavefish are extremely attractive as study systems owing to the fact that they are vertebrates, inhabit a wide range of geographical regions, and represent at least 86 species across 18 different taxonomic families (Romero and Paulson 2001). To date, some aspect of circadian rhythmicity has been evaluated in at least 13 of these species. However, much of the published research has focused on three primary species: the hill stream loach, Nemacheilus evezardi (Balitoridae), the
Somalian cavefish, Phreatichthys andruzzii (Cyprinidae), and the blind Mexican tetra, Astyanax mexicanus (Characidae; Friedrich 2013).
Astyanax mexicanus has found favor as a study system for regressive traits and cave evolution owing to the fact that this species is tractable in a lab setting, includes 29 different cave populations and consists of both a derived, cave morphotype and an extant surface morphotype that can be readily hybridized (Mitchell et al. 1977; Jeffery 2009; Gross et al. 2015). In the decades following its initial discovery in 1936 (Hubbs and Innes), Astyanax was the subject of a number of studies aimed at 1) examining sensitivity and behavioral responses to light (Breder and Gresser 1941a; Breder and Gresser 1941b; Breder 1944; Breder and Rasquin 1947; Kuhn
11! ! and Kähling 1954; Lüling 1954; Kähling 1961; Gertychowa 1970; Romero 1984; Romero 1985;
Langecker 1989) and 2) determining whether Astyanax cavefish displayed circadian rhythmicity
(Thines and Wolff-Van Ermengem 1965; Erckens and Weber 1976; Thines and Weyers 1978;
Erckens and Martin 1982a; Erckens and Martin 1982b). For some time, the work of Erckens and
Martin (1982a; 1982b) stood as the authority on circadian rhythmicity in this species, despite small sample sizes (n = 6 and n = 1 for video analysis and infrared photocell-based trials, respectively, in both cave and surface fish). The results they presented were so nuanced that multiple subsequent authors have inadvertently misrepresented their findings (Boujard and
Leatherland 1992; Volpato and Trajano 2005; Zhdanova and Reebs 2006; Menna-Barreto and
Trajano 2015).
Recently, a growing community of researchers, aided by advances in genetic resources, laboratory techniques and assay methods, has shown renewed interest in studying these questions in Astyanax (Romero et al. 2003; Espinasa and Jeffery 2006; Yoshizawa and Jeffery 2008), with a dramatic increase in the number of publications on related topics in the last four years alone
(Duboué et al. 2011; Duboué and Borowsky 2012; Beale et al. 2013; Moran et al. 2014;
Caballero-Hernández et al. 2015; Yoshizawa et al. 2015).
In this study, we present the results of high-resolution video tracking assays of the
Astyanax mexicanus cave (Pachón) and surface morphotypes over multiple days under both a
12:12hr light/dark cycle and constant darkness. We examined both the locomotor and spatial components of activity in this species, allowing both aspects to be examined independently in a manner that has not been previously employed in this system. Detailed descriptions of the activity profiles observed, analysis of the presence or absence of circadian rhythmicity within the data and comparisons both between and within morphotypes are provided and then discussed
12! ! within the context of previously published work in this species. Finally, we provide evidence that apparent rhythms in activity are the result of a combination of endogenous, light-entrainable oscillation and simple responses to environmental cues in both the surface morphotype and the
Pachón cave population of Astyanax mexicanus.
MATERIALS AND METHODS
Fish and animal husbandry
Fish used in these studies were part of a laboratory population maintained at the
University of Cincinnati. All specimens were laboratory-bred Astyanax mexicanus belonging to lines originally sourced from the Sierra de El Abra region of northeastern Mexico. Cavefish used in this study are members of a lab-cultured line originally sourced from the Pachón cave population. Fish used in 72hr trials were bred on-site and were 6-12 months old and of similar size when assayed.
All fish were kept in a dedicated animal room maintained at 21.7 ± 1˚C under a 12:12hr light/dark cycle (~160 lux/0 lux). According to convention, zeitgeber time (ZT) 00:00 corresponds to the time that lights turn on in the morning, in this case 07:00 (EST). Surface and cave individuals used in this study were housed in 5-gallon aquaria on our animal husbandry system as described elsewhere (Gross et al. 2013). Fish were fed TetraMin Tropical Flakes
(Tetra) once per day between ZT 01:00 and ZT 03:00.
All husbandry procedures and experimental protocols were approved by the Institutional
Animal Care and Use Committee (IACUC) of the University of Cincinnati (Protocol Number 10-
01-21-01).
13! ! Rearing and trial conditions
A month in advance of the commencement of 72hr trials, 12:12hr light/dark surface and cavefish were moved to the top rack of our husbandry system, their tanks adjacent to one another and located 0.5m below a fluorescent lighting fixture in the ceiling. Dark-reared fish were housed in glass finger bowls covered by boxes from the morning after fertilization until large enough to be moved onto the husbandry system. Once on the system, these fish were kept on the bottom shelf in tanks that had been blacked out using black posterboard secured to the sides and lids of 5-gallon glass aquaria. Several layers of opaque laboratory tape were used to secure the posterboard and prevent light from entering at the seams. Posterboard on aquarium lids was first wrapped in clear packing tape to keep it dry, then affixed to the plastic lids. Tanks were placed next to each other on the bottom shelf of our husbandry system. Wherever possible, measures were taken to limit exposure of these specimens to light; sorting, cleaning and feeding of embryos and very young fish were performed quickly and under very dim light, and blackout tanks were only opened for maintenance or to retrieve fish for assay in total darkness, aided by night vision goggles (Night Cougar LT, American Technologies Network Corp.). However, due to the flow-through nature of our husbandry system, it was necessary that two small (2.5cm diameter) holes remained open in the tops of these blackout tanks to facilitate filtration and feeding.
12:12hr light/dark trials were conducted in the main area of the animal room on the same lighting schedule under which the animals had been maintained prior to being assayed. During the “light phase” the room was lit by several fluorescent fixtures in the ceiling (40W Philips
Colortone 50 fluorescent bulbs), with the nearest fixture located approximately 2.5m away from the trial tanks. Trials in constant darkness were conducted in a separate area in which lighting
14! ! could be independently controlled. Trials involving fish reared in blackout tanks were set up in total darkness, aided by night vision goggles, to ensure that these fish were not exposed to any light during this process. All trials were conducted at room temperature on the same table using the same experimental apparatus (described below).
Fish to be used in trials were fed in the morning prior to acclimation as normal. Prior to trial commencement, fish were transferred to the plastic assay tanks (described below) and allowed to acclimate. Fish were not fed at any point during trials in order to avoid potential issues with video tracking and to ensure that the activity patterns observed were not distorted by behaviors associated with feeding.
Automated video tracking
Fish were placed in small plastic tanks (10.1cm wide x 10.1cm deep x 12.6cm high) made of clear plastic and were filled with ~1.05L of system water. White contact paper was used to cover the back, base and sides to evenly distribute the IR lighting (see below), but also served to prevent fish in adjacent tanks from interacting with one another. Tanks were arrayed two high and three or four across, allowing between six and eight fish to be assayed simultaneously; each fish was used only once per experiment. Shallow channels were cut just above the waterline on the sides of designated “bottom” tanks to ensure that air was free to flow over the surface of the water in those tanks.
Automated video tracking was conducted using an experimental rig set up on a static, level table. The system consisted of three IR LED arrays (IR-ROOM Ultra Covert IR
Illuminator, NightVisionExperts.com) situated at one end of the table and used to evenly illuminate the trial tanks placed ~1.05m away, near the opposite end of the table. An ultra (IR)
15! ! sensitive monochrome camera (Ikegami Model ICD-49; Ikegami Tsushinki Co.) with a manual iris veri-focal lens (Computar H3Z4512CS-IR 1/2" Varifocal Day/Night Lens 4.5-12.5mm; CBC
Group) equipped with an IR filter (Heliopan E 35.5 RG 850, Heliopan Lichtfilter-Technik
Summer) was placed ~0.65m from the front (i.e., non-covered side) of the trial tanks. Images recorded by the camera were sent to a digitizing board in a nearby computer and subsequently fed into the EthoVision XT program (v.8.0.516, Noldus Information Technology) for analysis.
Both the results of the analysis and the digitized video files were saved to an external hard drive.
The IR illumination served to backlight the fish being assayed, resulting in the fish appearing as a dark spot contrasted against a light background. This allowed the Ethovision XT software to easily identify the fish within a given frame, define a center point and track the movement of that center point over time as the fish moved within a user-defined “arena” (i.e., the area inside a given trial tank that is filled with water). The superimposition of “zones” over this arena allowed the program to report how much time a specimen spent in a given portion of the tank, in this case the top, middle or bottom third. In this study, frames of video were sampled at a rate of 15 samples per second using the dynamic subtraction method. Sampling was triggered automatically at a specified time (ZT 12:00 for all 72hr trials) and continued until the specified trial length had elapsed. Raw data was binned at 15min intervals and used to report each specimen’s mean velocity (cm/s) over that span, as well as how much time (in seconds; one data point spans 900 seconds) the specimen spent in each of the three zones (top, middle and bottom); reported “usage” of a particular zone refers to the latter type of metric. All descriptive statistics used in this study were subsequently calculated based on this data.
16! ! Descriptive statistics and rhythm analysis
To facilitate quantitative comparison of assay data, it was necessary to develop a set of metrics to capture and describe the major features and potential rhythms observed in sets of individuals assayed under similar trial conditions. Three of the metrics (mean for the entire trial, subjective day mean and subjective night mean) were used to describe velocity and zone usage over the course of a trial as a whole, or specifically during periods when the lights were on or off
(or the corresponding periods of trials conducted in constant darkness; ZT 00:00-12:00 and ZT
12:00-00:00, respectively).
The remaining metrics used were based on cosinor rhythmometry, which involves fitting a cosine wave to a data set and using features of the cosine function (e.g., amplitude, period, midline and acrophase) to describe the rhythms that might be present. This was accomplished using a spreadsheet-based method that makes use of the solver feature in Microsoft Excel (v.
14.0.0; Microsoft Corporation), adapted from methods described and validated elsewhere
(Bourdon et al. 1995). Briefly, the cosine function is described using the equation
! ! = !! ×!cos !!!! + ! + !! ! where a = amplitude, b = period (i.e., 24hr), c = (acro)phase and d = midline estimated statistic of rhythm (mesor), or the midline around which the function oscillates. Setting the initial values of variables a, c and d to zero and setting the value of variable b to 24 (i.e., a 24hr period), the solver function is used to adjust the values of the first three variables in order to minimize the residual sum of squares (RSS) obtained by comparing the values obtained using the function to the observed data at each point. Holding b constant at a value of 24 forces the solver to fit a cosine function that describes the data with the assumption that the data contains a circadian rhythm with a period of 24hrs. In cases where this strategy returned a negative value for
17! ! amplitude and/or the phase value represented the position of the bathyphase (i.e., trough), rather than the acrophase (i.e., peak) of the cosine function, values were adjusted as described elsewhere (Bourdon et al. 1995). “Percent rhythm” values serve as a general description of how a given cosine function accounts for the variation seen around the mean of a given data set.
Percent rhythm is similar to an r2 value for the fit cosine and was calculated using
!"" ! = ! 1 −! !"" !×!100% where TSS = the total sum of squares for the data set when compared to the mean.
Comparisons of velocity and tank usage between morphotypes within experiments were made using Wilcoxon-Mann-Whitney tests. Dwass-Steel-Critchlow-Fligner post-hoc analyses following Kruskal-Wallis tests were used to compare values within morphotypes between experiments; JMP Pro 11 (SAS Institute Inc., Cary, NC) was used to conduct these analyses.
Mesor, amplitude and acrophase values were compared using F-tests based on methods described by Nelson et al. (1979). Circadian rhythmicity was assessed using zero-amplitude tests as described in Bourdon et al. (1995) and autocorrelation analyses following methods outlined by Pecar (In preparation).
RESULTS
Astyanax cavefish demonstrate higher overall swimming velocity compared to surface fish irrespective of lighting conditions
To facilitate the characterization and comparison of the activity phenotypes displayed by the Astyanax mexicanus cave and surface morphs, cave (Pachón; n = 16) and surface fish (n =
20) were assayed using a purpose-built automated video tracking assay rig (Fig. 2.1). Initial comparisons were made based on velocity data collected over 72hrs under our standard
18! ! laboratory lighting regimen (12:12hr light/dark cycle). The cavefish demonstrated a higher overall level of activity relative to surface fish throughout the 72hr trial period (Fig. 2.2a,c).
Comparison of mean velocity data (cm/s; Table 2.1) confirmed that these differences were significant regardless of whether the trial was considered as a whole (z = 4.600, p < 0.0001), or data from subjective day (ZT 00:00-12:00; z = 5.078, p < 0.0001) and subjective night (ZT
12:00-00:00; z = 3.677, p = 0.0002) phases of the light cycle were evaluated separately.
In order to determine whether these differences persist under “cave-like” conditions, cave
(Pachón; n = 17) and surface fish (n = 18) reared and housed under our standard lab lighting regimen were assayed for 72hr under constant darkness. The transition to constant darkness was timed to coincide with the beginning of the dark phase of the cycle in order to facilitate direct comparison of discrete portions of the 72hr trial with corresponding data from the light and dark phases of the 12:12hr light/dark trials. Under these conditions, the cavefish again display a higher level of activity throughout the trial period (Fig. 2.2b,d). As in the previous set of trials, mean velocity was significantly higher whether the trial was considered as a whole (z = 3.746, p
= 0.0002), or subjective day (ZT 00:00-12:00; z = 3.713, p = 0.0002) and subjective night (ZT
12:00-00:00; z = 3.977, p < 0.0001) were examined separately. When mean velocity data were compared within morphotypes between the two lighting conditions, no significant differences were detected in either morphotype.
Despite the absence of a visual system, cavefish demonstrate a robust response to illumination
In order characterize the apparent rhythms present in the velocity data obtained and to assess whether or not it would be proper to define them as “circadian” (i.e., having a period of
~24hr), cosine functions with a period of 24h were fit to detrended velocity data for each set of
19! ! trials. The midline estimated statistic of rhythm (mesor), amplitude and acrophase (i.e., the position of the peak amplitude) describe the cosine function fit to each data set (Table 2.2).
Under 12:12hr light/dark conditions, both the surface and cave fish showed significant circadian rhythmicity when subjected to zero-amplitude testing (F2,285 = 56.584, p < 0.0001; F2,285 =
169.656, p < 0.0001, respectively) and time lag autocorrelation (z = 5.664, p < 0.0001; z = 6.419, p < 0.0001). Despite the absence of a visual system, the cavefish display a robust response to lighting stimulus, displaying a rhythmic pattern of activity that differs from surface fish assayed under identical conditions not only in terms of amplitude (F1,570 = 27.160, p < 0.0001), but also polarity. The acrophase for the cavefish rhythm differs significantly from that exhibited by their surface counterparts (F1,570 = 24.075, p < 0.0001), with the cavefish acrophase (-108.8˚) corresponding to ZT 06:55, just after the midpoint of the subjective day; the surface fish acrophase (-252.7˚) corresponds to ZT 16:51, just over an hour before the midpoint of subjective night.
Under constant darkness, both morphotypes retain a significant level of circadian rhythmicity (F2,285 = 43.237, p < 0.0001; z = 2.566, p = 0.0103 and F2,285 = 145.710, p < 0.0001; z = 6.754, p < 0.0001 for surface and cavefish, respectively). While the cavefish mesor and amplitude remain higher than those of the surface fish and their acrophases remain significantly different (F1,570 = 10044.559, p < 0.0001; F1,570 = 16.747, p < 0.0001; F1,570 = 5.903, p = 0.0154 for mesor, amplitude and acrophase, respectively), surface fish under constant darkness exhibit a
118.1˚ shift in acrophase, relative that observed under 12:12hr light/dark conditions (F1,570 =
46.304, p < 0.0001). This 7.88hr shift changes the timing of peak velocity data to ZT 08:59, which falls during subjective day, rather than subjective night. As a result, the timing of this peak is more similar to that seen in cavefish under either lighting regimen (ZT 06:55 and ZT 07:52 for
20! ! 12:12hr light/dark and constant darkness, respectively) than to that seen in surface fish under
12:12hr light/dark conditions (ZT 16:51).
Surface fish reared in darkness from birth demonstrate activity patterns that resemble cavefish
In order to ascertain the effects, if any, of development and housing under “cave-like” conditions prior to assay, surface (n = 17) and cavefish (n = 18) that had been kept in darkness from very early development onward were assayed under constant darkness (Fig. 2.3).
Comparison of mean velocity data under these conditions between dark-reared surface and cavefish differed significantly only when comparing values during subjective night (ZT 12:00-
00:00), during which cavefish exhibited higher velocity than the surface morphotype (z = -2.195, p = 0.0282). While there were no differences between cavefish data under any of the trial conditions, there were differences within the surface morphotype. Dwass-Steel-Critchlow-
Fligner post-hoc analyses indicate that mean velocity was significantly higher for dark-reared surface fish in constant darkness than either of the other two experiments when considering data for either the trial as a whole or subjective day (all p ≤ 0.0215); there were no significant differences between experiments for values during subjective night.
Unlike cavefish reared under 12:12hr light/dark conditions, velocity data for cavefish reared and assayed in constant darkness does not show a significant level of circadian rhythmicity. In contrast, surface fish under these conditions show strong circadian rhythmicity
(F2,285 = 359.125, p < 0.0001; z = 8.473, p < 0.0001), with a higher mesor and amplitude than those seen in fish reared under light/dark conditions and assayed under either a 12:12hr light/dark cycle (F1,570 = 1785.768, p < 0.0001; F1,570 = 56.646, p < 0.0001 for mesor and amplitude, respectively) or constant darkness (F1,570 = 1445.473, p < 0.0001; F1,570 = 123.002, p
21! ! < 0.0001 for mesor and amplitude, respectively). While the acrophase (-114.3˚) of the cosine function fit to the velocity data for these dark-reared surface fish differed significantly from that seen in fish reared under a light/dark cycle and then assayed under constant darkness (F1,570 =
10.094, p = 0.0016), the timing of this peak of activity (ZT 07:37) similarly falls during subjective day.
Notably, upon visual comparison the activity profile of dark-reared surface fish appears to show a higher degree of similarity with that of cavefish reared under a light/dark cycle.
Comparison of the rhythms characterized in these trials revealed that, while the rhythms seen in dark-reared surface fish are significantly different from those seen in other surface fish data (as discussed above), the amplitude of this rhythm is not statistically distinguishable from that of cavefish raised and assayed under a 12:12hr light/dark cycle (F1,570 = 3.014, p = 0.0831), nor is the acrophase distinguishable from that of cavefish reared under a light/dark cycle and assayed under constant darkness (F1,570 = 0.825, p = 0.3641).
Cave and surface morphs demonstrate dramatically different patterns of tank usage
In order to examine whether patterns in activity extend to spatial usage of the trial tank, the amount of time that fish spent in each third of the tank (top, middle and bottom “zones”) was also examined (Fig. 2.4; Tables 2.3-2.5). Under 12:12hr light/dark conditions, cavefish spent significantly more time in the top and middle zones (z = 4.696, p < 0.0001; z = 3.295, p =
0.0010, respectively) when compared to surface fish, which spent more time than cavefish in the bottom zone (t34 = -9.702, p < 0.0001). This pattern held when data for subjective day and night were considered separately (all p ≤ 0.0004), with the exception that during subjective night, time
22! ! spent in the middle zone was no longer statistically distinguishable between morphotypes (z =
1.353, p = 0.1760).
Under constant darkness, a similar pattern emerged. Cavefish again spent significantly more time than surface fish in the top zone throughout the trial period (z = 4.373, p < 0.0001) and when subjective day and night were considered separately (z = 4.307, p < 0.0001; z = 4.835, p < 0.0001, respectively). Similarly, surface fish spent significantly more time than cavefish in the bottom zone throughout the trial, as well as specifically during subjective day and night when considered separately (z = -2.888, p = 0.0039; z = -2.129, p = 0.0333 and z = -3.317; p = 0.0009, respectively). However, differences in time spent in the middle zone were no longer significant using any of these three metrics (all p ≥ 0.6799).
Data for dark-reared fish reared under constant darkness yielded results quite different from those seen in other trials (Fig. 2.5). Under these conditions, there was no longer a significant difference between the time spent in the top zone by the cave and surface morphotypes, regardless of whether the trials were considered as a whole or subjective day and night were evaluated separately (all p ≥ 0.1923). However, surface fish spent significantly more time in the middle zone than cavefish, during the trial as a whole (z = 3.878, p = 0.0001) and during both subjective day and night when considered separately (z = 3.647, p = 0.0003; z =
3.977, p < 0.0001, respectively). Cavefish may have spent more time in the bottom zone than surface fish, however this just barely reaches significance and only during subjective day (z =
-1.963, p = 0.0496).
Data for all comparisons between experiments within morphotypes can be found in the accompanying tables (Tables 2.3-2.5), however, there are several salient trends worth noting, including: 1) Cavefish differ very little between experiments, with the exception that they tend to
23! ! use the bottom zone less and the top zone more during subjective day under constant darkness, relative to the results seen under 12:12hr light/dark conditions. 2) There were no significant differences between light/dark-reared and dark-reared cavefish under constant darkness. 3)
Under constant darkness, light/dark-reared fish used the top and middle zones more and the bottom zone less, relative to light/dark-reared fish under 12:12hr conditions, over the trial as a whole and during subjective day; dark-reared surface fish showed the same pattern, with the exception that significant differences were seen during subjective night as well. 4) Under constant darkness, surface fish reared in darkness spent significantly more time in the top zone and less time in the bottom zone than light/dark-reared surface fish; there were no differences in usage of the middle zone.
Rhythmic patterns of tank usage are altered or disappear entirely under constant darkness
Under 12:12hr light/dark conditions, surface fish show robust patterns in tank usage for both the top (F2,285 = 99.157, p < 0.0001; z = 9.555, p < 0.0001) and bottom zones (F2,285 =
104.204, p < 0.0001; z = 10.110, p < 0.0001; Tables 2.6 and 2.7). Cavefish also demonstrate significant rhythmicity in their usage of the top (F2,285 = 247.677, p < 0.0001; z = 8.343, p <
0.0001) and bottom zones (F2,285 = 240.148, p < 0.0001; z = 8.379, p < 0.0001) under these conditions. Rhythms in cavefish differ in amplitude (F1,570 = 72.997, p < 0.0001; F1,570 =
132.661, p < 0.0001 for top and bottom zones, respectively) and mesor (F1,570 = 13336.377, p <
0.0001; F1,570 = 6529.019, p < 0.0001 for top and bottom zones, respectively), as expected.
However, their acrophases, although different (F1,570 = 4.057, p = 0.0444; F1,570 = 9.063, p =
0.0027 for top and bottom zones, respectively), are similarly timed; rhythms in top zone usage
24! ! reach their respective peaks during subjective night, while rhythms in bottom zone usage reach their peaks during subjective night.
When light/dark-reared fish were assayed under constant darkness, no evidence of significant circadian rhythmicity was observed in cavefish. Surface fish retained rhythmicity in their usage of both the top (F2,285 = 58.958, p < 0.0001; z = 2.802, p = 0.0051) and bottom (F2,285
= 104.204, p < 0.0001; z = 3.960, p < 0.0001) zones, but the rhythms were significantly altered.
The amplitude of the rhythm describing usage of the bottom zone decreased significantly (F1,570
= 217.437, p < 0.0001) and the mesors for top and bottom data increased and decreased, respectively (F1,570 = 751.874, p < 0.0001 and F1,570 = 2281.771, p < 0.0001). The most notable change, however, is that the acrophase of the rhythms in usage of these two zones each shifted by ~180˚ (F1,570 = 246.991, p < 0.0001 and F1,570 = 108.652, p < 0.0001 for top and bottom, respectively), such that the inverse relationship between the two rhythms was maintained, but the peak for top zone activity coincided with subjective day and the peak for bottom zone activity occurred during subjective night.
No evidence of significant rhythmicity in patterns of tank usage was seen in dark-reared fish of either morphotype when assayed under constant darkness.
DISCUSSION
Light/dark-reared cavefish display a higher level of activity than light/dark-reared surface fish
In an early study, Breder and Gresser (1941a) described the behavior of surface fish as
“quiescent” unless feeding or responding to some evident external stimulus. Conversely, Erckens and Martin (1982b) described “permanent swimming” in Pachón, though differences in swimming speed were observable. Therefore, it is not surprising that under both a 12:12hr
25! ! light/dark cycle and under constant darkness, light/dark-reared cavefish had a higher level of activity than surface fish, as reflected by differences in mean velocity data. However, this pattern did not hold when cave and surface fish reared and assayed under constant darkness were compared; only during subjective night (ZT 12:00-00:00) was cavefish velocity significantly higher. Interestingly, while velocity data for cavefish does not differ significantly between experiments, both the mean trial velocity and the mean for subjective day (ZT 00:00-12:00) were higher in dark-reared surface fish than in light/dark-reared surface fish under either lighting regimen. It is possible, therefore, that something is causing the surface fish to act more “cave- like” under these conditions, thereby eliminating the difference in activity seen in other trials.
Studies by Duboué et al. (2011) and Yoshizawa et al. (2015) suggest that the higher mean velocity values data may reflect differences in the number and/or duration of sleep bouts between cave and surface fish rather than actual differences in velocity while swimming. While we did not assess sleep in this study and can therefore neither confirm nor rule out sleep as a factor, our data provide no evidence to the contrary. However, the fact that the observed differences in velocity data break down when dark-reared fish are assayed under constant darkness would seem to indicate that 1) rearing surface fish under constant darkness somehow results in fewer and/or shorter sleep bouts, 2) dark-reared surface fish swim faster, on average, than those reared under light/dark conditions or 3) some other factor influencing locomotor activity is at play. This possibility warrants further investigation, as any of these scenarios, or combinations thereof, would provide relevant insights into the behavior of this species.
26! ! Cavefish retain the ability to perceive and integrate photic cues
The results of this study add to the body of research suggesting that Astyanax cavefish retain the ability to both sense light and to alter their locomotor activity in response to photic stimulus. This was not surprising, given that Wilkens (1988) reported that some evidence of the ability to sense light had been found in all Astyanax cavefish populations tested. While the conditions under which this study was conducted do not allow us to definitively eliminate all other possibilities, changes in cavefish activity correspond very closely with changes in light state for both mean velocity and tank usage data. These activity profiles are also consistent with those seen under a 12:12hr light/dark regimen in previously published studies where the authors reached similar conclusions (Erckens and Weber 1976; Erckens and Martin 1982b; Beale et al.
2013; Yoshizawa et al. 2015).
Exactly which photosensitive organ(s) and/or cell populations may be mediating this response and the extent to which each influences these responses has yet to be conclusively determined. For example, studies in this species have shown that adult surface fish show a strongly photophobic response to light (Breder and Rasquin 1947; Romero 1985; Kowalko et al.
2013). Pachón cavefish also display some level of photophobic behavior when provided a choice between a light and a dark portion of a trial tank. However, in order to achieve an attenuated level of negative phototaxis comparable to that seen in the cave population in surface fish, it is necessary to remove both the eyes and the pineal gland (Langecker 1989). Further, inducing the development of eyes in cavefish is not sufficient to rescue the surface phenotype (Romero et al.
2003). Taken together, these results would seem to suggest, as Wilkens (1988) noted, that extraocular, extrapineal photoreceptors do not seem to have regressed in Astyanax and that they play a roll in the light response displayed by both cave and surface fish. They also suggest that 1)
27! ! cavefish are not relying on whatever rudiments of the regressed eye may remain in order to help mediate that response and 2) the photoreceptors in the cavefish pineal gland may not play an important role in triggering the photophobic behavior observed. However, the work of
Yoshizawa and Jeffery (2008) seems to contradict the latter possibility. They showed that the pineal “eye” is indeed functional in very young (≤ 7.5 days post fertilization) Pachón cavefish and demonstrated that the pineal gland is responsible for mediating the shadow response seen in these larvae. This matter is further complicated by the fact that responses to light differ between populations (Breder and Rasquin 1947), they appear to change with growth/age (Romero 1985) and may be influenced by the lighting conditions under which fish are kept during development
(Omura 1975; Herwig 1976; Tabata 1982). Further research examining these possibilities in greater detail is warranted and promises to provide much needed clarification of the factors involved in the cavefish response to light.
Cavefish retain a light-entrainable circadian oscillator capable of sustaining behavioral rhythms
Previous studies in Astyanax have suggested that locomotor activity is controlled by a passive system that can be entrained by exposure to cyclic photic cues (Thines and Wolff-Van
Ermengem 1965; Thines and Weyers 1978; Erckens and Martin 1982a; Erckens and Martin
1982b; Duboué et al. 2011; Beale et al. 2013; Caballero-Hernández et al. 2015). This conclusion is further supported by the fact that the results of this study demonstrated significant rhythms in mean velocity and usage of both the bottom and the top portion of the trial tank in both surface and cavefish under l2:12hr light/dark conditions. However, some debate remains as to whether or not cavefish retain a self-sustaining circadian oscillator similar to that seen in the surface fish.
Early studies by Thines and Wolff-Van Ermengem (1965), Thines and Weyers (1978) and a
28! ! much more recent study by Beale et al. (2013) all failed to detect any significant circadian rhythmicity in locomotor activity data for cavefish under constant darkness. However, in all but the last case, it is unclear what population the cavefish assayed represented and whether or not they were ever exposed to a light/dark cycle prior to being assayed. In contrast, Erckens and
Martin (1982b) detected significant rhythms in activity near the top and/or bottom of the tank that persisted after Pachón cavefish were switched from a 12:12hr light/dark cycle to constant darkness.
The assays described here support the conclusions reached by Erckens and Martin
(1982b). While the current study shows no evidence of entrained patterns of spatial tank usage in cavefish, significant circadian rhythmicity was detected in the velocity data for both morphotypes after fish were switched from a 12:12 light/dark cycle to constant darkness.
Although this study and previous work offer both complementary and conflicting results, it is likely that much of this can be attributed to a combination of differences in experimental design
(e.g., assaying single fish versus groups of fish, analyzing individual data versus mean data, etc.), assay techniques (e.g., employing infrared sensors versus video analysis, differing approaches to data normalization, etc.) and natural variation between assayed individuals. For example, while the design of the current study included automated measurement of mean velocity and mean time spent in each zone, Erckens and Martin (1982a; 1982b) collected no velocity data and time spent in each part of the tank was measured by video analysis for only one specimen per morphotype; all other data was collected using infrared sensors. Given that tank usage data cannot be decoupled from velocity in that study, it is entirely possible that the results of the current study reflect the same biological phenomenon observed by Erckens and Martin, but captured using different metrics.
29! ! Self-sustaining circadian oscillation is unlikely to operate in the natural cave environment
While data for light/dark-reared cavefish under constant darkness suggests that Pachón fish retain a weakly-entrainable, self-sustaining circadian oscillator that is tied to locomotor activity, failure to detect any significant rhythmicity in dark-reared cavefish assayed under constant darkness suggests that such coordinated rhythmicity, though inducible in the lab, would likely not be present in wild populations. Failure to exhibit a discernable rhythm under these conditions could be the result of either a lack of oscillation or a lack of coordination of the phase of oscillation between the individuals comprising a given population. Beale et al. (2013) addressed both of these possibilities within the context of gene expression within the core clock mechanism of the Chica cavefish. They concluded that, not only were non-photic environmental cues (e.g., rhythmic bat activity) insufficient to entrain circadian rhythmicity, but that the cave environment appeared to suppress core clock function to an even greater extent than exposure to constant darkness in the lab. Although that study did not include any analysis conducted in the
Pachón cave, similarity of lab-based results between the two populations and the fact that the
Pachón cave environment includes even fewer potential zeitgebers than the Chica cave led Beale et al. (2013) to assert that their conclusions are likely generalizable to the Pachón population.
The results presented here, taken together with the existing literature, support the conclusion reached by Erckens and Martin (1982b) that the weakly-entrainable self-sustaining oscillator underlying the weak circadian rhythmicity of locomotor activity seen in cavefish exposed to
“surface-like” lighting conditions, while noteworthy, is likely an evolutionary relic.
30! ! Patterns in cavefish tank usage are likely driven by exogenous cues, rather than endogenous rhythms
Looking beyond the presence or absence of circadian rhythmicity in the data obtained, the results of this study suggest that other environmental and/or biological factors may be at play.
Observed rhythms in behavior may be the result of a simple, passive response to exogenous stimulus, rather than the output of an entrained, endogenous rhythm. A truly endogenous rhythm will persist under constant conditions with a period of ~24hr (Johnson 2004). While velocity data for cavefish exhibited sustained circadian rhythmicity under constant conditions in this study, no such rhythmicity was seen in top or bottom zone usage in cavefish under constant conditions.
Therefore, while the data for cavefish zone usage under 12:12hr light/dark conditions clearly exhibited circadian rhythmicity, there is no evidence indicating that these rhythms were anything more than passive responses to cyclic photic cues present in the lab setting.
Instead, this may be a simple manifestation of the photophobic behavior well documented in this species (Breder and Gresser 1941a; Breder and Gresser 1941b; Breder 1944; Breder and
Rasquin 1947; Kuhn and Kähling 1954; Lüling 1954; Kähling 1961; Gertychowa 1970; Erckens and Martin 1982b; Tabata 1982; Romero 1985; Romero et al. 2003; Kowalko et al. 2013). If spending extended periods of time in the bottom zone of the tank (and/or avoidance the top zone) can be viewed, at least in part, as photophobic behavior, it is reasonable that exposure to a
12:12hr light/dark lighting regimen would create both the cyclic patterns in tank usage data seen in both cave and surface fish, as well as the inverse relationship between the time spent in the top and bottom zones as described here and by Erckens and colleagues (1976; 1982a; 1982b).
Further, differences between the amplitude of the responses seen in the cave and surface
31! ! morphotype may reflect the differing roles that visual systems, the photosensitive pineal gland and extraocular, extrapineal photoreceptors play in mediating this response (Wilkens 1988).
Surface fish data shows evidence of putative masking effect under 12:12hr light/dark cycle
Surface fish are neither diurnal nor nocturnal in the wild, but crepuscular (Wilkens 1988), and peaks in activity similar to those reported by Thines and Wolff-Van Ermengem (1965) can be seen around “dawn” and “dusk” (i.e., the periods around lighting transitions) in the mean velocity data for surface fish under 12:12hr light/dark conditions. However, in contrast with recent studies where a diurnal pattern of activity has been reported for the Astyanax mexicanus surface morphotype under a 12:12hr light/dark cycle (Duboué et al. 2011; Beale et al. 2013;
Yoshizawa et al. 2015), the acrophase of the cosine function fit to the data presented here was indicative of a more nocturnal pattern of behavior overall. Given that studies by Duboué et al.
(2011) and Yoshizawa et al. (2015) indicate that surface fish sleep more during the dark period of the light/dark cycle, this result is suspect. Further, when photic stimulus was not present, the acrophase of mean velocity rhythms in surface fish shifted to subjective day (ZT 00:00-12:00), regardless of whether they were reared in light/dark conditions or constant darkness. Similar results were seen in tank usage data, in which the potentially endogenous rhythms exhibited by surface fish under constant darkness were different in both intensity and polarity (i.e., amplitude and acrophase) from those exhibited under a 12:12hr light/dark cycle.
Taken together, this may be evidence of deviation from endogenous rhythms mediated by a separate behavioral response to the lighting stimulus provided, a response that bypasses the circadian clock entirely; this phenomenon referred to as “masking” (Takahashi and Zatz 1982;
Mrosovsky 1999). Masking can be either positive or negative, depending upon whether the
32! ! masking response increases or decreases the amplitude of observed rhythms in behavior. While the masking mechanism is complementary to the circadian entrainment mechanism, they represent two behaviorally and physiologically distinct mechanisms, with the entrainment mechanism mediating synchronization of endogenous clocks and the masking mechanism mediating acute responses to stimulus, usually light (Mrosovsky et al. 2001). If present, such negative masking both implicates the photophobic behavior previously discussed and suggests that the response is strong enough to obscure or even alter normal circadian rhythmicity in locomotor activity. Additionally, the process of shifting back to an endogenous, free-running circadian rhythm of locomotor activity when the lighting stimulus is removed may also account for other features of the data obtained. For example, this may explain why mean velocity data for light/dark-reared fish switched to constant darkness does not show a significant level of autocorrelation when a 24hr lag is applied, but does show a significant autocorrelation when a
48hr lag is applied.
Evaluation of zeitgebers potentially influencing circadian rhythmicity in dark-reared surface fish
While dark-reared cavefish do not show rhythmicity in any metrics used in this study under constant darkness, dark-reared surface fish exhibited strong, significant circadian rhythmicity in mean velocity data under the same conditions. This provides support for the existence of an endogenous, free-running circadian oscillator linked to locomotor activity in this species. However, the persistence and strength of the rhythmicity exhibited by dark-reared surface fish also calls into question what environmental cue(s), if any, may be responsible for entraining these rhythms and/or synchronizing free-running rhythms among members of this experimental group.
33! ! An effort was made to greatly reduce if not entirely eliminate exposure of dark-reared fish to sources of light at any point during their lives prior to the trials presented here. Despite this, the behavioral patterns seen in these fish may have been entrained by a very small amount of light entering their blackout tanks during the light phases of the normal 12:12hr light/dark schedule. Erckens and Martin (1982a) reported that surface fish were entrainable by light/dark cycles in which illumination during the light phase was 1 lux, or about the maximum intensity of light given off by a full moon on a clear night (Bünning and Moser 1969). This level of light was not sufficient to entrain sustained rhythms in Pachón cavefish behavior; however, and Erckens and Martin (1982b) conducted all subsequent trials in cavefish under light cycles with an illumination of 200 lux during the light phase. This difference could explain why rhythmicity was seen in dark-reared surface fish, but not dark-reared cavefish included in this study.
It is also possible that these fish were exposed to a pulse of light sufficient to synchronize endogenous rhythms that were already present in each individual sometime well before being placed in the blackout tanks. In fact, studies in zebrafish have shown that exposure to pulses of light just hours after fertilization are sufficient to synchronize clock-controlled circadian rhythms in the pineal gland during the third and fourth day of development (Ziv and Gothilf 2006).
Along with light cycles, temperature cues are often implicated in entrainment of circadian rhythms (Ki et al. 2015). For example, Thines and Weyers (1978) documented entrainment of
Astyanax cavefish (likely from the Chica cave population) to temperature cycles with an amplitude of 3˚C. In fact, it appears that cyclic thermal cues with a period of 24hr and amplitudes of as little as 1-2˚C can function as strong zeitgebers capable of entraining circadian rhythmicity in all ectothermic (poikilothermic) organisms (Rensing and Ruoff 2002).
34! ! Our animal husbandry facilities are temperature controlled, but it is possible that low amplitude shifts in temperature are causing or contributing to the results seen. However, given that low amplitude fluctuations in temperature are capable of entraining a wide variety of organisms and that sensitivity to these temperature shifts has been demonstrated in at least one population of Astyanax cavefish, it is unlikely that thermal cues such as these would produce a rhythm in dark-reared surface fish, but not dark-reared cavefish housed in the same room, on the same husbandry system.
The last of the principal zeitgebers that may be involved with entraining the rhythms seen in this study is the timing of food availability. Biswas and Ramteke (2008) showed that, under constant darkness, feeding at 18:00 was sufficient to entrain circadian rhythms in vertical swimming activity at the population level in the cave loach, Nemacheilus evezardi, with the acrophase of the rhythm occurring just after feeding time. Interestingly, however, a different experimental group fed at 06:00 as part of the same trial did not show evidence of entrained rhythms of swimming activity; the authors could not immediately account for this difference.
Similarly, Cavallari et al. (2011) demonstrated that the blind Somalian cavefish, Phreatichthys andruzzii, retains a food-entrainable oscillator with circadian period that drives rhythms in locomotor activity when synchronized by feeding at a specific time. This result was particularly significant in light of the fact that rhythms in locomotor activity in P. andruzzii cannot be entrained by cyclic photic cues. In Astyanax, Zafar and Morgan (1992) reported that cavefish were able to entrain rhythms in food-anticipatory locomotor activity when kept under constant conditions (3 lux, 24 ± 1˚C) and fed at the exact same time daily. These rhythms persisted with a period of approximately 24hr even after feeding was discontinued. Unfortunately, Zafar and
Morgan (1992) do not appear to have assayed surface fish to facilitate comparison between
35! ! morphotypes. Further, these studies were conducted in cavefish purchased from a commercial source and, although they are likely sourced from the Chica population, it is unwise to make any assumptions about which cave population(s) is/are represented in the genetic background of the fish assayed.
While feeding may have some degree of influence on the activity rhythms seen in this study, the failure of Beale et al. (2013) to detect a response to rhythmic deposition of bat guano
(a nutrient source) within the core genetic clockwork of the wild Chica cavefish population casts doubt upon the strength of feeding time as a zeitgeber in Astyanax. Moreover, the acrophases of rhythms observed under constant darkness in the present study do not seem to coincide with the usual feeding time like those reported by Zafar and Morgan (1992), Biswas and Ramteke (2008) or Cavallari et al. (2011), but rather occur several hours later. Therefore, it is unlikely that food- mediated entrainment is a major component of the apparent rhythmicity documented here.
Photic, thermal and feeding cues are the best-documented modes of entrainment of circadian rhythms of behavior in fish, but these are not the only exogenous factors potentially influencing the behavioral profiles seen in this study. Social cues, for example, have been shown to have an effect on circadian rhythmicity in fish, with fish housed in groups showing stronger rhythmicity than individual fish (Kavaliers 1980a; Kavaliers 1980b; Kavaliers 1981; Bolliet et al. 2001); however it remains unclear whether this is due to a simple increase in the chance of rhythm detection and no studies have demonstrated synchronization of activity rhythms in fish by social cues alone (Zhdanova and Reebs 2006). Additionally, electromagnetic fields have been shown to influence behavior and melatonin synthesis in fish, although it has yet to be determined whether these cues can entrain self-sustaining rhythms in the tested species (Lerchl et al. 1998;
Jain and Shedpure 2013; Lee and Yang 2014). Finally, and perhaps most interestingly, Delaunay
36! ! et al. (2000) demonstrated that rhythmic accumulation of Per3 transcripts occurs in zebrafish oocytes and persists in developing embryos, suggesting a possible method of circadian phase inheritance from mother to offspring. While none of these alternate zeitgebers have been evaluated in the current study, they present excellent opportunities for further research using the
Astyanax system.
CONCLUSIONS
We employed a high-resolution, automated video tracking assay method to assess locomotor rhythms in Astyanax mexicanus. Cavefish exhibited responsiveness to light, as well as a higher level of activity and differences in spatial usage, relative to surface fish. Both surface and cavefish displayed significant circadian rhythmicity in both velocity and tank usage; however, only patterns in velocity appear to be under the control of a light-entrainable, endogenous circadian oscillator like that hypothesized by Erckens and Martin (1982b). We also demonstrated evidence of masking effects similar, but not equivalent, to those previously suggested by Beale et al. (2013). Finally, studies in dark-reared specimens suggest that the lighting conditions under which fish develop and/or are housed prior to assay may have lasting effect on the patterns of activity that they exhibit. The results presented in this study represent an important contribution to our understanding of circadian rhythms in Astyanax and provide insights that promise to inform the design and execution of future work in this species. Here we have largely limited our discussion to issues concerning the interaction of behavior and environmental cues; potential genetic basis and evolutionary rationale for changes in activity and tank usage will be addressed elsewhere (see Chapter 4).
37! !
ACKNOWLEDGEMENTS
The authors would like to thank Wendy Lu and Tyler Bussian for their invaluable assistance with data analysis, as well as Jeremy Didion for assistance with measuring light intensity. Julie Carlson and Rachel Schafer provided helpful edits and suggestions. We would also like to thank Amanda Powers, Bethany Stahl and other past and present members of the
Gross Lab for their assistance, ideas and advice.
REFERENCES
Beale, A., C. Guibal, T. K. Tamai, L. Klotz, S. Cowen et al., 2013 Circadian rhythms in Mexican
blind cavefish Astyanax mexicanus in the lab and in the field. Nat Commun 4: 2769.
Bell-Pedersen, D., V. M. Cassone, D. J. Earnest, S. S. Golden, P. E. Hardin et al., 2005
Circadian rhythms from multiple oscillators: Lessons from diverse organisms. Nat Rev
Genet 6: 544-556.
Biswas, J., and A. K. Ramteke, 2008 Timed feeding synchronizes circadian rhythm in vertical
swimming activity in cave loach, Nemacheilus evezardi. Biological Rhythm Research 39:
405-412.
Bolliet, V., A. Aranda and T. Boujard, 2001 Demand-feeding rhythm in rainbow trout and
European catfish: Synchronisation by photoperiod and food availability. Physiol Behav
73: 625-633.
Boujard, T., and J. Leatherland, 1992 Circadian rhythms and feeding time in fishes. Environ Biol
Fishes 35: 109-131.
38! ! Bourdon, L., A. Buguet, M. Cucherat and M. W. Radomski, 1995 Use of a spreadsheet program
for circadian analysis of biological/physiological data. Aviat Space Envir Md 66: 787-
791.
Breder, C. M., 1944 Ocular anatomy and light sensitivity studies on the blind fish from Cueva de
los Sabinos, Mexico. Zoologica 29: 131-144.
Breder, C. M., and E. B. Gresser, 1941a Correlations between structural eye defects and behavior
in the Mexican blind characin. Zoologica 26: 123-131.
Breder, C. M., and E. B. Gresser, 1941b Further studies on the light sensitivity and behavior of
the Mexican blind characin. Zoologica 26: 289-296.
Breder, C. M., and P. Rasquin, 1947 Comparative studies in the light sensitivity of blind
characins from a series of Mexican caves. B Am Mus Nat Hist 89: 325-351.
Bünning, E., and I. Moser, 1969 Interference of moonlight with the photoperiodic measurement
of time by plants, and their adaptive reaction. Proc Natl Acad Sci U S A 62: 1018-1022.
Caballero-Hernández, O., M. Hernández-Patricio, I. Sigala-Regalado, J. B. Morales-Malacara
and M. Miranda-Anaya, 2015 Circadian rhythms and photic entrainment of swimming
activity in cave-dwelling fish Astyanax mexicanus (Actinopterygii: Characidae), from El
Sotano La Tinaja, San Luis Potosi, Mexico. Biol Rhythm Res 46: 579-586.
Cavallari, N., E. Frigato, D. Vallone, N. Frohlich, J. F. Lopez-Olmeda et al., 2011 A blind
circadian clock in cavefish reveals that opsins mediate peripheral clock photoreception.
PLoS Biol 9: e1001142.
Delaunay, F., C. Thisse, O. Marchand, V. Laudet and B. Thisse, 2000 An inherited functional
circadian clock in zebrafish embryos. Science 289: 297-300.
39! ! Duboué, E. R., and R. L. Borowsky, 2012 Altered rest-activity patterns evolve via circadian
independent mechanisms in cave adapted balitorid loaches. PLoS One 7: e30868.
Duboué, E. R., A. C. Keene and R. L. Borowsky, 2011 Evolutionary convergence on sleep loss
in cavefish populations. Curr Biol 21: 671-676.
Erckens, W., and W. Martin, 1982a Exogenous and endogenous control of swimming activity in
Astyanax mexicanus (Characidae, Pisces) by direct light response and by a circadian
oscillator I. Analyses of the time-control systems of an epigean river population. Z
Naturforsch 37c: 1253-1265.
Erckens, W., and W. Martin, 1982b Exogenous and endogenous control of swimming activity in
Astyanax mexicanus (Characidae, Pisces) by direct light response and by a circadian
oscillator II. Features of time-controlled behaviour of a cave population and their
comparison to an epigean ancestral form. Z Naturforsch 37c.
Erckens, W., and F. Weber, 1976 Rudiments of an ability for time measurement in the
cavernicolous fish Anoptichthys jordani Hubbs and Innes (Pisces, Characidae).
Experientia 32: 1297-1299.
Espinasa, L., and W. R. Jeffery, 2006 Conservation of retinal circadian rhythms during cavefish
eye degeneration. Evol Dev 8: 16-22.
Friedrich, M., 2013 Biological clocks and visual systems in cave-adapted animals at the dawn of
speleogenomics. Integr Comp Biol 53: 50-67.
Gertychowa, R., 1970 Studies on the ethology and space orientation of the blind cave fish
Anoptichthys jordani Hubbs et Innes 1936 (Characidae). Folia Biol-Krakow 18: 9-69.
Gross, J. B., A. Furterer, B. M. Carlson and B. A. Stahl, 2013 An integrated transcriptome-wide
analysis of cave and surface dwelling Astyanax mexicanus. PLoS One 8: e55659.
40! ! Gross, J. B., B. Meyer and M. Perkins, 2015 The rise of Astyanax cavefish. Dev Dyn: Early
online. DOI: 10.1002/dvdy.24253.
Herwig, H. J., 1976 Comparative ultrastructural investigations of the pineal organ of the blind
cave fish, Anoptichthys jordani, and its ancestor, the eyed river fish, Astyanax mexicanus.
Cell Tiss Res 167: 297-324.
Hubbs, C. L., and W. T. Innes, 1936 The first known blind fish of the family Characidae: A new
genus from Mexico. Occas Pap Mus Zool Univ Mich 342: 1-7.
Jain, N., and M. Shedpure, 2013 Effect of reproductive status, magnetic field exposure and
photoperiod on temporal organization in air-gulping activity in Clarias batrachus. Biol
Rhythm Res 44: 63-72.
Jeffery, W. R., 2009 Regressive evolution in Astyanax cavefish. Annu Rev Genet 43: 25-47.
Johnson, C., 2004 Chronobiology: Biological Timekeeping. Sinauer Associates, Inc.,
Sunderland, Massachusetts.
Kähling, J., 1961 Untersuchungen über den Lichtsinn und dessen Lokalisation bei dem
Höhlenfisch Anoptichthys jordani Hubbs and Innes (Characidae). Biol Zbl 80: 439-451.
Kavaliers, M., 1980a Circadian activity of the white sucker, Catostomus commersoni:
comparison of individual and shoaling fish. Can J Zool 58: 1399-1403.
Kavaliers, M., 1980b Social groupings and circadian activity of the killifish, Fundulus
heteroclitus. Biol Bull 158: 69-76.
Kavaliers, M., 1981 Period lengthening and disruption of socially facilitated circadian activity
rhythms of goldfish by lithium. Physiol Behav 27: 625-628.
Ki, Y., H. Ri, H. Lee, E. Yoo, J. Choe et al., 2015 Warming up your tick-tock: Temperature-
dependent regulation of circadian clocks. Neuroscientist: 1-16.
41! ! Kowalko, J. E., N. Rohner, S. B. Rompani, B. K. Peterson, T. A. Linden et al., 2013 Loss of
schooling behavior in cavefish through sight-dependent and sight-independent
mechanisms. Curr Biol 23: 1874-1883.
Kuhn, O., and J. Kähling, 1954 Augenrückbildung und Lichtstinn bei Anoptichthys jordani
Hubbs und Innes. Experientia 10: 385 - 388.
Langecker, T. G., 1989 Studies on the light reaction of epigean and cave populations of Astyanax
fasciatus (Characidae, Pisces). Mem Biospeleol 16: 169-176.
Lee, W., and K. L. Yang, 2014 Using medaka embryos as a model system to study biological
effects of the electromagnetic fields on development and behavior. Ecotoxicol Environ
Saf 108: 187-194.
Lerchl, A., A. Zachmann, M. A. Ali and R. J. Reiter, 1998 The effects of pulsing magnetic fields
on pineal melatonin synthesis in a teleost fish (brook trout, Salvelinus fontinalis).
Neurosci Lett 26.
Lu, W., Q. J. Meng, N. J. Tyler, K. A. Stokkan and A. S. Loudon, 2010 A circadian clock is not
required in an arctic mammal. Curr Biol 20: 533-537.
Lüling, K. H., 1954 Untersuchungen am Blindfisch Anoptichthys jordani Hubbs and Inns
(Characidae). II. Beobachtungen und Experimente an Anoptichthys jordani zur Prüfung
der Einstellung zum Futter, zum Licht und zur Wasserturbulenz. Zool Jahrb Abt Zool
Phys 65: 9-62.
Menna-Barreto, L., and E. Trajano, 2015 Biological rhythmicity in subterranean animals: A
function risking extinction?, pp. 55-68 in Mechanisms of Circadian Systems in Animals
and their Clinical Relevance, edited by R. Aguilar-Roblero, M. Díaz-Muñoz and M. L.
Fanjul-Moles. Springer International Publishing.
42! ! Mitchell, R. W., W. H. Russell and W. R. Elliot, 1977 Mexican Eyeless Characin Fishes, Genus
Astyanax: Environment, Distribution and Evolution. Texas Tech Press, Lubbock, Texas.
Moran, D., R. Softley and E. J. Warrant, 2014 Eyeless Mexican cavefish save energy by
eliminating the circadian rhythm in metabolism. PLoS One 9: e107877.
Mrosovsky, N., 1999 Masking: history, definitions, and measurement. Chronobiol Int 16: 415-
429.
Mrosovsky, N., R. J. Lucas and R. G. Foster, 2001 Persistence of masking responses to light in
mice lacking rods and cones. J Biol Rhythm 16: 585-587.
Nelson, W., Y. L. Tong, J. Lee and F. Halberg, 1979 Methods for cosinor-rhythmometry.
Chronobiologia 6: 305-323.
Omura, Y., 1975 Influence of light and darkness on the ultrastructure of the pineal organ in the
blind cave fish, Astyanax mexicanus. Cell Tiss Res 160: 99-112.
Oster, H., A. Avivi, A. Joel, U. Albrecht and E. Nevo, 2002 A switch from diurnal to nocturnal
activity in S. ehrenbergi is accompanied by an uncoupling of light input and the circadian
clock. Curr Biol 12: 1919-1922.
Pecar, B., In preparation Association between and within time series, pp. 1-25 in Excel
Forecasting Models for Business, Finance and IT.
Rensing, L., and P. Ruoff, 2002 Temperature effect on entrainment, phase shifting, and
amplitude of circadian clocks and its molecular bases. Chronobiol Int 19: 807-864.
Romero, A., 1984 Behavior in an 'intermediate' population of the subterranean-dwelling characid
Astyanax fasciatus. Environ Biol Fishes 10: 203-207.
Romero, A., 1985 Ontogenetic change in phototactic responses of surface and cave populations
of Astyanax fasciatus (Pisces: Characidae). Copeia 1985: 1004-1011.
43! ! Romero, A., S. M. Green, A. Romero, M. M. Lelonek and K. C. Stropnicky, 2003 One eye but
no vision: Cave fish with induced eyes do not respond to light. J Exp Zool B Mol Dev
Evol 300: 72-79.
Romero, A., and K. M. Paulson, 2001 It’s a wonderful hypogean life: A guide to the
troglomorphic fishes of the world. Environ Biol Fishes 62: 13-41.
Tabata, M., 1982 Persistence of pineal photosensory function in blind cave fish, Astyanax
mexicanus. Comp Biochem Physiol 73A: 125-127.
Takahashi, J. S., and M. Zatz, 1982 Regulation of circadian rhythmicity. Science 217: 1104-
1111.
Thines, G., and M. Weyers, 1978 Réponses locomotrices du poisson cavernicole Astyanax
jordani (Pisces, Characidae) à des signaux périodiques et apériodiques de lumière et de
température. Int J Speleol 10: 35-55.
Thines, G., and F. Wolff-Van Ermengem, 1965 Activity patterns in an epigean characidae and its
degenerated cave-relative. Anim Behav 13: 585.
Volpato, G. L., and E. Trajano, 2005 Biological Rhythms, pp. 101-153 in The Physiology of
Tropical Fishes, edited by A. L. Val, V. M. Fonseca de Almeida e Val and D. J. Randall.
Elsevier Academic Publishing, San Diego, California.
Wagner, H. J., K. Kemp, U. Mattheus and I. G. Priede, 2007 Rhythms at the bottom of the deep
sea: Cyclic current flow changes and melatonin patterns in two species of demersal fish.
Deep Sea Res Pt I 54: 1944-1956.
Wilkens, H., 1988 Evolution and genetics of epigean and cave Astyanax fasciatus (Characidae,
Pisces): Support for the neutral mutation theory, pp. 271-367 in Evolutionary Biology,
edited by M. K. Hecht and B. Wallace. Plenum Press, New York.
44! ! Yerushalmi, S., and R. M. Green, 2009 Evidence for the adaptive significance of circadian
rhythms. Ecol Lett 12: 970-981.
Yoshizawa, M., and W. R. Jeffery, 2008 Shadow response in the blind cavefish Astyanax reveals
conservation of a functional pineal eye. J Exp Biol 211: 292-299.
Yoshizawa, M., B. G. Robinson, E. R. Duboue, P. Masek, J. B. Jaggard et al., 2015 Distinct
genetic architecture underlies the emergence of sleep loss and prey-seeking behavior in
the Mexican cavefish. BMC Biol 13: 15.
Zafar, N. P., and E. Morgan, 1992 Feeding entrains an endogenous rhythm of swimming activity
in the blind Mexican cave fish. J Interdisc Cycle Res 23: 165-166.
Zhdanova, I. V., and S. G. Reebs, 2006 Circadian rhythms in fish, pp. 197-238 in Behaviour and
Physiology of Fish, edited by K. A. Sloman, R. W. Wilson and S. Balshine. Elsevier
Academic Press, San Diego, California.
Ziv, L., and Y. Gothilf, 2006 Circadian time-keeping during early stages of development. Proc
Natl Acad Sci U S A 103: 4146-4151.
45! ! FIGURE LEGENDS
Figure 2.1: Assay rig and video tracking. In our experimental set-up (A), an array of infrared
(IR) LEDs is used to illuminate an acrylic trial tank, backed with opaque contact paper (to facilitate more even illumination). A black and white CCD camera with an IR filter records the movement of the specimens (B) and relays video data to a PC for storage and analysis.
Automated video tracking (C; single tank shown for clarity) is performed in real time using
EthoVision XT software (v.8.0.516, Noldus Information Technology), which monitors changes in the surface area (yellow) and movement of the center-point (red) of the specimen as it moves through the defined “arena” (green). Further, the division of the arena into several horizontal (D) and vertical (E) zones facilitates downstream analysis of how various areas of the tank are being used over time.
Figure 2.2: 72hr assays reveal altered activity profiles in the Astyanax mexicanus cave morph. Mean velocity values (cm/s, 15 minute bins) for surface and cave specimens over 72hr under a 12:12 light/dark cycle (L/D) and constant darkness (D/D) are shown. Under L/D conditions (black and yellow bars represent periods of darkness and light, respectively), the surface form (A) displays a relatively low level of activity, with spikes in velocity often seen around subjective “dawn” and “dusk”. The cave form (C), however, is more active overall and displays sustained periods of increased activity when exposed to light. When fish normally kept under L/D conditions are subjected to constant darkness (black and gray bars represent subjective “night” and “day”, respectively), the surface form (B) displays a very weak pattern in activity, while the cave form (D) still displays an apparent, rhythmic increase in activity, with
46! ! maximum values during what would have been the middle of the “day” under L/D conditions.
Time scale is in zeitgeber time (ZT 00:00-12:00 = subjective day; ZT 12:00 = subjective night).
Figure 2.3: Absence of light during development may affect activity profiles. Mean velocity values (cm/s, 15 minute bins) over 72hr under constant darkness for surface and cave specimens reared under a 12:12 light/dark cycle (L/D) and constant darkness (D/D) are shown. Available data suggest surface fish raised under D/D conditions (A) display an activity profile that more closely resembles that of cavefish raised under L/D conditions (D), than surface fish raised under
L/D conditions (B). Cavefish raised under D/D conditions (C) have mean velocity values similar to L/D reared cavefish, but do not appear to display the same overall pattern of activity. Time scale is in zeitgeber time (ZT 00:00-12:00 = subjective day; ZT 12:00 = subjective night).
Figure 2.4: 72hr assays reveal altered spatial usage patterns in the Astyanax mexicanus cave morph. Mean times (in seconds, 900 second bins) spent in the top (green), middle (red) or bottom (blue) zones by surface and cave specimens over 72hr under a 12:12 light/dark cycle
(L/D) and constant darkness (D/D) are shown. Under L/D conditions (black and yellow bars represent periods of darkness and light, respectively), the surface form (A) is most often located in the bottom zone, though sustained periods of even higher affinity for the bottom of the tank correspond to subjective “day”. While cave form (C) also increases time spent in the bottom zone while the lights are on, it does not appear to share the same overall affinity for the bottom of the tank. When fish normally kept under L/D conditions are subjected to constant darkness
(black and gray bars represent subjective “night” and “day”, respectively), the surface form (B) shows reduced overall affinity for the bottom zone, with periods of slightly elevated usage of the
47! ! bottom zone occurring during what would have been subjective “night”, rather than “day”. The cave form (D), however, shows no discernable pattern of spatial usage. Time scale is in zeitgeber time (ZT 00:00-12:00 = subjective day; ZT 12:00 = subjective night).
Figure 2.5: Absence of light during development may affect spatial usage patterns. Mean times (in seconds, 900 second bins) spent in the top (green), middle (red) or bottom (blue) zones over 72hr under constant darkness for surface and cave specimens reared under a 12:12 light/dark cycle (L/D) and constant darkness (D/D) are shown. Available data suggest surface fish raised under D/D conditions (A) display “troglomorphic” tank usage profiles, more closely resembling those of cavefish raised under constant darkness (C) and L/D conditions (D) than surface fish raised under L/D conditions (B); this contrast is particularly apparent in data obtained for usage of the top zone. Cavefish raised under D/D conditions (C) have spatial usage patterns similar to L/D reared cavefish. Time scale is in zeitgeber time (ZT 00:00-12:00 = subjective day; ZT 12:00 = subjective night).
48! ! TABLES
Table 2.1: Mean velocity for surface and cavefish activity assays. No significance = n.s.
Number of Rearing Assay Portion of Mean Velocity Compare to Compare to Compare to Morphotype Specimens Condition Condition Triala (cm/s) Surfaceb L/Dc D/Dd Surface n = 20 12:12 L/D 12:12 L/D Entire Trial 1.235 ± 0.153 - - - Day Only 0.862 ± 0.102 - - -
Night Only 1.608 ± 0.225 - - -
Cave n = 16 12:12 L/D 12:12 L/D Entire Trial 3.954 ± 0.415 p < 0.0001 - - Day Only 4.302 ± 0.466 p < 0.0001 - -
Night Only 3.605 ± 0.421 p = 0.0002 - -
Surface n = 18 12:12 L/D 24 D/D Entire Trial 1.486 ± 0.221 - n.s. - Day Only 1.568 ± 0.280 - n.s. -
Night Only 1.403 ± 0.180 - n.s. -
Cave n = 17 12:12 L/D 24 D/D Entire Trial 4.119 ± 0.565 p = 0.0002 n.s. - Day Only 4.263 ± 0.582 p = 0.0002 n.s. -
Night Only 3.975 ± 0.554 p < 0.0001 n.s. -
Surface n = 17 24 D/D 24 D/D Entire Trial 2.507 ± 0.294 - p = 0.0020 p = 0.0215 Day Only 2.750 ± 0.330 - p < 0.0001 p = 0.0120
Night Only 2.264 ± 0.274 - n.s. n.s.
Cave n = 18 24 D/D 24 D/D Entire Trial 3.222 ± 0.303 n.s. n.s. n.s. Day Only 3.116 ± 0.301 n.s. n.s. n.s.
Night Only 3.328 ± 0.312 p = 0.0282 n.s. n.s. a “Day” and “night” refer to subjective day (ZT 00:00-12:00) and subjective night (ZT 12:00-00:00) respectively. b Results of statistical comparisons between the listed value and the corresponding value in surface fish within the same experiment. c Results of statistical comparisons between the listed value and the corresponding value in the same morphotype under light/dark (L/D) conditions. d Results of statistical comparisons between the listed value and the corresponding value for light/dark-reared fish of the same morphotype under constant darkness (D/D).
49! ! Table 2.2: Results of cosinor rhythmometry for surface and cavefish velocity data. No significance = n.s.
Zero- Rearing Assay 24hr Time Lag Acrophase Percent Morphotype Amplitude Mesor Amplitude Acrophase Condition Condition Autocorrelation Timing (ZT) Rhythm Test
Surface 12:12 L/D 12:12 L/D p < 0.0001 p < 0.0001 1.235 ± 0.128 0.362 ± 0.083 -252.7 ± 51.08˚ 16:51 ± 03:24 53.38%
Cave 12:12 L/D 12:12 L/D p < 0.0001 p < 0.0001 3.954 ± 0.230 0.610 ± 0.081 -103.7 ± 29.32˚ 06:55 ± 01:57 62.28%
Surface 12:12 L/D 24 D/D p < 0.0001 n.s./p = 0.0103a 1.486 ± 0.141 0.261 ± 0.069 -134.6 ± 58.60˚ 08:59 ± 03:54 23.28%
Cave 12:12 L/D 24 D/D p < 0.0001 p < 0.0001 4.119 ± 0.235 0.413 ± 0.059 -118.0 ± 31.65˚ 07:52 ± 02:07 59.79%
Surface 24 D/D 24 D/D p < 0.0001 p < 0.0001 2.507 ± 0.183 0.683 ± 0.062 -114.3 ± 20.12˚ 07:37 ± 01:20 83.84%
Cave 24 D/D 24 D/D n.s. n.s. - - - - - a Level of autocorrelation not significant with a 24hr time lag, but significant with a 48hr time lag.
50! ! Table 2.3: Mean top zone usage for surface and cavefish activity assays. No significance = n.s.
Number of Rearing Assay Portion of Mean Time in Compare to Compare to Compare to Morphotype Specimens Condition Condition Triala Top Zone (s)b Surfacec L/Dd D/De Surface n = 20 12:12 L/D 12:12 L/D Entire Trial 39.68 ± 4.194 - - - Day Only 15.43 ± 3.738 - - -
Night Only 63.94 ± 7.870 - - -
Cave n = 16 12:12 L/D 12:12 L/D Entire Trial 236.5 ± 22.87 p < 0.0001 - - Day Only 196.4 ± 28.59 p < 0.0001 - -
Night Only 276.6 ± 22.52 p < 0.0001 - -
Surface n = 18 12:12 L/D 24 D/D Entire Trial 84.84 ± 12.70 - p = 0.0360 - Day Only 95.09 ± 18.24 - p < 0.0001 -
Night Only 74.58 ± 9.609 - n.s. -
Cave n = 17 12:12 L/D 24 D/D Entire Trial 313.0 ± 26.64 p < 0.0001 n.s. - Day Only 307.2 ± 27.38 p < 0.0001 p = 0.0052 -
Night Only 318.9 ± 26.17 p < 0.0001 n.s. -
Surface n = 17 24 D/D 24 D/D Entire Trial 292.1 ± 27.38 - p < 0.0001 p < 0.0001 Day Only 268.2 ± 27.33 - p < 0.0001 p = 0.0002
Night Only 280.2 ± 26.72 - p < 0.0001 p < 0.0001
Cave n = 18 24 D/D 24 D/D Entire Trial 322.4 ± 27.85 n.s. n.s. n.s. Day Only 313.6 ± 26.17 n.s. p = 0.0092 n.s.
Night Only 318.0 ± 26.85 n.s. n.s. n.s. a “Day” and “night” refer to subjective day (ZT 00:00-12:00) and subjective night (ZT 12:00-00:00) respectively. b Reported times are in seconds out of 900 second (15 minute) time bins. c Results of statistical comparisons between the listed value and the corresponding value in surface fish within the same experiment. d Results of statistical comparisons between the listed value and the corresponding value in the same morphotype under light/dark (L/D) conditions. e Results of statistical comparisons between the listed value and the corresponding value for light/dark-reared fish of the same morphotype under constant darkness (D/D).
51! ! Table 2.4: Mean middle zone usage for surface and cavefish activity assays. No significance = n.s.
Number of Rearing Assay Portion of Mean Time in Compare to Compare Compare Morphotype Specimens Condition Condition Triala Middle Zone (s)b Surfacec to L/Dd to D/De Surface n = 20 12:12 L/D 12:12 L/D Entire Trial 158.8 ± 22.09 - - - Day Only 65.40 ± 13.29 - - -
Night Only 252.2 ± 35.99 - - -
Cave n = 16 12:12 L/D 12:12 L/D Entire Trial 265.4 ± 16.41 p = 0.0010 - - Day Only 251.1 ± 21.10 p < 0.0001 - -
Night Only 279.7 ± 14.75 n.s. - -
Surface n = 18 12:12 L/D 24 D/D Entire Trial 291.4 ± 39.67 - p = 0.0222 - Day Only 302.5 ± 42.64 - p < 0.0001 -
Night Only 280.3 ± 38.39 - n.s. -
Cave n = 17 12:12 L/D 24 D/D Entire Trial 250.6 ± 15.52 n.s. n.s. - Day Only 251.6 ± 16.97 n.s. n.s. -
Night Only 249.6 ± 14.64 n.s. n.s. -
Surface n = 17 24 D/D 24 D/D Entire Trial 343.3 ± 17.91 - p < 0.0001 n.s. Day Only 365.4 ± 26.44 - p < 0.0001 n.s.
Night Only 354.3 ± 20.79 - p = 0.0477 n.s.
Cave n = 18 24 D/D 24 D/D Entire Trial 242.4 ± 9.540 p = 0.0001 n.s. n.s. Day Only 242.8 ± 9.061 p = 0.0003 n.s. n.s.
Night Only 242.6 ± 9.115 p < 0.0001 n.s. n.s. a “Day” and “night” refer to subjective day (ZT 00:00-12:00) and subjective night (ZT 12:00-00:00) respectively. b Reported times are in seconds out of 900 second (15 minute) time bins. c Results of statistical comparisons between the listed value and the corresponding value in surface fish within the same experiment. d Results of statistical comparisons between the listed value and the corresponding value in the same morphotype under light/dark (L/D) conditions. e Results of statistical comparisons between the listed value and the corresponding value for light/dark-reared fish of the same morphotype under constant darkness (D/D).
52! ! Table 2.5: Mean bottom zone usage for surface and cavefish activity assays. No significance = n.s.
Number of Rearing Assay Portion of Mean Time in Compare to Compare Compare Morphotype Specimens Condition Condition Triala Bottom Zone (s)b Surfacec to L/Dd to D/De Surface n = 20 12:12 L/D 12:12 L/D Entire Trial 701.4 ± 24.23 - - - Day Only 819.1 ± 15.39 - - -
Night Only 583.7 ± 40.39 - - -
Cave n = 16 12:12 L/D 12:12 L/D Entire Trial 397.9 ± 31.88 p < 0.0001 - - Day Only 452.3 ± 42.44 p < 0.0001 - -
Night Only 343.5 ± 28.89 p = 0.0004 - -
Surface n = 18 12:12 L/D 24 D/D Entire Trial 523.7 ± 45.65 - p = 0.0101 - Day Only 502.3 ± 52.53 - p < 0.0001 -
Night Only 545.0 ± 41.19 - n.s. -
Cave n = 17 12:12 L/D 24 D/D Entire Trial 336.2 ± 37.49 p = 0.0039 n.s. - Day Only 341.1 ± 39.21 p = 0.0333 p = 0.0363 -
Night Only 331.3 ± 36.00 p = 0.0009 n.s. -
Surface n = 17 24 D/D 24 D/D Entire Trial 264.5 ± 22.08 - p < 0.0001 p = 0.0008 Day Only 266.2 ± 26.50 - p < 0.0001 p = 0.0071
Night Only 265.3 ± 22.10 - p < 0.0001 p = 0.0001
Cave n = 18 24 D/D 24 D/D Entire Trial 335.1 ± 27.54 n.s. n.s. n.s. Day Only 343.5 ± 26.14 p = 0.0496 n.s. n.s.
Night Only 339.3 ± 26.58 n.s. n.s. n.s. a “Day” and “night” refer to subjective day (ZT 00:00-12:00) and subjective night (ZT 12:00-00:00) respectively. b Reported times are in seconds out of 900 second (15 minute) time bins. c Results of statistical comparisons between the listed value and the corresponding value in surface fish within the same experiment. d Results of statistical comparisons between the listed value and the corresponding value in the same morphotype under light/dark (L/D) conditions. e Results of statistical comparisons between the listed value and the corresponding value for light/dark-reared fish of the same morphotype under constant darkness (D/D).
53! ! Table 2.6: Results of cosinor rhythmometry for surface and cavefish top zone usage data. No significance = n.s.
Zero- Rearing Assay 24hr Time Lag Acrophase Percent Morphotype Amplitude Mesor Amplitude Acrophase Condition Condition Autocorrelation Timing (ZT) Rhythm Test
Surface 12:12 L/D 12:12 L/D p < 0.0001 p < 0.0001 38.30 ± 0.717 21.39 ± 3.739 -264.6 ± 38.43˚ 17:39 ± 02:34 62.17%
Cave 12:12 L/D 12:12 L/D p < 0.0001 p < 0.0001 236.5 ± 1.783 42.14 ± 4.659 -254.2 ± 24.24˚ 16:57 ± 01:37 78.56%
Surface 12:12 L/D 24 D/D p < 0.0001 p = 0.0051 84.84 ± 1.068 20.17 ± 4.571 -119.9 ± 50.02˚ 08:00 ± 03:20 29.34%
Cave 12:12 L/D 24 D/D n.s. n.s. - - - - -
Surface 24 D/D 24 D/D p = 0.0001 n.s. - - - - -
Cave 24 D/D 24 D/D p = 0.0246 n.s. - - - - -
54! ! Table 2.7: Results of cosinor rhythmometry for surface and cavefish bottom zone usage data. No significance = n.s.
Zero- Rearing Assay 24hr Time Lag Acrophase Percent Morphotype Amplitude Mesor Amplitude Acrophase Condition Condition Autocorrelation Timing (ZT) Rhythm Test
Surface 12:12 L/D 12:12 L/D p < 0.0001 p < 0.0001 710.3 ± 3.091 122.2 ± 11.69 -92.99 ± 20.96˚ 06:12 ± 01:24 77.41%
Cave 12:12 L/D 12:12 L/D p < 0.0001 p < 0.0001 397.9 ± 2.313 59.25 ± 6.653 -80.43 ± 24.62˚ 05:22 ± 01:38 73.93%
Surface 12:12 L/D 24 D/D p < 0.0001 p < 0.0001 523.7 ± 2.654 40.73 ± 6.943 -277.1 ± 37.48˚ 18:28 ± 02:30 48.28%
Cave 12:12 L/D 24 D/D p = 0.0023 n.s. - - - - -
Surface 24 D/D 24 D/D p = 0.0007 n.s. - - - - -
Cave 24 D/D 24 D/D p = 0.0029 n.s. - - - - -
55! ! FIGURES
Figure 2.1 A Assay(Rig(Set6Up( B Camera(View(
To(Computer(
C Live(Tracking( D Horizontal(Zones( E VerCcal(Zones(
56! ! Figure 2.2
A Surface 72hr L/D Trials
8 7 n&=&20& 6 5 4 3 2
Mean Velocity (cm/s) MeanVelocity 1 0 0:00 2:00 4:00 6:00 8:00 0:00 2:00 4:00 6:00 8:00 0:00 2:00 4:00 6:00 8:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00
B Surface 72hr D/D Trials
8 7 n&=&18& 6 5 4 3 2
Mean Velocity (cm/s) MeanVelocity 1 0 0:00 2:00 4:00 6:00 8:00 0:00 2:00 4:00 6:00 8:00 0:00 2:00 4:00 6:00 8:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00
C Cave 72hr L/D Trials
8 7 6 5 4 3 2
Mean Velocity (cm/s) MeanVelocity 1 n&=&16& 0 0:00 2:00 4:00 6:00 8:00 0:00 2:00 4:00 6:00 8:00 0:00 2:00 4:00 6:00 8:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00
D Cave 72hr D/D Trials
8 7 6 5 4 3 2
Mean Velocity (cm/s) MeanVelocity 1 n&=&17& 0 0:00 2:00 4:00 6:00 8:00 0:00 2:00 4:00 6:00 8:00 0:00 2:00 4:00 6:00 8:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00
57! ! Figure 2.3
A Surface 72hr D/D-D/D Trials 8 7 n&=&17& 6 5 4 3 2
Mean Velocity (cm/s) MeanVelocity 1 0 0:00 2:00 4:00 6:00 8:00 0:00 2:00 4:00 6:00 8:00 0:00 2:00 4:00 6:00 8:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00
B Surface 72hr D/D Trials
8 7 n&=&18& 6 5 4 3 2
Mean Velocity (cm/s) MeanVelocity 1 0 0:00 2:00 4:00 6:00 8:00 0:00 2:00 4:00 6:00 8:00 0:00 2:00 4:00 6:00 8:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 C Cave 72hr D/D-D/D Trials 8 7 6 5 4 3 2 Mean Velocity (cm/s) MeanVelocity 1 n&=&18& 0 0:00 2:00 4:00 6:00 8:00 0:00 2:00 4:00 6:00 8:00 0:00 2:00 4:00 6:00 8:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00
D Cave 72hr D/D Trials
8 7 6 5 4 3 2
Mean Velocity (cm/s) MeanVelocity 1 n&=&17& 0 0:00 2:00 4:00 6:00 8:00 0:00 2:00 4:00 6:00 8:00 0:00 2:00 4:00 6:00 8:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00
58! ! Figure 2.4
A Surface 72hr L/D Trials
900 800 700 600 500 400 300 200
Mean Time in Zone (s) Zone in MeanTime 100 n"="20" 0 0:00 2:00 4:00 6:00 8:00 0:00 2:00 4:00 6:00 8:00 0:00 2:00 4:00 6:00 8:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00
B Surface 72hr D/D Trials
900 800 700 600 500 400 300 200
Mean Time in Zone (s) Zone in MeanTime 100 n"="18" 0 0:00 2:00 4:00 6:00 8:00 0:00 2:00 4:00 6:00 8:00 0:00 2:00 4:00 6:00 8:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 C Cave 72hr L/D Trials
900 800 700 600 500 400 300 200
Mean Time in Zone (s) Zone in MeanTime 100 n"="16" 0 0:00 2:00 4:00 6:00 8:00 0:00 2:00 4:00 6:00 8:00 0:00 2:00 4:00 6:00 8:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 D Cave 72hr D/D Trials
900 800 700 600 500 400 300 200
Mean Time in Zone (s) Zone in MeanTime 100 n"="17" 0 0:00 2:00 4:00 6:00 8:00 0:00 2:00 4:00 6:00 8:00 0:00 2:00 4:00 6:00 8:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00
59! ! Figure 2.5
A Surface 72hr D/D-D/D Trials
900 800 700 600 500 400 300 200
Mean Time in Zone (s) Zone in MeanTime 100 n"="17" 0 0:00 2:00 4:00 6:00 8:00 0:00 2:00 4:00 6:00 8:00 0:00 2:00 4:00 6:00 8:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00
B Surface 72hr D/D Trials
900 800 700 600 500 400 300 200
Mean Time in Zone (s) Zone in MeanTime 100 n"="18" 0 0:00 2:00 4:00 6:00 8:00 0:00 2:00 4:00 6:00 8:00 0:00 2:00 4:00 6:00 8:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00
C Cave 72hr D/D-D/D Trials
900 800 700 600 500 400 300 200
Mean Time in Zone (s) Zone in MeanTime 100 n"="18" 0 0:00 2:00 4:00 6:00 8:00 0:00 2:00 4:00 6:00 8:00 0:00 2:00 4:00 6:00 8:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00
D Cave 72hr D/D Trials
900 800 700 600 500 400 300 200
Mean Time in Zone (s) Zone in MeanTime 100 n"="17" 0 0:00 2:00 4:00 6:00 8:00 0:00 2:00 4:00 6:00 8:00 0:00 2:00 4:00 6:00 8:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 10:00 12:00 !
60! !
CHAPTER THREE
A high-density linkage map for Astyanax mexicanus using genotyping-by-sequencing
technology
Brian M. Carlson, Samuel W. Onusko* and Joshua B. Gross*
Department of Biological Sciences University of Cincinnati Cincinnati, OH 45221-0006
*The work described here, as well as the composition of the narrative for this manuscript, was principally completed by Brian M. Carlson with the assistance of the listed co-authors. This manuscript can be viewed in its published form at: http://www.g3journal.org/content/5/2/241
! 61! ABSTRACT
The Mexican tetra, Astyanax mexicanus, is a unique model system consisting of cave- adapted and surface-dwelling morphotypes which diverged >1My ago. This remarkable natural experiment has enabled powerful genetic analyses of cave adaptation. Here, we describe the application of next-generation sequencing technology to the creation of a high-density linkage map. Our map comprises over 2200 markers populating 25 linkage groups constructed from genotypic data generated from a single genotyping-by-sequencing project. We leveraged emergent genomic and transcriptomic resources to anchor hundreds of anonymous Astyanax markers to the genome of the zebrafish (Danio rerio), the most closely related model organism to our study species. This facilitated the identification of 784 distinct connections between our linkage map and the Danio rerio genome, highlighting several regions of conserved genomic architecture between the two species despite ~150My of divergence. Using a Mendelian cave- associated trait as a proof-of-principle, we successfully recovered the genomic position of the albinism locus near the gene Oca2. Further, our map successfully informed the positions of unplaced Astyanax genomic scaffolds within particular linkage groups. This ability to identify the relative location, orientation and linear order of unaligned genomic scaffolds will facilitate ongoing efforts to improve upon the current early draft and assemble future versions of the
Astyanax physical genome. Moreover, this improved linkage map will enable higher resolution genetic analyses and catalyze the discovery of the genetic basis for cave-associated phenotypes.
! 62! INTRODUCTION
The blind Mexican cave tetra is a powerful system for understanding the evolutionary mechanisms governing regressive phenotypes. These animals were discovered in 1936 and initially assigned to a new genus – Anoptichthys (lit. “bony fish without eyes”; Hubbs and Innes
1936). Breeding studies in the 1940s led to the discovery of viable hybrid offspring resulting from crosses between the (derived) blind cave-dwelling forms and (ancestral) surface-dwelling forms from the same geographical region of NE Mexico (Breder 1943a; Breder 1943b). Both morphotypes are now regarded as members of the same (or a closely-related) species, Astyanax mexicanus. This system has spurred well over half a century of comparative research (Şadoǧlu
1956) focusing on unresolved problems in evolution (Jeffery 2001), development (Pottin et al.
2011), genetics (Schemmel 1980), physiology (Salin et al. 2010) and behavior (Burchards et al.
1986).
Classical and quantitative genetic approaches have provided clear evidence that many troglomorphic (cave-associated) phenotypes evolved through heritable genetic changes. These studies centered on both Mendelian and complex phenotypes, including eye regression
(Yamamoto et al. 2004; Protas et al. 2007; Yoshizawa et al. 2012; O'Quin et al. 2013), feeding- related behaviors (Schemmel 1980; Yoshizawa et al. 2012), sleep loss (Duboué et al. 2011), schooling behavior (Kowalko et al. 2013), pigmentation loss (reviewed in Jeffery 2009), and intra-specific aggression (Elipot et al. 2013). QTL studies have identified candidate genes mediating a variety of these traits, such as retinal degeneration (O'Quin et al. 2013), rib number, eye size (Gross et al. 2008), albinism (Oca2; Protas et al. 2006) and the brown phenotype (Mc1r;
Gross et al. 2009).
! 63! Genomic resources for this model system, however, have historically been limited. The first linkage map was calculated based on recombination frequencies of an experimental F1 x
Pachón cave backcross pedigree using markers generated from random amplified polymorphic
DNA (RAPD) fingerprinting (Borowsky and Wilkens 2002). This map was supplanted by a higher resolution map, using more individuals and markers composed of polymorphic microsatellites identified using ~CAN dinucleotide repeats (Protas et al. 2006). Using this second-generation linkage map, Protas et al. (2008) discovered a genetic basis for several cave- associated phenotypic changes including pigmentation regression, reduced rib numbers, slower weight loss, and increased chemical sensitivity. Early comparative genomic analyses utilizing this map first demonstrated extensive synteny conserved between Astyanax and Danio rerio, despite ~150My of divergence (Gross et al. 2008). In 2013, O’Quin et al. published the first next generation sequencing (NGS)-based linkage map using restriction-associated DNA sequencing
(RAD-seq) technology. This map, comprising 698 markers on 25 linkage groups strengthened the evidence for vast regions of synteny between the genomes of Astyanax and zebrafish, and identified several critical loci associated with retinal degeneration (O’Quin et al. 2013).
Here, we present the most dense, comprehensive linkage map to date using genotyping- by-sequencing (GBS) technology. This technology enables accurate and high-throughput collection of massive amounts of sequence data (Davey et al. 2011), including thousands of single-nucleotide polymorphisms (SNPs) segregating between cave- and surface-dwelling morphs. GBS utilizes deep Illumina sequencing of restriction enzyme-nicked genomic DNA libraries that are uniquely barcoded for each member of an experimental pedigree. This technique is optimized to avoid inclusion of repetitive portions of the genome, and is extremely specific and highly reproducible (Elshire et al. 2011). Fish are well-represented among studies
! 64! employing GBS and other RAD-seq based methodologies (Rowe et al. 2011). However, a majority of GBS studies in fish have focused on species of commercial (Everett et al. 2012;
Houston et al. 2012; Li et al. 2014) or conservational concern (Hecht et al. 2013; Ogden et al.
2013; Hess et al. 2014; Larson et al. 2014). Here, we adapted this technology to construct a high- density linkage map for evolutionary and developmental studies in our emerging model system.
The resulting linkage map will enable higher resolution genomic studies and inform the assignment of chromosomal builds for the ongoing Astyanax genome sequencing project
(McGaugh et al. 2014).
MATERIALS AND METHODS
Pedigree and animal husbandry
Linkage mapping and QTL studies were carried out using genotypic and phenotypic data obtained from two separate F2 hybrid mapping populations (n = 129; n = 41) bred from a male surface fish and female cavefish from the Pachón cave. In addition, surface (n = 4), Pachón cave
(n = 4) and surface x Pachón F1 hybrid (n = 4) specimens were used to evaluate and code GBS markers for use with JoinMap software (v. 4.1; Kyazma; see below), but were not included in linkage mapping calculations. Parental specimens belonged to laboratory populations originally sourced from the El Abra region of northeastern Mexico and generously provided to our lab by
Dr. Richard Borowsky (New York University). All live fish used in this study were maintained as previously described (see Gross et al. 2013). Every individual from the “Asty66” F2 population (n = 129) was individually reared in a 1-liter tank. All phenotypic data from the
“Asty12” F2 population (n = 41) was obtained from paraformaldehyde-preserved specimens.
! 65! Genotyping-by-sequencing
Genomic DNA was extracted from caudal tail fin tissue of live surface, cave, F1 and F2 hybrid Astyanax mexicanus specimens, using the DNeasy Blood and Tissue Kit (Qiagen) as previously described (Gross et al. 2013). Twenty genomic samples were digested with EcoRI, subjected to gel electrophoresis and imaged to verify that sample quality, concentration and restriction fragment size distributions were suitable for use in downstream analyses. DNA samples were then pipetted into individual wells of 96-well plates and diluted to a final volume of 30µl (100ng/µl). Samples were processed by the Institute for Genomic Diversity (Cornell
University), where genomic libraries were constructed and GBS was performed as described elsewhere (Elshire et al. 2011; Lu et al. 2013).
GBS marker selection
Genotypes for each of 7956 GBS markers (each consisting of a single SNP in a 64-bp long sequence fragment) were screened in cave and surface (parental) forms to assign the morphotypic origin of each allele. F1 individuals were then evaluated to confirm heterozygosity at each locus. The morphotypic origin of each allele was assigned by consensus – if three or more (out of four) surface or cave individuals had the identical nucleotide at a particular locus, the genotype was assigned to the consensus parental population. Likewise, a true “hybrid” genotype was assigned if three or more F1 individuals harbored the same heterozygous condition
(e.g., M, R, S, W, Y, K SNP code) at a given locus. Those genotypes with an ambiguous morphotypic origin were denoted “NA”.
Markers were then screened for suitability in linkage calculations. Markers were deemed unsuitable, and discarded from further analysis, if 1) neither parental genotype could be assigned
! 66! (i.e., both the surface and cave genotypes were scored “NA”) or 2) the assigned surface and cave genotypes were identical. 6006 genomic markers were deemed suitable and prepared for linkage map calculation using the “cross-pollination” (CP) segregation coding used in JoinMap. At this stage, 107 markers were found to be uninformative (i.e., a single genotype was shared by all F2 individuals) and discarded from further analysis. We screened the remaining set (n = 5899) to identify markers failing to conform to predicted genotypic ratios (e.g., 1:2:1 ratios across the entire pedigree). 2896 markers demonstrated a χ2 value greater than 50, implying significant departure from the predicted genotype ratio and were discarded from further analysis. Our final
GBS marker set included 3003 markers, evaluated in 170 F2 individuals.
Linkage map construction and QTL analysis
Linkage map calculations were carried out using JoinMap (v.4.1, Kyazma). Our workflow employed program default settings, with the following exceptions: 1) the maximum grouping independence LOD value was set to 50.0; 2) linkage groups were calculated using regression mapping; and 3) linkage mapping was performed using the Kosambi method
(Kosambi 1943). Linkage groups were assigned based on independence LOD scores. We increased the maximum grouping independence LOD value to 50.0, as the default value of 10.0 did not allow sufficient subdivision of our data into an appropriate number of groups. Initial groupings identified 29 groups populated with between 10 – 225 markers, with independence
LOD scores ranging from 7.0 – 21.0. These groups were then processed for formal mapping calculations.
The first round of mapping produced 28 linkage groups comprising a total map length of
2956 cM. At this stage, one linkage group (comprising 10 markers, independence LOD = 19.0)
! 67! failed to assemble into a consolidated group and was therefore dropped from further analysis.
The remaining individual linkage groups ranged in length from 27.25 – 187.46 cM, containing between 10 and 225 markers with an average inter-marker distance between 0.51 – 6.40 cM.
After this initial round of mapping, we further screened existing linkages to target the most optimal 25 groups (Astyanax mexicanus has karyotypic number of 25; Kirby et al. 1977), and reduce the average inter-marker distance to a target of ~1 cM. Accordingly, nine groups (10 ≤ n
≤ 45 markers) were removed due to low marker number and/or unusually high average inter- marker distance. The five largest groups (154 ≤ n ≤ 225 markers) were then subdivided at the lowest independence LOD value resulting in two linkage groups comprising 20+ markers.
Throughout mapping, we limited the inflation of the overall map length by eliminating certain markers sparsely populating distal ends of otherwise densely populated linkage groups. This resulted in size reduction of the five longest remaining linkage groups (142.041 ≤ n ≤ 187.458 cM), by splitting them at the lowest independence LOD score at which a group (comprising 10+ markers) was separated. In these cases, the larger of the two resulting groups was retained. The resulting 25 linkage groups (independence LOD scores 10.0 ≤ n ≤ 24.0) were subjected to additional mapping. Groupings of markers dropped during this or a subsequent round of mapping were excluded from further analysis.
The second round of mapping produced a 2556.6 cM linkage map composed of 25 linkage groups, each consisting of 25 – 171 markers, ranging in length from 31.18 – 142.78 cM with mean inter-marker distances ranging from 0.47 – 3.66 cM. Using the same criteria described above, an additional group (comprising 25 markers and an average inter-marker distance of 3.658) was eliminated. A densely populated group with a high independence LOD
! 68! (153 markers; 135.73 cM; independence LOD of 24.0) was split and 12 linkage groups (103.982
≤ n ≤ 142.783 cM) were trimmed.
The result of this third and final round of mapping was then analyzed for genomic synteny shared between Astyanax mexicanus and the zebrafish genome and used to map albinism as a proof-of-concept. Albinism was scored as a binary phenotype wherein presence (0) of melanin or absence of melanin (1) was assigned to each of the members of our experimental F2 pedigree. All QTL analyses of albinism were conducted using R/qtl (Broman et al. 2003), run for each of three scan-one mapping methods: marker regression (MR), expect maximum (EM) and
Haley-Knott (HK), according to the methodology in Gross et al. (2014).
Assignment of genomic synteny between the Astyanax mexicanus and Danio rerio genomes
At present, physical genome resources for Astyanax mexicanus are in their early draft phases (McGaugh et al. 2014). Therefore, we anchored our GBS-based linkage map to the physical genome of the most closely-related fish model system with comprehensive resources,
Danio rerio. Astyanax and Danio are members of the superorder Ostariophysii, which diverged
~150My ago (Briggs 2005). In spite of this distance, significant genome-level synteny remains between these species (Gross et al. 2008; O'Quin et al. 2013). Our GBS marker set was derived from endonuclease restriction site-based libraries, and was therefore anonymous. We first identified all GBS markers that could be directly localized to a conserved region in the D. rerio genome. Accordingly, we performed BLAST searches of the 64-bp sequences comprising our marker sequences directly against the Danio genome (downloaded from the Ensembl genome browser; www.ensembl.org).
! 69! These and all subsequent searches were performed using a BLASTN script run on the
Ohio Supercomputing Cluster (OSC). All quality control defaults, including an expect value (e- value) cutoff of 10, were maintained. The script permitted the return of alignments between a given 64-bp marker sequence and regions of up to three distinct targets (e.g., three different
Danio rerio chromosomes). In cases where a single marker sequence aligned multiple times with the same target, raw results were filtered by e-value, retaining the lowest e-value alignment for each marker-target pairing. There are two 64-bp sequences for each GBS marker, differing only in that each contains one of the two alleles for the imbedded SNP. As both of these sequences were included when BLAST searches using the 64-bp marker sequences were conducted, this filtering step also served to collapse these results into a single set of results, retaining the better of the two alignments for each marker-target pairing.
In some instances, a single queried sequence returned alignments with multiple targets.
These instances were resolved by sorting results to determine the “top hit” which was defined as having the lowest e-value and highest percent identity (in case of an e-value tie) to a particular target sequence. If the target of the top hit (i.e., the alignment with the lowest e-value) for a given marker sequence agreed with the target reported for one or more other markers on the same linkage group that returned only a single, robust hit, then the top hit for the marker in question was considered “supported” and retained. If the top hit was not supported in this fashion, but a different BLAST result was, then the latter “not top hit, supported” result was retained instead. If none of the results returned for a marker sequence were supported, then the top hit was retained, despite the lack of support. In rare cases, there was no way to resolve which result should be retained. Results for these “unresolved” markers were discarded.
! 70! When using BLAST searches to align our 64-bp markers directly to the Danio rerio genome returned relatively few high-quality hits, we developed a strategy whereby we first aligned our GBS marker sequences to the Astyanax mexicanus genome and transcriptome data.
This information was then used to identify homologous Danio genomic and transcriptomic sequences. Current genomic resources in Astyanax consist of >10,000 unplaced genomic scaffolds (Bioproject PRJNA89115). The collective sequence data for the Astyanax genome
(GenBank Assembly ID GCA_000372685) was downloaded from Ensembl, along with the transcript sequences for 23,042 predicted genes. BLAST searches were used to determine putative locations for the 64-bp sequences of the 2235 GBS markers comprising our final linkage map in both the Astyanax genomic and transcriptomic data sets. After results of initial searches were performed as described, ~2000-bp stretches of genomic sequence harboring our 64-bp GBS marker sequences were aligned with the Danio genome. Similarly, full sequences for predicted
Astyanax transcripts to which our GBS markers aligned were queried against a Danio cDNA database downloaded from Ensembl. Both data sets were then filtered (as described), yielding a single, “best” Danio alignment for each informative query. This process enabled us to leverage draft genomic and transcriptomic data to augment the amount of sequence information associated with our 64-bp GBS markers and identify homologous genomic positions in a well-characterized model system.
After BLAST searches using the direct, genomic and transcriptomic alignment methods were completed, the filtered results for all three were combined. Where multiple methods returned results for the same marker, a single result was chosen and retained using the same filtering process applied to single data sets (above). The Circos program (Krzywinski et al. 2009)
! 71! was used to visualize comparative genomic positions between our linkage map and the Astyanax and Danio rerio genomes.
Position identification for previously published markers in the Astyanax genome
Previous maps published by Gross et al. (2008) and O’Quin et al. (2013) were employed to examine synteny between Astyanax and Danio, and to provide a comparison between this study and prior studies. These authors provided predicted Danio positions for the markers used in their analyses, but positions in the draft Astyanax genome were not determined since these studies predated available genomic resources. Our GBS-based map does not share any markers with the two previous maps, so it was necessary to identify positions of previously generated markers in Astyanax to enable comparison between previous mapping efforts and those described here. Accordingly, microsatellite and RAD-seq marker sequences (where available) for each data set were aligned with Astyanax genome scaffolds using the same BLAST and filtering protocols employed for our own data (above). Both previous studies included markers located in candidate genes. The locations of Astyanax orthologs of these candidate genes were identified using
Ensembl.
GBS marker sequences and genotyping data are available from the Dryad Digital
Repository: http://dx.doi.org/10.5061/dryad.6s718
RESULTS AND DISCUSSION
A high-density linkage map in Astyanax mexicanus
Here, we present a dense linkage map for Astyanax mexicanus, generated using genotyping-by-sequencing technology. This map was created using 170 experimental F2
! 72! individuals, based on genotypic information for 3003 loci. The construction of this map ultimately yielded 25 linkage groups (the karyotypic number for Astyanax) comprising 2235 markers spanning 2110.7 cM, with an average inter-marker distance of 1.052 cM (Figure 3.1;
Table S1 at http://www.g3journal.org/content/5/2/241/suppl/DC1). The strategy we employed enables application of powerful, cost-effective next-generation sequencing technology to facilitate genetic studies in emerging or non-model systems.
Cross-genera marker identification was greatly facilitated by alignment first to draft
Astyanax genomic and transcriptomic resources, followed by searches of the homologous sequences in Danio (Figure 3.2A-C). While direct BLAST searches of our 64-bp GBS marker sequences returned results for few of the markers in our map (1.2%), success rates were much higher when using Astyanax genomic (26.5%) or transcriptomic (13.3%) sequences as an intermediary (Table 3.1). Each Danio rerio chromosome was represented in our comparative genomic analysis, with Astyanax linkage groups containing 14 – 52 markers (average = 30.84) comprising ancient syntenic blocks shared with each of 25 zebrafish chromosomes (Figure
3.2D). Of the 2235 GBS markers that constitute our linkage map, 784 marker sequences (35.1%) were successfully identified in the Danio rerio genome (Figure 3.3A).
We performed a proof-of-concept analysis using the albinism phenotype to validate the utility of our GBS-based linkage map (Figure 3.3B-D). Accordingly, we mapped the monogenic trait of albinism, using the R/qtl package to evaluate phenotypic and genotypic data for the 170
F2 hybrid individuals used to construct our map. We identified a peak LOD score of 20.68 on linkage group 13, associated with marker TP71406. This marker and the surrounding region form a syntenic block within a region of Danio rerio chromosome 6. This genomic interval contains the gene Oca2, previously demonstrated to be the causative locus for albinism in
! 73! Astyanax cavefish. This supports previous findings of conserved synteny inclusive of significant portions of chromosome 6 in Danio (Gross et al. 2008; O'Quin et al. 2013), and implies our densely-populated map will enable future QTL studies of trait evolution in Astyanax.
Conserved genomic architecture between Astyanax and Danio based on GBS markers
Our analysis of synteny between Astyanax and Danio illustrates variable levels of genomic conservation across linkage groups (Figures 3.2D; 3.3A). Certain chromosomes, for instance, appear to have changed little since the divergence of these teleost species (e.g., Danio chromosomes 6 and 23, in Astyanax linkage groups 13 and 15, respectively). However, other
Danio chromosomes appear scattered across several linkage groups, without a consensus representation for any particular group (e.g., Danio chromosomes 2 and 5).
We believe these findings most likely reflect genomic rearrangements that have occurred since the divergence of these two species. However, this finding could also be attributed to low representation of particular Danio chromosomes within our GBS marker set. We examined this possibility by assessing the number of syntenic links between our GBS-based linkage map and each Danio chromosome. We would anticipate that longer chromosomes would naturally harbor more syntenic links. Values were therefore expressed as a ratio of syntenic links per megabase
(mean = 0.59 GBS markers/Mb). While the mean value for chromosomes that were not strongly represented on any particular linkage group in our map (i.e., had fewer than 10 syntenic links with each linkage group, mean = 0.52 GBS markers/Mb, n = 8) was lower than that for chromosomes demonstrating strong synteny with a particular linkage group (mean = 0.61 GBS markers/Mb, n = 17), there was not a significant difference between the two groups (t23 = 0.5809, p = 0.5670). This leads us to conclude that, while representation of particular chromosomes in
! 74! our data set may be a contributing factor, it is unlikely that this is the primary cause of the differences in chromosomal representation patterns observed.
Alternatively, BLAST results for Astyanax GBS markers (or the larger Astyanax sequences to which they were aligned) may include paralogous genes or otherwise ambiguous results that could lead to erroneous links between a linkage group and a Danio chromosome.
While we cannot rule out this possibility, we feel our strategy prioritized the “optimal” BLAST result among multiple hits for a single marker leading to alignments that agree with nearby, unambiguous results (Table 3.1). As a result, of the 784 markers in our map for which a putative
Danio position was determined, only 15.9% (n = 125) of final calls were unsupported by the results for other markers belonging to the same linkage group (Table S1). Given that chromosomal arrangements have occurred over the ~150My since divergence, we feel our systematic approach best identifies paralogous genes and other potential sources of ambiguity.
Erroneous or ambiguous genotyping data may have led to incorrect assignment of “cave” and “surface” alleles for particular markers. These erroneous assignments could have adversely affected downstream efforts, causing markers to be incorrectly placed during the grouping and/or mapping stages of linkage map construction. All efforts were made to ensure allelic identification was accurate using a stringent screening process (see Methods), however we relied on a relatively small number of cave, surface and F1 hybrid individuals (n = 4 each) to identify parental allelic origin. Similarly, the relatively small number of meiotic events represented by the
170 F2 individuals may have resulted in linkage map inaccuracies (Gross et al. 2008). Future comparisons between the map we present here and a finished-grade Astyanax genome will clarify if regions lacking synteny between Astyanax and Danio are attributable to errors in our linkage map or genomic rearrangements that have occurred since the divergence of these taxa.
! 75!
Unplaced Astyanax genome scaffolds can be anchored to our new linkage map
Positional locations in the current draft of the Astyanax genome were established for
93.6% (n = 2091) of the 2235 GBS markers present in our map. These markers were localized to positions spread across 598 different Astyanax genome scaffolds. Our 25 Astyanax linkage groups contain markers representing between 12 (linkage groups 8 and 22) and 55 (linkage group
3) genome scaffolds each, with a map-wide average of 27.64 scaffolds/linkage group. Individual genome scaffolds contained between 1 and 31 GBS markers appearing in our final map, with an average of 3.50 markers per scaffold. GBS markers located on the same genomic scaffold co- localized to a single linkage group 87.3% of the time. This suggests that our recombination mapping successfully recapitulated the true genomic positions of the markers used to construct our map.
Improved linkage mapping resources in Astyanax
We sought to compare our linkage map with maps previously published by Gross et al.
(2008) and O’Quin et al. (2013) that also examined synteny between Astyanax and Danio.
Metrics such as the number of linkage groups, total map length, number of markers and marker density are commonly used to compare linkage maps within species. Both our GBS-based map and the RAD-seq and microsatellite-based map published by O’Quin et al. (2013) consist of 25 linkage groups, matching the Astyanax mexicanus karyotype number of 25. The microsatellite- based map presented by Gross et al. (2008) contains 28 groups (Table 3.2). While our map is of comparable length, it represents a dramatic increase in marker number (+559% v. Gross et al.
2008; +320% v. O’Quin et al. 2013) and marker density (+473% v. Gross et al. 2008; +279% v.
! 76! O’Quin et al. 2013), relative to previously published linkage maps for this system. As a result, we saw a substantial increase in the number of syntenic links between our map and Danio
(+506% v. Gross et al. 2008; +453% v. O’Quin et al. 2013) and an increase in the number of unplaced Astyanax scaffolds that can be anchored to our map (+263% v. Gross et al. 2008;
+171% v. O’Quin et al. 2013).
Our map contains a total of 784 links between our linkage groups and the Danio rerio genome and an average of 30.84 links (minimum = 14, maximum = 52) per Danio rerio chromosome (Table 3.3). This represents a considerable improvement over the results presented by Gross et al. (2008; 155 total links, average links per Danio chromosome = 6.20, minimum =
0, maximum = 15) and O’Quin et al. (2013; 173 total links, average links per Danio chromosome = 6.92, minimum = 1, maximum = 20). Additionally, while instances of synteny strongly represented in previous maps were also identified in this analysis, our map demonstrated increased representation of certain Danio chromosomes poorly represented in previous maps. For example, Gross et al. (2008) did not identify links between their map and Danio rerio chromosome 11, however, we identified 36 links between our map and chromosome 11.
Similarly, Danio chromosomes 17 and 19 are each represented once in the map of O’Quin et al.
(2013). We identified substantial links between these chromosomes and our linkage groups 9 (n
= 21) and 23 (n = 15), respectively.
Our linkage map uses an entirely different marker set than those used in previous maps.
Therefore, it was not possible to make direct comparisons with the linkage groups across prior studies. However, we could indirectly compare maps by examining connections between
Astyanax genomic scaffolds and each linkage map. We examined the five strongest syntenic links between single linkage groups in our GBS-based map and single Danio chromosomes and
! 77! then identified analogous connections between those chromosomes and specific linkage groups in the maps presented by Gross et al. (2008) and O’Quin et al. (2013).
Astyanax genomic scaffolds harboring markers associating each linkage group to a particular Danio chromosome were then compared (Table 3.4). We found that many of the identified Astyanax genomic scaffolds co-localize to putatively analogous linkage groups in both our GBS-based map and those of Gross et al. (2008) and/or O’Quin et al. (2013). However, in every case examined, our linkage groups were inclusive of a much higher number of Astyanax genomic scaffolds compared with prior studies. Thus, while the linkage groups in our map represent genomic intervals similar to those represented in prior maps, our map achieves a higher level of detail and resolution. These results also suggest that future mapping efforts in Astyanax may benefit by combining GBS marker discovery with those markers employed by Gross et al.
(2008) and O’Quin et al. (2013) to generate the most comprehensive linkage mapping resource.
High-density GBS-based linkage mapping will inform the Astyanax genome sequencing project
Preliminary Astyanax genomic resources enabled us to locate 64-bp, anonymous GBS markers, and assess the quality and reliability of our Astyanax linkage map. This emerging resource did not allow us to determine how well the 25 Astyanax chromosomes are represented in our map. However, these resources allowed us to determine if markers predicted to occur in the same genome scaffolds also co-occur in our GBS-based linkage map. Overall, we observed a high level of agreement between our linkage groups and one or more unplaced Astyanax genomic scaffolds.
In many cases, markers present on the same scaffold clustered together over a portion of a linkage group with little or no interruption from unplaced markers or markers from other
! 78! scaffolds (Figure 3.4). We expect these results will help inform chromosomal positions of scaffolds, given that linkage maps have been successfully used to augment genomic resources in other fish species, including several species of catfish (Liu 2011; Ninwichian et al. 2012), rainbow trout (Palti et al. 2011; Palti et al. 2012) and Atlantic salmon (Lorenz et al. 2010). We believe our high-density GBS-based map resources will both provide a resource for more refined
QTL analyses, and inform the genomic architecture of the Astyanax genome sequencing project.
CONCLUSIONS
We constructed a high-density linkage map for Astyanax mexicanus based on high- throughput genotyping-by-sequencing data. We leveraged emerging Astyanax genomic and transcriptomic resources and Danio rerio genomic and transcriptomic data to locate syntenic regions shared between our map and the Danio genome. These findings were based on the physical position of homologous (64-bp) GBS marker sequences. As expected, based on the significant divergence between these species, we recovered varying levels of synteny between portions of our Astyanax linkage groups and regions of the Danio genome. As a proof of concept, we successfully mapped a strong QTL associated with albinism, and demonstrated significant conserved genomic architecture in the regions surrounding the gene Oca2, between
Astyanax and Danio. We successfully anchored emerging Astyanax genomic information to our
GBS-based linkage map, identifying the putative location of thousands of anonymous GBS marker sequences within unplaced Astyanax genome scaffolds. This strategy revealed significant co-linearity between genomic scaffolds and our linkage map and demonstrates the utility of high- density, GBS-based linkage maps to inform and improve nascent genomic resources. Multiple comparisons with previously published maps suggest that our GBS-based map offers a higher
! 79! level of resolution and a greater number of connections between Astyanax and Danio genomes.
We hope that this resource and technology will accelerate the search and identification of genes mediating cave-associated traits in Astyanax, facilitate the genomic assembly for this system, and prove useful to other natural model systems of evolutionary and biomedical relevance.
ACKNOWLEDGEMENTS
The authors wish to thank Amanda Krutzler, Bethany Stahl and members of the Gross lab for valuable effort and input. We are also grateful to Wesley Warren, Suzanne McGaugh and the
Genome Institute at Washington University for providing access to the draft genome assembly
(Bioproject PRJNA89115 NCBI accession number APWO00000000; supported by NIH grant
R24 RR032658-01 to WW). Additionally, we would like to thank Suzanne McGaugh for providing BMC and other members of the Gross Lab with instruction in script-based BLAST search methods. This project was supported by National Institutes of Health (National Institute of
Dental and Craniofacial Research) grant DE022403 to JBG and a Cave Research Foundation
Graduate Student Research Grant to BMC.
REFERENCES
Borowsky, R., and H. Wilkens, 2002 Mapping a cave fish genome: polygenic systems and
regressive evolution. J. Hered. 93: 19-21.
Breder, C. M., Jr., 1943a Apparent changes in phenotypic ratios of the Characins at the type
locality of Anoptichthys jordani Hubbs and Innes. Copeia 1943: 26-30.
Breder, C. M., Jr., 1943b Problems in the behavior and evolution of a species of blind cave fish.
T. New York Acad. Sci. 5: 168-176.
! 80! Briggs, J., 2005 The biogeography of otophysan fishes (Ostariophysi: Otophysi): A new
appraisal. J. Biogeogr. 32: 287-294.
Broman, K. W., H. Wu, S. Sen and G. A. Churchill, 2003 R/qtl: QTL mapping in experimental
crosses. Bioinformatics 19: 889-890.
Burchards, H., Dölle, A., Parzefall, J. (1985). Aggressive behaviour of an epigean population of
Astyanax mexicanus (Characidae, Pisces) and some observations of three subterranean
populations. Behav. Proc. 11: 225-235.
Davey, J. W., P. A. Hohenlohe, P. D. Etter, J. Q. Boone, J. M. Catchen et al., 2011 Genome-
wide genetic marker discovery and genotyping using next-generation sequencing. Nat.
Rev. Genet. 12: 499-510.
Duboué, E. R., A. C. Keene and R. L. Borowsky, 2011 Evolutionary convergence on sleep loss
in cavefish populations. Curr. Biol. 21: 671-676.
Elipot, Y., H. Hinaux, J. Callebert and S. Retaux, 2013 Evolutionary shift from fighting to
foraging in blind cavefish through changes in the serotonin network. Curr. Biol. 23: 1-10.
Elshire, R. J., J. C. Glaubitz, Q. Sun, J. A. Poland, K. Kawamoto et al., 2011 A robust, simple
genotyping-by-sequencing (GBS) approach for high diversity species. PLoS One 6:
e19379.
Everett, M. V., M. R. Miller and J. E. Seeb, 2012 Meiotic maps of sockeye salmon derived from
massively parallel DNA sequencing. BMC Genomics 13: 521.
Gross, J. B., M. Protas, M. Conrad, P. E. Scheid, O. Vidal et al., 2008 Synteny and candidate
gene prediction using an anchored linkage map of Astyanax mexicanus. Proc. Natl. Acad.
Sci. U. S. A. 105: 20106-20111.
! 81! Gross, J. B., R. Borowsky, and C. J. Tabin, 2009 A novel role for Mc1r in the parallel evolution
of depigmentation in independent populations of the cavefish Astyanax mexicanus. PLoS
Genet. 5(1): e1000326. DOI: 10.1371/journal.pgen.1000326.
Gross, J. B., A. Furterer, B. M. Carlson, and B. A. Stahl, 2013 An integrated transcriptome-wide
analysis of cave and surface dwelling Astyanax mexicanus. PLoS ONE 8(2): e55659.
DOI: 10.1371/journal.pone.0055659.
Gross, J. B., A. J. Krutzler and B. M. Carlson, 2014 Complex craniofacial changes in blind cave-
dwelling fish are mediated by genetically symmetric and asymmetric loci. Genetics 196:
1303-1319.
Hecht, B. C., N. R. Campbell, D. E. Holecek and S. R. Narum, 2013 Genome-wide association
reveals genetic basis for the propensity to migrate in wild populations of rainbow and
steelhead trout. Mol. Ecol. 22: 3061-3076.
Hess, J. E., N. R. Campbell, M. F. Docker, C. Baker, A. Jackson et al., 2014 Use of genotyping-
by-sequencing data to develop a high-throughput and multi-functional SNP panel for
conservation applications in Pacific lamprey. Mol. Ecol. Resour. Advance online
publication. doi: 10.1111/1755-0998.12283.
Houston, R. D., J. W. Davey, S. C. Bishop, N. R. Lowe, J. C. Mota-Velasco et al., 2012
Characterisation of QTL-linked and genome-wide restriction site-associated DNA (RAD)
markers in farmed Atlantic salmon. BMC Genomics 13: 244.
Hubbs, C. L., and W. T. Innes, 1936 The first known blind fish of the family Characidae: A new
genus from Mexico. Occas. Pap. Mus. Zool. Univ. Mich. 342: 1-7.
Jeffery, W. R., 2001 Cavefish as a model system in evolutionary developmental biology. Dev.
Biol. 231: 1-12.
! 82! Jeffery, W. R., 2009 Regressive evolution in Astyanax cavefish. Annu. Rev. Genet. 43: 25-47.
Kirby, R. F., K. W. Thompson and C. Hubbs, 1977 Karyotypic similarities between the Mexican
and blind tetras. Copeia 1977: 578-580.
Kosambi D. D., 1943 The estimation of map distances from recombination values. Ann.
Eugenics 12:172-175.
Kowalko, J. E., N. Rohner, S. B. Rompani, B. K. Peterson, T. A. Linden et al., 2013 Loss of
schooling behavior in cavefish through sight-dependent and sight-independent
mechanisms. Curr. Biol. 23: 1874-1883.
Krzywinski, M., J. Schein, I. Birol, J. Connors, R. Gascoyne et al., 2009 Circos: An information
aesthetic for comparative genomics. Genome Res. 19: 1639-1645.
Larson, W. A., L. W. Seeb, M. V. Everett, R. K. Waples, W. D. Templin et al., 2014 Genotyping
by sequencing resolves shallow population structure to inform conservation of Chinook
salmon (Oncorhynchus tshawytscha). Evol. Appl. 7: 355-369.
Li, C., G. Waldbieser, B. Bosworth, B. H. Beck, W. Thongda et al., 2014 SNP discovery in wild
and domesticated populations of blue catfish, Ictalurus furcatus, using genotyping-by-
sequencing and subsequent SNP validation. Mol. Ecol. Resour. Advance online
publication. doi: 10.1111/1755-0998.12272.
Liu, Z., 2011 Development of genomic resources in support of sequencing, assembly, and
annotation of the catfish genome. Comp. Biochem. Physiol. D 6: 11-17.
Lorenz, S., S. Brenna-Hansen, T. Moen, A. Roseth, W. S. Davidson et al., 2010 BAC-based
upgrading and physical integration of a genetic SNP map in Atlantic salmon. Anim.
Genet. 41: 48-54.
! 83! Lu, F., A. E. Lipka, J. Glaubitz, R. Elshire, J. H. Cherney et al., 2013 Switchgrass genomic
diversity, ploidy, and evolution: Novel insights from a network-based SNP discovery
protocol. PLoS Genet. 9: e1003215.
McGaugh, S. E., J. B. Gross, B. Aken, M. Blin, R. Borowsky et al., 2014 The cavefish genome
reveals candidate genes for eye loss. Nat. Commun. 5: 5307.
Ninwichian, P., E. Peatman, H. Liu, H. Kucuktas, B. Somridhivej et al., 2012 Second-generation
genetic linkage map of catfish and its integration with the BAC-based physical map. G3-
Genes Genom. Genet. 2: 1233-1241.
O'Quin, K. E., M. Yoshizawa, P. Doshi and W. R. Jeffery, 2013 Quantitative genetic analysis of
retinal degeneration in the blind cavefish Astyanax mexicanus. PLoS One 8: e57281.
Ogden, R., K. Gharbi, N. Mugue, J. Martinsohn, H. Senn et al., 2013 Sturgeon conservation
genomics: SNP discovery and validation using RAD sequencing. Mol. Ecol. 22: 3112-
3123.
Palti, Y., C. Genet, G. Gao, Y. Hu, F. M. You et al., 2012 A second generation integrated map of
the rainbow trout (Oncorhynchus mykiss) genome: Analysis of conserved synteny with
model fish genomes. Mar. Biotechnol. 14: 343-357.
Palti, Y., C. Genet, M. C. Luo, A. Charlet, G. Gao et al., 2011 A first generation integrated map
of the rainbow trout genome. BMC Genomics 12: 180.
Pottin, K., Hinaux, H., Rétaux, S., 2011 Restoring eye size in Astyanax mexicanus blind cavefish
embryos through modulation of the Shh and Fgf8 forebrain organising centres.
Development 138: 2467-2476.
Protas, M., M. Conrad, J. B. Gross, C. Tabin and R. Borowsky, 2007 Regressive evolution in the
Mexican cave tetra, Astyanax mexicanus. Curr. Biol. 17: 452-454.
! 84! Protas, M., I. Tabansky, M. Conrad, J. B. Gross, O. Vidal et al., 2008 Multi-trait evolution in a
cave fish, Astyanax mexicanus. Evol. Dev. 10: 196-209.
Protas, M. E., C. Hersey, D. Kochanek, Y. Zhou, H. Wilkens et al., 2006 Genetic analysis of
cavefish reveals molecular convergence in the evolution of albinism. Nat. Genet. 38:
107-111.
Rowe, H. C., S. Renaut and A. Guggisberg, 2011 RAD in the realm of next-generation
sequencing technologies. Mol. Ecol. 20: 3499-3502.
Şadoǧlu, P., 1956 A preliminary report on the genetics of the Mexican cave characins. Copeia
1956: 113-114.
Salin, K., Voituron, Y., Mourin, J., Hervant, F., 2010 Cave colonization without fasting
capacities: An example with the fish Astyanax fasciatus mexicanus. Comp. Biochem.
Physiol. A 156: 451-457.
Schemmel, C., 1980 Studies on the genetics of feeding behaviour in the cave fish Astyanax
mexicanus f. anoptichthys. An example of apparent monofactorial inheritance by
polygenes. Z. Tierpsychol. 53: 9-22.
Yamamoto, Y., D. W. Stock and W. R. Jeffery, 2004 Hedgehog signalling controls eye
degeneration in blind cavefish. Nature 431: 844-847.
Yoshizawa, M., Y. Yamamoto, K. E. O'Quin and W. R. Jeffery, 2012 Evolution of an adaptive
behavior and its sensory receptors promotes eye regression in blind cavefish. BMC Biol.
10: 108.
! 85! FIGURE LEGENDS
Figure 3.1: A GBS-based linkage map in the Mexican cave tetra, Astyanax mexicanus. We analyzed 3003 SNP markers in 170 individuals using genotyping-by-sequencing technology.
This linkage map consists of 2235 markers in 25 linkage groups (A. mexicanus karyotype number = 25), spanning a total distance of 2110.7 cM (mean inter-marker distance = 1.052 cM).
Astyanax linkage group 8 (red box) illustrates typical marker density observed in most groups.
This group consists of 52 GBS markers spanning 67.061 cM with a mean inter-marker distance of 1.315 cM.
Figure 3.2: Short GBS sequences identify syntenic stretches between two Ostariophysian freshwater fish species. To reveal syntenic regions between Astyanax mexicanus and Danio rerio, we first identified stretches of the Danio genome harboring homologous sequences to our anonymous GBS marker sequences (A). Individual 64-bp sequences were compared with the
Danio genome both directly and by first aligning to larger Astyanax genomic scaffolds and predicted gene transcripts (B) followed by alignment of some or all of the larger sequence to the
Danio genome based on BLAST sequence analysis (C). This resulted in identification of homologous sequences for 784 Astyanax GBS markers within the Danio genome. The markers shared between Danio chromosomes and Astyanax linkage groups are represented using an
Oxford plot (D).
Figure 3.3: Whole-genome synteny between Astyanax and Danio and a proof-of-concept analysis of albinism. Syntenic links between our GBS map and the Danio genome were visualized using Circos (A). Each line represents a connection between the position of a
! 86! particular marker in our linkage map (black; scale in cM) and a homologous sequence in Danio
(various colors; scale in Mb). We scored albinism, a Mendelian trait associated with the Oca2 gene in cave-dwelling Astyanax (C), and performed QTL analysis using R/qtl. Each of three mapping methods (MR in red; EM in blue; HK in black) revealed peak LOD scores of ~20 (LOD at 0.001α threshold = 6.75) at, or adjacent to, GBS marker TP71406 on Astyanax linkage group
13 (B). Homologous sequences to TP71406 and several of its neighbors on Astyanax linkage group 13 are clustered together on Danio chromosome 6 near the Oca2 gene. A phenotypic effect plot for marker TP71406 revealed the predicted association between the homozygous
“cave” condition (genotype CC) and albinism in F2 individuals (D).
Figure 3.4: Colinearity between Astyanax linkage groups and genome scaffolds. We visualized the “anchoring” of seven unplaced Astyanax genome scaffolds (various colors) to linkage group 23 (black) in our Astyanax linkage map. For clarity, only scaffolds harboring ≥4
GBS markers were included. Scaffolds correspond to discrete, colinear sections of the linkage group with minimal overlap. The linear arrangement of markers is largely preserved between the scaffold and the linkage group. The scale for Astyanax scaffolds is in Mb; linkage group 23 is shown in cM.
! 87!
Table 3.1: Summary of BLAST results and identification of markers used in Astyanax-to-Danio syntenic analysis.
Astyanax GBS Markers to Astyanax Genome GBS Markers to Transcriptome to GBS Markers to Astyanax to Danio Astyanax Danio Danio Genomea Genomeb Genomec Transcriptomed Transcriptomee
Total Number BLAST Queries 2235 2235 2088 2235 572
BLAST Result Categories
Single robust hit 14 (0.6%) 1838 (82.2%) 255 (12.2%) 508 (22.7%) 110 (19.2%) Top hit, with positional support 0 (0.0%) 173 (7.7%) 92 (4.4%) 15 (0.7%) 120 (21.0%) Top hit, without positional support 10 (0.4%) 71 (3.2%) 138 (6.6%) 60 (2.7%) 61 (10.7%) Not top hit, with positional support 2 (<0.1%) 6 (0.3%) 108 (5.2%) 2 (<0.1%) 7 (1.2%) Unresolved 4 (0.2%) 14 (0.6%) 4 (0.2%) 12 (0.5%) 0 (0.0%) No result 2205 (98.7%) 133 (6.0%) 1491 (71.4%) 1638 (73.3%) 274 (47.9%)
Identified Syntenic Markers Between 26 N/A 593 N/A 298 Astyanax and Danio aResults of 64-bp GBS markers BLASTed directly to the Danio rerio genome. bResults of 64-bp GBS markers BLASTed directly to the Astyanax genome draft assembly. cResults of ~2-kb genomic intervals harboring 64-bp GBS markers BLASTed to the Danio rerio genome. dResults of 64-bp GBS markers BLASTed directly to the Astyanax predicted transcriptome. eResults of Astyanax transcripts harboring 64-bp GBS markers BLASTed to the Danio rerio transcriptome.
! 88! Table 3.2: Comparison of Astyanax linkage maps and syntenic studies with Danio rerio.
Gross et al. O'Quin et al. Current 2008 2013 Analysis
Total Number of Linkage Groups 28 25 25
Total Number of Genomic Markers 400 698 2235 Linkage Map Length 1783 cM 1835.5 cM 2110.7 cM
Marker Density 0.224 per cM 0.380 per cM 1.06 per cM
Microsatellite + Genotyping-By- Marker Type Microsatellite RAD-seq Sequencing Number of Astyanax Genomic Scaffolds 227 350 598 Represented by Map
Number of Syntenic Markers Identified 155 173 784 Between Astyanax and Danio
! 89! Table 3.3: Comparison of syntenic analyses between Astyanax linkage maps and their association with the Danio rerio genome across multiple studies.
Gross et al. 2008 O'Quin et al. 2013 Current Analysis
Number of Number of Number of represented represented represented Number of Represented Astyanax Number of Represented Astyanax Number of Represented Astyanax Danio rerio syntenic linkage genome syntenic linkage genome syntenic linkage genome chromosome linksa group(s)b scaffoldsc links group(s) scaffolds links group(s) scaffolds 4, 5, 9, 18, 1, 5, 8, 9, 12, 13, 1 13 5, 8, 21 7 15 11 42 26 21, 23 14, 18, 19, 21, 25 7, 12, 13, 1, 3, 5, 14, 16, 17, 2 6 2, 14, 15, 22 2 6 5 23 20 16, 23 18, 19, 22, 24, 25 2, 6, 10, 12, 13, 3 6 1, 4, 19 6 4 4, 15, 25 3 28 14, 15, 19, 22, 23, 23 25 3, 5, 7, 9, 12, 14, 4 3 6, 7 2 4 3 4 24 17 20, 24 1, 5, 9, 10, 2, 8, 16, 17, 2, 3, 7, 8, 9, 11, 5 15 9 13 11 22 17 20 19 15, 19, 22, 24, 25 1, 2, 11, 16, 4, 6, 12, 13, 16, 6 9 4, 13 4 20 16 37 23 18 17, 22, 24 17, 22, 24, 13, 22, 23, 4, 5, 7, 9, 10, 11, 7 11 10 6 6 31 25 26 25 13, 16, 20, 24 3, 6, 7, 9, 10, 13, 8 4 9, 12 4 7 7, 14, 17 6 42 14, 20, 21, 22, 23, 29 24 1, 2, 5, 9, 12, 13, 9 5 3, 17 4 8 10, 11 7 31 23 15, 19, 22, 25 3, 5, 7, 16, 20, 23, 10 3 17, 18 3 4 8, 10, 14 4 14 12 25 3, 4, 5, 7, 9, 15, 11 0 - 0 5 14, 17, 22 5 36 17, 18, 21, 22, 24, 24 25
! 90! 1, 2, 3, 4, 5, 8, 11, 12 7 10, 16 4 6 24 4 26 12, 13, 14, 15, 18, 18 20 1, 2, 4, 7, 10, 13, 13 11 1, 5 6 7 4, 12 4 34 21 14, 21, 22, 25 2, 5, 8, 10, 13, 14, 14 6 6, 7 4 9 3, 6, 15, 19 6 24 16 24 1, 2, 6, 9, 13, 16, 15 5 2 5 8 1, 7, 12, 14 6 27 17 20, 24, 25 2, 8, 10, 11, 16, 16 3 13 3 7 8, 19 7 23 17 17, 19, 20, 22, 23 1, 2, 3, 5, 6, 9, 10, 17 6 3, 23 2 1 20 1 52 12, 14, 15, 20, 22, 34 23 3, 5, 6, 7, 13, 14, 18 8 11 3 7 5 3 32 15, 16, 18, 19, 21, 26 24 5, 6, 9, 10, 11, 15, 19 3 19 3 1 25 1 42 28 17, 20, 23, 25 2, 3, 9, 13, 14, 18, 20 7 1, 2 7 3 1, 2 3 21 15 19, 21 21 3 15, 17 1 6 2, 7 6 16 12, 14, 18, 19 12 1, 3, 5, 6, 7, 9, 10, 22 4 12, 20 4 7 14, 18, 22 6 42 11, 12, 13, 14, 15, 32 16, 18, 19, 22, 23 14, 16, 18, 3, 4, 7, 10, 14, 15, 23 6 26 3 7 6 33 17 25 16, 22 1, 2, 3, 7, 10, 11, 24 8 1, 13, 15 6 9 2, 3, 8, 11 8 38 25 13, 15, 16, 18, 23 3, 5, 7, 8, 9, 13, 25 3 6, 7 3 3 3, 6 3 31 19 16, 17, 20, 24 Bold type indicates that a listed linkage group harbors five or more links with a given Danio chromosome. aIndicates the number of syntenic links identified between Astyanax linkage maps and each listed Danio rerio chromosome. bIndicates the identity of syntenic Astyanax linkage groups harboring syntenic links with each listed Danio rerio chromosome. 3Indicates the number of Astyanax genome scaffolds anchored to a given linkage map that harbor connections with a given Danio rerio chromosome.
! 91! Table 3.4: Representative analysis of linkage group equivalence and quality based on highly syntenic chromosomes in Danio rerio and linkage groups in Astyanax mexicanus.
Gross et al. 2008 O’Quin et al. 2013 Current Analysis Identity of Identity of Identity of Principal represented Principal represented Principal represented represented Number of Astyanax represented Number of Astyanax represented Number of Astyanax Danio rerio linkage syntenic genome linkage syntenic genome linkage syntenic genome chromosome groupa linksb scaffoldsc group links scaffolds group links scaffolds KB882256.1, KB871670.1, KB882253.1, KB871811.1, KB882235.1, KB871878.1, KB882230.1, KB871811.1, KB872044.1, KB882185.1, KB882115.1, KB872200.1, KB882171.1, KB882122.1, KB882115.1, KB882161.1, 6 4 8 1 16 13 28 KB882161.1, KB882120.1, KB882152.1, KB882172.1, KB882122.1, KB882122.1, KB882176.1 KB882161.1, KB882120.1, KB882172.1, KB882115.1, KB882176.1, KB882082.1, KB882185.1 KB872595.1, KB871670.1
KB882289.1, KB882113.1, KB871601.1, KB882105.1, KB871816.1, KB871607.1, KB872252.1, 8 9 3 KB871923.1, 17 5 KB871684.1, 6 17 KB871939.1, KB882105.1 KB871923.1, KB871817.1, KB872214.1 KB871684.1, KB871601.1, KB871595.1
! 92! KB882261.1, KB882210.1, KB882154.1, KB871819.1, KB882118.1, KB872081.1, KB881455.1, KB882109.1, KB872296.1, KB872296.1, 13 5 9 4 6 25 17 KB882107.1, KB882107.1, KB882107.1, KB872296.1, KB882118.1, KB872081.1 KB872081.1, KB882125.1 KB871838.1, KB871652.1, KB871591.1
KB882265.1, KB882243.1, KB882233.1, KB882179.1, KB882084.1, KB882158.1, 17 23 5 KB882233.1, 20 1 KB882265.1 9 21 KB882153.1, KB882265.1 KB882117.1, KB882084.1, KB872047.1, KB871726.1, KB871695.1
KB882214.1, KB882138.1, KB872166.1, KB882128.1, KB882098.1, KB880082.1, KB882102.1, 23 26 6 18 4 KB882102.1, 15 20 KB882102.1, KB882098.1, KB882128.1 KB882242.1 KB872132.1, KB872075.1, KB871985.1 Bold lettering indicates genomic scaffolds containing syntenic markers on the principal represented linkage group in our GBS-based map and one or more previous maps. Italic lettering indicates scaffolds that contain a sytenic marker in the GBS-based map and are associated with the principal linkage group(s) in previous map(s), but do not contain a syntenic marker (and vice versa). aIndicates the most common (i.e., “principal”) linkage group anchoring to the indicated Danio rerio chromosome. bIndicates the number of points of synteny between the principal linkage group from this paper and the indicated Danio rerio chromosome. cLists the identity of Astyanax genomic scaffolds to which each point of synteny identified.
! 93! FIGURES
Figure 1.
! 94! Figure 2.
! 95! Figure 3.
! 96! Figure 4.
! 97!
CHAPTER FOUR
Genetic analysis reveals candidate genes potentially underlying altered activity profiles in the blind Mexican tetra, Astyanax mexicanus
Brian M. Carlson, Ian B. Klingler*, Bradley J. Meyer* and Joshua B. Gross*
Department of Biological Sciences University of Cincinnati Cincinnati, OH 45221-0006
*The analysis and narrative presented here are solely the work of Brian M. Carlson. It is expected that this chapter will be adapted and submitted for publication with the listed co-authors providing assistance and input at that time.
! 98! ABSTRACT
Animal models provide useful tools for the examination of the genetic basis of morphological, physiological and behavioral phenotypes of interest. Cave-adapted species serve as powerful natural models in which to examine a wide array of phenotypic changes with evolutionary, developmental and clinical relevance. In this study, we explored the genetic underpinnings of previously characterized differences in locomotor activity patterns between the surface and Pachón cave populations of Astyanax mexicanus. We successfully identified multiple novel QTL underlying patterns in both overall levels of activity (velocity), as well as how the fish used space within the trial tank over time (time spent near the top or bottom of the tank).
Further, we demonstrate that observed patterns in velocity and tank usage are likely mediated by different regions of the genome. We interrogated eight genomic intervals underlying observed
QTL distributed across six different linkage groups and employed available genomic and transcriptomic data to generate and evaluate a list of 36 potential candidate genes. Interestingly, our data support the candidacy of a number of genes, but do not suggest that differences in the patterns of behavior observed here are the result of alterations to teleost multiple tissue opsins, melanopsins or members of the core circadian clockwork, as has described in other species. This study expands our knowledge of the genetic architecture underlying patterns in activity in this system and informs future investigations into the role of specific genes in shaping complex behavioral phenotypes in Astyanax and other vertebrate taxa.
! 99! INTRODUCTION
Since its discovery (Hubbs and Innes 1936), the blind Mexican tetra, Astyanax mexicanus, has proven to be an excellent system for studying the genetic basis of both simple and complex traits (Gross et al. 2015). The presence of 29 different cave-adapted populations that can be interbred with extant surface populations to generate viable hybrid offspring has made Astyanax a valuable model in which to investigate the evolutionary and developmental processes leading to the suite of regressive and constructive changes observed in these and other cave organisms (Jeffery 2001; Jeffery 2009). Increasingly, however, strong arguments are being made for the use of Astyanax and similar species as natural models for the study of phenotypes with human medical implications (Albertson et al. 2009), such as craniofacial defects (Gross et al. 2014) and diseases affecting the eye and retina (O'Quin et al. 2013), as well as non- morphological traits, such as alterations in biological rhythmicity (Menna-Barreto and Trajano
2015), sleep (Duboué et al. 2011; Duboué and Borowsky 2012) and behavior (Elipot et al. 2013;
Yoshizawa et al. 2015).
“Simpler” organisms have proven valuable in efforts to understand the genetics underlying human behavior (Kendler and Greenspan 2006) and fish species provide an important vertebrate model for understanding neural development, function and disease (Guo 2004). As a first step toward understanding the genetic basis of behavior in any species, it is critical to identify regions of the genome harboring genes or regulatory sequences that impact phenotypes of interest, thereby narrowing the search to one or more critical genetic intervals that may be more thoroughly examined. One way to accomplish this is through the aid of quantitative trait locus (QTL) analysis (Flint 2003).
! 100! Over the past decade, QTL analysis has successfully identified genomic intervals associated with a variety of behavioral traits in a number of fish species. For example, studies have demonstrated QTL for boldness in zebrafish (Wright et al. 2006a; Wright et al. 2006b), feeding, exploration, risk taking and schooling in species of stickleback (Greenwood et al. 2013;
Laine et al. 2014; Greenwood et al. 2015), startle response in medaka (Tsuboko et al. 2014) and anti-predator behavior, response to crowding stress and behaviors related to spawning and migration in rainbow trout (Colihueque et al. 2010; Hecht et al. 2012; Rexroad et al. 2013;
Christensen et al. 2014). Recent studies in Astyanax have identified QTL for several behavioral traits, including feeding angle (Kowalko et al. 2013a), schooling behavior (Kowalko et al.
2013b), vibration attraction (Yoshizawa et al. 2012b; Yoshizawa et al. 2015) and locomotor activity (Yoshizawa et al. 2015).
In this study, we investigated the genetic basis of differences in activity patterns between the surface and cave morphotype of Astyanax mexicanus. Analysis of data from 24hr assays of locomotor activity in an F2 surface x cavefish hybrid pedigree revealed multiple QTL associated with metrics for levels of overall activity, as well as spatial components of locomotor activity patterns. We leveraged available genomic and transcriptomic data to screen genes in the genomic intervals putatively underlying these QTL and generated a set of potential candidates for further study. Our results indicate several genes that may play a role in mediating changes in locomotor activity and related behaviors.
! 101! MATERIALS AND METHODS
Fish and animal husbandry
Fish used in these studies were part of a laboratory population maintained at the
University of Cincinnati. All specimens were laboratory-bred Astyanax mexicanus belonging to lines originally sourced from the Sierra de El Abra region of northeastern Mexico. F1 and F2 hybrids were generated by crossing male surface fish with female (Pachón) cavefish. All fish used in 24hr activity assays and subsequent QTL analyses were adult fish generously provided to our lab by Dr. Richard Borowsky (New York University).
All fish were kept in a dedicated animal room maintained at 21.7 ± 1˚C under a 12:12hr light/dark cycle (~160 lux/0 lux). Surface, cave and F1 hybrid individuals used in this study were housed in 5-gallon aquaria on our animal husbandry system as described elsewhere (Gross et al.
2013). F2 hybrid individuals belonged to a single F2 mapping pedigree (“Asty66”; n = 129) and were reared individually, at room temperature, in 1L tanks filled with water from our husbandry system. Individual housing facilitated identification of individual fish so that phenotypic data could be linked to previously collected genotypic data for each specimen. All fish were fed
TetraMin tropical flake food (Tetra) once per day.
All husbandry procedures and experimental protocols were approved by the Institutional
Animal Care and Use Committee (IACUC) of the University of Cincinnati (Protocol Number 10-
01-21-01).
Automated video tracking
The 24hr activity assays described in this study were conducted in the same manner and using the same apparatus as described elsewhere (see Chapter 2), except as follows: Fish were
! 102! placed in plastic tanks (30.1cm wide x 12.2cm deep at rim, 28.0cm wide x 10.7cm deep at base,
22.5cm high) filled with approximately 4.5L of water from our husbandry system. These tanks were covered on the back, base and sides by a thin layer of opaque, white mylar to facilitate even distribution of the infrared (IR) lighting used. One fish was assayed during each 24hr trial. Trials began at 15min intervals between zeitgeber time (ZT) 06:00 and 10:00; this allowed time to reset the assay rig between trials, thereby enabling trials to be conducted on successive days.
Descriptive metrics and statistics
Three metrics (mean for the entire trial, subjective day mean and subjective night mean) were used to describe velocity (cm/s) and zone usage (s/900s time bin) over the course of a trial as a whole, or specifically during periods when the lights were on or off (ZT 00:00-12:00 and ZT
12:00-00:00, respectively). Evaluations and metrics relative to circadian rhythmicity (as used elsewhere, see Chapter 2) were not included in this study as the 24hr duration of trials did not allow for statistical validation of fit cosine functions or testing of 24hr time-lag autocorrelation.
Instead, a fourth metric, the difference between the mean value for subjective day and subjective night, was used. This metric provides a rough approximation of the relationship between values under light and dark conditions. For example, a low-amplitude diurnal rhythm would result in a small positive value for this metric, while a high-amplitude nocturnal rhythm would result in a larger, negative value.
The presence of albinism in F2 hybrids was scored as a binary trait based on visual assessment, as was the presence of an eye on the right and left sides of the head. Sex was also scored as a binary trait (1 = female, 0 = male) based on visual assessment of body shape and anal fin morphology as described elsewhere (Borowsky 2008). The sizes of the right and left eye and
! 103! right and left pupil (in pixels) were measured from images (7.81x magnification) using ImageJ
(National Institutes of Health, Bethesda, MD) according to methods described elsewhere (Gross et al. 2014). Statistical comparisons of velocity and tank usage data between groups based on sex, albinism and presence/absence of eyes were made using Wilcoxon-Mann-Whitney tests in
JMP Pro 11 (SAS Institute Inc., Cary, NC).
Genotyping and quantitative trait locus analyses
Using genomic DNA isolated from tail fin clips (as previously described; Gross et al.
2013), all members of the Asty66 F2 hybrid pedigree were genotyped during a single round of genotyping-by-sequencing (GBS) conducted at the Cornell University Institute of Biotechnology according to methods described elsewhere (Elshire et al. 2011; Lu et al. 2013). Discovery of anonymous SNP markers (contained within 64bp sequences) was completed as part of the GBS procedure. Genotype data for these and other fish were reviewed, suitable markers were identified and a 2235 marker, high-density GBS-based linkage map for Astyanax mexicanus was constructed as described elsewhere (Carlson et al. 2015). All genetic analyses described in this study are based on this map and genotypic data set.
QTL analyses were conducted in R (R Project for Statistical Computing, http://www.r- project.org/) using the R/qtl package (Broman et al. 2003). Traits with binary or normal distributions were analyzed using each of three scan-one mapping methods: marker regression
(MR), expectation maximization (EM), and Haley-Knott (HK), as earlier described (Gross et al.
2014). Continuous data with non-normal distributions were subjected to a non-parametric (NP) scan-one analysis. Permutation tests (1000 permutations) were used to establish 0.05α LOD
! 104! significance thresholds for each phenotype-mapping method combination, as well as a genome- wide significance threshold.
Comparative genomic and transcriptomic analyses
For each of the activity QTL identified, the region of the linkage map associated with the
QTL peak was defined as beginning at the marker with the highest LOD value and extending in both directions until either LOD values dropped to a level consistently below the significance threshold, in which case the first marker below the significant LOD value was included, or the end of the linkage group had been reached. In cases where multiple methods returned the same peak marker (or a pseudomarker for which the closest genotyped marker was the same) for a given trait, LOD scores generated using the method that returned the QTL peak with the highest
LOD score were used to determine this interval. For several QTL, LOD scores dipped below significance for portions of the linkage group, but then “peaked” again further away, suggesting the presence of two or more distinct QTL on a single linkage group. These “secondary peaks” were treated as if they had been separately identified, rather than treating all significant LOD scores as part of a single peak, which would require inclusion of long stretches of markers without significant association with the phenotype of interest.
Earlier BLAST analyses anchored 598 unplaced, annotated genome scaffolds (McGaugh et al. 2014) to the GBS-based Astyanax linkage map by identifying putative locations for 2091 of the 2235 GBS markers of which the map was comprised (Carlson et al. 2015). This information facilitated the identification of portions of the genome associated with the intervals of the linkage map identified as containing activity QTL. The list of genes harbored by these scaffolds was then used as a starting point for candidate gene identification.
! 105! BioMart (v. 0.7) was used to query the Ensembl genome browser (release 78; www.ensembl.org) and retrieve gene ontology (GO) terms associated with each of the genes found on Astyanax genomic scaffolds anchored to the region of our linkage map associated with each identified QTL peak. In cases where ZFIN IDs for zebrafish homologs were provided, GO annotations for that species were queried as well. Genes annotated using one or more relevant
GO terms were considered provisional candidates and subjected to further analysis.
Gene expression was examined using RNA-seq data. This data set included a developmental series consisting of pooled samples from both cave and surface fish at 10 hours,
24 hours, 1.5 days and 3 days post-fertilization (50 embryos per sample, three technical replicates), as well as pooled samples (three biological replicates, two technical replicates) from juvenile (3-4 month old) cave and surface fish raised under both 12:12hr light/dark conditions and in total darkness (see description of dark-reared fish in Chapter 2). Total RNA was isolated using an RNeasy Kit (Qiagen; Valencia, CA) and samples were sequenced by the Cincinnati
Children’s Hospital Medical Center DNA Sequencing and Genotyping Core facility using
Illumina HiSeq technology (v. 2 kit). SeqMan NGen (v. 11; DNAstar, Madison WI) was used to align RNAseq reads to an Astyanax transcriptome template and ArrayStar (v. 11; DNAstar,
Madison WI) was used to normalize read counts using the RPKM method (Mortazavi et al.
2008), calculate fold change comparisons between samples and determine statistical significance using the Student t-test controlled for false-discovery rate (Benjamini and Hochberg 1995).
Differences in gene expression between surface and cavefish for genes with relevant GO annotations were then examined.
RNA-seq reads from cave and surface fish were also aligned against the Astyanax reference genome (McGaugh et al. 2014) using SeqMan NGen (v. 11; DNAstar, Madison WI).
! 106! For each QTL examined, SNP reports were generated for all scaffolds anchored to the associated interval. Results were then filtered to limit potential variants to those where alleles segregated between cave (Pachón) and surface in all (or nearly all; >90%) reads. Variant calls in genes with relevant GO annotations were examined and manually verified.
RESULTS AND DISCUSSION
Cave x surface hybrid individuals demonstrate diverse patterns of locomotor activity
Previously, we demonstrated that the surface and cave (Pachón) morphotypes of
Astyanax mexicanus exhibit differences in patterns of overall activity, as well as in the spatial component of locomotor behavior (see Chapter 2). Given that these differences were observed at the population level, we hypothesized that the activity profiles observed had a heritable genetic basis. In order to test this, we assayed a small number of surface, cave and F1 hybrid fish for 24hr under 12:12hr light/dark conditions. Generally speaking, F1 individuals displayed overall patterns similar to those seen in surface fish, but with “cave-like” influences, such as higher mean velocity and a less extreme bias toward usage of the bottom zone (Fig. 4.1). This mirrors the fact that, morphologically, these F1 hybrids look very similar to surface fish, but with subtle differences that make them slightly more cave-like (e.g., slightly smaller eyes).
We then assayed a large F2 hybrid population under the same conditions. F2 surface x cavefish hybrid pedigrees typically include some individuals that look very similar to surface fish, some that look very similar to cavefish and a large number of individuals that have intermediate phenotypes resulting from various combinations of cave-like and surface-like traits.
In the same way, we observed a wide range of locomotor activity patterns in our F2 pedigree.
Some fish displayed activity profiles very similar to one of the parental populations, others
! 107! showed a clear combination of elements of the behavior of both surface and cavefish and many showed patterns that were not recognizable as belonging to either morphotype (Fig. 4.2). These results are indicative of the complex nature of locomotor activity and suggest that the patterns of activity (or lack thereof) displayed by individual specimens are influenced by a number of different genetic loci.
Genetic analyses reveal a complex genetic basis for activity differences between cave and surface fish
In order to determine whether a genetic basis for any of the elements of the activity profiles observed could be identified, we conducted QTL analyses using the data from members of our F2 surface x cave (Pachón) hybrid pedigree, each assayed for 24h under a 12:12 light/dark cycle (Table 4.S1; genotypic data provided in Carlson et al. 2015). Our analysis revealed a number of putative associations between regions of our linkage map and metrics for both mean velocity and tank usage (Table 4.1). At least one QTL was found for each of three different velocity metrics and five different tank usage metrics, with these QTL being distributed over a total of six linkage groups from our GBS-based linkage map (Carlson et al. 2015). In several instances, our efforts to determine the interval covered by a given QTL revealed that there were actually two or more distinct LOD “peaks” on the linkage group in question. In these cases, secondary peaks were investigated independently, as if they had been identified on different linkage groups.
Interestingly, our results suggest that while both mean velocity and tank usage are under genetic control, different regions of the genome mediate these aspects of locomotor behavior.
Linkage groups 2, 3, 13 and 15 each harbor QTL for at least one velocity metric; linkage groups
! 108! 14 and 25 each harbor QTL for at least one tank usage metric. This demonstrates the value of evaluating activity level (e.g., mean velocity) and the spatial component of locomotor behavior
(e.g., time spent in different zones) independently, rather than dealing with combined metrics
(Erckens and Weber 1976; Erckens and Martin 1982a; Erckens and Martin 1982b) or employing methods that do not account for the spatial component at all (Thines and Wolff-Van Ermengem
1965; Thines and Weyers 1978; Duboué et al. 2011; Beale et al. 2013; Yoshizawa et al. 2015).
Co-localizing QTL suggest potential relationships between patterns in locomotor behavior, sex and cave-associated traits
It has long been theorized that there is a link between regression of the visual system and loss of rhythmic behavior in cave species. In a recent survey of the relevant literature in over forty cave-adapted species, Friedrich (2013) provided support for this idea, noting that behaviorally arrhythmic species are almost universally “primary anophthalmic” species (i.e., species that never develop an eye); adult-specific and population-specific anophthalmic species appear to retain some level of behavioral rhythmicity. Acknowledging the potential role that eye loss and other cave-associated traits may play in influencing patterns in locomotor activity, recent studies have looked for associations between activity data and traits such as eye size, pupil size, thickness of the inner nuclear layer of the retina and albinism (Duboué et al. 2011;
Yoshizawa et al. 2015). Additionally, studies examining other behavioral phenotypes in this species have explored potential links between sex and the behaviors observed (Yoshizawa et al.
2012a; Elipot et al. 2013; Kowalko et al. 2013b). With the exception that tendency to school was shown to differ between sexes in a surface x cave (Tinaja) F2 hybrid population (Kowalko et al.
2013b), no other such links have been shown between these factors and behavior in this species.
! 109! In this study, we looked for overlap between our behavioral QTL and regions of the genome associated with presence/absence of eyes, eye size and pupil size on either side of the head, as well as albinism and sex (Fig. 4.3; Table 4.S2). As suggested by the results of previous studies (Duboué et al. 2011; Yoshizawa et al. 2015), we saw no correspondence between our behavioral results and QTL for the size of the eye or pupil on either the right or left side; there were no significant QTL for eye size and QTL for pupil size were confined to linkage group 20.
However, when scored as a binary trait, QTL for presence/absence of an eye on either side of the head were found on linkage group 3 near a QTL for mean velocity during subjective day.
Statistical analysis shows that there is a significant difference between the mean “day” velocity of F2 hybrids that have a right (z = 2.677, p = 0.0074) or left (z = 3.227, p = 0.0013) eye and those who do not. Similarly, the QTL for albinism previously described on linkage group 13 of this map (Carlson et al. 2015) has its peak LOD score at a marker very close to another QTL peak for mean “day” velocity; scores for this activity metric differ significantly between albino and pigmented members of our F2 pedigree (z = 3.417, p = 0.0006).
While the exact nature of the relationship between eye loss, albinism and the behavioral
QTL described in this study remains unclear, any association beyond simple linkage due to proximity is likely influenced by selective pressures acting on either the morphological trait, the associated behavioral trait, or both. Initially, it was theorized that there may be selective pressure to lose the morphological features rendered useless in the darkness, perhaps as a response to some advantage conferred by energy conservation (Poulson and White 1969; Culver 1982).
However, the current body of literature does not seem to support this hypothesis in the case of the loss/reduction of either eyes (Jeffery 2005) or pigmentation (Bilandžija et al. 2013). Instead, it appears more likely that the loss of eyes and pigmentation is associated with beneficial
! 110! morphological, physiological and/or behavioral changes and that positive selective pressures acting on the latter are responsible for indirect selection upon the former (see Wright 1964). For example, a link between eye reduction and increases in vibration attraction behavior has been proposed (Yoshizawa et al. 2012b), although the precise nature of this connection is a matter of some debate (Borowsky 2013). Additionally, Bilandžija et al. (2013) suggest that albinism may be the result of selection upon behavioral changes that result from down-regulation of oca2 and the associated increase in tyrosine and catecholamine levels that occurs in the absence of melanin synthesis. Therefore, potential links between changes in locomotor activity patterns such as those presented in this study and regressive cave-associated traits provide excellent opportunities to further examine the mechanisms underlying the constellation of constructive and regressive traits exhibited by these and other cave-adapted species.
In addition to potential links with cave-associated traits, the LOD peak of a robust QTL for sex on linkage group 15 is situated one marker away from the peak values of QTL for both mean velocity throughout entirety of the trial and mean velocity during subjective night.
Statistical analyses indicate a significant difference between male and female members of our F2 pedigree for both mean trial velocity (z = -2.751, p = 0.0059) and mean “night” velocity (z =
-3.362, p = 0.0008). While the link between patterns of locomotor activity and sex, if present, is noteworthy and should be kept in mind during experimental design, it is more likely that it is the result of sex-based differences in behavioral traits such as boldness (Dahlbom et al. 2011; Irving and Brown 2013; King et al. 2013; Ingley et al. 2014), rather than a consequence of some aspect of cave adaptation.
! 111! QTL for locomotor activity are distinct from those previously identified
A recent study by Yoshizawa et al. (2015) presented two distinct QTL associated with locomotor activity. The results presented in that study are based on the linkage map published by
O'Quin et al. (2013), which does not share any markers with the linkage map used in this study.
Direct comparison of QTL results between the current study and that of Yoshizawa et al. (2015), while desirable, is therefore impossible. However, it is possible to make inferences about the relationships between QTL presented in these two studies by using the draft Astyanax genome
(McGaugh et al. 2014) as a reference, as described elsewhere (Carlson et al. 2015). When the locations of genomic scaffolds that can be anchored to both maps are compared, clear relationships between the linkage groups comprising each of these two linkage maps emerge
(Table 4.S3). As a result, it appears quite likely that the QTL for locomotor activity reported on linkage groups 3 and 22 by Yoshizawa et al. (2015) correspond with regions of linkage groups
7/8 and 5, respectively, in the map employed here. Given that none of the QTL identified in this study are found on those linkage groups, it is reasonable to assume that the QTL described in these two studies represent the influence of different loci on patterns in locomotor activity.
Interestingly, the same method of comparison reveals that the QTL for eye size and vibration attraction behavior (VAB) that Yoshizawa et al. (2015) noted on linkage group 2 of the map that they employed may be related to one or more of the velocity QTL that we demonstrate on linkage group 3 of the map used here. Further, comparison of the original study by O'Quin et al. (2013) with the results presented here suggests that there may be a relationship between the
QTL identified for the thickness of the inner nuclear layer (linkage group 2) and the outer plexiform layer of the eye (linkage group 7) and the velocity and tank usage QTL described on linkage groups 3 and 14, respectively, in this study. However, greater caution must be exercised
! 112! in making such comparisons because even if the assumption that these QTL would occur on the same linkage groups is true, there is no way of knowing precisely where these QTL would be located within a given linkage group, relative to the QTL demonstrated here. Nonetheless, these comparisons provide interesting insights that may warrant further research.
Comparative genomic and transcriptomic analyses identify putative candidate genes mediating differential activity in cavefish
Having identified a number of genomic regions putatively associated with various elements of the activity profiles observed in our 24hr assays, we sought to explore the genes believed to be located in these intervals and to ascertain whether or not the available data support one or more of them as candidates potentially influencing the behaviors observed. We focused on genes underlying QTL peaks for which the effect plots showed a clear genetic effect; those
QTL that were not investigated typically had effect plots that showed a high degree of variability among specimens possessing two “surface” alleles. As a result, we examined a total of eight intervals in our map, all but one of which contained a primary QTL peak for at least one locomotor activity trait. These intervals ranged from 1.587 to 10.001 cM in length (mean = 4.286 cM) and contained between three and seven GBS markers (mean = 4.5 markers). The sequences for the markers in each interval were used to anchor between two and six genomic scaffolds
(mean = 3.375 scaffolds), thereby associating between 3.256 and 20.761 Mb (mean = 9.086 Mb) of genomic sequence and between 75 and 413 genes (mean = 192.75 genes) with each interval examined. We therefore began our analysis with an initial list of 1542 genes spread out over
72.685 Mb of the Astyanax genome.
! 113! In order to cut this list down to a more manageable size, we first screened for genes with relevant gene ontology (GO) terms. This initial screen reduced the list of potential candidates to
36 genes (Table 4.S4). These 36 genes were then examined for both sequence variation and differences in expression level. RNA-seq data revealed statistically significant differential expression in 27 of these genes, with the majority of expression differences occurring during development only (Table 4.S5). Only 3 genes showed significant differential expression in both juvenile samples and the developmental series and a single gene showed differential expression only in juveniles, however this disparity is likely influenced by the difference in the number of technical replicates between the developmental and juvenile samples. Analysis of sequence variation based on alignment of RNA-seq reads to the Astyanax draft genome (McGaugh et al.
2014) resulted in the identification of 16 genes that include exonic SNPs with alleles segregating between surface and Pachón cave samples. 13 genes possessed one or more variants in the coding sequence and 6 genes showed variation in the 5’ or 3’ untranslated regions (UTR; Table
4.S6). There were no obvious indels or splice variations in the genes examined. Of the 13 genes showing variation in the coding sequence, only 5 genes possessed non-synonymous changes.
Interestingly, 5 genes showed variation in regions annotated as introns, 3 of which are genes unique to this category. This is most likely due to either incorrect alignment or retained introns in uncharacterized splice variants. Taken together, these analyses provide some level of additional support for 29 of the 36 genes identified by screening GO terms. A summary of these results is provided in Table 4.2.
The approach employed here has certain limitations, namely that it cannot identify genes on genomic scaffolds that were not anchored to the examined intervals of our linkage map or genes that are not yet associated with relevant GO terms. Further, it does not enable us to
! 114! examine potential regulatory mutations in sequences outside of transcribed regions or differences in gene expression outside of the specific time points examined. However, while these limitations mean that our list of potential candidates is far from exhaustive, we believe that our analysis does provide evidence that further examination is warranted into the potential role that several genes may play in mediating the elements of locomotor activity examined in this study.
These particularly strong candidates are discussed below:
Regulator of G-protein signaling 4 (rgs4), found on a scaffold anchored to the region underlying the QTL on linkage group 13, was the only gene to be expressed at significantly different levels across all four time points in our developmental series. Our analysis suggests that rgs4 is likely overexpressed in cavefish, relative to surface fish, throughout development. In zebrafish, rgs4 has been shown to play a role in axonogenesis of neurons that play a role in motor activity and responses to touch (Cheng et al. 2013). In other species, it has also been shown to play a role in serotonin signaling (Gu et al. 2007) and modulation of melatonin receptor signaling in retinal ganglion cells (Ji et al. 2011). Given the importance of melatonin in sleep and circadian rhythmicity (Iuvone et al. 2005; Gandhi et al. 2015), the proposed role of serotonin signaling in the cavefish “behavioral syndrome” (Elipot et al. 2013; Elipot et al. 2014) and the differences in locomotor activity seen here, further investigation into the potential effects of elevated rgs4 in cavefish is warranted.
Similarly, atonal homolog 7 (atoh7), which is found on a scaffold anchored to the interval underlying the QTL on linkage group 25, stands out as a potential candidate due to the fact that, out of the five genes in which non-synonymous sequence variation was observed, atoh7 contained the only mutation in which cavefish display an amino acid change at a highly conserved amino acid residue (Figure 4.4). Additionally, this gene is significantly under-
! 115! expressed in cavefish at three days post-fertilization, as well in juveniles reared under 12:12hr light/dark conditions; no difference was seen between expression in surface and cavefish juveniles reared under constant darkness. In zebrafish, atoh7 (formerly known as lakritz or atonal homolog 5) has been shown to be essential for differentiation of retinal ganglion cells
(Kay et al. 2001). Functionally blind atoh7 null mutant zebrafish larvae also show small but significant defects in certain locomotor behaviors (Mirat et al. 2013). Further analysis could elucidate the effects, if any, of the atoh7 sequence variation and expression differences observed in Pachón cavefish.
Opsin 5 (opn5; neuropsin) is found on a scaffold anchored to the critical region underlying a QTL for mean day velocity on linkage group 3 and is significantly under-expressed in cavefish at 24hpf and 3dpf. The protein product of the opn5 gene is UV-sensitive in zebrafish
(Yamashita et al. 2014), although there is little information about the role that it plays in zebrafish or even where it is expressed. Manchenkov et al. (2015) noted “developmental phenotypic defects” in zebrafish subjected to morpholino knockdown of opn5, but they did not elaborate upon the nature of the observed defects. In other species, opn5 is present in the retina and skin, has been implicated in deep brain photoreception and is found in both the pineal gland and serotonin-positive cells in the hypothalamus (Yamashita et al. 2014; Haltaufderhyde et al.
2015). The presence of opn5 in the retina, pineal gland and serotonin-positive cells in the hypothalamus, together with the general lack of information about its expression patterns and function in fish, makes this gene a prime candidate for further examination.
While the above candidates were particularly well supported, the following results were also noteworthy:
! 116! 1) Insulin-like growth factor 1b receptor (igf1rb), crumbs family member 2a (crb2a) and zinc finger protein 503 (znf503) each contain multiple SNPs with alleles segregating between surface and Pachón cave samples, including at least one non-synonymous mutation, and show significant differential expression during at least one developmental stage. They may therefore warrant further investigation.
2) Fibroblast growth factor 8 (fgf8) is present on a scaffold anchored within the QTL interval on linkage group 25 and shows significant overexpression in cavefish, relative to surface fish at 24hpf and 3dpf. Elevated levels of this gene have been implicated in retinal defects in
Astyanax (Pottin et al. 2011).
3) Choline O-acetyltransferase a (chata), located on the same genomic scaffold as fgf8, is highly over-expressed in cavefish at 10hpf. In zebrafish, chata plays a part in the role of acetylcholine as an important neuromodulator in early development; cells in the brain and retina as well as the optic nerve and spinal motorneurons are ChAT-immunoreactive from various early stages of development (Arenzana et al. 2005). Additionally, the zebrafish bajan mutant line shows compromised motility and fatigue as a result of a point mutation in the chata gene (Wang et al. 2008). Therefore, it is possible that the early over-expression of chata seen in the Pachón population may contribute to the increased level of activity observed in cavefish.
4) Teleost multiple tissue opsin b (tmtopsb) is present in our data set but shows no evidence of sequence variation or significant expression differences. Additionally, none of the four known cavefish melanopsin (opn4) genes appear to be anchored within the QTL intervals examined here. Truncation of members of these two opsin families has been demonstrated in
Phreatichthys andruzzii and implicated in that species’ inability to entrain locomotor activity with exogenous lighting cues (Cavallari et al. 2011). Beale et al. (2013) similarly found no
! 117! evidence to implicate these genes as playing a part in the behavioral alterations observed in
Astyanax.
5) Similarly, none of the genes present on scaffolds anchored to the QTL intervals examined here are part of the core circadian clockwork or the stabilizing loop described in zebrafish (Vatine et al. 2011), nor were any of the genes annotated with GO terms indicating an obvious role in maintenance of circadian rhythms.
CONCLUSIONS
In this study, we successfully identified several novel QTL underlying patterns of locomotor activity in Astyanax mexicanus. Further, our results suggest that spatial components of locomotor activity, such as patterns in usage of the top and bottom of the trial tank, have a genetic basis that is distinct from loci underlying patterns in the overall level of activity captured by velocity measurements. Finally, examination of the relevant genetic intervals using a combination of genomic and transcriptomic data enabled us to build a list of potential candidate genes for further study in this species and highlight several genes with particularly strong support. The results presented here serve to further our understanding of the genetic underpinnings locomotor activity patterns in this species and help to lay the groundwork for additional studies that will elucidate how particular genes contribute to the development, maintenance or modulation of complex behaviors in vertebrate taxa.
ACKNOWLEDGEMENTS
The authors would like to thank Wendy Lu and Tyler Bussian for their invaluable assistance with data analysis and Kyle Dougherty for providing us with eye measurements. Julie
! 118! Carlson and Jenny Sung provided extremely helpful edits and suggestions on this work, which were sincerely appreciated. We would also like to thank Amanda Powers, Bethany Stahl, Allison
Furterer and other past and present members of the Gross Lab for their assistance, ideas and advice. Finally, we would like to thank Dr. Richard Borowsky (New York University) for generously providing us with the fish used in this study.
REFERENCES
Albertson, R. C., W. Cresko, H. W. Detrich, 3rd and J. H. Postlethwait, 2009 Evolutionary
mutant models for human disease. Trends Genet 25: 74-81.
Arenzana, F. J., D. Clemente, R. Sanchez-Gonzalez, A. Porteros, J. Aijon et al., 2005
Development of the cholinergic system in the brain and retina of the zebrafish. Brain Res
Bull 66: 421-425.
Beale, A., C. Guibal, T. K. Tamai, L. Klotz, S. Cowen et al., 2013 Circadian rhythms in Mexican
blind cavefish Astyanax mexicanus in the lab and in the field. Nat Commun 4: 2769.
Benjamini, Y., and Y. Hochberg, 1995 Controlling the false discovery rate: A practical and
powerful approach to multiple testing. J Roy Stat Soc B Met 57: 289-300.
Bilandžija, H., L. Ma, A. Parkhurst and W. R. Jeffery, 2013 A potential benefit of albinism in
Astyanax cavefish: Downregulation of the oca2 gene increases tyrosine and
catecholamine levels as an alternative to melanin synthesis. PLoS One 8: e80823.
Borowsky, R., 2008 Determining the sex of adult Astyanax mexicanus. CSH Protoc 3: 1070-
1071.
Borowsky, R., 2013 Eye regression in blind Astyanax cavefish may facilitate the evolution of an
adaptive behavior and its sensory receptors. BMC Biol 11: 81.
! 119! Broman, K. W., H. Wu, S. Sen and G. A. Churchill, 2003 R/qtl: QTL mapping in experimental
crosses. Bioinformatics 19: 889-890.
Carlson, B. M., S. W. Onusko and J. B. Gross, 2015 A high-density linkage map for Astyanax
mexicanus using genotyping-by-sequencing technology. G3 Genes Genom Genet 5: 241-
251.
Cavallari, N., E. Frigato, D. Vallone, N. Frohlich, J. F. Lopez-Olmeda et al., 2011 A blind
circadian clock in cavefish reveals that opsins mediate peripheral clock photoreception.
PLoS Biol 9: e1001142.
Cheng, Y. C., P. J. Scotting, L. S. Hsu, S. J. Lin, H. Y. Shih et al., 2013 Zebrafish rgs4 is
essential for motility and axonogenesis mediated by Akt signaling. Cell Mol Life Sci 70:
935-950.
Christensen, K. A., J. P. Brunelli, P. A. Wheeler and G. H. Thorgaard, 2014 Antipredator
behavior QTL: Differences in rainbow trout clonal lines derived from wild and hatchery
populations. Behav Genet 44: 535-546.
Colihueque, N., R. Cárdenas, L. Ramírez, F. Estay and C. Araneda, 2010 Analysis of the
association between spawning time QTL markers and the biannual spawning behavior in
rainbow trout (Oncorhynchus mykiss). Genet Mol Biol 33: 578-582.
Culver, D., 1982 Cave life. Evolution and ecology. Harvard University Press, Cambridge, MA.
Dahlbom, S. J., D. Lagman, K. Lundstedt-Enkel, L. F. Sundstrom and S. Winberg, 2011
Boldness predicts social status in zebrafish (Danio rerio). PLoS One 6: e23565.
Duboué, E. R., and R. L. Borowsky, 2012 Altered rest-activity patterns evolve via circadian
independent mechanisms in cave adapted balitorid loaches. PLoS One 7: e30868.
! 120! Duboué, E. R., A. C. Keene and R. L. Borowsky, 2011 Evolutionary convergence on sleep loss
in cavefish populations. Curr Biol 21: 671-676.
Elipot, Y., H. Hinaux, J. Callebert, J.-M. Launay, M. Blin et al., 2014 A mutation in the enzyme
monoamine oxidase explains part of the Astyanax cavefish behavioural syndrome. Nat
Commun 5: 3647.
Elipot, Y., H. Hinaux, J. Callebert and S. Retaux, 2013 Evolutionary shift from fighting to
foraging in blind cavefish through changes in the serotonin network. Curr Biol 23: 1-10.
Elshire, R. J., J. C. Glaubitz, Q. Sun, J. A. Poland, K. Kawamoto et al., 2011 A robust, simple
genotyping-by-sequencing (GBS) approach for high diversity species. PLoS One 6:
e19379.
Erckens, W., and W. Martin, 1982a Exogenous and endogenous control of swimming activity in
Astyanax mexicanus (Characidae, Pisces) by direct light response and by a circadian
oscillator I. Analyses of the time-control systems of an epigean river population. Z
Naturforsch 37c: 1253-1265.
Erckens, W., and W. Martin, 1982b Exogenous and endogenous control of swimming activity in
Astyanax mexicanus (Characidae, Pisces) by direct light response and by a circadian
oscillator II. Features of time-controlled behaviour of a cave population and their
comparison to a epigean ancestral form. Z Naturforsch 37c: 1266-1273.
Erckens, W., and F. Weber, 1976 Rudiments of an ability for time measurement in the
cavernicolous fish Anoptichthys jordani Hubbs and Innes (Pisces, Characidae).
Experientia 32: 1297-1299.
Flint, J., 2003 Analysis of quantitative trait loci that influence animal behavior. J Neurobiol 54:
46-77.
! 121! Friedrich, M., 2013 Biological clocks and visual systems in cave-adapted animals at the dawn of
speleogenomics. Integr Comp Biol 53: 50-67.
Gandhi, A. V., E. A. Mosser, G. Oikonomou and D. A. Prober, 2015 Melatonin is required for
the circadian regulation of sleep. Neuron 85: 1193-1199.
Greenwood, A. K., R. Ardekani, S. R. McCann, M. E. Dubin, A. Sullivan et al., 2015 Genetic
mapping of natural variation in schooling tendency in the threespine stickleback. G3
Genes Genom Genet 5: 761-769.
Greenwood, A. K., A. R. Wark, K. Yoshida and C. L. Peichel, 2013 Genetic and neural
modularity underlie the evolution of schooling behavior in threespine sticklebacks. Curr
Biol 23: 1884-1888.
Gross, J. B., A. Furterer, B. M. Carlson and B. A. Stahl, 2013 An integrated transcriptome-wide
analysis of cave and surface dwelling Astyanax mexicanus. PLoS One 8: e55659.
Gross, J. B., A. J. Krutzler and B. M. Carlson, 2014 Complex craniofacial changes in blind cave-
dwelling fish are mediated by genetically symmetric and asymmetric loci. Genetics 196:
1303-1319.
Gross, J. B., B. Meyer and M. Perkins, 2015 The rise of Astyanax cavefish. Dev Dyn: Early
online. DOI: 10.1002/dvdy.24253.
Gu, Z., Q. Jiang and Z. Yan, 2007 RGS4 modulates serotonin signaling in prefrontal cortex and
links to serotonin dysfunction in a rat model of schizophrenia. Mol Pharmacol 71: 1030-
1039.
Guo, S., 2004 Linking genes to brain, behavior and neurological diseases: What can we learn
from zebrafish? Genes Brain and Behav 3: 63-74.
! 122! Haltaufderhyde, K., R. N. Ozdeslik, N. L. Wicks, J. A. Najera and E. Oancea, 2015 Opsin
expression in human epidermal skin. Photochem Photobiol 91: 117-123.
Hecht, B. C., F. P. Thrower, M. C. Hale, M. R. Miller and K. M. Nichols, 2012 Genetic
architecture of migration-related traits in rainbow and steelhead trout, Oncorhynchus
mykiss. G3 Genes Genet Genom 2: 1113-1127.
Hubbs, C. L., and W. T. Innes, 1936 The first known blind fish of the family Characidae: A new
genus from Mexico. Occas Pap Mus Zool Univ Mich 342: 1-7.
Ingley, S. J., J. Rehm and J. B. Johnson, 2014 Size doesn't matter, sex does: A test for boldness
in sister species of Brachyrhaphis fishes. Ecol Evol 4: 4361-4369.
Irving, E., and C. Brown, 2013 Examining the link between personality and laterality in a feral
guppy Poecilia reticulata population. J Fish Biol 83: 311-325.
Iuvone, P. M., G. Tosini, N. Pozdeyev, R. Haque, D. C. Klein et al., 2005 Circadian clocks,
clock networks, arylalkylamine N-acetyltransferase, and melatonin in the retina. Prog
Retin Eye Res 24: 433-456.
Jeffery, W. R., 2001 Cavefish as a model system in evolutionary developmental biology. Dev
Biol 231: 1-12.
Jeffery, W. R., 2005 Adaptive evolution of eye degeneration in the Mexican blind cavefish. J
Hered 96: 185-196.
Jeffery, W. R., 2009 Regressive evolution in Astyanax cavefish. Annu Rev Genet 43: 25-47.
Ji, M., W. J. Zhao, L. D. Dong, Y. Miao, X. L. Yang et al., 2011 RGS2 and RGS4 modulate
melatonin-induced potentiation of glycine currents in rat retinal ganglion cells. Brain Res
1411: 1-8.
! 123! Kay, J. N., K. C. Finger-Baier, T. Roeser, W. Staub and H. Baier, 2001 Retinal ganglion cell
genesis requires lakritz, a zebrafish atonal homolog. Neuron 30: 725-736.
Kendler, K. S., and R. J. Greenspan, 2006 The nature of genetic influences on behavior: Lessons
from "simpler" organisms. Am J Psychiatry 163: 1683-1694.
King, A. J., I. Furtbauer, D. Mamuneas, C. James and A. Manica, 2013 Sex-differences and
temporal consistency in stickleback fish boldness. PLoS One 8: e81116.
Kowalko, J. E., N. Rohner, T. A. Linden, S. B. Rompani, W. C. Warren et al., 2013a
Convergence in feeding posture occurs through different genetic loci in independently
evolved cave populations of Astyanax mexicanus. Proc Natl Acad Sci U S A 110: 16933-
16938.
Kowalko, J. E., N. Rohner, S. B. Rompani, B. K. Peterson, T. A. Linden et al., 2013b Loss of
schooling behavior in cavefish through sight-dependent and sight-independent
mechanisms. Curr Biol 23: 1874-1883.
Laine, V. N., G. Herczeg, T. Shikano, J. Vilkki and J. Merila, 2014 QTL analysis of behavior in
nine-spined sticklebacks (Pungitius pungitius). Behav Genet 44: 77-88.
Lu, F., A. E. Lipka, J. Glaubitz, R. Elshire, J. H. Cherney et al., 2013 Switchgrass genomic
diversity, ploidy, and evolution: novel insights from a network-based SNP discovery
protocol. PLoS Genet 9: e1003215.
Manchenkov, T., M. P. Pasillas, G. G. Haddad and F. B. Imam, 2015 Novel genes critical for
hypoxic preconditioning in zebrafish are regulators of insulin and glucose metabolism.
G3 Genes Genet Genom: Early online. DOI:10.1534/g1533.1115.018010.
McGaugh, S. E., J. B. Gross, B. Aken, M. Blin, R. Borowsky et al., 2014 The cavefish genome
reveals candidate genes for eye loss. Nat Commun 5: 5307.
! 124! Menna-Barreto, L., and E. Trajano, 2015 Biological rhythmicity in subterranean animals: A
function risking extinction?, pp. 55-68 in Mechanisms of Circadian Systems in Animals
and Their Clinical Relevance, edited by R. Aguilar-Roblero, M. Díaz-Muñoz and M. L.
Fanjul-Moles. Springer International Publishing.
Mirat, O., J. R. Sternberg, K. E. Severi and C. Wyart, 2013 ZebraZoom: An automated program
for high-throughput behavioral analysis and categorization. Front Neural Circuits 7: 107.
Mortazavi, A., B. A. Williams, K. McCue, L. Schaeffer and B. Wold, 2008 Mapping and
quantifying mammalian transcriptomes by RNA-seq. Nat Methods 5: 621-628.
O'Quin, K. E., M. Yoshizawa, P. Doshi and W. R. Jeffery, 2013 Quantitative genetic analysis of
retinal degeneration in the blind cavefish Astyanax mexicanus. PLoS One 8: e57281.
Pottin, K., H. Hinaux and S. Retaux, 2011 Restoring eye size in Astyanax mexicanus blind
cavefish embryos through modulation of the Shh and Fgf8 forebrain organising centres.
Development 138: 2467-2476.
Poulson, T. L., and W. B. White, 1969 The cave environment. Science 165: 971-981.
Rexroad, C. E., R. L. Vallejo, S. Liu, Y. Palti and G. M. Weber, 2013 Quantitative trait loci
affecting response to crowding stress in an F2 generation of rainbow trout produced
through phenotypic selection. Mar Biotechnol 15: 613-627.
Thines, G., and M. Weyers, 1978 Réponses locomotrices du poisson cavernicole Astyanax
jordani (Pisces, Characidae) à des signaux périodiques et apériodiques de lumière et de
température. Int J Speleol 10: 35-55.
Thines, G., and F. Wolff-Van Ermengem, 1965 Activity patterns in an epigean characidae and its
degenerated cave-relative. Anim Behav 13: 585.
! 125! Tsuboko, S., T. Kimura, M. Shinya, Y. Suehiro, T. Okuyama et al., 2014 Genetic control of
startle behavior in medaka fish. PLoS One 9: e112527.
Vatine, G., D. Vallone, Y. Gothilf and N. S. Foulkes, 2011 It's time to swim! Zebrafish and the
circadian clock. FEBS Lett 585: 1485-1494.
Wang, M., H. Wen and P. Brehm, 2008 Function of neuromuscular synapses in the zebrafish
choline-acetyltransferase mutant bajan. J Neurophysiol 100: 1995-2004.
Wright, D., R. K. Butlin and O. Carlborg, 2006a Epistatic regulation of behavioural and
morphological traits in the zebrafish (Danio rerio). Behav Genet 36: 914-922.
Wright, D., R. Nakamichi, J. Krause and R. K. Butlin, 2006b QTL analysis of behavioral and
morphological differentiation between wild and laboratory zebrafish (Danio rerio).
Behav Genet 36: 271-284.
Wright, S., 1964 Pleiotropy in the evolution of structural reduction and of dominance. Am Nat
98: 65-69.
Yamashita, T., K. Ono, H. Ohuchi, A. Yumoto, H. Gotoh et al., 2014 Evolution of mammalian
Opn5 as a specialized UV-absorbing pigment by a single amino acid mutation. J Biol
Chem 289: 3991-4000.
Yoshizawa, M., G. Ashida and W. R. Jeffery, 2012a Parental genetic effects in a cavefish
adaptive behavior explain disparity between nuclear and mitochondrial DNA. Evolution
66: 2975-2982.
Yoshizawa, M., B. G. Robinson, E. R. Duboue, P. Masek, J. B. Jaggard et al., 2015 Distinct
genetic architecture underlies the emergence of sleep loss and prey-seeking behavior in
the Mexican cavefish. BMC Biol 13: 15.
! 126! Yoshizawa, M., Y. Yamamoto, K. E. O'Quin and W. R. Jeffery, 2012b Evolution of an adaptive
behavior and its sensory receptors promotes eye regression in blind cavefish. BMC Biol
10: 108.
FIGURE LEGENDS
Figure 4.1: F1 surface x Pachón hybrids show surface-like activity patterns with cave-like influences. Data shown are mean values for a small number of surface (n = 5), cave (n = 8) and
F1 hybrid (n = 5) individuals assayed for 24hr under 12:12hr light/dark conditions. Trial starts were staggered, so data shown covers more than 24hrs and displays greater variability at the very beginning and near the end. Data binned at 15min (900s) intervals. Green = time in top zone, red
= time in middle zone, blue = time in bottom zone. Times shown are in zeitgeber time (ZT); ZT
00:00 = lights on, ZT 12:00 = lights off. Yellow bars = subjective day, black bars = subjective night.
Figure 4.2: F2 surface x Pachón hybrids show a wide array of parental and non-parental activity patterns. Each panel includes data for one F2 individual assayed for 24hr under 12:12hr light/dark conditions. Trial starts were staggered. Times shown are in zeitgeber time (ZT); ZT
00:00 = lights on, ZT 12:00 = lights off. Yellow bars = subjective day, black bars = subjective night.
Figure 4.3: Locomotor activity QTL examined. Velocity (red), top usage (blue) and bottom usage (green) QTL data are shown with the results of multiple mapping methods included where the location of QTL peaks differed between methods (A). QTL for right/left pupil size (LG 20;
B), presence/absence of right/left eye (LG 3; C), sex (LG15; D) and presence/absence of
! 127! albinism (LG 13; E) are shown in subsequent panels, with co-localizing activity QTL included.
In all panels, arrows indicate the location of primary or secondary (a single peak at ~35 cM on
LG 3) QTL intervals examined after evaluating effect plots for each peak seen in our data.
Dotted black lines represent a genome-wide 0.05α LOD threshold of 4.65; some indicated QTL do not quite meet this threshold, but all indicated peaks have significant LOD scores based on permutation tests calculated for the particular combination of activity metric and mapping method that are represented.
Figure 4.4: Non-synonymous mutation causes amino acid substitution at highly conserved position in cavefish ATOH7. Amino acids that match the consensus are shaded in black. The potential effect of this mutation has not been evaluated.
! 128! TABLES
Table 4.1: Results of QTL analysis for metrics of locomotor activity. Results based on 24hr assay data for Astyanax surface x Pachón F2 hybrids (n = 127) under 12:12hr light/dark conditions. For all listed peaks, p ≤ 0.05 based on 1000 permutations.
Activity Peak Linkage Position LOD Closest Variance Data Set Methoda Metricb Marker Groupc (cM) Score Marker Explained Velocity MR Day Mean TP73890 2 69.50 5.67 - 18.6% Velocity EM Day Mean TP73890 2 69.50 4.76 - 15.9% Velocity HK Day Mean TP73890 2 69.50 4.63 - 15.5% Velocity MR Trial Mean TP74449 3 -4.05 5.06 - 16.8% Velocity HK Trial Mean TP74449 3 -4.05 4.82 - 16.0% Velocity EM Trial Mean TP60697 3 8.60 4.92 - 16.3% Velocity EM Day Mean TP86516 3 17.90 5.25 - 17.3% Velocity HK Day Mean TP86516 3 17.90 5.27 - 17.4% Velocity MR Day Mean TP11322 3 79.80 5.52 - 18.1% Velocity EM Day Mean TP44190 13 60.00 5.17 - 17.1% Velocity HK Day Mean TP44190 13 60.00 5.14 - 17.0% Top Usage NP Day Mean TP65410 14 2.60 5.39 - 17.8% Bottom Usage NP Day Mean TP65410 14 2.60 5.70 - 18.7% Bottom Usage EM Trial Mean TP53209 14 10.60 6.09 - 19.8% Bottom Usage HK Trial Mean TP53209 14 10.60 5.95 - 19.4% Bottom Usage MR Trial Mean TP19556 14 10.90 6.93 - 22.2% Top Usage NP Trial Mean TP36991 14 19.40 4.55 - 15.2% Velocity MR Night Mean TP73771 15 78.40 6.30 - 20.4% Velocity EM Night Mean TP73771 15 78.40 5.19 - 17.2% Velocity HK Night Mean TP73771 15 78.40 5.42 - 17.8% Velocity MR Trial Mean TP73771 15 78.42 5.76 - 18.8% Velocity HK Trial Mean TP73771 15 78.42 4.64 - 15.5% Velocity EM Trial Mean c15.loc79 15 79.00 4.51 TP73771 15.1% Bottom Usage HK Day Minus Night c25.loc40 25 40.00 5.01 TP57986 16.6% Bottom Usage EM Day Minus Night c25.loc43 25 43.00 4.99 TP9488 16.6% Bottom Usage MR Day Minus Night TP9488 25 43.10 4.96 - 16.5% a MR = Marker Regression; EM = Expectation Maximization; HK = Haley-Knott; NP = Non-Parametric b “Day” refers to the period when lights were on (ZT 00:000-12:00); “night” refers to the period when lights were off (ZT 12:00-00:00). c Linkage groups listed are from the map presented in Carlson et al. (2015).
! 129! Table 4.2: Summary of candidate gene assessment. For specific details, refer to Tables 4.S4-6.
Differential Expression Sequence Variation Astyanax Relevant GO 5' or 3' Non- Ensembl ID Gene Scaffold Terms? Development Juvenile UTR Synonymous Synonymous ENSAMXG00000000759 col15a1b KB871754.1 Yes Yes Yes No Yes No ENSAMXG00000000885 xcr1a.1 (1 of 2) KB871754.1 Yes No No No No Yes ENSAMXG00000000886 xcr1a.1 (2 of 2) KB871754.1 Yes No No No No No ENSAMXG00000000888 cc2d2a KB871754.1 Yes Yes No No Yes No ENSAMXG00000005193 cxcl32b.1 KB871766.1 Yes Yes No No No No ENSAMXG00000001564 znf503 (1 of 2) KB872296.1 Yes Yes No Yes No Yes ENSAMXG00000002007 pitx3 KB872296.1 Yes Yes No No Yes No ENSAMXG00000002296 sfrp5 KB872296.1 Yes Yes No Yes No No ENSAMXG00000012711 dmbx1b KB882082.1 Yes Yes No No Yes No ENSAMXG00000013049 vapb KB882082.1 Yes Yes No No No No ENSAMXG00000013417 novel gene KB882082.1 Yes No No No No No ENSAMXG00000014705 prickle1a KB882082.1 Yes No No No No No ENSAMXG00000014945 cerkl KB882082.1 Yes No No No No No ENSAMXG00000009797 mcm3 KB882090.1 Yes Yes No Yes No No ENSAMXG00000010179 opn5 KB882090.1 Yes Yes No No No No ENSAMXG00000021415 igf1rb KB882097.1 Yes Yes No No Yes Yes ENSAMXG00000021574 lim2.2 KB882097.1 Yes No Yes No No No ENSAMXG00000021593 uhrf1 KB882097.1 Yes Yes No No No No ENSAMXG00000019838 crb2a KB882104.1 Yes Yes Yes Yes Yes Yes ENSAMXG00000020264 htr2cl2 KB882104.1 Yes Yes No No No No ENSAMXG00000005992 novel gene KB882105.1 Yes No No No No No ENSAMXG00000010609 lepr KB882115.1 Yes Yes No Yes Yes No ENSAMXG00000012018 fgf8 KB882125.1 Yes Yes No No No No ENSAMXG00000012130 chata KB882125.1 Yes Yes No No No No ENSAMXG00000012172 rgra KB882125.1 Yes Yes No Yes No No ENSAMXG00000012545 novel gene KB882125.1 Yes Yes No No No No
! 130! ENSAMXG00000002537 b9d2 KB882129.1 Yes Yes No No Yes No ENSAMXG00000005363 rb1 KB882129.1 Yes No No No Yes No ENSAMXG00000005527 dlat KB882129.1 Yes Yes No No Yes No ENSAMXG00000025967 atoh7 KB882154.1 Yes Yes Yes No No Yes ENSAMXG00000008135 tmtopsb KB882155.1 Yes No No No No No ENSAMXG00000016819 scinla KB882172.1 Yes Yes No No No No ENSAMXG00000002437 novel gene KB882228.1 Yes No No No No No ENSAMXG00000011539 rgs4 KB882253.1 Yes Yes No No No No ENSAMXG00000014314 marcksa KB882283.1 Yes Yes No No No No ENSAMXG00000005947 tfap2a KB882287.1 Yes Yes No No No No
! 131! Table 4.S1: Activity data for Asty66 F2 surface x Pachón hybrid pedigree assayed for 24hr under 12:12hr light/dark conditions. Here n = 127; two specimens were deceased before assays could be conducted.
Velocity Data (cm/s) Top Zone Usage Data (s/900s) Bottom Zone Usage Data (s/900s)
Specimen Trial Mean Day Mean Night Mean Day - Night Trial Mean Day Mean Night Mean Day - Night Trial Mean Day Mean Night Mean Day - Night Asty66-001 7.295445491 7.939838938 6.651052044 1.288786893 233.9377566 243.1215929 224.7539203 18.36767263 289.8589554 289.4943545 290.2235563 -0.729201833
Asty66-002 3.701063739 0.603390786 6.798736693 -6.195345907 253.0172531 2.152150813 503.8823554 -501.7302046 495.4739467 882.5659004 108.3819931 774.1839073
Asty66-003 7.479609775 4.963537137 9.995682413 -5.032145276 129.1823078 17.61761785 240.7469977 -223.1293799 573.7817669 750.9711097 396.5924242 354.3786854
Asty66-004 7.987692302 6.079070386 9.896314219 -3.817243833 193.3099765 25.64161356 360.9783395 -335.3367259 452.6213725 757.8349183 147.4078266 610.4270917
Asty66-005 5.648295587 1.819395915 9.477195259 -7.657799344 67.50083419 0.119564 134.8821044 -134.7625404 656.5951375 893.1952785 419.9949965 473.200282
Asty66-006 2.291887594 2.541555256 2.042219933 0.499335322 255.5972645 56.70323233 454.4912966 -397.7880643 461.09547 713.7588957 208.4320444 505.3268513
Asty66-007 4.91287673 6.089907894 3.735845567 2.354062327 202.5685402 154.9931871 250.1438933 -95.15070619 540.6740076 574.1964189 507.1515964 67.0448225
Asty66-008 6.441225124 3.468676118 9.413774129 -5.945098011 76.16018845 16.17867881 136.1416981 -119.9630193 686.6449792 827.0965425 546.1934159 280.9031266
Asty66-009 4.069436972 3.69973399 4.439139955 -0.739405965 142.8160791 133.9130791 151.7190791 -17.80600006 405.641058 466.2544485 345.0276676 121.226781
Asty66-010 8.225417118 7.237583625 9.213250612 -1.975666988 108.314912 57.87801652 158.7518075 -100.873791 542.2352904 625.3329726 459.1376082 166.1953644
Asty66-011 ------
Asty66-012 3.062399852 1.879301217 4.245498487 -2.36619727 156.5211043 16.89328229 296.1489263 -279.255644 551.4504086 681.7866478 421.1141695 260.6724784
Asty66-013 2.214045068 2.671073912 1.757016224 0.914057689 158.4160547 127.4740014 189.358108 -61.88410663 499.4508402 616.3726239 382.5290565 233.8435673
Asty66-014 6.314358674 5.686666858 6.94205049 -1.255383632 186.4993478 125.009734 247.9889615 -122.9792275 382.8714137 449.3966182 316.3462093 133.0504089
Asty66-015 4.913681553 5.091289525 4.736073581 0.355215944 223.4832053 192.038567 254.9278436 -62.8892766 492.7528233 487.1364438 498.3692029 -11.23275917
Asty66-016 6.544809552 6.742451276 6.347167829 0.395283447 128.9098122 187.4047654 70.41485908 116.9899063 549.9155405 496.591034 603.240047 -106.649013
Asty66-017 4.836573959 0.937097322 8.736050596 -7.798953274 23.09809829 1.399315958 44.79688063 -43.39756467 743.9317794 851.9874043 635.8761545 216.1112498
Asty66-018 7.618853397 8.759593276 6.478113518 2.281479758 144.3752781 110.5786336 178.1719225 -67.59328885 537.0669282 564.2017025 509.9321539 54.26954858
Asty66-019 4.120807061 1.793147898 6.448466223 -4.655318325 272.9396052 4.945917875 540.9332925 -535.9873746 476.471611 842.2255592 110.7176629 731.5078963
Asty66-020 6.037611369 4.577961233 7.497261505 -2.919300271 207.9555251 169.8636135 246.0474368 -76.18382333 354.7043565 365.7317023 343.6770107 22.05469165
Asty66-021 4.544219412 1.595263448 7.493175376 -5.897911928 65.0143198 0.619369375 129.4092702 -128.7899009 753.6241102 894.5445444 612.7036761 281.8408683
Asty66-022 2.5452212 1.251795578 3.838646821 -2.586851242 31.94826785 6.526665417 57.36987029 -50.84320488 812.5066038 866.497748 758.5154597 107.9822882
Asty66-023 3.216902417 2.775351303 3.658453531 -0.883102228 361.3599706 144.6522913 578.0676499 -433.4153585 259.9884609 414.6396402 105.3372817 309.3023585
Asty66-024 4.883081335 1.903605219 7.862557452 -5.958952233 123.1321596 11.8264095 234.4379097 -222.6115002 634.973168 839.4672461 430.4790899 408.9881562
Asty66-025 4.913911574 1.373173706 8.454649441 -7.081475735 170.1204673 9.754893542 330.486041 -320.7311475 549.7400178 861.8569971 237.6230386 624.2339585
! 132! Asty66-026 11.88369531 10.37734355 13.39004707 -3.012703522 165.6837388 42.79209608 288.5753816 -245.7832855 525.1397231 680.9629086 369.3165375 311.6463711
Asty66-027 6.295022977 7.00919289 5.580853064 1.428339826 147.8419389 90.84223092 204.8416468 -113.9994159 460.569945 547.6469521 373.492938 174.1540141
Asty66-028 7.646265897 5.235136659 10.05739514 -4.822258477 333.5304059 123.6896625 543.3711493 -419.6814868 309.762192 564.3344724 55.18991167 509.1445607
Asty66-029 6.857569222 3.531821363 10.18331708 -6.651495717 144.1535285 11.99185348 276.3152035 -264.32335 527.1869099 789.9489786 264.4248411 525.5241375
Asty66-030 10.40256402 10.47177166 10.33335637 0.138415293 198.6781934 79.95078523 317.4056016 -237.4548164 414.4217143 488.090869 340.7525596 147.3383093
Asty66-031 3.018212066 4.440555983 1.595868148 2.844687834 111.0144867 70.90632319 151.1226503 -80.21632706 656.5840139 644.5243837 668.6436441 -24.11926042
Asty66-032 8.705738036 9.503788935 7.907687138 1.596101796 138.7185795 26.43963454 250.9975245 -224.55789 567.8164283 719.3408711 416.2919854 303.0488857
Asty66-033 4.044857943 1.941280144 6.148435742 -4.207155598 150.987793 27.76665558 274.2089305 -246.4422749 579.4739185 800.1953336 358.7525035 441.4428301
Asty66-034 5.567201634 2.850138808 8.28426446 -5.434125652 263.8388386 61.42462088 466.2530563 -404.8284354 454.5493398 720.6963878 188.4022918 532.2940961
Asty66-035 6.658118128 3.501640926 9.81459533 -6.312954404 125.9895305 33.21029419 218.7687668 -185.5584726 483.5582115 693.9321273 273.1842956 420.7478317
Asty66-036 8.982398944 6.412600613 11.55219728 -5.139596664 56.76266519 9.166805771 104.3585246 -95.19171883 698.9298329 817.0879225 580.7717432 236.3161793
Asty66-037 7.249903749 2.759729589 11.74007791 -8.980348319 209.8383102 25.58391765 394.0927028 -368.5087851 524.0302798 794.7106809 253.3498786 541.3608023
Asty66-038 6.170313922 3.98214266 8.358485184 -4.376342523 32.52036728 12.75303088 52.28770369 -39.53467281 662.1447832 762.7738845 561.5156819 201.2582026
Asty66-039 6.09161426 6.542529558 5.640698963 0.901830595 152.5904372 117.4431373 187.737737 -70.29459975 513.8920168 564.553442 463.2305916 101.3228504
Asty66-040 2.213233206 1.486573819 2.939892593 -1.453318774 90.8745552 8.522411292 173.2266991 -164.7042878 704.0071323 859.6714771 548.3427875 311.3286896
Asty66-041 5.645599239 7.675361097 3.615837381 4.059523715 85.71453373 52.15493244 119.274135 -67.11920258 625.4650486 642.2401575 608.6899398 33.55021773
Asty66-042 11.79353291 12.2303039 11.35676191 0.873541991 244.6481227 171.748835 317.5474104 -145.7985754 406.2381828 468.5087876 343.9675781 124.5412095
Asty66-043 5.244420786 7.780061062 2.708780509 5.071280553 193.0604223 241.9183076 144.2025371 97.71577052 511.7089312 390.4286234 632.989239 -242.5606156
Asty66-044 8.961871138 3.497173778 14.4265685 -10.92939472 95.84271755 11.57755008 180.107885 -168.5303349 655.6518334 830.3803805 480.9232863 349.4570942
Asty66-045 7.828399164 7.039001749 8.617796579 -1.578794831 137.6793462 67.77263421 207.5860581 -139.8134239 567.8897653 672.3994824 463.3800483 209.0194341
Asty66-046 3.948355068 3.386010223 4.510699913 -1.12468969 144.2995073 81.41822319 207.1807913 -125.7625681 567.6687105 685.2588708 450.0785502 235.1803206
Asty66-047 6.247853245 2.25528437 10.24042212 -7.98513775 77.22201383 1.361083292 153.0829444 -151.7218611 627.1924707 861.1437823 393.2411591 467.9026231
Asty66-048 4.730619665 2.42217815 7.03906118 -4.616883029 94.50526916 32.716745 156.2937933 -123.5770483 560.9081299 700.9808411 420.8354187 280.1454224
Asty66-049 6.816768935 4.749290666 8.884247205 -4.134956539 108.8227116 44.20114531 173.4442779 -129.2431326 561.8959238 665.8241579 457.9676896 207.8564683
Asty66-050 4.528673682 5.209467968 3.847879396 1.361588571 115.8241574 44.2463299 187.4019849 -143.155655 577.7621366 606.480785 549.0434883 57.43729671
Asty66-051 5.840381534 1.914535539 9.766227528 -7.851691988 55.66399753 5.742548188 105.5854469 -99.84289869 725.1345096 864.5457954 585.7232237 278.8225718
Asty66-052 4.475809657 2.541338035 6.410281279 -3.868943244 19.96475633 15.45712371 24.47238896 -9.01526525 704.7064422 699.7984081 709.6144762 -9.816068104
Asty66-053 8.499944853 3.789887932 13.21000177 -9.420113843 99.66077109 3.876793208 195.444749 -191.5679558 639.2607886 842.6468135 435.8747636 406.7720499
Asty66-054 7.889708853 5.443397326 10.33602038 -4.892623053 88.65288924 50.46018271 126.8455958 -76.38541306 614.6153102 683.5919263 545.6386941 137.9532322
Asty66-055 5.769715933 1.948002439 9.591429428 -7.643426988 159.5790236 12.60844217 306.549605 -293.9411629 566.8557446 838.6435049 295.0679842 543.5755207
Asty66-056 2.495158766 2.708319963 2.281997569 0.426322393 219.5591424 130.7029249 308.4153598 -177.7124349 527.2988268 654.984151 399.6135026 255.3706484
! 133! Asty66-057 5.524989501 8.049740465 3.000238538 5.049501926 421.1065225 250.8856076 591.3274374 -340.4418298 256.8266194 363.3028886 150.3503501 212.9525385
Asty66-058 2.850121156 2.490195144 3.210047167 -0.719852024 101.5352157 27.17995688 175.8904746 -148.7105177 649.8877351 765.949283 533.8261872 232.1230958
Asty66-059 12.07740035 10.76059837 13.39420233 -2.633603958 90.65940933 58.88249356 122.4363251 -63.55383154 634.7635841 681.2048161 588.3223521 92.88246406
Asty66-060 10.17142658 10.63470062 9.708152543 0.926548077 172.274358 222.8137873 121.7349287 101.0788586 512.9859033 436.5977099 589.3740966 -152.7763867
Asty66-061 7.81833284 8.490731642 7.145934037 1.344797605 250.9676341 66.29615671 435.6391115 -369.3429548 328.9821066 423.5527187 234.4114944 189.1412242
Asty66-062 4.800140727 3.986602753 5.613678702 -1.62707595 178.6380129 38.15065048 319.1253753 -280.9747248 430.4710961 684.0674019 176.8747904 507.1926114
Asty66-063 7.518992497 8.608513847 6.429471147 2.179042699 201.8782663 176.7788613 226.9776713 -50.19881 436.5139453 412.9601835 460.0677072 -47.10752371
Asty66-064 4.474171612 2.333163944 6.61517928 -4.282015336 163.3310388 31.33758735 295.3244902 -263.9869028 524.7202067 790.033089 259.4073244 530.6257646
Asty66-065 1.86154203 1.976081492 1.747002567 0.229078925 154.560463 23.29412719 285.8267989 -262.5326717 501.9046828 796.0467412 207.7626244 588.2841168
Asty66-066 4.078976497 1.560858653 6.597094341 -5.036235689 121.3046374 3.043321042 239.5659538 -236.5226327 589.2750389 838.8993162 339.6507617 499.2485544
Asty66-067 12.16868903 13.18226441 11.15511365 2.027150759 101.6064673 64.64798029 138.5649544 -73.9169741 544.7986183 624.8588884 464.7383482 160.1205402
Asty66-068 9.334704725 8.71862084 9.950788611 -1.232167771 177.4795632 126.5077583 228.4513682 -101.9436099 499.7925009 554.8110594 444.7739424 110.037117
Asty66-069 4.475373632 5.367543956 3.583203308 1.784340647 295.439884 154.2764991 436.6032689 -282.3267698 317.3784899 481.3028302 153.4541496 327.8486806
Asty66-070 5.467497322 6.784474082 4.150520563 2.633953519 69.44444417 13.30080044 125.5880879 -112.2872875 623.6632462 727.666555 519.6599373 208.0066177
Asty66-071 7.558624322 7.038037858 8.079210787 -1.041172929 272.670935 103.8802697 441.4616003 -337.5813306 361.3783921 537.0912576 185.6655267 351.4257309
Asty66-072 6.725936536 8.741431195 4.710441878 4.030989316 326.6537364 272.6545979 380.6528749 -107.998277 208.2513063 263.6115269 152.8910857 110.7204413
Asty66-073 9.069422208 7.046806413 11.092038 -4.045231591 165.5085641 88.17845677 242.8386714 -154.6602146 466.7007296 616.1661662 317.2352931 298.9308731
Asty66-074 5.147146325 4.4450244 5.849268249 -1.404243849 120.8364611 51.46396379 190.2089584 -138.7449946 530.6368881 671.6431731 389.630603 282.0125701
Asty66-075 3.409864607 3.598135889 3.221593325 0.376542564 129.2865781 34.62142685 223.9517293 -189.3303024 609.0493276 730.1704485 487.9282066 242.2422419
Asty66-076 4.446073395 1.763854977 7.128291813 -5.364436836 93.09309313 21.8968971 164.2892891 -142.392392 641.3747087 818.9696643 463.7797531 355.1899112
Asty66-077 6.653421803 4.016974313 9.289869294 -5.272894982 122.617061 25.07090371 220.1632183 -195.0923145 593.5459773 770.2049269 416.8870277 353.3178992
Asty66-078 6.317334396 8.564791153 4.069877638 4.494913515 126.3583022 112.8858019 139.8308025 -26.94500054 546.7109467 459.4525085 633.9693849 -174.5168765
Asty66-079 3.131309939 3.076996314 3.185623564 -0.108627251 65.8582196 47.3765431 84.3398961 -36.963353 637.4530777 638.2146017 636.6915537 1.523048
Asty66-080 3.751664078 3.624237678 3.879090479 -0.254852801 214.7814484 220.6421697 208.9207271 11.72144258 498.9413015 460.4312644 537.4513385 -77.0200741
Asty66-081 3.661289118 0.717559046 6.60501919 -5.887460145 131.796727 3.447197104 260.1462568 -256.6990597 622.2413385 844.202536 400.280141 443.922395
Asty66-082 7.381144567 3.471571374 11.29071776 -7.819146386 113.4822317 8.775442021 218.1890214 -209.4135794 634.4448613 824.7546153 444.1351073 380.6195081
Asty66-083 5.942618871 3.204806814 8.680430929 -5.475624115 56.60382634 8.249221521 104.9584312 -96.70920965 683.3187352 778.3700361 588.2674343 190.1026018
Asty66-084 5.456691342 5.184566426 5.728816259 -0.544249832 62.01896358 7.891918292 116.1460089 -108.2540906 685.8042077 791.3357814 580.2726339 211.0631475
Asty66-085 3.19750832 3.452868367 2.942148274 0.510720093 76.1740909 63.58024735 88.76793444 -25.18768708 663.5006532 688.9500606 638.0512458 50.89881485
Asty66-086 9.140756872 8.542329937 9.739183807 -1.19685387 300.075423 285.5814163 314.5694298 -28.98801342 349.4640464 361.8542138 337.0738791 24.78033473
Asty66-087 7.387903215 4.246506967 10.52929946 -6.282792497 172.9934795 8.634328438 337.3526305 -328.718302 534.9884608 836.7179678 233.2589538 603.459014
! 134! Asty66-088 5.206222871 1.656918516 8.755527227 -7.098608711 113.2312867 2.079162458 224.383411 -222.3042485 638.2917643 875.0006952 401.5828334 473.4178618
Asty66-089 6.542887529 4.746721127 8.33905393 -3.592332803 136.7051075 49.53773125 223.8724837 -174.3347525 569.1507482 735.7211391 402.5803574 333.1407817
Asty66-090 7.792867561 6.748340982 8.83739414 -2.089053157 315.8759453 133.0017515 498.7501392 -365.7483878 211.6494968 296.2413798 127.0576137 169.1837661
Asty66-091 11.37598852 11.68214056 11.06983649 0.612304075 416.939509 490.6170064 343.2620116 147.3549948 247.6716294 159.6575739 335.6856849 -176.0281109
Asty66-092 3.65642903 5.364394041 1.948464019 3.415930021 294.9477264 390.6281304 199.2673225 191.3608079 376.0639101 260.1803185 491.9475018 -231.7671833
Asty66-093 10.29802425 10.2026244 10.3934241 -0.190799707 190.4108969 140.8477927 239.974001 -99.12620829 413.2465804 454.1444233 372.3487374 81.79568588
Asty66-094 4.445790329 2.142420485 6.749160173 -4.606739688 125.2613726 3.872622625 246.6501226 -242.7774999 670.9125798 866.4824545 475.342705 391.1397495
Asty66-095 6.816636959 7.557751131 6.075522787 1.482228345 287.9956344 257.123096 318.8681728 -61.74507679 332.6958897 336.7610644 328.6307149 8.130349521
Asty66-096 4.991130687 4.788561108 5.193700266 -0.405139158 126.589089 16.60410423 236.5740738 -219.9699695 420.4527449 706.8450396 134.0604503 572.7845893
Asty66-097 5.017193395 3.439584439 6.594802351 -3.155217912 55.3994963 18.7096819 92.08931071 -73.37962881 643.5751728 764.7494723 522.4008734 242.348599
Asty66-098 6.572411972 4.481956781 8.662867162 -4.180910381 40.90757414 60.3839946 21.43115367 38.95284094 699.2760129 664.4102438 734.1417821 -69.73153835
Asty66-099 4.218787278 1.494916656 6.942657901 -5.447741245 88.75611725 6.311172563 171.2010619 -164.8898894 626.6241939 839.3101431 413.9382448 425.3718983
Asty66-100 8.372544264 3.871936296 12.87315223 -9.001215937 92.2060947 6.687937667 177.7242517 -171.0363141 638.7456895 843.5421544 433.9492246 409.5929298
Asty66-101 3.612785748 3.684155802 3.541415694 0.142740109 174.4862914 34.770187 314.2023959 -279.4322089 472.463436 653.247692 291.67918 361.568512
Asty66-102 8.587950168 10.07445959 7.101440747 2.973018843 181.6948889 139.4623792 223.9273987 -84.4650195 476.1118756 457.6277651 494.595986 -36.96822094
Asty66-103 6.624373687 5.807835241 7.440912133 -1.633076892 145.7262818 80.22188898 211.2306746 -131.0087856 461.1656804 517.7594255 404.5719353 113.1874902
Asty66-104 9.59450675 9.844057198 9.344956302 0.499100896 176.8024973 9.293320875 344.3116738 -335.0183529 508.9805081 764.8328868 253.1281294 511.7047575
Asty66-105 7.614379805 7.946057234 7.282702376 0.663354858 86.53792672 6.374429708 166.7014237 -160.326994 632.3851638 759.5720724 505.1982552 254.3738173
Asty66-106 8.729527803 9.353758062 8.105297544 1.248460518 75.3861495 48.08975625 102.6825428 -54.5927865 683.2186356 681.6691686 684.7681026 -3.098933979
Asty66-107 11.37380651 10.97070034 11.77691269 -0.806212346 175.8449415 104.1207871 247.5690959 -143.4483089 507.7998143 597.8937294 417.7058992 180.1878302
Asty66-108 6.694417561 6.256856531 7.13197859 -0.875122059 423.0998351 384.1383039 462.0613664 -77.92306256 217.7038154 288.5990158 146.808615 141.7904008
Asty66-109 5.953768458 6.857362161 5.050174755 1.807187406 167.5446284 77.35721817 257.7320386 -180.3748204 408.6430184 459.3127848 357.9732519 101.3395329
Asty66-110 10.87757356 10.78128284 10.97386427 -0.192581424 475.4799934 297.5725725 653.3874143 -355.8148418 118.1914556 185.3353358 51.04757546 134.2877603
Asty66-111 7.721540654 6.313595364 9.129485945 -2.815890581 73.2204425 21.8767375 124.5641475 -102.68741 714.1273219 800.6833224 627.5713213 173.1120011
Asty66-112 4.313833624 4.580217572 4.047449676 0.532767896 260.9908535 166.395565 355.586142 -189.190577 487.962615 549.342397 426.582833 122.759564
Asty66-113 6.24515071 6.994305534 5.495995886 1.498309648 195.4840256 137.2789459 253.6891054 -116.4101595 479.864935 516.121678 443.608192 72.51348602
Asty66-114 10.28933087 10.50204603 10.07661571 0.425430325 177.268936 113.6761756 240.8616965 -127.1855209 530.3584838 581.6086925 479.1082751 102.5004174
Asty66-115 5.08240728 5.466502105 4.698312455 0.76818965 665.6399457 667.4876278 663.7922636 3.695364188 96.14579849 48.49015667 143.8014403 -95.31128365
Asty66-116 7.110073086 8.349323072 5.8708231 2.478499972 347.9316121 146.3345295 549.5286947 -403.1941652 249.7462744 378.1232629 121.3692859 256.753977
Asty66-117 6.363956605 5.801961437 6.925951772 -1.123990335 192.7163263 89.3080575 296.1245951 -206.8165376 462.1347042 553.2734124 370.9959959 182.2774165
Asty66-118 7.795952796 8.57705109 7.014854501 1.562196589 285.663789 294.1045209 277.223057 16.88146388 394.4628661 352.7312059 436.1945262 -83.46332027
! 135! Asty66-119 9.568166706 9.755264208 9.381069203 0.374195004 94.19280426 87.40337554 100.982233 -13.57885744 458.9141232 463.849266 453.9789804 9.870285583
Asty66-120 6.075302642 6.591021759 5.559583525 1.031438234 128.2848827 110.6975048 145.8722606 -35.17475585 672.7199416 691.1606036 654.2792795 36.88132412
Asty66-121 9.769593578 10.67586178 8.863325375 1.812536405 436.8858438 329.9257595 543.8459281 -213.9201686 164.6945558 188.4551203 140.9339914 47.52112896
Asty66-122 7.403321499 7.712744676 7.093898322 0.618846354 159.9988879 107.9106893 212.0870866 -104.1763973 437.3738323 527.9849302 346.7627343 181.2221959
Asty66-123 9.199735571 13.09406379 5.305407351 7.788656441 100.2505974 91.42892794 109.0722668 -17.64333883 574.6625088 508.0948979 641.2301197 -133.1352218
Asty66-124 5.09478274 2.683023729 7.506541752 -4.823518023 141.6409468 2.569235917 280.7126576 -278.1434217 578.397147 872.0157662 284.7785279 587.2372383
Asty66-125 5.677126357 7.754904531 3.599348183 4.155556348 176.4511033 146.2080111 206.6941955 -60.48618442 385.7482486 409.7409904 361.7555068 47.98548358
Asty66-126 6.602361589 7.847573119 5.35715006 2.490423059 361.875416 135.2449659 588.5058661 -453.2609002 282.6381933 415.8026078 149.4737789 266.3288289
Asty66-127 10.14122223 10.56117127 9.721273195 0.839898077 329.0283342 291.7083748 366.3482935 -74.63991873 319.6734921 330.401234 308.9457502 21.45548373
Asty66-128 6.315798386 7.485634491 5.14596228 2.339672211 297.1631359 255.4443348 338.8819371 -83.43760227 381.6865475 357.5047269 405.8683681 -48.36364127
Asty66-129 ------
! 136! Table 4.S2: Sex, albinism and eye data for Asty66 F2 surface x Pachón hybrid pedigree. All members of the pedigree (n = 129) were scored for sex (1 = female, 0 = male), presence of albinism and presence of right and left eye (1 = present, 0 = absent) in a binary fashion. Eye and pupil size measurements (in pixels) were made from images where possible; in some cases an eye was technically present, but was too disorganized to measure the eye and/or pupil accurately.
Right Eye Left Eye Specimen Sex Albinism Right Eye Left Eye Right Pupil Left Pupil Area Area Asty66-001 1 0 1 1 10530 8583 61397 57146 Asty66-002 1 0 1 1 10931 9646 52644 50421 Asty66-003 1 0 1 1 12040 12085 53709 52138 Asty66-004 1 0 1 1 16556 15157 80908 75006 Asty66-005 0 0 1 1 12171 11042 45482 45360 Asty66-006 0 0 1 1 10293 11055 39487 40970 Asty66-007 1 0 1 1 14338 13416 59899 53822 Asty66-008 0 0 1 1 12754 12571 59177 59308 Asty66-009 1 0 1 1 - - - - Asty66-010 1 0 1 1 13173 12059 64882 66087 Asty66-011 1 0 1 1 - - 42371 39638 Asty66-012 0 0 1 1 6562 5582 33057 34579 Asty66-013 1 0 1 1 14866 13815 73217 75362 Asty66-014 0 0 1 1 - - - - Asty66-015 0 0 1 1 13588 11268 60026 53979 Asty66-016 1 0 1 1 - - - - Asty66-017 0 0 1 1 11180 10743 55990 53825 Asty66-018 1 0 1 1 7109 8050 30597 37801 Asty66-019 1 0 1 1 7760 7313 44710 42482 Asty66-020 1 0 1 1 16387 14830 81795 79853 Asty66-021 0 0 1 1 13376 12565 52803 48975
! 137! Asty66-022 1 0 1 1 15700 15529 77039 73945 Asty66-023 1 0 1 1 9483 10644 53506 45213 Asty66-024 1 0 1 1 10126 10148 58929 59620 Asty66-025 1 0 1 1 12222 13558 60791 55919 Asty66-026 0 0 1 1 5744 5962 34748 35940 Asty66-027 1 0 1 1 16706 15507 60775 53335 Asty66-028 0 0 1 1 8457 9604 40225 42390 Asty66-029 1 0 1 1 4752 7275 37674 32054 Asty66-030 0 0 1 1 8195 8190 33992 29952 Asty66-031 1 0 1 1 - - - - Asty66-032 1 0 1 1 8450 7117 50859 40261 Asty66-033 1 0 1 1 17357 16646 72396 71512 Asty66-034 0 0 1 1 11795 11903 57848 56702 Asty66-035 0 0 1 1 4216 3508 35743 33154 Asty66-036 0 0 1 1 11766 12884 60872 62075 Asty66-037 1 0 1 1 5307 7217 28321 45083 Asty66-038 0 0 1 1 - - - - Asty66-039 0 0 1 1 15640 14691 64646 60485 Asty66-040 1 0 1 1 - - 41411 38768 Asty66-041 0 0 1 1 6274 4291 29203 37840 Asty66-042 0 0 1 1 13550 14317 67901 65818 Asty66-043 0 0 1 1 7227 5708 42250 41810 Asty66-044 0 0 1 1 8425 8899 44775 45584 Asty66-045 0 0 1 1 12167 11694 55132 57682 Asty66-046 0 0 1 1 20860 20243 86604 83838 Asty66-047 0 0 1 1 11990 10948 51592 49364 Asty66-048 1 0 1 1 11028 9966 46183 43592 Asty66-049 1 0 1 1 11373 11456 61584 57901 Asty66-050 0 0 1 1 4370 4366 31706 34125 Asty66-051 0 0 1 1 - - - - Asty66-052 0 0 1 1 14739 15012 56401 56512
! 138! Asty66-053 0 0 1 1 13880 13580 50820 50032 Asty66-054 0 0 1 1 5716 5566 35115 34977 Asty66-055 1 0 1 1 9215 9756 56606 57175 Asty66-056 0 0 1 1 - - - - Asty66-057 1 0 0 0 - - - - Asty66-058 1 0 0 0 - - - - Asty66-059 1 0 0 0 - - - - Asty66-060 1 0 1 1 - - - - Asty66-061 0 0 0 0 - - - - Asty66-062 0 0 1 1 - - - - Asty66-063 1 0 1 1 - - - - Asty66-064 0 0 1 1 - - - - Asty66-065 1 0 0 1 - - - - Asty66-066 0 0 1 1 - - - - Asty66-067 0 0 1 1 - - - - Asty66-068 0 0 0 0 - - - - Asty66-069 1 0 0 1 - - - - Asty66-070 1 0 1 1 - - - - Asty66-071 1 0 1 1 - - - - Asty66-072 0 0 0 0 - - - - Asty66-073 0 0 1 0 - - - - Asty66-074 1 0 1 1 - - - - Asty66-075 1 0 0 0 - - - - Asty66-076 1 0 1 1 - - - - Asty66-077 1 0 1 1 - - - - Asty66-078 0 0 1 1 - - - - Asty66-079 1 0 1 1 - - - - Asty66-080 0 0 1 1 - - - - Asty66-081 1 0 1 1 - - - - Asty66-082 0 0 1 1 - - - - Asty66-083 0 0 1 1 - - - -
! 139! Asty66-084 0 0 1 1 - - - - Asty66-085 1 0 1 1 - - - - Asty66-086 0 0 1 0 - - - - Asty66-087 0 1 1 1 - - - - Asty66-088 0 0 1 1 - - - - Asty66-089 0 0 1 1 - - - - Asty66-090 0 0 1 1 - - - - Asty66-091 0 0 1 1 - - - - Asty66-092 0 0 1 1 - - - - Asty66-093 0 1 0 0 - - - - Asty66-094 0 0 1 1 - - - - Asty66-095 1 0 1 1 - - - - Asty66-096 1 1 1 1 - - - - Asty66-097 0 0 1 1 - - - - Asty66-098 0 0 1 1 16507 16132 74233 69511 Asty66-099 0 0 1 1 8342 8001 37599 35817 Asty66-100 0 0 1 1 - - - - Asty66-101 1 1 0 0 - - - - Asty66-102 1 1 1 1 5586 4566 31782 33017 Asty66-103 1 1 1 1 8089 8132 40118 41105 Asty66-104 0 1 1 1 9144 10499 54721 59577 Asty66-105 0 1 1 1 5991 7017 36767 38054 Asty66-106 0 1 0 0 - - - - Asty66-107 0 1 1 1 5721 5202 45584 42682 Asty66-108 1 1 1 1 - - - - Asty66-109 1 1 1 1 - - - - Asty66-110 0 1 1 1 - - - - Asty66-111 0 1 1 1 - - - - Asty66-112 0 1 1 1 - - - - Asty66-113 0 1 1 1 9883 9811 47507 50993 Asty66-114 0 1 1 1 - 730 18890 31111
! 140! Asty66-115 1 1 1 1 13118 14501 54078 56009 Asty66-116 1 1 0 1 - 2764 - 25166 Asty66-117 1 0 0 0 - - - - Asty66-118 1 0 0 0 - - - - Asty66-119 0 0 0 1 - 1884 - 26899 Asty66-120 0 0 0 0 - - - - Asty66-121 1 0 1 1 5194 5485 41036 40745 Asty66-122 0 0 0 0 - - - - Asty66-123 0 0 0 0 - - - - Asty66-124 1 0 0 0 - - - - Asty66-125 1 1 1 1 - - - - Asty66-126 0 1 1 1 5636 5655 43843 39129 Asty66-127 0 0 1 0 7076 - 34090 - Asty66-128 1 0 0 0 - - - - Asty66-129 1 0 1 1 5827 7613 50518 56949
! 141! Table 4.S3: Distribution of commonly-anchored genomic scaffolds between linkage maps published by Carlson et al. (2015) and O’Quin et al. (2013). Numbers indicate the number of genomic scaffolds anchored to both indicated linkage groups.
Carlson et al. (2015) Linkage Groups
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
1 ------1 1 - - - - 18 - - 1 - 1 1 ------2 - 1 16 ------1 ------1 - - - - 3 - 1 - - - - 10 5 - 1 - - 2 - - 1 1 - 1 ------4 1 1 ------1 ------16 5 - - 1 1 1 ------5 - - - - - 1 - 6 - - - - 1 1 ------10 -
7 - - 1 ------11 1 ------1 - 8 ------16 8 ------1 - - - - - 9 - - - - - 1 - 1 - 1 ------3 1 - - - 10 - - - - - 1 ------9 - - - - - 11 ------1 1 - - - - 1 - - 1 - - 12 - - - 1 - -
Linkage Groups Linkage 12 6 5 ------1 ------
) 13 ------1 1 - 11 ------
2013 14 - - - - - 1 1 ------1 - - - 1 3 - - - ( 15 - 1 ------1 ------al.
et 16 - - 1 ------1 - 5 ------17 - 1 - - 1 14 ------1 - - 1 - - 1 1 - - 1 - 18 - - 1 - - - 1 - - 1 1 - - - 9 ------
O'Quin O'Quin 19 - - - - 1 ------1 ------1 - 20 ------5 ------21 ------6 2 1 - - - - 1 ------22 - 1 - - 10 1 ------1 - - - 1 ------23 ------1 ------1 24 - 1 1 5 ------1 - 25 ------1 ------5 - -
! 142! Table 4.S4: Results of gene ontology-based candidate gene screen. Listed GO terms based on annotations of Astyanax genes and/or zebrafish homologs.
Astyanax Ensembl Gene ID Gene Symbol Gene Name GO Term Scaffold ENSAMXG00000000759 col15a1b collagen, type XV, alpha 1b KB871754.1 eye development
ENSAMXG00000000885 xcr1a.1 (1 of 2) chemokine (C motif) receptor 1a, duplicate 1 KB871754.1 locomotion
ENSAMXG00000000886 xcr1a.1 (2 of 2) chemokine (C motif) receptor 1a, duplicate 1 KB871754.1 locomotion photoreceptor cell outer segment ENSAMXG00000000888 cc2d2a coiled-coil and C2 domain containing 2A KB871754.1 organization ENSAMXG00000005193 cxcl32b.1 chemokine (C-X-C motif) ligand 32b, duplicate 1 KB871766.1 locomotion embryonic camera-type eye ENSAMXG00000001564 znf503 (1 of 2) zinc finger protein 503 KB872296.1 morphogenesis ENSAMXG00000002007 pitx3 paired-like homeodomain 3 KB872296.1 camera-type eye development lens development in camera-type ENSAMXG00000002007 pitx3 paired-like homeodomain 3 KB872296.1 eye ENSAMXG00000002296 sfrp5 secreted frizzled-related protein 5 KB872296.1 eye development retina development in camera- ENSAMXG00000012711 dmbx1b diencephalon/mesencephalon homeobox 1b KB882082.1 type eye VAMP (vesicle-associated membrane protein)- ENSAMXG00000013049 vapb KB882082.1 swimming behavior associated protein B and C ENSAMXG00000013417 novel gene Uncharacterized protein KB882082.1 locomotion
ENSAMXG00000014705 prickle1a prickle homolog 1a (Drosophila) KB882082.1 neural retina development
ENSAMXG00000014945 cerkl ceramide kinase-like KB882082.1 photoreceptor cell development
ENSAMXG00000014945 cerkl ceramide kinase-like KB882082.1 retina layer formation retina development in camera- ENSAMXG00000009797 mcm3 minichromosome maintenance complex component 3 KB882090.1 type eye
! 143! ENSAMXG00000010179 opn5 opsin 5 KB882090.1 photoreceptor activity
ENSAMXG00000021415 igf1rb insulin-like growth factor 1b receptor KB882097.1 camera-type eye development embryonic camera-type eye ENSAMXG00000021415 igf1rb insulin-like growth factor 1b receptor KB882097.1 development ENSAMXG00000021574 lim2.2 lens intrinsic membrane protein 2.2 KB882097.1 structural constituent of eye lens lens development in camera-type ENSAMXG00000021593 uhrf1 ubiquitin-like with PHD and ring finger domains 1 KB882097.1 eye ENSAMXG00000019838 crb2a crumbs family member 2a KB882104.1 camera-type eye development embryonic retina morphogenesis ENSAMXG00000019838 crb2a crumbs family member 2a KB882104.1 in camera-type eye 5-hydroxytryptamine (serotonin) receptor 2C, G ENSAMXG00000020264 htr2cl2 KB882104.1 feeding behavior protein-coupled-like 2 5-hydroxytryptamine (serotonin) receptor 2C, G ENSAMXG00000020264 htr2cl2 KB882104.1 locomotory behavior protein-coupled-like 2 ENSAMXG00000005992 novel gene uncharacterized protein KB882105.1 regulation of behavior
ENSAMXG00000010609 lepr leptin receptor KB882115.1 camera-type eye development embryonic camera-type eye ENSAMXG00000012018 fgf8 fibroblast growth factor 8 KB882125.1 development embryonic retina morphogenesis ENSAMXG00000012018 fgf8 fibroblast growth factor 8 KB882125.1 in camera-type eye retina development in camera- ENSAMXG00000012018 fgf8 fibroblast growth factor 8 KB882125.1 type eye ENSAMXG00000012130 chata choline O-acetyltransferase a KB882125.1 locomotory behavior
ENSAMXG00000012172 rgra retinal G protein coupled receptor a KB882125.1 phototransduction
ENSAMXG00000012172 rgra retinal G protein coupled receptor a KB882125.1 locomotion
ENSAMXG00000012545 novel gene uncharacterized protein KB882125.1 locomotion
ENSAMXG00000002537 b9d2 B9 protein domain 2 KB882129.1 opsin transport
ENSAMXG00000005363 rb1 retinoblastoma 1 KB882129.1 optic nerve development
! 144! retinal ganglion cell axon ENSAMXG00000005363 rb1 retinoblastoma 1 KB882129.1 guidance ENSAMXG00000005527 dlat dihydrolipoamide S-acetyltransferase KB882129.1 detection of light stimulus detection of light stimulus ENSAMXG00000005527 dlat dihydrolipoamide S-acetyltransferase KB882129.1 involved in visual perception ENSAMXG00000025967 atoh7 atonal homolog 7 KB882154.1 camera-type eye development
ENSAMXG00000025967 atoh7 atonal homolog 7 KB882154.1 eye development
ENSAMXG00000025967 atoh7 atonal homolog 7 KB882154.1 retina layer formation
ENSAMXG00000025967 atoh7 atonal homolog 7 KB882154.1 swimming behavior
ENSAMXG00000008135 tmtopsb teleost multiple tissue opsin b KB882155.1 photoreceptor activity
ENSAMXG00000008135 tmtopsb teleost multiple tissue opsin b KB882155.1 phototransduction
ENSAMXG00000016819 scinla scinderin like a KB882172.1 eye development
ENSAMXG00000002437 novel gene uncharacterized protein KB882228.1 photoreceptor activity
ENSAMXG00000011539 rgs4 regulator of G-protein signaling 4 KB882253.1 locomotory behavior
ENSAMXG00000014314 marcksa myristoylated alanine-rich protein kinase C substrate a KB882283.1 retina layer formation retina development in camera- ENSAMXG00000005947 tfap2a transcription factor AP-2 alpha KB882287.1 type eye
! 145! Table 4.S5: Results of gene expression analysis. Listed fold changes are for Pachón cavefish, relative to surface fish. Only genes with significant expression differences for at least one time point are shown. Hpf = hours post fertilization; dpf = days post fertilization; *p ≤ 0.05.
Astyanax Fold Change Fold Change Fold Change Fold Change Fold Change in Fold Change in Ensembl Transcript ID Gene Scaffold at 10hpf at 24hpf at 1.5dpf at 3dpf L/D Juveniles D/D Juveniles ENSAMXT00000000806 col15a1b KB871754.1 1.632 up 2.403 up* 1.244 up 1.631 down* 1.584 down* 1.023 up ENSAMXT00000000907 cc2d2a KB871754.1 1.269 down 1.419 down* 1.599 down 1.217 down* 1.206 down 1.195 down ENSAMXT00000005317 cxcl32b.1 KB871766.1 none 15.495 down 37.715 up* 44.778 up* 1.353 down 1.422 up ENSAMXT00000001586 znf503 (1 of 2) KB872296.1 1.375 up* 1.200 up 1.123 down 1.618 down* 1.524 up 1.085 down ENSAMXT00000002039 pitx3 KB872296.1 1.236 up 1.434 down* 1.419 down 2.394 down* 1.290 up 1.395 down ENSAMXT00000002347 sfrp5 KB872296.1 1.569 up* 1.235 up 1.233 up* 1.088 up 1.005 down 1.018 up ENSAMXT00000013064 dmbx1b KB882082.1 1.384 down* 1.367 down* 1.101 down 2.169 down* 1.036 down 3.593 down ENSAMXT00000013423 vapb KB882082.1 1.306 up* 1.010 down 1.113 down 1.223 up* 1.114 up 1.007 down ENSAMXT00000010067 mcm3 KB882090.1 1.561 down* 1.168 down* 1.252 down 1.170 down 1.024 up 1.836 up ENSAMXT00000010451 opn5 KB882090.1 3.679 down 6.341 down* 1.933 down 3.681 down* 1.170 up 2.169 down ENSAMXT00000022053 igf1rb KB882097.1 1.537 up 1.278 up 1.299 up 1.528 up* 1.101 down 1.242 down ENSAMXT00000022220 lim2.2 KB882097.1 1.228 down 12.428 down 18.190 up 4.197 up 78.344 down 181.240 down* ENSAMXT00000022239 uhrf1 KB882097.1 1.029 up 1.230 down* 1.202 down 1.079 up 1.582 down 1.475 up ENSAMXT00000020429 crb2a KB882104.1 1.344 up 1.695 down* 1.367 down 1.710 down* 4.724 down 7.837 down* ENSAMXT00000020869 htr2cl2 KB882104.1 39.977 up 1.096 up 2.188 up 7.862 up* 7.454 up 3.285 down ENSAMXT00000010901 lepr KB882115.1 1.493 up 1.463 down 1.060 up 1.209 down* 1.008 down 1.157 down ENSAMXT00000012355 fgf8 KB882125.1 1.311 down 1.635 up* 1.360 up 1.419 up* 1.339 up 1.018 down ENSAMXT00000012506 chata KB882125.1 35.993 up* 1.350 up 1.204 up 1.298 down 1.268 down 2.173 down ENSAMXT00000012519 rgra KB882125.1 2.550 up* 2.003 up* 1.077 up 2.655 down* 1.552 up 1.043 down ENSAMXT00000012896 novel gene KB882125.1 3.288 up* 6.177 up* 5.560 up 4.456 up* 1.339 up 2.353 up ENSAMXT00000002596 b9d2 KB882129.1 1.265 down* 1.140 down 1.129 down 1.439 down 1.755 down 1.908 down ENSAMXT00000005728 dlat KB882129.1 1.696 up 1.061 down 1.197 down 1.209 up* 1.141 up 1.283 up ENSAMXT00000026693 atoh7 KB882154.1 none 4.119 up 38.702 down 11.083 down* 74.041 down* 1.044 down ENSAMXT00000017315 scinla KB882172.1 1.387 down* 2.026 up* 1.167 up 1.123 down 1.178 down 1.280 down ENSAMXT00000011876 rgs4 KB882253.1 2.272 up* 1.204 up* 1.919 up* 2.004 up* 1.385 down 1.806 down ENSAMXT00000014726 marcksa KB882283.1 1.427 down* 1.096 down 1.208 down 1.248 down* 1.019 up 1.006 up ENSAMXT00000006092 tfap2a KB882287.1 1.692 down* 1.259 down* 1.169 down 1.253 down* 1.722 up 1.431 down
! 146! Table 4.S6: Results of sequence variation analysis. CDS = Coding sequence; UTR =
Untranslated region.
Astyanax Portion Genomic Protein Ensembl ID Gene Scaffold of Gene Change Change ENSAMXG00000000759 col15a1b KB871754.1 CDS g.69818A>G - ENSAMXG00000000759 col15a1b KB871754.1 Intron g.81811G>C - ENSAMXG00000000885 xcr1a.1 (1 of 2) KB871754.1 CDS g.1085A>C p.His277Pro ENSAMXG00000000888 cc2d2a KB871754.1 CDS g.1407T>C - ENSAMXG00000000888 cc2d2a KB871754.1 CDS g.1883A>G - ENSAMXG00000000888 cc2d2a KB871754.1 CDS g.3139G>A - ENSAMXG00000000888 cc2d2a KB871754.1 CDS g.13847G>A - ENSAMXG00000000888 cc2d2a KB871754.1 CDS g.20242G>A - ENSAMXG00000000888 cc2d2a KB871754.1 CDS g.20272G>A - ENSAMXG00000001564 znf503 (1 of 2) KB872296.1 3' UTR g.89A>G - ENSAMXG00000001564 znf503 (1 of 2) KB872296.1 3' UTR g.399A>G - ENSAMXG00000001564 znf503 (1 of 2) KB872296.1 3' UTR g.487T>A - ENSAMXG00000001564 znf503 (1 of 2) KB872296.1 3' UTR g.492C>T - ENSAMXG00000001564 znf503 (1 of 2) KB872296.1 CDS g.2184G>A p.Ala262Val ENSAMXG00000001564 znf503 (1 of 2) KB872296.1 5' UTR g.3498T>C - ENSAMXG00000001564 znf503 (1 of 2) KB872296.1 5' UTR g.3570A>G - ENSAMXG00000002007 pitx3 KB872296.1 CDS g.1056G>A - ENSAMXG00000002296 sfrp5 KB872296.1 3' UTR g.25305A>T - ENSAMXG00000012711 dmbx1b KB882082.1 CDS g.4566C>T - ENSAMXG00000009797 mcm3 KB882090.1 3' UTR g.8193T>G - ENSAMXG00000021415 igf1rb KB882097.1 CDS g.88C>T - ENSAMXG00000021415 igf1rb KB882097.1 CDS g.16248C>T p.Arg793Gln ENSAMXG00000021415 igf1rb KB882097.1 CDS g.21599G>A - ENSAMXG00000021415 igf1rb KB882097.1 CDS g.24682A>C - ENSAMXG00000019838 crb2a KB882104.1 5' UTR g.2318G>T - ENSAMXG00000019838 crb2a KB882104.1 CDS g.5815C>T - ENSAMXG00000019838 crb2a KB882104.1 CDS g.5894A>G p.Val1014Ala ENSAMXG00000019838 crb2a KB882104.1 CDS g.9101A>G - ENSAMXG00000019838 crb2a KB882104.1 CDS g.10605C>T p.Glu557Lys ENSAMXG00000019838 crb2a KB882104.1 CDS g.12316G>A - ENSAMXG00000019838 crb2a KB882104.1 CDS g.12484G>T - ENSAMXG00000019838 crb2a KB882104.1 CDS g.17234C>T - ENSAMXG00000019838 crb2a KB882104.1 CDS g.18107A>C p.Ser58Ala ENSAMXG00000010609 lepr KB882115.1 5' UTR g.7A>G - ENSAMXG00000010609 lepr KB882115.1 CDS g.10195G>T - ENSAMXG00000010609 lepr KB882115.1 3' UTR g.24538T>G - ENSAMXG00000012172 rgra KB882125.1 3' UTR g.2534G>A - ENSAMXG00000002537 b9d2 KB882129.1 CDS g.3893G>A - ENSAMXG00000005363 rb1 KB882129.1 CDS g.2798C>T - ENSAMXG00000005363 rb1 KB882129.1 Intron g.15718C>A - ENSAMXG00000005527 dlat KB882129.1 CDS g.11697C>A - ENSAMXG00000025967 atoh7 KB882154.1 CDS g.95T>A p.Tyr110Phe ENSAMXG00000011539 rgs4 KB882253.1 Intron g.2193G>A - ENSAMXG00000014314 marcksa KB882283.1 Intron g.1615C>G - ENSAMXG00000005947 tfap2a KB882287.1 Intron g.5998G>C -
! 147! FIGURES
Figure 4.1
Velocity"by"Morphotype" Tank"Usage"by"Morphotype"
16" 900" 14" 800" 12" 700" 10" 600" 500" 8" 400" Surface" 6" Surface" 300" 4" Time"in"Zone"(s)" 200"
Mean"Velocity"(cm/s)" 2" 100" 0" 0" 6:00" 8:00" 0:00" 2:00" 4:00" 6:00" 8:00" 6:00" 8:00" 0:00" 2:00" 4:00" 6:00" 8:00" 10:00" 12:00" 14:00" 16:00" 18:00" 20:00" 22:00" 10:00" 10:00" 12:00" 14:00" 16:00" 18:00" 20:00" 22:00" 10:00"
16" 900" 14" 800" 12" 700" 10" 600" 500" 8" Hybrid"
"Hybrid" 400" 1" 6" 1 F F 300" 4"
Time"in"Zone"(s)" 200" 2"
Mean"Velocity"(cm/s)" 100" 0" 0" 6:00" 8:00" 0:00" 2:00" 4:00" 6:00" 8:00" 6:00" 8:00" 0:00" 2:00" 4:00" 6:00" 8:00" 10:00" 12:00" 14:00" 16:00" 18:00" 20:00" 22:00" 10:00" 10:00" 12:00" 14:00" 16:00" 18:00" 20:00" 22:00" 10:00"
16" 900" 14" 800" 12" 700" 10" 600" 500" 8"
Cave" Cave" 400" 6" 300" 4"
Time"in"Zone"(s)" 200" 2"
Mean"Velocity"(cm/s)" 100" 0" 0" 6:00" 8:00" 0:00" 2:00" 4:00" 6:00" 8:00" 6:00" 8:00" 0:00" 2:00" 4:00" 6:00" 8:00" 10:00" 12:00" 14:00" 16:00" 18:00" 20:00" 22:00" 10:00" 10:00" 12:00" 14:00" 16:00" 18:00" 20:00" 22:00" 10:00" Zeitgeber"Time" Zeitgeber"Time"
! 148! Figure 4.2
Selected"Individual"F2"Cave"x"Surface"Hybrid"Velocity"Over"24hr"" 16" 16" 14" 14" 12" 12" 10" 10" 8" 8" "Hybrid" "Hybrid" 2 2
F 6" 6" F 4" 4"
Mean"Velocity"(cm/s)" 2" 2" Mean"Velocity"(cm/s)" 0" 0" 6:00" 8:00" 10:0 12:0 14:0 16:0 18:0 20:0 22:0 0:00" 2:00" 4:00" 6:00" 8:00" 10:0 6:00" 8:00" 10:0 12:0 14:0 16:0 18:0 20:0 22:0 0:00" 2:00" 4:00" 6:00" 8:00" 10:0
16" 16" 14" 14" 12" 12" 10" 10" 8" 8" "Hybrid" "Hybrid" 2 2
F 6" 6" F 4" 4"
Mean"Velocity"(cm/s)" 2" 2" Mean"Velocity"(cm/s)" 0" 0" 6:00" 8:00" 0:00" 2:00" 4:00" 6:00" 8:00" 6:00" 8:00" 10:0 12:0 14:0 16:0 18:0 20:0 22:0 0:00" 2:00" 4:00" 6:00" 8:00" 10:0 10:00" 12:00" 14:00" 16:00" 18:00" 20:00" 22:00" 10:00"
16" 16" 14" 14" 12" 12" 10" 10" 8" 8" Hybrid" Hybrid" 2" 2"
F 6" 6" F 4" 4"
Mean"Velocity"(cm/s)" 2" 2" Mean"Velocity"(cm/s)" 0" 0" 6:00" 8:00" 0:00" 2:00" 4:00" 6:00" 8:00" 6:00" 8:00" 0:00" 2:00" 4:00" 6:00" 8:00" 10:00" 12:00" 14:00" 16:00" 18:00" 20:00" 22:00" 10:00" 10:00" 12:00" 14:00" 16:00" 18:00" 20:00" 22:00" 10:00" Zeitgeber"Time" Zeitgeber"Time"
! 149! Figure 4.3 Locomotor"AcTvity"QTL"in"Astyanax(mexicanus" A" Velocity"" Top"Zone"Usage" Bo7om"Zone"Usage"
Astyanax(Linkage"Group"20"" Astyanax(Linkage"Group"3"" Right/LeF"Pupil"Size"" Velocity"" B" C" Presence"of"Right/LeF"Eye" LOD" LOD" LOD" 4.92" 4.48" 5.52"
Astyanax(Linkage"Group"15"" Astyanax(Linkage"Group"13"" Velocity"" LOD" Velocity"" LOD" D" Sex" 6.30" E" Albinism" 5.17"
! 150! Figure 4.4 Alignment Report of Untitled ClustalV (PAM250) Page 1 Friday, May 29, 2015 8:38 AM Majority MKX------RRPS-CADSGSXXDS-RDXEKFESAARRRMAANARERKRMQGLNTA 10 20 30 40 50 60 Human atoh7.pro MKS-CKPSGPPAGARVAPPCAGGTECAGTCAGAGRLESAARRRLAANARERRRMQGLNTA 59 Mouse atoh7.pro MKSACKPHGPPAGARGAPPCAGAAERAVSCAGPGRLESAARRRLAANARERRRMQGLNTA 60 Spotted gar atoh7.pro MKS------CRPSSCADSGSDSDS-KCSEKQENAARRRMAANARERKRMQGLNTA 48 Zebrafish atoh7.pro MKP------RRPS-CADSGSDSDS-RDPEKFESAMRRRMAANARERKRMQGLNTA 47 Surface fish atoh7.pro MKP------RRPS-CADSGSECDP-RDLEKFESAARRRMAANARERKRMQGLNTA 47 Cavefish atoh7.pro MKP------RRPS-CADSGSECDP-RDLEKFESAARRRMAANARERKRMQGLNTA 47
Majority FDRLRKVVPQWGQDKKLSKYETLQMALSYIMALNRILTDASRHSTPXRXWLDLQFXXL-Q 70 80 90 100 110 120 Human atoh7.pro FDRLRRVVPQWGQDKKLSKYETLQMALSYIMALTRILAEAERFG-SERDWVGLHCEHFGR 118 Mouse atoh7.pro FDRLRRVVPQWGQDKKLSKYETLQMALSYIIALTRILAEA------ERDWVGLRCEQRGR 114 Spotted gar atoh7.pro FDRLRKVVPQWGQDKKLSKYETLQMALSYIMALNRILTDAQRHSTSQRQWLDLHIEHF-Q 107 Zebrafish atoh7.pro FDRLRKVVPQWGQDKKLSKYETLQMALSYIMALNRILSDAGRHVDPQKDWLNLQFDGL-Q 106 Surface fish atoh7.pro FDRLRKVVPQWGQDKKLSKYETLQMALSYIMALNRILTDSSRHSTPHRQWLDLQFDSL-Q 106 Cavefish atoh7.pro FDRLRKVVPQWGQDKKLSKYETLQMALSYIMALNRILTDSSRHSTPHRQWLDLQFDSL-Q 106
Majority TESYXXFXXYSSPXEXE-YMHXSFSYXYEXLQVXT
130 140 150 Human atoh7.pro DH-YLPFPGAKLPGESELYSQRLFGFQPEPFQMAT 152 Mouse atoh7.pro DHPYLPFPGARLQVDPEPYGQRLFGFQPEPFPMAS 149 Spotted gar atoh7.pro AESYSTYLEQSSAGNEE-YVHSAFPYQCDTYQVTN 141 Zebrafish atoh7.pro TEGY--FMHYDSPVESD-CMLSSFSYHYESL 134 Surface fish atoh7.pro TESYPCIVRYSSPMEHE-HMHPSFSYHYEGLQVPT 140 Cavefish atoh7.pro TESFPCIVRYSSPMEHE-HMHPSFSYHYEGLQVPT 140
Decoration 'Decorationp.Tyr110Phe++ #1': Shade (with solid black) residues that match the Consensus exactly. !
!
! 151!
CHAPTER FIVE
General Conclusions
Brian M. Carlson
Department of Biological Sciences University of Cincinnati Cincinnati, OH 45221-0006
! 152! The principle aims of my doctoral research were 1) to characterize the differences in locomotor activity between surface and cave populations of the blind Mexican tetra, Astyanax mexicanus, 2) to develop improved linkage mapping resources for use in this species, 3) to identify QTL associated with locomotor activity by assaying members of an F2 surface x cave hybrid mapping pedigree, and 4) to identify candidate genes potentially mediating the behavioral phenotypes observed. The work presented in this dissertation represents the successful completion of these aims and promises to greatly facilitate both ongoing and future research into the genetic basis of locomotor activity, circadian rhythmicity and associated behavioral phenotypes in this study system.
Efforts to assay, characterize, evaluate and compare activity patterns between the surface and cave morphotypes of Astyanax mexicanus proved fruitful and provide further support for the notion that Pachón cavefish retain a weakly-entrainable oscillator underlying locomotor activity and that this mechanism has a limited ability to self-sustain rhythms once rhythmic environmental cues have been discontinued. However, this work also highlights the importance of careful consideration in experimental design, with particular attention needing to be paid to 1) the genetic background of the fish being used, 2) how these specimens have been maintained prior to being assayed, 3) the range of trial conditions under which assays are conducted, 4) which activity metrics are employed, and 5) how the observations and measurements are made.
If these assays had been conducted or analyzed differently, it is entirely possible that the results obtained would have suggested an entirely different story than that which has been described here. With this in mind, it is much easier to understand the level of disagreement that exists within the body of literature on circadian rhythmicity and associated behaviors in this species.
Further, the apparent effects of rearing fish in total darkness suggest that one must be particularly
! 153! cautious about attempting to generalize results seen in lab-based studies to the cave environment.
Despite these complications, however, Astyanax remains an excellent model for studying the genetic architecture of behavior, such as patterns of locomotor activity. In particular, assaying additional cave populations that converge on the phenotypes observed here may facilitate identification of potential “weak points” in the complex systems underlying behavior and locomotor activity. This, in turn, may help explain aberrant behavior and departure from normal biorhythms in this and other species.
The development of high-density linkage maps like the one constructed and utilized in these studies promises to assist in the identification of critical genetic intervals containing genes or regulatory elements underlying complex phenotypes including, but certainly not limited to, patterns in locomotor activity. On one hand, lowering inter-marker distances can help to reduce the occurrence of “ghost QTL” that can appear when interval mapping methods (e.g., expectation maximization and Haley-Knott) are employed. This can also serve to narrow the regions of the genome that must be interrogated to identify potential candidate genes for a phenotype of interest by allowing the definition of critical QTL intervals on a much finer scale. This allows for more effective allocation of time and resources during the identification and potential validation of genes of interest. On the other hand, building a high-density map like the one employed here increased the chance of detecting loci of small effect, like those predicted to underlie behavioral phenotypes in Astyanax. In this case, having a large number of markers also facilitated anchoring scaffolds that represented approximately 80% of the total length of the Astyanax genome to our map. This, in turn, was critical in both candidate gene discovery and drawing comparisons between our map and previously published maps. The linkage map presented here has already proven invaluable in identifying regions of the genome mediating differential locomotor activity.
! 154! However, it promises to have an even greater impact on Astyanax research as it is applied to a wider array of phenotypes with evolutionary, developmental and clinical significance. Further, the success of this approach serves as proof-of-concept that a linkage map of this quality and density can be constructed using data from a single genotyping-by-sequencing (GBS) run. This has the potential to spur advancement in the development of linkage mapping resources in non- model systems and may serve as an impetus to explore new lines of inquiry in a broader range of study organisms.
The ultimate goal of this line of research was to investigate the genetic architecture putatively underlying altered activity patterns in Astyanax mexicanus. Building upon the successes and insights gleaned from the 72hr activity assays and construction of our high-density
GBS-based linkage map, QTL analysis facilitated the successful identification of multiple regions of the genome associated with elements of locomotor activity patterns. Not only were these efforts successful, they also demonstrated 1) that the phenotypes observed have a complex genetic basis, 2) while some activity QTL co-localize with regions implicated in eye loss and albinism, other activity QTL were independent of associations with QTL for cave-associated phenotypes, 3) that levels of activity (as measured by mean velocity) and patterns in tank usage are likely both under genetic control, but 4) tank usage and overall activity are mediated by different regions of the genome, and 5) the locomotor activity QTL identified here are distinct from those previously identified by other studies. Finally, by leveraging available transcriptomic and genomic data, we successfully developed a list of 36 candidate genes, including several that are prime targets for further research. Importantly, the results suggest that the differences in circadian rhythmicity and locomotor activity between the surface and Pachón cave populations
! 155! of Astyanax mexicanus are unlikely to be caused by some catastrophic defect in a member of the core molecular clockwork or a well-characterized photoreceptor.
The work presented here was not able to go so far as to test or confirm a role for any particular gene, regulatory element or molecular pathway in mediating the behavioral phenotypes observed. However, these studies represent a significant contribution to our understanding of circadian rhythmicity, locomotor activity and related behaviors in Astyanax mexicanus. It is hoped that the experimental methods, genetic tools and set of candidate genes developed here will provide opportunities for future discoveries by providing an excellent starting point for new lines of research in this or other cave-adapted species.
!
!
!
! 156!