UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Biochemical and molecular traits underlying hypoxia tolerance in sculpins Mandic, Milica 2015

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2015_may_mandic_milica.pdf [ 3.61MB ]
Metadata
JSON: 24-1.0166251.json
JSON-LD: 24-1.0166251-ld.json
RDF/XML (Pretty): 24-1.0166251-rdf.xml
RDF/JSON: 24-1.0166251-rdf.json
Turtle: 24-1.0166251-turtle.txt
N-Triples: 24-1.0166251-rdf-ntriples.txt
Original Record: 24-1.0166251-source.json
Full Text
24-1.0166251-fulltext.txt
Citation
24-1.0166251.ris

Full Text

BIOCHEMICAL AND MOLECULAR TRAITS UNDERLYING HYPOXIA TOLERANCE IN SCULPINS  by  MILICA MANDIC   B.Sc.H. The University of British Columbia, 2005 M.Sc. The University of British Columbia, 2008   A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY  in  THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES  (Zoology)  THE UNIVERSITY OF BRITISH COLUMBIA  (Vancouver)  April, 2015  © Milica Mandic, 2015  ii Abstract   Variation in environmental conditions can sometimes impose severe limitations on organismal survival and reproductive success and organisms that live in variable environments have evolved complex traits to cope with or buffer against the environmental conditions. In my dissertation I examined how adaptive evolution and phenotypic plasticity play a role in the ability of nearshore species of sculpin to tolerate low levels of O2 or hypoxia. I showed that constitutively expressed biochemical traits in the brain correlated with hypoxia tolerance among the different species independent of phylogenetic relationships, suggesting that these traits may have evolved via natural selection in response to hypoxia. A similar correlation was not seen in either the liver or the white muscle.   Next, I showed that transcriptionally mediated phenotypic plasticity is likely associated with the difference in hypoxia tolerance between two species of sculpin. The hypoxia-tolerant tidepool sculpin (Oligocottus maculosus) did not alter gene transcription during the ecologically relevant time-frames of hypoxia exposure (up to 8 hours), in contrast to the hypoxia-intolerant silverspotted sculpin (Blepsias cirrhosus). This suggests that the tolerant species may not rely on phenotypic plasticity during a typical environmental hypoxia exposure and instead may rely on constitutively expressed or fixed traits for survival. Only if hypoxia persists do the hypoxia tolerant tidepool sculpins alter gene transcription, for which a large set of genes showed transcriptional patterns that were divergent to the hypoxia intolerant silverspotted sculpin.  Lastly, I examined if similar transcriptional responses occur among three species of sculpin all with the same measured hypoxia tolerance. While a high proportion (65%) of clones showed similar transcription patterns among the species, a majority of genes associated with metabolism and protein production showed differences in both short and long exposures to hypoxia. As metabolism and  iii protein production both play a major role in hypoxic survival, transcriptional differences in genes belonging to these biological processes suggests that the species likely use different mechanism to achieve similar overall hypoxia tolerance phenotype. Combined, this work demonstrates how phenotypic plasticity and adaptive evolution play a role in the variation of hypoxia tolerance among species of sculpin living in the nearshore environment.     iv Preface   A	  version	  of	  Chapter	  2	  has	  been	  published	  as	  “Mandic	  M., B. Speers-Roesch and J.G. Richards. 2013. Hypoxia tolerance is associated with high anaerobic enzyme activity in brain but not in liver or muscle. Physiological and Biochemical Zoology 86: 92-105”.	  I	  designed	  and	  conducted	  all	  experiments,	  carried	  out	  all analyses and wrote the manuscript under the supervision of Dr. J.G. Richards. Dr. B. Speers-Roesch assisted in design and helped carry out the loss of equilibrium experiment. All authors provided editorial input to the manuscript.   A version of Chapter 3	  has	  been	  published	  as	  “Mandic	  M.,	  M.L.	  Ramon,	  A.Y.	  Gracey and J.G. Richards. 2014. Divergent transcriptional patterns are related to differences in hypoxia tolerance between the intertidal and the subtidal sculpins. Molecular Ecology 23: 6091-6103”.	  I	  designed all the experiments, carried out the analyses and wrote the manuscript under the supervision of Dr. J.G. Richards. Experiments were primarily conducted by me with assistance from Dr. M.L. Ramon in the preparation and hybridization of samples to microarrays. Dr. A.Y. Gracey provided technical expertise on microarrays. All authors provided editorial input to the manuscript.  Chapter 4 work was a collaboration with Drs. M.L. Ramon, A.C. Gerstein, A.Y. Gracey and J.G. Richards. I designed the experiments, primarily conducted all experiments with assistance from Dr. M.L. Ramon in preparation and hybridization of samples to microarrays and wrote the manuscript under the supervision of Dr. J.G. Richards. Dr. A.C. Gerstein provided expertise on the maximum-likelihood model analysis and Dr. A.Y. Gracey provided advice on microarray experiments. All experiments in this dissertation were approved by the UBC animal care committee (A09-0611).     v Table	  of	  Contents  Abstract ........................................................................................................................................ ii Preface ......................................................................................................................................... iv Table of Contents ...................................................................................................................... v List of Tables ............................................................................................................................. ix List of Figures ............................................................................................................................. x Acknowlodgements............................................................................................................... xii Dedication ................................................................................................................................xiv Chapter 1 General Introduction ....................................................................................... 1 1.1 Environmental Stress - Hypoxia ........................................................................................ 3 1.2 Time-Scale of the Hypoxia Response ............................................................................... 4 1.3 Levels of Biological Organization and the Hypoxic Response ............................... 7 1.4 Model Species: Sculpins (Superfamily Cottoidea) ...................................................... 8 1.5 Research Goals ...................................................................................................................... 10 Chapter 2 Hypoxia Tolerance in Sculpins is Associated with High Anaerobic Enzyme Activity in Brain, but Not in Liver or Muscle ............................................... 12 2.1 Summary ................................................................................................................................. 12 2.2 Introduction ........................................................................................................................... 13 2.3 Materials and Methods ....................................................................................................... 16 2.3.1 Experimental animals ................................................................................................ 16 2.3.2 Hypoxia tolerance experiments ............................................................................. 17 2.3.3 Analytical procedures ................................................................................................ 18 2.3.4 Statistical analysis ....................................................................................................... 21  vi 2.4 Results ...................................................................................................................................... 22 2.4.1 Phylogeny ....................................................................................................................... 22 2.4.2 Hypoxia tolerance ........................................................................................................ 23 2.4.3 Scaling .............................................................................................................................. 23 2.4.4 Brain ................................................................................................................................. 24 2.4.5 Liver .................................................................................................................................. 24 2.4.6 Muscle .............................................................................................................................. 25 2.5 Discussion ............................................................................................................................... 26 2.5.1 Hypoxia tolerance ........................................................................................................ 26 2.5.2 Brain enzymes ............................................................................................................... 27 2.5.3 Standardization of enzyme activity ...................................................................... 28 2.5.4 Muscle and liver enzymes ........................................................................................ 29 2.5.5 Metabolites ..................................................................................................................... 30 2.5.6 Conclusions .................................................................................................................... 31 Chapter 3 Divergent Transcriptional Patterns are Related to Differences in Hypoxia Tolerance Between the Intertidal and the Subtidal Sculpins ............... 40 3.1 Summary ................................................................................................................................. 40 3.2 Introduction ........................................................................................................................... 41 3.3 Materials and Methods ....................................................................................................... 44 3.3.1 Experimental animals ................................................................................................ 44 3.3.2 Hypoxia time-course .................................................................................................. 44 3.3.3 Microarray hybridization ......................................................................................... 45 3.3.4 Microarray probe sequences ................................................................................... 47 3.3.5 Statistical analyses ...................................................................................................... 48 3.4 Results ...................................................................................................................................... 48  vii 3.4.1 Experiment 1: Relative hypoxia exposure time course ................................ 49 3.4.2 Experiment 2: Absolute environmental hypoxia............................................. 51 3.5 Discussion ............................................................................................................................... 52 3.5.1 Transcription of genes similarly induced or repressed by hypoxia in both species .................................................................................................................................. 53 3.5.2 Transcriptionally divergent response between the species ....................... 54 3.5.3 Relative vs. absolute levels of hypoxia ................................................................ 56 3.5.4 Conclusions .................................................................................................................... 57 Chapter 4 Variation Underlying Similar Phenotypes: Transcriptional Divergence in Hypoxia Responsive Genes Associated with Metabolism and Protein Synthesis in Three Species of Sculpin with the Same Hypoxia Tolerance…. ............................................................................................................................. 66 4.1 Summary ................................................................................................................................. 66 4.2 Introduction ........................................................................................................................... 67 4.3 Materials and Methods ....................................................................................................... 70 4.3.1 Experimental animals ................................................................................................ 70 4.3.2 Critical O2 tension ........................................................................................................ 70 4.3.3 Hypoxia exposure ........................................................................................................ 71 4.3.4 Microarray experiments ........................................................................................... 71 4.3.5 Statistical analysis ....................................................................................................... 71 4.4 Results and Discussion ....................................................................................................... 73 4.4.1 Energy metabolism ..................................................................................................... 74 4.4.2 Protein production, protein localization and protein folding .................... 76 4.4.3 Transcription patterns of hypoxia intermediate-tolerant species versus a hypoxia-tolerant species ...................................................................................................... 77 4.4.4 Wild caught versus lab-reared species ................................................................ 78  viii 4.4.5 Short-term versus long-term responses ............................................................. 78 4.4.6 Conclusion ...................................................................................................................... 79 Chapter 5 General Discussion ........................................................................................ 91 5.1 Major Findings and Implications ................................................................................... 91 5.1.1 Measures of hypoxia tolerance ............................................................................... 91 5.1.2 Multiple candidate traits underlie variation in hypoxia tolerance among sculpins........................................................................................................................................... 92 5.1.3 Short- versus long-term hypoxia exposure ....................................................... 95 5.1.4 The role of constitutive differences in phenotype versus phenotypic plasticity in the intertidal sculpins ...................................................................................... 96 5.1.5 Convergent evolution of hypoxia tolerance ...................................................... 98 5.2 Future Directions ................................................................................................................. 98 5.3 Concluding Thoughts ........................................................................................................ 100 Bibliography .......................................................................................................................... 103 Appendix A ............................................................................................................................. 130 Appendix B ............................................................................................................................. 131 Appendix C ............................................................................................................................. 133      ix List	  of	  Tables Table 2.1  Body weights of species of sculpin used in the LOE50 experiment and in analytical procedures. ........................................................................................................................ 37 Table 2.2  Tissue enzymes in 12 species of sculpin. ............................................................... 38 Table 2.3  Tissue metabolites in 12 species of sculpin. ......................................................... 39 Table 3.1  Experimental design of O2 exposures and sampling times for each species ..................................................................................................................................................................... 64 Table 3.2  Euclidean distance of the transcriptional response between the tidepool sculpin and the silverspotted sculpin measured for the 72 hour time course. ........... 65 Table 4.1  Summary of clones with similar and different transcription patterns among the three species of sculpin. .............................................................................................. 90 Table 5.1  Summary of traits examined in 5 species. ........................................................... 101     x List	  of	  Figures Figure 1.1 Time-scale of a stress response ................................................................................ 11 Figure 2.1 Phylogenetic relationship among 15 species of sculpin based on a maximum-likelihood tree using cytochrome b sequences. ................................................. 33 Figure 2.2  Relationship between Pcrit and LOE50. ................................................................... 34 Figure 2.3 Relationships between Pcrit and maximal enzyme activities of lactate dehydrogenase (LDH), pyruvate kinase (PK), creatine phosphokinase (CPK) and citrate synthase (CS) in the brain of 12 species of sculpin. ................................................. 34 Figure 2.4  Relationship between Pcrit and soluble protein concentration in brain, liver, and muscle. ................................................................................................................................. 35 Figure 3.1  Venn diagram depicting the number of clones whose transcription significantly changed in response to hypoxia. .......................................................................... 58 Figure 3.2  Trajectory of transcription in the tidepool sculpin and the silverspotted sculpin exposed to relative environmental hypoxia over 72 hours and absolute environmental hypoxia at 24 hours. ............................................................................................ 59 Figure 3.3  Transcription levels of genes associated with metabolism (A-D), regulation of cell number (E-H) and immunity (I) in the tidepool sculpin and silverspotted sculpin. ......................................................................................................................... 60 Figure 3.4  Transcription levels of genes associated with protein production (A-E), protein localization (F) and protein folding (G-H) in the tidepool sculpin and the silverspotted sculpin. ......................................................................................................................... 62 Figure 4.1  A sculpin phylogenetic tree based on cytochrome b sequences. ................ 80 Figure 4.2  Schematic representation of genes associated with metabolism. .............. 81 Figure 4.3  Schematic representation of genes associated with protein production, localization and folding. .................................................................................................................... 82 Figure 4.4  Transcript levels of genes associated with metabolism in the smoothhead sculpin (black line, square symbol), sailfin sculpin (blue line, diamond symbol) and Pacific staghorn sculpin (red line, inverted triangle symbol) exposed to 72 hours of hypoxia. .................................................................................................................................................... 83  xi Figure 4.5  Genes for which all three species showed similar significant changes in transcription in response to 72 hours of hypoxia. .................................................................. 85 Figure 4.6  Similar transcription patterns in genes for which only a subset of species showed a significant effect of time in hypoxia. ......................................................................... 86 Figure 4.7  Transcript levels of genes associated with protein production, localization and folding in the smoothhead sculpin (black line, square symbol), sailfin sculpin (blue line, diamond symbol) and Pacific staghorn sculpin (red line, inverted triangle symbol) exposed to 72 hours of hypoxia................................................. 88         xii Acknowlodgements  Writing acknowledgements is difficult. And in this case it is particulary difficult as I am not sure I will be able to adequately convey the extent of how fortunate, grateful and appreciative I am to have met wonderful and amazing people during my rather long tenure at UBC as an ungraduate,	  Masters’	  and	  Doctorate	  student. A deep and heartful thanks to my supervisor Jeffrey Richards, who has shaped and influenced the way I think and write and without his guidance I would be a much poorer scientist indeed. I have also been very fortunate to have had Patricia Schulte, whom I greatly admire, as a mentor. The	  Richards’	  lab,	  both	  past	  and present, have been an incredible group of people for science, friendship and my well-being, so a very big thanks to Ben Speers-Roesch, Gigi Lau, Matt Regan, Rush Dhillon, Andrew Thompson, Tammy Rodela, Cris Gomez, Derek Somo, Josh Emerman, Derrick Groom, Gina Galli, Lili Yao and Mark Scott.   During my graduate time, more so than any other period of my life, I have met and become close to a number of people who have had an incredible impact on my life. These are the friends that have enriched me beyond measure and whose friendship I will enjoy for the rest of my life. To my extraordinary women: Anne Dalziel, who is and will always be a part of my family, Aleeza Gerstein, a force of nature whose fierceness and loyalty I adore and admire, Georgie Cox whose generosity in friendship is unparalled, Yvonne Dzal, the perfect balance of sweet kindness and fiery nature, Gigi Lau whose short stature hides a lion’s	  strength, Katelyn Tovey who has sneakingly crept into my heart, Carla Crossman my temporary roommate and permanent friend, Jasmine Ono who is a tough nut to crack but incredibly worth it, Gina Conte my poised friend whose smile brightens my day, Julie Lee-Yaw the strong and intelligent woman that views life under a different lens, Kim Sujvadzic the rock in my life and Ada Smailbegovic the longest friendship that will never wane. To my remarkable men: Ben Speers-Roesch, my  xiii older red-headed brother, Matt Regan, my challenger and confidante, Mike Sackville whom I treasure more than he realizes (even on the days I want to throttle him), J.S. Moore who let me into his life and made mine all the richer for it, Rush Dhillon the long-standing and forever continuing Christmas time date, Matt Siegle the protector who keeps me inline, Dave Toews my scientific partner-in-crime and always friend and Steve Haynes a part of my life unmarred by graduate work stress. And to Seth Rudman, Cris Gomez, Ken Thorne, John Yensen, Andrew Thompson, Taylor Gibbons, Jessica Meir, Benny Goller, Tammy Rodela, Angie	  O’Neill,	  Emily Gallagher, Michelle Ou, Cathy Harman, Jessica Hill, Jon Mee, Sam Yeaman, Alana Schick, Till Harter and Stephanie Avery-Gomm thank you for letting me be a part of your life.  За маму и	  тату ово не	  би	  било	  могуће	  без	  вас. Хвала	  за	  безусловну	  подршку	  и	  љубав	  коју	  сте	  ми	  показали,	  и	  за	  ваше	  разумјевање	  током	  стреснога	  периода. To Aco and Bojan, you are truly the very best older brothers in the world. Your unwaivering support and willingness to support me in whatever I do throughout my life has meant more to me than I can ever truly express.     xiv Dedication         This thesis is dedicated to Момчило and Cлободанка Мандић   1  Chapter 1  General Introduction A fundamental goal of comparative physiology is to understand how environmental stresses have shaped the functional physiological responses of an organism and how these responses contribute to defining the geographical distribution	  of	  a	  species.	  Although	  “stress”	  is	  defined	  in	  different	  ways	  in	  different	  scientific fields, environmental stress can be defined as an environmental factor that results in a loss of fitness of an organism or a population, i.e. results in loss of reproductive output (Bijlsma and Loeschcke 2005, Schulte 2014). As organisms most often experience environmental conditions that are not optimal for maximum fitness, the definition of stress is more strictly applied to environmental conditions that have a significant impact on the fitness of individuals in a population (Bijlsma and Loeschcke 2005). Thus environmental stresses impose a strong influence on a species’	  ability to survive in a habitat. By observing the environments where organisms exist, we can acquire insight on their evolved stress responses that compensate or buffer against the (potentially) deleterious effects of the environmental conditions.     There are different aspects of environmental stress that can influence the adaptive trajectory of an organism, mainly the frequency, intensity and the time-scale of exposure to the environmental stress. Focusing on the temporal aspect, an environmental stress may be of short or long duration, or sustained within or across generations, with the differences in the length of time of exposure to stress influencing the mechanisms used to alleviate the effects of stress (see Fig. 1.1). For example, immediate or acute responses of an organism may be sufficient for initial survival to an environmental stress or if the length of the stress exposure is short.  2 Longer	  durations	  of	  stress	  within	  an	  individual’s	  lifetime	  may	  lead	  to	  improvement	  in the performance of the individual via reversible phenotypic plasticity or via irreversible developmental plasticity if the exposure to stress occurs during developmental stages (Stearns 1989, Gabriel 2005, Gabriel 2006, Ghalambor et al. 2007). Epigenetic changes may improve compensatory responses of individuals across a few generations, while across many generations evolutionary change may shift populations or species closer to adaptive optimum for the environmental stress (Burggren 2014, Schulte 2014). Thus, compensatory mechanisms that work at different time-scales can decrease the negative effect of environmental stress and improve survival and fitness of individuals within a population or species.  Phenotypic plasticity may impact adaptive evolution and in turn, adaptation may play a significant role in the evolution of phenotypic plasticity. One view-point regarding the impact of phenotypic plasticity on adaptation is that environmentally-induced phenotypic variation constrains adaptive evolution by decreasing the effect of natural selection on genotype (Stebbens 1977, Grant 1991). In contrast, others propose that phenotypic plasticity may facilitate adaptation through a process known	  as	  ‘genetic	  assimilation’,	  whereby	  the	  phenotypic	  change	  that	  allows	  survival	  and reproduction of individuals in response to a new environmental stress becomes constitutively expressed in the population even if the environmental stress is no longer present (Robinson and Dukas 1999, Ghalambor et al. 2007). Therefore phenotypic plasticity, whether through constraint or facilitation of adaption, plays a role in the evolutionary trajectory of a population or a species in response to environmental stress.  Adaptation, in turn, can impact the evolution of phenotypic plasticity as natural selection may favor either environmentally-induced plastic traits or fixed, constitutively expressed traits (Berrigan and Scheiner 2004, Callahan et al. 2008). More specifically, several other strategies may be selected over phenotypic plasticity	  in	  heterogeneous	  environments	  (‘heterogeneous’	  encompassing	  either	  changes from non-stress to stress conditions or from one type of environmental stress to a different one). These include the 1) specialist: single phenotype is produced that is optimal for one set of environmental conditions, 2) generalist:  3 single phenotype is produced that is moderately successful in most environments; also	  termed	  ‘conservative	  bet-hedging’	  and	  3)	  diversified	  bet-hedger: equal probability of producing alternate phenotypes (Philippi and Seger 1989, Wilson and Yoshimura 1994, Dewitt and Langerhans 2004, Botero et al. 2015). Theoretical modeling and empirical studies have examined the environmental conditions that may favor one strategy over another in a heterogeneous environment and these include predictability of environmental change, reliability of environmental cues, frequency and intensity of change, length of environmental stress and costs associated with phenotypic plasticity (Moran 1992, Padilla and Adolph 1996, Schlichting and Smith 2002, Gabriel 2005, Botero et al. 2015). If environmental change is predictable, environmental cues are reliable or there is low cost associated with altering phenotype, selection should favor phenotypic plasticity in response to an environmental stress. As the role of adaptation and phenotypic plasticity on the stress response of a population or species are interlinked, my doctoral research examined the interplay between these two processes in hopes of gaining a better understanding how organisms evolved to cope with an environmental stress, with a specific focus on low O2 availability or hypoxia in the environment. 1.1 Environmental Stress - Hypoxia Periodic hypoxia is a widespread abiotic stress, occurring across diverse aquatic environments such as ponds, lakes, rivers, marine coastal areas and regions of intermediate oceanic depth (400-1000 meters; Diaz and Breitburg 2009). Water is 1000 times denser than air and 50 times more viscous; O2 diffuses 10,000 times slower through water than air and there is 20 to 40 times less O2 contained in water than in the same volume of air, depending on temperature and salinity (Graham 1990). As a consequence, in part, of these physical properties of water, aquatic ecosystems are particularly susceptible to fluctuations in O2 and hypoxia occurs when respiration exceeds the production and diffusion of O2 in the environment. Occurrence of hypoxic events is also attributed to many other interacting components of the aquatic ecosystem, including density of vegetation, loads of  4 organic materials and the degree and duration to which bodies of water become isolated. Together, these elements create complex, heterogeneous aquatic environments that are marked with intermittent decreases in O2 levels to varying degrees of magnitude, frequency and duration (Burggren and Roberts 1991, Val 1999). This has led to the evolution of a variety of behavioral, morphological, physiological, biochemical and molecular responses among fishes (Nikinmaa 2002). Hypoxic events can last for hours, days or even months. Short-term hypoxia is prevalent and often severe in coastal waters such as salt marshes, estuaries and intertidal zones (Truchot and Duhamel-Jouve, 1980). Long-term hypoxia can occur in river systems such as the Amazon during the dry season when water levels decrease and there is low water mixing coupled with high water plant cover and high organic decomposition (Val, 1999), as well as in shallow temperate lakes and ponds that are prone to ice coverage in the winter (van den Thillart et al. 1989). While it is clear that hypoxia is a natural occurrence in aquatic habitats, the extensiveness, severity and duration of hypoxic zones has dramatically increased since around the 1950s as a result of anthropogenic eutrophication of both the freshwater and coastal ecosystems (Diaz and Breitburg, 2009). A number of these hypoxic areas have been classified as dead zones due to massive mortality of vertebrate and invertebrate organisms, leading to increasing global concern about the impact of hypoxia, especially anthropogenic-induced hypoxia, on the survival of productive ecosystems (Diaz 2008, Vaquer-Sunyer and Duarte 2008). Clearly, it is critical that we understand the mechanisms organisms use to cope with hypoxic stress in order to identify the effect of ever-increasing hypoxic challenges on biological diversity. 1.2 Time-Scale of the Hypoxia Response Rapid changes in the environment necessitate immediate behavioural, physiological and metabolic responses in order to maintain function, and many acute modifications to hypoxia exposure have been noted across diverse fish taxa. Behavioural responses in fishes include the avoidance of low O2 zones (Dalla Via et al. 1998, Lefrancois and Domenici 2006), reduction of spontaneous swimming  5 activity (Chapman and McKenzie 2009), aquatic surface respiration and air-breathing (Martin 1995, Yoshiyama et al. 1995, Graham 1997, Watters and Cech 2003). A common acute respiratory modification to low O2 conditions is the hypoxic ventilatory response, which is the increase in ventilation volume as a result of changes in breathing frequency and/or amplitude in order to increase arterial PO2 (reviewed in Perry et al. 2009). Bradycardia has also been shown to occur in a number of fish species within minutes of hypoxia exposure, the benefits of which include increased time for O2 diffusion in the myocardium and a reduction in myocardial O2 demand (Farrell 2007, Gamperl and Driedzic 2009). Hypoxia-induced acute responses may be sufficient for short-term survival, which in turn may allow phenotypic changes to occur that can enhance the capacity of the organism to endure hypoxia exposure (Richards 2009).   Phenotypic plasticity or the modification of phenotypic expression in response to changing environmental conditions is viewed as an important strategy in adaptation to variable, heterogeneous environments (e.g. Stearns 1989, Moran 1992, Via et al. 1995, Gabriel 2005, Ghalambor et al. 2007, Pfenning et al. 2010). Indeed, reversible phenotypic plasticity, irreversible developmental plasticity and epigenetics have all been noted to play a role in hypoxic stress response of fishes. For example, transcription of genes involved in glycolysis, oxidative phosphorylation, iron/heme metabolism cell proliferation and protein metabolism have been found to change in fish species in response to hypoxic exposure (Gracey et al. 2001, Ju et al. 2007, Everett et al. 2011, Leveelahti et al. 2011). These hypoxia-induced gene transcription patterns have been presumed to alter phenotype and therefore are considered an important part of the plastic response. Other reversible phenotypic changes as a result of long-term hypoxic events include gill remodeling to increase respiratory surface area (Sollid et al. 2003, Nilsson 2007), decreases in hemoglobin (Hb) modulators that enhance Hb-O2 binding affinity (Weber and Lykkeboe 1978, Tetens and Lykkeboe, 1981) and changes in traits associated with both energy supply and energy demand (Hochachka 1986). Energy supply during hypoxic exposure becomes limited but changes in the concentration of enzymes associated with carbohydrate metabolism over time may increase capacity of  6 pathways of ATP production (Martinez et al. 2006, Martinez 2011). To conserve the lower energy supply, fishes such as the goldfish (Carassius auratus) turn down processes with high energy expenditure, profoundly reducing metabolic energy demand (van Waversveld et al. 1989). This can occur rapidly at the onset of hypoxia but is maintained throughout the hypoxic exposure and lowering metabolic demand to match hypoxia-induced reduction in energy supply is one of the classic strategies hypoxia-tolerant species use to combat the effects of long-term hypoxia (van Waversveld et al. 1989, Lutz and Nilsson 1997, Bickler and Buck 2007, Lewis et al. 2007, Hochachka 1986, Hochachka et al. 1996, Boutilier 2001). In addition to reversible phenotypic plasticity, alteration of phenotypic expression over generations via epigenetics has been noted in zabrafish (Danio rerio) where parental exposure to hypoxia increased offspring hypoxia tolerance as measured by time to loss of equilibrium (Ho and Burggren 2012), while changes in adult critical oxygen tension were also found in zebrafish as a result of developmental exposure to hypoxia (Robertson et al. 2014). The many noted examples of phenotypic changes in response to low O2 across diverse fish lineages suggests that phenotypic plasticity is an important strategy for increasing survival and reducing the impact of hypoxic stress.        Intraspecific and interspecific variation in fixed physiological traits in different taxa have been thought to have evolved in response to O2 availability in the environment. Specifically, higher gill surface area and greater total gill filament length, have been found in populations living in O2-poor environments versus well-oxygenated environments in sailfin mollies (Poecilia latipinna; Timmerman and Chapman 2001) and African cyprinids, Barbus neumayeri (Chapman et al. 1999) respectively. It has been thought that a number of changes at the level of the hemoglobin have evolved in response to hypoxic selective pressure, including high concentration of Hb and hematocrit among fishes in the Lake Victoria region (Chapman et al. 2002), higher number of Hb isoforms with a greater proportion of the higher Hb-O2 affinity between intertidal and deeper water triplefin fishes (Brix et al. 1999), and amino acid substitution leading to higher Hb-O2 affinity in the high altitude bar-headed goose in comparison to low altitude geese species (Perutz 1983,  7 Jessen et al. 1991). These studies are informative examples of how populations or species that vary in O2 regimes have potentially evolved differences in fixed physiological traits. However, many studies examining traits underlying hypoxia tolerance have not employed methods to test if the traits have evolved by natural selection in response to hypoxic pressure. The most direct method to ascertain if a trait is adaptive is to measure the fitness of alternate alleles in a population (Endler 1986, Dalziel et al. 2009), a task that may be difficult to accomplish in many taxa. Alternately, comparative phylogenetic analysis may be used to determine if candidate traits are potential adaptations (Pierce and Crawford 1997, Feder et al. 2000, Darveau et al. 2005a, Garland et al. 2005). These include, for example, a phylogenetically corrected examination of traits in closely related species with varying levels of hypoxia tolerance or a comparison of traits in populations or species with independently evolved similar level of hypoxia tolerance (instances of parallel or convergent evolution are thought to be a result of natural selection; Endler 1986). The use of these methods is critical in determining if candidate traits are potential adaptations to environments with periodic decreases in O2. 1.3 Levels of Biological Organization and the Hypoxic Response  Coordination and integration of many traits contribute to complex phenotypes such as hypoxia tolerance. As such, many investigators in the field of comparative physiology have called on a systematic examination of levels of biological hierarchy as an integrative, mechanistic approach to understanding how species respond to environmental challenges (Dalziel et al. 2009, Gaston et al. 2009, Mykles	  et	  al.	  2012,	  Somero	  2012).	  This	  ‘vertical	  integration’	  of	  traits	  (Mykles	  et	  al.	  2012), from gene transcription, protein function or amount, biochemical and cellular pathways, physiological traits, morphological characteristics to behavioural responses, is important in order to understand how variation in environmental stress leads to variation in stress response among different populations or species. Technological advances and	  the	  advent	  of	  the	  ‘–omic’	  era	  in	  particular	  have	  provided an array of tools, such as whole-genome sequencing data, transcriptomics, epigenomics, proteomics and metabolomics, to address the relationship between  8 genetic variation, complex phenotypes and the environment (Ritchie et al.  2015), which can be applied to non-model systems (Cossins and Crawford 2005, Buckley 2007, Forne et al. 2010). The levels of biological organization underlying hypoxia tolerance have been examined across diverse fish species, from behaviour (Watters and Cech 2003), morphology (Sollid et al. 2003, Chapman et al. 2002), physiology (Nonnotte et al. 1993, Rutjes et al. 2007), biochemistry (DiAngelo and Heath 1987, Chippari-Gomes et al. 2005, Farwell et al. 2007, Richards et al. 2008), cellular processes (Kocha et al. 2014, Onukwufor et al. 2014), proteomics (Smith et al. 2009, Chen et al. 2013), gene transcription (Gracey et al. 2001, Ton et al. 2003, Everett et al. 2010) to the role of specific genes (Rytkonen et al. 2007). Combining this multi-level approach to a study system is a powerful way to examine the relationship of how variation in underlying traits affects complex phenotypes in face of hypoxic challenge. 1.4 Model Species: Sculpins (Superfamily Cottoidea)  August Krogh has famously	  stated	  “For	  a	  large	  number	  of	  problems	  there	  will	  be some animal of choice, or a few such animals, on which it can be most conveniently	  studied”	  (Krogh	  1929).	  This	  quote	  exemplifies	  the	  comparative	  physiology approach of carefully considering the most appropriate organisms to study given the questions under investigation. The comparative approach of examining populations and species that have adapted to different environmental conditions has been a staple of physiology as it allows us to infer the evolutionary processes that had shaped the physiological and ecological landscape (Garland et al. 2005). Included in this approach is the explicit incorporation of phylogenetic information and the application of phylogenetically independent contrasts (PIC) in order to take into account phylogenetic relationships and species relatedness when drawing inference from correlative data (Felsenstein 1985, Garland et al. 1992). To address the question of how hypoxia tolerance evolved several criteria need to be considered when selecting model systems: 1) organisms must show variation in hypoxia tolerance, 2) the level of hypoxia tolerance matches the likelihood of hypoxic exposure in the habitat of the chosen species and 3) phylogenetic  9 relationships need to be elucidated in order to take into account non-independence in the data as a result of species relatedness. As such, one ideal model study system to address important questions regarding evolution of hypoxia tolerance are sculpins, small benthic fish species living in the nearshore environment.     The marine intertidal zone is an aquatic environment susceptible to O2 fluctuations as a result of the ebb and flow of tides, with hypoxic and even anoxic (no O2) events occurring in isolated tidepools at night (Truchot and Duhamel-Jouve 1980, Richards 2011). In contrast, hypoxia is less frequent or severe in the subtidal zone and O2 levels remain stable and high in the pelagic and deeper coastal waters. Marine sculpins, a group of fishes belonging to the superfamily Cottoidea, are distributed along the nearshore environment with species inhabiting the intertidal zone, subtidal zone and deeper waters. As predicted by the habitat of the species, sculpins living in the intertidal zone are more hypoxia-tolerant than subtidal species and sculpins from both environments are more hypoxia-tolerant than the deeper water species (Mandic et al. 2009b). A number of behavioural responses and fixed physiological traits have been found to underlie variation in hypoxia tolerance among the different sculpin species. Gill morphology and hemoglobin properties significantly correlated with critical O2 tension (Pcrit; a proxy measure of hypoxia tolerance) independent of phylogenetic relationships, where the more hypoxia-tolerant species were found to have larger gill surface area and higher Hb-O2 binding affinity compared to the more intolerant species of sculpin (Mandic et al. 2009b). These traits suggest that the variation in O2 extraction and delivery may partially underlie the variation in hypoxia tolerance among sculpin species. In response to acute hypoxia exposure, the hypoxia-tolerant intertidal species consistently displayed behavioural modifications, mainly aquatic surface respiration followed by aerial respiration, while less tolerant species did not exhibit these behaviours or performed them at low frequencies (Mandic et al. 2009a). Overall, these previous studies have shown that the sculpins are a tractable model system as they have begun to address the question of what underlying traits may be important in adaptation to environmental hypoxia. In particular, the past studies and this thesis have begun to apply the approach of using multiple biological levels  10 (behavioural, physiological, biochemical and gene transcription) with temporal time-scales (acute responses, phenotypic plasticity, fixed putative adaptations) to examine how variation in a complex phenotype may have evolved in response to an environmental stress.  1.5 Research Goals  The overall objective of my doctoral research was to examine how adaptation and phenotypic plasticity have shaped the hypoxic response in sculpin species distributed along the nearshore environment. Specifically, I focused on how variation in fixed biochemical traits and transcriptionally-induced phenotypic plasticity correlated with hypoxia tolerance among the different sculpin species. My first objective was to determine if the previously assessed critical O2 tension (Pcrit) in species (Mandic et al. 2009b) is related to hypoxia tolerance in sculpins. I quantified loss of equilibrium of 50% of a population (LOE50) in 11 species of sculpin and correlated this measure of hypoxia tolerance to the previously assessed Pcrit values (Chapter 2). My second objective was to determine whether the variation in biochemical traits associated with energy metabolism, explained, at least in part, the differences in the ability of the species to tolerate low levels of O2. To address this objective, I quantified maximal activity of enzymes and metabolite concentrations associated with ATP production in brain, liver and white muscle of 12 species of sculpin under normoxic conditions (Chapter 2). Phylogenetically independent contrast analysis (Felsenstein 1985) was applied to all correlations to take into account phylogenetic relationships, allowing for more confidence that the significant correlations were less likely a result of relatedness and instead can be considered putative adaptive traits.   My subsequent objectives focused on defining the relationship between levels of hypoxia tolerance and phenotypic plasticity via changes in gene transcription. Specifically, my third objective was to determine how species with different levels of hypoxia tolerance transcriptionally respond to hypoxic stress over time. To address this objective, I assessed gene transcription patterns of the hypoxia-tolerant tidepool sculpin (Oligocottus maculosus) and the hypoxia- 11 intolerant silverspotted sculpin (Blepsias cirrhosis) during exposure to hypoxia (Chapter 3). Both species were exposed to two sets of hypoxic conditions; a relative hypoxia exposure where individuals were exposed to O2 tensions scaled to the species’	  hypoxia	  tolerance	  (65%	  Pcrit) and an absolute hypoxia exposure where individuals were exposed to a single environmental O2 tension. My fourth objective was to determine if species of sculpin with similar tolerances to environmental hypoxia also exhibit convergence at the level of gene transcription. Specifically, gene transcription patterns over short- and long-term hypoxia exposure were assessed in three similarly hypoxia-tolerant species of sculpin, in order to provide insight into whether the species evolved similar hypoxia tolerance via convergent or divergent mechanisms (Chapter 4).  Figure 1.1 Time-scale of a stress response      12 Chapter 2  Hypoxia Tolerance in Sculpins is Associated with High Anaerobic Enzyme Activity in Brain, but Not in Liver or Muscle1 2.1 Summary We assessed hypoxia tolerance in 11 species of fish from the superfamily Cottoidea (commonly called sculpins) that are known to differ in their critical O2 tensions (Pcrit) and examined whether hypoxia tolerance correlated with larger substrate stores and higher maximal activity of enzymes associated with anaerobic ATP production (especially glycolysis). Among the sculpins studied, there was large variation in time to loss of equilibrium (LOE50) at 6.4±0.1 torr with values ranging between 25 and 538 minutes and the variation in LOE50 was correlated with Pcrit. Our measures of time to LOE50 and Pcrit were regressed against maximal enzyme activities of lactate dehydrogenase (LDH), pyruvate kinase (PK), creatine phosphokinase (CPK) and citrate synthase (CS) as well as the concentrations of glycogen, glucose and creatine phosphate in the brain, liver and white muscle. In the brain, there was a phylogenetically independent relationship between Pcrit and tissue LDH, PK, CPK and CS activities expressed relative to tissue mass. Hypoxia-tolerant sculpins (those with low Pcrit values) had higher levels of brain LDH, PK,                                                         1 A version of this chapter has been published: Mandic M., B. Speers-Roesch and J.G. Richards. 2013. Hypoxia tolerance is associated with high anaerobic enzyme activity in brain but not in liver or muscle. Physiological and Biochemical Zoology 86: 92-105  13 CPK, and CS than hypoxia-intolerant sculpins. Similarly, LOE50 regressed against brain LDH, PK and CPK activities expressed relative to tissue mass, with the more hypoxia-tolerant species (ie. higher LOE50) having higher enzyme activities. However, when the phylogenetic relationship among our sculpins was taken into account, only the relationship between hypoxia tolerance and LDH activity remained significant. When enzyme activities were expressed relative to total soluble protein in the tissue, the only relationships that remained were between brain LDH activity and Pcrit and LOE50. In liver and white muscle, there were no relationships between the measures of hypoxia tolerance and enzyme activity or metabolite content. Overall, our analysis suggests that hypoxia-tolerant sculpins maintain higher maximal activities of some of the enzymes involved in anaerobic metabolism in the brain and this may be an adaptation to hypoxia.  2.2 Introduction In order to survive periods of low environmental O2 (hypoxia), fish must be able to acquire more O2 from the hypoxic environment or respond at the biochemical level by reducing energetic demands or enhancing O2-independent means of ATP production. Initial compensatory mechanisms during environmental hypoxia are geared towards enhancing O2 extraction and delivery to the tissues in order to maintain aerobic energy production. However, if the severity of hypoxia increases and tissue O2 levels drop, fish face limitations to their metabolic energy supply as a result of reduced capacity to fully oxidize carbohydrate, lipid and amino acid fuels. To maintain metabolic energy balance and prevent hypoxia-induced cellular death, hypoxia-tolerant animals are thought to down-regulate ATP consuming pathways and shift ATP generation toward substrate-level phosphorylation (i.e., anaerobic or O2-independent energy production) by increasing glycolytic flux as well as creatine phosphate (CrP) hydrolysis (e.g. in Astronotus ocellatus, Richards et al. 2007). Hypoxia-intolerant species are thought to be unable to reduce ATP demand during hypoxia and therefore quickly exhaust their limited substrates for O2-independent energy production (Boutilier and St-Pierre 2000; Boutilier 2001). As such, at low O2 tensions, hypoxia survival is  14 dictated by the capacity to reduce cellular ATP demands and up-regulate or sustain prolonged periods of ATP generation via substrate-level phosphorylation. It has been postulated, therefore, that hypoxia tolerance in fishes and other ectotherms is associated with a greater capacity to generate ATP via substrate-level phosphorylation (Richards 2009).  Higher substrate levels for anaerobic metabolism have long been associated with hypoxia tolerance (Van den Thillart and Van Raaij 1995; Chippari-Gomes et al. 2005). For example, the impressive hypoxia tolerance of goldfish (Carassius auratus) and Crucian carp (Carassius carassius) is, in part, attributed to large hepatic glycogen stores, which are accumulated prior to the onset of severe hypoxia during wintering under the ice (Walker and Johansen 1977; Hyvarinen et al. 1985). The winter glycogen levels in the goldfish and Crucian carp are 10 to 45-fold higher than in the hypoxia-intolerant Atlantic cod (Gadus morhua) and rainbow trout (Oncorhynchus mykiss; Van den Thillart and Van Raaij 1995), suggesting that high glycogen stores may be an important adaptive trait for hypoxia survival in fishes. Hypoxia tolerance is also associated with higher maximal activities of some glycolytic enzymes, particularly lactate dehydrogenase (LDH; Almeida-Val et al. 2000; Chippari-Gomes et al. 2005; Farwell et al. 2007). In centrarchids, the more hypoxia-tolerant pumpkinseed, Lepomis gibbosus, has higher levels of LDH in white muscle than the hypoxia-intolerant bluegill, Lepomis macrochirus (Farwell et al. 2007). While variation in maximal enzyme activity appears to be related to hypoxia tolerance, it is unclear what effect this variation has on overall metabolic flux since in vivo flux rates are regulated primarily by allosteric and covalent modification of regulatory enzymes (e.g. hexokinase, phosphofructokinase and pyruvate dehydrogenase; Newsholme and Crabtree 1986). Despite the lack of a clear mechanistic understanding of how maximal enzyme activity affects metabolic flux rates, variation in maximal enzyme activities seen among species and populations of fish has been attributed to natural selection (e.g. Crawford and Powers 1989; Pierce and Crawford 1997), thus suggesting that this variation is functionally important. In particular, Pierce and Crawford (1997) showed a significant correlation between habitat temperature and maximal enzyme activity of LDH, glyceraldehyde 3- 15 phosphate dehydrogenase and pyruvate kinase (PK) in the heart of 11 species of fish from the genus Fundulus, and the authors argued strongly that maximal enzyme activity was under selection.  In order to determine if environmental hypoxia acts as a selective pressure on substrate concentration and maximal activity of enzymes involved in anaerobic metabolism, we chose to examine these variables in sculpins, a group of small benthic fishes belonging to superfamily Cottoidea. We have previously shown that there is variation in critical O2 tension, or Pcrit, which is a whole animal measure of O2 extraction capacity, among the species of sculpin studied and that Pcrit correlates with habitat. The more hypoxia-tolerant species (lower Pcrit) inhabit the O2-variable tidepools and the more hypoxia-intolerant species (higher Pcrit) are located in the more stable subtidal and deeper waters (Mandic et al. 2009). A multi-species comparison is only effective, however, if the phylogenetic relationship is taken into account	  since	  species’	  means	  cannot	  be	  treated	  as	  statistically	  independent	  due	  to	  the differences in the degree of relatedness among the species (Felsenstein 1985; Garland et al. 2005). Therefore, application of the phylogenetically independent contrast (PIC) analysis, which uses phylogenetic information to transform species data into standardized independent contrasts effectively eliminates non-independence from the data set (Felsenstein, 1985). Taking into account the phylogenetic relationship among sculpins, we hypothesize that the hypoxia-tolerant species have larger anaerobic substrate stores and higher maximal activity of enzymes associated with O2-independent ATP generating pathways compared with the more hypoxia-intolerant species of sculpin.            In this study, we quantified the concentrations of the three main substrates of anaerobic ATP production (glycogen, glucose and CrP) and the maximal enzyme activities of LDH, PK, creatine phosphokinase (CPK), and citrate synthase (CS) in brain, liver and muscle of 12 species of sculpin acclimated to normoxia. We chose to examine LDH and PK because maximal activity of these enzymes has previously been shown to be under selection in fish exposed to environmental stress (Pierce and Crawford 1997).  Creatine phosphokinase was examined because it is part of an enzyme system that plays an important role in anaerobic generation of ATP via CrP  16 hydrolysis. An enzyme from aerobic metabolism, CS, was also selected for analysis in order to compare variation in anaerobic enzymes with an enzyme involved in aerobic metabolism. Brain, liver, and muscle were selected for analysis because it is likely that tissue-specific profiles have evolved in response to differences in O2 supply, energetic demand, and metabolic role of the tissue (Martinez et al. 2006). Brain is well known as a hypoxia-sensitive tissue (Boutilier 2001) and the maintenance of stable cellular ATP is critical to survival. Liver and muscle were chosen because previous studies have shown that these two tissues respond differently to long-term hypoxia exposure (Martinez et al. 2006), with liver showing increases and muscle showing decreases in glycolytic enzyme activity. Analysis was performed in normoxia acclimated sculpins (precluding measurements of lactate) in order to assess whether standing variation in substrate levels and maximal enzyme activities were correlated with independent measures of hypoxia tolerance (Pcrit and time to loss of equilibrium analysis; LOE50). Using PIC analysis, hypoxia tolerance was regressed against substrate concentrations and maximal enzyme activity to ascertain if environmental hypoxia has been a selective pressure in defining the ability to generate ATP under low O2 conditions. 2.3 Materials and Methods 2.3.1 Experimental animals   The sculpins used in this study were collected between June and August of 2006, 2007 and 2008. Sculpins collected in 2006 and 2007 were terminally sampled (Mandic et al. 2009), and muscle, liver and brain tissue from those animals were used in this study for analysis of metabolites and maximal enzyme activities. Sculpins collected in 2008 were used for the determination of LOE50 and additional Pcrit analysis on two species of sculpin, Myoxocephalus polyacanthocephalus and Hemilepidotus hemilepidotus. Marine sculpins, including Oligocottus maculosus, Oligocottus snyderi, Clinocottus globiceps, Enophrys bison, Artedius fenestralis, Artedius lateralis, Leptocottus armatus, Blepsias cirrhosus, Scorpaenichthys marmoratus, Myoxocephalus scorpius, M. polyacanthocephalus and H. hemilepidotus  17 were caught using handheld nets or seines during the lowest tidal cycle at Ross Islets	  (48º52.4′N;	  125º9.7′W)	  and	  Wizard’s	  Rock	  (48º51.5′N;	  125º9.4′W),	  near	  the	  Bamfield Marine Sciences Centre (BMSC), Bamfield, British Columbia, Canada and held in flow-through seawater (12°C, 33 ppt salinity). The freshwater sculpin (Cottus asper)	  was	  caught	  using	  baited	  minnow	  traps	  in	  Pachena	  Lake	  (48°50’11”	  N;	  125°01’44”	  W)	  during	  the	  same	  time	  periods	  and	  were	  housed	  in	  freshwater	  re-circulating system held at 12°C. All sculpins were allowed to recover from capture for at least one week before experimentation. In our experiments, variable sample sizes across the different species of sculpin exist as a consequence of year-to-year variation in our ability to collect sufficient numbers of individuals of species that are found at low abundance in the wild. Fish were fed daily with mussels, bloodworms and frozen fish bait, except 24 hours prior to experimental trials. All experiments were conducted according to guidelines set out by the Canadian Council for Animal Care and approved institutional protocols. 2.3.2 Hypoxia tolerance experiments Hypoxia tolerance is a complex trait and warrants a comparison of multiple methods of measurement. We have employed two approaches to assess hypoxia tolerance in sculpins, including the analysis of the time for 50% of a group of sculpins to show loss of equilibrium (LOE50) and Pcrit (water PO2 at which an animal no longer sustains routine O2 consumption and the rate declines as PO2 decreases in the water). For the majority of the species used in this study, Pcrit was taken directly from Mandic et al. (2009). The present study included two new species, M. polyacanthocephalus and H. hemilepidotus and Pcrit was determined using the same methods as described in Mandic et al. (2009).  2.3.2.1 Time to loss of equilibrium (LOE50)  To determine LOE50, individuals from two randomly selected species were placed into separate 40 L aquaria and allowed to recover overnight under flow-through conditions. A small submersible pump was placed in each aquarium to ensure adequate mixing and an Oxyguard Handy MKIII O2 probe was suspended in  18 each aquarium to monitor water PO2. The aquaria were partially submerged in a wet table with flow-through water to keep the temperature at 12°C.  At the end of the recovery period, water flow was stopped and air bubbling into the tanks was switched to N2 to induce hypoxia. The rate of water PO2 decrease was constant among all the trials and within 30 min water PO2 stabilized at 6.4±0.1 torr (n = 24). Water PO2 was monitored every 10 minutes and adjusted by bubbling either N2 or air into the aquaria as required. Fish behavior was also monitored every 10 min and the time an individual showed loss of equilibrium and unresponsiveness to prodding was recorded. Once a fish reached these endpoints, it was removed from the tank and immediately euthanized in an overdose of benzocaine (125 mg mL-1, Sigma-Aldrich, USA).   For species in which individuals were collected in abundance, multiple trials were conducted with eight individual fish per trial. Four trials were carried out for O. maculosus and L. armatus, three trials for C. globiceps, B. cirrhosis and C. asper, and two trials for A. fenestralis. For A. lateralis, E. bison, H. hemilepidotus, M. polyacanthocephalus and O. snyderi, a single trial of six individuals was conducted as a consequence of a much lower capture success from the wild.       2.3.3 Analytical procedures 2.3.3.1 Tissue collection and metabolite assays  Resting, normoxic tissue samples of brain, muscle and liver were obtained as described in Mandic et al. (2009). Briefly, individual fish were isolated overnight in sampling chambers which were submerged in well-aerated 580 L tanks. A sampling chamber was removed from a tank at which time the fish was immediately exposed to an overdose of benzocaine (125 mg mL-1, Sigma-Aldrich, USA) and within one minute the fish lost equilibrium and was quickly sampled. Liver, brain, and muscle were rapidly dissected from the fish, frozen in liquid N2, and stored at -80°C for up to 8 months until analysis. Sampling of each tissue, from dissection to freezing, took less than 10 sec, so all tissues were sampled within 30 sec of the loss of equilibrium.   Aliquots of frozen liver, brain, and muscle were weighed into 2 mL tubes and immediately sonicated on ice in 6 to 10 volumes of 8% perchloric acid using three, 6  19 sec bursts of a Kontes Micro Ultrasonic Cell Disrupter (Kontes, Vineland, New Jersey) set on its highest setting. An aliquot of the homogenate was taken for later glycogen digestion and frozen at -80°C. The remaining homogenate was centrifuged at 20,000 g for 5 minutes at 4°C, and the supernatant neutralized with 3M K2CO3. Neutralized extracts were centrifuged to remove the precipitates and frozen at -80°C until later analysis.   Neutralized extracts were thawed on ice and assayed for ATP, CrP and glucose according to protocols outlined in Bergmeyer (1983). The homogenate set aside for glycogen analysis was thawed, partially neutralized with 3M K2CO3, buffered in 0.45M sodium acetate (pH 4.8) and digested to glucose using amyloglucosidase (Bergmeyer 1983). The homogenate was then neutralized with 3M K2CO3 and assayed for glucose according to the protocols outlined in Bergmeyer (1983). Total glycogen content was calculated by subtracting free glucose measured in neutralized extracts from glucose obtained in digested samples. Glycogen was expressed as µmol glycosyl units/g wet weight. 2.3.3.2 Enzyme activity assays Aliquots of frozen liver, brain, and muscle were quickly weighed to prevent the tissues from thawing and placed in 10 to 20 volumes of ice-cold homogenizer buffer (50 mM HEPES, 5 mM dipotassium ethylene diaminetetraacetic acid, 0.1% Triton X-100, pH 7.4) and immediately sonicated on ice using three, 6 sec bursts set on the highest speed setting. The homogenate was centrifuged at 10,000 g for 2 minutes at 4qC and the resulting supernatant was divided in separate aliquots for analysis of lactate dehydrogenase (LDH), pyruvate kinase (PK), creatine phosphokinase (CPK) and citrate synthase (CS). The high-speed centrifugation may have resulted in some loss of activity due to enzymes bound to subcellular structures, but since all our samples were treated identically, this allows for a valid relative comparison of enzyme activity among the different species of sculpin.   Maximal enzyme activities were determined spectrophotometrically (VersaMax, Molecular Devices, Sunnyvale, California) by measuring the accumulation or disappearance of NADH at 340 nm (LDH, PK and CPK) and the  20 appearance of 5-thio-2-nitrobenzoic acid (TNB) as a result of the reaction of free CoA	  with	  5,5’-dithiobis(2-nitrobenzoic acid) (DTNB) at 412 nm (CS) at 25qC over a 10 minute incubation period. The assay conditions were as follows (in mM): LDH (50 Tris, pH 7.4, 2.5 pyruvate, 0.6 NADH), PK (50 Tris, pH 7.4, 0.5 NADH, 5 ADP, 0.01 fructose-1,6-biphosphate, 100 KCl, 10 MgCl2, 5 phosphoenolpyruvate, 50 U/mL LDH), CPK (50 Tris, pH 7.4, 20 creatine phosphate, 1.5 ADP, 12 AMP, 1.5 NAD, 20 D-glucose, 6.5 dithiothreitol, 25 MgCl2, 2 U/mL hexokinase and 1.5 U/mL glucose-6-phosphate dehydrogenase), and CS (50 Tris, pH 8.0, 0.5 oxaloacetate, 0.3 acetyl-CoA, 0.15 5,5-dithiobis-2-nitrobenzoic acid). Substrates such as pyruvate (LDH), phosphoenolpyruvate (PK), creatine phosphate (CPK) and oxaloacetate (CS) were at saturating levels. Kinase and phosphatase inhibitors were not used in this study to avoid inadvertent modification to two of the kinase enzymes under study, PK and CPK. Kinases and phosphatases are known to modify PK by increasing the Michaelis-Menten (Km) of the enzyme with increasing phosphorylation and to avoid the complicating factors of phosphorylation, saturating levels of substrate PEP were used to measure maximal enzyme activity (Riou et al. 1976). There is also some evidence that the activity of CPK may be regulated by phosphorylation and de-phosphorylation, but the extent of regulation appears small (Lin et al. 2009). Background reaction rates were determined by omitting the following substrates from the control reaction rates: pyruvate (LDH), phosphoenolpyruvate (PK), creatine phosphate (CPK) and oxaloacetate (CS). We used empirically determined extinction coefficients to quantify maximal activity for each enzyme. Total soluble protein was determined in each homogenate using the Bradford protein assay [Sigma-Aldrich (Bradford 1976)]. Maximal enzyme activities were normalized to tissue weight and total soluble protein.  Functional properties of an enzyme can be modified through binding with subcellular components such as membranes, mitochondria and F-actin (Brooks and Storey 1995), and hypoxia has been shown to alter the ratio of free vs. bound enzymes in muscle, liver and brain (Lushchak et al. 1998). To ascertain if percent of bound enzymes differs among species of sculpin we tested the activity of free and  21 bound LDH in muscle, liver and brain of 3 species of sculpin that differ in hypoxia tolerance. Tissue samples were homogenized as described above and the homogenate was spun in a centrifuge at 10,000 g for 2 minutes at 4qC. The supernatant was removed and set aside to test the maximal activity of free LDH, while the pellet was re-suspended in 1 mL of 50 mM Tris buffer and centrifuged. The supernatant was drawn off and the pellet was re-suspended	  in	  200	  μL	  of	  50	  mM	  Tris buffer, frozen at -80°C, thawed and assayed for maximal activity of bound LDH. In the liver there was no detectable bound LDH activity, while in the brain only 1.6% was bound for all 3 species (data not shown). In the muscle, there was 15.9±0.8%, 20.1±1.9% and 22.0±2.4% bound enzyme activity in O. maculosus, M. scorpius and C. asper respectively, although the differences among the species was not significant (ANOVA, P=0.095).  2.3.4 Statistical analysis Data are presented as means r SEM. The Kaplar-Meier survival curve analysis in SigmaStat 3.0 was used to determine the LOE50 median value for each sculpin species. There was no significant difference in LOE50 between trials within a species and consequently data from individual trials were combined for each species to establish the LOE50. To compare maximal enzyme activity of free vs. bound enzymes among species we used a one-way ANOVA (SigmaStat 3.0). Regressions were determined using the least squares regression analysis and correlations were determined using Pearson’s	  correlation	  analysis	  (SigmaStat 3.0) as well as the phylogenetically independent contrast (PIC) analysis using PAUP (Midford et al. 2005) in Mesquite (Maddison and Maddison 2004) in order to remove the effect of phylogeny from the regressions and correlations. For PIC analysis, phylogeny based on cytochrome b sequences originally published in Mandic et al. (2009) was modified to include H. hemilepidotus (accession no. EF521369) and M. polyacanthocephalus (accession no. AB114909; Fig. 2.1). Analysis used to create the trees follows the steps outlined in Mandic et al. 2009. Briefly, a maximum-likelihood tree was constructed with bootstrap analysis of 1000 pseudoreplicates in PAUP (v. 4, Sinauer Sunderland, United States) and the  22 maximum-likelihood tree was imported into Mesquite for PIC analysis. In addition, Mr. Bayes v. 3.0 was used to generate a consensus tree from 10 000 Bayesian trees (Huelsenbeck and Ronquist 2003). The maximum-likelihood and the Bayesian consensus tree show similar phylogenetic relationships among species, therefore maximum-likelihood tree is presented with both bootstrap and posterior probability values for each clade (Fig. 2.1). In order to determine if variation in tree topology and branch length has an effect on PIC analysis (Martins and Housworth 2002) we used both the maximum-likelihood tree and the Bayesian consensus tree for all analysis. Analysis of PIC using the Bayesian consensus tree are not discussed unless a difference in significance of a regression occurred as a result of the two types of trees.   All the comparisons conducted in this study were planned prior to statistical analysis, therefore accounting for multiple comparisons of the results using techniques such as sequential Bonferroni analysis and false discovery rate are not reported. The goal of this study was to explore the data maximally and to minimize the possibility of introducing a Type I statistical error. Consequently, there may be an inflated Type I error associated with our analysis and as such we caution the reader that further analysis and explicit testing of the significant results reported herein should be performed. Statistical significance was assumed at P<0.05. 2.4 Results 2.4.1 Phylogeny  In the present study, cytochrome b sequences from 15 species of sculpin were used to construct a well-resolved phylogeny (Fig. 2.1) and the relationships among the species are in general agreement with published molecular phylogenies (e.g. Ramon and Knope 2008, Mandic et al. 2009). Both the maximum-likelihood tree presented in Fig. 2.1 and Bayesian consensus tree are used in PIC analysis of the following correlations and regressions between hypoxia tolerance and enzyme and metabolites of energy producing pathways in sculpins.  23 2.4.2 Hypoxia tolerance  Time to loss of equilibrium at 6.4±0.1 torr ranged between 25 and 538 min in our study species. Blepsias cirrhosus was the most hypoxia-intolerant species with the lowest LOE50 at 25 min and O. maculosus was the most hypoxia-tolerant species with the highest LOE50 at 538 min. There was a negative relationship between LOE50 and Pcrit, where species with low Pcrit had correspondingly high LOE50 values. This relationship	  was	  significant	  when	  analyzed	  with	  Pearson’s	  correlation	  analysis	  (dashed line on Fig. 2.2); however, when corrected for the phylogenetic relationship among species using PIC, the correlation was no longer significant (solid line on Fig. 2.2). The critical O2 tensions used in this analysis were previously published (Mandic et al. 2009) with the exception of M. polyacanthocephalus (Pcrit 28.2r1.3 torr; n = 6) and H. hemilepidotus (Pcrit 38.5r1.5 torr; n= 4).   2.4.3 Scaling  Allometic scaling is well known to affect metabolic rate and other physiological and biochemical processes (e.g. Somero and Childress 1980; Darveau et al. 2005a; Darveau et al. 2005b). In our sculpins, there was significant variation in body weight (Table 2.1), which could influence the measures of hypoxia tolerance and maximal enzyme activities. In order to determine if the variation in body weight impacted LOE50, we regressed LOE50 against body weight across species and no significant relationship was observed. Furthermore, we previously demonstrated that variation in body weight was not related to Pcrit (Mandic et al. 2009). When tissue maximal enzymes were regressed against body weight there were no significant relationships, with the exception of muscle PK activity expressed as total soluble protein, which increased with increasing body weight (0.15x-0.58, r2=0.42, P=0.022). However, when Pcrit or LOE50 were regressed against residual PK activity the relationships were not significant (Pcrit: y=17.31x+36.65, r2=0.02, P=0.659; LOE50: -523.43x+278.58, r2=0.07, P=0.450). Overall, our analysis suggests that the variation in body weight seen among the sculpins used in the present study does not affect our determination of LOE50, Pcrit  or maximal enzyme activity.  24 2.4.4 Brain  Both Pcrit (Fig. 2.3A) and LOE50 (4.31x-332.08, r2=0.66, P=0.004, least squares regression; 3.21x, r2=0.49, P=0.025, PIC regression) regressed to brain LDH activity when standardized to tissue mass and analyzed using least squares and PIC regressions. Hypoxia-tolerant sculpins (low Pcrit, high LOE50) had higher brain LDH activity compared with the hypoxia-intolerant species. The same trend was seen between Pcrit and activities of PK, CPK, and CS in the brain (expressed relative to tissue mass), with all three enzymes showing higher activities in species of sculpin with lower Pcrit. The regressions were significant when analyzed with least squares and PIC regression analyses (Fig. 2.3B,D). Using least squares regression analysis, there was a positive regression between brain PK (7.65x-393.35, r2=0.41, P=0.046, least squares regression) and CPK (0.55x-217.48, r2=0.45, P=0.035, least squares regression) activities and LOE50 but not when phylogeny was taken into account. There was no relationship between brain CS activity and LOE50.  When enzyme activities were expressed relative to total soluble protein (Appendix A), only brain LDH activity showed a relationship to Pcrit (-34.98x+74.69, r2=0.36, P=0.041) and LOE50 (730.78x-524.09, r2=0.69, P=0.003) using least squares regression analysis. When corrected for the phylogenetic relationships among species, LOE50 (620.00x, r2=0.65, P=0.005) regressed to brain LDH activity but not Pcrit although the P-value approached significance (-35.20x, r2=0.31, P=0.058). Brain PK, CPK, and CS standardized to total soluble protein did not show a relationship with Pcrit. There was a negative relationship between total soluble protein in the brain and Pcrit (Fig. 2.4A), but not LOE50.  There was variation in brain ATP and CrP levels among the species of sculpin in normoxia (Table 2.3), but there was no relationship between these metabolite levels and Pcrit or LOE50. Due to a limitation in tissue size, we were unable to measure glycogen levels in the brain.  2.4.5 Liver Species of sculpin exhibited a range of LDH, PK, CPK and CS activities in the liver (Table 2.2; Appendix A); however, there was no relationship between Pcrit or  25 LOE50 and the activity of any enzyme expressed relative to tissue mass or total soluble protein. There was no relationship between soluble liver protein and LOE50 or Pcrit using least squares regression analysis. The phylogenetically corrected correlation between soluble liver protein and Pcrit was significant (Fig. 2.4B), suggesting that species of sculpin with higher Pcrit also exhibited higher levels of total soluble protein in their liver. However when the data were analyzed with PIC using the Bayesian consensus tree instead of the maximum likelihood tree, there was no relationship between soluble liver protein and Pcrit. This inconsistency in our PIC correlations is likely attributed to the differences in branch lengths between the two types of trees, indicating that the correlation may be an artifact of a specific tree topology.   Liver ATP, CrP and glycogen varied among the different species of sculpin (Table 2.3), but despite this variation, there was no relationship between the liver metabolites and Pcrit or LOE50.    2.4.6 Muscle Activities of LDH, PK, CPK and CS varied among the sculpins examined (Table 2.2, Appendix A). Despite this variability, there was no relationship between our measures of hypoxia tolerance (Pcrit or LOE50) and muscle enzyme activities either when expressed relative to tissue mass or when expressed relative to total soluble protein. Soluble muscle protein did not correlate with Pcrit (Fig. 2.4C) or with LOE50.  Similar to liver and brain, there were differences in the levels of muscle ATP, CrP, and glycogen among species of sculpin (Table 2.3). There was a negative relationship between muscle CrP and Pcrit, where species of sculpin with low Pcrit exhibited high levels of muscle CrP (-1.19x+63.937, r2=0.33 P=0.050, least squares regression; -1.47x, r2=0.46, P=0.016, PIC regression). When LOE50 was regressed against muscle CrP levels, there was no significant pattern among sculpins. Neither Pcrit nor LOE50 regressed against muscle ATP and glycogen.  26 2.5 Discussion 2.5.1 Hypoxia tolerance  The comparative approach in tandem with phylogenetic analysis is a powerful tool in providing evidence of selection on phenotypic traits once it is determined that there is variation in tolerance among species as a consequence of an environmental pressure. Environmental hypoxia may be an important agent of selection among sculpins, as different species are exposed to varied frequency, duration and magnitude of low levels of O2. Loss of equilibrium and unresponsiveness to stimuli pose severe impediments to survival in the wild, therefore an important indicator of tolerance to hypoxia is the length of time that is required for individuals of a species to reach these endpoints, which is expressed as LOE50. Sculpins exhibited a range of LOE50 values (Fig. 2.2), indicating that variation in hypoxia tolerance exists in this group of closely related species.  The ability of an animal to acquire O2 from its environment is considered to be critical to hypoxia survival. Indeed, among the species of sculpin investigated herein, we observed a strong relationship between LOE50 and Pcrit, supporting the notion that Pcrit is an indirect measure of hypoxia tolerance as previously suggested in the literature (Hagerman 1998; Chapman et al. 2002; Mandic et al. 2009; Seibel 2011; Speers-Roesch et al. 2012). In sculpins, there was a correlation between LOE50 and Pcrit, where species with high LOE50 values exhibited correspondingly low Pcrit values (Fig. 2.2). However, the r2 value for the correlation indicated that only 46% of the variability in Pcrit was associated with the variability in LOE50, and the correlation became non-significant with the application of PIC. This would suggest that beyond O2 extraction, other characteristics, such as metabolic rate suppression and capacity to generate energy via substrate-level phosphorylation, may play important roles in contributing to the variation in hypoxia tolerance among sculpins.       As O2 levels drop below the point where oxidative phosphorylation can generate adequate ATP to sustain metabolic functions (ie. below a species Pcrit) there is an increased reliance on glycolysis and CrP hydrolysis for the production of  27 ATP (DiAngelo and Heath 1987; Boutilier et al. 1988; Van Waarde et al. 1990). This switch in energy production during low O2 exposure occurs in both hypoxia-tolerant and hypoxia-intolerant species. Therefore the distinguishing characteristic between hypoxia-tolerant and -intolerant species is not the recruitment of anaerobic pathways, but instead it is thought to be the ability to generate sufficient flux through these pathways during exposure to hypoxia. As such, we predicted that the hypoxia-tolerant species would have high tissue levels of substrates to support these pathways or that the hypoxia-tolerant species would possibly exhibit higher maximal enzyme activities as suggested in previous studies (e.g. Martinez et al. 2011).  2.5.2 Brain enzymes  The brain is an energetically demanding organ and as such is highly susceptible to decreases in O2 (Lutz 1992; Erecinska and Silver 2001; Nilsson 2001). To prevent rapid neural death during exposure to hypoxia, tolerant species increase their reliance on glycolysis and increase blood flow to the brain in order to maintain a constant supply of ATP (Storey 1987; Nilsson et al. 1994; Johansson et al. 1995). Our data are consistent with this idea as higher PK and LDH activities expressed relative to tissue mass suggest that maximal capacity for glycolysis is greater in the more hypoxia-tolerant sculpins, indicating that these species may be capable of producing more ATP during a sudden drop in environmental O2. Although the brain primarily relies on glycogen and glucose for energy supply (reviewed in Bolanos et al. 2009), CrP hydrolysis also occurs to buffer ATP consumption (Lutz et al. 2003) and high CPK activity expressed relative to tissue mass, as seen in sculpins with low Pcrit, could confer a comparable benefit of higher ATP production during hypoxia. Similar to the trend observed among the enzymes involved in the anaerobic pathways, there was a significant regression between CS, an enzyme involved in the tricarboxylic acid cycle, and hypoxia tolerance as measured by Pcrit. Sculpins with a low Pcrit had greater CS activity expressed relative to tissue mass than the species with a high Pcrit and this could be advantageous as it is likely that glucose oxidation still occurs in the brain during hypoxia due to increased blood flow, as seen in the  28 crucian carp (Nilsson et al. 1994). Increased blood flow to the brain would allow the delivery of any O2 that the animal is able to extract from the environment as well as glucose mobilized from sources such as the liver.  Up-regulating the amount of enzymes and hence enzymatic activities can take time, therefore it could be beneficial to maintain high levels of these enzymes even under normoxic conditions. This is particularly relevant in sculpins, where they can experience sudden bouts of short-term hypoxia in their natural environment, such as for sculpins isolated in tidepools at night. Indeed, in the present study we show in sculpins that hypoxia tolerance, quantified by measures of Pcrit, regressed against LDH, PK, CPK and CS activities expressed relative to tissue mass (Fig. 2.3). The relationships remained significant when phylogenetic independent contrast analysis was conducted, indicating that the variation in enzyme activities among sculpin species may be the result of adaptation to hypoxia. However, when LOE50 was used as an index of hypoxia tolerance in sculpins, only brain LDH exhibited a phylogenetically independent relationship to hypoxia tolerance. There was no relationship between brain CS and LOE50, and while LOE50 significantly regressed against brain PK and CPK, the relationships were not significant when phylogeny was taking into account, possibly attributing the variation in these enzymes to phylogenetic distance among species.          Our data suggests that Pcrit rather than LOE50 regressed more strongly against brain maximal enzyme activities expressed relative to tissue mass in sculpins. However, the discrepancy between Pcrit and LOE50 in the regressions with brain enzyme activities may be a result of decreased statistical power in the LOE50 regressions. Maximal enzyme values for M. scorpius and S. marmoratus could not be included in the regressions with LOE50 since LOE50 values are not known for these two species, potentially leading to a loss in the ability to detect significant relationships between maximal enzyme activities and LOE50.  2.5.3 Standardization of enzyme activity Standardization of enzyme activity to a common denominator may be achieved in multiple ways. For example, enzyme activity can be expressed relative  29 to tissue mass, total soluble protein, or in case of mitochondrial enzymes, relative to cytochrome oxidase (COX) or many other denominators depending on the specific goal of the analysis (Leary et al. 2003). In the current study, enzyme activities discussed thus far have been expressed relative to tissue mass. When total soluble protein was used as a denominator, Pcrit and LOE50 only regressed significantly against brain LDH activity. Interestingly, there was a phylogenetically independent regression between Pcrit and total soluble protein, where sculpins with low Pcrit displayed high soluble brain protein levels as opposed to the lower levels of soluble brain protein exhibited by species with high Pcrit (Fig. 2.4). Although the underlying reason behind this trend is unclear, one possible explanation is that the high soluble protein levels in the brain increase protein stability, which may be important in animals that are prone to environmental perturbations such as hypoxia. Proteins may have low intrinsic stabilities as a trade-off for increased degree of flexibility to undergo rapid catalytic conformational changes during periods of environmental stress and are therefore greatly dependent on high cellular protein concentration to maintain functional protein structure (Hochachka and Somero 2002). Alternatively, reversible brain swelling has been known to occur during hypoxia in some fishes such as the common carp (Van der Linden et al. 2001), and it is possible that high levels of soluble proteins in hypoxia-tolerant species of sculpin may be a mechanism to negate the negative effects of large intracellular volume increases on protein function. However, there was no relationship between brain soluble proteins and LOE50 and further investigation is necessary to confirm if there are high levels of soluble proteins in the brains of tolerant species of sculpin and if so, to test whether these high levels of soluble proteins play a role in maintaining protein function during exposure to hypoxia.   2.5.4 Muscle and liver enzymes   There were no significant regressions among different enzyme activities and Pcrit or LOE50 in either the muscle or the liver. During hypoxia exposure, energy demand in the muscle can decrease as a result of reduction in spontaneous locomotor activity (Nilsson et al. 1993) and a rapid decrease in protein synthesis  30 rates (Smith et al. 1996). Therefore, unlike the brain, energy demand can be significantly reduced in the muscle negating the need to up-regulate activities of enzymes associated with anaerobic pathways. In support of this idea, muscle glycolytic enzyme activities in Fundulus grandis were seen to significantly decrease (Martinez et al. 2006) and muscle CPK mRNA levels were down-regulated in Gillichthys mirabilis (Gracey et al. 2001) in response to low O2 exposures. Lack of a relationship between hypoxia tolerance and enzyme activities in the muscle could also be attributed to other selection pressures acting on glycolysis and CrP hydrolysis such as burst swimming for predator avoidance or prey capture. In the liver, many glycolytic enzymes are also common to gluconeogenesis, a process where lactate transported from other areas of the body is re-synthesized into glucose (Knox et al. 1980). Many of the enzymes are readily reversible and therefore, a lack of regression to hypoxia tolerance may reflect the dual function of these enzymes.      2.5.5 Metabolites  Sustaining flux during hypoxia requires not only elevated enzyme capacities but sufficient substrate reserves as well. The remarkable hypoxia/anoxia tolerance of goldfish and crucian carp have been partly attributed to large reserves of liver glycogen (Van den Thillart et al. 1980; Hyvarinen et al. 1985) and in general, level of glycogen deposition is thought to be affected by hypoxia (Van den Thillart and Van Raaij 1995). Therefore it is plausible that higher glycogen or CrP levels may occur in the more hypoxia-tolerant species. The results for muscle CrP appear to support this prediction, as there was a phylogenetically independent regression between hypoxia tolerance and muscle CrP, although this was only the case when Pcrit was used as a measure of hypoxia tolerance and not LOE50. Although elevated levels of muscle CrP in hypoxia-tolerant sculpin species may be beneficial for maintaining ATP during exposure to hypoxia, without further measurements of creatine levels and pH buffering capacity it is difficult to ascertain the mechanism contributing to the variation of CrP levels among the sculpin species. Unlike the muscle, there was no relationship between CrP and hypoxia tolerance in the brain or the liver,  31 suggesting that this mode of energy production is not likely sustained long in these tissues during a bout of hypoxia. Hydrolysis of CrP may primarily be utilized in the initial stages of hypoxia until exogenous glucose can be delivered to the tissue or tissue glycogen mobilized.  There was high variation in liver glycogen among the differentially tolerant species of sculpin. However, neither Pcrit or LOE50 regressed against liver glycogen. The lack of trend between hypoxia tolerance and liver and muscle glycogen in sculpins is contrary to the prediction that species more frequently exposed to environmental hypoxia store greater amounts of glycogen in their tissues than species that rarely experience low O2. This may be attributed to the highly labile glycogen stores that can be influenced by many factors, including temperature, diet, and burst swimming activity (Van den Thillart and Raaij 1995). Although the amount of substrate available may be an important factor for hypoxic survival, there was no clear relationship in sculpins that the more hypoxia-tolerant species exhibit higher levels of glycogen or CrP than the more intolerant species. 2.5.6 Conclusions  Increased reliance on substrate-level phosphorylation for ATP production is a critical component of hypoxia survival. In the brain, sculpins with low Pcrit and high LOE50 expressed high LDH activities regardless of whether the maximal enzyme activity was expressed relative to tissue mass or total soluble protein. As these relationships remained significant when phylogeny was taken into account, one possibility is that selection is acting on maximum capacity for flux through anaerobic pathways for ATP production during hypoxia in the brain. Although these significant correlations in the brain are intriguing, they must be viewed cautiously because of the inflated Type II error associated with the multiple comparisons. The results observed in the brain are in contrast with the results seen in liver and the white muscle, where there were no relationships between hypoxia tolerance and maximal enzyme activities. A significant, phylogenetically independent relationship was found between Pcrit and total soluble protein in the brain, although the functional significance of this is unknown. Further investigation is necessary to  32 ascertain if high total soluble protein in the brain of hypoxia-tolerant species of sculpin is an important factor to survival in hypoxia.     33 Figure 2.1 Phylogenetic relationship among 15 species of sculpin based on a maximum-likelihood tree using cytochrome b sequences.  Satyrichthys amiscus is included as an out-group. A similar tree has been previously published (Mandic et al. 2009) but did not include Hemilepidotus hemilepidotus and Myoxocephalus polyacanthocephalus. Numbers above lines show percent of 1000 maximum-likelihood bootstrap replicates with values shown for groups with more than 50% support. Italicized numbers below lines show posterior probability from Bayesian Markov Chain Monte Carlo analysis.    34  Figure 2.2  Relationship between Pcrit and LOE50.  Pearson’s	  correlation	  (dashed	  line;	  -0.04x+46.57, r2=0.46, P=0.023) and phylogenetically independent contrast (PIC) correlation (solid line; -0.03x, r2=0.31, P=0.073) are plotted on species data that are presented as mean r SE. Dotted lines represent 95% prediction lines for the PIC correlation. Numbers represent different species: 1) Oligocottus maculosus, (2) Clinocottus globiceps, (3) Oligocottus snyderi, (4) Myoxocephalus polyacanthocephalus, (5) Enophrys bison, (6) Artedius fenestralis, (7) Artedius lateralis, (8) Leptocottus armatus, (9) Blepsias cirrhosus, (10) Cottus asper, and (11) Hemilepidotus hemilepidotus.   Figure 2.3 Relationships between Pcrit and maximal enzyme activities of lactate dehydrogenase (LDH), pyruvate kinase (PK), creatine phosphokinase (CPK) and citrate synthase (CS) in the brain of 12 species of sculpin.  Pcrit and brain LDH (A; -0.29x+76.55, r2=0.75, P<0.001, least squares regression; -0.30x, r2=0.80, P<0.001, phylogenetically independent contrast (PIC) regression), Pcrit and brain PK (B; -0.44x+74.58, r2=0.34, P=0.045, least squares regression; -0.54x, r2=0.45, P=0.016, PIC regression), Pcrit and brain CPK (C; -0.04x+70.32, r2=0.57, P=0.004, least squares regression; -0.04x, r2=0.60, P=0.003, PIC) and Pcrit and brain CS (D; -8.45x+71.22, r2=0.36, P=0.039, least squares regression; -8.67x,  35 r2=0.62, P=0.002, PIC). Least squares regression (dashed lines) and PIC regressions (solid lines) are plotted on species data that are presented as mean r SE. Dotted lines represent 95% prediction lines for the PIC regressions. Numbers represent different species: (1) Oligocottus maculosus, (2) Clinocottus globiceps, (3) Oligocottus snyderi, (4) Myoxocephalus polyacanthocephalus, (5) Enophrys bison, (6) Artedius fenestralis, (7) Artedius lateralis, (8) Leptocottus armatus, (9) Blepsias cirrhosus, (10) Cottus asper, (11) Scorpaenichthys marmoratus, and (12) Myoxocephalus scorpius.  Figure 2.4  Relationship between Pcrit and soluble protein concentration in brain, liver, and muscle. Brain (A; -0.29x+75.84, r2=0.52, P=0.008, least squares regression; -0.29x, r2=0.71, P<0.001 phylogenetically independent contrast (PIC) regression), liver (B; 0.12x+19.23, r2=0.09, P=0.337, least squares regression; 0.29x, r2=0.36, P=0.040, PIC regression), and muscle (C; -0.23x+53.99, r2=0.12, P=0.262, least squares Pcrit (torr)Enzyme Activity	  (µmol/min/g	  wet tissue)A806040200806040200500 1300600 700 800 900 1000 1100 1200CBDLDHCPKPKCS2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 6.080 100 200180120 140 160 60 80 12010070 90 1101109 11 8234567121109 11 823456 712110911 82 3456 712110911 823456712 36 regression; -0.29x, r2=0.18, P=0.173, PIC regression). See figure 2.3 captions for more detail.  Tissue Protein (mg/g wet tissue)Pcrit (torr)40 50 1009060 70 80 110806040200-20100806040200-20100806040200-2010080 100 200180120 140 160 22080 100 200180120 140 160 220ABCBrainLiverMuscle1109 1182 3456 7121109118234 567121109118 23456712 37 Table 2.1  Body weights of species of sculpin used in the LOE50 experiment and in analytical procedures.  Body weights are expressed as grams. The first data column refers to body weights of fish used in the LOE50 experiment and the second data column refers to body weights of fish used to assess maximal enzyme activities and metabolites. Data are means ± SE.  Species Body weights LOE50 experiment Analytical experiment O. maculosus 2.92±0.22 4.50±0.50 C. globiceps 1.21±0.06 1.59±0.20 O. snyderi 3.85±1.30 2.54±0.71 M. polyacanthocephalus 14.42±2.32 10.50±1.30 E. bison 4.28±1.61 10.91±5.29 A. fenestralis 6.52±2.32 16.91±2.14 A. lateralis 14.48±3.58 15.16±1.38 L. armatus 32.26±1.58 48.22±2.07 S. marmoratus N/A 20.84±2.88 B. cirrhosis 7.20±0.54 2.63±0.37 M. scorpius N/A 9.73±2.65 C. asper 30.47±2.10 19.34±1.84 H. hemilepidotus 8.13±1.28 N/A   38 Table 2.2  Tissue enzymes in 12 species of sculpin.  Tissue enzymes are presented in Pmol/min/g wet tissue. Numbers in parentheses represent sample size.  Data are means±SE.   Species Liver  Muscle LDH PK CPK CS  LDH PK  CPK CS O. maculosus 3.06±0.81 (16) 5.62±0.64 (16) 31.65±9.42 (16) 0.75±0.07 (16)  304.60±25.47 (8) 32.08±2.56 (8) 3149.96±215.47 (8) 0.77±0.09 (8) C. globiceps 6.01±1.73 (10) 4.27±0.99 (10) 66.53±17.85 (10) 1.34±0.24 (10)  311.54±26.50 (8) 31.91±2.17 (8) 2802.02±144.76 (8) 0.89±0.05 (8) O. snyderi 4.76±1.02 (8) 3.14±0.62 (8) 76.99±23.10 (8) 1.41±0.41 (8)  244.64±19.67 (8) 24.75±1.89 (8) 3176.79±152.64 (8) 0.65±0.07 (8) M. polyacanthocephalus 1.59±0.11 (8) 4.06±0.71 (8) 39.67±12.61 (8) 0.45±0.03 (8)  404.13±16.67 (8) 43.83±3.68 (8) 2544.41±335.57 (8) 1.00±0.13 (8) E. bison 6.85±1.67 (11) 6.48±1.16 (11) 15.61±4.85 (11) 1.06±0.13 (11)  219.18±45.18 (9) 28.02±6.01 (9) 2020.06±260.30 (9) 1.13±0.11 (9) A. fenestralis 1.64±0.20 (8) 3.68±0.42 (8) 10.46±1.20 (8) 0.64±0.07 (8)  478.64±42.63 (8) 41.22±3.74 (8) 2931.43±286.33 (8) 0.62±0.08 (8) A. lateralis 2.01±0.48 (7) 3.49±0.51 (7) 28.73±5.41 (7) 0.50±0.06 (7)  344.06±36.90 (8) 41.05±3.96 (8) 3442.93±327.40 (8) 0.35±0.08 (8) L. armatus 3.98±0.50 (12) 4.12±0.37 (12) 34.54±2.09 (12) 0.73±0.05 (12)  328.07±35.63 (8) 30.28±3.48 (8) 2453.38±365.28 (8) 0.65±0.07 (8) S. marmoratus 3.38±0.49 (12) 6.06±0.54 (12) 23.13±3.02 (12) 0.97±0.12 (12)  301.53±24.47 (8) 29.65±2.52 (8) 3486.20±234.35 (8) 0.93±0.09 (8) B. cirrhosis 1.81±0.20 (12) 4.86±0.36 (12) 14.26±2.13 (12) 1.31±0.10 (12)  63.90±6.66 (10) 12.42±0.99 (10) 1112.49±206.38 (10) 0.39±0.05 (10) M. scorpius 3.09±0.47 (7) 4.94±0.37 (7) 23.50±1.54 (7) 0.68±0.08 (7)  140.54±39.67 (8) 23.80±3.89 (8) 1872.41±180.71 (8) 0.90±0.09 (8) C. asper 1.94±0.17 (8) 2.55±0.32 (8) 32.05±3.32 (8) 0.99±0.1 (8)  471.73±13.85 (8) 50.24±3.25 (8) 3422.87±151.11 (8) 1.01±0.10 (8)  39  Table 2.3  Tissue metabolites in 12 species of sculpin.  Tissue metabolites are presented in Pmol/g wet tissue. Numbers in parentheses represent sample size. Data are means ± SE.  Brain  Liver  Muscle ATP CrP  ATP CrP Glycogen  ATP CrP Glycogen O. maculosus 0.7±0.1 (10) 1.7±0.4 (10)  1.5±0.2 (8) 3.5±0.6 (8) 250.4±38.2 (8)  2.8±0.6 (7) 23.2±2.6 (7) 6.4±1.2 (7) C. globiceps 0.4±0.1 (8) 0.6±0.2 (8)  0.6±0.2 (6) 0.7±0.1 (6) 80.0±23.5 (6)  1.1±0.3 (8) 25.8±2.2 (8) 8.9±3.3 (8) O. snyderi 0.8±0.3 (2) 1.8±0.4 (2)  0.9±0.2 (6) 0.9±0.3 (6) 229.8±52.5 (6)  1.0±0.4 (8) 24.4±3.4 (8) 11.1±4.6 (8) M. polyacanthocephalus 0.8±0.1 (4) 2.0±0.2 (4)  1.1±0.1 (8) 0.3±0.1 (8) 385.6±43.1 (8)  3.3±0.6 (8) 27.6±3.1 (8) 27.6±4.8 (8) E. bison 1.1±0.1 (7) 4.4±0.5 (7)  0.7±0.1 (9) 1.6±0.4 (9) 159.3±27.1 (9)  2.1±0.6 (9) 29.5±3.9 (9) 7.9±4.8 (9) A. fenestralis 1.0±0.1 (8) 4.9±0.7 (8)  1.9±0.3 (8) 3.1±1.2 (8) 352.4±57.4 (8)  4.0±0.5 (8) 18.6±1.5 (8) 14.2±6.4 (8) A. lateralis 0.8±0.1 (5) 3.3±1.1 (5)  1.4±0.3 (8) 2.2±0.6 (8) 328.7±64.4 (8)  4.2±0.3 (8) 21.1±3.1 (8) 10.4±1.6 (8) L. armatus 1.2±0.1 (8) 4.1±0.5 (8)  1.0±0.1 (8) 0.6±0.1 (8) 98.2±17.0 (8)  2.9±0.7 (8) 21.7±2.2 (8) 6.2±1.5 (8) S. marmoratus 0.9±0.1 (8) 2.3±0.4 (8)  1.3±0.1 (8) 0.4±0.1 (8) 394.2±34.6 (8)  6.0±0.3 (8) 26.5±1.8 (8) 12.8±2.2 (8) B. cirrhosis 0.6±0.1 (10) 0.8±0.2 (10)  0.8±0.1 (9) 1.3±0.5 (9) 126.3±29.0 (9)  0.7±0.1 (10) 19.7±4.4 (10) 0.3±0.2 (10) M. scorpius 0.8±0.1 (4) 2.1±0.8 (4)  1.1±0.1 (8) 0.4±0.1 (8) 296.2±37.3 (8)  2.2±0.6 (8) 24.5±1.8 (8) 21.1±4.2 (8) C. asper 1.1±0.1 (10) 4.3±0.5 (10)  1.7±0.2 (8) 0.5±0.1 (8) 278.7±74.9 (8)  2.7±0.3 (8) 12.3±3.3 (8) 14.4±4.9 (8)  40 Chapter 3  Divergent Transcriptional Patterns are Related to Differences in Hypoxia Tolerance Between the Intertidal and the Subtidal Sculpins2 3.1 Summary Transcriptionally mediated phenotypic plasticity as a mechanism of modifying traits in response to an environmental challenge remains an important area of study. We compared the transcriptional responses to low oxygen (hypoxia) of the hypoxia-tolerant intertidal fish, the tidepool sculpin (Oligocottus maculosus) with the closely related hypoxia-intolerant subtidal fish, the silverspotted sculpin (Blepsias cirrhosus) to determine if these species use different mechanisms to cope with hypoxia. Individuals from each species were exposed to environmental O2 tensions chosen to yield a similar level of tissue hypoxia and gene transcription was assessed in the liver over time. There was an effect of time in hypoxia, where the greatest transcriptional change in the silverspotted sculpin occurred between 3 to 24 hours in contrast to the tidepool sculpin where the largest transcriptional change occurred between 24 and 72 hours of hypoxia. A number of genes showed similar hypoxia-induced transcription patterns in both species (e.g. genes associated with glycolysis and apoptosis) suggesting they are involved in a conserved hypoxia                                                         2 A version of this chapter has been published: Mandic M., M.L. Ramon, A.Y. Gracey and J.G. Richards. 2014. Divergent transcriptional patterns are related to differences in hypoxia tolerance between the intertidal and the subtidal sculpins. Molecular Ecology 23: 6091-6103  41 response. A large set of genes showed divergent transcriptional patterns in the two species, including fatty acid oxidation and oxidative phosphorylation, suggesting that these biological processes may contribute to explaining variation in hypoxia tolerance in these species. When both species were exposed to a single environmental O2 tension, large transcriptional responses were seen in the hypoxia-intolerant silverspotted sculpin while almost no response was observed in the hypoxia-tolerant tidepool sculpin. Overall, divergent transcription patterns in response to both magnitude and duration of hypoxia provide insights into the processes	  that	  may	  determine	  an	  animal’s	  capacity	  to	  tolerate	  frequent	  bouts	  of	  hypoxia in the wild. 3.2 Introduction The evolution of phenotypic plasticity is an important evolutionary mechanism for adaptation of animals to variable environments (see reviews West-Eberhard 1989, Ghalambor et al. 2007). In many aquatic ecosystems, periodic decreases in O2 tension (hypoxia) are common and create spatial and temporal environmental heterogeneity. As a result, hypoxia is considered to be an important selective pressure in the evolution of aquatic organisms (Nikinmaa 2002). In particular, the marine rocky intertidal zone is susceptible to rapid and pronounced changes in O2 tension due to the predictable influence of the tidal cycle. When tidepools are isolated from the ocean at night, O2 levels drop to near anoxic levels as a consequence of flora and fauna respiration. While the severity and frequency of hypoxia is significantly reduced in the connecting subtidal environment, periodic hypoxia still occurs as a result of the lower mixing of water associated with the protected nearshore environment (Truchot and Duhamel-Jouve 1980, Richards 2011). In contrast, O2 tensions are generally high and stable in the pelagic zone and up to mid-range depths of ~400 m (reviewed in Diaz and Breitburg 2009). This spatial and temporal variation of hypoxic events, along with variation in other abiotic factors, heavily influences species ranges resulting in a strong vertical zonation pattern along the nearshore (Trussell and Etter 2001, Altieri 2006).   42 Sculpins (superfamily Cottoidea) are a group of predominantly benthic fish species that inhabit environments from the intertidal and nearshore to deeper, mid-range water depths and the greatest diversity of sculpin species occur in the Northeastern Pacific (Nelson 2006, Froese and Pauly 2013). Phylogenetic analysis indicates that the direction of evolution of sculpins is from the subtidal to the intertidal environment, suggesting that the ancestral species are found primarily in the O2 stable subtidal and the more derived species are found in the O2 variable intertidal (Ramon and Knope 2008). The distribution of sculpins along the nearshore environment is related to their hypoxia tolerance with the more hypoxia-tolerant species inhabiting the intertidal and the less hypoxia-tolerant species inhabiting deeper subtidal waters. Variation in hypoxia tolerance among sculpins is  explained, at least in part, by variation in physiological traits involved in O2 extraction and biochemical traits involved in energy production (Mandic et al. 2009, Chapter 2). However, the extent to which phenotypic plasticity, versus constitutive differences in phenotype, play a role in hypoxic survival in the heterogeneous nearshore environment is unknown among sculpins that differ in hypoxia tolerance.  Identifying the mechanisms that result in phenotypic plasticity remains an important question in ecology and evolution (Aubin-Horth and Renn 2009). Environmentally induced changes in phenotypic traits may be achieved through multiple mechanisms, including changes in allelic sensitivity and gene transcription (Via et al. 1995, Dewitt et al. 1998), as well translational and post-translation modification. Particular attention has been given to understanding the role of environmentally induced changes in gene transcript levels and the characterization of transcriptomic reaction norms (Schlichting and Pigliucci 1993, Schlichting and Smith 2002, Aubin-Horth and Renn 2009) because changes in gene transcription have been shown to result in phenotypic plasticity in response to environmental change (as reviewed by Schlichting and Pigliucci 1993) and because of the availability of high-throughput methods (e.g. microarrays and RNA-seq; Aubin-Horth and Renn 2009). Multiple studies have examined transcription patterns during exposure to short-term (Gracey et al. 2001, Ju et al. 2007, Everett et al. 2011, Leveelahti et al. 2011), long-term (van der Meer et al. 2005) and developmental  43 hypoxia (Ton et al. 2003) and these studies provide insight into the genes and pathways that are potentially involved in hypoxia-induced phenotypic plasticity. However, despite our general understanding of the effects of hypoxia on gene transcription in fish, no study has attempted to assess how natural variation in hypoxia tolerance among fishes affects the gene transcription patterns that form the basis of phenotypic plasticity. The goal of this study was to compare the hypoxia-induced gene transcription patterns between the hypoxia-tolerant tidepool sculpin (Oligocottus maculosus) and the hypoxia-intolerant silverspotted sculpin (Blepsias cirrhosus) to determine if, and how, differences in transcriptional plasticity are associated with differences in hypoxia tolerance. We have conducted this work to gain insight into the biological processes that may change in response to hypoxia exposure and contribute to defining differences in whole animal hypoxia tolerance and species distribution patterns across the rocky intertidal zone. Since the ability to acquire O2 from the environment differs between the two study species due to differences in gill surface area and hemoglobin-O2 binding affinity (Mandic et al. 2009), we used an experimental treatment that scaled our hypoxia exposures to the species-specific hypoxia tolerance. Therefore, both species experienced the same level of hypoxemia and tissue hypoxia (see Speers-Roesch et al. 2012; this treatment is referred to as the	  “relative	  hypoxia	  exposure”)	  and	  we	  examined	  changes	  in	  transcript	  levels	  for	  up to 72 hours of hypoxia. While this approach may lead to fewer differences in gene transcription between the two species than may occur if both species were exposed to a single environmental O2 tension, any differences in transcriptional patterns are more likely to be involved in defining hypoxia tolerance. In addition, we also exposed the two species of sculpin to a single environmental O2 tension for 24 hours to examine the direct effect of environmental hypoxia, which is an important factor that	  determines	  species	  distribution	  (termed	  the	  “absolute	  environmental hypoxia”).	  In	  both	  hypoxia	  exposures,	  we	  examined	  the	  transcript	  profile	  in	  the	  liver, which is a tissue critical to energy metabolism and as a result is vital in an animal’s	  defense	  against	  hypoxia	  (van	  den	  Thillart	  et	  al.	  1980).	  To	  narrow	  the	  focus of our analyses, we posed the following questions: 1) Is there temporal variation in  44 gene transcript levels and how does it vary between the two species of sculpin? 2) Are there similar transcriptional patterns in genes between the two species?  3) What are the genes or biological processes for which these species exhibit a divergent transcriptional pattern? 4) How does gene transcription differ between these species when exposed to absolute environmental hypoxia, especially in comparison to relative environmental hypoxia of the same duration? 3.3 Materials and Methods 3.3.1 Experimental animals Sculpins were collected in July and August of 2010 near the Bamfield Marine Sciences Centre (BMSC), Bamfield, British Columbia, Canada. Tidepool sculpins (Oligocottus maculosus), a fish species found along the intertidal zone and most densely populated in the upper mid tidepool region (Green 1971), were caught using handheld nets while beach seines were used to capture silverspotted sculpins (Blepsias cirrhosus), a fish species typically found in the kelp and eelgrass beds of the subtidal zone (Dean et al. 2000, Elfird and Konar 2013). Collection occurred during	  the	  lowest	  tidal	  cycle	  at	  Wizard’s	  Rock	  (48°51.5′N;	  125°9.4′W).	  The	  sculpins	  were transported to The University of British Columbia and held in fully aerated recirculating seawater at 12°C and 30 ppt salinity, replaced monthly with natural, sand-filtered seawater from the Vancouver Aquarium. The fish were allowed to acclimate to constant laboratory conditions for two months prior to experimentation. During this time, fish were fed daily with a combination of bloodworms and krill, except 24 hours before an experiment.  3.3.2 Hypoxia time-course  Fish were transferred to a custom-built recirculating system, the Low Oxygen Control and Monitoring Aquatic Research System (Integrated Aqua Systems, Inc.), designed to closely regulate O2 levels and temperature. Individuals from each species were split into four separate, 36 gallon insulated tanks capable of independent control of O2 tensions, with a front-facing window. Fish were allowed to recover for 24 hours under full air saturation, following which access to the air- 45 water interface was restricted by a Plexiglas cover. We exposed the two species of sculpin to hypoxia at O2 tensions that corresponded to 65% of the species-specific Pcrit, which is a measurement of O2 extraction and a proxy for hypoxia tolerance (Pcrit of 25.9 Torr for tidepool sculpin and 44.4 Torr for silverspotted sculpin, Mandic et al. 2009). A previous study of gene transcription in Gillichthys mirabilis (Gracey et al. 2001) found that this level of hypoxia induced hypoxic stress while allowing fish to survive for up to 6 days of hypoxia. Specifically, two tanks of the hypoxia-tolerant tidepool sculpin, were exposed to O2 tensions set at 17 Torr and two tanks of the hypoxia-intolerant	  silverspotted	  sculpin,	  were	  exposed	  to	  29	  Torr	  (“experiment	  1:	  relative	  hypoxia	  exposure”).	  Additionally,	  O2 tensions were set to 23 Torr in the third	  tank	  of	  each	  species	  (“experiment	  2:	  absolute	  environmental	  hypoxia	  exposure”).	  Normoxic	  conditions	  (~158	  Torr)	  were	  maintained	  in	  the	  fourth	  tank	  of	  individuals for both species (Table 3.1).   For experiment 1, liver, muscle, heart, brain and gill from individuals of both species were sampled (only liver was used in the microarray experiments and other tissues were archived) at 3, 8, 24, 48 and 72 hours of relative hypoxia exposure, with two normoxic fish sampled at each hypoxic time point. For experiment 2, the same tissues were sampled at 24 hours. As experiment 1 and 2 were conducted simultaneously we used the same normoxic control samples. Briefly, fish were removed from the tank and immediately euthanized by exposure to an overdose of benzocaine (125 mg mL-1; Sigma-Aldrich). Tissues were dissected, flash frozen in liquid N2, and stored at -80°C until analysis.  3.3.3 Microarray hybridization   For each species, total RNA was isolated from liver of five normoxic and four hypoxic samples per time point, for a total of 29 samples per species, using TRIzol (Invitrogen)	  according	  to	  the	  manufacturer’s	  protocol.	  The	  total	  RNA	  was	  purified	  further using glass-fiber filter columns (Qiagen)	  according	  to	  the	  manufacturer’s	  instructions. Double-stranded cDNA was prepared by reverse-transcribing 1μg of liver RNA in a 20 μl reaction containing 20 pmols of T7-dT15VN primer, 2 μl MMLV HP RT 10X Reaction Buffer, 2 μl 100 mM DTT, 2 μl 50x aminoallyl dNTP mix (25 mM  46 dATP, 25 mM dCTP, 2 5mM dGTP, 15 mM dTTP, 10 mM aminoallyl-dUTP), 20 U of RNaseOut (Invitrogen) and 100 U of MMLV-reverse transcriptase (Epicentre). The RNA and primer mixture was heat-denatured at 65°C for 10 minutes, placed on ice for 10 minutes and the remaining components were added. The reaction was incubated at 40°C for 2 hours and stopped by heating to 65°C for 15 minutes. Next, the reaction was neutralized by the addition of 10 μl 1M HEPES, 3 μl NaOAc and purified using a Qiagen QIAquick PCR purification kit (Qiagen). Half the cDNA was labeled with Cy5 and the other half with Cy3. Fluorescently-labeled cDNA samples were resuspended in 50 μl of hybridization buffer (358.8 mM LiCl, 200 mM LiMes, 50 mM EDTA, 0.1% LDS, 0.1% BSA) with the addition of 1 µl of poly A and 1 µl of tRNA, they were then competitively hybridized for 18 hours at 65°C to a custom cDNA array from woolly sculpin (Clinocottus analis;	  see	  section	  below,	  “Microarray probe sequences”)	  using	  an	  interwoven	  loop	  design (Kerr and Churchill 2001, Vinciotti et al. 2005). Each array then underwent a SDS wash (0.5 x SSC, 0.01% SDS at 50°C) for 10 minutes followed by a 0.06 X SSC wash for 10 minutes to remove unhybridized cDNA. Competitive hybridization was performed only between time point samples within a species and no direct cross-species hybridizations were performed. We used a loop design where each RNA sample was hybridized to either 2 or 4 arrays with fluor-reversal (two loops, one for each species of sculpin, Appendix B). The additional two technical replicates for some of the biological replicates build internal spokes of a hybridization loop in order to connect the loop to create more reliable comparison among the non-adjacent samples (Kerr and Churchill 2001).     The heterologous, or cross-species, microarray platform used for this analysis has been validated by numerous studies (e.g. Renn et al. 2004, Boswell et al. 2009, Chelaifa et al. 2010), but great care must be taken when designing and interpreting heterologous arrays as certain experimental designs can confound estimates of gene transcript levels (Kassahn 2008, Machado et al. 2009). For example, in experiments that competitively hybridize two different species on an array (e.g. Cy3 label for species A and Cy5 label for species B onto microarray platform developed for species C), relative transcription levels between species  47 could emerge from differences in transcript abundance or from a difference in sequence divergence that could unknowingly affect how well each species hybridizes to the array (Kassahn 2008, Machado et al. 2009). However, issues with sequence divergence can be avoided in heterologous experiments by competitively hybridizing samples from the same species, rather than between species (e.g. Cy3 label for species A and Cy5 label for species A onto microarray platform developed for species C). Under these analytical conditions, both samples will be similarly affected by sequence divergence and relative expression should solely reflect transcript abundance (Kassahn 2008). In this study, we only conducted competitive hybridization on samples from the same species that differed by hypoxia exposure time. This approach avoids false positives caused by the skewing of the hybridization ratios, which would occur as a result of large sequence divergence when two different species are hybridized together on an array. Further, the current study focused only on clones whose transcript levels were significantly altered in response to hypoxia exposure in both species and we avoid drawing any conclusions regarding the clones that were found to significantly change in only one of the species. This conservative approach of analysis eliminates the possibility of drawing conclusions from false negatives that may be due to poor hybridization.  3.3.4 Microarray probe sequences  The custom woolly sculpin array was an anonymous array consisting of approximately 9600 cDNA clones (array elements). A previous study (Ramon and Gracey, unpublished) generated approximately 800 EST (expression sequence tag) sequences from the cDNA clones on the array and an additional 900 were sequenced for this study. The 900 clones were chosen based on a preliminary data analysis. We selected a number of clones belonging to each of the broad clusters that were generated via hierarchical clustering. The clones were sequenced using the following protocols.  Briefly, clones were grown overnight in 96-well plates, their inserts amplified using flanking vector-specific	  primers,	  and	  the	  5’	  end	  sequence	  of	  the	  cDNA inserts were	  obtained	  by	  Sanger	  sequencing.	  The	  resulting	  5’-end EST sequences were then  48 clustered using two consecutive rounds of CAP3 assembly (Huang and Madan 1999). The resulting non-redundant contigs were annotated as the genes for which they had the greatest identity using SwissProt when the EST sequences yielded a hit with an E-value of less than 1e-4. We determined after sequencing ten 96-well plates that we had approximately 1% redundancy within our normalized libraries. 3.3.5 Statistical analyses Raw data were log2 transformed and lowess normalized using the maanova package (Wu et al. 2013) in R (R Core Team 2013). Differential expression for each clone	  across	  time	  points	  was	  detected	  using	  linear	  mixed	  models	  (‘array’	  as	  a	  random	  effect,	  and	  ‘dye’	  and	  ‘time	  points’	  as	  fixed	  effects)	  with	  100	  permutations	  and the F test statistic. Significance was estimated via false discovery rate (Storey 2002) using the qvalue package (Dabney and Storey 2013). Clones showing a main effect of time in hypoxia (q<0.01) were identified for each species and clones significant to both species (overlap dataset) were further analyzed (2408 clones, Fig. 3.1). We chose to use a high level of significance (q<0.01) rather than the more standard q<0.05 to account for minor effects on error estimates of variance due to uneven technical replicates across the biological samples. Genes (annotated ESTs) from the overlap dataset were assigned to gene ontology biological process categories using searches in EMBL-EBI QuickGO (www.ebi.ac.uk/QuickGo/) and Uniprot (http://www.uniprot.org) databases. A principal component analysis was preformed in EMA (Easy Microarray data Analysis) package (Servant et al. 2013) in R on the overlap dataset for both species simultaneously. Venn diagrams were created using VENNY (Oliveros 2007).  3.4 Results  Transcript levels of 2408 clones significantly changed in response to hypoxia exposure in both the hypoxia-tolerant tidepool sculpin and the hypoxia-intolerant silverspotted sculpin (Fig. 3.1). A principal component analysis, performed on these hypoxia-responsive clones, showed a clear separation between three clusters of time points. One cluster included the late time points for the relative hypoxia  49 exposure (24hr, 48hr, 72hr) in the tidepool sculpin, while the second cluster included the early time-points for the relative hypoxia exposure (3hr, 8hr, 24hr) and the absolute hypoxia exposure time point (24hr) in the silverspotted sculpin. The third cluster included the 0hr time point for both species as well as the 3hr relative hypoxia and 24hr absolute hypoxia exposure in the tidepool sculpin and the late time-points for relative hypoxia exposure (48hr, 72hr) in the silverspotted sculpin (Fig. 3.2).    3.4.1 Experiment 1: Relative hypoxia exposure time course  A total of 235 genes were assigned a UniProt identification from clones whose transcription significantly changed over time in both species. The annotated genes were broadly categorized into a number of gene ontology biological processes. The biological processes roughly fell into general categories of metabolism, regulation of cell number, immunity, protein production, protein localization and protein folding (Figs. 3.3, 3.4).  In the tidepool sculpin, relative hypoxia exposure resulted in a general increase in transcript abundance for genes associated with glycolysis (Fig. 3.3A), positive regulation of apoptosis (Fig. 3.3E), negative regulation of cell proliferation (Fig. 3.3H), chromosome organization (Fig. 3.4A), regulation of translational initiation (Fig. 3.4C) and protein localization (Fig. 3.4F). In particular, there were generally small changes in transcript level during the initial stages of the time-course, followed by large increases in transcript levels between 24hr and 72hr of hypoxia exposure. The silverspotted sculpin also exhibited an increase in transcript levels of genes involved in glycolysis, positive regulation of apoptosis, regulation of translational initiation and protein localization, but the transcript levels changed during the first 24hr of hypoxia and subsequently generally returned to levels that were close to those seen in the normoxic samples. Unlike the tidepool sculpin, the silverspotted sculpin exhibited a variable transcriptional response in the genes belonging to negative regulation of cell proliferation (Fig. 3.3H) and an increase or no change followed by a decrease in transcript levels of genes associated with chromosome organization (Fig. 3.4A).   50 There was an overall decrease in the transcript levels of genes assigned to oxidative phosphorylation (Fig. 3.3B), fatty acid oxidation (Fig. 3.3C), response to reactive O2 species (Fig. 3.3D), negative regulation of apoptosis (Fig. 3.3F) and positive regulation of cell proliferation (Fig. 3.3G) between 24hr and 72hr of hypoxia exposure in the tidepool sculpin, with the exception of a transient increase at 48hr in oxidative phosphorylation genes. In contrast, transcript levels of oxidative phosphorylation, fatty acid oxidation and negative regulation of apoptosis genes decreased in the silverspotted sculpin only in the initial stages of hypoxia, at which point there was a steady increase in transcript levels for the duration of hypoxia exposure (Fig. 3.3B,C). The transcriptional response in the silverspotted sculpin was more variable for genes associated with responses to reactive O2 species and positive regulation of cell proliferation, where some of the genes exhibited an increase in transcript level while others a decrease (Fig. 3.3D,G). Genes associated with RNA processing (Fig. 3.4B), translational elongation (Fig. 3.4D) and protein folding (Fig. 3.4G,H) showed a decrease in transcript levels during the first 24hr hypoxia exposure in the tidepool sculpin, followed by an increase for the rest of the hypoxia exposure. On the other hand, in the silverspotted sculpin there was a decrease in transcript abundance of genes belonging to RNA processing, heat shock proteins (HSPs, protein folding, Fig. 3.4H) and translational elongation (Fig. 3.4D) for the majority of the hypoxia exposure, with the exception of a transient increase at 48hr in translational elongation genes. The remainder of the protein folding genes in the silverspotted sculpin exhibited either a decrease or increase in transcript levels in response to hypoxia (Fig. 3.4G).    In the tidepool sculpin, there was variation in transcript abundance of hypoxia-induced immune response genes and protein maturation genes during the initial 24hr of hypoxia. This was followed by a larger divergence in the transcript levels of the genes, where the transcript levels for some genes decreased while others increased for the remainder of the hypoxia exposure (Fig. 3.3I, Fig. 3.4E). In the silverspotted sculpin, the hypoxia-induced immune response genes and protein maturation genes exhibited a similar pattern of transcription to the tidepool sculpin; however, in the silverspotted sculpin a change in transcript level of the genes  51 occurred in the first 24 hours with an increase in transcript levels for some genes and a decrease in others. By 48 hours of hypoxia exposure, gene transcript abundance returned close to the normoxic levels (Fig. 3.3I). To identify the genes that showed the greatest difference in transcriptional response among species in response to hypoxia exposure, we calculated Euclidean distance of the transcriptional time-course curves between the tidepool and silverspotted sculpins. The two genes with the highest Euclidean distances were DNA damage-inducible transcription 4 protein and Diablo homolog, both genes which are associated with positive regulation of apoptosis. Five other genes with high Euclidean distance measures were also associated with regulation of cell number. Other biological processes represented by genes with high Euclidean distance scores were glycolysis, regulation of translation initiation, immune response, chromosome organization and protein folding (Table 3.2, complete list of genes in Appendix C).  3.4.2 Experiment 2: Absolute environmental hypoxia  There was little or no difference in gene transcript levels between relative environmental hypoxia exposure (29 Torr) and absolute environmental hypoxia exposure (23 Torr) at 24hr for the silverspotted sculpin. In the tidepool sculpin, however, gene transcript levels were different between 24hr relative (17 Torr) and absolute (23 Torr) environmental hypoxia exposures. Genes for most of the biological processes in the tidepool sculpin either increased or decreased during relative environmental hypoxia, but the same genes did not change relative to normoxic levels in the absolute environmental hypoxia exposure (e.g. glycolysis, response to reactive O2 species, regulation of translational initiation, protein maturation) or transcription occurred in the opposite direction than that determined for relative environmental hypoxia exposure (e.g. oxidative phosphorylation, translational elongation, protein folding). Principal component analysis also demonstrated that the 24hr relative and absolute hypoxia time-points clustered together in the silverspotted sculpin, while in the tidepool sculpin, the two time-points were grouped in separate clusters (Fig. 3.2).  52 3.5 Discussion   The intertidal environment has many of the characteristics that have been proposed to favor the evolution of phenotypic plasticity, including frequent environmental change, temporal variation and reliable environmental cues (Moran 1992, Schlichting and Smith 2002). To identify genes or pathways that may have evolved to mediate phenotypic plasticity in response to environmental change, it is common to use transcriptomic analysis (Schlichting and Pigliucci 1993). In the present study, both the hypoxia-tolerant tidepool sculpin and the hypoxia-intolerant silverspotted sculpin altered the transcription of a large number of genes in response to relative environmental hypoxia. However, there were temporal differences in the transcriptional response between the two species. In the silverspotted sculpin, the greatest transcriptional change occurred rapidly, within 3 to 24 hours of hypoxia exposure. In the tidepool sculpin, on the other hand, the greatest transcriptional response for the most part occurred later at 24 to 72 hours of hypoxia exposure, which is beyond the typical timeframe that tidepool sculpins experience hypoxia in the intertidal zone (approximately 6 to 8 hours; Richards 2011).  Although protein translation is thought to occur quickly following transcription (e.g. in sea urchins; Ben-Tabou de-Leon and Davidson 2009), it must be noted that changes in gene transcription levels are not always matched by changes in protein abundance or activity, necessitating functional analysis. Despite this inconsistent link between gene transcription and function, it is thought that genes which show rapid and transient changes in transcription may be most associated with phenotypic plasticity (Aubin-Horth and Renn, 2009). The species-specific temporal patterns suggest that rapid changes in gene transcription, and therefore phenotypic plasticity mediated via gene regulation, are not likely an important component of survival in initial stages of hypoxia for the tidepool sculpin, which inhabits the most O2 variable environment. However, it is possible that this species does display phenotypic plasticity as a result of translational or post-translational regulation.  53 It is also possible that the tidepool sculpin is a phenotypic generalist and does not alter phenotype in response to fluctuations in O2 over environmentally realistic timeframes, in spite of living in an environment that appears to be conducive to the evolution of plasticity. If the lack of a transcriptional response within ecologically relevant timeframes is an indicator that the tidepool sculpin is a phenotypic generalist, several factors may contribute to favoring a fixed generalist phenotype strategy over phenotypic plasticity. As hypoxia exposure in the intertidal is a cyclic and often a daily event, the metabolic costs associated with continually altering phenotype may be greater than the benefit received via matching phenotype to the environmental condition. The lag time between the cue for environmental change and the production of a new phenotype could also cause a mismatch between the environment and the phenotype (Padilla and Adolph 1996) and it is possible that a time lag coupled with the short duration of a tidal cycle may constrain the benefits of phenotypic plasticity in the intertidal and instead favor the evolution	  of	  a	  ‘resilient’	  generalist. 3.5.1 Transcription of genes similarly induced or repressed by hypoxia in both species   Both species of sculpin similarly increased or decreased transcription of genes belonging to a number of biological processes, including glycolysis, positive regulation of apoptosis, immune response and protein maturation, albeit the greatest change in transcription of these genes occurred during different stages of hypoxia in each species. The transcription patterns of the later time-course of the tidepool sculpin shared similarities to the transcription patterns of the silverspotted sculpin during the initial stages of hypoxia. This was also seen in genes associated with negative regulation of apoptosis, chromosome organization, regulation of translational initiation and protein localization, except in these biological processes the direction of transcriptional change switched in the later half of the hypoxia exposure in the silverspotted sculpin. Therefore, while the tidepool sculpin exhibited little transcription change during initial hypoxia exposure followed by large changes in the later stages of hypoxia, in the silverspotted sculpin  54 transcription was initially altered first in one direction then reversed for the remainder of the hypoxia exposure.     Similar induction or repression patterns of gene transcription between the two species may be a result of either shared phylogenetic history or simply be a widespread, common response to hypoxia exposure that is observed in all fishes regardless of hypoxia tolerance. For example, both species of sculpin induced transcription of genes associated with glycolysis. The increase in transcription of glycolytic enzymes has been shown in a number of other fish species (e.g. Gracey et al. 2001, Ton et al. 2003, van der Meer et al. 2005, Ju et al. 2007) and indeed increased reliance on glycolysis during hypoxia has been well documented in fishes, reptiles and mammals and appears to be independent of hypoxia tolerance (Kelly and Storey 1988, Churchill et al. 1994, van den Thillart and van Raaij 1995, Webster 2003). Far less is known about the responses of the other biological processes to hypoxia in fishes, including apoptosis and protein maturation and therefore it is difficult to ascertain whether similar transcriptional changes between the two sculpin species during low O2 exposure is wide-spread among fishes or if it is a consequence of the sculpin-specific ancestry.  3.5.2 Transcriptionally divergent response between the species Transcription of genes involved in oxidative phosphorylation, fatty acid oxidation, cell proliferation, translational elongation, protein folding and antioxidant defenses differed between the two species, such that, for the most part, gene transcription decreased in the tidepool sculpin during hypoxia while in the silverspotted sculpin the transcriptional response was more variable. In contrast to the tidepool sculpin, transcription of genes associated with oxidative phosphorylation and fatty acid oxidation primarily increased during hypoxia in the silverspotted sculpin, while a split response of both an increase and decrease was seen among the genes associated with antioxidant defense and protein folding. These fundamentally divergent responses, despite the fact that they occur over different time frames in the species, may still highlight important biological processes that contribute to explaining the variation in hypoxia tolerance among the  55 fish species. However, as noted previously, transcriptional change does not always relate to a functional change at higher levels of biological organization, thus further studies are need to elucidate the role of these biological processes in the evolution of hypoxia tolerance.    The ability to suppress metabolic energy demands during hypoxia exposure may be a potential functional consequence of the transcriptional differences between the two species of sculpin. Energy suppression, which has been linked to decreases in sodium pump activity, protein synthesis, urea synthesis and gluconeogenesis, has been postulated to be the hallmark of hypoxia tolerance (Boutilier and St-Pierre 2000, Boutilier 2001, Bickler and Buck 2007, Richards 2009). If an animal is able to decrease energy demand it may be able to maintain energy balance despite a decrease in energy supply, thus prolonging survival during hypoxia. In the tidepool sculpin, transcriptional data suggests hypoxia-induced decreases in energetically expensive biological processes such as cell proliferation and translational elongation or biological processes dependent on the presence of O2, such as fatty acid oxidation, oxidative phosphorylation and ROS. Conversely, transcription of these genes did not decrease in the silverspotted sculpin, likely indicating the species may not down-regulate energy demand and therefore be unable to conserve their limited energy supply during hypoxia. As a result, the difference in hypoxia tolerance between the two species may be partially explained by the ability of the species to decrease energetically expensive biological processes, perhaps partially mediated by changes in gene transcription. Energy producing pathways requiring O2 are compromised during hypoxia and as a result there is an increased reliance on anaerobic glycolysis (as shown previously, e.g. in Carassius auratus, van den Thillart et al. 1980). In the tidepool sculpin, a decrease in the transcription of fatty acid oxidation genes and oxidative phosphorylation genes coupled with increased glycolytic transcription suggests a shift to O2-independent production of energy during hypoxia. In the silverspotted sculpin, despite an increase in transcription of glycolytic genes, there was also an overall induction of genes involved in oxidative phosphorylation and fatty acid oxidation (after a brief, transient decrease at 3 hours of hypoxia). This  56 counterintuitive response in the silverspotted sculpin is likely detrimental if the changes in transcript level are associated with function changes as they could result in increased production of ROS and fatty acid metabolites. Furthermore, the metabolic costs of increasing transcript levels of genes from futile biological pathways could result in a more severe disruption of metabolic energy balance in the silverspotted sculpins than in tidepool sculpin and contribute to their lower hypoxia tolerance. Silverspotted sculpins have roughly half the hepatic glycogen stores than tidepool sculpins and they use these glycogen stores at a higher rate when exposed to hypoxia (Chapter 2, Speers-Roesch et al. 2013), further supporting the notion that silverspotted sculpins have high metabolic demands that are not suppressed during hypoxia exposure. Previous functional studies and the transcriptional data from the current study suggest that the lower capacity of O2-independent production of energy in the silverspotted sculpin precludes the species from being able to down-regulate oxidative phosphorylation and fatty acid oxidation and thus conserve the limited energy supply during hypoxia.    3.5.3 Relative vs. absolute levels of hypoxia When both species were exposed to the same environmental O2 tension (23 Torr), there was a notable lack of a transcriptional response in the tidepool sculpin compared with large transcriptional changes in the silverspotted sculpin. The silverspotted sculpin showed similar transcriptional patterns at 24 hours between the relative (29 Torr) and absolute (23 Torr) hypoxia exposures. In contrast, substantial differences were seen in the tidepool sculpin at 24 hours between the relative (17 Torr) and absolute (23 Torr) hypoxia exposure, as demonstrated by the separation of clustering patterns of the time-course points for this species in the principal component analysis. For the tidepool sculpin, a more severe environmental hypoxia was required to induce changes in gene transcription than in the silverspotted sculpin, highlighting the greater ability of the tidepool sculpin to utilize other mechanisms of combating hypoxia, such as greater O2 extraction capacity. Likely, the transcriptional differences between the two species of sculpin reflect the patterns of species distribution across the marine nearshore  57 environment. Collectively, this study highlights not only large differences in the induction or repression of genes involved in certain biological processes between hypoxia-tolerant and -intolerant species, but the large variation in the initiation of a molecular response, both temporally and in response to different magnitudes of hypoxia.  3.5.4 Conclusions There was significant temporal variation in the transcriptional response to relative hypoxia exposure between the two sculpins. Despite living in a highly variable intertidal environment, the tidepool sculpin does not alter phenotype via changes in transcription unless hypoxia persists for longer than a typical tidal cycle or O2 levels become too low. In contrast, the silverspotted sculpin very quickly altered gene transcription during exposure to hypoxia, suggesting that the phenotypic traits already expressed were unable to support survival in low O2 environments. During sustained hypoxia, genes associated with a number of biological processes exhibited large transcriptional divergences between the two species. In the tidepool sculpin, the predominant pattern was a decrease in transcription while in the silverspotted sculpin the transcriptional response was more variable. This would suggest that the tidepool sculpin is better able to conserve limited energy stores by lowering energy expenditure associated with metabolically expensive biological processes during periods of extended hypoxia exposure, although further functional studies will be required to corroborate this finding. Ultimately, these transcriptional differences may play a significant role in determining the capacities of the species to tolerate fluctuating O2 in the environment and therefore determine the species distribution along the nearshore environment.    58 Figure 3.1  Venn diagram depicting the number of clones whose transcription significantly changed in response to hypoxia.  The number in the black circle represents the tidepool sculpin, the number in the light gray circle represents the silverspotted sculpin and the number in the dark gray area represents the common clones whose transcription significantly changed in response to hypoxia in both species. Significance set at q<0.01.    Tidepool Silverspotted3470 2408 904 59 Figure 3.2  Trajectory of transcription in the tidepool sculpin and the silverspotted sculpin exposed to relative environmental hypoxia over 72 hours and absolute environmental hypoxia at 24 hours.  Relative time points are indicated by square symbols and absolute time points are indicated by diamond symbols and (Abs). The tidepool sculpin is represented with black and the silverspotted sculpin is represented with red. The numbers in the brackets indicate the percent of variation in transcription among the time points explained by each of the respective principal components. The 0hr time points for both species overlap, but were offset for clarity.      60 Figure 3.3  Transcription levels of genes associated with metabolism (A-D), regulation of cell number (E-H) and immunity (I) in the tidepool sculpin and silverspotted sculpin.  Both species were exposed to relative environmental hypoxia for 72 hours (time-course, 17 Torr and 29 Torr, respectively). Additionally the tidepool sculpin and the silverspotted sculpin were exposed to absolute environmental hypoxia (23 Torr) for 24 hours and the resulting transcription of genes are represented in the light blue strips for each of the species. Genes represented in panel A are 6-phosphofructokinase liver type (red), 6-phosphofructokinase muscle type (black), alpha-enolase (blue), glucose-6-phosphate isomerase (green) and pyruvate kinase isozymes M1/M2 (purple); in panel B, ATP synthase subunit d (red), ATP synthase subunit e (black), ATP synthase subunit gamma (green), ATP synthase subunit O (purple), cytochrome b-c1 complex subunit 10 (blue); in panel C, 3-hydroxybutyrate dehydrogenase type 2 (blue), enoyl-CoA hydratase (black), hydroxyacyl-coenzyme A dehydrogenase (green), medium-chain specific acyl-CoA dehydrogenase (red); in panel D, mitochondrial uncoupling protein 2 (blue), peroxiredoxin-1 (black), phospholipid hydroperoxide glutathione peroxidase (purple), protein DJ-1 (green), tropomyosin alpha-1 chain (red); in panel E, catenin beta-1 (yellow), cell death activator CIDE-B (blue), death-associated protein-like 1 (coral), Diablo homolog (light blue), DNA damage-inducible transcript 4 protein (red), eukaryotic translation initiation factor 4 gamma 2 (tan), eukaryotic translation initiation factor 5A-1 (orange), poly(U)-binding-splicing factor PUF60 (black), kinase isozymes M1/M2 (purple), RAF proto-oncogene serine/threonine-protein kinase (magenta), serine/threonine-protein kinase Sgk1 (gray); in panel F, annexin A4 (red), annexin A5 (blue), Keratin, type I cytoskeletal 18 (black), macrophage migration inhibitory factor (green); in panel G, coronin-1A (red), eukaryotic translation initiation factor 5A-1 (blue), macrophage migration inhibitory factor (gray), peptidyl-prolyl cis-trans isomerase NIMA-interacting 1 (purple), peroxiredoxin-1 (red), peroxisome proliferator-activated receptor delta (black), ubiquitin-conjugating enzyme E2 A (green); in panel H, cystathionline gamma-lyase (black), fatty acid-binding protein, heart (red), RAF proto-oncogene serine/threonine-protein kinase (blue); in panel I,  61 beta-2-microglobulin (tan), C-C motif chemokine 25 (yellow), collectin-12 (blue), complement C1q subcomponent subunit C (black), complement C3 (red), complement factor H (magenta), coronin-1A (orange), cytochrome b-245 light chain (purple), glucose-6-phosphate isomerase (pink), macrophage migration inhibitory factor (gray), major histocompatibility complex class I-related gene protein (coral), mannan-binding lectin serine protease 1 (light blue), myeloid differentiation primary response protein MyD88 (green).     62 Figure 3.4  Transcription levels of genes associated with protein production (A-E), protein localization (F) and protein folding (G-H) in the tidepool sculpin and the silverspotted sculpin.  Genes represented in panel A are phospholipid hydroperoxide glutathione peroxidase (blue), protein SET (black), ubiquitin-conjugating enzyme E2 A (red); in panel B, 40S ribosomal protein S7 (blue), methylosome protein 50 (green), poly(U)-binding-splicing factor PUF60 (black), ribosome biogenesis protein NSA2 homolog (red); in panel C, eukaryotic translation initiation factor 1b (black), eukaryotic translation initiation factor 4 gamma 2 (red), eukaryotic translation initiation factor 4E-binding protein 3 (blue); in panel D, 40S ribosomal protein S21 (green), 40S ribosomal protein S7 (purple), 60S ribosomal protein L10 (magenta), 60S ribosomal protein L13 (blue), 60S ribosomal protein L30 (red), 60S ribosomal protein L39 (gray), elongation factor 1-gamma (black), eukaryotic translation initiation factor 5A-1 (yellow); in panel E, complement C1q subcomponent subunit C (black), complement C3 (blue), complement factor H (green), mannan-binding lectin serine protease 1 (magenta), peptidyl-prolyl cis-trans isomerase FKBP1A (purple), signal peptidase complex catalytic subunit SEC11A (red); in panel F, ADP-ribosylation factor 5 (black), ADP-ribosylation factor 6 (red), catenin beta-1 (purple), centrosomal protein of 57 kDa (black), clathrin light chain A (green), eukaryotic translation initiation factor 5A-1 (light blue), importin-7 (gray), keratin type I cytoskeletal 18 (yellow), translocation protein SEC63 homolog (blue); in panel G, peptidyl-prolyl cis-trans isomerase F (blue), peptidyl-prolyl cis-trans isomerase FKBP1A (green), peptidyl-prolyl cis-trans isomerase NIMA-interacting 1 (magenta), T-complex protein 1 subunit eta (gray), T-complex protein 1 subunit theta (purple), T-complex protein 1 subunit zeta (black); in panel H, 10 kDa heat shock protein (red), heat shock protein 75 kDa (green), heat shock protein HSP 90-alpha (black), heat shock protein HSP 90-beta (blue). See figure 3.3 legend for more detail.  63   64 Table 3.1  Experimental design of O2 exposures and sampling times for each species  Normoxia  Relative Hypoxia  Absolute Hypoxia Species   (3, 8, 24, 48, 72 hours)  (24 hours) Tidepool 158 Torr  17 Torr  23 Torr Silverspotted 158 Torr  29 Torr  23 Torr  65 Table 3.2  Euclidean distance of the transcriptional response between the tidepool sculpin and the silverspotted sculpin measured for the 72 hour time course.  Values are presented for the top twenty genes from biological processes presented in figures 3.3 and 3.4 and the remaining Euclidean distances can be found in the supplementary material.  Gene Name Gene UniProt ID Gene Ontology Name Euclidean distance DNA damage-inducible transcript 4 protein DDT4 GO:0043065~positive regulation of apoptosis 4.027 Diablo homolog, mitochondrial DBLOH GO:0043065~positive regulation of apoptosis 3.219 Fatty acid-binding protein, heart FABPH GO:0008285~negative regulation of cell proliferation 2.662 Serine/threonine-protein kinase Sgk1 SGK1 GO:0043065~positive regulation of apoptosis 2.659 Eukaryotic translation initiation factor 4 gamma 2 IF4G2 GO:0043065~positive regulation of apoptosis,  GO:0006446~regulation of translational initiation 2.598 Ubiquitin-conjugating enzyme E2 A UBE2A GO:0051276~chromosome organization 2.402 6-phosphofructokinase, liver type K6PL GO:0006096~glycolysis 2.339 Cystathionine gamma-lyase CGL GO:0008285~negative regulation of cell proliferation 2.139 Eukaryotic translation initiation factor 1b EIF1B GO:0006446~regulation of translational initiation 2.078 6-phosphofructokinase, muscle type K6PF GO:0006096~glycolysis 1.746 Complement factor H CFAH GO:0006955~immune response,  GO:0051604~protein maturation 1.590 Cell death activator CIDE-B CIDEB GO:0043065~positive regulation of apoptosis 1.501 3-hydroxybutyrate dehydrogenase type 2 BDH2 GO:0019395~fatty acid oxidation 1.481 Myeloid differentiation primary response protein MyD88 MYD88 GO:0006955~immune response 1.430 T-complex protein 1 subunit zeta TCPZ GO:0006457~protein folding 1.346 Heat shock protein HSP 90-beta HS90B GO:0006457~protein folding 1.334 Protein SET SET GO:0051276~chromosome organization 1.294 Major histocompatibility complex class I-related gene protein HMR1 GO:0006955~immune response 1.290 Tropomyosin alpha-1 chain TPM1 GO:0000302~response to reactive oxygen species 1.257 T-complex protein 1 subunit theta TCPQ GO:0006457~protein folding 1.251  66 Chapter 4  Variation Underlying Similar Phenotypes: Transcriptional Divergence in Hypoxia Responsive Genes Associated with Metabolism and Protein Synthesis in Three Species of Sculpin with the Same Hypoxia Tolerance 4.1 Summary The evolution of phenotypic convergence among taxa can emerge from similar or differing genetic mechanisms, making it difficult to predict the evolutionary change underlying phenotypes. In this study, we exposed three species of closely related marine fish from the superfamily Cottoidea (smoothhead sculpin [Artedius lateralis], sailfin sculpin [Nautichthys oculofasciatus] and Pacific staghorn sculpin [Leptocottus armatus]) to 72 hours of hypoxia to assess if transcriptional patterns differed among species with similar hypoxia tolerances. Individuals of each species were exposed to 23 torr and sampled	  at	  3	  and	  8	  hours	  (termed	  ‘short-term exposure’)	  and	  24,	  48	  and	  72	  hours	  (termed	  ‘long-term	  exposure’)	  and	  mRNA	  transcription was assessed with a custom designed microarray. A high proportion (65%) of clones on the microarray were shown not to differ in their hypoxia-induced transcriptional patterns in the three species. However, the majority of genes associated with metabolism and protein production differed in transcription patterns among the species in both the short- and long-term hypoxia expoures. These transcriptional data suggest that the three species of sculpin utilize metabolic pathways, specifically amino acid catabolism, glycolysis and fatty acid oxidation, differently to generate ATP in hypoxia. Transcription of genes involved in different steps of protein production, including mRNA processing, protein translation and  67 protein localization increased in the smoothhead sculpin, decreased in the sailfin sculpin and was variable in the Pacific staghorn sculpin. As changes in metabolism and protein production are considered integral to hypoxic survival, differences in the transcription patterns of genes associated with these processes suggests that the three sculpin species may have evolved similar hypoxia tolerance, in part, through different mechanisms.   4.2 Introduction Natural selection has long been implicated as the evolutionary cause underlying convergent evolution. This is based on the premise that repeated evolution of similar traits in response to an environmental pressure is unlikely to be a product of genetic drift (Endler 1986). Phenotypic convergence may be achieved via similar or different genetic changes depending if there are few or many paths for selection to act on. For example, genetic biases and constraints, such as pleiotropy or limited available genetic variation may decrease the number of ways in which an animal can evolve in response to an environmental stressor (Chevin et al. 2010, Christin et al. 2010, Losos 2011). In these instances it is more likely that similar genetics underlie convergent phenotypes. In turn, if genetic constraint is minimal and genetic variation high, more paths are available for evolution to proceed along, leading to greater instances of differing genetic basis of convergent phenotypes. As phenotypic similarity among taxa can be achieved in multiple ways, there has been a substantial effort in identifying instances of trait convergence and the underlying genetic and molecular mechanisms in order to understand the principles governing the evolution of phenotypic convergence (Hoekstra et al. 2006, Gross et al. 2009, O’Quin	  et	  al.	  2010,	  Rosenblum	  et	  al.	  2010,	  Conte	  et	  al.	  2012).	   Studies on ecological and physiological genomics have begun to uncover how genomes function and respond to produce complex phenotypes in response to environmental challenges. Analysis of transcriptomics is important as it provides an understanding of the variation in genome regulation in response to the environment, ultimately providing an important link in elucidating genotype-environment interactions, phenotypic plasticity and phenotypic evolution (Aubin- 68 Horth and Renn 2009, Whitehead 2012, Alvarez et al. 2015). Previous transcriptomic studies, such as those conducted using microarray platforms in lake whitefish (Coregonus clupeaformis; Derome et al. 2006) and Atlantic salmon (Salmo salar; Roberge et al. 2006) have examined whether similar transcriptional patterns underlie convergent phenotypes. Sympatric dwarf (limnetic) and normal (benthic) whitefish ecotypes have diverged in many traits, including bioenergetics and growth, in response to ecological pressures and competitive interactions (reviewed in Bernatchez et al. 2010). These phenotypic differences were found to be consistent across independent, isolated populations and are reflected by parallel differences in transcription of genes associated with swimming activity and energy metabolism between the whitefish ecotype pairs (Derome et al. 2006). Likewise, differences in the transcription of metabolic genes were found to underlie phenotypic differences in growth rate and age to sexual maturity between wild and farmed Atlantic salmon and the transcriptional patterns were consistent among multiple strains (Roberge et al. 2006). These transcriptomic studies provide instances of evolution of parallel molecular mechanisms in different populations or strains of a species in response to similar selective pressures.  Hypoxia (low O2 levels) is an environmental factor that has significant ecological and physiological consequences in animals, particularly in aquatic habitats due to the relatively low capacitance of water for O2 and the added pressure of anthropogenic eutrophication (Diaz and Breitburg 2009). As a result, there have been multiple, independent origins of hypoxia tolerance among diverse teleost lineages including perciformes, cypriniformes and characiformes (Hochachka and Somero 2002). Numerous researchers have attempted to identify the underlying traits contributing to whole animal hypoxia tolerance across different levels of biological organization, from the role of specific genes (Rytkonen et al. 2008, Andersen et al. 2009, Rytkonen et al. 2012), gene transcription (Gracey et al. 2001, Ton et al. 2003), biochemical pathways (van den Thillart et al. 1980, Martinez et al. 2006, Davis and Moyes 2007), physiological traits (Sollid et al. 2003, Rutjes et al. 2007) to behavioural responses (Sloman et al. 2006). However, although many traits have been implicated in the evolution of hypoxia tolerance, no  69 study has directly examined whether species with similar levels of hypoxia tolerance exhibit convergent transcriptional patterns.  Marine coastal systems are highly heterogeneous environments with areas prone to frequent O2 fluctuations, such as dense marshes and isolated tidepools, in close proximity to the O2 stable, deeper near-shore waters. As a result, there is great diversity in hypoxia tolerance among coastal species, such as in the group of fishes belonging to the superfamily Cottoidea. These fishes, known as sculpins, show a large range of hypoxia tolerance that matches the likelihood of hypoxia exposure in their habitat (Mandic et al. 2009, Chapter 2). We recently demonstrated that sculpin species that differ in hypoxia tolerance show highly divergent patterns of gene transcription in both the timing of the hypoxia-induced transcriptional change (initial time-frame of hypoxia exposure for the intolerant silverspotted sculpin, Blesias cirrhosis, and later time-frame of hypoxia exposure for the tolerant tidepool sculpin, Oligocottus maculosus) as well as the induction or repression patterns of the transcriptional response (Chapter 3). In order to determine if species with similar hypoxia tolerance exhibited convergent transcription patterns, we examined three species of sculpin: smoothhead sculpin (Artedius lateralis), sailfin sculpin (Nautichthys oculofasciatus) and Pacific staghorn sculpin (Leptocottus armatus). All three species inhabit the high subtidal, which is susceptible to periodic decreases in O2, and exhibit intermediate levels of hypoxia tolerance in relation to other sculpin species. The species were chosen because of their similar level of hypoxia tolerance, their location on separate clades of the sculpin phylogenetic tree, which suggests independent evolution of intermediate-tolerance (Fig. 4.1) and the relative abundance of individuals of each species. The three sculpin species were exposed to the same environmental hypoxia level for 72 hours and the liver was sampled for a microarray analysis of changes in the transcriptome. Overall, the goal of this study was to determine if three closely-related species of sculpin with similar hypoxia tolerance showed convergent transcriptional responses to hypoxia exposure.    70 4.3 Materials and Methods 4.3.1 Experimental animals Individuals of Artedius lateralis (smoothhead sculpin) and Leptocottus armatus (Pacific staghorn sculpin) were collected in July and August of 2010 near the Bamfield Marine Sciences Centre (BMSC), Bamfield, British Columbia. Smoothhead sculpins were captured at Ross Islets (48°52'24.0"N 125°09'42.0"W) using seine nets during low tide and Pacific staghorn sculpins in the Bamfield Inlet (48°49'06.1"N 125°08'30.9"W) using baited minnow traps. The sculpins were collected from habitats typical of these species (Marliave 1997), with the smoothhead sculpin collected from a shallow, sand and pebble beach and the Pacific staghorn from a beach with a muddy substrate. The two species were transported to The University of British Columbia and a third species of sculpin, Nautichthys oculofasciatus (sailfin sculpin) was acquired from the Vancouver Aquarium, British Columbia (BC), Canada. The sailfin sculpins were first generation offspring from wild fish caught in the waters around Stanley Park, Vancouver, BC. All three species of sculpin where held in fully aerated recirculating seawater at 12°C and 30 ppt salinity for at least 2 months prior to experimentation. Fish were fed daily with prawns and krill, except 24 hours before the experiments.     4.3.2 Critical O2 tension Critical O2 tension (Pcrit; water PO2 at which an animal no longer maintains routine O2 consumption rate and the rate decreases with a decline in water PO2) can be considered a proxy for hypoxia tolerance in sculpins (Chapter 2). The smoothhead sculpin and the Pacific staghorn sculpin Pcrit values of 35.7±6.9 Torr and 37.4±1.2 Torr respectively were obtained from Mandic et al. (2009). Following protocols described in detail in Henriksson et al. (2008), we determined that the Pcrit for the sailfin sculpin was 35.5±2.3 Torr (n=6) at 12°C and 30 ppt salinity. The Pcrit values among the species were not statistically different (ANOVA, p = 0.900).  71 4.3.3 Hypoxia exposure  For each species, individuals were transferred into four separate, 136 litre tanks of a custom built recirculating system, the Low Oxygen Control and Monitoring Aquatic Research System (Integrated Aqua Systems, Inc.). Fish were allowed to recover in air-saturated water for 24 hours, after which O2 levels were decreased to 23 Torr, a level of hypoxia corresponding to 65% Pcrit of all three species of sculpin. Fish were sampled for liver, muscle, heart, brain and gill at 3, 8, 24, 48 and 72 hours of hypoxia and two normoxic fish were also sampled at each of the hypoxic time points. All tissues were flash frozen in liquid N2 and stored in a -80°C freezer until analysis, but only liver samples were used in the present microarray experiments.  4.3.4 Microarray experiments  The protocols used for RNA extraction, reverse-transcription of RNA to cDNA, fluorescent labeling of the cDNA and hybridization of the microarrays are described in detail in Chapter 3. Briefly, RNA was extracted from the liver of each of the species using TRIzol (Invitrogen). Total RNA was then purified using glass-fibre filter columns (Qiagen) and reverse transcribed using MMLV-reverse transcriptase (Epicentre). The cDNA samples were purified using QIAquick PCR purification kits (Qiagen) and each sample was labeled separately with Cy5 and Cy3 dyes. Fluorescently labeled cDNA samples were hybridized to a custom cDNA array from woolly sculpin (Clinocottus analis). A total of 25 samples per species were used: 5 individuals sampled at normoxia and 4 individuals sampled at each hypoxic time-point. Hybridizations were only performed between 2 samples within the same species hybridized on the woolly sculpin array. We used an interwoven loop design (Kerr and Churchill 2001), where each sample was hybridized to either 2 or 4 arrays with fluor-reversal, creating a separate loop for each species.       4.3.5 Statistical analysis   The custom woolly sculpin array is comprised of approximately 9600 cDNA clones (array elements; Chapter 3). A total of 1700 clones were sequenced and the  72 resulting non-redundant contigs were annotated with the gene name only if the matched sequences yielded a hit with an E-value of  <1e-4 in SwissProt (for details see Chapter 3).	  For	  clarification,	  the	  term	  ‘clones’	  refers	  to	  all	  (sequenced	  and	  non-sequenced)	  array	  elements,	  while	  the	  term	  ‘genes’	  strictly	  refers	  to	  the	  sequenced	  and identified clones. Sequence divergence between the woolly sculpin used to create the microarray platform and the sculpin species under investigation may result in poor hybridization at certain clones. To minimize the effect of poor hybridization, clones with expression levels <2 standard deviations above background for any species were removed leaving 3294 clones for further analysis.  Raw data were log2-transformed and lowess normalized using the MAANOVA package (Wu et al. 2013) in R (R Core Team 2013). To test for the effect of time in hypoxia, differential expression for each clone across time-points was detected using	  linear	  mixed	  models	  (‘array’	  as	  a	  random	  effect	  and	  ‘dye’	  and	  ‘time’	  as	  fixed	  effects) with 100 permutations and the F-test statistic. False discovery rate (Storey 2002) in the q value package (Dabney and Storey 2013) in R was used to determine significance level (set at q<0.01). Clones showing a main effect of time in hypoxia in at least one of the species of sculpin (2734 clones in total; 1670 clones for smoothhead sculpin, 1183 clones for sailfin sculpin and 1778 clones for Pacific staghorn sculpin) were further analyzed to determine if the transcription patterns differed among the species.  Specifically, we tested whether the slope of the transcription response for short-term hypoxia response (0, 3hr, 8hr) and long-term hypoxia response (0, 24hr, 48hr, 72hr) differed among the species. The time-course was partitioned into short- and long-term responses because: 1) daily tidal events are the primary cause of hypoxic exposure in the environments inhabited by these species, resulting in ecologically relevant hypoxic exposure of up to 8 hours; 2) shallow coastal environments are increasingly influenced by anthropogenic eutrophication resulting in longer-term, physiologically challenging hypoxia exposures (Diaz and Breitburg 2009); and 3) our previous work in hypoxia-tolerant and -intolerant sculpins demonstrated a strong effect of time on transcription with differences between the early and the late hypoxia exposure (Chapter 3). A maximum-likelihood  73 model was applied to the data to determine the best fit intercept and linear slope for each species over short- and long-term hypoxia, using the subplex method of optim function, as implemented in the find.mle routine of the diversitree package (FitzJohn 2009). For each species pair (smoothhead sculpin vs. sailfin sculpin, smoothhead sculpin vs. Pacific staghorn sculpin, and sailfin sculpin vs. Pacific staghorn sculpin), we then used a likelihood-ratio test to determine whether the slopes were significantly different by comparing whether the full likelihood model (which allowed both species to have different slopes) to a constrained model with a single slope. If the drop in log- likelihood between the full and the constrained model was > 1.92 (the critical value for a chi-square distribution with 1 degree of freedom and an alpha level of 0.05) then we rejected the hypothesis that the slopes are the same for the two species. From this analysis, we broadly classified the data into 4 categories of transcriptional response: differences among the species during short-term hypoxia exposure, differences among the species during long-term hypoxia exposure, differences among the species in both the short- and long-term hypoxia exposure, and no differences among the species. For data presentation, we focused exclusively on annotated genes, which were assigned to gene ontology biological process categories using searches in EMBL-EBI QuickGO (www.ebi.ac.uk/QuickGo/) and Uniprot (http://www.uniprot.org).   4.4 Results and Discussion  To date, very few studies have examined whether phenotypic similarity among species corresponds to transcriptional similarities or differences in wild populations. Therefore, in this study we compared the transcriptional response to hypoxia (i.e., changes in gene transcription over the time in hypoxia) among three closely-related species of sculpin that show a similar level of hypoxia tolerance. Of the total number of clones on the microarray whose transcript levels significantly changed in at least one species over time in hypoxia (2556 clones), the patterns of transcription did not differ among the species in 64% of the clones, suggesting an overall similar biological response to hypoxia (Table 1). However, certain biological processes were enriched for genes that differed in transcription patterns among the  74 species, including metabolism, protein production, cell count (apoptosis and cell proliferation) and immune response (Table 1). Here, we focus specifically on biological processes considered extremely relevant to hypoxic survival, mainly metabolism and protein production, for which transcription of 59% and 73% of the genes respectively differed among the three species (Table 1, Fig. 4.2-4.3). These differences in transcription patterns of genes involved in biological processes considered essential to hypoxic survival indicate that the underlying molecular basis of the phenotype differs among the species.  Hypoxic survival is critically linked to the maintenance of energy supply and demand (Boutilier 2001). Hypoxia-tolerant animals are able to maintain cellular energy balance during hypoxia exposure by augmenting energy supply through a greater reliance on O2-independent ATP production (e.g. glycolysis) and higher concentrations of substrates to maintain these pathways (e.g. glycogen), as well decreasing energy demand through a coordinated suppression of energetically expensive processes, such as protein production, in order to conserve limited ATP stores (Hochachka et al. 1986, Hochachka et al. 1996, Bickler and Buck 2007). These hypoxia defense strategies have been noted in diverse vertebrate groups including sharks (Routley et al. 2002), various teleost fishes (van den Thillart et al. 1980, Via et al. 1994, Smith et al. 1996), reptiles (Land et al. 1993) and some mammals (Preedy et al. 1985). Mechanisms underlying the hypoxia defense strategies include changes in post-translational modifications (e.g. decrease in protein synthesis via phosphorylation of initiation factor eIF2a; Koumenis et al. 2002) and gene transcription (increase in glycolytic genes and decrease in protein synthesis genes during hypoxic exposure; Gracey et al. 2001, van der Meer et al. 2005). 4.4.1 Energy metabolism  Transcription of metabolic genes varied among the three sculpin species, suggesting that the species may rely on different metabolic pathways to generate ATP during hypoxia exposure. In the smoothhead sculpin there was a decrease in the transcript levels of glycolytic genes (e.g. fructose-bisphosphate aldolase B [ALDOB] and glyceraldehyde-3-phosphate dehydrogenase [GAPDH]) and genes  75 associated with glycogen degradation (glycogen debranching enzyme [GDE]; Fig. 4.4) compared to the other two species. These changes in transcript levels suggest that the smoothhead sculpin shifts glycogen turnover toward synthesis rather than degradation and decreases reliance on glycolysis during hypoxia. There was also a decrease in transcript levels for genes associated with fatty acid oxidation during short- and long-term hypoxia exposure (e.g. hydroxyacyl-coenzyme A dehydrogenase [HADH]; Fig. 4.4). Thus, it would appear that the smoothhead sculpin relies on enhanced amino acid degradation (rather than glycolysis or fatty acid oxidation) to generate substrates for the tricarboxylic acid (TCA), and thereby support ATP generation by oxidative phosphorylation (Fig. 4.2). Indeed, transcript levels of genes associated with amino acid degradation and the TCA cycle showed an increase in transcript levels during long-term hypoxia, while transcript levels of genes associated with amino acid synthesis decreased (Fig. 4.4-4.6). If gene transcript levels relate to biological function, these data suggest that in addition to the typical hypoxia-induced reliance on glycolysis for ATP generation, prolonged hypoxia exposure in smoothhead sculpins induces a greater reliance on amino acid catabolism as a source of ATP. At first glance, this increased reliance on O2-dependent ATP production via amino acid catabolism may seem like a surprising response for an organism to have in an O2 deprived environment, but previous studies have shown enhanced amino acid catabolism during hypoxia in at least two other fish species (Tilapia mossambica; Kutty 1972 and Oncorhynchus mykiss; Medale et al. 1987).     In contrast to the smoothhead sculpin, both the sailfin sculpin and the Pacific staghorn sculpin showed a general increase in the transcription of glycolytic genes during short- and long-term hypoxia exposure (e.g. GAPDH; Fig. 4.4) and both species showed a short-term increase in transcript levels of genes associated with fatty acid oxidation (e.g. HADH) followed by a decrease in transcript level during long-term hypoxia (e.g. medium-chain specific acyl-CoA dehydrogenase [ACADM]; Fig. 4.4). However, these two species diverged in their response for other metabolic processes. Transcription patterns of genes associated with amino acid synthesis and degradation in the sailfin sculpin were similar to those of the smoothhead sculpin,  76 with a decrease in beta-ureidopropionase [UPB1] and an increase in 4-hydroxyphenylpyruvate dioxygenase [HPD] (Fig. 4.4). The transcriptional response of amino acid-related genes in the sailfin and smoothhead sculpin differed from the Pacific staghorn sculpin, which showed transcriptional patterns that suggest both amino acid synthesis and degradation increased (Fig. 4.4). We also identified a significant increase in genes involved in glycogen degradation in the sailfin sculpin while in the Pacific staghorn sculpin there was shift from glycogen synthesis in the short-term to glycogen degradation during long-term hypoxia exposure. The transcriptional data strongly suggests that the three sculpin species utilize different metabolic pathways for ATP generation during exposure to low O2 (Fig 4.2).  4.4.2 Protein production, protein localization and protein folding  Multiple steps are involved in cellular protein production, including gene transcription, RNA processing and translation (Fig. 4.3). Differences in the transcriptional patterns of genes associated with these processes suggest that the smoothhead sculpin increased protein production during hypoxia exposure, while the sailfin sculpin decreased protein production and the Pacific staghorn had a variable response (Fig. 4.7). In the smoothhead sculpin, this was demonstrated by increased transcript levels of genes associated with translation (e.g. eukaryotic translation initiation factor 4 gamma 2 [EIF4G2] and eukaryotic translation initiation factor 4E-1A [EIF4E1A]; Fig. 4.7). In contrast, the sailfin sculpin showed a decrease in transcription of genes associated with translation, while the Pacific staghorn’s	  response was variable, as highlighted by transcriptional patterns of both the genes associated with RNA processing (e.g. ribonuclease inhibitor [RNHI] versus RNA-binding motif, single-stranded-interacting protein 2 [RBMS2]) and genes associated with translation (e.g. EIF4G2 versus EIF4E1A; Fig. 4.7). To fully examine protein production, it is important to not only include transcription, RNA processing and translation, but also consider steps such as protein folding and protein targeting or localization. While for large proportion of genes associated with transcription, RNA processing and translation patterns of transcription differed among the species, a majority of genes associated with protein localization and protein folding  77 were transcriptionally similar among the species (Fig. 4.3). In particular, all three species showed an increase in transcription of genes associated with protein folding, especially during long-term hypoxia exposure. This is not surprising since hypoxia exposure has been shown to increase protein-folding chaperones, such as the heat shock protein 70, in the liver of rats (Aoe et al. 1997) and heat shock protein 90 in the Chinese shrimp (Fenneropenaeus chinensis; Li et al 2009). 4.4.3 Transcription patterns of hypoxia intermediate-tolerant species versus a hypoxia-tolerant species We previously showed that the transcription patterns in the very hypoxia-tolerant tidepool sculpin aligned with the characteristics attributed to tolerant animals, such as an increase in transcript levels of glycolytic genes and a decrease in transcripts of genes associated with energetically expensive processes such as cell proliferation (Chapter 3). Unlike the tidepool sculpin, the three intermediate–tolerant sculpin species in this study showed only some of the characteristics considered important to hypoxia tolerance, and these characteristics differed among the species. For example, the smoothhead sculpin showed a decrease in transcription of genes involved in some of the pathways that require O2, such as fatty acid oxidation, while sailfin sculpins showed a decrease in protein production, and both the sailfin sculpin and Pacific staghorn sculpin showed an increase in glycolysis (Fig. 4.2 and 4.3). However, despite the considerable variation in transcriptional patterns for genes that were a priori predicted to be important in hypoxia, the three sculpin species exhibit a very similar overall tolerance to hypoxia. Species differences in transcription of some of the genes involved in metabolism and protein production may be offset by additional variation in other traits leading to a similar ability to tolerate hypoxia. The transcriptional patterns thus suggest that the three species of sculpin have evolved similar levels of hypoxia tolerance via different underlying mechanisms.   78 4.4.4 Wild caught versus lab-reared species  In this study, individuals of the smoothhead sculpin and the Pacific staghorn sculpin were wild-caught, while individuals of the sailfin sculpin were first generation lab-reared. This may complicate analyses as the wild-caught species, unlike the sailfin sculpin, may have been exposed to hypoxia during development, which can have consequences on the adult hypoxic phenotype (Robertson et al. 2014). However, an examination of the patterns of all transcriptionally-responsive clones using heat maps across the three species produced no clear indication to suggest that the overall transcriptional response of the sailfin sculpin differed to a great degree from the other two species (data not shown). Of course, this does not exclude the possibility that some of the patterns described in this study are a result of differences in developmental plasticity. One can get around the effects of developmental plasticity by examining first generation lab-reared individuals for all species, although epigenetic effects can persist across many generations increasing the difficulty of unraveling the genetic mechanisms of survival in hypoxia.  4.4.5 Short-term versus long-term responses Time is an important factor to consider when examining changes in gene transcription (Aubin-Horth and Renn 2009). In our previous work (Chapter 3), we showed that time in hypoxia had a strong effect on gene transcription and that hypoxia-tolerant and -intolerant sculpins showed different temporal patterns of gene transcript levels. In general, the hypoxia-tolerant tidepool sculpin showed changes in gene transcript levels during longer-term exposure (24 to 72 hr) versus the hypoxia-intolerant silverspotted sculpin, which changed gene transcript levels during early (short-term hypoxia; 3 to 8 hr) hypoxia exposure (Chapter 3) and sometimes in opposite directions of the hypoxia-tolerant sculpin. The initial 8 hr hypoxia exposure is an ecologically relevant time-frame for species inhabiting the intertidal, while the longer duration of hypoxia lasting for several days is a physiological challenge	  testing	  the	  species’	  capacities	  for	  hypoxic	  survival.	  Therefore, it is not surprising that different mechanisms may shape the hypoxic response between the short and long exposures to hypoxia and we predicted that in  79 the current study differences in transcription patterns among the species would occur in both the short- and long-term hypoxia exposure. Indeed, there was an effect of time on transcript levels, with roughly equal number of clones showing differences in transcription patterns among species during short-term hypoxia exposure as during long-term hypoxia exposure (41% and 43% respectively), while a small proportion of clones (16%) showed transcriptional differences across both time scales. Overall, transcription of glycolytic genes, as an example, differed in the species both in the short- and long-term hypoxia, although for most genes the differences occurred only in one time-frame or the other. It would appear that the underlying aspects of hypoxia tolerance in sculpins are influenced by divergent transcriptional responses as well as strong temporal effects. Taken together, this suggests that the capacity to acclimate may differ among the species and despite a similar starting level of hypoxia tolerance the reaction norms may be remarkably different among the three sculpin species when comparing the trajectory from normoxia, through short-term to long-term hypoxia exposure.  4.4.6 Conclusion  In this study, we found that three species of sculpin with similar hypoxia tolerance had very different patterns of transcription in metabolism and protein production genes. These biological processes are thought to influence hypoxia tolerance, suggesting that these species use alternate mechanisms to achieve a similar overall phenotype. Our results argue against convergence as a result of constraint, and decrease the likelihood that similarities in hypoxia tolerance are a result of species relatedness. The divergent transcriptional patterns indicate that different genetic changes were selected upon during the evolution of intermediate hypoxia tolerance in the three species of sculpin.    80 Figure 4.1  A sculpin phylogenetic tree based on cytochrome b sequences. The data are unpublished courtesy of M.L. Ramon. Species names in red represent hypoxia-tolerant sculpins, in blue represent intermediate hypoxia tolerant sculpins, in green represent hypoxia-intolerant sculpins and in black represent sculpins with unknown hypoxia tolerance. Names of the three species of sculpin examined in this study are boxed (Artedius lateralis or smoothhead sculpin, Nautichthys oculofasciatus or sailfin sculpin and Leptocottus armatus or Pacific staghorn sculpin).                        Cottus bairdiiLeptocottus armatusJordania zonopeBlepsias cirrhosusScorpaenichthys marmoratusStellerina xyosternaRhaphocottus richardsoniiNautichthys oculofasciatusMyoxocephalus polyacanthocephalusMyoxocephalus scorpiusEnophrys bisonDasycottus setigerAscelichthys rhodorusIcelinus borealisChitonotus pugetensisRuscarius creaserRuscarius meanyiClinocottus embryumClinocottus recalvusClinocottus globicepsOrthonopias triacisHemilepidotus hemilepidotusClinocottus analisLeiocottus hirundoOligocottus snyderiOligocottus maculosusOligocottus rubellioOligocottus rimensisArtedius notospilotusClinocottus acuticepsArtedius fenestralisArtedius harringtoniArtedius lateralisCorallineCottus asper 81 Figure 4.2  Schematic representation of genes associated with metabolism.  Arrows represent gene transcription changes in response to hypoxia in the smoothhead sculpin (black), sailfin sculpin (blue) and Pacific staghorn sculpin (red).  Gray arrows represent genes with similar transcription patterns among the species. Short arrows indicate transcription patterns during short-term hypoxia, long arrows indicate transcription patterns during long-term hypoxia, double arrows indicate transcription patterns during short- and long-term hypoxia, a vertical line indicates	  no	  change	  from	  normoxia	  and	  the	  letter	  ‘v’	  represents	  a	  variable	  response.	  Full names for each gene abbreviation are listed in either Fig. 4.3-4.5 legends.    82 Figure 4.3  Schematic representation of genes associated with protein production, localization and folding.  See Fig. 4.4-4.6 legends for more details.    CytosolNucleusDNAPrimary RNAmRNATranscriptionRNA processingmRNAAmino acidsGeneSAP18CTDSP2RNHIRBMS2NSA2SPCS2EIF4EBP3EIF4E1ARPL30EIF4G2ARF5EIF1BTranslationProtein LocalizationSEC63IPO7SSR3Protein FoldingQSOX1HSP90AB1FKBP1ACCT6APDIA4V VRibosomebiogenesisEEF1GEIF5ACCT7HSP90AA1HSPE1SEC61ARPL10RPL39RPS21RPS27RPS7NR1D2VVVVVVV 83 Figure 4.4  Transcript levels of genes associated with metabolism in the smoothhead sculpin (black line, square symbol), sailfin sculpin (blue line, diamond symbol) and Pacific staghorn sculpin (red line, inverted triangle symbol) exposed to 72 hours of hypoxia.  Solid lines represent the genes for which transcription significantly changed in response to hypoxia for a given species (q < 0.01) and dashed lines represent genes for which transcription did not significantly change in response to hypoxia. Opaque lines represent the portion of the time-course that is the focus of the difference among the species, while transparent lines represent the portion of the time-course for which transcription does not differ among the species. Letters represent significant difference in transcript levels between species (A-B represent difference in the short term hypoxia and X-Z represent differences in the long term hypoxia). Data are represented as mean ± SE. The genes in the panels are classified in the following categories: glycogen metabolic process (protein phosphatase 1 regulatory subunit 3C-B [ppp1r3cb], glycogen debranching enzyme [GDE] and phosphoglucomutase-1 [PGM1]), glycolysis (fructose-bisphosphate aldolase B [ALDOB], alpha-enolase [ENO1] and glyceraldehyde-3-phosphate dehydrogenase [GAPDH], polyol pathway (aldose reductase [AKR1B1]), the pentose-phosphate shunt (GDH/6PGL endoplasmic bifunctional protein [H6PD]), pyruvate metabolic process (pyruvate dehydrogenase kinase isozyme 4 [PDK] and mitochondrial pyruvate carrier 1 [MPC1]), fatty acid oxidation (hydroxyacyl-coenzyme A dehydrogenase [HADH]), tricarboxylic acid cycle (ATP-citrate synthase [ACYL]), oxidative phosphorylation (cytochrome b-c1 complex subunit 6 [UQCRH], ATPase inhibitor [ATPIF1], NADH dehydrogenase iron-sulfur protein 8 [NDUFS8] and HIG1 domain family member 1A [HIGD1A]), amino acid degradation (homogentisate 1,2-dioxygenase [HGD] and 4-hydroxyphenylpyruvate dioxygenase [HPD]) and amino acid biosynthesis (betaine—homocysteine S-methyltransferase 1 [BHMT] and beta-ureidopropionase [UPB1]).  84  85 Figure 4.5  Genes for which all three species showed similar significant changes in transcription in response to 72 hours of hypoxia.  The smoothhead sculpin is represented by the black line and square symbol, the sailfin sculpin is represented by the blue line and diamond symbol and the Pacific staghorn sculpin is represented by the red line and inverted triangle symbol. Data are represented as mean ± SE. The genes in the panels are classified in the following categories: pentose-phosphate shunt (ribulose-phosphate 3-epimerase [RPE]), fatty acid oxidation (medium-chain specific acyl-CoA dehydrogenase [ACADM]), oxidative phosphorylation (NADH dehydrogenase 1 alpha subcomplex subunit 1 [NDUFA1] and ATP synthase subunit d [ATP5H]), translational initiation (eukaryotic translation initiation factor 1b [EIF1B]), protein localization (importin-7 [IPO7] and translocon-associated protein subunit gamma [SSR3]) and protein folding (peptidyl-prolyl cis-trans isomerase FKBP1A [FKBP1A], protein disulfide-isomerase A4 [PDIA4] and T-complex protein 1 subunit zeta [CCT6A]).     86 Figure 4.6  Similar transcription patterns in genes for which only a subset of species showed a significant effect of time in hypoxia.  The smoothhead sculpin is represented by the black line and square symbol, the sailfin sculpin is represented by the blue line and diamond symbol and the Pacific staghorn sculpin is represented by the red line and inverted triangle symbol. Solid lines represent the genes for which transcription significantly changed in response to hypoxia for a given species (q < 0.01) and dashed lines represent genes for which transcription did not significantly change in response to hypoxia. Data are represented as mean ± SE. The genes in the panels are classified in the following categories: glycolysis (glucose-6-phosphate isomerase [GPI]), tricarboxylic acid cycle (glutamate dehydrogenase [GLUD1]), oxidative phosphorylation (cytochrome c oxidase assembly protein COX14 [COX14], ATP synthase subunit e [ATP5I] and ATP synthase subunit O [ATP5O]), amino acid degradation (alanine aminotransferase 2 [GPT2], methylmalonate-semialdehyde dehydrogenase, acylating [ALDH6A1] and D-amino-acid oxidase [DAO]), transcriptional repression (nuclear receptor subfamily 1 group D member 2 [NR1D2]), ribosomal biogenesis (60S ribosomal protein L10 [RPL10], 60S ribosomal protein L39 [RPL39], 40S ribosomal protein S21 [RPS21], 40S ribosomal protein S27 [RPS27] and 40S ribosomal protein S7 [RPS7]), translational elongation (elongation factor 1-gamma [EEF1G] and eukaryotic translation initiation factor 5A-1 [EIF5A]), protein folding (10 kDa heat shock protein [HSPE1], heat shock protein HSP 90-alpha [HSP90AA1] and T-complex protein 1 subunit eta [CCT7]) and protein localization (protein transport protein Sec61 subunit alpha [SEC61A]).    87  88 Figure 4.7  Transcript levels of genes associated with protein production, localization and folding in the smoothhead sculpin (black line, square symbol), sailfin sculpin (blue line, diamond symbol) and Pacific staghorn sculpin (red line, inverted triangle symbol) exposed to 72 hours of hypoxia.  The genes in the panels are classified in the following categories: transcriptional repression (histone deacetylase complex subunit SAP18 [SAP18] and carboxy-terminal domain RNA polymerase II polypeptide A small phosphatase 2 [CTDSP2]), RNA metabolic process (ribonuclease inhibitor [RNHI]), RNA processing (RNA-binding motif, single-stranded-interacting protein 2 [RBMS2] and ribosome biogenesis protein NSA2 homolog [NSA2] and 60S ribosomal protein L30 [RPL30]), translation (signal peptidase complex sunbunit 2 [SPCS2]), translational initiation (eukaryotic translation initiation factor 4E-binding protein 3 [EIF4EBP3], eukaryotic translation initiation factor 4E-1A [EIF4E1A] and eukaryotic translation initiation factor 4 gamma 2 [EIF4G2]), protein localization (ADP-ribosylation factor 5 [ARF5] and translocation protein SEC63 homolog [SEC63]), and protein folding (sulfhydryl oxidase 1 [QSOX1] and heat shock protein HSP 90-beta [HSP90AB1]). See Fig. 4.3 legend for more detail.  89     90 Table 4.1  Summary of clones with similar and different transcription patterns among the three species of sculpin.  Clones are organized into three broad categories: non-sequenced clones, sequenced but unidentified clones, and sequenced and identified clones (genes), which are further divided into biological processes.    Similar Transcription Patterns   Different Transcription Patterns Categories Number of genes or clones Percent  Number of genes or clones Percent Metabolism 27 41%  39 59% Chromosome organization 3 100%  0 0% Protein production 3 27%  8 73% Ribosome biogenesis 5 71%  2 29% Protein folding 5 71%  2 29% Protein localization 3 60%  2 40% Protein modification 10 50%  10 50% Cell count 9 37%  16 63% Response to reactive O2 species 3 50%  3 50% Immune response 11 37%  19 63% Other biological processes 20 61%  12 39% Unknown biological processes 19 66%  10 34% Unidentified clones 213 55%  174 45% Non-sequenced clones 1328 69%  609 31% Total  1659 65%  907 35%   91 Chapter 5  General Discussion Understanding the traits and mechanisms underlying complex phenotypes that have evolved in response to organisms living in heterogeneous environments ultimately aids us in determining the limitations that set species boundaries and niches and is one of the key themes of integrative biology. Combining evolutionary, physiological and ecological understanding of species provides essential information to enable predictions of how anthropogenic-induced environmental challenges (e.g. eutrophication and climate change) will have impacts on biodiversity in the future. For my doctoral research, my overall objective was to explore how phenotypic plasticity and adaptive evolution have contributed to the ability of different sculpin species to survive bouts of hypoxia. To address my overall objective, I examined if variation in fixed or non-plastic biochemical traits in multiple tissues and transcriptional-based plastic responses in the liver explained the variation in hypoxia tolerance in sculpins. Sculpins are a group of closely-related species that experience different O2 regimes resulting in variation in tolerance, and as such are a tractable model system to understand what traits underlie the complex phenotype of hypoxia tolerance in wild fishes living in the nearshore marine environment.  5.1 Major Findings and Implications 5.1.1 Measures of hypoxia tolerance  Critical O2 tension (Pcrit) has long been used as an indirect measure of hypoxia tolerance (Hagerman 1998, Chapman et al. 2002, Mandic et al. 2009, Seibel 2011, Speers-Roesch et al. 2012). The primary reason for this is that Pcrit is  92 considered an indicator of O2 extraction and delivery and an animal with lower Pcrit potentially has greater hypoxia tolerance as a result of greater O2 uptake and transport to tissues. In sculpin, the correlation between Pcrit and LOE50, a measure of hypoxia tolerance, supports the idea that Pcrit may be used as a proxy for hypoxia tolerance (Chapter 2). This is also reinforced by a study on hypoxia-tolerant epaulette shark (Hemiscyllium ocellatum) and hypoxia-intolerant shovelnose ray (Aptychotrema rostrata) which found that a comparable percentage of Pcrit produced similar levels of arterial hypoxemia (Speers-Roesch et al. 2012). However, killifish (Fundulus heteroclitus), a well known hypoxia-tolerant species (Stierhoff et al. 2003), was found to have high Pcrit values of approximately 63.9 Torr, indicating that Pcrit does not represent an accurate assessment of tolerance in this species (Richards et al. 2008). Therefore, while Pcrit is an appropriate proxy measure of hypoxia tolerance in sculpins, this is not universally applicable to all fish species, indicating that careful quantification of several measures of hypoxia tolerance are necessary to provide an accurate estimation of hypoxia tolerance.       5.1.2 Multiple candidate traits underlie variation in hypoxia tolerance among sculpins   The hypoxic response of an organism is a complex phenotype with many underlying traits that span across different levels of biological organization. As such, I have examined a number of traits in sculpins, summarized in Table 5.1, from behavior (Mandic et al. 2009a), morphology and physiology (Mandic et al. 2009b) to biochemistry (Chapter 2) and transcriptional responses (Chapter 3 and 4). I have previously shown that aquatic surface respiration, aerial emergence, gill surface area and hemoglobin-oxygen (Hb-O2) binding affinity are putative adaptive traits in the evolution of hypoxia tolerance in sculpins. These traits are involved with the extraction and delivery of O2, and therefore partially determine whole-animal performance through their impact on cellular and tissue function. However, organ systems, from the brain, to heart, liver and muscle, perform different roles during hypoxia exposure, likely resulting in tissue-specific selective pressures. Therefore, it  93 is necessary to examine the effect of hypoxia on different tissues in order to begin to understand the coordination of traits in combating periodic decreases in O2.    In this thesis, I found a relationship between biochemical traits and hypoxia tolerance in sculpins in the brain. Specifically, maximal activity of brain metabolic enzymes measured under normoxic conditions correlated with hypoxia tolerance. The most tolerant species had higher activity than the less tolerant species in enzymes associated with anaerobic metabolism, particularly lactate dehydrogenase which correlated with both Pcrit and LOE50 in a phylogenetically independent manner (Chapter 2). Brain is a hypoxia-sensitive tissue and O2 lack in the brain rapidly leads to organismal death (Nilsson 1996, Boutilier 2001, Nilsson 2001). The high constitutive activity of metabolic enzymes in species frequently exposed to hypoxia may help protect the brain from hypoxia-induced energy imbalance when environmental O2 rapidly decreases in tidepools at night.  Maximal enzyme activity and metabolic substrates (e.g. glycogen, glucose and creatine phosphate) involved in anaerobic metabolism in the liver and in the white muscle differed across the sculpin species. However, the variation in these tissues, unlike the brain, did not correlate with hypoxia tolerance (Chapter 2). In response to hypoxic exposure, fish have been noted to decrease swimming activity (Chapman and McKenzie 2009), lowering the energetic demand of the white muscle to match the hypoxia-induced decrease in energy supply. Lower energetic demand of the white muscle during hypoxia may negate the necessity of increasing capacity of anaerobic pathways and may explain the lack of relationship between maximal activity of metabolic enzymes in sculpins and hypoxia tolerance. Based on a study on Fundulus grandis which showed that there was an increase in maximal activity of glycolytic enzymes in the liver during exposure to long-term hypoxia, in contrast to a decrease seen in the muscle (Martinez et al. 2006), I predicted that liver function may be important to hypoxic survival and therefore the activities of anaerobic enzymes would be higher in the liver of hypoxia-tolerant versus hypoxia-intolerant sculpin species. However, the lack of relationship suggests that liver function may be less important in the evolution of hypoxia tolerance in sculpins than originally predicted. Or, it is possible that while fixed, constitutively-expressed biochemical  94 traits in the liver do not correspond to hypoxia tolerance, changes in cellular phenotype of the liver in response to low O2 may help contribute to overall hypoxia tolerance. Indeed, in estuarine long-jaw mudsucker (Gillichthys mirabilis) exposure to hypoxia induced increases in transcription of many genes in the liver, in contrast to the muscle where there was primarily a decrease in transcription or the brain which showed minimal changes in gene transcription (Gracey et al. 2001). This makes the liver a good candidate to examine if, and how, transcriptional-based plasticity is related to hypoxic survival in sculpins.      I exposed two species of sculpin (tidepool sculpin and silverspotted sculpin; Chapter 3) and three species of sculpin (smoothhead sculpin, sailfin sculpin and Pacific staghorn sculpin; Chapter 4) to hypoxia and found changes in transcription of genes associated with a number of biological processes in the liver. In response to hypoxic exposure, both the tolerant tidepool sculpin and the intolerant silverspotted sculpin increased transcription of glycolytic genes, apoptotic genes, protein localization genes and many genes associated with the immune response. However, many differences were also noted between the species, where the tidepool sculpin decreased transcription of genes belonging to oxidative phosphorylation, fatty acid oxidation, cell proliferation and antioxidant defense, while the silverspotted sculpin either increased transcript levels or showed a variable transcriptional response (Chapter 3). If the transcription patterns are linked to functional changes, the tolerant tidepool sculpin, in contrast to the intolerant silverspotted sculpin, may lower energy expenditure of the liver through a decrease in either metabolically expensive processes or pathways that require O2 if hypoxia persists for longer than 24 hours. The transcriptionally-induced phenotypic change in the liver is a potential mechanism underlying differences in survival of the two species only when, or if, the species are exposed to longer bouts of hypoxia. In chapter 4, three sculpin species with intermediate hypoxia tolerance were exposed to low O2 in order to assess gene transcription changes in the liver. While it is difficult to predict how transcription patterns in species with intermediate tolerance will compare to the species with high and low tolerance levels, I expected that transcriptional changes would be induced in similar biological processes across  95 the same timeframes in all species. This was the case as changes in transcription in the intermediate-tolerant species occurred for many of the same processes, such as glycolysis, oxidative phosphorylation, fatty acid oxidation, protein translation, protein localization and protein folding (Table 5.1). While transcription patterns for some of the biological processes in the three intermediate-tolerant species were similar to the hypoxia-tolerant tidepool sculpin (e.g. decrease in transcription of fatty acid oxidation genes in the smoothhead sculpin, decrease in transcription of translational initiation genes in the sailfin sculpin and an increase in transcript levels of glycolytic genes in the Pacific staghorn sculpin), some were more similar to the hypoxia-intolerant silverspotted sculpin (e.g. increase in transcription of fatty acid oxidation genes in the sailfin and the Pacific staghorn sculpin). The intermediate-tolerant species also exhibited a combination of transcriptional response of the two extreme tolerant species in genes belonging to biological processes such as oxidative phosphorylation and ribosome biogenesis. While it is difficult to predict the extent of functional consequence of these transcription patterns, similarities of the molecular responses to both hypoxia-tolerant and -intolerant species may contribute to the intermediate hypoxia tolerance phenotype in the three sculpin species.  5.1.3 Short- versus long-term hypoxia exposure A focus on ecologically appropriate time-frames allows us to gain an understanding of the mechanisms that may have been selected for in response to hypoxia. A typical hypoxic bout for an intertidal species is dictated by changes in the tidal cycle and therefore the environmentally relevant exposure of sculpins is up to 6 to 8 hours of hypoxia. Yet, the time-course of hypoxic exposure for sculpin species in both Chapter 3 and 4 was 72 hours with sampling occuring at 3, 8, 24, 48 and 72 hours. The ecologically relevant 3 and 8 hour samples are critical in determining if transcriptional-based plasticity plays a role in hypoxic survival that contributes to the distribution patterns observed for the species along the nearshore environment. While the remainder of the time-course is an arbitrary length of time, it is important to examine as it provides insights into the traits that may contribute to the upper  96 limits	  of	  the	  species’	  capacities	  to	  survive	  hypoxia	  bouts	  (i.e. deaths of some individuals were noted in different sculpin species by 72 hours of exposure). As anthropogenic eutrophication of the coastal waters has caused a steady increase in the duration and severity of hypoxia bouts, which are predicted to continue to increase, it is critical that we understand how traits change in different species in response to longer bouts of hypoxia. 5.1.4 The role of constitutive differences in phenotype versus phenotypic plasticity in the intertidal sculpins  Heterogeneous environments, especially if the environmental variation is temporal, are considered to be conducive to the evolution of phenotypic plasticity (Moran 1992), suggesting that organisms inhabiting the intertidal zone may evolve plasticity to combat the frequent and severe changes in stresses such as hypoxia. However, in the case of the hypoxia-tolerant tidepool sculpin transcriptionally-induced phenotypic plasticity does not occur during the ecologically relevant time-frame of 6 to 8 hours and transcriptional changes occur only if hypoxia persists for longer than 24 hours (Chapter 3). While it is possible that the phenotype is altered via other mechanisms, e.g. post-translational modification, it is also possible that the tidepool sculpin is a resilient generalist and expresses a single phenotype that is adequate across alternative environments. This phenotype includes a low routine metabolic rate, high Hb-O2 binding and high maximal activity of brain glycolytic enzymes (Table 5.1). These constitutive or fixed traits (traits expressed in the absence of hypoxia) contribute to hypoxic survival and have likely evolved in response to hypoxic pressure. While there may be cost associated with maintaining these traits across the different intertidal conditions, such as the trade-off associated with high O2 extraction and delivery during hypoxic versus hyperoxic exposures, it may be less than the cost associated with phenotypic plasticity. Factors contributing to cost of phenotypic plasticity include energetic cost of alternating phenotype in changing environmental conditions or the cost that comes from a mismatch between phenotype and environment as a result of lag time between cue for environmental change and production of new phenotype. This would favor the  97 evolution of fixed traits over plastic traits in the tidepool sculpin, suggesting that the evolution of fixed traits may be the primary mechanism of survival during short exposures to hypoxia in the tidepool sculpin. However, if hypoxic conditions persist, phenotypic plasticity may play a role in combating low O2 stress.  The speciation trajectory of sculpins has been thought to be from the deeper subtidal waters to the intertidal zone, suggesting that the deeper subtidal species represent an ancestral state and the intertidal species a derived state (Ramon and Knope 2008). In contrast to the tidepool sculpin, the deeper subtidal silverspotted sculpin (Chapter 3) and the shallow subtidal smoothhead sculpin, sailfin sculpin and Pacific staghorn sculpin (Chapter 4) alter transcription patterns within the initial 8 hours of hypoxia exposure, suggesting that phenotypic plasticity occurs in these species in response to short-term hypoxia. This would raise the possibility that the ancestor to the tidepool sculpin, which likely inhabited the subtidal, also exhibited phenotypic changes in the initial stages of hypoxia. If this was the case, then how and why do the present-day tidepool sculpins lack short-term phenotypic plasticity in response to hypoxia? Plasticity is considered an important first step in establishment of a species in new environment and may lead to the second step of canalization of the plastic traits into non-plastic constitutive or fixed traits, termed genetic assimilation (reviewed in Ghalambor et al. 2007). It is therefore possible that phenotypic plasticity allowed the ancestral intertidal sculpin to initially colonize the new environment and over time, perhaps as a result of high cost of maintenance of plastic traits, there was a genetic assimilation of the environmentally-induced phenotypic changes. While this is very speculative and there is very little evidence to support this, it is interesting to note that most of the subtidal and deeper water species, unlike the tidepool sculpin, showed an increase in transcription of glycolytic genes during short-term exposure to hypoxia (see Table 5.1). However, maximal activity of glycolytic enzymes (considered a good proxy for enzyme concentration) were found to be higher in the liver of the tidepool sculpin than in the other subtidal and deeper water species (Table 5.1), suggesting a tentative link that higher concentrations of glycolytic enzymes may be a result of  98 genetic assimilation from environmentally-induced phenotypes to constitutively expressed traits.     5.1.5 Convergent evolution of hypoxia tolerance  Species that inhabit the high subtidal occasionally experience periods of low O2  and these species have evolved similar levels of hypoxia tolerance. Hypoxia tolerance is a complex phenotype comprised of many interrelated traits and in Chapter 4 I investigated whether the underlying transcriptional response was similar among species with similar levels of hypoxia tolerance including the smoothhead sculpin, sailfin sculpin and Pacific staghorn sculpin. Transcription patterns of many of the genes associated with energy metabolism and protein production differed among the three sculpin species (Chapter 4). There was also a difference in the maximal activities of enzymes associated with energy metabolism in the brain and liver, where the enzyme activities of the Pacific staghorn sculpin reflected those of the hypoxia-tolerant tidepool sculpin, while the enzyme activities of the smoothhead sculpin were more closely aligned with the hypoxia-intolerant silverspotted sculpin (Chapter 2, Table 5.1). Together, this would suggest that among the closely related sculpin species convergence of hypoxia tolerance phenotype was likely achieved via multiple evolutionary paths, possibly indicating lower genetic constraint. These species may become increasingly exposed to hypoxic conditions, which are predicted to become longer and more frequent in the coastal waters (Diaz and Breitburg 2009). Lower genetic constraint associated with genes deemed important in hypoxia tolerance may be beneficial as it may allow greater evolutionary flexibility for the adaptation of the species to the increasing hypoxic conditions.  5.2 Future Directions  In my dissertation, I studied traits underlying hypoxia tolerance in a number of wild-caught species of sculpin. Examining species from the wild can be advantageous as it provides ecological context that can better inform the mechanistic and evolutionary study of how animals function. However, the  99 drawback of not using laboratory-reared animals is the unknown contribution of developmental plasticity and transgenerational epigenetic effects to the variation in traits in sculpins. Therefore, to understand if the differences in traits are genetically-based, experiments will need be performed on individuals that have been reared in the laboratory under common garden conditions. While this may disentangle the effect of developmental plasticity, epigenetic effects can persist for many generations. Therefore one avenue of research is the use of bioinformatics tools, such as methylation arrays, methylation sequencing or chromatin immune-precipitation followed by sequencing (Ritchie et al. 2015), to examine how epigenetics contributes to the complex phenotype of hypoxia tolerance in sculpins.   In this thesis I examined molecular response to environmental hypoxia in a number of species using transcriptomics, a powerful tool that can provide information on potential mechanisms underlying complex organismal responses to the environment. A major benefit of this approach is that it is not limited to only studying traits that have previously been shown to be important in hypoxia tolerance. Instead transcriptomics allows for a broad examination of changes in many biological processes, which can lead to the uncovering of other potentially important mechanisms. However, gene transcription is not the only determinant of phenotype and as the relationship between transcription and function is complex, further examination is required. One set of approaches to providing connection between gene transcription and phenotype is to study the subsequent steps of the pathway leading from gene to phenotype, such as proteomes and metabolomes or by using techniques to manipulate transcription, such as transgenics and RNAi, to determine the effect on phenotype (Alvarez et al. 2015). Another avenue of research is to examine trans-regulatory elements, such as transcription factors, that affect gene transcription and may explain the variation in patterns and ultimately phenotype. I am currently looking at the hypoxia inducible factor (HIF), an important transcription factor regulating transcription of many genes in hypoxia and prolyl hydroxylase (PHD), an important regulator of HIF (Semenza and Wang 1992, Semenza 1998, Bruick and McKnight 2001, Ivan et al. 2001, Jaakkola et al. 2001), in a number of sculpin species to determine if there is variation in these  100 trans-regulatory elements that may explain, at least in part, the variation in gene transcription patterns in sculpins. Overall, I have listed a few of many more approaches that can be taken to further explore how variation in hypoxia tolerance evolved in the closely-related species of sculpin.     5.3 Concluding Thoughts  My doctoral dissertation as well as previous research have focused on examining how species that experience different environmental O2 profiles have evolved to cope with hypoxic stress. I have assessed hypoxia tolerance in wild species of fish and correlated morphological, physiological and biochemical traits using phylogenetically-corrected methods in order to determine candidate traits that may be adaptations to hypoxia in sculpins. I have also examined the role of transcriptionally-based phenotypic plasticity in hypoxic response of the intertidal and subtidal fishes and found complex transcriptional patterns with a strong effect of exposure time to environmental challenge. Diverse responses of species to hypoxia pose a significant challenge in predicting the effect of anthropogenic-induced increases in hypoxic events along the coastal waterways. This task can be accomplished, however, by utilizing a multidisciplinary, integrative approach to examine organismal stress response to environmental challenges in the context of physiology, evolution and ecology.           101  Table 5.1  Summary of traits examined in 5 species. Information taken from Mandic et al. 2009a, 2009b, Chapter 2, Chapter 3 and Chapter 4. Red represents the hypoxia-tolerant species, blue the intermediate-tolerant species and green hypoxia-intolerant species. The unit for maximal enzyme activity is μmol/min/g	  wet	  tissue	  and	  the	  unit	  for	  metabolite	  concentration	  is	  μmol/g	  wet	  tissue.	  Up	  arrows	  represent	  increase	  in	  transcript	  levels,	  down	  arrows	  represent decrease in transcript levels, dashes represent no change in transcription and ‘v’	  represents	  a	  variable	  response.	  Abbreviations in the table: Pcrit (critical oxygen tension), LOE50 (loss of equilibrium in 50% of the individuals), ASR Freq. (aquatic surface respiration frequency), Air Emerge Freq. (air emergence frequency), Gill S.A. (gill surface area), Hb-O2 (hemoglobin-oxygen), PK (pyruvate kinase), LDH (lactate dehydrogenase), CPK (creatine phosphokinase), CS (citrate synthase), ATP (adenosine triphosphate), CrP (creatine phosphokinase), OxPhos (oxidative phosphorylation), FAO (fatty acid oxidation), N/O (not observed) and N/A (data not available).   102     103 Bibliography Almeida-Val V.M.F., A.L. Val, W.P. Duncan, F.C.A. Souza, M.N. Paula-Silva, and S. Land.  2000. Scaling effects on hypoxia tolerance in the Amazon fish Astronotus ocellatus (Perciformes: Cichlidae): contribution of tissue enzyme levels. Comparative Biochemistry and Physiology B 125: 219-225   Altieri A.H. 2006 Inducible variation in hypoxia tolerance across the intertidal- subtidal distribution of the blue mussel Mytilus edulis. Marine Ecology Progress Series 325: 295-300  Alvarez M., A.W. Schrey and C.L. Richards. 2015. Ten years of transcriptomics in wild  populations: what have we learned about their ecology and evolution? Molecular Ecology  doi: 10.1111/mec.13055  Andersen O., O.F. Wetten, M.C. De Rosa, C. Andrew, C.C. Alinovi, M. Colafranseschi,  O. Brix and A. Colosimo. 2009. Haemoglobin polymorphisms affect the oxygen-binding properties in Atlantic cod populations. Proceedings of the Royal Society B 276: 833-841    Aoe T., H. Inaba, S. Kon, M. Imai, M. Aono, T. Mizuguchi, T. Saito and T. Nishino.  1997. Heat shock protein 70 messanger RNA reflects the severity of ischemia/hypoxia-reperfusion injury in the perfused rat liver. Critical Care medicine 25: 324-329.  Aubin-Horth N. and S.C.P. Renn. 2009. Genomic reaction norms: using integrative  biology to understand molecular mechanisms of phenotypic plasticity. Molecular Ecology 18: 3763-3780  Ben-Tabou de-Leon S. and E.H. Davidson. 2009. Modeling the dynamics of   104 transcriptional gene regulatory networks for animal development. Developmental Biology 325: 317-328  Bergmeyer H.U. 1983. Methods of Enzymatic Analysis. Academic Press, New York.  Bernatchez L., S Renaut, A.R. Whiteley, N. Derome, J. Juekens, L. Landry, G. Lu, A.W.  Nolte, K. Ostbye, S.M. Rogers and J. St-Cyr. 2010. On the origin of species: insights from ecological genomics of lake whitefish. Philosophical Transactions of the Royal Society B 365: 1783-1800  Berrigan D. and S.M. Scheiner. 2004. Modeling the evolution of phenotypic plasticity.  In: Phenotypic Plasticity (eds. DeWitt TJ, Scheiner SM), pp. 82-97. Oxford University Press, New York.    Bickler P.E. and L.T. Buck. 2007. Hypoxia tolerance in reptiles, amphibians, and  fishes: life with variable oxygen availability. Annual Review of Physiology 69: 145-170   Bijlsma R. and V. Loeschcke. 2005. Environmental stress, adaptation and evolution:  an overview. Journal of Evolutionary Biology 18: 744-749  Bolanos J.P., A. Almeida, and S. Moncada. 2009. Glycolysis: a bioenergetics or a  survival pathway? Trends in Biochemal Sciences 35: 145-149  Boswell M.G., M.C. Wells, L.M. Kirk, Z. Ju, Z. Zhang, R.E. Booth and R.B. Walter.  2009. Comparison of gene expression responses to hypoxia in viviparous (Xiphophorus) and oviparous (Oryzias) fishes using a medaka microarray. Comparative Biochemistry and Physiology C 149: 258-265  Botero C.A., F.J. Weissing, J. Wright and D.R. Rubenstein. 2015. Evolutionary tipping   105 points in the capacity to adapt to environmental change. Proceedings of the National Academy of Sciences of the United States of America 121: 184-189  Boutilier R.G. 2001. Mechanisms of cell survivial in hypoxia and hypothermia.  The Journal of Experimental Biology 204: 3171-3181   Boutilier R.G., G. Dobson, U. Hoeger, and D.J. Randall. 1988. Acute exposure to  graded levels of hypoxia in rainbow trout (Salmo gairdneri): metabolic and respiratory adapatations. Respiration Physiology 71: 69-82  Boutilier R.G. and J. St-Pierre. 2000. Surviving hypoxia without really diving.  Comparative Biochemistry and Physiology A 126: 481-490  Bradford M.M. 1976. A rapid and sensitive method for the quantitation of  microgram quantities of protein utilizing the principle of protein-dye binding. Analytical Biochemistry 72: 248-254  Brix O., K.D. Clements and R.M.G. Wells. 1999. Haemoglobin components and oxygen  transport in relation to habitat selection in triplefin fishes (Tripterygiidae). Journal of Comparative Physiology B 169: 329-334   Brooks S.P.J. and K.B. Storey. 1995. Is glycolytic rate controlled by the reversible  binding of enzymes to subcellular structures? In: Biochemistry and Molecular Biology of Fishes. Vol. 4. (eds. P.W. Hochachka and T.P. Mommsen), pp. 1-23 Elsevier Academic Press, Amsterdam.   Bruick R.K. and S.L. McKnight. 2001. A conserved family of prolyl-4-hydroxylases  that modify HIF. Science 294: 1337-1340  Buckley B.A. 2007. Comparative environmental genomics in non-model species:   106 using heterologous hybridization to DNA-based microarrays. The Journal of Experimental Biology 209: 1602-1606  Burggren W.W. 2014. Epigenetics as a source of variation in comparative animal  physiology – or – Lamark	  is	  lookin’	  pretty	  good	  these	  days.	  The	  Journal	  of	  Experimental Biology 217: 682-689  Burggren, W. and Roberts, J. 1991. Respiration and metabolism. In Environmental  and metabolic animal physiology: comparative animal physiology, 4th edition. (ed. C. L. Prosser) pp. 353-435. Wiley-Liss, New York.   Callahan, H.S., H. Maughan and U.K. Steiner. 2008. Phenotypic plasticity, costs of  phenotypes, and costs of plasticity: toward an integrative view. Annals of the  New York Academy of Sciences 1133: 44-66  Chapman L.J., C.A. Chapman, D.A. Brazeau, B. McLaughlin and M. Jordan. 1999.  Papyrus swamps, hypoxia, and faunal diversification: variation among populations of Barbus neumayeri. Journal of Fish Biology 54: 310-327  Chapman L.J., C.A. Chapman, F.G. Nordlie, and A.E. Rosenberger. 2002. Physiological  refugia: swamps, hypoxia tolerance and maintenance of fish diversity in the Lake Victoria region. Comparative Biochemistry and Physiology A 133: 421-437  Chapman L.J. and D.J. McKenzie. 2009. Behavioral responses and ecological  consequences. In: Hypoxia (eds. J.G. Richards, A.P. Farrell and C.J. Brauner), pp.26-77. Elsevier Academic Press, San Diego.  Chelaifa H., F. Mahe and M. Ainouche. 2010. Transcriptome divergence between the  hexaploid salt-marsh sister species Spartina maritime and Spartina alterniflora (Poaceae). Molecular Ecology 19: 2050-2063   107  Chen K., R.B. Cole and B.B. Rees. 2013. Hypoxia-induced changes in the zebrafish  (Danio rerio) skeletal muscle proteome. Journal of Proteomics 78: 477-485  Chevin L.-M.,	  G.	  Martin	  and	  T.	  Lenormand.	  2010.	  Fisher’s	  model	  and	  the	  genomics	  of	   adaptation: restricted pleiotropy, heterogenous mutation, and parallel evolution. Evolution 64: 3213-3231  Chippari-Gomes A.R., L.C. Gomes, N.P. Lopes, A.L. Val, and V.M.F. Almeida-Val. 2005.  Metabolic adjustments in two Amazonian cichlids exposed to hypoxia and anoxia. Comparative Biochemistry and Physiology B 141: 347-355  Christin P.-A., D.M. Weinreich and G. Besnard. 2010. Causes and evolutionary  significance of genetic convergence. Trends in Genetics 26: 400-405  Churchill T.A., K.M. Cheetham, S. Simpkin, C.J. Green, L.C.H. Wang and B.J. Fuller.  1994. Liver metabolism in cold hypoxia: a comparison of energy metabolism and glycolysis in cold-sensitive and cold-resistant mammals. Journal of Comparative Physiology Part D 164: 396-404  Conte G.L., M.E. Arnegard, C.L. Peichel and D. Schluter. 2012. The probability of  genetic parallelism and convergence in natural populations. Proceedings of the Royal Society B 279: 5039-5047  Cossins A.R. and D.L. Crawford. 2005. Fish as models for environmental genomics.  Nature Reviews Genetics 6: 324-340  Crawford D.L. and D.A. Powers. 1989. Molecular basis of evolutionary adaptation at  the lactate dehydrogenase-B locus in the fish Fundulus heteroclitus. Proceedings of the National Academy of Sciences of the United States of America 86: 9365-9369  108  Dabney A., J.D. Storey and with assistance from G.R. Warnes. 2013. qvalue: Q-value  estimation for false discovery rate control. R package version 1.34.0.  Dalziel A.C., S.M. Rogers and P.M. Schulte. 2009. Linking genotypes to phenotypes  and fitness: how mechanistic biology can inform molecular ecology. Molecular Ecology 18: 4997-5017  Darveau C.A., P.W. Hochachka, D.W. Roubik, and R.K. Suarez. 2005a. Allometric  scaling of flight energetics in orchid bees: evolution of flux capacities and flux rates. The Journal of Experimental Biology 208: 3593-3602  Darveau C.A., P.W. Hochachka, K.C. Welch Jr., D.W. Roubik, and R.K. Suarez. 2005b.  Allometric scaling of flight energetics in Panamanian orchid bees: a comparative phylogenetic approach. The Journal of Experimental Biology 208: 3581-3591  Davies R. and C.D. Moyes. 2007. Allometric scaling in centrarchid fish: origins of  intra- and inter-specific variation in oxidative and glycolytic enzyme levels in muscle. Journal of Experimental Biology 210: 3798-3804  Dean T.A., L. Haldorson, D.R. Laur, S.C. Jewett and A. Blanchard. 2000. The  distribution of nearshore fishes in kelp and eelgrass communities in Prince William Sound, Alaska: associations with vegetation and physical habitat characteristics. Environmental Biology of Fishes 57: 271-287  Della Via J., G. Van den Thillart, O. Cattani and P. Cortesi. 1998. Behavioural  responses and biochemical correlates in Solea solea to gradual hypoxia exposures. Canadian Journal of Zoology 76: 2108-2113  Derome N., P. Duchesne and L. Bernatchez. 2006. Parallelism in gene transcription   109 among sympatric lake whitefish (Coregonus clupeaformis Mitchill) ecotypes. Molecular Ecology 15: 1239-1249  Dewitt T.J. and R.B. Langerhans. 2004. Integrated solutions to environmental  heterogeneity: theory of multimoment reaction norms. In: Phenotypic Plasticity (eds. DeWitt TJ, Scheiner SM), pp. 98-111. Oxford University Press, New York.    DeWitt T.J., A. Sih and D.S. Wilson. 1998. Costs and limits of phenotypic plasticity.  Trends in Ecology and Evolution 13: 77-81   DiAngelo C.R. and A.G. Heath. 1987. Comparison of in vivo energy metabolism in the  brain of rainbow trout, Salmo gairdneri and bullhead catfish, Ictalurus nebulosus during anoxia. Comparative Biochemistry and Physiology B 88: 297-303   Diaz, R.J. 2008. Spreading dead zones and consequences for marine ecosystems.  Science 321: 926-929.  Diaz R.J. and D.L. Breitburg. 2009. The hypoxic environment. In: Hypoxia (eds. J.G.  Richards, A.P. Farrell and C.J. Brauner), pp.1-23. Elsevier Academic Press, San  Diego.    Efird T.P. and B. Konar. 2013. Habitat characteristics can influence fish assemblages  in high latitude kelp forests. Environmental Biology of Fishes 97: 1253-1263  Endler J.A. 1986. Natural selection in the wild. Princeton University Press, Princeton.  Erecinska M. and I.A. Silver. 2001. Tissue oxygen tension and brain sensitivity to  hypoxia. Respiration Physiology 128: 263-276   110 Everett M.V., C.E. Antal and D.L. Crawford. 2011. The effect of short-term hypoxic  exposure on metabolic gene expression. Journal of Experimental Zoology 313A: 1-15  Farrell A.P. 2007. Tribute to Peter Lutz: a message from the heart – why hypoxic  bradycardia in fishes? The Journal of Experimental Biology 210: 1715-1725  Farwell M., M.G. Fox, C.D. Moyes, and G. Burness. 2007. Can hypoxia tolerance  explain differences in distribution of two co-occurring north temperate sunfishes? Environmental Biology of Fishes 78: 83-90  Feder M.E., A.F. Bennett and R.B. Huey. 2000. Evolutionary physiology. Annual  Review of Ecology, Evolution, and Systemics 31: 315-341   Felsenstein J. 1985. Phylogenies and the comparative methods. The American  Naturalist 125: 1-15  FitzJohn R.G. 2012. Diversitree: comparative phylogenetics analyses of  diversification in R. Methods in Ecology and Evolution 3: 1084-1092  Forne I., J. Abian and J. Cerda. 2010. Fish proteome analysis: model organisms and  non-sequenced species. Proteomics 10: 858-872  Froese R. and D. Pauly. 2013. FishBase. www.fishbase.org. Data retrieved from the  internet February 2013  Gabriel W. 2005. How stress selects for reversible phenotypic plasticity. Journal of  Evolutionary Biology. 18: 873-883  Gabriel W. 2006. Selective advantage of irreversible and reversible phenotypic  plasticity. Archiv fur Hydrobiologie 167: 1-20  111  Gamperl and Driedzic 2009. Cardiovascular function and cardiac metabolism. In:  Hypoxia (eds. J.G. Richards, A.P. Farrell and C.J. Brauner), pp.302-360. Elsevier Academic Press, San Diego.  Garland T.Jr., A.F. Bennett, and E.L. Rezende. 2005. Phylogenetic approaches in  comparative physiology. The Journal of Experimental Biology 208: 3015- 3035  Garland T.Jr., P.H. Harvey and A.R. Ives. 1992. Procedures for the analysis of  comparative data using phylogenetically independent contrasts. Systematic Biology 41: 18-32  Gaston K.J., S.L. Chown, P. Calosi, J. Bernardo, D.T. Bilton, A. Clarke, S. Clusella- Trullas, C.K. Ghalambor, M. Konarzewski, L.S. Peck, W.P. Porter, H.O. Portner, E.L. Rezende, P.M. Schulte, J.I. Spicer, J.H. Stillman, J.S. Terblanche and M. van Kluenen. 2009. Macrophysiology: a conceptual reunification. The American Naturalist 174: 595-612  Ghalambor C.K., J.K. McKay, S.P Carroll and D.N. Reznick. 2007. Adaptive versus non- adaptive phenotypic plasticity and the potential for contemporary adaptation  in new environments. Functional Ecology 21: 394-407  Gracey A.Y., J.V. Troll and G.N. Somero. 2001. Hypoxia-induced gene expression  profiling in the euryoxic fish Gillichthys mirabilis. Proceedings of the National Academy of Sciences of the United States of America 98: 1993-1998  Graham J.B. 1990. Ecological, evolutionary, and physical factors influencing aquatic  animal respiration. American Zoologist 30: 137-146  Graham J.B. 1997. Air-breathing fishes: evolution, diversity, and adaptation.   112 Academic Press, San Diego.   Grant V. 1991. The evolutionary process: a critical study of evolutionary theory.  Columbia University Press, New York  Green J.M. 1971 Local distribution of Oligocottus maculosus Girard and other  tidepool cottids of the west coast of Vancouver Island, British Columbia.  Canadian Journal of Zoology 49: 1111-1128  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 Genetics 5: e1000326  Hagerman L. 1998. Physiological flexibility; a necessity for life in anoxic and  sulphidic habitats. Hydrobiologia 375/376: 241-254  Henriksson P., M. Mandic and J.G. Richards. 2008. The osmorespiratory compromise  in sculpins: impaired gas exchange is associated with freshwater tolerance. Physiological and Biochemical Zoology 81: 310-319  Ho, D.H. and W.W. Burggren. 2012. Parental hypoxic exposure confers offspring  hypoxia resistance in zebrafish (Danio rerio). The Journal of Experimental Biology. 215: 4208-4216  Hochachka P.W. 1986. Defense strategies against hypoxia and hypothermia. Science  231: 234-241  Hochachka P.W., L.T. Buck, C.J. Doll and S.C. Land. 1996. Unifying theory of hypoxia  tolerance: molecular/metabolic defense and rescue mechanisms for surviving oxygen lack. Proceedings of the National Academy of Sciences of the United States of America 93: 9493-9498   113  Hochachka, P.W and G.N. Somero. 2002. Biochemical adaptation: mechanism and   process in physiological evolution. Oxford University Press, New York.   Hoekstra H.E., R.J. Hirschmann, R.A. Bundey, P.A. Insel and J.P. Crossland. 2006. A  single amino acid mutation contributes to adaptive beach mouse color pattern. Science 313: 101-104  Huang X. and A. Madan. 1999. CAP3: A DNA sequence assembly program. Genome  Research 9: 868-877  Huelsenbeck J. P. and F. Ronquist. 2003. MrBayes 3: Bayesian phylogenetics  inference under mixed models. Bioinformatics 19: 1572-1574  Hyvarinen H., I.J. Holopainen, and J. Piironen. 1985. Anaerobic wintering of crucian  carp (Carassius carassius L.) – I. Annual dynamics of glycogen reserves in nature. Comparative Biochemistry and Physiology A 82: 797-803  Ivan M., K. Kondo, H. Yang, W. Kim, J. Valiando, M. Ohh, A. Salic, J.M. Asara, W.S. Lane  and W.G.Jr. Kaelin. 2001. HIF-α targeted for VHL-mediated destruction by proline hydroxylation: implications for O2 sensing. Science 292: 464-468  Jaakkola P., D.R. Mole, Y. Tian, M.I. Wilson, J. Gielbert, S.J. Gaskell, A. von Kriegsheim,  H.F. Hebestreit, M. Mukherji, C.J. Schofield, P.H. Maxwell, C.W. Pugh and P.J. Ratcliffe. 2001. Targeting of HIF-α to the von Hippel-Lindau ubiquitylation complex by O2-regulated prolyl hydroxylation. Science 292: 468-472  Jessen T.-H., R.E. Weber, G. Fermi, J. Tame and G. Braunitzer. 1991. Adaptation of  bird hemoglobins to high altitudes: demonstration of molecular mechanism by protein engineering. Proceedings of the National Academy of Sciences of the United States of America 88: 6519-6522   114  Johansson D., G.E. Nilsson, and E. Tornblom. 1995. Effects of anoxia on energy  metabolism in crucian carp brain slices studied with microcalorimetry. The Journal of Experimental Biology 198: 853-859   Ju Z., M.C. Wells, S.J. Heater and R.B. Walter. 2007. Multiple tissue gene expression  analyses in Japanese medaka (Oryzias latipes) exposed to hypoxia. Comparative Biochemistry and Physiology Part C 145: 134-144   Kassahn K.S. 2008. Microarrays for comparative and ecological genomics: beyond  single-species applications of array technologies. Journal of Fish Biology 72: 2407-2434   Kelly D.A. and K.B. Storey. 1988. Organ-specific control of glycolysis in anoxic  turtles. American Journal of Physiology: Regulatory, Integrative and Comparative Physiology 255: R774-R779   Kerr M.K. and G.A. Churchill. 2001. Experimental design for gene expression  microarrays. Biostatistics 2: 183-201  Knox D., M.J. Walton, and C.B. Cowey. 1980. Distribution of enzymes of glycolysis  and gluconeogenesis in fish tissues. Marine Biology 56: 7-10  Kocha K.M., K. Reilly, D.S.M. Porplycia, J. McDonald, T. Snider and C.D. Moyes. 2014.  Evolution of the oxygen sensitivity of cytochrome c oxidase subunit 4. American Journal of Physiology: Regulatory, Integrative and Comparative Physiology DOI: 10.1152/ajpregu.00281.2014  Koumenis C., C. Naczki, M. Koritzinsky, S. Rastani, A. Diehl, N. Sonenberg, A.  Koromilas and B.G. Wouters. 2002. Regulation of protein synthesis by hypoxia via activation of the endoplasmic reticulum kinase PERK and  115 phosphorylation of the translation initiation factor eIF2a. Molecular and Cellular Biology 22: 7405-7416  Krogh A. 1929. The progress of physiology. American Journal of Physiology 90: 242- 251   Kutty M.N. 1972. Respiratory quotient and ammonia excretion in Tilapia  mossambica. Marine Biology 16: 126-133   Land S.C., L.T. Buck and P.W. Hochachka. 1993. Response of protein synthesis to  anoxia and recovery in anoxia-tolerant hepatocytes. American Journal of Physiology – Regulatory, Integrative and Comparative Physiology 34: R41-R48   Leary S.C., C.N. Lyons, A.G. Rosenberger, J.S. Ballantyne, J. Stillman, and C.D. Moyes.  2003. Fiber-type differences in muscle mitochondrial profiles. American Journal of Physiology – Regulatory, Integrative and Comparative Physiology 285: R817-R826   Lefrancois C. and P. Domenici. 2006. Locomotor kinematics and behavior in the  escape response of European sea bass, Dicentrarchus labrax L., exposed to hypoxia. Marine Biology 149: 969-977  Leveelahti L., P. Leskinen, E.H. Leder, W. Waser and M. Nikinmaa. 2011. Responses  of threespine stickleback (Gasterosteus aculeatus, L) transcriptome to hypoxia. Comparative Biochemistry and Physiology Part D 6: 370-381    Lewis J.M., I. Costa, A.L. Val, V.M. Almeida-Val, A.K. Gamperl and W.R. Driedzic. 2007.  Responses to hypoxia and recovery: repayment of oxygen debt is not associated with compensatory protein synthesis in the Amazonian cichlid, Astronotus ocellatus. The Journal of Experimental Biology 210: 1935-1943  116  Li F., W. Luan, C. Zhang, J. Zhang, B. Wang, Y. Xie, S. Li and J. Xiang. 2009. Cloning  of cytoplasmic heat shock protein 90 (FcHSP90) from Fenneropenaeus chinensis and its expression response to heat shock and hypoxia. Cell Stress and Chaperones 14: 161-172  Lin G., Y. Liu, and K.M. MacLeod. 2009. Regulation of muscle creatine kinase by  phosphorylation in normal and diabetic hearts. Cellular and Molecular Life Sciences 66: 135-144  Losos J.B. 2011. Convergence, adaptation, and constraint. Evolution 65: 1827-1840.  Lushchak V.I., T.V. Bahnjukova, and K.B. Storey. 1998. Effect of hypoxia on the  activity and binding of glycolytic and associated enzymes in sea scorpion tissues. Brazilian Journal of Medical and Biological Research 31: 1059-1067  Lutz P.L. 1992. Mechanisms for anoxic survival in the vertebrate brain. Annual  Review of Physiology 54: 601-618  Lutz P.L. and G.E. Nilsson. 1997. Contrasting strategies for anoxic brain survival –  glycolysis up or down. The Journal of Experimental Biology 200: 411-419  Lutz P.L., G.E. Nilsson, and H.M. Prentice. 2003. The brain without oxygen: causes of  failure-physiological and molecular mechanisms for survival. Kluwer Academic Publishers, Dordrecht.  Machado H.E., A.A Pollen, H.A. Hofmann and S.C.P Renn. 2009. Interspecific profiling  of gene expression informed by comparative genomic hybridization: a review and a novel approach in African cichlid fishes. Integrative and Comparative Biology 49: 644-659   117 Maddison W. P. and D.R. Maddison. 2004. Mesquite: A modular system for  evolutionary analysis. Version 1.02. http://mesquiteproject.org. Data  retrieved from internet August 2009  Mandic M., K.A. Sloman and J.G. Richards. 2009a. Escaping to the surface: a  phylogenetically independent analysis of hypoxia-induced respiratory behaviours in sculpins. Physiological and Biochemical Zoology 82: 730-738  Mandic M., A.E. Todgham, and J.G. Richards. 2009b. Mechanisms and evolution of  hypoxia tolerance in fish. Proceedings of the Royal Society B 276: 735-744  Marliave J.B. 1997. Substratum preferences of settling larvae of marine fishes reared  in the laboratory. Journal of Experimental Marine Biology and Ecology 27: 47-60  Martin K.L.M. 1995. Time and tide wait for no fish: intertidal fishes out of water.  Environmental Biology of Fishes 44: 165-181  Martinez M.L., C. Landry, R. Boehm, S. Manning, A.O. Cheek, and B.B. Rees. 2006.  Effects of long-term hypoxia on enzymes of carbohydrate metabolism in the Gulf killifish, Fundulus grandis. The Journal of Experimental Biology 209: 3851-3861   Martinez, M.L., E.L. Raynard, B.B. Rees, and L.J. Chapman. 2011. Oxygen limitation  and tissue metabolic potential of the African fish Barbus neumayeri: roles of native habitat and acclimatization. BMC Ecology 11: 2-9  Martins E.P. and E.A. Housworth. 2002. Phylogeny shape and the phylogenetic  comparative method. Systematic Biology 51: 873-880  Medale F., J.P. Parent and F. Vellas. 1987. Responses to prolonged hypoxia by   118 rainbow trout (Salmo gairdneri) I. Free amino acids and protein in plasma, liver and white muscle. Fish Physiology and Biochemistry 3: 183-189.  Midford P.E., T.Jr. Garland, and W.P. Maddison. 2005. PDAP package of Mesquite.  Version 1.07 http://mesquiteproject.org/pdap_mesquite/index.html. Data  retrieved from the internet June 2009  Moran N.A. 1992. The evolutionary maintenance of alternative phenotypes. The  American Naturalist 139: 971-989  Mykles D.L., C.K. Ghalambor, J.H. Stillman and L. Tomanek. 2012. Grand challenges in  comparative physiology: integration across disciplines and across levels of biological organization. Integrative and Comparative Biology 50: 6-16  Nelson J. 2006. Fishes of the world 4th edition. John Wiley and Sons, Hoboken, New  Jersey   Newsholme E.A. and B. Crabtree. 1986. Maximum catalytic activity of some key  enzymes in provision of physiologically useful information about metabolic fluxes. The Journal of Experimental Biology 239: 159-167    Nikinmaa M. 2002 Oxygen-dependent cellular functions – why fishes and their  aquatic environment are a prime choice of study. Comparative Biochemistry and Physiology A 133: 1-16    Nilsson G.E. 1996. Brain and body oxygen requirements of Gnathonemus petersii, a  fish with an exceptionally large brain. The Journal of Experimental Biology 199: 603-607  Nilsson G.E. 2001. Surviving anoxia with the brain turned on. Physiology 16: 217- 221  119  Nilsson G.E. 2007. Gill remodeling in fish – a new fashion or an ancient secret? The  Journal of Experimental Biology 210: 2403-2409  Nilsson G.E., P. Hylland, and C.O. Lofman. 1994. Anoxia and adenosine induce  increased cerebral blood flow rate in crucian carp. American Journal of  Physiology – Regulatory, Integrative and Comparative Physiology 36: R590-595   Nilsson G.E., P. Rosen, and D. Johansson. 1993. Anoxic depression of spontaneous  locomotor activity in crucian carp quantified by a computerized imaging technique. The Journal of Experimental Biology 180: 153-162  Nonnotte G., V. Maxime, J.P. Truchot, P. Williot and C. Peyraud. 1993. Respiratory  responses to progressive ambient hypoxia in the sturgeon, Acipenser baeri. Respiration Physiology 91: 71-82    Oliveros J.C. 2007. VENNY. An interactive tool for comparing lists with Venn  Diagrams. http://bioinfogp.cnb.csic.es/tools/venny/index.html. Data  retrieved from internet July 2011  Onukwufor J.O., N. MacDonald, F. Kibenge, D. Stevens and C. Kamunde. 2014. Effects  of hypoxia-cadmium interactions on rainbow trout (Oncorhynchus mykiss) mitochondrial bioenergetics: attenuation of hypoxia-induced proton leak by low doses of cadmium. The Journal of Experimental Biology 217: 831-840    O’Quin	  K.E.,	  C.M.	  Hofmann,	  H.A.	  Hofmann	  and	  K.L.	  Carleton.	  2010.	  Parallel	  evolution	   of opsin gene expression in African cichlid fishes. Molecular Biology and Evolution 27: 2839-2854   Padilla D.K. and S.C. Adolph. 1996. Plastic inducible morphologies are not always   120 adaptive: the importance of time delays in a stochastic environment. Evolutionary Ecology 10: 105-117     Perry S.F., M.G. Jonz and K.M. Gilmour. 2009. Oxygen sensing and the hypoxic  ventilatory response. In: Hypoxia (eds. J.G. Richards, A.P. Farrell and C.J. Brauner), pp.193-253. Elsevier Academic Press, San Diego.  Perutz M.F. 1983. Species adaptation in a protein molecule. Molecular Biology and  Evolution 1: 1-28  Pfenning D.W., M.A. Wund, E.C. Snell-Rood, T. Cruickshank, C.D. Schlichting and A.P.  Moczek.	  2010.	  Phenotypic	  plasticity’s	  impacts	  on	  diversification	  and	  speciation. Trends in Ecology and Evolution 25: 459-467  Philippi	  T.	  and	  J.	  Seger.	  1989.	  Hedging	  one’s	  evolutionary	  bets,	  revisited.	  Trends	  in	   ecology and evolution 4: 41-44  Pierce V.A. and D.L. Crawford. 1997. Phylogenetic analysis of glycolytic enzyme  expression. Science 276: 256-259  Preedy V.R., D.M. Smith and P. H. Sugden. 1985. The effects of 6 hours of hypoxia on  protein synthesis in rat tissues in vivo and in vitro.  Biochemical Journal 228: 179-185  R Core Team. 2013. R: A language and environment for statistical computing. R  Foundation for Statistical Computing, Vienna, Austria. URL  http://www.R-project.org/. Data retrieved from the internet October 2010  Ramon M.L. and M.J. Knope. 2008. Molecular support for marine sculpin (Cottidae;   121 Oligocottinae) diversification during the transition from the subtidal to intertidal habitat in the Northeaster Pacific Ocean. Molecular Phylogenetics and Evolution 46: 475-483   Renn S.C.P., N. Aubin-Horth and H.A. Hoffman. 2004. Biologically meaningful  expression profiling across species using heterologous hybridization to a cDNA microarray. BMC Genomics 5: 42-55  Richards J.G. 2009. Metabolic and molecular responses of fish to hypoxia. In:  Hypoxia (eds. J.G. Richards, A.P. Farrell and C.J. Brauner), pp.443-485. Elsevier Academic Press, San Diego.  Richards J.G. 2011. Physiological, behavioral and biochemical adaptations of  intertidal fishes. The Journal of Experimental Biology 214: 191-199   Richards J.G., B.A. Sardella and P.M. Schulte. 2008. Regulation of pyruvate  dehydrogenase in the common killifish, Fundulus heteroclitus, during hypoxia exposure. American Journal of Physiology – Regulatory, Integrative and Comparative Physiology 295: R979-R990   Richards J.G., Y.S. Wang, C.J. Brauner, R.J. Gonzalez, M.L. Patrick, P.M. Schulte, A.R.  Choppari-Gomes, V.M. Almeida-Val, and A.L. Val. 2007. Metabolic and ionoregulatory responses of the Amazonian cichlid, Astronotus ocellatus, to severe hypoxia. Journal of Comparative Physiology B 177: 361-374   Riou J.P., T.H. Claus, and S.J. Pilkis. 1976. Control of pyruvate kinase activity by  glucagon in isolated hepatocytes. Biochemical and Biophysical Research Communications 73: 591-599  Ritchie M.D., E.R. Hozinger, R. Li, S.A. Pendergrass and D. Kim. 2015. Methods of   122 integrative data to uncover genotype-phenotype interactions. Nature Reviews Genetics 16: 85-97  Roberge C., S. Einum, H. Guderley and L. Bernatchez. 2006. Rapid parallel  evolutionary changes of transcription profiles in farmed Atlantic salmon. Molecular Ecology 15: 9-20    Robertson, C.E., P.A. Wright, L. Koblitz and N.J. Bernier. 2014. Hypoxia-inducible  factor-1 mediates adaptive developmental plasticity of hypoxia tolerance in zebrafish, Danio rerio. Proceedings of the Royal Society B 281: 1-9  Robinson B.W. and R. Dukas. 1999. The influence of phenotypic modification on  evolution: the Baldwin effect and modern perspective. Oikos 85: 582-589  Rosenblum E.B., H. Rompler, T. Schoneberg and H.E. Hoekstra. 2010. Molecular and  functional basis of phenotypic convergence in white lizards at White Sands. Proceedings of the National Academy of Sciences of the United States of America 107: 2113-2117  Routley M.H., G.E. Nilsson and G.M.C. Renshaw. 2002. Exposure to hypoxia primes  the respiratory and metabolic responses of the epaulette shark to  progressive hypoxia. Comparative Biochemistry and Physiology Part A 131: 313-321   Rutjes H.A., M.C. Nieveen, R.E. Weber, F. Witte and G.E.E.J.M. van Thillart. 2007.  Multiple strategies of Lake Victoria cichlids to cope with lifelong hypoxia include hemoglobin switching. American Journal of Physiology – Regulatory, Integrative and Comparative Physiology 293: R1376-R1383   Rytkonen K.T., A. Akbarzadeh, H.K. Miandare, H. Kamei, C. Duan, E.H. Leder, T.A.   123 Williams and M. Nikinmaa. 2012. Subfunctionalization of cyprinid hypoxia-inducible factors for roles in development and oxygen sensing. Evolution 67: 873-882  Rytkonen K.T., H.J. Ryynanen, M. Nikinmaa and C.R. Primmer. 2008. Variable  patterns in the molecular evolution of the hypoxia-inducible factor-1 alpha (HIF-1α)	  gene	  in	  teleost	  fishes	  and	  mammals.	  Gene	  420:	  1-10  Rytkonen K.T., K.A.M. Vuori, C.R. Primmer and M. Nikinmaa. 2007. Comparison of  hypoxia-inducible factor -1 alpha in hypoxia-sensitive and hypoxia-tolerant fish species. Comparative Biochemistry and Physiology D 2: 177-186  Schlichting C.D. and M. Pigliucci. 1993. Control of phenotypic plasticity via  regulatory genes. The American Naturalist 142: 366-370   Schlichting C.D. and H. Smith. 2002. Phenotypic plasticity: linking molecular  mechanisms with evolutionary outcomes. Evolutionary Ecology 16: 189-211  Schulte P.M. 2014. What is environmental stress? Insights from fish living in a  variable environment. The Journal of Experimental Biology 217: 23-34  Seibel B.A. 2011. Critical oxygen levels and metabolic suppression in oceanic oxygen  minimum zones. The Journal of Experimental Biology 214: 326-336  Semenza G.L. 1998. Hypoxia-inducible factor 1: master regulator of O2 homeostasis.  Current Opinion in Genetics and Development 8: 588-594  Semenza G.L. and G.L. Wang. 1992. A nuclear factor induced by hypoxia via de novo  protein synthesis binds to the human erythropoietin gene enhancer at a site required for transcriptional activation. Molecular and Cellular Biology 12: 5447-5454  124  Servant N., E. Gravier, P. Gestraud, C. Laurent, C. Paccard, A. Biton, I. Brito, J. Mandel,  B. Asselain, E. Barillot and P. Hupe. 2013. EMA: Easy Microarray data Analysis. R package version 1.4.3. http://CRAN.R-project.org/package=EMA. Data retrieved from internet March 2011  Simpson G.G. 1952. The meaning of evolution. Yale University Press  Sloman K.A., C.M. Wood, G.R. Scott, S. Wood, M. Kajimura, O.E. Johannsson, V.M.F.  Almeida-Val and A.L. Val. 2006. Tribute to R. G. Boutilier: the effect of size on the physiological and behavioural responses of Oscar, Astronotus ocellatus, to hypoxia. Journal of Experimental Biology 209: 1197-1205   Smith W.S., P. Cash, S. Ellefsen and G.E. Nilsson. 2009. Proteomic changes in the  crucian carp brain during exposure to anoxia. Proteomics 9: 2217-2229  Smith R.W., D.F. Houlihan, G.E. Nilsson and J.G. Brechin. 1996. Tissue-specific  changes in protein synthesis rates in vivo during anoxia in crucian carp. American Journal of Physiology – Regulatory, Integrative and Comparative Physiology 271: R897-R904  Sollid J., P. De Angelis, K. Gundersen and G.E. Nilsson. 2003. Hypoxia induces  adaptive and reversible gross morphological changes in crucian carp gills. Journal of Experimental Biology 206: 3667-3673  Somero G.N. 2012. The physiology of global change: linking patterns to mechanisms.  Annual Review of Marine Science 4: 39-61  Somero G.N. and J.J. Childress. 1980. A violation of the metabolism-size scaling  paradigm: activities of glycolytic enzymes in muscle increase in larger-size fish. Physiological Zoology 53: 322-337  125  Speers-Roesch B., J.G. Richards, C.J. Brauner, A.P. Farrell, A.J.R. Hickey, Y.S. Wang  and G.M.C. Renshaw. 2012. Hypoxia tolerance in elasmobranchs. I. Critical oxygen tension as a measure of blood oxygen transport during hypoxia exposure. The Journal of Experimental Biology 215: 93-102   Speers-Roesch B., M. Mandic , D.J.E. Groom and J.G. Richards. 2013. Critical oxygen  tensions as predictors of hypoxia tolerance and tissue metabolic responses during hypoxia exposure in fishes. Journal of Experimental Marine Biology and Ecology 449: 239-249  Stearns S.C. 1989. The evolutionary significance of phenotypic plasticity. Bioscience  39: 436-445  Stebbens G.L. 1977. Process of organic evolution. Prentice-Hall, Englewood Cliffs   Stierhoff K.L., T.E. Targett and P.A. Grecay. 2003. Hypoxia tolerance of the  mummichog: the role of access to the water surface. Journal of Fish Biology 63: 580-592  Storey J.D. 2002. A direct approach to false discovery rates. Journal of the Royal  Statistical Society Series B 64: 479-498  Storey K.B. 1987. Tissue-specific controls on carbohydrate catabolism during anoxia  in goldfish. Physiological Zoology 60: 601-607  Tetens V. and G. Lykkeboe. 1981. Blood respiratory properties of rainbow trout,  Salmo gairdneri: responses to hypoxia acclimation and anoxic incubation of blood in vitro. Journal of Comparative Physiology B 145: 117-125   Timmerman C.M. and L.J. Chapman. 2001. Hypoxia and interdemic variation in   126 Poecilia latipinna. Journal of Fish Biology 65: 635-650  Ton C., D. Stamatiou and C.C. Liew. 2003. Gene expression profile of zebrafish  exposed to hypoxia during development. Physiological Genomics 13: 97-106   Truchot J.-P. and A. Duhamel-Jouve. 1980. Oxygen and carbon dioxide in the marine  intertidal environment: diurnal and tidal changes in rockpools. Respiration physiology 39: 241-254  Trussell G.C. and R.J. Etter. 2001. Integrating genetic and environmental forces that  shape the evolution of geographic variation in a marine snail. Genetica 112-113: 321-337.  Val A.L. 1999. Hypoxia adaptation in fish of the Amazon: a never-ending task. South  African Journal of Zoology 33: 107-125  Van den Thillart G., F. Kesbeke and A. van Waarde. 1980. Anaerobic energy- metabolism of goldfish, Carassius auratus (L.): influence of hypoxia and anoxia on phosphorylated compounds and glycogen. Journal of Comparative Physiology 136: 45-52   Van den Thillart G. and M. van Raaij. 1995. Endogenous fuels; non-invasive versus  invasive approaches. In: Biochemistry and Molecular Biology of Fishes vol. 4 (eds. P.W. Hochachka and T.P. Mommsen), pp. 33-63, Elsevier, Amsterdam.   van den Thillart, G., van Waarde, A., Muller, H.J., Erkelens, C., Addink, A. and  Lugtenburg, J. 1989. Fish muscle energy metabolism measured by in vivo 31P-NMR during anoxia and recovery. Am. J. Physiol. 256: R922-R929.   Van der Linden A., M. Verhoye, and G.E. Nilsson. 2001. Does anoxia induce cell  swelling in carp brains? In vivo MRI measurements in crucian carp and  127 common carp. Journal of Neurophysiology 85: 125-133  Van der Meer D.L.M., G.E.E.J.M. van den Thillart, F. Witte, M.A.G. de Bakker, J.  Besser, M.K. Richardson, H.P. Spaink, J.T.D. Leito and C.P. Bagowski. 2005. Gene expression profiling of the long-term adaptive response to hypoxia in the gills of adult zebrafish. American Journal of Physiology – Regulatory, Integrative and Comparative Physiology 289: R1512-R1519  Van Waarde A., G. Van den Thillart, C. Erkelens, A. Addink, and J. Lugtenburg. 1990.  Functional coupling of glycolysis and phosphocreatine utilization in anoxic fish muscle: an in vivo 31P NMR study. Journal of Biological Chemistry 265: 914-923  van Waversveld J., A.D.F. Addink and G. van den Thillart. 1989. The anaerobic energy  metabolism of goldfish determined by simultaneous direct and indirect calorimeter during anoxia and hypoxia. Journal of Comparative Physiology B 159: 263-268  Vaquer-Sunyer R. and C.M. Duarte. 2008. Thresholds of hypoxia for marine  biodiversity. Science 105: 15452-15457  Via J.D., G. van den Thillart, O. Cattani and A. de Zwaan. 1994. Influence of long-term  hypoxia exposure on the energy metabolism of Solea solea. II. Intermediary metabolism in blood, liver and muscle. Marine Ecology Progress Series 111: 17-27  Via S., R. Gomulkiewicz, G. De Jong, S.M. Scheiner, C.D. Schlichting and P.H. van  Tienderen. 1995. Adaptive phenotypic plasticity: consensus and controversy. Trends in Ecology and Evolution 10: 212-217   Vinciotti	  V.,	  R.	  Khanin,	  D.	  D’Alimonte,	  X.	  Liu,	  N.	  Cattini,	  G.	  Hotchkiss,	  G.	  Bucca,	  O.	    128 de Jesus, J. Rasaiyaah, C.P. Smith, P. Kellam and E. Wit. 2006. An experimental evaluation of a loop versus a reference design for two-channel microarrays. Bioinformatics 21: 492-501  Walker R.M. and P.H. Johansen. 1977. Anaerobic metabolism in goldfish (Carassius  auratus). Canadian Journal of Zoology 55: 1304-1311  Watters J.V. and J.J. Cech Jr. 2003. Behavioural responses of mosshead and woolly  sculpins to increasing environmental hypoxia. Copeia 2003: 397-401  Weber R.E. and G. Lykkeboe. 1978. Respiratory adaptations in carp blood: influences  of hypoxia, red cell organic phosphates, divalent cations and CO2 on  hemoglobin-oxygen affinity. Journal of Comparative Physiology 128: 127-137  Webster K.A. 2003. Evolution of the coordinate regulation of glycolytic enzyme  genes by hypoxia. The Journal of Experimental Biology 206: 2911-2922  West-Eberhard M.J. 1989. Phenotypic plasticity and the origins of diversity. Annual  Review of Ecology and Systematics 20: 249-278  Wilson D.S. and J. Yoshimura. 1994. On the coexistence of specialists and generalists.  The American Naturalist 144: 692-707  Whitehead A. 2012. Comparative genomics in ecological physiology: toward a more  nuanced understanding of acclimation and adaptation. The Journal of Experimental Biology 215: 884-891  Wu H., modified by H. Yang H, K. Sheppard with ideas from G. Churchill, K. Kerr and   X. Cui. 2013. maanova: Tools for analyzing Micro Array experiments. R package version 1.30.0. http://research.jax.org/faculty/churchill. Data retrieved from internet March 2011  129  Yoshiyama R.M., C.J. Valpey, L.L. Schalk, N.M. Oswald, K.K. Vaness, D. Lauritzen and  M. Limm. 1995. Differential propensities for aerial emergence in intertidal sculpins (Teleostei: Cottidae). Journal of Experimental Marine Biology and Ecology 191: 195-207   130 Appendix	  A	   Tissue enzymes in 12 species of sculpin. Tissue enzymes are presented in Pmol/min/mg protein except brain CS, liver LDH, liver PK, liver CS and muscle CS are presented in nmol/min/mg protein. Numbers in parentheses represent sample size. Data are means ± SE.   Species Brain  Liver  Muscle LDH PK CPK CS  LDH PK CPK CS  LDH PK  CPK       CS O. maculosus 1.3±0.1 (8) 0.8±0.1 (8) 7.2±0.2 (8) 27.2±0.1 (8)  16.7±3.4 (16) 34.0±4.4 (16) 0.2±0.1 (16) 4.3±0.2 (16)  3.3±0.3 (8) 0.4±0.1 (8) 34.6±2.6 (8) 8.3±1.1        (8) C. globiceps 1.3±0.1 (8) 0.8±0.1 (8) 9.1±0.4 (8) 33.5±0.2 (8)  52.2±15.1 (10) 30.4±3.3 (10) 0.4±0.1 (10) 12.3±1.8 (10)  3.6±0.3 (8) 0.4±0.1 (8) 32.2±1.1 (8) 10.4±0.7        (8) O. snyderi 1.2±0.1 (8) 0.7±0.1 (8) 8.7±0.5 (8) 25.4±0.1 (8)  35.1±8.5 (8) 17.2±3.1 (8) 0.3±0.1 (8) 5.7±0.8 (8)  2.6±0.3 (8) 0.3±0.1 (8) 32.9±1.8 (8) 6.6±0.6        (8) M. polyacanthocephalus 1.1±0.1 (8) 0.6±0.1 (8) 6.0±0.7 (8) 33.8±0.4 (8)  13.6±0.9 (8) 29.0±2.5 (8) 0.2±0.1 (8) 3.8±0.2 (8)  4.4±0.2 (8) 0.5±0.1 (8) 27.7±3.7 (8) 10.7±1.1        (8) E. bison 1.0±0.1 (9) 0.6±0.1 (9) 6.4±0.3 (9) 32.4±0.2 (9)  43.0±10.6 (11) 40.3±8.2 (11) 0.1±0.1 (11) 6.6±0.7 (11)  2.9±0.4 (9) 0.4±0.1 (9) 28.1±2.8 (9) 16.5±1.8        (9) A. fenestralis 1.1±0.1 (8) 0.6±0.1 (8) 6.9±0.4 (8) 33.3±0.2 (8)  11.5±1.5 (8) 27.2±4.5 (8) 0.1±0.1 (8) 5.0±1.0 (8)  5.0±0.2 (8) 0.5±0.1 (8) 31.9±3.3 (8) 6.8±1.0        (8) A. lateralis 0.9±0.1 (7) 0.7±0.1 (7) 7.2±1.0 (7) 28.0±0.2 (7)  13.5±1.8 (7) 25.5±3.9 (7) 0.2±0.1 (7) 3.6±0.1 (7)  3.3±0.2 (8) 0.4±0.1 (8) 33.3±1.9 (8) 3.4±0.7        (8) L. armatus 1.3±0.1 (8) 0.8±0.1 (8) 8.1±0.4 (8) 37.7±0.2 (8)  28.0±4.8 (12) 30.0±3.5 (12) 0.2±0.1 (12) 5.1±0.3 (12)  4.7±0.3 (8) 0.4±0.1 (8) 36.0±4.5 (8) 9.5±0.8        (8) S. marmoratus 1.1±0.1 (8) 0.7±0.1 (8) 7.0±0.3 (8) 36.7±0.2 (8)  19.7±1.9 (12) 38.8±4.8 (12) 0.1±0.1 (12) 5.7±0.4 (12)  2.9±0.5 (8) 0.3±0.1 (8) 39.0±3.1 (8) 10.4±1.0        (8) B. cirrhosis 0.8±0.1 (10) 0.6±0.1 (10) 6.0±0.3 (10) 32.0±0.1 (10)  9.0±1.0 (12) 24.1±1.3 (12) 0.1±0.1 (12) 6.5±0.5 (12)  1.2±0.1 (10) 0.2±0.1 (10) 22.0±4.2 (10) 7.3±0.7       (10) M. scorpius 0.9±0.1 (8) 0.6±0.1 (8) 5.4±0.2 (8) 34.5±0.1 (8)  19.8±1.9 (7) 33.8±4.1 (7) 0.2±0.1 (7) 4.4±0.3 (7)  2.1±0.5 (8) 0.4±0.1 (8) 29.7±1.0 (8) 15.1±1.7        (8) C. asper 1.1±0.1 (8) 0.9±0.1 (8) 7.0±0.9 (8) 28.5±0.5 (8)  11.1±1.1 (8) 15.2±2.9 (8) 0.2±0.1 (8) 5.5±0.2 (8)  4.8±0.2 (8) 0.5±0.1 (8) 34.4±0.7 (8) 10.1±0.8        (8)  131 Appendix	  B A hybridization loop for each species of sculpin. The 29 individuals per species are indicated by the number that is next to every time-point.   132    133 Appendix	  C Euclidean distance for the remainder of the genes associated with biological processes presented in figures 3.3 and 3.4.  Gene Name Gene Ontology: Biological Process UniProt Euclidean Distance Hydroxyacyl-coenzyme A dehydrogenase, mitochonsdrial GO:0019395~fatty acid oxidation DDT4 1.251 Importin-7 GO:0008104~protein localization DBLOH 1.246 Heat shock protein HSP 90-alpha GO:0006457~protein folding FABPH 1.227 Eukaryotic translation initiation factor 4E-binding protein 3-like GO:0006446~regulation of translational initiation SGK1 1.219 Macrophage migration inhibitory factor GO:0006955~immune response IF4G2 1.193 Macrophage migration inhibitory factor GO:0043066~negative regulation of apoptosis IF4G2 1.193 Macrophage migration inhibitory factor GO:0008284~positive regulation of cell proliferation UBE2A 1.193 Thioredoxin GO:0008284~positive regulation of cell proliferation K6PL 1.145 Keratin, type I cytoskeletal 18 GO:0008104~protein localization CGL 1.116 Keratin, type I cytoskeletal 18 GO:0043066~negative regulation of apoptosis EIF1B 1.116 C-C motif chemokine 25 GO:0006955~immune response K6PF 1.079 Protein DJ-1 GO:0000302~response to reactive oxygen species CFAH 1.052 Enoyl-CoA hydratase, mitochondrial GO:0019395~fatty acid oxidation CFAH 1.040 Annexin A5 GO:0043066~negative regulation of apoptosis CIDEB 1.035 Phospholipid hydroperoxide glutathione peroxidase, mitochondrial GO:0051276~chromosome organization BDH2 1.034 Phospholipid hydroperoxide glutathione peroxidase, mitochondrial GO:0000302~response to reactive oxygen species MYD88 1.034 Peptidyl-prolyl cis-trans isomerase F GO:0006457~protein folding TCPZ 1.015  134 Gene Name Gene Ontology: Biological Process UniProt Euclidean Distance Annexin A4 GO:0043066~negative regulation of apoptosis HS90B 1.008 Collectin-12 GO:0006955~immune response SET 1.007 T-complex protein 1 subunit eta GO:0006457~protein folding HMR1 0.997 10 kDa heat shock protein, mitochondrial GO:0006457~protein folding TPM1 0.994 Beta-2-microglobulin GO:0006955~immune response TCPQ 0.968 Peptidyl-prolyl cis-trans isomerase FKBP1A GO:0006457~protein folding HCDH 0.958 Peptidyl-prolyl cis-trans isomerase FKBP1A GO:0051604~protein maturation IPO7 0.958 Death-associated protein-like 1 GO:0043065~positive regulation of apoptosis HS90A 0.931 Complement C1q subcomponent subunit C GO:0006955~immune response 4EBP3 0.930 Complement C1q subcomponent subunit C GO:0051604~protein maturation MIF 0.930 Medium-chain specific acyl-CoA dehydrogenase, mitochondrial GO:0019395~fatty acid oxidation MIF 0.929 Heat shock protein 75 kDa, mitochondrial GO:0006457~protein folding MIF 0.839 ADP-ribosylation factor 5 GO:0008104~protein localization THIO 0.826 Complement C3 GO:0006955~immune response K1C18 0.824 Complement C3 GO:0051604~protein maturation K1C18 0.824 Glucose-6-phosphate isomerase GO:0006096~glycolysis CCL25 0.805 Glucose-6-phosphate isomerase GO:0006955~immune response PARK7 0.805 Mannan-binding lectin serine protease 1  GO:0006955~immune response ECHM 0.802 Mannan-binding lectin serine protease 1  GO:0051604~protein maturation ANXA5 0.802 Clathrin light chain A GO:0008104~protein localization GPX4 0.796  135 Gene Name Gene Ontology: Biological Process UniProt Euclidean Distance Ribosome biogenesis protein NSA2 homolog GO:0006396~RNA processing GPX4 0.781 Peroxisome proliferator-activated receptor delta GO:0008284~positive regulation of cell proliferation PPIF 0.771 Mitochondrial uncoupling protein 2 GO:0000302~response to reactive oxygen species ANXA4 0.764 Peroxiredoxin-1 GO:0008284~positive regulation of cell proliferation COL12 0.668 Peroxiredoxin-1 GO:0000302~response to reactive oxygen species TCPH 0.668 40S ribosomal protein S7 GO:0006396~RNA processing CH10 0.665 40S ribosomal protein S7 GO:0006414~translational elongation B2MG 0.665 Cytochrome b-245 light chain GO:0006955~immune response FKB1A 0.664 Elongation factor 1-gamma GO:0006414~translational elongation FKB1A 0.656 Poly(U)-binding-splicing factor PUF60 GO:0043065~positive regulation of apoptosis DAPL1 0.653 Poly(U)-binding-splicing factor PUF60 GO:0006396~RNA processing C1QC 0.653 Cytochrome b-c1 complex subunit 10 GO:0006119~oxidative phosphorylation C1QC 0.646 ATP synthase subunit d, mitochondrial GO:0006119~oxidative phosphorylation ACADM 0.637 60S ribosomal protein L39 GO:0006414~translational elongation TRAP1 0.626 ATP synthase subunit e, mitochondrial GO:0006119~oxidative phosphorylation ARF5 0.622 Catenin beta-1 GO:0043065~positive regulation of apoptosis CO3 0.611 Catenin beta-1 GO:0008104~protein localization CO3 0.611 Eukaryotic translation initiation factor 5A-1 GO:0043065~positive regulation of apoptosis G6PI 0.600 Eukaryotic translation initiation factor 5A-1 GO:0008104~protein localization G6PI 0.600 Eukaryotic translation initiation factor 5A-1 GO:0008284~positive regulation of cell proliferation MASP1 0.600 Eukaryotic translation initiation factor 5A-1 GO:0006414~translational elongation MASP1 0.600  136 Gene Name Gene Ontology: Biological Process UniProt Euclidean Distance Alpha-enolase GO:0006096~glycolysis CLCA 0.587 ATP synthase subunit O, mitochondrial GO:0006119~oxidative phosphorylation NSA2 0.580 Pyruvate kinase isozymes M1/M2 GO:0043065~positive regulation of apoptosis PPARD 0.547 Pyruvate kinase isozymes M1/M2 GO:0006096~glycolysis UCP2 0.547 Peptidyl-prolyl cis-trans isomerase NIMA-interacting 1 GO:0006457~protein folding PRDX1 0.536 Peptidyl-prolyl cis-trans isomerase NIMA-interacting 1 GO:0008284~positive regulation of cell proliferation PRDX1 0.536 60S ribosomal protein L30 GO:0006414~translational elongation RS7 0.518 Coronin-1A GO:0006955~immune response RS7 0.513 Coronin-1A GO:0008284~positive regulation of cell proliferation CY24A 0.513 60S ribosomal protein L13 GO:0006414~translational elongation EF1G 0.505 40S ribosomal protein S21 GO:0006414~translational elongation PUF60 0.503 60S ribosomal protein L10 GO:0006414~translational elongation PUF60 0.500 RAF proto-oncogene serine/threonine-protein kinase GO:0043065~positive regulation of apoptosis QCR10 0.479 RAF proto-oncogene serine/threonine-protein kinase GO:0008285~negative regulation of cell proliferation ATP5H 0.479 ATP synthase subunit gamma, mitochondrial GO:0006119~oxidative phosphorylation RL39 0.478 Translocation protein SEC63 homolog GO:0008104~protein localization ATP5I 0.468 Centrosomal protein of 57 kDa GO:0008104~protein localization CTNB1 0.276 ADP-ribosylation factor 6 GO:0008104~protein localization CTNB1 0.257 Methylosome protein 50 GO:0006396~RNA processing IF5A1 0.240 Signal peptidase complex catalytic subunit SEC11A GO:0051604~protein maturation IF5A1 0.140  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.24.1-0166251/manifest

Comment

Related Items