Ecotype-specific transcriptomic differences in the telencephalon of Alaskan threespine stickleback, Gasterosteus aculeatus Sanna Pausio Ecology and Evolutionary Biology Master's thesis Credits: 40 ECTS Supervisors: Erica Leder Heidi Viitaniemi Irma Saloniemi 14.4.2022 Turku The originality of this thesis has been checked in accordance with the University of Turku quality assurance system using the Turnitin Originality Check service. Master's thesis Subject: Ecology and Evolutionary Biology Author: Sanna Pausio Title: Ecotype-specific transcriptomic differences in the telencephalon of Alaskan threespine stickleback, Gasterosteus aculeatus Supervisors: Erica Leder, Heidi Viitaniemi, Irma Saloniemi Number of pages: 60 pages, 5 appendices Date: 14.4.2022 __________________________________________________________________________________ Brain activity reacts to environmental complexity within an individual’s reaction norm. Brain activity is highly energy demanding and a target for energetic trade-offs. Hence, cognition is an honest but complex signal of selection landscape. ‘Clever foraging hypothesis’ predicts increased fitness value for cognition in patchy, moderately stable environments. In this thesis, I have studied telencephalon transcriptome of threespine stickleback, Gasterosteus aculeatus Linnaeus, 1758, females from four lakes with two levels of environmental complexity in subarctic Alaska. The threespine stickleback populations in these lakes represent benthic and limnetic ecotypes. Populations of these ecotypes have repeatedly and independently been established in North America after the last glaciation from the ancestral marine form. Despite their young evolutionary age, the ecotypes are phenotypically different. Because of the patchier nature of foraging in benthic environment, benthic populations should benefit more from investment in spatial cognition compared to limnetic populations. Therefore, I hypothesized increased expression of cognition-related genes in the telencephalons of benthic individuals. Contrary to the expectations, more genes involved in the regulation of neurogenesis and neuron projection were upregulated in the limnetic ecotype. The most significant upregulated biological processes in the benthic ecotype were lipid transport and innate immunity. This suggests greater investment in cognition in the limnetic ecotype and immune defense in the benthic ecotype. These results highlight the impact of immunology instead of spatial complexity as the main factor in the telencephalon gene expression in the study populations. More research is needed to discover the environmental and pathogen community differences among the lake types and the impact of genetic differentiation on the reaction norm of immunity expression among the threespine stickleback ecotypes. __________________________________________________________________________________ Key words: threespine stickleback, Gasterosteus aculeatus, transcriptome, habitat complexity, telencephalon, microarray Pro gradu tutkielma Oppiaine: Ekologia ja evoluutiobiologia Tekijä: Sanna Pausio Otsikko: Ecotype-specific transcriptomic differences in the telencephalon of Alaskan threespine stickleback, Gasterosteus aculeatus Ohjaajat: Erica Leder, Heidi Viitaniemi, Irma Saloniemi Sivumäärä: 60 sivua, 5 liitettä Päivämäärä: 14.4.2022 __________________________________________________________________________________ Aivojen ja hermoston toiminta mukautuu nopeasti ympäristön vaatimuksiin yksilön reaktionormin asettamissa rajoissa. Niiden toiminta kuluttaa runsaasti energiaa. Jotta energian allokointi aivotoimintaan olisi kannattavaa, sen on tuotettava yksilölle valintahyötyä. Siksi älykkyyttä ja aivotoimintaa voidaan käyttää mittareina ympäristön valintapaineista. ”Clever foraging” -hypoteesi ennustaa resurssien allokoinnin aivotoimintaan ja oppimiseen hyödyttävän yksilöä eniten kohtuullisen stabiileissa ja laikuttaisissa elinympäristöissä. Tässä pro gradu -työssä olen verrannut kolmipiikkinaaraiden Gasterosteus aculeatus Linne, 1758, aivojen nisäkkäiden hippokampusta vastaavan alueen, pääteaivojen, transkriptomeja neljässä subarktisen Alaskan järvipopulaatiossa. Tutkimuspopulaatiot edustavat benttistä (pohja) ja limneettistä (ulappa) ekotyyppiä ja niiden habitaatit kahta ympäristön laikuttaisuuden tasoa. Kolmipiikki on levinnyt Pohjois-Amerikan järviin viimeisimmän jäätiköitymisen jälkeen useiden erillisten leviämistapahtumien seurauksena. Populaatiot ovat toistuvasti ja toisistaan riippumattomasti kehittyneet alkuperäisestä mereisestä muodostaan kohti benttistä ja limneettistä ekotyyppiä. Huolimatta nuoresta evolutiivisesta iästään ekotyypit ovat fenotyypiltään erilaisia. Koska benttisissä elinympäristöissä resurssit ovat sijoittuneet limneettisiä ympäristöjä laikuttaisemmin, on todennäköistä, että benttisiä resursseja hyödyntävät kolmipiikit hyötyvät enemmän informaatiosta ja sen käsittelystä elinympäristössään. Tämän takia oletin tiedon käsittelyyn liittyvien geenien transkription olevan runsaampaa benttiseen ekotyyppiin kuuluvissa populaatioissa. Vastoin oletusta hermosolujen kasvun ja synnyn säätelyyn liittyvien geeniryhmien transkriptio oli runsaampaa limneettisissä populaatioissa. Benttisissä populaatioissa parhaiten edustettuina olivat rasvan kuljetusta ja synnynnäisen immuniteetin ilmenemistä säätelevät geeniryhmät. Vaikuttaakin siltä, että limneettisten populaatioiden pääteaivojen geeniekspressio painottuu tiedon käsittelyyn ja benttisten populaatioiden synnynnäisen immuniteetin ilmenemiseen. Tuloksissa painottuu ympäristön laikuttaisuuden sijaan immuunipuolustuksen merkitys tutkimuspopulaatioiden geeniekspressiossa. Lisää tutkimusta tarvitaan benttisten ja limneettisten elinympäristöjen ja niiden patogeeniyhteisöjen eroista sekä siitä, onko tutkimusekotyyppien geneettinen erilaistuminen vaikuttanut niiden synnynnäisen immuniteetin ilmenemisen reaktionormeihin. __________________________________________________________________________________ Avainsanat: kolmipiikki, Gasterosteus aculeatus, transkriptomi, ympäristön laikuttaisuus, pääteaivot, microarray Table of contents 1 Introduction 6 1.1 Adaptation to environmental complexity 6 1.1.1 The significance of adaptation 6 1.1.2 Cognition as adaptation to complexity 6 1.2 Physiological and molecular view to information processing and memory formation 7 1.2.1 The neural network for environmental information processing 7 1.2.2 Mechanisms for memory formation and neural plasticity 8 1.2.3 Molecular regulation of information processing 10 1.3 Physical approaches for cognition assessment 11 1.3.1 Brain size and morphology 11 1.3.2 Molecular measures for cognition 12 1.4 Threespine stickleback as model species 13 1.5 Study populations 15 1.6 Aim of the thesis 18 2 Material and methods 19 2.1 Sample collection and RNA extraction 19 2.2 Microarray analysis 20 2.3 Differential expression analysis 20 2.4 Gene set enrichment analysis 21 2.4.1 Probe sequence annotation 21 2.4.2 Functional enrichment analysis 22 3 Results 23 3.1 Differential gene expression 23 3.2 Functional enrichment analysis 28 4 Discussion 33 4.1 Population differentiation 33 4.2 Ecotype-specific patterns in neural regulation 34 4.2.1 Upregulated neurogenesis and transcription in the limnetic ecotype 34 4.2.2 Metabolic functions and immune defense enriched in the benthic ecotype 35 4.3 Potential sources for the ecotype-specific gene expression profiles 37 4.3.1 Limnetic populations 37 4.3.2 Benthic populations 38 4.4 Elements of uncertainty 41 4.4.1 Populations and tissue collection 41 4.4.2 Data analysis 41 5 Conclusions 43 Acknowledgments 44 References 45 Appendices 61 Appendix 1. RNA isolation protocol 61 Appendix 2. Microarray design and MA-plots 62 Appendix 3. Differentially expressed genes and their human orthologs. 63 Appendix 4. FGSEA enriched biological processes and molecular functions 81 Appendix 5. Expressed genes annotated to FGSEA enriched GO terms 85 6 1 Introduction 1.1 Adaptation to environmental complexity 1.1.1 The significance of adaptation Adaptation to fluctuating environmental conditions in space and time is essential for survival of both individuals and populations. The ability to adapt to altering conditions is also vital for dispersal (e.g., Sol et al., 2008) and a stepping stone for ecological speciation (Rundle, 2000; Savolainen et al., 2013). Understanding the processes, mechanisms, and prerequisites for adaptation is needed to assess the ecological and evolutionary consequences, that even subtle, small-scale changes in abiotic or biotic factors may induce (Urban et al., 2020). Studying molecular level responses to environmental factors aims to identify the genetic basis and molecular interactions regulating the emergence of a phenotypic outcome. Together, the three levels, genes, interactions among gene products, and phenotype as the interface between individual and environment, comprise a complex and reciprocal network. Insight into the function of this network and its interaction with ecological and non-ecological factors that drive systemic and population-level change is essential; not only for ecological and evolutionary research but also for more applied fields such as conservation biology or climate change research (Savolainen et al., 2013). 1.1.2 Cognition as adaptation to complexity Cognition is based on information and the ability to process it. From organismal perspective, it is a combination of three interacting aspects: perception, learning and memory (Arechavala‑Lopez et al., 2020). It is a highly plastic (multiple phenotypes may arise from a single genotype) trait, and its physiology is regulated by several delicate mechanisms, as discussed later in this introduction. While cognition itself is beneficial for fitness, the expenses coming from information gathering and processing may exceed the benefits (Dall et al., 2005). To fitness benefits outweigh the expenses, information has to be reliable and valuable enough to change an individual’s behavior in a beneficial manner (Dall et al., 2005). This threshold depends on organismal characteristics (Leavell and Bernal, 2019) together with environmental predictability, patchiness, and productivity (Olsson and Brown, 2010). Patchiness can also be described as complexity, which consists of the inter- and intraspecies interactions as well as 7 spatial and temporal patchiness in the physical environment. Social environment, group size and parental care, (Gonzalez-Voyer et al., 2009; Kotrschal et al., 2012; Samuk et al., 2014; Tsuboi et al., 2015; Ashton et al., 2018), as well as ecology, above all diet, (Gonzalez-Voyer et al., 2009; Henke-von der Malsburg et al., 2020) has been shown to explain cognitional differences among populations. Often the aspects of environmental complexity act together (Pollen, 2007). According to ‘statistical decision theory’ (Guitton et al., 1955; McNamara and Houston, 1980) and ‘optimal foraging theory’ (Emlen, 1966; MacArthur and Pianka, 1966; Charnov, 1976), the benefits from information-based strategies for an individual are highest in moderately predictable and complex environments. Hence, these environments should promote the evolution of cognition. ‘Clever foraging hypothesis’ (Parker and Gibson, 1977; Park and Bell, 2010), the theoretical basis of this thesis, is based on the presumptions of those theories. According to ‘clever foraging hypothesis’, cognitive evolution is mainly driven by the relative benefits of complex food searching strategy for an individual in an environment. Hence, it predicts increased information value and cognition in active predators with patchy food distribution (Park and Bell, 2010). This hypothesis has been tested in both intra- and interspecies cognition studies in e.g., threespine stickleback Gasterosteus aculeatus (Park and Bell, 2010; Ahmed et al., 2017), arctic char Salvelinus alpinus (Tamayo et al., 2020), sharks (Yopak et al., 2007) and carnivores (Gittleman, 1986). Finding a robust correlation between cognition and fitness is complicated by the complexity of environmental relationships and their temporal fluctuation (Rowe and Healy, 2014). Moreover, biases arising from animal personality, life history, and the complexity of the concept of cognition may hinder distinguishing cognitional differences (Rowe and Healy, 2014). Not only the ability to learn as such, but also the cognitional strategy, e.g., trade-off between initial learning efficiency and learning flexibility (Bebus et al., 2016), may have different fitness value in different environments (Croston et al., 2017). 1.2 Physiological and molecular view to information processing and memory formation 1.2.1 The neural network for environmental information processing Sensory organs are the first step in environmental information processing. Their neuron count and surface area is an indication of the type of cues an animal relies on in its environment 8 (Kotrschal et al., 1998; Keagy et al., 2018). Signals received by sensory organs are processed and the information is stored in specific sets of neural cells, neural circuits, that wire together to process specific information (Holtmaat and Svoboda, 2009; Chen et al., 2020). In amniotes, the most important brain regions for processing environmental information are the hippocampus and the cortex (Squire et al., 2004; Frankland and Bontempi, 2005). In addition to spatial mapping, amniote hippocampus is associated with several other processes, such as emotions and stress-related behavior (Park et al., 2016; Herold et al., 2019). The hippocampus is a crucial functional region in explicit memory that requires conscious recollection of facts or experiences. Hippocampus serves as initial memory storage for newly acquired memories (Tonegawa et al., 2018). According to well established systems consolidation theory, memories are gradually transferred from hippocampus to cortical network for final memory stabilization (Frankland and Bontempi, 2005). Still, hippocampus continues to contribute in remote memory management, recollection and cognitional flexibility (Abraham et al., 2002; Frankland and Bontempi, 2005; Herold et al., 2019), although hippocampus involvement in memory management after complete maturation of cortical storage is still debated (Tonegawa et al., 2018). In teleost fish, brain organization differs from that of amniotes because of different developmental process during ontogeny (Salas et al., 2003). Teleost telencephalon dorsolateral region is generally considered to be functionally equivalent to amniote hippocampus, while the dorsomedial region is considered to be responsible for amniote amygdala-like functions, such as fear and social behavior (Demski, 2013). Teleost telencephalon has been demonstrated to mediate a wide variety of behaviors, e.g., spatial and conditional learning, aggressivity, processing of sensory information, and even reproduction (Demski, 2013). 1.2.2 Mechanisms for memory formation and neural plasticity Cognition as a process consists of four phases: memory acquisition, consolidation, retrieval, and reconsolidation. These phases are enabled by neural plasticity: the ability of neural network to create novel connections and regulate the strength of existing ones to build and remodel memory engrams (sets of neurons that contribute to consolidation of a memory). Neural plasticity involves both functional (Oh and Smith, 2019) and structural (Holtmaat and 9 Svoboda, 2009; Zupanc and Sîrbulescu, 2011; Anggono and Huganir, 2012; Kol and Goshen, 2021) changes in synapses and neural networks. Neural plasticity is mainly achieved by four mechanisms: (1) differentiation and co-operation of neurons into inhibitory neurons, whose facilitation decreases the possibility of action potential in postsynaptic neuron; and excitatory neurons, whose facilitation induces action potential in postsynaptic neuron (Oh and Smith, 2019), (2) plasticity in the number and location of synapses, which is regulated by the number and location of dendritic spines and axonal boutons as well as clustering of neurotransmitter receptors (Holtmaat and Svoboda, 2009), (3) change in synaptic efficacy (synaptic strengthening (potentiation) and weakening (depression)) defined mainly by the status and number of neurotransmitter receptors but also neurotransmitter release (Malenka and Bear, 2004), and (4) proliferation and integration of new neural cells in neural circuits (von Krogh et al., 2010; Steadman et al., 2020). In addition to synaptic plasticity, memory formation is regulated by glial cells: oligodendrocytes, astrocytes, and microglia. Oligodendrocytes participate in the facilitation of memory consolidation (Steadman et al., 2020) and memory retrieval (Pan et al., 2020). Astrocytes have been shown to be capable of facilitating memory acquisition (Adamsky et al., 2018), long-term memory consolidation and memory allocation by secreting synaptic connection regulating factors (Kol and Goshen, 2021; Wang et al., 2021b), but see Agulhon et al. (2010) for opposing results. Microglia have immunological functions in brain, but they are also capable of promoting synaptogenesis, activity-dependent elimination of synapses and direct regulation of synaptic plasticity (Wang et al., 2021a). Neural stimulation, both learning and memory retrieval, lower the threshold for synaptic wiring (increase intrinsic excitability) in stimulated synapses (Parsons, 2018; Chen et al., 2020). Learning also stimulates neural cell proliferation, and promotes the survival of relatively mature and the death of more immature neurons (Dupret et al., 2007). Ultimately, signaling intensity (Medina et al., 2019), pattern (Lee et al., 2017), and duration (Lee et al., 2017; Tyssowski et al., 2018) as well as previous experiences (Parsons, 2018) and systemic homeostasis (Johansen et al., 2012) determine the trajectory and the result of information processing. 10 1.2.3 Molecular regulation of information processing Changes is synaptic efficacy initiate by voltage- and neurotransmitter-mediated opening of calcium-permeable AMPA-receptors (AMPARs) in glutamatergic excitatory neurons and chlorine-permeable GABA-receptors (GABARs) in GABAergic inhibitory neurons. The induction of long-term potentiation (LTP) usually further requires the opening of another glutamatergic calcium channel, NMDAR, although other forms of LTP exist (Frey and Morris, 1998). Besides calcium channel status, synaptic strength is regulated by their number: the LTP-related active transport of intracellular AMPARs to synaptic density and their long- term depression (LTD)-related endocytosis (Malenka and Bear, 2004). Early phase of LTP, as well as some forms of LTD can be maintained by local translation in dendrites, but later phases of LTP and memory consolidation requires somatic transcription and translation (Costa-Mattioli et al., 2009), which is accompanied by increased genome-wide chromatin accessibility and nuclear reorganization (Tyssowski et al., 2018; Marco et al., 2020). Change in synaptic strength is mediated to gene expression by signaling pathways, which can be activated by intracellular calcium content change, but also regulated by a variety of other signaling molecules and regulatory pathways (Zheng et al., 2009; Miningou and Blackwell, 2020). The first wave of molecular response to a stimulus can be triggered by synaptic activity dependent activation of Ras-mitogen-associated protein kinase (MAPK or ERK), calcium/calmodulin-dependent protein kinases (CaMKs), and calcineurin-mediated signaling pathways (Yap and Greenberg, 2018). These pathways interact and co-act with each other and other signaling pathways in a neural stimulus-specific manner (Zheng et al., 2009). The connectivity and reciprocity of signaling pathways enables the simultaneous sensitivity of transcriptional response to both internal (homeostasis) and external signals. MAPK signaling can be activated by several pathways (Miningou and Blackwell, 2020), including phosphatidylinositol 3 kinase/protein kinase B (pI3k/AKT) pathway, which in turn can be activated by variety of growth factors (Mendoza et al., 2011). Both MAPK and pI3k/AKT pathways participate in mammalian target of rapamycin (mTOR) pathway regulation (Mendoza et al., 2011), which in neural cells is involved with various brain functions from cell proliferation to synaptic plasticity and regulation of complex behaviors (Lipton and Sahin, 2014). It, as well as AKT signaling itself, has plethora of downstream targets (Brazil et al., 2004; Miningou and Blackwell, 2020). MAPK and AKT signaling have been demonstrated to both counter- and co-act, even compensate, each other (Mendoza et al., 11 2011). The three pathways, MAPK, pI3k/AKT and mTOR, are central for memory formation (Chen et al., 2005; Costa-Mattioli et al., 2009), but are accompanied with vast number of other crucial pathways, e.g., p38 MAPK pathway, which is involved with the regulation of inflammatory responses, but also synaptic plasticity (Falcicchia et al., 2020). The complexity of the neural regulatory mechanisms enables the differentiated response to environmental stimuli among cell types and brain regions (Malenka and Bear, 2004; Yap and Greenberg, 2018). Whilst the majority of the evidence for the mechanisms presented here comes from mammalian hippocampus, information processing is in many respects conserved across taxa (Salas et al., 2003; Costa-Mattioli et al., 2009). 1.3 Physical approaches for cognition assessment 1.3.1 Brain size and morphology Brain allometry and morphometry are commonly used as physical measures of cognition. Because of its energy demands, the size of a neural tissue can be considered to signal the value of the information it processes in an environment-individual interaction (Dunbar, 1998; Niven and Laughlin, 2008; Tsuboi et al., 2015). Hence, both total brain size and the size of functional brain regions, but also brain region morphometry (Park and Bell, 2010) has been used to study the relationship between cognition and environment. Positive correlation between ecological complexity and brain size has been demonstrated repeatedly in teleost fish (Pollen, 2007; Gonda et al., 2013; White and Brown, 2015; Keagy et al., 2018; Tamayo et al., 2020). This correlation has also been found in teleost telencephalon (Pollen, 2007; Gonda et al., 2009; White and Brown, 2015; Tamayo et al., 2020), the target tissue of this thesis. Both total brain (Gonzalez-Voyer et al., 2009; Kotrschal et al., 2012; Samuk et al., 2014; Tsuboi et al., 2015) and telencephalon (Pollen, 2007) sizes have also been shown to correlate with social complexity. Although the relationship with environmental complexity and brain size and morphometry is well established, some studies (Ahmed et al., 2017; Keagy et al., 2018; Axelrod et al., 2021) have failed to demonstrate this, which highlights the complexity of cognitive evolution. Park & Bell (2010) have discovered negative correlation between habitat complexity and telencephalon size in threespine stickleback, although the telencephalon shape indicated 12 positive correlation between ecological complexity and telencephalon dorsolateral region size in the study. The use of brain size as a proxy for cognition is supported by its experimentally established correlation with cognitive performance (Odling-Smee et al., 2008; Buechel et al., 2018; Keagy et al., 2018) and behavioral plasticity (Herczeg et al., 2019). Brain size also shows plastic response to environmental enrichment (Park et al., 2012; Fong et al., 2019), which has been demonstrated to have positive effects on learning ability (Salvanes et al., 2013; Carbia and Brown, 2019). However, morphometric measurement of brain functional regions instead of brain size may help to better distinguish the selective forces underlining cognitive evolution, as these regions wire together for specific cognitional tasks (Introduction 1.2.1). 1.3.2 Molecular measures for cognition Molecular approach to cognition study, i.e., measurement of protein or RNA expression instead of morphometry has the advantage of enabling the research of ongoing physiological processes. Especially RNA expression approach, being less technically challenging than proteomics (Hill et al., 2020), has been widely applied. Several studies have measured the transcript abundance of individual genes known to contribute to cognition (e.g., Salvanes et al., 2013). However, studying the entire transcriptome gives a more comprehensive view of processes that generate phenotypic expression, by allowing functional grouping of expressed genes. Transcriptomes can be studied by sequencing (RNA-seq) or by hybridizing mRNA to pre- selected gene oligonucleotides or open reading frames that are fixed on a slide or chip (i.e., microarray) (Malone and Oliver, 2011). RNA-seq studies have outnumbered microarrays in recent years (Mantione et al., 2014). Despite this, with appropriately selected, well-annotated candidate probes microarray studies can reveal ecologically and evolutionary meaningful molecular cascades by assaying mRNA abundances (e.g., Stanford et al., 2020). Several studies have demonstrated the connection between environmental complexity and cognition-related gene expression. For example genes related to neuronal signaling and neuron growth (Rampon et al., 2000), growth and learning (Evans et al., 2015), and synaptic plasticity (Podgorniak et al., 2015; Liu et al., 2020) have been found to be upregulated after environmental enrichment or correlated with environmental complexity. An intraspecies study in black-capped chickadees Poecile atricapillus has discovered positive correlation between 13 temporally increased environmental patchiness, stemming from harsher climate, and heritable upregulation of learning and neurogenesis related genes, previously shown to correlate with increased level of spatial learning (Pravosudov et al., 2013). 1.4 Threespine stickleback as model species Threespine stickleback is a small teleost fish, which is distributed around the northern hemisphere in the margins of the Atlantic and Pacific Oceans. It occupies marine environments and a wide range of freshwater habitats, forming marine, anadromous, and resident freshwater populations ranging from different fluvial environments to a variety of lentic habitats from ponds to large lakes (Bell and Foster, 1994). Ecological factors, such as salinity, temperature, the level of spatial complexity and species interactions vary among different threespine stickleback habitats. Threespine stickleback has several, morphologically distinct, habitat-specific ecotypes that have evolved repeatedly from the ancestral marine form in response for environmental factors (Bell and Foster, 1994). The ability to quickly adapt to novel environments has made it a popular model for exploring genomic features that enable the rapid adaptive radiation (Hohenlohe et al., 2010; Jones et al., 2012; Morris et al., 2014; Härer et al., 2021). The threespine stickleback genome was sequenced for the first time in 2006 (Stickleback assembly and gene annotation, 2010). The benthic-limnetic division is one of the best studied forms of adaptive radiation in threespine stickleback. Benthic (littoral) and limnetic (pelagic) lakes are structurally different (Foster et al., 2008; Park et al., 2012). The littoral zone of both benthic and limnetic lakes is complex, with landforms and rooted vegetation offering shelter from predation, contrary to the structurally monotonous, open pelagic area. Shallow, eutrophic benthic lakes have more littoral area than deep, oligotrophic limnetic lakes. In limnetic habitats, threespine sticklebacks feed mostly on open water, while in benthic habitats their diet consists of a mixture of benthic prey and plankton, although neither ecotype cannot be considered a specialist. Both ecotypes breed on benthos, and especially in allopatry, limnetic fish use littoral zone also for foraging (Park and Bell, 2010; Bolnick and Ballare, 2020). Benthic and limnetic threespine sticklebacks are morphologically different. The benthic ecotype has a deeper body, reduced defensive structures, longer upper jaw, and fewer and shorter gill rakers than limnetic morph (Bell and Foster, 1994; McPhail, 1994). They also differ in pigmentation (Gygax et al., 2018). The morphological differences are more 14 prominent in sympatric populations (Vamosi and Schluter, 2004). There are also behavioral differences in courtship behavior (Shaw et al., 2007), although in both ecotypes, males show territoriality and paternal care (Bell and Foster, 1994). Wild-caught limnetic fish show increased sociality compared to sympatric benthic ecotypes (Kozak and Boughman, 2008). Both limnetic and benthic ecotype have similar spatial learning strategies, but benthic individuals learn the tasks quicker and with fewer mistakes than limnetics (Odling-Smee et al., 2008; Park, 2013). Spatial performance has been discovered to have a heritable component (Martinez et al., 2016). Performance in reversal learning task (i.e., cognitional flexibility) has not been studied at the population level. Instead, at the individual level, personality type at the reactive/proactive axis has been shown to affect reversal learning performance (Bensky and Bell, 2020). Benthic threespine sticklebacks have been reported to have bigger brains than limnetics (Keagy et al., 2018). While the brain size difference is congruent with the presumptions of ‘clever foraging hypothesis’ in this ecotype pair, brain size fluctuation has been found to be inconsistent with environmental complexity gradient in another ecotype pair that inhabits environments at two complexity levels, namely lake-stream ecotype pair (Ahmed et al., 2017). Brain, including telencephalon, morphology is highly plastic in adult fish (Park et al., 2012; Buechel et al., 2019). The overall brain, but also telencephalon, size is bigger and more plastic in males than in females (Buechel et al., 2019). Over the breeding cycle, total brain and telencephalon volume grow in both sexes (Buechel et al., 2019). In Icelandic threespine stickleback populations, males, but not females, from structurally complex lava habitats has been reported to have bigger brain than those from more monotonous mud habitats (Kotrschal et al., 2012). Noreikiene et al. (2015) have discovered a small heritable component in telencephalon morphology in a full- and half-sibling experiment on a stickleback population from the Baltic Sea. However, in Alaskan lake populations, the morphological distinction between the telencephalons of benthic and limnetic sticklebacks has been demonstrated to disappear in the next generation reared in common garden (Park et al., 2012). 15 1.5 Study populations The samples for this thesis were collected from four Alaskan lakes in the Matanuska - Susitna and Kenai regions at the Gulf of Alaska (Figure 1, Table 1). The lakes have formed after the withdrawal of the Cordilleran ice sheet in the last glaciation ca 10 000 – 12 000 years ago (Bell and Foster, 1994), confirming the young evolutionary age of the study populations. All the study lakes locate in distinct drainages, suggesting the populations being established from distinct founder events. This is supported by genetic evidence from other freshwater populations in the area (Hohenlohe et al., 2010). The populations are from the same geographical area, which indicates phenotypic adaptive radiation to be based on shared ancestral genetic variation (Fang et al., 2020). Figure 1. Study lake locations. Mud Lake (Mud) and Tern Lake (Tern) are benthic lakes, shallow with high relative littoral area. Lynne Lake (Lynne) and South Rolly Lake (SRolley) are deeper and have more pelagic area (Park and Bell, 2010). Previously, the fish from these lakes has been morphologically determined as benthic and limnetic, respectively (Table 1) (Park and Bell, 2010). Also, the telencephalon shape of the limnetic and benthic fish has been shown to correspond to both 16 fish morphology and lake topography. Limnetic fish in the study lakes have laterally concave telencephalons in comparison to benthic fish convex ones, with fish from Tern and South Rolly lakes having the most extreme telencephalon shapes (Figure 2) (Park and Bell, 2010). This indicates smaller dorsolateral telencephalic regions in limnetic fish (Park and Bell, 2010). Cognitive performance in a spatial task has been shown to vary inconsistently in relation to ecotype in these populations, with the Mud population having the highest performance, followed by Lynne, Tern, and South Rolly populations (Park, 2013). Table 1. Properties of the study lakes and study populations Mat-Sus refers to Matanuska - Susitna region. *No information available. 1Benthic foraging reported (Park, Chase and Bell, 2012). Modified from (Park and Bell, 2010). Lynne South Rolly Mud Tern Topography Limnetic * Benthic Benthic Location Lat (N) 61.712 61.401 61.563 60.533 Location Long (W) 150.039 150.073 148.949 149.55 Drainage Little Susitna River Susitna River Knik River Kenai River Region Mat-Sus Mat-Sus Mat-Sus Kenai Fish body shape * Limnetic Benthic Benthic Foraging behavior Limnetic1 Limnetic * * Telencephalon shape Limnetic Limnetic Benthic Benthic Number of egg- carrying individuals sampled 1 3 * 2 Number of parasitized individuals sampled 1 - * 1 17 Figure 2. Telencephalon morphology. 1(a) A principal component analysis plot of the morphology of the telencephalons of the threespine sticklebacks sampled in (Park & Bell, 2010). The plot represents the sample means of the study populations. Only the populations shared with this study are included for clarity. 1(b) The telencephalon and body shape difference between benthic and limnetic ecotype. Generalist refers to limnetic ecotype. Modified from (Park and Bell, 2010). All the study lakes have native predatory fish (Park and Bell, 2010). The benthic stickleback in Mud lake is sympatric with, but reproductively isolated from, anadromous threespine stickleback (Karve et al., 2008). The fish used in this study were confirmed to be resident benthic ecotype. All the telencephalons were collected from females to avoid the confounding factors of sex differences. The number of confirmed reproductively active sampled individuals varied among the populations (Table 1). Two samples in this study, one individual from Tern Lake and one from Lynne Lake were infected with Glugea anomala (microsporidia) (Table 1). 18 1.6 Aim of the thesis In this thesis, I have examined transcriptome profile differences between benthic and limnetic threespine stickleback telencephalon and discussed the differences in ongoing functional processes in the study populations. I have also discussed the possible ecological reasons for the detected processes as well as their compatibility with ‘clever foraging hypothesis’. The aim of the thesis was not to test the impact of any specific selective force on the adaptive radiation of threespine stickleback. Instead, I aimed to examine how gene expression differs in the differentiated ecotypes from distinct levels of habitat complexity, using the assumptions of past studies as a background to explain the detected gene expression differences. Considering the well-established connection between environmental complexity and spatial cognition, I hypothesized up-regulated expression of neuron growth and connectivity related genes and functional pathways in the benthic ecotype of threespine stickleback. This expectation was supported by the earlier detected bigger telencephalon dorsolateral region (Park and Bell, 2010) and better overall performance in a spatial learning experiment (Park, 2013) in the benthic populations. 19 2 Material and methods 2.1 Sample collection and RNA extraction The samples for the thesis were collected prior the start of the thesis. Telencephalons from female threespine sticklebacks from benthic (Mud and Tern) and limnetic (Lynne and South Rolly) populations in Alaska were collected during breeding season mid-June 2010 between 1-7 pm, using non-baited minnow traps. The fish were euthanized by decapitation and the telencephalons were immediately dissected and placed in 1,5 µl RNase free Eppendorf tubes. The tubes were instantly placed in dry ice. The samples were kept at -70°C until extraction to avoid RNA degradation. All the equipment used in sample collection was treated with 70 % ethanol and RNase Away (ThermoFisher) prior use. The RNA was extracted in October 2010 - May 2011 with a phenol-based phase separation, following the RNA isolation protocol from the Tri-reagent manufacturer (Ambion, Applied Biosystems) (TRI Reagent® Solution Protocol (PN 9738M Rev D), 2010). The protocol is explained in detail in appendix 1. After extraction, the samples were treated with RNase-free DNase, RQ1 1U/µl (Promega), following the manufacture’s protocol (Appendix 1) and the isolation protocol was conducted again immediately after DNase treatment, with downscaling of reagent amounts for 400 µl Tri-reagent, to remove any traces of DNase. Finally, the RNA was dissolved in 15 µl NF water. After dissolving, the samples were stored at -70°C until processing for microarray. All reagents used for RNA isolation were molecular grade. The glassware used for storing the working solutions was treated with 0.1 % diethylpyrocarbonate (DEPC), incubated at room temperature for 12 hours, and autoclaved, to inactivate RNase enzymes. The pipette tips and Eppendorf tubes were RNase free. All the surfaces and equipment were wiped with 70 % ethanol and RNase Away before extraction to avoid contamination. RNA purity and yield were checked with the NanoDrop ND-1000 spectrophotometer (Thermofisher). RNA quality was further assessed with automated gel-based electrophoresis with Experion (Bio-Rad) in Turku Centre for Biotechnology. Only samples with spectrophotometer purity value 260/280 ~2.0 and RNA integrity value ≥7 were sent for microarray analysis. For sending, an aliquot of 600 – 800 ng RNA/sample was precipitated 20 with 3 M sodium acetate and stored in 100 µl of 75 % ethanol, to preserve RNA integrity during transport. For the microarray hybridization, 100 ng RNA/sample were used. 2.2 Microarray analysis Four samples from each of the study populations were used for the analysis. The samples were processed in the UHN microarray center, Toronto, Canada, and analyzed on a custom two-color, oligonucleotide, 8x60k Agilent array. The platform consists of 60 bp, custom- designed probes, which are designed based on previously sequenced DNA sequences. The microarray design was a modified version of a design described in (Leder et al., 2009). The study array is publicly available in the ArrayExpress repository, accession A-MTAB-573 (Agilent ID 041818). Briefly, the microarray consisted of 1319 controls and 61597 probes, which were present in 3 or 10 replicates. Instead of splice variant specific probes, consensus probes were chosen. To measure the gene expression level differences between the ecotypes, 100 ng RNA per sample was labeled with Cy3 or Cy5 and hybridized against a sample from another ecotype in a microarray. After that, the fluorescence intensity of the probe spots in green and red channels was measured and used as a proxy for the probe expression level in the hybridized sample (Two-Color Microarray-Based Gene Expression Analysis (Low Input Quick Amp Labeling) Protocol, 2015). The array was designed with dye swap between the ecotypes to control dye-specific bias (Appendix 2). 2.3 Differential expression analysis Before differential expression analysis, the technical quality of microarray experiments was evaluated. Agilent microarrays have a set of negative control probes for evaluating background noise levels due to unspecific cRNA binding and chemical emission. They also use positive control probes with the addition of spike-ins, the RNA sequences known to hybridize with the probes. These control features together with the signal intensity distribution of non-control probes are used for testing the microarrays’ technical performance (Agilent technologies Inc., 2015). The microarrays used in this study were labeled as ‘good’ technical quality in all evaluation metrics, except for array 25418180002 1-2, where the slope of the expected/observed log ratio of Spike In signal intensity was slightly below the ‘good’ technical quality threshold. 21 Limma -package (Ritchie et al., 2015), designed for statistical program R (R Core Team, 2014) especially for linear modelling of microarray data, was used to analyze the gene expression differences among the populations. Prior to model construction, the data was normalized between arrays and filtered based on signal intensity across probe replicates. All the saturated probes and control probes were removed from the data and the signals were normalized across arrays using the method Aquantile. There was no apparent bias in the MA- plots of log2FC x log2 average intensity in any of the arrays, showing no need for within array normalization (Appendix 2). Only probes with at least two replicates flagged as being well above the background signal and median log2 raw expression value being >5.4 were included in the final analysis to ensure data reliability. Population clustering was examined by creating a heatmap and a dendrogram from raw expression values using R-package Heatmaply (Galili et al., 2018). Subsequently, a principal component analysis (PCA) using raw expression values was performed with Limma-package, method plotMDS (gene.selection = ”common”), with all the expressed and with differentially expressed genes (DEG). A linear regression model was fitted to median signal intensities between the probes from each population comparison, using habitat as the factor. The model fit was evaluated by the empirical Bayes method. Finally, the p-values were corrected for multiple comparisons using Benjamini-Yekutieli correction under dependency (Yekutieli and Benjamini, 2001) to control false discovery rate. All the p-values reported in this thesis are adjusted, unless stated otherwise. 2.4 Gene set enrichment analysis 2.4.1 Probe sequence annotation The assignment of the probes to the Gasterosteus aculeatus genome was based on Ensembl release version 103 (Yates et al., 2020) and version 104 (Howe et al., 2021), stickleback genes (BROAD S1) reference genome using BLASTn (Basic Local Alignment Search Tool for nucleotides). From the blast, a full-length (60 base), 100 % identity, single alignment was present for 11 536 probe sequences and similar alignment with over 50 base length for additional 17 sequences of the total of 11 698 expressed probes. For 44 sequences, there were multiple alignments present. Most of these had the same gene product. For hits with different gene products, the ones with tissue-specific expression in the subject tissue were chosen. For 22 101 sequences, no alignment could be found or assigned reliably in the reference genome. From the 11 597 successfully aligned probes 253 assigned to duplicate genes. The human orthologs for threespine stickleback genes were annotated using BioMart, Ensembl. One to one orthologs were present for 5 866 stickleback genes, one to many for 2 933 genes, and many to many for 405 genes. In case of multiple matches, the best human ortholog match was chosen based on GOC-score and whole-genome alignment score. For the stickleback genes for which no ortholog were returned from Ensembl BioMart, the annotation was supplemented with the Basic Local Alignment Search tool for proteins, (BLASTp) from the stickleback gene encoded protein to human protein in the Uniprot database (Bateman et al., 2021). Only the matches with e-score below 1e-50 and reviewed protein function were accepted. This way, 9 341 human orthologs could be assigned to expressed stickleback genes. 2.4.2 Functional enrichment analysis Gene set enrichment analysis (GSEA) and the subsequent semantic similarity clustering was performed using ViSEAGO package (Brionne et al., 2019) from Bioconductor, R, following the pipeline from ViSEAGO vignette. The enrichment was performed with fgseamultilevel algorithm (Korotkevich et al., 2021), with minimum Gene Ontology (GO) term size 10 for biological process and 5 for molecular function, due to GO term count difference between the categories. The algorithm returns enriched GO terms with adjusted p-values and IC-value. The gene set for GSEA was organized to a ranked list by multiplying the -log10 of the adjusted p-value of each probe by the sign of log fold change of the probe between the ecotypes from the differential gene expression model. All the expressed genes were included in the analysis. The method has no background gene set. In the analysis, the gene sets in the positive and negative extremes of the ranked list are weighted in the enrichment score calculation to avoid high enrichment scores in the middle of the list (Subramanian et al., 2005). GO annotations for the analysis were retrieved from Ensembl version 104 database (Howe et al., 2021). Human orthologues were used for superior annotation (20 358 biological process GO terms in Ensembl) compared to stickleback genes (BROAD S1) database (12 055 biological process GO terms) (Howe et al., 2021). Hierarchical clustering based on semantic similarity of enriched GO terms was performed using Ward D2 aggregation method (Ward, 1963) with Resnik’s distance (Resnik, 1995). The distance between GO clusters was computed using Best-Match Average (BMA) method. 23 3 Results 3.1 Differential gene expression When exploring between-array normalized median raw expression values, most of the probes showed similar expression patterns in all the populations (Figure 4). Based on raw expression values, the three populations from Matanuska – Susitna valley (South Rolly, Mud, and Lynne) clustered together, apart from the Kenai peninsula population (Tern). Figure 3. Heatmap based on between-array normalized median raw expression values of all the expressed probes using method percentize and eucledian distance. The data shows little ecotype- specific difference in gene expression. The populations from Matanuska - Susitna (South Rolly, Mud, and Lynne) cluster together. Principal component analysis (PCA) with all the expressed probes showed ecotype-specific division in PC 2 (Figure 5a), apart from one outlier in the limnetic group. There was not much population-level clustering in the first two PCs. The method used here does not return the percentage of variation explained by the principal components. In the PCA based on DEGs, the populations from each ecotype clustered together on the PC 1 axis (Figure 5b). Lynne and Mud, and South Rolly and Tern, respectively, clustered in PC 2, apart from one outlier in both limnetic populations (Figure 5b). Overall, the limnetic populations showed tighter clustering than the benthic populations. 24 Figure 4. MDS-plots from between-array normalized median raw expression values. Figure 5a represents principal component analysis from all the expressed genes and 5b from differentially expressed genes adjusted p<0.05. Figure 5a shows weak separation between the ecotypes in PC 2. In figure 5b, the two ecotypes separate in PC 1. Limnetic populations cluster closer together, with one outlier in the Lynne population, whereas benthic populations form two clusters in PC 2. In total 11 698 probes (57 % of the total number hybridized in microarray) were included in the differential expression model. Of those, 434 (9.4 %) were differentially expressed (p<0.05). Of differentially expressed probes, 42 % were upregulated in the benthic populations and 58 % in the limnetic populations (Figure 6). 25 Figure 5. Volcano plot displaying differentially expressed genes between ecotypes, based on probe expression model p-values and fold change between ecotypes. The red line is the significance threshold p=0.05. Of the 22 top DEGs with p<0.01, nine were upregulated in benthic (Table 3) and 13 in limnetic ecotype (Table 4). Of those genes, three genes upregulated in the benthic group (ENSGACG00000011017, ENSGACG00000014279, and ENSGACG00000018099) and one gene in the limnetic group (ENSGACG00000007572), lacked human ortholog and GO annotation. The most significant DEG (logFC=4.51, p<0.0007), ENSGACG00000020394, MINK1 (MAP4K4, Mitogen-Activated Protein Kinase 6), was upregulated in the benthic ecotype (Table 3). It is a protein kinase, which gene product is an upstream regulator of JNK in neural cells (Larhammar et al., 2017). It has been associated with variety of other regulatory functions on other cell types as well (Yue et al., 2016). Two other well-characterized benthic upregulated genes, protein phosphatase PTEN and guanine nucleotide binding protein GNB2, are upstream regulators in protein kinase B (AKT) cascade (Lasarge and Danzer, 2014). Hence, all three are involved in MAPK signaling cascades that interact with p38 signaling pathway and participate in cell signaling. In addition to cell signaling, another function observed in the benthic top genes was compound transport. It is associated with gene products of EXOC6B, NAPB, and WASHC3. All the characterized benthic top genes (p<0.01), except for WASHC3, engaged in signal transduction or synaptic transmission. 26 Table 2. Top genes (adjusted p<0.01) upregulated in the benthic ecotype Go terms from Ensembl release 104 human ortholog annotations. *GO term from stickleback broad s1 annotation. Stickleback gene ID Human ortholog gene symbol Uniprot ID GO terms ENSGACG00000020394 MINK1 MINK protein kinase activity; transferase activity; phosphorylation; chemical synaptic transmission; neuron projection morphogenesis; actin cytoskeleton reorganization; positive regulation of JNK cascade; positive regulation of p38MAPK; regulation of AMPA receptor activity cascade ENSGACG00000020395 GNB2 GNB2 GTPase activity; G protein-coupled receptor signaling pathway; calcium channel regulator activity; signal transduction ENSGACG00000018808 EXOC6B EXOC6B exocytosis; Golgi to plasma membrane transport ENSGACG00000011818 NAPB NAPB SNARE complex disassembly; syntaxin binding; vesicle-mediated transport; synaptic transmission, glutamatergic ENSGACG00000009998 PTEN PTEN negative regulation of protein phosphorylation; positive/negative regulation of apoptotic process; neuron- neuron synaptic transmission; synapse assembly; regulation of neuron projection development; central nervous system myelin maintenance; protein kinase B signaling ENSGACG00000019865 WASHC3 CCDC53 WASH complex; protein transport; actin filament polymerization ENSGACG00000018099 integral component of membrane* The most upregulated gene in the limnetic group, ENSGACG00000012847, CATIP (Ciliogenesis Associated TTC17 Interacting Protein) (Table 4), is involved in ciliogenesis and actin filament organization (Bontems et al., 2014). Three other top genes upregulated in limnetic ecotype, COL5A3, GDPD2, and NTNG2 are also involved with neuron projection and cell morphological organization. XRCC2 and SPIN1 are annotated with cell cycle. Important GO term groups in DEGs of limnetic ecotype are also histone 27 methylation/demethylation and transcription regulation. AASDHPPT, ESD, SAMM50, and CRYM participate in reversible histone methylation process. SPIN1 and CRYM function as transcription activity regulators. Table 3. Top genes (adjusted p<0.01) upregulated in the limnetic ecotype Go terms from Ensembl release 104 human ortholog annotations. *GO term from stickleback broad s1 annotation. Stickleback gene ID Human ortholog gene symbol Uniprot ID GO terms ENSGACG00000012847 CATIP C2ORF62 actin filament polymerization; cilium organization; cell projection organization ENSGACG00000016535 COL5A3 COL5A3 collagen fibril organization; cell- matrix adhesion ENSGACG00000008096 ESD ESD formaldehyde catabolic process; methylumbelliferyl-acetate deacetylase activity; glutathione derivative biosynthetic process ENSGACG00000019330 SAMM50 SAMM50 SAM complex; MIB complex; protein insertion into mitochondrial outer membrane; integral component of membrane ENSGACG00000017135 GDPD2 GDPD2 lipid metabolic process; glycerophosphoinositol inositolphosphodiesterase activity; actin filament reorganization ENSGACG00000020819 NOP16 NOP16 RNA binding; ribosomal large subunit biogenesis ENSGACG00000007572 integral component of membrane* ENSGACG00000016021 NTNG2 NTNG2 synaptic membrane adhesion; intercellular bridge; regulation of neuron projection arborization; regulation of neuron migration; modulation of chemical synaptic transmission ENSGACG00000016430 SEC11A SEC11A proteolysis; signal peptide processing; serine-type endopeptidase activity ENSGACG00000008154 AASDHPPT AASDHPPT lysine biosynthetic process via aminoadipic acid; magnesium ion binding; transferase activity 28 3.2 Functional enrichment analysis Gene set enrichment analysis on all the expressed genes returned 59 enriched GO terms (Figure 7); 27 GO terms were enriched in the limnetic and 32 in the benthic ecotype. They clustered into 13 semantic similarity groups (Figure 8). After multiple testing adjustment, the six strongest supported GO terms remained significantly enriched (Table 5). These were lipid transport, GO:0006869, neutrophil degranulation, GO:0043312, innate immune response, GO:0045087, and negative regulation of protein kinase B signaling, GO:0051898 in the benthic ecotype and non-motile cilium assembly, GO:1905515 and regulation of smoothened signaling pathway, GO:0008589, in limnetic ecotype. Most of the semantic similarity clusters were shared among the ecotypes, indicating similar functions being active in both groups (Figure 8). However, two clusters in the benthic ecotype, ‘lipid transport’ and ‘endocytosis’ (clusters 7 and 8) were distinctive for benthic ecotype, while ‘cell cycle’, ‘cell projection organization’ and ‘multicellular organism development’ (clusters 4, 5 and 9) were specific for limnetic ecotype (Figure 8). In shared clusters, most limnetic enriched GO terms were cell cycle related. In total 16 of 27 enriched GO terms in the limnetic ecotype were directly associated with cell cycle or morphology and projection, and additional five terms with transcription (Figure 7). The benthic ecotype was enriched for more diverse functions, including immune reactions, electron transfer, and regulation of signaling cascades (Figure 7). Both ecotypes shared similar functions in cluster 12, ‘programmed cell death’. ENSGACG00000001170 XRCC2 XRCC2 positive regulation of neurogenesis; DNA-dependent ATPase activity; mitotic cell cycle; DNA repair; replication fork ENSGACG00000009554 CRYM CRYM negative regulation of transcription by RNA polymerase II; lysine catabolic process; NADP binding; oxidoreductase activity; thyroid hormone transport ENSGACG00000018280 SPIN1 SPIN1 positive regulation of Wnt signaling pathway; positive regulation of transcription, DNA-templated; methylated histone binding 29 Figure 6. Gene set enrichment heatmap of enriched biological process GO terms with Resnik method semantic similarity clustering. Figure 7. Multidimensional scaling plot from the semantic similarity of the enriched ‘biological process’ GO term clusters using Resnik clustering method and BMA distance. The clusters are labeled by the ecotype the GO term cluster is enriched in. B=benthic, L=limnetic. 30 GSEA for molecular functions returned 19 enriched terms (Figure 9). Three terms, DNA binding, GO:0003677, gamma-tubulin binding, GO:0043015, and tumor necrosis factor receptor binding, GO:0005164 were significantly enriched after multiple testing adjustment (Table 5). They were all enriched in limnetic ecotype. The most significantly enriched term in benthic ecotype was 4 iron, 4 sulfur cluster binding, GO:0051539 (Table 5). The molecular function enriched terms clustered in 5 functional groups, three of which were shared among ecotypes (Figure 10). The ecotype-specific clusters were cluster 2, ‘metal cluster binding’ for benthic ecotype and clusters 3 and 5, ‘signaling receptor binding’ and ‘DNA binding’ for the limnetic ecotype, respectively. In the shared clusters, the molecular functions were mainly DNA and chromatin binding in the limnetic ecotype, and protein and lipid binding and phosphatase activity regulation in the benthic ecotype (Figures 9 and 10). Figure 8. Gene set enrichment of molecular functions GO terms. The enriched terms are clustered by semantic similarity using Resnik method. 31 Figure 9. MDS plot of molecular function GO terms semantic similarity clusters using Resnik method and BMA distance. Table 4. FGSEA enriched GO terms at significance level p<0.1. GO class: BP=Biological process, MF=Molecular function. Ecotype: B=benthic, L=limnetic. GO cluster GO class Ecotype GO ID GO term Condition padj 7 BP B GO:0006869 lipid transport 0.01 10 BP B GO:0043312 neutrophil degranulation 0.03 10 BP B GO:0045087 innate immune response 0.03 13 BP B GO:0051898 negative regulation of protein kinase B signaling 0.04 2 BP B GO:0016226 iron-sulfur cluster assembly 0.05 7 BP B GO:0120009 intermembrane lipid transfer 0.07 6 BP B GO:0070584 mitochondrion morphogenesis 0.08 11 BP B GO:0090023 positive regulation of neutrophil chemotaxis 0.08 6 BP B GO:0032981 mitochondrial respiratory chain complex I assembly 0.1 5 BP L GO:1905515 non-motile cilium assembly 0.03 32 GO cluster GO class Ecotype GO ID GO term Condition padj 13 BP L GO:0008589 regulation of smoothened signaling pathway 0.03 3 BP L GO:0006260 DNA replication 0.09 2 MF B GO:0051539 4 iron, 4 sulfur cluster binding 0.07 5 MF L GO:0003677 DNA binding 0 4 MF L GO:0043015 gamma-tubulin binding 0.01 3 MF L GO:0005164 tumor necrosis factor receptor binding 0.01 33 4 Discussion 4.1 Population differentiation When examining the eucledian distance of raw expression values, there was weak clustering by populations and geographical regions, but not by geographical proximity within regions (Figure 1, Figure 4). There is no direct contact between the Matanuska – Susitna area study drainages, which excludes the possibility of gene flow between the study lakes. The separation of the Tern Lake population from the other study populations can be explained by its location at higher elevation and further inlands, as well as by its narrower genetic diversity (Park, 2013). Genetic clustering by watersheds and geographical proximity within watersheds has previously been detected in threespine stickleback populations from Vancouver Island, British Columbia (Bolnick and Ballare, 2020). Although the distinction between the gene expression profiles of the Matanuska - Susitna area populations and the Tern Lake population was not strongly supported (Figure 4 and 5a), it may have diminished the statistical power in detecting transcription level differences between the ecotypes. Despite the geographical factor, PCA with DEGs supported the presence of ecotype-specific gene expression pattern (Figure 5b). This is concordant with the differences in telencephalon shape and size (Figure 2), which has been previously reported in these populations (Park and Bell, 2010). The greater distribution between the benthic populations compared to the limnetic populations in the PCA with DEGs may relate to above discussed factors concerning geographical location of the benthic populations or differences in the reaction norms of the populations. Sympatric Mud population (Introduction 1.5) may experience more extreme habitat specialization compared to allopatric populations. The anadromous ecotype sympatric with the Mud population resembles limnetic ecotype (Karve et al., 2008). Non-hybridizing, sympatric populations are often more specialized than allopatric populations (Svanbäck and Schluter, 2012). The values in PC 2 were more extreme in the Mud Lake population than the in the Tern Lake population (Figure 5b). On the other hand, the lack of genetic diversity in the Tern Lake population may show here as a separating factor. These two populations have previously been found to differ in spatial performance (Park, 2013) (Introduction 1.5). Interestingly, the populations with poor spatial performance and the populations with good spatial performance in Park (2013) clustered in PC 2 based on DEGs in this study (Figure 5b). 34 4.2 Ecotype-specific patterns in neural regulation 4.2.1 Upregulated neurogenesis and transcription in the limnetic ecotype The enrichment of cell cycle and projection associated gene sets (Figure 8) indicates ongoing active engram formation in the limnetic ecotype. Engram formation is accompanied by neural cell proliferation and structural changes in neurons (Introduction 1.2.2). Gamma tubulin, the target for a significantly enriched molecular function in the limnetic ecotype, ‘gamma tubulin binding’, is a primary compound in the formation of mitotic spindle (Alvarado-Kristensson, 2018). It also forms nucleating centers that initiate microtubule formation (Alvarado-Kristensson, 2018), which in turn contribute to cytoskeleton formation and direct neuron growth and morphology (Kapitein and Hoogenraad, 2015). Non-motile cilia, the target for ‘non-motile cilium assembly’, which was significantly enriched in the limnetic ecotype (Table 5), act as cell antennas. They are essential for various biological processes ranging from non-synaptic neuronal signaling in mature neurons to cell differentiation and projection and axon guidance (Guemez-Gamboa et al., 2014). Microtubules are central structural elements in cilia (Kapitein and Hoogenraad, 2015). The most significant DEG in the limnetic ecotype, CATIP (Table 4), is involved in actin polymerization modulation and cilia initiation (Bontems et al., 2014). The importance of cell cycle regulation in the limnetic ecotype is further consolidated by the enrichment of the biological process ‘DNA replication’ and several non-significantly enriched biological processes (Table 5, Figures 7-8). The only limnetic enriched signaling pathway, ‘smoothened signaling’ (Table 5), is signal transduction in the ‘sonic hedgehog’ pathway, which is associated with neuron differentiation, proliferation, and axonal elongation (Delmotte et al., 2020). ‘Smoothened signaling’ activation itself promotes GABAergic neuron transition from excitatory to inhibitory stage during maturation (Delmotte et al., 2020), which is an important regulatory mechanism for synaptic signal transduction and is needed for systemic homeostasis. The enrichment of processes that are associated with neural circuit formation is further supported by the upregulation of two genes from the WNT-signaling pathway in the limnetic ecotype, SPIN1 (Table 4) and REG1, which has the largest fold change of the DEGs (Appendix 3). WNT signaling regulates a variety of neural circuit formation processes (Rosso and Inestrosa, 2013), and is an upstream regulator in mTOR pathway (Lipton and Sahin, 2014). 35 Transcription-related gene sets were non-significantly enriched in the limnetic ecotype (Figure 7). The enrichment of transcription-related processes solely in the limnetic ecotype indicates increased biological significance for transcription regulation in the limnetic group. Gene sets involved in transcription regulation have been demonstrated to be upregulated in memory consolidation and throughout memory formation (Marco et al., 2020). Adult neurogenesis is substantially more frequent in teleost fish than in mammalian brain (Zupanc and Sîrbulescu, 2011), which emphasizes its significance for fish behavior. To stress the connection between neurogenesis and learning, environmental enrichment has been shown to increase proliferation in teleost fish (Lema et al., 2005; von Krogh et al., 2010). Together the evidence from DEGs and enriched functional gene sets indicate the individuals in the limnetic ecotype experiencing conditions that promote learning and memory formation. 4.2.2 Metabolic functions and immune defense enriched in the benthic ecotype Functional groups associated with increased metabolism and immune defense were enriched in the benthic ecotype. Biological processes ‘Lipid transport’, ‘iron-sulfur cluster assembly’, ‘mitochondrion morphogenesis’ and ‘mitochondrial respiratory chain complex I assembly’ are associated with cellular energy production. The strongest supported biological process, ‘lipid transport’ (Table 5), is needed for both intake of fatty acids and localization of synthesized lipids. Twenty percent of the energy needed in the brain has been estimated to be produced by fatty acid oxidation (Tracey et al., 2018). Cellular energy production mainly takes place in mitochondrial respiratory complexes by electron transport in iron-sulfur complexes, which are also assembled in mitochondria (Ciofi-Baffoni et al., 2018; Khacho and Slack, 2018). The function of these complexes is regulated by morphological modifications in mitochondria (Khacho and Slack, 2018). Mitochondrial dynamics is important in the regulation of various energy-demanding neuronal processes because it directly regulates cellular metabolism (Khacho and Slack, 2018). Apart from energy production, lipids in brain have major roles as structural molecules in cellular membranes and in neural signaling (Tracey et al., 2018). Phosphatidylinositol (PPI) signaling was a major compounding factor in the benthic ecotype’s enrichment profile in this study. ‘Phosphatidylinositol biosynthetic process’ was non-significantly enriched in the benthic ecotype. PPIs are lipid species, that have minor roles as membrane structural proteins, but also act as signaling molecules in different signaling cascades and cellular types (Balla, 36 2013). Since PPI signaling is cellular-location dependent, lipid transport and transfer are important in PPI signaling. Phosphorylation state, together with cellular location, determines the regulatory impacts of PPIs (Balla, 2013). ‘Positive regulation of phosphorylation’ was enriched in the benthic ecotype. Phosphorylation of PPIs is regulated by a variety of kinases and phosphatases. The best-studied PPI pathway is pI3k/PTEN regulated pI3k/AKT pathway (Introduction 1.2.3). In this study, both positive and negative regulation of AKT signaling (‘regulation of protein kinase B signaling’) were enriched in the benthic ecotype, but only the enrichment of negative regulation was significant (Figure 7, Table 5). The enrichment of negative regulation of AKT was driven by PTEN (Table 3), a protein phosphatase which inhibits PPI-induced activation of AKT on plasma membrane (Balla, 2013) and downregulates the activation of downstream mTOR pathway (Introduction 1.2.3). Besides pI3k/PTEN regulation, AKT activity is regulated by variety of other signaling cascades (Balla, 2013) and in turn participates in downstream regulation of several processes, including neurodegeneration, apoptosis and synaptic strength (Brazil et al., 2004). AKT signaling in a MAPK-family protein kinase c-Jun N-terminal kinase (JNK) pathway promotes cell survival and prevents apoptosis (Brazil et al., 2004; Larhammar et al., 2017). Interestingly, the strongest upregulated DEG in the benthic ecotype, MINK, participates in activation of neurodegenerative signaling cascade in JNK pathway (Larhammar et al., 2017). Within this pathway, MINK is also involved in induction of depotentiation after LTP by mediating AMPA-receptor removal from synapses (Zhu et al., 2005) and suppressing dendritic spines and excitatory synapses (Han et al., 2010). JNK signaling was not enriched in this study, but it strongly interacts with p38MAPK pathway, a stress induced pathway involved in e.g., neurodegeneration and regulation of synaptic plasticity, notably the astrocyte-mediated induction of LTD (Navarrete et al., 2019; Falcicchia et al., 2020). ‘Positive regulation of p38MAPK cascade’ was non-significantly enriched in the benthic ecotype (Figure 7). PI3k/AKT pathway is also involved in the regulation of ‘neutrophil chemotaxis’ and ‘neutrophil degranulation’ (Table 5) enriched in the benthic ecotype (Heit et al., 2002; Lodge et al., 2020). Neutrophils, which migrate to brain via bloodstream, attend to innate immune defense together with astrocytes and microglia (Kim et al., 2020). They move to targets by chemotaxis and can destroy the target by releasing proteolytic enzymes, reactive oxidative 37 species (ROS), and stimulating inflammatory cytokines (Rane et al., 2005). Together with pI3k/AKT signaling, p38MAPK pathway participates in regulating neutrophil chemotaxis (Table 5) (Heit et al., 2002). Neutrophil chemotaxis is regulated by the mitochondrial electron transport chain and redox status (Zhou et al., 2018). Thus, the enrichment of neutrophil chemotaxis, mitochondrial dynamics, and the assembly of electron transfer machinery have a complementary function. The enrichment of immunological functions is also supported by the gene with the greatest fold change in the benthic group, GIMAP7 (logFC=5.0, p=0.46), a member of immune-associated subfamily of nucleotide-binding proteins. The pathways that were enriched in the benthic ecotype, together with mitochondrial functions, are essential for maintaining systemic homeostasis, but also for various organismal processes, including neuronal dynamics in engram formation (Chen et al., 2005; Khacho and Slack, 2018; Tracey et al., 2018; Falcicchia et al., 2020) and immune defense. Immune defense and memory formation share elements (Falcicchia et al., 2020; Cornell et al., 2022) and organismal stress has been demonstrated to negatively influence learning (Wood et al., 2011) and brain cell proliferation (Øverli and Sørensen, 2016). Hence, the environment- organism interaction does not appear to promote memory formation through active gene expression in the benthic ecotype during the sampling time. 4.3 Potential sources for the ecotype-specific gene expression profiles 4.3.1 Limnetic populations The upregulation of cognitive processes in the limnetic ecotype can be explained by both environmental and social complexity. Benthic populations forage in complex littoral environments throughout their lifespan, while limnetic populations have been reported to use mostly pelagic prey outside the breeding season, although the distinction is not stark (Introduction 1.4). It is possible, that limnetic individuals undergo habitat transition in the breeding season and as a result experience increased engram formation and restructuring compared to benthic individuals in familiar environments. Habitat generalists have indeed been suggested to outperform specialists in cognitional performance (Henke-von der Malsburg et al., 2020). Outside the breeding season, benthic populations live in patchier habitats, but given the temporal aspect, limnetic populations may experience more variable environment. However, the potential of habitat transition as a factor in the transcriptomic difference is diminished by the discovery of the recent diet change from pelagic feeding to 38 feeding on benthos in the Lynne population (Park et al., 2012), although the topography of Lynne Lake and the telencephalon shape of the Lynne population are classified as limnetic (Park and Bell, 2010). On the other hand, the enrichment of cognition-related functions can be explained by more variable habitat use throughout the year in the limnetic populations, as suggested by Park & Bell (2010). Besides feeding behavior, increased level of sociality in the limnetic populations in adulthood compared to the benthic populations, which are more solitary, may be a key factor in promoting learning and memory. Interestingly, in the comparison between male and female telencephalon, Park and Bell (2010) found the telencephalons of males, besides being bigger than in females morphologically resembled that of limnetics. Since threespine stickleback males show paternal care, they experience more socially complex environment than females. Park and Bell (2010) have suggested social factors as possible explanation for the telencephalon size fluctuation among populations and sexes. The triangular shape is concurrent with the idea of neuron growth in the dorsomedial region of the telencephalon, which is primarily considered to be responsible for social and emotional processes (Demski, 2013). Biological processes that are associated with neuron growth and projection have been found to be upregulated in socially reared threespine stickleback females (Greenwood and Peichel, 2015). As discussed in the introduction (1.1.3, 1.3.1), social interactions have repeatedly been shown to have major impact on cognitive processes. Together, the concave shape of limnetic telencephalon and enrichment of learning- and memory-related functions in this study emphasize the possibility of the impact of social interactions on cognition of limnetic females. 4.3.2 Benthic populations Several lake and organismal features may contribute to the detected enrichment of the immunological and metabolic functions in the benthic ecotype. First, parasite abundance, parasite community, and parasites’ ability to infect and exploit their hosts may vary among the lakes and the habitats. Stutz et al. (2015) have discovered significant lake- but also habitat-specific, plastic differences in immune gene expression between threespine stickleback populations from small and large lakes. The discovery underlines the impact of parasite community structure on host’s gene expression. Benthic and limnetic threespine sticklebacks have been suggested to differ in their exposure to limnetic and benthic parasites due to different proportions of consumed intermediate hosts (MacColl, 2009; Stutz et al., 39 2014), There is a complex relationship concerning exposure rates and parasite infections, though, perhaps because of avoidance behavior or resistance to infection (Stutz et al., 2014; Arostegui et al., 2018). Second, sympatric limnetic/benthic populations have been hypothesized to differ in their investment in immune defense (MacColl, 2009). Exposure to different parasites, together with host genetics and exposure history have been shown to impact the activation, strength, and mode of immune defense in threespine stickleback populations (Kalbe and Kurtz, 2006; Lenz et al., 2013; Piecyk et al., 2021). Functional groups that have previously been enriched in threespine stickleback upon parasite exposure include immunological and metabolic processes (Lenz et al., 2013). Infection with a parasite has been demonstrated to alter the expression of inositol pathway in threespine stickleback brain, simultaneously with a behavioral change (Grecias et al., 2020). While the inositol pathway was not enriched in this study, PPIs enriched in the benthic ecotype (Discussion 4.2.2) are an important part of inositol signaling. One individual from both ecotypes in this study was observed to be infected with a parasite. However, the early stages of infection are difficult to detect, and mere exposure to a parasite has been shown to be sufficient to alter gene expression in the brain (Grecias et al., 2020) and the head kidney (Piecyk et al., 2021). The head kidney has been suggested to be involved in the redistribution of energy substrates during immune response in teleost fish (Geven and Klaren, 2017). It must be noted that previous studies describing transcriptional response to parasite exposure in threespine stickleback are mainly transplant studies between novel parasite/host combinations and lab-bred, previously unexposed individuals. Thus, they are not readily comparable to wild populations in their native habitats where threespine stickleback populations are adapted to their local parasite communities. The other potential source for the need of immune defense stem from behavior and physiological stress associated with the breeding season. Stressful courtship behaviors, such as biting and chasing females, exist in both ecotypes, but are more common in the benthic ecotype (Rowland, 1994; Shaw et al., 2007). Social stress has been demonstrated to increase neutrophil mobilization in mice (Ishikawa et al., 2021). The third possible explanation for the enrichment of immunological functions is the consumption of algal toxins. The high ratio of surface area to water volume makes shallow lakes more susceptible to warming than deeper lakes (Moran et al., 2010). Warming, especially together with thermal stratification and eutrophication, exposes them to toxic algal 40 (including cyanobacteria) blooms (Vilas et al., 2018). While the lakes in the study area are mostly oligo-or mesotrophic (Jones et al., 2003), toxic algae are frequently encountered in Alaskan lakes. The impact of lake morphology and rooted vegetation on algal blooms is complex. Lakes with widespread rooted vegetation, as both benthic lakes in this study, are often less phytoplankton dominated (Bayley and Prather, 2003; Vilas et al., 2018). Shallow lakes are also less prone to thermal stratification than deeper lakes. On the other hand, infrequent thorough mixing promotes the release of nutrients from lake sediment for algae use (Vilas et al., 2018). Panmictic conditions also promote the mixing of toxic algae throughout the water column, thus making it difficult to avoid their consumption. Algae and cyanobacteria toxins have direct immuno- and neurotoxic effects on teleost fish (Banerjee et al., 2021). Exposure to cyanobacteria toxin microcystin has been shown to have multifaceted impacts on the fish immune system, depending on dosage and exposure route, including up- and downregulation of innate immune response, induction of inflammatory reaction, and apoptosis (Lin et al., 2021). Finally, a possible cause for the enrichment of lipid transport and metabolism is physiological stress that threespine stickleback females experience because of energy-demanding, repeated spawning and egg production (Wootton, 1994; Foster et al., 2008). Benthic fish have been suggested to produce greater clutch size per spawning than limnetics (Bell and Foster, 1994). The difference in egg production is predicted to even out during the spawning season (Bell and Foster, 1994), but the one-time energy investment and thus energy demand remains higher in benthic than in limnetic females. To meet the energy demand, females increase their energy intake, but also consume systemic lipid and glycogen reservoirs and decrease their growth rate (Wootton, 1994). While there is no evidence for food availability limitation in the benthic populations, oxygen depletion has been suggested to limit feeding ability in male threespine stickleback during the breeding season (Guderley, 1994). A similar mechanism may shift cellular energy production in brain from external sources to the use of systemic lipid reservoirs in benthic females and contribute to the detected enrichment of lipid metabolism; however, no directly hypoxia associated processes were detected in this study, although benthic functional enrichment shared similarities with hypoxia associated decrease in DNA replication and mitosis and upregulation of oxidase and oxygenase activity (Leveelahti et al., 2011). The enrichment of stress-related functions in geographically distant populations of Mud and Tern lakes is interesting. It is an intriguing question whether the ecotype-specific expression 41 of immune defense is triggered by environmental factors, which are either correlated with lake topography or pathogen community, or by genetic differences between the ecotypes in the immune response reaction norm. Genomic markers near genes linked to both immunology and cognition has been found to vary in frequency in threespine stickleback lake populations in relation to lake size (Bolnick and Ballare, 2020). 4.4 Elements of uncertainty 4.4.1 Populations and tissue collection Sampled populations in this study represent benthic and limnetic threespine stickleback populations. The ecological specialization (Park and Bell, 2010) in allopatric study populations is not as extreme as in sympatric populations of the same ecotypes and one of the limnetic populations has atypical feeding habits. Also, cognitional performance among the populations (Park, 2013) does not strictly follow the expectations of benthic/limnetic division. This adds uncertainty to the results. However, the telencephalon and body morphology of the study populations, as well as lake morphology, are typical for benthic and limnetic ecotypes, which justifies their use as sample populations in this study. The samples were collected using minnow traps. Although the samples were collected within an hour per site, the time the fish spent in the trap may have varied. This could potentially have had an impact on the stress level of individual fish. This impact could not be controlled. However, it is unlikely that the results are severely affected by the variation in trapping time because there is no evidence for systematic bias in trapping time between the lake types. Fish age may change its transcriptome. While the age of the sampled fish was not determined, a proportion of samples from every population except for Mud was confirmed sexually mature (with eggs). Hence, it is unlikely that these results are driven by age-dependent changes in RNA abundance. 4.4.2 Data analysis Functional enrichment is sensitive to gene set selection and the annotation coverage of selected genes (Primmer et al., 2013; Stanford et al., 2020). The use of annotations from distant phylogenic group is complicated by evolutionary differences among the groups (Primmer et al., 2013). While GO annotations are not species-specific, some uncertainty to their interpretation comes from the fact that many of the regulatory pathways have been 42 studied in mammals and invertebrates, and it is not known how conserved they are between the study organisms and teleost fish (Pravosudov et al., 2013; Sun and Lin, 2016). The use of human annotations in the functional enrichment instead of threespine stickleback or other teleost species increased annotation coverage but reduced the number of genes included in this study. To diminish the ambiguity from human ortholog use, the orthologs were chosen with strict selection criteria (Material and methods 2.4.1). Cognitive processes are relatively conserved in vertebrates. Still, it must be remembered, that the knowledge of regulatory pathways is incomplete. The gene set enrichment method applied in this thesis allows potentially biologically meaningful genes with subtle, non-significant changes in the expression levels to be included. The problem of the enrichment of gene sets with non-significant up- or downregulation that lack phenotypic correlation is resolved by weighting the positive and the negative values in the ranked list of genes (Subramanian et al., 2005). The gene set for the analysis can be ranked by fold change or significance. While the fold change is biologically meaningful, the significance value, which has been used here, better considers the consistency among samples. Because of the weighting of the extreme values, the ranking method may have influenced the enrichment. Since transcriptome is not equivalent to protein expression, it does not accurately describe the phenotypic outcome. Measuring mRNA contents also leaves out non-coding functional RNA (Stanford et al., 2020). Finally, the biologically most meaningful transcripts may not be among the most expressed ones (Stanford et al., 2020). Despite this, RNA profiling has been proven to reveal biologically meaningful differences between populations (Introduction 1.3.2). The use of FGSEA in this study is intended to minimize the effects of the difference between statistical and biological significance. 43 5 Conclusions In this thesis, I have discovered divergent gene expression in the telencephalons of benthic and limnetic threespine sticklebacks from allopatric populations with non-extreme foraging behavior. Contrary to my hypothesis, the limnetic populations in this study were enriched with more neural growth-associated functions compared to benthic populations. This contradicts the presumptions of ‘clever foraging hypothesis’, as benthic foraging habitat has complex elements compared to limnetic habitat. However, it is not clear whether the results in this study have been driven by the enhancement of cognitional processes in the limnetic ecotype or rather by environmentally or socially induced organismal stress in the benthic ecotype, severe enough to trigger energetic trade-off and counteract the impact of environmental complexity on gene expression. Considering the earlier research on the foraging habits and cognition of the study populations, it is plausible that the distinction in the foraging strategy among the ecotypes in these populations is too subtle and inconsistent to be the main driving force of the ecotype-specific transcriptome profiles. Hence, these results do not unambiguously neither support nor oppose ‘clever foraging hypothesis’ in explaining cognitional differences between benthic and limnetic ecotypes in general. Instead, they support the presence of trade-off between energy investment on immunology and cognition. Because of the transient nature of transcriptome and plasticity of both cognitional and immunological traits, these results reflect only a limited time span and a unique combination of environmental factors. Further research is needed to discover whether they can be applied to other populations in the benthic - limnetic axis. The presence of lake size correlated variation on the genetic markers of both traits indicates heritable differences between the ecotypes and potential importance of those traits for adaptive radiation of threespine stickleback in benthic and limnetic lakes (Discussion 4.3.2). Especially interesting question is the extent of environmental versus genetic impact on the expression of innate immune defense among the ecotypes. What is the role of immunology in the adaptive radiation of threespine stickleback? What are the unifying environmental factors that trigger the expression of innate immunology in lakes that are structurally similar, but differ in their geographical location? 44 Acknowledgments I would like to thank my supervisors Erica Leder, Heidi Viitaniemi, and Irma Saloniemi for their endless patience, support, and insightful comments. I would also like to thank Erica for an interesting topic and material for the thesis. A special thanks to Professor Michael A. Bell for sharing his expertise on Alaskan threespine stickleback populations during sample collection. Finally, many thanks to Heidi for everything, especially for teaching me the art of RNA handling, and Tiina Sävilammi for invaluable help with statistical analysis and R. 45 References Abraham, W. C., Logan, B., Greenwood, J. M., and Dragunow, M. (2002). Induction and experience-dependent consolidation of stable long-term potentiation lasting months in the Hippocampus. J. Neurosci. 22, 9626–9634. doi:10.1523/jneurosci.22-21- 09626.2002. Adamsky, A., Kol, A., Kreisel, T., Doron, A., Ozeri-Engelhard, N., Melcer, T., et al. (2018). Astrocytic Activation Generates De Novo Neuronal Potentiation and Memory Enhancement. Cell 174, 59–71. doi:10.1016/j.cell.2018.05.002. Agilent technologies Inc. (2015). Agilent Feature Extraction 11.5 Reference Guide. 328. Available at: https://www.agilent.com/cs/library/usermanuals/Public/G4460- 90045_FE_RefGuide.pdf [Accessed March 16, 2021]. Agulhon, C., Fiacco, T. A., and McCarthy, K. D. (2010). Hippocampal short- and long-term plasticity are not modulated by astrocyte Ca2+ signaling. Science 327, 1250–1254. doi:10.1126/science.1184821. Ahmed, N. I., Thompson, C., Bolnick, D. I., and Stuart, Y. E. (2017). Brain morphology of the threespine stickleback (Gasterosteus aculeatus) varies inconsistently with respect to habitat complexity: A test of the Clever Foraging Hypothesis. Ecol. Evol. 7, 3372– 3380. doi:10.1002/ece3.2918. Alvarado-Kristensson, M. (2018). γ-tubulin as a signal-transducing molecule and meshwork with therapeutic potential. Signal Transduct. Target. Ther. 3, 1–6. doi:10.1038/s41392-018-0021-x. Anggono, V., and Huganir, R. L. (2012). Regulation of AMPA receptor trafficking and synaptic plasticity. Curr. Opin. Neurobiol. 22, 461–469. doi:10.1016/j.conb.2011.12.006. Arechavala‑Lopez, P., Caballero‑Froilán, J. C., Jiménez‑García, M., Capó, X., Tejada, S., Saraiva, J. L., et al. (2020). Enriched environments enhance cognition, exploratory behaviour and brain physiological functions of Sparus aurata. Sci. Rep. 10. doi:10.1038/s41598-020-68306-6. Arostegui, M. C., Hovel, R. A., and Quinn, T. P. (2018). Schistocephalus solidus parasite prevalence and biomass intensity in threespine stickleback vary by habitat and diet in boreal lakes. Environ. Biol. Fishes 101, 501–514. doi:10.1007/s10641-018-0719-1. 46 Ashton, B. J., Ridley, A. R., Edwards, E. K., and Thornton, A. (2018). Cognitive performance is linked to group size and affects fitness in Australian magpies. Nature 554, 364–367. doi:10.1038/nature25503. Axelrod, C. J., Laberge, F., and Robinson, B. W. (2021). Interspecific and intraspecific comparisons reveal the importance of evolutionary context in sunfish brain form divergence. J. Evol. Biol. 34, 639–652. doi:10.1111/jeb.13763. Balla, T. (2013). Phosphoinositides: Tiny lipids with giant impact on cell regulation. Physiol. Rev. 93, 1019–1137. doi:10.1152/physrev.00028.2012. Banerjee, S., Maity, S., Guchhait, R., Chatterjee, A., Biswas, C., Adhikari, M., et al. (2021). Toxic effects of cyanotoxins in teleost fish: A comprehensive review. Aquat. Toxicol. 240, 105971. doi:10.1016/j.aquatox.2021.105971. Bateman, A., Martin, M. J., Orchard, S., Magrane, M., Agivetova, R., Ahmad, S., et al. (2021). UniProt: The universal protein knowledgebase in 2021. Nucleic Acids Res. 49, D480–D489. doi:10.1093/nar/gkaa1100. Bayley, S. E., and Prather, C. M. (2003). Do wetland lakes exhibit alternative stable states? Submersed aquatic vegetation and chlorophyll in western boreal shallow lakes. Limnol. Oceanogr. 48, 2335–2345. doi:10.4319/lo.2003.48.6.2335. Bebus, S. E., Small, T. W., Jones, B. C., Elderbrock, E. K., and Schoech, S. J. (2016). Associative learning is inversely related to reversal learning and varies with nestling corticosterone exposure. Anim. Behav. 111, 251–260. doi:10.1016/j.anbehav.2015.10.027. Bell, M. A., and Foster, S. A. (1994). “Introduction To the Evolutionary Biology of the Threespine Stickleback,” in The Evolutionary Biology of the Threespine Stickleback, eds. M. A. Bell and S. A. Foster (Oxford University Press), 1–27. Bensky, M. K., and Bell, A. M. (2020). Predictors of individual variation in reversal learning performance in three-spined sticklebacks. Anim. Cogn. 23, 925–938. doi:10.1007/s10071-020-01399-8. Bolnick, D. I., and Ballare, K. M. (2020). Resource diversity promotes among‐individual diet variation, but not genomic diversity, in lake stickleback. Ecol. Lett. 23, 495–505. doi:10.1111/ele.13448. Bontems, F., Fish, R. J., Borlat, I., Lembo, F. F., Chocu, S., Chalmel, F. F., et al. (2014). C2orf62 and TTC17 are involved in actin organization and ciliogenesis in zebrafish and human. PLoS One 9, 86476. doi:10.1371/journal.pone.0086476. 47 Brazil, D. P., Yang, Z. Z., and Hemmings, B. A. (2004). Advances in protein kinase B signalling: AKTion on multiple fronts. Trends Biochem. Sci. 29, 233–242. doi:10.1016/J.TIBS.2004.03.006. Brionne, A., Juanchich, A., and Hennequet-Antier, C. (2019). ViSEAGO: A Bioconductor package for clustering biological functions using Gene Ontology and semantic similarity. BioData Min. 12. doi:10.1186/s13040-019-0204-1. Buechel, S. D., Boussard, A., Kotrschal, A., van Der Bijl, W., and Kolm, N. (2018). Brain size affects performance in a reversal-learning test. Proc. R. Soc. B Biol. Sci. 285. doi:10.1098/rspb.2017.2031. Buechel, S. D., Noreikiene, K., DeFaveri, J., Toli, E., Kolm, N., and Merilä, J. (2019). Variation in sexual brain size dimorphism over the breeding cycle in the three-spined stickleback. J. Exp. Biol. 222. doi:10.1242/JEB.194464. Carbia, P. S., and Brown, C. (2019). Environmental enrichment influences spatial learning ability in captive-reared intertidal gobies (Bathygobius cocosensis). Anim. Cogn. 22, 89–98. doi:10.1007/s10071-018-1225-8. Charnov, E. L. (1976). Optimal Foraging, the Marginal Value Theorem. Theor. Popul. Biol. 9, 129–136. Chen, L., Cummings, K. A., Mau, W., Zaki, Y., Dong, Z., Rabinowitz, S., et al. (2020). The role of intrinsic excitability in the evolution of memory: Significance in memory allocation, consolidation, and updating. Neurobiol. Learn. Mem. 173, 107266. doi:10.1016/j.nlm.2020.107266. Chen, X., Garelick, M. G., Wang, H., Li, V., Athos, J., and Storm, D. R. (2005). PI3 kinase signaling is required for retrieval and extinction of contextual memory. Nat. Neurosci. 8, 925–931. doi:10.1038/nn1482. Ciofi-Baffoni, S., Nasta, V., and Banci, L. (2018). Protein networks in the maturation of human iron–sulfur proteins. Metallomics 10, 49–72. doi:10.1039/C7MT00269F. Cornell, J., Salinas, S., Huang, H. Y., and Zhou, M. (2022). Microglia regulation of synaptic plasticity and learning and memory. Neural Regen. Res. 17, 705–716. doi:10.4103/1673-5374.322423. Costa-Mattioli, M., Sossin, W. S., Klann, E., and Sonenberg, N. (2009). Translational Control of Long-Lasting Synaptic Plasticity and Memory. Neuron 61, 10–26. doi:10.1016/j.neuron.2008.10.055. Croston, R., Branch, C. L., Pitera, A. M., Kozlovsky, D. Y., Bridge, E. S., Parchman, T. L., et al. (2017). Predictably harsh environment is associated with reduced cognitive 48 flexibility in wild food-caching mountain chickadees. Anim. Behav. 123, 139–149. doi:10.1016/j.anbehav.2016.10.004. Dall, S. R. X., Giraldeau, L.-A., Olsson, O., McNamara, J. M., and Stephens, D. W. (2005). Information and its use by animals in evolutionary ecology. Trends Ecol. Evol. 20, 187–93. doi:10.1016/j.tree.2005.01.010. Delmotte, Q., Hamze, M., Medina, I., Buhler, E., Zhang, J., Belgacem, Y. H., et al. (2020). Smoothened receptor signaling regulates the developmental shift of GABA polarity in rat somatosensory cortex. J. Cell Sci. 133. doi:10.1242/jcs.247700. Demski, L. S. (2013). The pallium and mind/behavior relationships in teleost fishes. in Brain, Behavior and Evolution (Karger Publishers), 31–44. doi:10.1159/000351994. Dunbar, R. I. M. (1998). The social brain hypothesis. Evol. Anthropol. Issues, News, Rev. 6, 178–190. doi:10.1002/(SICI)1520-6505(1998)6:5<178::AID-EVAN5>3.3.CO;2-P. Dupret, D., Fabre, A., Döbrössy, M. D., Panatier, A., Rodríguez, J. J., Lamarque, S., et al. (2007). Spatial learning depends on both the addition and removal of new hippocampal neurons. PLoS Biol. 5, 1683–1694. doi:10.1371/journal.pbio.0050214. Emlen, J. M. (1966). The Role of Time and Energy in Food Preference. Am. Nat. 100, 611– 617. doi:10.1086/282455. Evans, M. L., Hori, T. S., Rise, M. L., and Fleming, I. A. (2015). Transcriptomic Responses of Atlantic Salmon (Salmo salar) to Environmental Enrichment during Juvenile Rearing. PLoS One 10, e0118378. doi:10.1371/journal.pone.0118378. Falcicchia, C., Tozzi, F., Arancio, O., Watterson, D. M., and Origlia, N. (2020). Involvement of p38 MAPK in Synaptic Function and Dysfunction. Int. J. Mol. Sci. 21, 5624. doi:10.3390/ijms21165624. Fang, B., Kemppainen, P., Momigliano, P., Feng, X., and Merilä, J. (2020). On the causes of geographically heterogeneous parallel evolution in sticklebacks. Nat. Ecol. Evol. 4, 1105–1115. doi:10.1038/s41559-020-1222-6. Fong, S., Buechel, S. D., Boussard, A., Kotrschal, A., and Kolm, N. (2019). Plastic changes in brain morphology in relation to learning and environmental enrichment in the guppy (Poecilia reticulata). J. Exp. Biol. 222. doi:10.1242/JEB.200402. Foster, S. A., Shaw, K. A., Robert, K. L., and Baker, J. A. (2008). Benthic, limnetic and oceanic threespine stickleback: Profiles of reproductive behaviour. Behaviour 145, 485–508. doi:10.1163/156853908792451421. Frankland, P. W., and Bontempi, B. (2005). The organization of recent and remote memories. Nat. Rev. Neurosci. 6, 119–130. doi:10.1038/nrn1607. 49 Frey, U., and Morris, R. G. M. (1998). Synaptic tagging: Implications for late maintenance of hippocampal long- term potentiation. Trends Neurosci. 21, 181–188. doi:10.1016/S0166-2236(97)01189-2. Galili, T., O’Callaghan, A., Sidi, J., and Sievert, C. (2018). heatmaply: an R package for creating interactive cluster heatmaps for online publishing. Bioinformatics 34, 1600– 1602. doi:10.1093/BIOINFORMATICS/BTX657. Geven, E. J. W., and Klaren, P. H. M. (2017). The teleost head kidney: Integrating thyroid and immune signalling. Dev. Comp. Immunol. 66, 73–83. doi:10.1016/j.dci.2016.06.025. Gittleman, J. L. (1986). Carnivore Brain Size, Behavioral Ecology, and Phylogeny. J. Mammal. 67, 23–36. doi:10.2307/1380998. Gonda, A., Herczeg, G., and Merilä, J. (2009). Adaptive brain size divergence in nine-spined sticklebacks (Pungitius pungitius)? J. Evol. Biol. 22, 1721–1726. doi:10.1111/j.1420- 9101.2009.01782.x. Gonda, A., Herczeg, G., and Merilä, J. (2013). Evolutionary ecology of intraspecific brain size variation: a review. Ecol. Evol. 3, 2751–2764. doi:10.1002/ece3.627. Gonzalez-Voyer, A., Winberg, S., and Kolm, N. (2009). Social fishes and single mothers: brain evolution in African cichlids. Proc. R. Soc. B Biol. Sci. 276, 161–167. doi:10.1098/rspb.2008.0979. Grecias, L., Hebert, F. O., Alves, V. A., Barber, I., and Aubin-Horth, N. (2020). Host behaviour alteration by its parasite: from brain gene expression to functional test. Proc. R. Soc. B Biol. Sci. 287. doi:10.1098/rspb.2020.2252. Greenwood, A. K., and Peichel, C. L. (2015). Social regulation of gene expression in threespine sticklebacks. PLoS One 10. doi:10.1371/JOURNAL.PONE.0137726. Guderley, H. E. (1994). “Physiological ecology and evolution in the threespine stickleback,” in The Evolutionary Biology of the threespine stickleback, eds. M. A. Bell and S. A. Foster (Oxford University Press), 85–113. Guemez-Gamboa, A., Coufal, N. G., and Gleeson, J. G. (2014). Primary cilia in the developing and mature brain. Neuron 82, 511. doi:10.1016/J.NEURON.2014.04.024. Guitton, H., Blackwell, D., and Girshick, M. A. (1955). Theory of Games and Statistical Decisions. Rev. économique 6, 508. doi:10.2307/3497265. Gygax, M., Rentsch, A. K., Rudman, S. M., and Rennison, D. J. (2018). Differential predation alters pigmentation in threespine stickleback ( Gasterosteus aculeatus ). J. Evol. Biol. 31, 1589–1598. doi:10.1111/jeb.13354. 50 Han, S., Nam, J., Li, Y., Kim, S., Cho, S.-H., Cho, Y. S., et al. (2010). Regulation of Dendritic Spines, Spatial Memory, and Embryonic Development by the TANC Family of PSD-95-Interacting Proteins. J. Neurosci. 30, 15102–15112. doi:10.1523/JNEUROSCI.3128-10.2010. Härer, A., Bolnick, D. I., and Rennison, D. J. (2021). The genomic signature of ecological divergence along the benthic‐limnetic axis in allopatric and sympatric threespine stickleback. Mol. Ecol. 30, 451–463. doi:10.1111/mec.15746. Heit, B., Tavener, S., Raharjo, E., and Kubes, P. (2002). An intracellular signaling hierarchy determines direction of migration in opposing chemotactic gradients. J. Cell Biol. 159, 91–102. doi:10.1083/jcb.200202114. Henke-von der Malsburg, J., Kappeler, P. M., and Fichtel, C. (2020). Linking ecology and cognition: does ecological specialisation predict cognitive test performance? Behav. Ecol. Sociobiol. 74. doi:10.1007/s00265-020-02923-z. Herczeg, G., Urszán, T. J., Orf, S., Nagy, G., Kotrschal, A., and Kolm, N. (2019). Brain size predicts behavioural plasticity in guppies (Poecilia reticulata): An experiment. J. Evol. Biol. 32, 218–226. doi:10.1111/jeb.13405. Herold, C., Schlömer, P., Mafoppa-Fomat, I., Mehlhorn, J., Amunts, K., and Axer, M. (2019). The hippocampus of birds in a view of evolutionary connectomics. Cortex 118, 165– 187. doi:10.1016/j.cortex.2018.09.025. Hill, M. S., Vande Zande, P., and Wittkopp, P. J. (2020). Molecular and evolutionary processes generating variation in gene expression. Nat. Rev. Genet., 1–13. doi:10.1038/s41576-020-00304-w. Hohenlohe, P. A., Bassham, S., Etter, P. D., Stiffler, N., Johnson, E. A., and Cresko, W. A. (2010). Population genomics of parallel adaptation in threespine stickleback using sequenced RAD tags. PLoS Genet. 6, e1000862. doi:10.1371/journal.pgen.1000862. Holtmaat, A., and Svoboda, K. (2009). Experience-dependent structural synaptic plasticity in the mammalian brain. Nat. Rev. Neurosci. 10, 647–658. doi:10.1038/nrn2699. Howe, K. L., Achuthan, P., Allen, J., Allen, J., Alvarez-Jarreta, J., Ridwan Amode, M., et al. (2021). Ensembl 2021. Nucleic Acids Res. 49, D884–D891. doi:10.1093/nar/gkaa942. Ishikawa, Y., Kitaoka, S., Kawano, Y., Ishii, S., Suzuki, T., Wakahashi, K., et al. (2021). Repeated social defeat stress induces neutrophil mobilization in mice: maintenance after cessation of stress and strain-dependent difference in response. Br. J. Pharmacol. 178, 827–844. doi:10.1111/BPH.15203. 51 Johansen, I. B., Sørensen, C., Sandvik, G. K., Nilsson, G. E., Höglund, E., Bakken, M., et al. (2012). Neural plasticity is affected by stress and heritable variation in stress coping style. Comp. Biochem. Physiol. Part D Genomics Proteomics 7, 161–171. doi:10.1016/J.CBD.2012.01.002. Jones, F. C., Grabherr, M. G., Chan, Y. F., Russell, P., Mauceli, E., Johnson, J., et al. (2012). The genomic basis of adaptive evolution in threespine sticklebacks. Nature 484, 55– 61. doi:10.1038/nature10944. Jones, J. R., Bell, M. A., Baker, J. A., and Koenings, J. P. (2003). General limnology of lakes near cook inlet, southcentral alaska. Lake Reserv. Manag. 19, 141–149. doi:10.1080/07438140309354080. Kalbe, M., and Kurtz, J. (2006). Local differences in immunocompetence reflect resistance of sticklebacks against the eye fluke Diplostomum pseudospathaceum. Parasitology 132, 105–116. doi:10.1017/S0031182005008681. Kapitein, L. C., and Hoogenraad, C. C. (2015). Building the Neuronal Microtubule Cytoskeleton. Neuron 87, 492–506. doi:10.1016/J.NEURON.2015.05.046. Karve, A. D., Von Hippel, F. A., and Bell, M. A. (2008). Isolation between sympatric anadromous and resident threespine stickleback species in Mud Lake, Alaska. Environ. Biol. Fishes 81, 287–296. doi:10.1007/s10641-007-9200-2. Keagy, J., Braithwaite, V. A., and Boughman, J. W. (2018). Brain differences in ecologically differentiated sticklebacks. Curr. Zool. 64, 243–250. doi:10.1093/cz/zox074. Khacho, M., and Slack, R. S. (2018). Mitochondrial dynamics in the regulation of neurogenesis: From development to the adult brain. Dev. Dyn. 247, 47–53. doi:10.1002/DVDY.24538. Kim, Y. R., Kim, Y. M., Lee, J., Park, J., Lee, J. E., and Hyun, Y.-M. (2020). Neutrophils Return to Bloodstream Through the Brain Blood Vessel After Crosstalk With Microglia During LPS-Induced Neuroinflammation. Front. Cell Dev. Biol. 8. doi:10.3389/FCELL.2020.613733. Kol, A., and Goshen, I. (2021). The memory orchestra: the role of astrocytes and oligodendrocytes in parallel to neurons. Curr. Opin. Neurobiol. 67, 131–137. doi:10.1016/j.conb.2020.10.022. Korotkevich, G., Sukhov, V., Budin, N., Shpak, B., Artyomov, M., and Sergushichev, A. (2021). Fast gene set enrichment analysis. bioRxiv, 060012. doi:10.1101/060012. 52 Kotrschal, A., Räsänen, K., Kristjánsson, B. K., Senn, M., and Kolm, N. (2012). Extreme sexual brain size dimorphism in sticklebacks: A consequence of the cognitive challenges of sex and parenting? PLoS One 7. doi:10.1371/journal.pone.0030055. Kotrschal, K., Staaden, M. J. Van, and Huber, R. (1998). Fish brains: evolution and environmental relationships. Rev. Fish Biol. Fish. 8, 373–408. doi:10.1023/A:1008839605380. Kozak, G. M., and Boughman, J. W. (2008). Experience influences shoal member preference in a species pair of sticklebacks. Behav. Ecol. 19, 667–676. doi:10.1093/beheco/arn022. Larhammar, M., Huntwork-Rodriguez, S., Rudhard, Y., Sengupta-Ghosh, A., and Lewcock, J. W. (2017). The Ste20 family kinases MAP4K4, MINK1, and TNIK converge to regulate stress-induced JNK signaling in neurons. J. Neurosci. 37, 11074–11084. doi:10.1523/JNEUROSCI.0905-17.2017. Lasarge, C. L., and Danzer, S. C. (2014). Mechanisms regulating neuronal excitability and seizure development following mTOR pathway hyperactivation. Front. Mol. Neurosci. 7. doi:10.3389/fnmol.2014.00018. Leavell, B. C., and Bernal, X. E. (2019). The Cognitive Ecology of Stimulus Ambiguity: A Predator–Prey Perspective. Trends Ecol. Evol. 34, 1048–1060. doi:10.1016/j.tree.2019.07.004. Leder, E. H., Merilä, J., and Primmer, C. R. (2009). A flexible whole-genome microarray for transcriptomics in three-spine stickleback (Gasterosteus aculeatus). BMC Genomics 10, 426. doi:10.1186/1471-2164-10-426. Lee, P. R., Cohen, J. E., Iacobas, D. A., Iacobas, S., and Fields, R. D. (2017). Gene networks activated by specific patterns of action potentials in dorsal root ganglia neurons. Sci. Rep. 7. doi:10.1038/srep43765. Lema, S. C., Hodges, M. J., Marchetti, M. P., and Nevitt, G. A. (2005). Proliferation zones in the salmon telencephalon and evidence for environmental influence on proliferation rate. Comp. Biochem. Physiol. Part A Mol. Integr. Physiol. 141, 327–335. doi:10.1016/j.cbpb.2005.06.003. Lenz, T. L., Eizaguirre, C., Rotter, B., Kalbe, M., and Milinski, M. (2013). Exploring local immunological adaptation of two stickleback ecotypes by experimental infection and transcriptome-wide digital gene expression analysis. Mol. Ecol. 22, 774–786. doi:10.1111/j.1365-294X.2012.05756.x. 53 Leveelahti, L., Leskinen, P., Leder, E. H., Waser, W., and Nikinmaa, M. (2011). Responses of threespine stickleback (Gasterosteus aculeatus, L) transcriptome to hypoxia. Comp. Biochem. Physiol. - Part D Genomics Proteomics 6, 370–381. doi:10.1016/j.cbd.2011.08.001. Lin, W., Hung, T.-C., Kurobe, T., Wang, Y., and Yang, P. (2021). Microcystin-Induced Immunotoxicity in Fishes: A Scoping Review. Toxins (Basel). 13, 765. doi:10.3390/toxins13110765. Lipton, J. O., and Sahin, M. (2014). The Neurology of mTOR. Neuron 84, 275–291. doi:10.1016/J.NEURON.2014.09.034. Liu, Y., Jones, C. D., Day, L. B., Summers, K., and Burmeister, S. S. (2020). Cognitive Phenotype and Differential Gene Expression in a Hippocampal Homologue in Two Species of Frog. Integr. Comp. Biol. 60, 1007–1023. doi:10.1093/icb/icaa032. Lodge, K. M., Cowburn, A. S., Li, W., and Condliffe, A. M. (2020). The impact of hypoxia on neutrophil degranulation and consequences for the host. Int. J. Mol. Sci. 21, 1183. doi:10.3390/ijms21041183. MacArthur, R. H., and Pianka, E. R. (1966). On Optimal Use of a Patchy Environment. Am. Nat. 100, 603–609. doi:10.1086/282454. MacColl, A. D. C. (2009). Parasite burdens differ between sympatric three-spined stickleback species. Ecography (Cop.). 32, 153–160. doi:10.1111/j.1600-0587.2008.05486.x. Malenka, R. C., and Bear, M. F. (2004). LTP and LTD: An embarrassment of riches. Neuron 44, 5–21. doi:10.1016/j.neuron.2004.09.012. Malone, J. H., and Oliver, B. (2011). Microarrays, deep sequencing and the true measure of the transcriptome. BMC Biol. 9. doi:10.1186/1741-7007-9-34. Mantione, K. J., Kream, R. M., Kuzelova, H., Ptacek, R., Raboch, J., Samuel, J. M., et al. (2014). Comparing bioinformatic gene expression profiling methods: microarray and RNA-Seq. Med. Sci. Monit. Basic Res. 20, 138–142. doi:10.12659/MSMBR.892101. Marco, A., Meharena, H. S., Dileep, V., Raju, R. M., Davila-Velderrain, J., Zhang, A. L., et al. (2020). Mapping the epigenomic and transcriptomic interplay during memory formation and recall in the hippocampal engram ensemble. Nat. Neurosci. 23, 1606– 1617. doi:10.1038/s41593-020-00717-0. Martinez, J., Keagy, J., Wurst, B., Fetzner, W., and Boughman, J. W. (2016). The relative roles of genes and rearing environment on the spatial cognitive ability of two sympatric species of threespine stickleback. Evol. Ecol. Res. 17, 565–581. 54 McNamara, J., and Houston, A. (1980). The application of statistical decision theory to animal behaviour. J. Theor. Biol. 85, 673–690. doi:10.1016/0022-5193(80)90265-9. McPhail, J. D. (1994). “Speciation and the evolution of reproductive isolation in the sticklebacks (Gasterosteus) of south-western British Columbia,” in The evolutionary biology of the threespine stickleback, eds. M. A. Bell and S. A. Foster (Oxford University Press), 399–437. Medina, A. C., Torres-García, M. E., Rodríguez-Serrano, L. M., Bello-Medina, P. C., Quirarte, G. L., McGaugh, J. L., et al. (2019). Inhibition of transcription and translation in dorsal hippocampus does not interfere with consolidation of memory of intense training. Neurobiol. Learn. Mem. 166, 107092. doi:10.1016/j.nlm.2019.107092. Mendoza, M. C., Er, E. E., and Blenis, J. (2011). The Ras-ERK and PI3K-mTOR pathways: cross-talk and compensation. Trends Biochem. Sci. 36, 320–328. doi:10.1016/j.tibs.2011.03.006. Miningou, N., and Blackwell, K. T. (2020). The road to ERK activation: Do neurons take alternate routes? Cell. Signal. 68, 109541. doi:10.1016/j.cellsig.2020.109541. Moran, R., Harvey, I., Moss, B., Feuchtmayr, H., Hatton, K., Heyes, T., et al. (2010). Influence of simulated climate change and eutrophication on three-spined stickleback populations: a large scale mesocosm experiment. Freshw. Biol. 55, 315–325. doi:10.1111/J.1365-2427.2009.02276.X. Morris, M. R. J., Richard, R., Leder, E. H., Barrett, R. D. H., Aubin-Horth, N., and Rogers, S. M. (2014). Gene expression plasticity evolves in response to colonization of freshwater lakes in threespine stickleback. Mol. Ecol. 23, 3226–3240. doi:10.1111/mec.12820. Navarrete, M., Cuartero, M. I., Palenzuela, R., Draffin, J. E., Konomi, A., Serra, I., et al. (2019). Astrocytic p38α MAPK drives NMDA receptor-dependent long-term depression and modulates long-term memory. Nat. Commun. 10. doi:10.1038/s41467- 019-10830-9. Niven, J. E., and Laughlin, S. B. (2008). Energy limitation as a selective pressure on the evolution of sensory systems. J. Exp. Biol. 211, 1792–1804. doi:10.1242/jeb.017574. Noreikiene, K., Herczeg, G., Gonda, A., Balázs, G., Husby, A., and Merilä, J. (2015). Quantitative genetic analysis of brain size variation in sticklebacks: Support for the mosaic model of brain evolution. Proc. R. Soc. B Biol. Sci. 282. doi:10.1098/RSPB.2015.1008. 55 Odling-Smee, L. C., Boughman, J. W., and Braithwaite, V. A. (2008). Sympatric species of threespine stickleback differ in their performance in a spatial learning task. Behav. Ecol. Sociobiol. 2008 6212 62, 1935–1945. doi:10.1007/S00265-008-0625-1. Oh, W. C., and Smith, K. R. (2019). Activity-dependent development of GABAergic synapses. Brain Res. 1707, 18–26. doi:10.1016/J.BRAINRES.2018.11.014. Olsson, O., and Brown, J. S. (2010). Smart, smarter, smartest: foraging information states and coexistence. Oikos 119, 292–303. doi:10.1111/j.1600-0706.2009.17784.x. Øverli, Ø., and Sørensen, C. (2016). On the Role of Neurogenesis and Neural Plasticity in the Evolution of Animal Personalities and Stress Coping Styles. Brain. Behav. Evol. 87, 167–174. doi:10.1159/000447085. Pan, S., Mayoral, S. R., Choi, H. S., Chan, J. R., and Kheirbek, M. A. (2020). Preservation of a remote fear memory requires new myelin formation. Nat. Neurosci. 23, 487–499. doi:10.1038/s41593-019-0582-1. Park, P. J. (2013). Spatial learning ability of the threespine stickleback (Gasterosteus aculeatus) in relation to inferred ecology and ancestry. Evol. Ecol. Res. 15, 213–239. Park, P. J., and Bell, M. A. (2010). Variation of telencephalon morphology of the threespine stickleback (Gasterosteus aculeatus) in relation to inferred ecology. J. Evol. Biol. 23, 1261–1277. doi:10.1111/j.1420-9101.2010.01987.x. Park, P. J., Chase, I., and Bell, M. A. (2012). Phenotypic plasticity of the threespine stickleback Gasterosteus aculeatus telencephalon in response to experience in captivity. Curr. Zool. 58, 189–210. doi:10.1093/czoolo/58.1.189. Park, S., Kramer, E. E., Mercaldo, V., Rashid, A. J., Insel, N., Frankland, P. W., et al. (2016). Neuronal Allocation to a Hippocampal Engram. Neuropsychopharmacology 41, 2987– 2993. doi:10.1038/npp.2016.73. Parker, S. T., and Gibson, K. R. (1977). Object manipulation, tool use and sensorimotor intelligence as feeding adaptations in cebus monkeys and great apes. J. Hum. Evol. 6, 623–641. doi:10.1016/S0047-2484(77)80135-8. Parsons, R. G. (2018). Behavioral and neural mechanisms by which prior experience impacts subsequent learning. Neurobiol. Learn. Mem. 154, 22–29. doi:https://doi.org/10.1016/j.nlm.2017.11.008. Piecyk, A., Hahn, M. A., Roth, O., Dheilly, N. M., Heins, D. C., Bell, M. A., et al. (2021). Cross-continental experimental infections reveal distinct defence mechanisms in populations of the three-spined stickleback Gasterosteus aculeatus. Proc. R. Soc. B 288. doi:10.1098/RSPB.2021.1758. 56 Podgorniak, T., Milan, M., Pujolar, J. M., Maes, G. E., Bargelloni, L., De Oliveira, E., et al. (2015). Differences in brain gene transcription profiles advocate for an important role of cognitive function in upstream migration and water obstacles crossing in European eel. BMC Genomics 16. doi:10.1186/s12864-015-1589-y. Pollen, A. A. (2007). Environmental complexity and social organization sculpt the brain in Lake Tanganyikan cichlid fish. Brain. Behav. Evol. 70, 21–39. doi:DOI: 10.1159/000101067. Pravosudov, V. V., Roth, T. C., Forister, M. L., Ladage, L. D., Kramer, R., Schilkey, F., et al. (2013). Differential hippocampal gene expression is associated with climate-related natural variation in memory and the hippocampus in food-caching chickadees. Mol. Ecol. 22, 397–408. doi:10.1111/mec.12146. R Core Team (2014). R: A Language and Environment for Statistical Computing. Rampon, C., Jiang, C. H., Dong, H., Tang, Y. P., Lockhart, D. J., Schultz, P. G., et al. (2000). Effects of environmental enrichment on gene expression in the brain. Proc. Natl. Acad. Sci. U. S. A. 97, 12880–12884. doi:10.1073/pnas.97.23.12880. Rane, M. J., Gozal, D., Butt, W., Gozal, E., Pierce, W. M., Guo, S. Z., et al. (2005). γ-Amino Butyric Acid Type B Receptors Stimulate Neutrophil Chemotaxis during Ischemia- Reperfusion. J. Immunol. 174, 7242–7249. doi:10.4049/JIMMUNOL.174.11.7242. Resnik, P. (1995). Using Information Content to Evaluate Semantic Similarity in a Taxonomy. in Proceedings of the 14th international joint conference on Artificial intelligence - Volume 1 (IJCAI’95), ed. C. S. Mellish (San Francisco, CA, USA: Morgan Kaufmann Publishers Inc.), 448–453. Rosso, S. B., and Inestrosa, N. C. (2013). WNT signalling in neuronal maturation and synaptogenesis. Front. Cell. Neurosci. 7. doi:10.3389/fncel.2013.00103. Rowe, C., and Healy, S. D. (2014). Measuring variation in cognition. Behav. Ecol. 25, 1287– 1292. doi:10.1093/beheco/aru090. Rowland, W. J. (1994). “Proximate determinants of stickleback behavior: an evolutionary perspective,” in The Evolutionary Biology of the threespine stickleback, eds. M. A. Bell and S. A. Foster (Oxford University Press), 297–344. Rundle, H. D. (2000). Natural Selection and Parallel Speciation in Sympatric Sticklebacks. Science (80-. ). 287, 306–308. doi:10.1126/science.287.5451.306. Salas, C., Broglio, C., and Rodríguez, F. (2003). Evolution of forebrain and spatial cognition in vertebrates: Conservation across diversity. Brain. Behav. Evol. 62, 72–82. doi:10.1159/000072438. 57 Salvanes, A. G. V., Moberg, O., Ebbesson, L. O. E., Nilsen, T. O., Jensen, K. H., and Braithwaite, V. A. (2013). Environmental enrichment promotes neural plasticity and cognitive ability in fish. Proc. R. Soc. B Biol. Sci. 280, 20131331. doi:10.1098/rspb.2013.1331. Samuk, K., Iritani, D., and Schluter, D. (2014). Reversed brain size sexual dimorphism accompanies loss of parental care in white sticklebacks. Ecol. Evol. 4, 3236–3243. doi:10.1002/ece3.1175. Savolainen, O., Lascoux, M., and Merilä, J. (2013). Ecological genomics of local adaptation. Nat. Rev. Genet. 14, 807–820. doi:10.1038/nrg3522. Shaw, K. A., Scotti, M. L., and Foster, S. A. (2007). Ancestral plasticity and the evolutionary diversification of courtship behaviour in threespine sticklebacks. Anim. Behav. 73, 415–422. doi:10.1016/j.anbehav.2006.09.002. Sol, D., Bacher, S., Reader, S. M., and Lefebvre, L. (2008). Brain size predicts the success of mammal species introduced into novel environments. Am. Nat. 172, s63–s71. doi:10.1086/588304. Squire, L. R., Stark, C. E. L., and Clark, R. E. (2004). The medial temporal lobe. Annu. Rev. Neurosci. 27, 279–306. doi:10.1146/annurev.neuro.27.070203.144130. Stanford, B. C. M., Clake, D. J., Morris, M. R. J., and Rogers, S. M. (2020). The power and limitations of gene expression pathway analyses toward predicting population response to environmental stressors. Evol. Appl. 13, 1166–1182. doi:10.1111/eva.12935. Steadman, P. E., Xia, F., Ahmed, M., Mocle, A. J., Penning, A. R. A., Geraghty, A. C., et al. (2020). Disruption of Oligodendrogenesis Impairs Memory Consolidation in Adult Mice. Neuron 105, 150–164. doi:10.1016/j.neuron.2019.10.013. Stickleback assembly and gene annotation (2010). Gasterosteus_aculeatus - Ensembl genome Brows. 103. Available at: http://www.ensembl.org/Gasterosteus_aculeatus/Info/Annotation [Accessed March 9, 2021]. Stutz, W. E., Lau, O. L., and Bolnick, D. I. (2014). Contrasting patterns of phenotype- dependent parasitism within and among populations of threespine stickleback. Am. Nat. 183, 810–825. doi:10.1086/676005. Stutz, W. E., Schmerer, M., Coates, J. L., and Bolnick, D. I. (2015). Among-lake reciprocal transplants induce convergent expression of immune genes in threespine stickleback. Mol. Ecol. 24, 4629–4646. doi:10.1111/mec.13295. 58 Subramanian, A., Tamayo, P., Mootha, V. K., Mukherjee, S., Ebert, B. L., Gillette, M. A., et al. (2005). Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles. Proc. Natl. Acad. Sci. U. S. A. 102, 15545–15550. doi:10.1073/pnas.0506580102. Svanbäck, R., and Schluter, D. (2012). Niche Specialization Influences Adaptive Phenotypic Plasticity in the Threespine Stickleback. Am. Nat. 180, 50–59. doi:10.1086/666000. Tamayo, A.-M. P., Devineau, O., Præbel, K., Kahilainen, K. K., and Østbye, K. (2020). A brain and a head for a different habitat: Size variation in four morphs of Arctic charr (Salvelinus alpinus (L.)) in a deep oligotrophic lake. Ecol. Evol. 10, 11335–11351. doi:10.1002/ECE3.6771. Tonegawa, S., Morrissey, M. D., and Kitamura, T. (2018). The role of engram cells in the systems consolidation of memory. Nat. Rev. Neurosci. 19, 485–498. doi:10.1038/s41583-018-0031-2. Tracey, T. J., Steyn, F. J., Wolvetang, E. J., and Ngo, S. T. (2018). Neuronal lipid metabolism: Multiple pathways driving functional outcomes in health and disease. Front. Mol. Neurosci. 11, 10. doi:10.3389/fnmol.2018.00010. TRI Reagent® Solution Protocol (PN 9738M Rev D) (2010). Available at: https://assets.thermofisher.com/TFS-Assets/LSG/manuals/9738M_D.pdf [Accessed March 12, 2021]. Tsuboi, M., Husby, A., Kotrschal, A., Hayward, A., Buechel, S. D., Zidar, J., et al. (2015). Comparative support for the expensive tissue hypothesis: Big brains are correlated with smaller gut and greater parental investment in Lake Tanganyika cichlids. Evolution (N. Y). 69, 190–200. doi:10.1111/evo.12556. Two-Color Microarray-Based Gene Expression Analysis (Low Input Quick Amp Labeling) Protocol (2015). Available at: https://www.agilent.com/cs/library/usermanuals/Public/G4140- 90050_GeneExpression_TwoColor_6.9.pdf [Accessed March 16, 2021]. Tyssowski, K. M., DeStefino, N. R., Cho, J. H., Dunn, C. J., Poston, R. G., Carty, C. E., et al. (2018). Different Neuronal Activity Patterns Induce Different Gene Expression Programs. Neuron 98, 530–546. doi:10.1016/j.neuron.2018.04.001. Urban, M. C., Strauss, S. Y., Pelletier, F., Palkovacs, E. P., Leibold, M. A., Hendry, A. P., et al. (2020). Evolutionary origins for ecological patterns in space. Proc. Natl. Acad. Sci. U. S. A. 117, 17482–17490. doi:10.1073/pnas.1918960117. 59 Vamosi, S. M., and Schluter, D. (2004). Character shifts in the defensive armor of sympatric sticklebacks. Evolution (N. Y). 58, 376–385. doi:10.1111/j.0014-3820.2004.tb01653.x. Vilas, M. P., Marti, C. L., Oldham, C. E., and Hipsey, M. R. (2018). Macrophyte-induced thermal stratification in a shallow urban lake promotes conditions suitable for nitrogen-fixing cyanobacteria. Hydrobiologia 806, 411–426. doi:10.1007/s10750-017- 3376-z. von Krogh, K., Sørensen, C., Nilsson, G. E., and Øverli, Ø. (2010). Forebrain cell proliferation, behavior, and physiology of zebrafish, Danio rerio, kept in enriched or barren environments. Physiol. Behav. 101, 32–39. doi:10.1016/j.physbeh.2010.04.003. Wang, C., Wang, L., and Gu, Y. (2021a). Microglia, synaptic dynamics and forgetting. Brain Res. Bull. 174, 173–183. doi:10.1016/J.BRAINRESBULL.2021.06.005. Wang, Y., Fu, W. Y., Cheung, K., Hung, K. W., Chen, C., Geng, H., et al. (2021b). Astrocyte-secreted IL-33 mediates homeostatic synaptic plasticity in the adult hippocampus. Proc. Natl. Acad. Sci. U. S. A. 118, e2020810118. doi:10.1073/pnas.2020810118. Ward, J. H. (1963). Hierarchical Grouping to Optimize an Objective Function. J. Am. Stat. Assoc. 58, 236. doi:10.2307/2282967. White, G. E., and Brown, C. (2015). Microhabitat Use Affects Brain Size and Structure in Intertidal Gobies. Brain. Behav. Evol. 85, 107–116. doi:http://dx.doi.org/10.1159/000380875. Wood, L. S., Desjardins, J. K., and Fernald, R. D. (2011). Effects of stress and motivation on performing a spatial task. Neurobiol. Learn. Mem. 95, 277–285. doi:10.1016/J.NLM.2010.12.002. Wootton, R. J. (1994). “Energy allocation in the threespine stickleback,” in The Evolutionary Biology of the Threespine Stickleback, eds. M. A. Bell and S. Forster (Oxford University Press), 116–140. Yap, E. L., and Greenberg, M. E. (2018). Activity-Regulated Transcription: Bridging the Gap between Neural Activity and Behavior. Neuron 100, 330–348. doi:10.1016/j.neuron.2018.10.013. Yates, A. D., Achuthan, P., Akanni, W., Allen, J., Allen, J., Alvarez-Jarreta, J., et al. (2020). Ensembl 2020. Nucleic Acids Res. 48, D682–D688. doi:10.1093/nar/gkz966. Yekutieli, D., and Benjamini, Y. (2001). The control of the false discovery rate in multiple testing under dependency. Ann. Stat. 29, 1165–1188. doi:10.1214/aos/1013699998. 60 Yopak, K. E., Lisney, T. J., Collin, S. P., and Montgomery, J. C. (2007). Variation in Brain Organization and Cerebellar Foliation in Chondrichthyans: Sharks and Holocephalans. Brain. Behav. Evol. 69, 280–300. doi:10.1159/000100037. Yue, M., Luo, D., Liu, S. Y. P., Zhou, Q., Hu, M., Liu, Y., et al. (2016). Misshapen/NIK- related kinase (mink1) is involved in platelet function, hemostasis, and thrombus formation. Blood 127, 927–937. doi:10.1182/blood-2015-07-659185. Zheng, F., Luo, Y., and Wang, H. (2009). Regulation of BDNF-mediated transcription of immediate early gene Arc by intracellular calcium and calmodulin. J. Neurosci. Res. 87, 380. doi:10.1002/JNR.21863. Zhou, W., Cao, L., Jeffries, J., Zhu, X., Staiger, C. J., and Deng, Q. (2018). Neutrophil- specific knockout demonstrates a role for mitochondria in regulating neutrophil motility in zebrafish. Dis. Model. Mech. 11. doi:10.1242/dmm.033027. Zhu, Y., Pak, D., Qin, Y., Mccormack, S. G., Kim, M. J., Baumgart, J. P., et al. (2005). Rap2- JNK removes synaptic AMPA receptors during depotentiation. Neuron 46, 905–916. doi:10.1016/j.neuron.2005.04.037. Zupanc, G. K. H., and Sîrbulescu, R. F. (2011). Adult neurogenesis and neuronal regeneration in the central nervous system of teleost fish. Eur. J. Neurosci. 34, 917–929. doi:10.1111/j.1460-9568.2011.07854.x. 61 Appendices Appendix 1. RNA isolation protocol 1000 µl Tri-reageant (Ambion, Applied Biosystems) were added to a sample immediately after removing it from dry ice. A carbide bead was added to the tube, and the tissue was homogenized by shaking at 30 shakes/second in Tissue Lyser (Qiagen) for 60 seconds, in 30 second cycles, until completely homogenized. The bead was removed by magnet and the lysis was incubated at room temperature for 10 minutes to ensure homogenization. Because of the small amount and soluble nature of the tissue, the additional centrifugation phase was not necessary. 100 µl Bromo-3-chloropropane (Sigma) was added to lysis and mixed by vigorous shaking by hand for 15 seconds followed by a 10-minute incubation in room temperature. The mixture was centrifuged in +4°C at 12,000 x g for 15 minutes. At the centrifugation, the mixture was separated into three layers, with the middle, aqueous, layer containing RNA and lower, red, phase DNA and proteins. The aqueous layer was carefully transferred to a fresh tube by pipetting, avoiding phase mixture. 500 µl isopropanol was added to aqueous solution and mixed by shaking, followed by a 10 - minute incubation in room temperature. The samples were centrifuged for 15 minutes at 12,000 x g in +4°C, to precipitate and collect the RNA. After centrifugation, the supernatant was discarded, and the pellet was washed twice with 1 ml of 75 % ethanol. After the second wash, the samples were incubated in 75 % ethanol in +4°C over night. The ethanol was removed from the samples by pouring after centrifugation in +4°C at 7,500 x g for 5 minutes. The residual ethanol was removed by pipetting with fine tip and brief drying on a paper towel, until no traces of ethanol were visible, avoiding the complete drying of the pellet to ensure RNA solubility. DNAse treatment The master mix of 2.5 µl enzyme and 3 µl 10 x reaction buffer were prepared. The precipitated RNA was dissolved in 25 µl nuclease free H₂O on ice. 5.5 µl master mix was added and mixed by careful pipetting. The samples were incubated in a 37°C water for 45 minutes and cooled on ice. 62 Appendix 2. Microarray design and MA-plots Microarray design Array: GacV4_2012, ID:041818 Cy5 Cy3 Barcode Mud 3 Lynne 1 25418180002 1-1 Tern 4 South Rolly 1 25418180002 1-2 Lynne 4 Mud 4 25418180002 1-3 South Rolly 4 Tern 5 25418180002 1-4 Tern 2 Lynne 2 25418180002 2-1 Mud 2 South Rolly 2 25418180002 2-2 Lynne 3 Tern 3 25418180002 2-3 South Rolly 5 Mud 1 25418180002 2-4 Microarray MA-plots. 63 Appendix 3. Differentially expressed genes and their human orthologs. ProbeName Log FC padj Gasterosteus aculeatus Ensembl ID Human ortholog Ensembl ID Human ortholog gene symbol Gac_32274_PI222095 114 4.51 0.000 7 ENSGACG00000020 394 ENSG000001415 03 MINK1 Gac_53227_PI225017 757 -2.37 0.002 0 ENSGACG00000016 535 ENSG000000805 73 COL5A3 Gac_25502_PI225017 757 -1.86 0.002 0 ENSGACG00000012 847 ENSG000001584 28 CATIP Gac_40437_PI225017 757 2.59 0.002 0 ENSGACG00000020 395 ENSG000001723 54 GNB2 Gac_15926_PI225017 757 -1.63 0.003 0 ENSGACG00000008 096 ENSG000001396 84 ESD Gac_16650_PI225017 757 -1.91 0.003 4 ENSGACG00000019 330 ENSG000001003 47 SAMM50 Gac_46919_PI225017 757 -2.62 0.006 2 ENSGACG00000020 819 ENSG000000481 62 NOP16 Gac_11884_PI225017 757 -2.48 0.006 2 ENSGACG00000017 135 ENSG000001300 55 GDPD2 Gac_6186_PI2250177 57 2.12 0.006 2 ENSGACG00000018 808 ENSG000001440 36 EXOC6B Gac_15447_PI225017 757 1.96 0.007 4 ENSGACG00000011 818 ENSG000001258 14 NAPB Gac_2808_PI2250177 57 1.45 0.007 9 ENSGACG00000009 998 ENSG000001718 62 PTEN Gac_28462_PI225017 757 2.24 0.008 5 ENSGACG00000019 865 ENSG000001208 60 WASHC3 Gac_44710_PI225017 757 -1.42 0.008 5 ENSGACG00000016 430 ENSG000001406 12 SEC11A Gac_6455_PI2250177 57 -1.26 0.008 5 ENSGACG00000008 154 ENSG000001493 13 AASDHPP T Gac_27832_PI225017 757 -1.85 0.008 5 ENSGACG00000016 021 ENSG000001963 58 NTNG2 Gac_52854_PI222095 114 -3.74 0.008 6 ENSGACG00000009 554 ENSG000001033 16 CRYM Gac_39320_PI225017 757 -1.79 0.008 6 ENSGACG00000001 170 ENSG000001965 84 XRCC2 Gac_38489_PI225017 757 -1.52 0.009 0 ENSGACG00000018 280 ENSG000001067 23 SPIN1 Gac_47680_PI225017 757 1.17 0.010 0 ENSGACG00000016 607 ENSG000001062 46 PTCD1 Gac_39550_PI225017 757 -1.17 0.010 0 ENSGACG00000018 088 ENSG000002428 52 ZNF709 64 ProbeName Log FC padj Gasterosteus aculeatus Ensembl ID Human ortholog Ensembl ID Human ortholog gene symbol Gac_24528_PI222095 114 1.37 0.010 3 ENSGACG00000007 021 ENSG000001578 51 DPYSL5 Gac_51475_PI225017 757 -1.88 0.011 5 ENSGACG00000020 288 ENSG000000356 87 ADSS2 Gac_43044_PI225017 757 1.99 0.011 5 ENSGACG00000007 602 ENSG000001196 86 FLVCR2 Gac_23428_PI225017 757 -1.26 0.012 8 ENSGACG00000013 296 ENSG000001376 92 DCUN1D5 Gac_54100_PI225017 757 1.18 0.014 4 ENSGACG00000014 142 ENSG000000539 00 ANAPC4 Gac_3404_PI2250177 57 -1.59 0.014 4 ENSGACG00000012 321 ENSG000001048 24 HNRNPL Gac_52175_PI225017 757 -1.31 0.014 4 ENSGACG00000015 229 ENSG000001141 07 CEP70 Gac_49223_PI225017 757 -1.42 0.014 4 ENSGACG00000004 936 ENSG000001152 82 TTC31 Gac_6546_PI2250177 57 -1.47 0.014 4 ENSGACG00000011 844 ENSG000001511 35 TMEM263 Gac_T09453xG07120_ 753 1.40 0.014 4 ENSGACG00000007 120 ENSG000001641 82 NDUFAF2 Gac_3967_PI2250177 57 -1.46 0.014 4 ENSGACG00000010 417 ENSG000001661 28 RAB8B Gac_29870_PI225017 757 -1.92 0.014 4 ENSGACG00000004 903 ENSG000001690 32 MAP2K1 Gac_40893_PI225017 757 1.58 0.014 4 ENSGACG00000018 932 ENSG000001846 72 RALYL Gac_26794_PI222095 114 -1.18 0.014 7 ENSGACG00000016 450 ENSG000001985 86 TLK1 Gac_43969_PI225017 757 0.98 0.014 9 ENSGACG00000007 296 ENSG000001984 08 OGA Gac_27306_PI222095 114 -7.74 0.016 9 ENSGACG00000015 818 ENSG000001153 86 REG1A Gac_34558_PI222095 114 -1.48 0.017 1 ENSGACG00000020 284 ENSG000001621 88 GNG3 Gac_40410_PI225017 757 2.58 0.017 2 ENSGACG00000011 359 ENSG000001723 66 MCRIP2 Gac_6999_PI2250177 57 -1.02 0.017 4 ENSGACG00000004 234 ENSG000001697 27 GPS1 Gac_5956_PI2250177 57 -1.25 0.019 4 ENSGACG00000020 505 ENSG000001891 43 CLDN4 Gac_21488_PI225017 757 1.25 0.019 6 ENSGACG00000003 218 ENSG000000646 01 CTSA 65 ProbeName Log FC padj Gasterosteus aculeatus Ensembl ID Human ortholog Ensembl ID Human ortholog gene symbol Gac_Gene_T23700_22 15 -1.06 0.019 6 ENSGACG00000017 901 ENSG000002424 19 PCDHGC4 Gac_12079_PI225017 757 2.23 0.019 7 ENSGACG00000015 943 ENSG000000515 23 CYBA Gac_22618_PI225017 757 1.82 0.019 7 ENSGACG00000010 118 ENSG000000715 53 ATP6AP1 Gac_9438_PI2250177 57 -1.12 0.019 7 ENSGACG00000019 329 ENSG000000727 78 ACADVL Gac_27698_PI225017 757 1.49 0.019 7 ENSGACG00000007 011 ENSG000000847 64 MAPRE3 Gac_46173_PI225017 757 -1.41 0.019 7 ENSGACG00000006 249 ENSG000000874 60 GNAS Gac_T10167xG07609_ 4805 -1.13 0.019 7 ENSGACG00000007 609 ENSG000001168 52 KIF21B Gac_5808_PI2250177 57 -1.09 0.019 7 ENSGACG00000007 747 ENSG000001176 32 STMN1 Gac_463_PI22209511 4 -1.10 0.019 7 ENSGACG00000012 623 ENSG000001178 77 POLR1G Gac_137_PI22209511 4 -1.91 0.019 7 ENSGACG00000005 981 ENSG000001259 52 MAX Gac_33481_PI225017 757 -2.64 0.019 7 ENSGACG00000010 775 ENSG000001288 86 ELL3 Gac_36919_PI225017 757 0.96 0.019 7 ENSGACG00000017 521 ENSG000001356 32 SMYD5 Gac_54235_PI225017 757 -0.93 0.019 7 ENSGACG00000005 068 ENSG000001381 62 TACC2 Gac_42912_PI225017 757 0.90 0.019 7 ENSGACG00000010 813 ENSG000001514 13 NUBPL Gac_26283_PI225017 757 -1.26 0.019 7 ENSGACG00000008 117 ENSG000001628 13 BPNT1 Gac_T20445xG15460_ 607 -1.21 0.019 7 ENSGACG00000015 460 ENSG000001638 64 NMNAT3 Gac_T27086xG20444_ 5636 1.14 0.019 7 ENSGACG00000020 444 ENSG000001734 42 EHBP1L1 Gac_42837_PI225017 757 1.36 0.019 7 ENSGACG00000010 841 ENSG000001817 04 YIPF6 Gac_16985_PI225017 757 -0.94 0.019 7 ENSGACG00000001 149 ENSG000001874 16 LHFPL3 Gac_23096_PI222095 114 -0.98 0.019 7 ENSGACG00000000 467 ENSG000002577 27 CNPY2 Gac_34391_PI225017 757 -0.87 0.019 7 ENSGACG00000011 044 ENSG000002743 86 TMEM269 66 ProbeName Log FC padj Gasterosteus aculeatus Ensembl ID Human ortholog Ensembl ID Human ortholog gene symbol Gac_34965_PI225017 757 -1.28 0.020 5 ENSGACG00000013 509 ENSG000001086 02 ALDH3A1 Gac_23507_PI225017 757 1.35 0.020 5 ENSGACG00000013 295 ENSG000001090 89 CDR2L Gac_41226_PI225017 757 1.07 0.020 5 ENSGACG00000017 747 ENSG000001561 70 NDUFAF6 Gac_26599_PI225017 757 1.27 0.020 5 ENSGACG00000014 987 ENSG000001821 73 TSEN54 Gac_54605_PI225017 757 -0.98 0.020 5 ENSGACG00000001 974 ENSG000002537 29 PRKDC Gac_40496_PI225017 757 -1.75 0.020 5 ENSGACG00000020 388 ENSG000001063 84 MOGAT3 Gac_2992_PI2220951 14 1.83 0.020 5 ENSGACG00000013 619 ENSG000001675 36 DHRS13 Gac_54661_PI222095 114 -2.13 0.020 8 ENSGACG00000012 888 ENSG000000808 24 HSP90AA 1 Gac_19543_PI225017 757 -1.02 0.020 8 ENSGACG00000000 750 ENSG000001138 51 CRBN Gac_54911_PI225017 757 -1.40 0.020 8 ENSGACG00000003 928 ENSG000001281 59 TUBGCP6 Gac_33929_PI222095 114 1.40 0.020 8 ENSGACG00000016 157 ENSG000001370 94 DNAJB5 Gac_2306_PI2250177 57 -1.06 0.020 8 ENSGACG00000012 991 ENSG000001372 75 RIPK1 Gac_22805_PI225017 757 -1.08 0.020 8 ENSGACG00000006 602 ENSG000001551 11 CDK19 Gac_1446_PI2250177 57 1.08 0.020 8 ENSGACG00000018 583 ENSG000001650 29 ABCA1 Gac_52822_PI225017 757 -3.45 0.020 8 ENSGACG00000007 077 ENSG000002415 63 CORT Gac_10332_PI225017 757 -1.15 0.021 0 ENSGACG00000018 256 ENSG000001472 34 FRMPD3 Gac_38119_PI225017 757 -1.04 0.021 0 ENSGACG00000016 401 ENSG000001078 97 ACBD5 Gac_15067_PI222095 114 -1.22 0.021 0 ENSGACG00000010 415 ENSG000001567 95 NTAQ1 Gac_50463_PI222095 114 1.24 0.021 0 ENSGACG00000000 950 ENSG000001647 94 KCNV1 Gac_11438_PI225017 757 -1.17 0.021 0 ENSGACG00000009 527 ENSG000001761 71 BNIP3 Gac_6190_PI2250177 57 -2.20 0.021 4 ENSGACG00000018 805 ENSG000001494 76 TKFC 67 ProbeName Log FC padj Gasterosteus aculeatus Ensembl ID Human ortholog Ensembl ID Human ortholog gene symbol Gac_27217_PI222095 114 0.81 0.021 5 ENSGACG00000006 384 ENSG000000010 84 GCLC Gac_34613_PI225017 757 1.01 0.021 5 ENSGACG00000010 619 ENSG000000808 23 MOK Gac_45842_PI225017 757 1.35 0.021 5 ENSGACG00000013 081 ENSG000001007 49 VRK1 Gac_37433_PI225017 757 -1.66 0.021 5 ENSGACG00000011 396 ENSG000001050 58 FAM32A Gac_28890_PI225017 757 -1.25 0.021 5 ENSGACG00000000 064 ENSG000001321 82 NUP210 Gac_6815_PI2250177 57 1.27 0.021 5 ENSGACG00000000 623 ENSG000001514 90 PTPRO Gac_30511_PI225017 757 1.47 0.021 5 ENSGACG00000011 930 ENSG000001570 77 ZFYVE9 Gac_15402_PI225017 757 1.26 0.021 5 ENSGACG00000011 815 ENSG000001634 21 PROK2 Gac_34511_PI225017 757 -1.10 0.021 5 ENSGACG00000010 635 ENSG000001866 87 LYRM7 Gac_23035_PI225017 757 -1.02 0.021 5 ENSGACG00000000 492 ENSG000002044 63 BAG6 Gac_698_PI22501775 7 -1.40 0.021 5 ENSGACG00000017 643 ENSG000000141 38 POLA2 Gac_16589_PI222095 114 0.75 0.021 5 ENSGACG00000004 853 ENSG000001318 44 MCCC2 Gac_44606_PI222095 114 -0.82 0.021 5 ENSGACG00000017 580 ENSG000001850 85 INTS5 Gac_43143_PI225017 757 -1.39 0.021 6 ENSGACG00000005 197 ENSG000000066 25 GGCT Gac_T00532xG00416_ 431 1.51 0.021 6 ENSGACG00000000 416 ENSG000000803 71 RAB21 Gac_6498_PI2250177 57 -0.98 0.021 6 ENSGACG00000020 900 ENSG000001375 00 CCDC90B Gac_Gene_07735 -0.99 0.022 1 ENSGACG00000007 735 ENSG000001773 02 TOP3A Gac_52894_PI225017 757 1.29 0.022 1 ENSGACG00000013 288 ENSG000001861 11 PIP5K1C Gac_45396_PI225017 757 -0.88 0.022 1 ENSGACG00000013 213 ENSG000002318 24 AKAIN1 Gac_18920_PI222095 114 -2.01 0.022 7 ENSGACG00000020 758 ENSG000000111 43 MKS1 Gac_2855_PI2250177 57 1.95 0.022 7 ENSGACG00000009 990 ENSG000001248 18 OPN5 68 ProbeName Log FC padj Gasterosteus aculeatus Ensembl ID Human ortholog Ensembl ID Human ortholog gene symbol Gac_5555_PI2250177 57 -0.88 0.022 9 ENSGACG00000010 339 ENSG000001198 01 YPEL5 Gac_32436_PI225017 757 -1.44 0.022 9 ENSGACG00000005 331 ENSG000001778 30 CHID1 Gac_47830_PI225017 757 1.01 0.023 1 ENSGACG00000005 761 ENSG000001314 67 PSME3 Gac_28991_PI225017 757 3.32 0.023 4 ENSGACG00000018 138 ENSG000001715 87 DSCAM Gac_24883_PI225017 757 -1.05 0.024 2 ENSGACG00000001 958 ENSG000000503 93 MCUR1 Gac_48659_PI225017 757 -0.77 0.024 2 ENSGACG00000001 042 ENSG000001120 81 SRSF3 Gac_29298_PI225017 757 -1.07 0.024 2 ENSGACG00000016 291 ENSG000001258 21 DTD1 Gac_49604_PI225017 757 0.87 0.024 2 ENSGACG00000005 124 ENSG000001593 63 ATP13A2 Gac_47557_PI225017 757 -0.69 0.024 2 ENSGACG00000010 222 ENSG000001627 55 KLHDC9 Gac_10503_PI225017 757 1.01 0.024 2 ENSGACG00000007 738 ENSG000002571 03 LSM14A Gac_27791_PI225017 757 -0.74 0.024 8 ENSGACG00000013 536 ENSG000001197 07 RBM25 Gac_T06996xG05264_ 904 -2.56 0.024 8 ENSGACG00000005 264 ENSG000001354 37 RDH5 Gac_40364_PI222095 114 0.92 0.024 8 ENSGACG00000009 122 ENSG000001382 31 DBR1 Gac_31286_PI225017 757 1.16 0.024 8 ENSGACG00000013 420 ENSG000001678 89 MGAT5B Gac_30294_PI225017 757 -1.98 0.025 1 ENSGACG00000018 447 ENSG000001558 58 LSM11 Gac_22139_PI222095 114 -0.87 0.025 1 ENSGACG00000018 458 ENSG000001711 69 NAIF1 Gac_47121_PI225017 757 1.00 0.025 1 ENSGACG00000010 251 ENSG000001988 29 SUCNR1 Gac_46690_PI225017 757 1.35 0.025 6 ENSGACG00000008 392 ENSG000001642 52 AGGF1 Gac_1621_PI2220951 14 -0.94 0.025 6 ENSGACG00000018 561 ENSG000001151 63 CENPA Gac_11429_PI225017 757 -1.65 0.025 6 ENSGACG00000009 536 ENSG000001627 02 ZNF281 Gac_42136_PI225017 757 1.00 0.025 6 ENSGACG00000014 382 ENSG000001821 08 DEXI 69 ProbeName Log FC padj Gasterosteus aculeatus Ensembl ID Human ortholog Ensembl ID Human ortholog gene symbol Gac_3314_PI2220951 14 -1.26 0.025 9 ENSGACG00000016 703 ENSG000002044 38 GPANK1 Gac_39130_PI225017 757 1.71 0.026 0 ENSGACG00000019 975 ENSG000000064 59 KDM7A Gac_T15334xG11552_ 1371 -1.06 0.026 0 ENSGACG00000011 552 ENSG000001675 53 TUBA1C Gac_49294_PI222095 114 -0.95 0.026 0 ENSGACG00000019 396 ENSG000001702 96 GABARAP Gac_21164_PI225017 757 -0.98 0.026 0 ENSGACG00000008 942 ENSG000001770 30 DEAF1 Gac_10352_PI222095 114 -1.45 0.026 1 ENSGACG00000018 175 ENSG000001133 28 CCNG1 Gac_8716_PI2250177 57 -1.14 0.026 5 ENSGACG00000011 734 ENSG000001854 18 TARS3 Gac_32127_PI225017 757 -1.12 0.026 7 ENSGACG00000000 858 ENSG000000037 56 RBM5 Gac_38942_PI222095 114 -1.21 0.026 7 ENSGACG00000018 248 ENSG000000911 57 WDR7 Gac_11576_PI222095 114 1.01 0.026 7 ENSGACG00000009 858 ENSG000001056 55 ISYNA1 Gac_29543_PI225017 757 -0.93 0.026 7 ENSGACG00000013 187 ENSG000001122 82 MED23 Gac_5048_PI2220951 14 -1.41 0.026 7 ENSGACG00000010 310 ENSG000001255 34 PPDPF Gac_40946_PI225017 757 0.75 0.026 7 ENSGACG00000007 387 ENSG000001448 40 RABL3 Gac_53571_PI225017 757 2.35 0.026 7 ENSGACG00000000 116 ENSG000001530 29 MR1 Gac_44249_PI225017 757 -1.32 0.026 7 ENSGACG00000001 248 ENSG000001682 37 GLYCTK Gac_19947_PI225017 757 -1.60 0.026 7 ENSGACG00000000 341 ENSG000001981 13 TOR4A Gac_11182_PI225017 757 -1.75 0.027 1 ENSGACG00000006 959 ENSG000001158 28 QPCT Gac_4970_PI2250177 57 1.19 0.027 3 ENSGACG00000011 641 ENSG000001436 27 PKLR Gac_20173_PI225017 757 -1.02 0.027 5 ENSGACG00000006 958 ENSG000001474 19 CCDC25 Gac_6095_PI2250177 57 -0.77 0.027 7 ENSGACG00000008 438 ENSG000001048 83 PEX11G Gac_53988_PI225017 757 0.81 0.027 7 ENSGACG00000005 632 ENSG000001107 56 HPS5 70 ProbeName Log FC padj Gasterosteus aculeatus Ensembl ID Human ortholog Ensembl ID Human ortholog gene symbol Gac_45855_PI225017 757 -0.87 0.028 3 ENSGACG00000015 461 ENSG000001349 87 WDR36 Gac_40870_PI225017 757 -0.79 0.028 3 ENSGACG00000018 936 ENSG000001713 11 EXOSC1 Gac_99_PI225017757 0.76 0.028 7 ENSGACG00000012 089 ENSG000001603 23 ADAMTS1 3 Gac_37554_PI225017 757 -1.02 0.028 7 ENSGACG00000013 783 ENSG000001705 71 EMB Gac_T22298xG16852_ 2347 -0.85 0.029 1 ENSGACG00000016 852 ENSG000000998 89 ARVCF Gac_46029_PI225017 757 0.73 0.029 1 ENSGACG00000004 619 ENSG000001183 63 SPCS2 Gac_37472_PI225017 757 -1.14 0.029 1 ENSGACG00000011 385 ENSG000001606 33 SAFB Gac_Gene_T04434_24 2 -1.01 0.029 1 ENSGACG00000003 355 ENSG000001651 69 DYNLT3 Gac_Gene_10285_a_2 0 1.35 0.029 3 ENSGACG00000010 285 ENSG000001486 88 RPP30 Gac_52026_PI222095 114 0.73 0.029 3 ENSGACG00000012 733 ENSG000001988 22 GRM3 Gac_2810_PI2250177 57 -2.09 0.029 4 ENSGACG00000009 978 ENSG000000769 24 XAB2 Gac_Gene_20432 -1.14 0.029 4 ENSGACG00000020 432 ENSG000000222 67 FHL1 Gac_55099_PI225017 757 -1.52 0.029 4 ENSGACG00000017 478 ENSG000001023 93 GLA Gac_35187_PI225017 757 -0.89 0.029 4 ENSGACG00000010 020 ENSG000001220 08 POLK Gac_29012_PI225017 757 -2.06 0.029 4 ENSGACG00000000 047 ENSG000001334 66 C1QTNF6 Gac_43841_PI225017 757 -0.79 0.029 4 ENSGACG00000007 353 ENSG000001445 59 TAMM41 Gac_5553_PI2250177 57 -1.21 0.029 4 ENSGACG00000010 340 ENSG000001598 40 ZYX Gac_26564_PI225017 757 0.92 0.029 4 ENSGACG00000017 430 ENSG000001891 34 NKAPL Gac_5387_PI2250177 57 -2.48 0.029 8 ENSGACG00000018 220 ENSG000001415 27 CARD14 Gac_T06417xG04826_ 4268 -0.94 0.030 7 ENSGACG00000004 826 ENSG000000770 44 DGKD Gac_34197_PI225017 757 1.08 0.030 7 ENSGACG00000014 965 ENSG000001278 24 TUBA4A 71 ProbeName Log FC padj Gasterosteus aculeatus Ensembl ID Human ortholog Ensembl ID Human ortholog gene symbol Gac_307_PI22209511 4 -1.46 0.030 7 ENSGACG00000003 637 ENSG000001323 30 SCLY Gac_9259_PI2250177 57 1.41 0.030 7 ENSGACG00000015 399 ENSG000001522 13 ARL11 Gac_43622_PI225017 757 -1.54 0.030 7 ENSGACG00000016 526 ENSG000001668 51 PLK1 Gac_23544_PI225017 757 -1.49 0.030 7 ENSGACG00000004 091 ENSG000001646 45 TEX47 Gac_36324_PI225017 757 1.67 0.031 0 ENSGACG00000001 337 ENSG000001351 64 DMTF1 Gac_40029_PI225017 757 -1.15 0.031 2 ENSGACG00000018 162 ENSG000001047 65 BNIP3L Gac_4395_PI2250177 57 0.97 0.031 2 ENSGACG00000013 812 ENSG000001112 41 FGF6 Gac_50637_PI225017 757 -1.32 0.031 2 ENSGACG00000006 084 ENSG000001241 93 SRSF6 Gac_9324_PI2250177 57 -1.14 0.031 2 ENSGACG00000015 389 ENSG000001696 83 LRRC45 Gac_33359_PI225017 757 -1.39 0.031 2 ENSGACG00000020 663 ENSG000001962 18 RYR1 Gac_43036_PI225017 757 -1.26 0.031 3 ENSGACG00000007 591 ENSG000001055 07 CABP5 Gac_49062_PI225017 757 2.16 0.031 3 ENSGACG00000013 996 ENSG000001347 16 CYP2J2 Gac_46574_PI225017 757 -0.71 0.031 3 ENSGACG00000017 548 ENSG000001387 77 PPA2 Gac_23643_PI225017 757 -2.35 0.031 3 ENSGACG00000013 064 ENSG000001393 04 PTPRQ Gac_41174_PI225017 757 2.56 0.031 3 ENSGACG00000003 287 ENSG000001592 08 CIART Gac_35988_PI225017 757 -0.76 0.031 3 ENSGACG00000018 402 ENSG000001736 11 SCAI Gac_38936_PI222095 114 -1.17 0.031 3 ENSGACG00000006 743 ENSG000001849 24 PTRHD1 Gac_35132_PI225017 757 -0.98 0.031 3 ENSGACG00000010 024 ENSG000001391 54 AEBP2 Gac_13762_PI222095 114 1.04 0.031 5 ENSGACG00000004 145 ENSG000000874 70 DNM1L Gac_14343_PI225017 757 -0.77 0.031 5 ENSGACG00000017 794 ENSG000001294 84 PARP2 Gac_29685_PI225017 757 1.17 0.031 5 ENSGACG00000004 128 ENSG000001321 99 ENOSF1 72 ProbeName Log FC padj Gasterosteus aculeatus Ensembl ID Human ortholog Ensembl ID Human ortholog gene symbol Gac_9657_PI2250177 57 1.40 0.031 5 ENSGACG00000008 302 ENSG000001490 43 SYT8 Gac_3229_PI2250177 57 -0.98 0.031 5 ENSGACG00000016 720 ENSG000001810 04 BBS12 Gac_40850_PI225017 757 -2.42 0.031 6 ENSGACG00000018 939 ENSG000001671 91 GPRC5B Gac_8545_PI2250177 57 -0.67 0.031 6 ENSGACG00000011 804 ENSG000001562 56 USP16 Gac_T14264xG10741_ 1968 -1.17 0.031 6 ENSGACG00000010 741 ENSG000001507 64 DIXDC1 Gac_7006_PI2250177 57 -0.84 0.031 6 ENSGACG00000010 949 ENSG000001628 73 KLHDC8A Gac_23381_PI225017 757 -0.75 0.031 9 ENSGACG00000013 313 ENSG000000133 75 PGM3 Gac_47144_PI225017 757 2.55 0.032 0 ENSGACG00000019 282 ENSG000001667 10 B2M Gac_43411_PI225017 757 -1.70 0.032 6 ENSGACG00000001 380 ENSG000000327 42 IFT88 Gac_19686_PI222095 114 0.71 0.032 6 ENSGACG00000018 791 ENSG000000686 54 POLR1A Gac_4094_PI2220951 14 -0.79 0.032 6 ENSGACG00000010 353 ENSG000000906 61 CERS4 Gac_42082_PI225017 757 0.88 0.032 6 ENSGACG00000014 386 ENSG000001032 74 NUBP1 Gac_5170_PI2220951 14 -1.60 0.032 6 ENSGACG00000002 628 ENSG000001488 48 ADAM12 Gac_4143_PI2250177 57 -0.90 0.032 6 ENSGACG00000008 609 ENSG000001565 21 TYSND1 Gac_25688_PI225017 757 0.72 0.032 6 ENSGACG00000007 842 ENSG000001834 31 SF3A3 Gac_4285_PI2250177 57 1.24 0.032 7 ENSGACG00000017 630 ENSG000001101 04 CCDC86 Gac_39547_PI225017 757 0.72 0.032 9 ENSGACG00000018 089 ENSG000001712 02 TMEM126 A Gac_54703_PI225017 757 0.75 0.033 6 ENSGACG00000018 960 ENSG000001306 38 ATXN10 Gac_T23532xG17768_ 2549 -0.71 0.033 6 ENSGACG00000017 768 ENSG000001326 92 BCAN Gac_459_PI22501775 7 -0.69 0.033 9 ENSGACG00000018 061 ENSG000001761 01 SSNA1 Gac_12062_PI225017 757 0.88 0.034 0 ENSGACG00000004 393 ENSG000001435 95 AQP10 73 ProbeName Log FC padj Gasterosteus aculeatus Ensembl ID Human ortholog Ensembl ID Human ortholog gene symbol Gac_38384_PI225017 757 -0.98 0.034 8 ENSGACG00000017 285 ENSG000000395 60 RAI14 Gac_49541_PI225017 757 -0.96 0.034 8 ENSGACG00000016 673 ENSG000001386 86 BBS7 Gac_50689_PI222095 114 1.07 0.034 8 ENSGACG00000020 577 ENSG000001669 79 EVA1C Gac_27055_PI225017 757 1.18 0.034 8 ENSGACG00000015 835 ENSG000001515 75 TEX9 Gac_3321_PI2250177 57 1.08 0.035 1 ENSGACG00000012 340 ENSG000001517 02 FLI1 Gac_31529_PI225017 757 0.69 0.035 1 ENSGACG00000017 579 ENSG000001692 30 PRELID1 Gac_17609_PI225017 757 -1.01 0.035 4 ENSGACG00000017 363 ENSG000001070 20 PLGRKT Gac_55179_PI225017 757 -0.79 0.036 8 ENSGACG00000008 409 ENSG000001004 18 DESI1 Gac_49042_PI225017 757 -1.29 0.036 8 ENSGACG00000011 079 ENSG000001044 42 ARMC1 Gac_49878_PI225017 757 1.45 0.036 8 ENSGACG00000009 224 ENSG000001242 57 NEURL2 Gac_25316_PI225017 757 -1.06 0.036 8 ENSGACG00000012 971 ENSG000001390 44 B4GALNT 3 Gac_48643_PI225017 757 0.83 0.036 9 ENSGACG00000019 099 ENSG000000706 10 GBA2 Gac_9621_PI2220951 14 0.83 0.036 9 ENSGACG00000002 229 ENSG000001326 00 PRMT7 Gac_41235_PI222095 114 -1.13 0.037 1 ENSGACG00000000 829 ENSG000001635 28 CHCHD4 Gac_30404_PI222095 114 1.56 0.037 1 ENSGACG00000017 996 ENSG000001642 70 HTR4 Gac_16216_PI225017 757 -1.12 0.037 2 ENSGACG00000018 710 ENSG000001091 58 GABRA4 Gac_23425_PI225017 757 1.05 0.037 2 ENSGACG00000013 304 ENSG000001162 51 RPL22 Gac_14320_PI225017 757 -0.66 0.037 2 ENSGACG00000008 765 ENSG000001347 79 TPGS2 Gac_33570_PI225017 757 0.75 0.037 2 ENSGACG00000019 825 ENSG000001360 21 SCYL2 Gac_8178_PI2250177 57 -0.69 0.037 2 ENSGACG00000000 917 ENSG000001386 98 RAP1GDS 1 Gac_35113_PI225017 757 -0.78 0.037 2 ENSGACG00000004 434 ENSG000001435 75 HAX1 74 ProbeName Log FC padj Gasterosteus aculeatus Ensembl ID Human ortholog Ensembl ID Human ortholog gene symbol Gac_33164_PI222095 114 0.83 0.037 2 ENSGACG00000013 146 ENSG000001876 64 HAPLN4 Gac_40871_PI225017 757 -0.91 0.037 5 ENSGACG00000007 400 ENSG000001199 29 CUTC Gac_19093_PI225017 757 0.94 0.037 6 ENSGACG00000002 312 ENSG000001753 05 CCNE2 Gac_52378_PI225017 757 -0.71 0.037 9 ENSGACG00000011 903 ENSG000000727 56 TRNT1 Gac_13023_PI225017 757 1.18 0.037 9 ENSGACG00000012 149 ENSG000001007 31 PCNX1 Gac_39375_PI225017 757 -1.04 0.037 9 ENSGACG00000013 899 ENSG000001029 31 ARL2BP Gac_27030_PI225017 757 0.63 0.037 9 ENSGACG00000014 081 ENSG000001030 47 TANGO6 Gac_6574_PI2250177 57 0.60 0.037 9 ENSGACG00000011 836 ENSG000001253 75 DMAC2L Gac_25997_PI225017 757 -0.75 0.037 9 ENSGACG00000004 659 ENSG000001258 71 MGME1 Gac_20427_PI225017 757 1.04 0.037 9 ENSGACG00000020 325 ENSG000001325 10 KDM6B Gac_28561_PI225017 757 -1.08 0.037 9 ENSGACG00000016 628 ENSG000001369 33 RABEPK Gac_10390_PI225017 757 0.95 0.037 9 ENSGACG00000002 371 ENSG000001383 03 ASCC1 Gac_41358_PI225017 757 -1.85 0.037 9 ENSGACG00000006 904 ENSG000001517 78 SERP2 Gac_47734_PI225017 757 -1.66 0.037 9 ENSGACG00000014 845 ENSG000001627 45 OLFML2B Gac_19249_PI225017 757 -0.83 0.037 9 ENSGACG00000007 706 ENSG000001641 90 NIPBL Gac_37943_PI225017 757 -0.80 0.037 9 ENSGACG00000015 632 ENSG000001666 89 PLEKHA7 Gac_34034_PI222095 114 -0.84 0.037 9 ENSGACG00000014 400 ENSG000001680 40 FADD Gac_16284_PI225017 757 1.11 0.037 9 ENSGACG00000009 678 ENSG000001696 41 LUZP1 Gac_33774_PI225017 757 -0.89 0.037 9 ENSGACG00000014 633 ENSG000001883 16 ENO4 Gac_16256_PI225017 757 -1.60 0.037 9 ENSGACG00000009 661 ENSG000001890 43 NDUFA4 Gac_32981_PI225017 757 -1.45 0.038 6 ENSGACG00000011 096 ENSG000002643 64 DYNLL2 75 ProbeName Log FC padj Gasterosteus aculeatus Ensembl ID Human ortholog Ensembl ID Human ortholog gene symbol Gac_14656_PI222095 114 0.67 0.038 8 ENSGACG00000016 519 ENSG000001289 89 ARPP19 Gac_51467_PI225017 757 -1.46 0.038 9 ENSGACG00000020 292 ENSG000000116 77 GABRA3 Gac_6276_PI2220951 14 -3.33 0.039 4 ENSGACG00000018 765 ENSG000001643 05 CASP3 Gac_4299_PI2250177 57 0.79 0.039 4 ENSGACG00000006 823 ENSG000001685 69 TMEM223 Gac_16041_PI225017 757 0.71 0.039 5 ENSGACG00000006 347 ENSG000001790 83 FAM133A Gac_28048_PI225017 757 -0.74 0.039 6 ENSGACG00000011 536 ENSG000001052 51 SHD Gac_9487_PI2250177 57 0.82 0.040 1 ENSGACG00000019 308 ENSG000001790 94 PER1 Gac_T16900xG12768_ 793 -0.97 0.040 5 ENSGACG00000012 768 ENSG000001048 05 NUCB1 Gac_20919_PI225017 757 1.80 0.040 5 ENSGACG00000002 649 ENSG000001525 27 PLEKHH2 Gac_12444_PI222095 114 -1.22 0.040 6 ENSGACG00000019 755 ENSG000001112 24 PARP11 Gac_53908_PI225017 757 1.25 0.040 6 ENSGACG00000016 466 ENSG000001372 73 FOXF2 Gac_5604_PI2250177 57 -1.10 0.041 1 ENSGACG00000010 332 ENSG000001663 21 NUDT13 Gac_24938_PI225017 757 -0.82 0.041 1 ENSGACG00000010 954 ENSG000001868 89 TMEM17 Gac_45693_PI225017 757 -0.85 0.041 4 ENSGACG00000009 458 ENSG000000516 20 HEBP2 Gac_T03511xG02673_ 2093 -1.32 0.041 4 ENSGACG00000002 673 ENSG000001154 15 STAT1 Gac_32145_PI225017 757 -0.97 0.041 4 ENSGACG00000000 859 ENSG000001338 78 DUSP26 Gac_27640_PI222095 114 -1.03 0.041 4 ENSGACG00000011 710 ENSG000001580 14 SLC30A2 Gac_41905_PI225017 757 -2.20 0.041 4 ENSGACG00000001 667 ENSG000001668 48 TERF2IP Gac_47608_PI225017 757 1.13 0.041 4 ENSGACG00000010 199 ENSG000001715 52 BCL2L1 Gac_15353_PI225017 757 1.11 0.041 4 ENSGACG00000020 869 ENSG000001751 75 PPM1E Gac_32857_PI225017 757 -0.81 0.041 4 ENSGACG00000020 179 ENSG000001883 75 H3-5 76 ProbeName Log FC padj Gasterosteus aculeatus Ensembl ID Human ortholog Ensembl ID Human ortholog gene symbol Gac_4066_PI2250177 57 -0.61 0.041 7 ENSGACG00000010 389 ENSG000001085 10 MED13 Gac_27607_PI225017 757 -0.56 0.041 7 ENSGACG00000017 826 ENSG000001096 06 DHX15 Gac_24256_PI225017 757 -0.82 0.041 7 ENSGACG00000015 260 ENSG000001842 26 PCDH9 Gac_40257_PI225017 757 -0.84 0.041 8 ENSGACG00000003 886 ENSG000001431 79 UCK2 Gac_38321_PI222095 114 -0.97 0.042 1 ENSGACG00000013 826 ENSG000001387 64 CCNG2 Gac_50213_PI225017 757 0.70 0.042 4 ENSGACG00000018 871 ENSG000002626 60 AC139530 .2 Gac_6544_PI2250177 57 2.11 0.042 4 ENSGACG00000011 841 ENSG000002654 91 RNF115 Gac_35624_PI225017 757 1.26 0.042 8 ENSGACG00000004 345 ENSG000000553 32 EIF2AK2 Gac_43517_PI225017 757 0.92 0.042 8 ENSGACG00000005 738 ENSG000000888 08 PPP1R13 B Gac_16025_PI225017 757 -1.01 0.042 8 ENSGACG00000006 361 ENSG000001067 33 NMRK1 Gac_31435_PI225017 757 -0.96 0.042 8 ENSGACG00000012 151 ENSG000001103 18 CEP126 Gac_22750_PI222095 114 -0.99 0.042 8 ENSGACG00000010 073 ENSG000001516 40 DPYSL4 Gac_16469_PI222095 114 0.90 0.042 8 ENSGACG00000001 224 ENSG000001579 11 PEX10 Gac_19580_PI225017 757 0.78 0.042 8 ENSGACG00000000 309 ENSG000001639 38 GNL3 Gac_Gene_T07924_11 1 1.15 0.042 8 ENSGACG00000005 975 ENSG000001679 96 FTH1 Gac_29761_PI225017 757 -0.89 0.042 8 ENSGACG00000014 037 ENSG000001756 43 RMI2 Gac_44767_PI222095 114 -1.35 0.042 8 ENSGACG00000001 975 ENSG000001821 62 P2RY8 Gac_24299_PI225017 757 -1.65 0.042 8 ENSGACG00000003 741 ENSG000001989 59 TGM2 Gac_47880_PI225017 757 1.50 0.042 9 ENSGACG00000005 778 ENSG000001129 77 DAP Gac_17567_PI225017 757 2.53 0.042 9 ENSGACG00000002 886 ENSG000001408 53 NLRC5 Gac_12296_PI225017 757 -1.32 0.042 9 ENSGACG00000019 226 ENSG000001833 07 TMEM121 B 77 ProbeName Log FC padj Gasterosteus aculeatus Ensembl ID Human ortholog Ensembl ID Human ortholog gene symbol Gac_31152_PI225017 757 -0.64 0.042 9 ENSGACG00000019 996 ENSG000001976 20 EOLA1 Gac_53729_PI222095 114 0.98 0.043 2 ENSGACG00000010 986 ENSG000001695 99 NFU1 Gac_1281_PI2250177 57 -0.82 0.043 3 ENSGACG00000004 523 ENSG000001002 16 TOMM22 Gac_17280_PI222095 114 -0.72 0.043 3 ENSGACG00000004 572 ENSG000001258 48 FLRT3 Gac_50175_PI225017 757 -1.58 0.043 4 ENSGACG00000000 821 ENSG000001739 14 RBM4B Gac_37899_PI225017 757 -1.59 0.043 4 ENSGACG00000004 803 ENSG000001151 28 SF3B6 Gac_20867_PI225017 757 -1.21 0.043 4 ENSGACG00000011 667 ENSG000001630 29 SMC6 Gac_21765_PI225017 757 -1.57 0.043 5 ENSGACG00000016 240 ENSG000001522 29 PSTPIP2 Gac_9311_PI2250177 57 -1.00 0.043 8 ENSGACG00000013 606 ENSG000001211 52 NCAPH Gac_900_PI22501775 7 -0.74 0.043 8 ENSGACG00000019 571 ENSG000001245 87 PEX6 Gac_36490_PI225017 757 -0.73 0.043 8 ENSGACG00000014 042 ENSG000001317 78 CHD1L Gac_47439_PI222095 114 -0.83 0.043 8 ENSGACG00000011 573 ENSG000001405 21 POLG Gac_14064_PI225017 757 1.51 0.043 8 ENSGACG00000018 660 ENSG000001454 28 RNF175 Gac_42245_PI225017 757 -0.65 0.043 8 ENSGACG00000005 536 ENSG000001512 47 EIF4E Gac_23352_PI225017 757 1.16 0.043 8 ENSGACG00000013 318 ENSG000001608 18 GPATCH4 Gac_41239_PI225017 757 0.83 0.043 8 ENSGACG00000017 741 ENSG000001731 20 KDM2A Gac_50548_PI225017 757 0.94 0.043 8 ENSGACG00000007 860 ENSG000001827 68 NGRN Gac_47383_PI225017 757 -1.28 0.043 8 ENSGACG00000002 579 ENSG000001874 75 H1-6 Gac_49618_PI225017 757 0.89 0.043 8 ENSGACG00000005 108 ENSG000001977 46 PSAP Gac_45862_PI225017 757 -0.78 0.043 9 ENSGACG00000015 463 ENSG000001361 69 SETDB2 Gac_39336_PI225017 757 -2.41 0.044 2 ENSGACG00000001 165 ENSG000001063 27 TFR2 78 ProbeName Log FC padj Gasterosteus aculeatus Ensembl ID Human ortholog Ensembl ID Human ortholog gene symbol Gac_42763_PI225017 757 2.22 0.044 5 ENSGACG00000018 734 ENSG000001100 63 DCPS Gac_23541_PI222095 114 0.75 0.044 8 ENSGACG00000009 598 ENSG000001083 12 UBTF Gac_8079_PI2220951 14 2.17 0.044 8 ENSGACG00000009 911 ENSG000001639 47 ARHGEF3 Gac_7842_PI2250177 57 0.84 0.044 9 ENSGACG00000008 196 ENSG000001596 85 CHCHD6 Gac_41744_PI225017 757 -0.62 0.045 0 ENSGACG00000016 968 ENSG000001094 72 CPE Gac_T21148xG15996_ 1209 -1.72 0.045 0 ENSGACG00000015 996 ENSG000001835 69 SERHL2 Gac_45113_PI225017 757 1.23 0.045 3 ENSGACG00000007 869 ENSG000000537 70 AP5M1 Gac_13549_PI225017 757 -1.61 0.045 4 ENSGACG00000010 655 ENSG000001605 63 MED27 Gac_16991_PI225017 757 0.93 0.045 5 ENSGACG00000001 145 ENSG000001020 03 SYP Gac_Gene_T01558_57 1 4.96 0.045 7 ENSGACG00000001 198 ENSG000001791 44 GIMAP7 Gac_12016_PI225017 757 0.91 0.046 0 ENSGACG00000015 955 ENSG000001032 48 MTHFSD Gac_Gene_03206_633 0.86 0.046 0 ENSGACG00000003 206 ENSG000001125 31 QKI Gac_24372_PI225017 757 1.07 0.046 0 ENSGACG00000008 909 ENSG000001362 70 TBRG4 Gac_2076_PI2250177 57 -1.03 0.046 2 ENSGACG00000012 291 ENSG000001886 12 SUMO2 Gac_21902_PI225017 757 -0.80 0.046 3 ENSGACG00000018 509 ENSG000001018 43 PSMD10 Gac_20362_PI225017 757 -0.80 0.046 5 ENSGACG00000011 692 ENSG000001542 40 CEP112 Gac_2690_PI2250177 57 -0.63 0.046 7 ENSGACG00000008 279 ENSG000001184 32 CNR1 Gac_29495_PI225017 757 1.51 0.046 7 ENSGACG00000001 439 ENSG000001516 93 ASAP2 Gac_26051_PI225017 757 -0.70 0.046 7 ENSGACG00000004 663 ENSG000001707 79 CDCA4 Gac_286_PI22501775 7 -0.87 0.046 7 ENSGACG00000012 734 ENSG000001970 77 KIAA1671 Gac_49072_PI225017 757 2.03 0.046 7 ENSGACG00000014 009 ENSG000002043 05 AGER 79 ProbeName Log FC padj Gasterosteus aculeatus Ensembl ID Human ortholog Ensembl ID Human ortholog gene symbol Gac_25443_PI225017 757 -0.98 0.046 7 ENSGACG00000003 827 ENSG000002043 92 LSM2 Gac_6857_PI2250177 57 -0.63 0.047 3 ENSGACG00000013 358 ENSG000001319 43 C19orf12 Gac_T17650xG13329_ 1515 -0.60 0.047 6 ENSGACG00000013 329 ENSG000000660 44 ELAVL1 Gac_20515_PI225017 757 -0.69 0.047 6 ENSGACG00000011 255 ENSG000001001 54 TTC28 Gac_41448_PI225017 757 -0.83 0.047 6 ENSGACG00000013 037 ENSG000002042 56 BRD2 Gac_42727_PI225017 757 1.23 0.047 8 ENSGACG00000009 724 ENSG000001413 37 ARSG Gac_21679_PI222095 114 -0.92 0.048 2 ENSGACG00000008 584 ENSG000001090 72 VTN Gac_Gene_04780b -2.18 0.048 2 ENSGACG00000004 780 ENSG000001145 54 PLXNA1 Gac_274_PI22501775 7 -0.94 0.048 2 ENSGACG00000012 738 ENSG000001399 77 NAA30 Gac_11678_PI225017 757 1.54 0.048 2 ENSGACG00000015 312 ENSG000001431 53 ATP1B1 Gac_26818_PI225017 757 -1.66 0.048 2 ENSGACG00000017 237 ENSG000001472 74 RBMX Gac_10155_PI225017 757 -0.91 0.048 2 ENSGACG00000002 950 ENSG000001981 98 SZT2 Gac_33631_PI225017 757 0.56 0.048 2 ENSGACG00000019 812 ENSG000002134 65 ARL2 Gac_21867_PI222095 114 0.73 0.048 3 ENSGACG00000010 760 ENSG000001637 54 GYG1 Gac_15553_PI225017 757 -0.79 0.048 3 ENSGACG00000002 186 ENSG000000822 58 CCNT2 Gac_49207_PI225017 757 0.66 0.048 3 ENSGACG00000015 744 ENSG000001131 63 CERT1 Gac_7030_PI2250177 57 0.83 0.048 3 ENSGACG00000010 947 ENSG000001731 41 MRPL57 Gac_44179_PI225017 757 -1.36 0.048 7 ENSGACG00000002 399 ENSG000001454 91 ROPN1L Gac_20950_PI222095 114 0.97 0.048 7 ENSGACG00000003 016 ENSG000001448 34 TAGLN3 Gac_14062_PI225017 757 -2.87 0.049 1 ENSGACG00000018 668 ENSG000001668 01 FAM111A Gac_28203_PI225017 757 -0.81 0.049 6 ENSGACG00000001 725 ENSG000001720 07 RAB33B 80 ProbeName Log FC padj Gasterosteus aculeatus Ensembl ID Human ortholog Ensembl ID Human ortholog gene symbol Gac_51945_PI225017 757 -0.73 0.049 6 ENSGACG00000008 449 ENSG000001968 76 SCN8A Gac_18132_PI225017 757 1.17 0.049 9 ENSGACG00000002 021 ENSG000001118 34 RSPH4A 81 Appendix 4. FGSEA enriched biological processes and molecular functions Negative NES indicates enrichment in limnetic ecotype, positive NES in benthic ecotype. Gene frequency is the ratio of genes expressed in this study in an enriched GO term versus all the genes annotated to the GO term in the Ensembl data base. GO class GO cluster GO ID GOterm padj NES genes frequency BP 1 GO:0006629 lipid metabolic process 0.25 1.32 27.052% (89/329) BP 1 GO:0006661 phosphatidylinositol biosynthetic process 0.22 1.64 28.571% (14/49) BP 1 GO:0006665 sphingolipid metabolic process 0.26 1.65 28.571% (10/35) BP 1 GO:0006687 glycosphingolipid metabolic process 0.26 1.69 32.143% (9/28) BP 1 GO:0006646 phosphatidylethanolamine biosynthetic process 0.27 1.74 50% (6/12) BP 1 GO:0001676 long-chain fatty acid metabolic process 0.25 1.78 66.667% (8/12) BP 2 GO:0051301 cell division 0.26 -1.39 26.57% (55/207) BP 2 GO:0006749 glutathione metabolic process 0.25 1.73 21.429% (6/28) BP 2 GO:0016226 iron-sulfur cluster assembly 0.05 2.04 36.842% (7/19) BP 2 GO:0007339 binding of sperm to zona pellucida 0.14 1.90 61.538% (8/13) BP 3 GO:0071897 DNA biosynthetic process 0.25 -1.69 40.909% (9/22) BP 3 GO:0000381 regulation of alternative mRNA splicing. via spliceosome 0.10 -1.74 39.474% (15/38) BP 3 GO:0016070 RNA metabolic process 0.25 -1.64 45.161% (14/31) BP 3 GO:0042795 snRNA transcription by RNA polymerase II 0.26 -1.57 35.185% (19/54) BP 3 GO:0006260 DNA replication 0.09 -1.70 29.07% (25/86) BP 3 GO:0006366 transcription by RNA polymerase II 0.26 -1.42 33.036% (37/112) BP 3 GO:0006357 regulation of transcription by RNA polymerase II 0.25 -1.28 24.696% (122/494) BP 3 GO:0022900 electron transport chain 0.11 1.66 51.515% (34/66) BP 3 GO:0051603 proteolysis involved in cellular protein catabolic process 0.26 1.63 52.632% (20/38) 82 GO class GO cluster GO ID GOterm padj NES genes frequency BP 3 GO:0006801 superoxide metabolic process 0.25 1.83 42.857% (6/14) BP 4 GO:0000070 mitotic sister chromatid segregation 0.22 -1.71 40% (6/15) BP 4 GO:0070317 negative regulation of G0 to G1 transition 0.26 -1.62 34.783% (8/23) BP 4 GO:0000278 mitotic cell cycle 0.10 -1.64 26.829% (22/82) BP 4 GO:0007052 mitotic spindle organization 0.21 -1.57 45.946% (34/74) BP 4 GO:0000086 G2/M transition of mitotic cell cycle 0.27 -1.48 27.397% (20/73) BP 5 GO:0045724 positive regulation of cilium assembly 0.26 -1.67 36.364% (4/11) BP 5 GO:1905515 non-motile cilium assembly 0.03 -1.95 31.034% (9/29) BP 5 GO:0097711 ciliary basal body-plasma membrane docking 0.26 -1.56 27.778% (15/54) BP 5 GO:0030030 cell projection organization 0.26 -1.46 20.952% (22/105) BP 6 GO:0006325 chromatin organization 0.25 -1.41 25.389% (49/193) BP 6 GO:0032981 mitochondrial respiratory chain complex I assembly 0.10 1.80 28.846% (15/52) BP 6 GO:0070584 mitochondrion morphogenesis 0.08 2.02 42.857% (6/14) BP 7 GO:0006869 lipid transport 0.01 2.00 50.649% (39/77) BP 7 GO:0015914 phospholipid transport 0.25 1.82 41.379% (12/29) BP 7 GO:0120009 intermembrane lipid transfer 0.07 1.99 46.154% (12/26) BP 8 GO:0048488 synaptic vesicle endocytosis 0.20 1.81 53.571% (15/28) BP 8 GO:0006911 phagocytosis. engulfment 0.25 1.82 57.143% (8/14) BP 9 GO:0060173 limb development 0.11 -1.80 47.368% (9/19) BP 9 GO:0042733 embryonic digit morphogenesis 0.18 -1.78 33.333% (6/18) BP 9 GO:0007368 determination of left/right symmetry 0.15 -1.75 42.857% (12/28) BP 9 GO:0001843 neural tube closure 0.11 -1.67 32.653% (16/49) 83 GO class GO cluster GO ID GOterm padj NES genes frequency BP 10 GO:0006974 cellular response to DNA damage stimulus 0.22 -1.36 32.119% (97/302) BP 10 GO:0043312 neutrophil degranulation 0.03 1.55 32.4% (81/250) BP 10 GO:0045087 innate immune response 0.03 1.66 31.915% (45/141) BP 10 GO:0006955 immune response 0.11 1.67 31.507% (23/73) BP 11 GO:0050769 positive regulation of neurogenesis 0.11 -1.81 57.143% (8/14) BP 11 GO:0032722 positive regulation of chemokine production 0.25 1.78 66.667% (8/12) BP 11 GO:0090050 positive regulation of cell migration involved in sprouting angiogenesis 0.26 1.74 40% (4/10) BP 11 GO:0042327 positive regulation of phosphorylation 0.10 1.97 35.714% (5/14) BP 11 GO:0090023 positive regulation of neutrophil chemotaxis 0.08 1.99 69.231% (9/13) BP 12 GO:0060544 regulation of necroptotic process 0.26 -1.67 54.545% (6/11) BP 12 GO:0097191 extrinsic apoptotic signaling pathway 0.11 -1.82 36.842% (7/19) BP 12 GO:2001244 positive regulation of intrinsic apoptotic signaling pathway 0.26 1.77 40% (6/15) BP 12 GO:2001235 positive regulation of apoptotic signaling pathway 0.25 1.78 75% (9/12) BP 13 GO:0008589 regulation of smoothened signaling pathway 0.03 -1.96 63.636% (7/11) BP 13 GO:0051897 positive regulation of protein kinase B signaling 0.11 1.66 34.286% (24/70) BP 13 GO:0014068 positive regulation of phosphatidylinositol 3-kinase signaling 0.25 1.59 27.907% (12/43) BP 13 GO:0051898 negative regulation of protein kinase B signaling 0.04 2.05 44% (11/25) BP 13 GO:1900745 positive regulation of p38MAPK cascade 0.21 1.85 53.333% (8/15) MF 1 GO:0003756 protein disulfide isomerase activity 0.33 -1.70 54.545% (6/11) MF 1 GO:0009055 electron transfer activity 0.44 1.55 52.83% (28/53) MF 1 GO:0004438 phosphatidylinositol-3- phosphatase activity 0.44 1.69 25% (3/12) 84 GO class GO cluster GO ID GOterm padj NES genes frequency MF 2 GO:0051536 iron-sulfur cluster binding 0.15 1.77 56.25% (27/48) MF 2 GO:0051539 4 iron. 4 sulfur cluster binding 0.07 1.92 61.29% (19/31) MF 3 GO:0005164 tumor necrosis factor receptor binding 0.01 -1.91 37.5% (3/8) MF 3 GO:0035257 nuclear hormone receptor binding 0.15 -1.81 46.154% (6/13) MF 4 GO:0043015 gamma-tubulin binding 0.01 -2.04 25% (4/16) MF 4 GO:0035064 methylated histone binding 0.44 -1.62 38.71% (12/31) MF 4 GO:0043621 protein self-association 0.33 1.70 51.724% (15/29) MF 4 GO:0019955 cytokine binding 0.44 1.66 46.154% (6/13) MF 4 GO:0034236 protein kinase A catalytic subunit binding 0.44 1.70 37.5% (3/8) MF 4 GO:0031720 haptoglobin binding 0.44 1.62 100% (5/5) MF 5 GO:0003684 damaged DNA binding 0.33 -1.70 36.585% (15/41) MF 5 GO:0003677 DNA binding 0.00 -1.47 26.011% (193/742) MF 6 GO:0003682 chromatin binding 0.33 -1.41 34.222% (77/225) MF 6 GO:0008035 high-density lipoprotein particle binding 0.44 1.72 42.857% (3/7) MF 6 GO:0019212 phosphatase inhibitor activity 0.44 1.65 80% (4/5) MF 6 GO:0097001 ceramide binding 0.33 1.77 50% (3/6) 85 Appendix 5. Expressed genes annotated to FGSEA enriched GO terms GO cla ss GO clu ster GO term ID Human ortholog gene symbols BP 1 GO:000662 9 ABCB4;ST3GAL1;ABHD5;AGPS;PNPLA6;IDI1;PLPP1;GBA2;HACD3; HADHA;HSD17B14;CETP;SCD;TECR;MTMR3;DDHD1;PLTP;RUBC NL;GDPD3;ASAH1;EPHX3;ISYNA1;CPT1A;CHKA;PLCH1;HDLBP;EL OVL4;GPAM;EPHX2;AKR1D1;ACSL3;MBOAT7;INSIG2;ECHS1;PLD 2;APOE;SLC27A1;PNPLA7;PRKAB2;ERG28;CEPT1;CYP2J2;FADS2 ;PNPLA8;HADHB;ABHD2;ETNK2;OSBPL10;PLA1A;ERLIN2;IPMK;A CSL1;LPCAT1;PLCL2;SLC16A1;PTDSS1;LRP8;AASDH;G6PD;AGPA T3;FDPS;ACOX1;TPRA1;LIPH;ENPP6;ABCA1;FAAH2;ECI1;AGPAT2 ;FASN;FABP6;PTEN;FADS6;PTDSS2;ACER2;ACOT4;PSAPL1;PLCX D1;PLCB1;ACSM2A;HACD4;GM2A;FAR1;SLC22A24;PSAP;HACD2; CERS1;PISD;HTD2 BP 1 GO:000666 1 MTMR7;SH3YL1;PITPNM2;MTMR3;VAC14;PIK3R2;PIP5K1B;SLC27 A1;INPP5K;PIK3R5;PIP5K1A;PTPN13;PTEN;PIP5K1C BP 1 GO:000666 5 AGK;GBA2;CERK;ASAH1;ACER2;PSAPL1;TH;GM2A;PSAP;CERS1 BP 1 GO:000668 7 CTSA;GBA2;CERK;ASAH1;ESYT1;GLB1;GM2A;PSAP;HEXA BP 1 GO:000664 6 CHKA;SLC27A1;CEPT1;ETNK2;PHOSPHO1;PISD BP 1 GO:000167 6 CPT1A;ACSL3;SLC27A1;ACSL1;CPT2;SLC27A4;ACOT4;ACAD9 BP 2 GO:005130 1 CCDC124;KIF2A;CDC42;NDE1;CCNT2;KLHL42;BIRC5;TRIOBP;TTC 28;KATNAL1;UBE2I;FSD1;CDC37;TIMELESS;RNF8;CCNG1;ACTR8; GNAI2;NEK4;CENPA;BIRC6;ARHGEF2;NEK6;NCAPH;CDK7;BORA; SETDB2;DCTN3;NEK1;NUSAP1;CCNG2;POGZ;JTB;DYNC1LI1;TAC C1;EML3;INCENP;USP16;PMF1;DYNLT3;PLK1;TUBA1C;CDCA4;PR KCE;TPPP;CKS1B;LIG4;ANKLE2;CETN1;FIGN;KNTC1;SAPCD2;TU BB;DCTN1;PPP1CB BP 2 GO:000674 9 GCLC;GGT5;GSTZ1;ARL6IP5;G6PD;GSTT4 BP 2 GO:001622 6 NUBP1;ISCU;NUBPL;XDH;CIAO2B;NFU1;NFS1 BP 2 GO:000733 9 CCT4;TCP1;ALDOA;CCT5;CLGN;CCT8;CCT2;ZP3 BP 3 GO:007189 7 POLB;POLE4;POLK;REV1;POLG;POLE3;POLL;TK2;LIG4 BP 3 GO:000038 1 RBM5;YTHDC1;CELF4;FMR1;HNRNPL;RBM25;SRSF6;ZC3H10;TRA 2B;WTAP;RBMX;CELF3;RBM15;RBM10;RNPS1 BP 3 GO:001607 0 HNRNPM;POLR2E;HNRNPL;POLR2I;GTF2F1;HNRNPR;NCBP1;HN RNPD;DUSP11;POLR2D;WTAP;RBMX;POLR2H;HNRNPK BP 3 GO:004279 5 CCNT2;SRRT;POLR2E;INTS9;POLR2I;RPAP2;GTF2F1;ELL3;CDK7; NCBP1;INTS3;POLR2D;TAF5;RPRD2;POLR2H;PHAX;GTF2A1;POL R2G;INTS5 86 GO cla ss GO clu ster GO term ID Human ortholog gene symbols BP 3 GO:000626 0 POLA2;RFC1;POLB;SUPT16H;GINS1;TIMELESS;ORC4;RPA2;POLK ;DTD1;GINS2;REV1;BARD1;POLG;POLE3;KIN;POLL;MCM7;FAM11 1A;LIG4;RMI2;TOP3A;PRIM1;TOP1;NT5M BP 3 GO:000636 6 ETV1;GTF2IRD1;CCNT2;SUPT16H;POLR2E;TRIP11;NELFCD;RPAP 1;POLR2I;SUPT6H;GTF2H3;RNGTT;TAF12;HIF3A;GTF2F1;ELL3;CD K7;NCBP1;WDR61;NFIC;POLR2D;RBMX;TAF5;PKNOX1;PMF1;IWS 1;POLR2H;GTF2A1;DDX21;POLR2G;STAT3;DRAP1;DEAF1;JUN;MA F;TCEA1;GTF2H4 BP 3 GO:000635 7 ZFX;ETV1;TARBP1;MED29;RORA;TCF3;SMARCE1;RFX3;MEF2C;P GR;CCNT2;NFE2L1;ATXN7L3;KDM2B;IKZF5;HIF1A;MYEF2;TLE5;L HX2;MED13;KAT2A;DLX4;DBX1;CHD4;PPARD;MED23;FHL2;EPAS1 ;CHD5;TFAP2E;ARID1A;ATF6;BCL11A;MXI1;ENY2;ZNF644;INHBA; TWIST1;TTC21B;BAZ2B;RBPJL;HIF3A;MAX;STAT5A;UXT;BCL11B; KLF2;FEZF1;MED18;ZNF236;RFXAP;TFCP2;EMX1;VEZF1;TBR1;D MRT1;BBS7;AEBP2;TCF12;ZFHX3;MED9;SMAD4;NFIC;DMRTA2;H DGF;CIAO1;OGT;ST18;THRB;ZIC1;TGIF2LX;BATF;USP16;ZNF362; PKNOX1;MED27;SAFB;GMEB1;LHX8;NIPBL;UNCX;GBX1;HNRNPK; PRKCB;CTNNB1;IRF2BP2;STAT3;ZEB2;HIC2;NFXL1;IRF2BP1;ZBT B26;BPTF;CAMTA1;KLF17;PHF8;TADA2B;ESRRA;STOX2;NR2F1;M AMSTR;DEAF1;JUN;MAF;MED14;BHLHA15;BHLHE22;NPM1;ZBTB2 0;RFX7;TSHZ2;POU3F1;ZFP91;ZNF292;MYT1;HES5;FOXO6;BRD2; EHMT2;SP9;ZNF709;SEBOX BP 3 GO:002290 0 CIAPIN1;CYBA;NDUFB4;ME1;P4HA2;ME2;PHGDH;DHRS2;ETFB;C OX7A2;NDUFS7;COX7A2L;NCF2;SDHB;AKR1A1;COX7C;COX4I1;A OC2;FDX1;ETFA;SDHC;SURF1;XDH;NCF1;DHRS3;NDUFS4;CYBB; ETFDH;GLRX;COX5A;NDUFA4L2;ADH5;MT-ND3;MT-ND1 BP 3 GO:005160 3 PSMB1;PSMA4;PSMA3;PSMB5;CTSZ;CTSH;FZR1;CTSC;CTSL;MD M2;CTSV;CTSK;PSMA8;PSMB4;CTSS;RNF123;CTSB;CLPX;PSMB1 0;PSMB9 BP 3 GO:000680 1 CYBA;SOD2;NCF1;CYBB;SH3PXD2B;NOXA1 BP 4 GO:000007 0 WRAP73;TUBG1;ESPL1;NUSAP1;INCENP;PLK1 BP 4 GO:007031 7 YAF2;L3MBTL2;RNF2;MAX;UXT;SUZ12;EHMT1;EHMT2 BP 4 GO:000027 8 HGF;PPP1R12A;TUBB1;NEK4;DNMT3A;RNF2;TUBGCP6;TUBG1;E SPL1;SETDB2;DCTN3;JTB;USP16;NOLC1;PLK1;TUBA1C;TUBB6;C ETN1;TUBB;XRCC2;DCTN1;TUBA4B BP 4 GO:000705 2 MAD1L1;NUP160;INTS13;KIF2A;NDE1;SEH1L;BIRC5;NUDC;KIF4A; RANGAP1;DCTN6;CEP126;NUP107;CENPA;STMN1;NEK6;CENPL; B9D2;NUP85;TUBG1;DYNC1LI2;TACC2;DYNC1LI1;TACC1;INCENP; BUB3;PMF1;CLASP2;MAD2L1;PLK1;DCTN2;CENPS;KNTC1;DYNLL 2 BP 4 GO:000008 6 PPP1R12A;PRKACA;HSP90AA1;CCP110;FBXL15;SKP1;CEP70;OP TN;TUBG1;BORA;DCTN3;PLK1;YWHAG;DCTN2;SSNA1;TUBB;CEP 290;DCTN1;CEP43;PPP1CB BP 5 GO:004572 4 FUZ;CROCC;CCP110;DZIP1 87 GO cla ss GO clu ster GO term ID Human ortholog gene symbols BP 5 GO:190551 5 FUZ;MKS1;IFT88;CEP126;IFT172;BBS7;TMEM17;DYNC2H1;DCTN1 BP 5 GO:009771 1 MKS1;PRKACA;HSP90AA1;CCP110;CEP70;B9D2;TUBG1;PLK1;YW HAG;DCTN2;SSNA1;TUBB;CEP290;DCTN1;CEP43 BP 5 GO:003003 0 FUZ;MKS1;IFT88;CROCC;CCP110;CEP126;WRAP73;B9D2;CEP20; NEK1;BBS7;DCDC2;TMEM237;CATIP;DNAAF3;GPR22;UNC119B;T MEM17;DYNC2H1;DNAJB13;CEP290;CEP43 BP 6 GO:000632 5 HLTF;SMARCE1;BAZ2A;TDRD3;ATXN7L3;KDM2B;BRPF3;L3MBTL2 ;RNF40;SPIN1;EZH1;KMT5B;CHD4;RNF8;CHD5;RBBP5;ARID1A;DN MT3A;HAT1;RRP8;DPF2;SETDB2;AEBP2;SETD1B;CBX4;BRD4;KD M6A;OGT;RNF20;N6AMT1;USP16;SAFB;DPY30;SMARCAD1;MEAF 6;BAP1;PRKCB;RCCD1;BPTF;PHF8;SUZ12;EHMT1;SMYD3;KDM4D ;TLK1;BRD2;EHMT2;BAG6;C17orf49 BP 6 GO:003298 1 NDUFB7;TAZ;FOXRED1;NDUFS7;NDUFAF4;NDUFB5;NUBPL;NDU FAF6;NDUFAF2;TMEM126A;ACAD9;NDUFV2;NDUFA6;MT-ND3;MT- ND4 BP 6 GO:007058 4 POLDIP2;DNM1L;CERT1;NUBPL;MFF;BCL2L1 BP 7 GO:000686 9 ABCB4;CETP;GRAMD1A;OSBPL8;PLTP;ATP11C;OSBP;CHKA;TME M30A;CERT1;HDLBP;ESYT2;LBP;APOE;SLC27A1;OSBPL2;ATP8A2 ;ANO3;ESYT1;PRELID3A;OSBPL1A;ABCG8;OSBPL10;ATP10D;AN O7;VLDLR;PEX19;ABCA1;SLC27A4;MFSD2A;PRELID1;SPNS1;FAB P6;APOLD1;APOO;GM2A;MFSD2B;APOC2;YJEFN3 BP 7 GO:001591 4 CETP;PITPNM2;OSBPL8;PLTP;ATP11C;OSBPL2;ATP8A2;PRELID3 A;ABCG8;ATP10D;ABCA1;PRELID1 BP 7 GO:012000 9 CETP;PITPNM2;OSBPL8;PLTP;OSBP;CERT1;APOE;OSBPL2;PRELI D3A;ABCG8;ABCA1;PRELID1 BP 8 GO:004848 8 ACTB;DNM1L;SYP;DNM1;PACSIN1;SYT2;SYT8;AP2M1;CDK5;ARF6 ;NLGN1;PIP5K1C;NLGN3;STON1;NLGN2 BP 8 GO:000691 1 MARCO;MYH9;BIN2;MFGE8;ELMO1;ABCA1;ARHGAP12;IGLL5 BP 9 GO:006017 3 WNT3;KAT2A;KAT2B;IFT172;BBS7;RNF165;SLC39A3;CTNNB1;BM PR2 BP 9 GO:004273 3 CREBBP;MKS1;B9D1;TWIST1;SMAD4;CTNNB1 BP 9 GO:000736 8 MKS1;RFX3;IFT74;SUFU;NPHP3;ACVR2A;IFT172;DYNC2LI1;BBS7; DPCD;DYNC2H1;DNAAF4 BP 9 GO:000184 3 FUZ;MKS1;PRKACA;LHX2;SUFU;KAT2A;TWIST1;IFT172;PRICKLE1 ;KDM6A;SPINT1;RPS7;TRAF6;DEAF1;RGMA;BRD2 BP 10 GO:000697 4 BAZ1B;ERCC1;NUCKS1;POLB;MLH1;UNG;XAB2;UBE2A;XRCC5;SI RT4;SUPT16H;ASCC2;SUSD6;APEX1;XIAP;FMR1;UBR5;FBXW7;G TF2H3;TIMELESS;TDP2;RNF8;PHF1;ACTR8;WDR48;NEK4;GADD4 5A;RBBP5;RPA2;MLH3;NRDE2;POLK;OARD1;RNF113A;MGME1;PA RP2;TOP2A;CHD1L;CDK7;FBH1;DDB2;REV1;TTC5;CDK9;XPA;APT X;DCUN1D5;BARD1;NEIL1;BRD4;INTS3;FANCD2;SETD7;SHPRH;T LK2;POLE3;KIN;DCLRE1C;BATF;USP16;PAXIP1;RNF111;UBQLN4; MAPKAPK2;SMC6;SMARCAD1;DTX3L;NIPBL;CASP3;SPIDR;POLL; 88 GO cla ss GO clu ster GO term ID Human ortholog gene symbols MCM7;FAM111A;INO80E;USP47;NFRKB;ZDHHC16;ZMAT3;FBXO45 ;LIG4;TRAF6;CENPS;ERCC4;PARP10;SLX1B;MORF4L1;KDM4D;NH EJ1;MCRS1;XRCC2;TOPORS;TLK1;GTF2H4;PARG;SHLD3;PRKDC; EID3 BP 10 GO:004331 2 MGST1;ALOX5;DERA;ATP6V0A1;AGA;CYBA;SLC2A3;CTSA;PKM;P TPRC;CMTM6;UNC13D;ADA2;TOM1;CTSZ;MAGT1;PSMD7;HMOX2; CTSH;ASAH1;RAB3D;ALDOC;CTSC;TCIRG1;BIN2;PTPN6;FAF2;LA MTOR2;PADI2;CTSD;VAMP8;PLAU;HVCN1;NCKAP1L;RAP2C;FGL2 ;OSTF1;TMBIM1;ADAM10;RHOF;SRP14;IQGAP1;CD53;CREG1;ILF2 ;IQGAP2;SURF4;ALDOA;LPCAT1;CCT8;EEF1A1;NBEAL2;PSMC2;C TSS;PTX3;GYG1;CTSB;CYBB;CCT2;B2M;EEF2;FTH1;PGM2;AGPA T2;GLB1;RAB37;HPSE;PSMD1;RHOG;ARHGAP45;CXCR2;RAP2B;A TP6V0C;TUBB4B;HMGB1;GM2A;ATG7;PSAP;PLEKHO2;HBB;PECA M1 BP 10 GO:004508 7 MARCO;CYBA;EIF2AK2;IFI35;DDX1;HCK;USP14;TLR8;CORO1A;C1 QBP;DHX58;C7;TRIM23;CAPZA1;TRIM32;NMI;LBP;CRP;APCS;RSA D2;PARP9;NLRC5;APP;LYAR;MFHAS1;MR1;SLA;NCF1;C1QC;TNF AIP8L2;SLC15A2;PTX3;CYBB;IFI27;CLEC4E;B2M;PARP14;CSF1R; POLR3C;HMGB1;ILRUN;SRC;AGER;TTC4;IGLL5 BP 10 GO:000695 5 CD74;LCP2;TLR8;C1QBP;CTSC;C7;CERT1;EXOSC9;S1PR4;CTSL; CTSV;IFI44;IRF8;MR1;C1QC;CTSS;B2M;SP2;FTH1;PRELID1;CMKL R1;CXCR2;TAPBP BP 11 GO:005076 9 SPEN;XRCC5;SRRT;OPRM1;ELL3;LIG4;XRCC2;TGM2 BP 11 GO:003272 2 CD74;EIF2AK2;HMOX1;SELENOK;LBP;APP;CSF1R;AGER BP 11 GO:009005 0 HDAC9;HMOX1;VEGFA;AKT3 BP 11 GO:004232 7 VEGFA;NCKAP1L;APP;AR;DSCAM BP 11 GO:009002 3 CD74;DAPK2;DNM1L;C1QBP;RIPOR2;NCKAP1L;RAC2;LBP;CXCR2 BP 12 GO:006054 4 HSP90AA1;CDC37;BIRC2;RIPK1;OGT;FADD BP 12 GO:009719 1 KRT18;INHBA;PARP2;ACVR1B;RIPK1;FADD;KRT8 BP 12 GO:200124 4 IL20RA;DNM1L;NKX3-1;BCL2L1;BOK;BCAP31 BP 12 GO:200123 5 CTSH;TGFBR1;CTSC;TPD52L1;TRAF7;DAB2IP;NKX3-1;ING5;PTEN BP 13 GO:000858 9 CREBBP;FUZ;MKS1;RORA;TTC21B;IFT172;ZIC1 BP 13 GO:005189 7 MYOC;FGFR1;MAZ;AKT2;PIK3R2;C1QBP;FGF6;VEGFA;HBEGF;RT N4;RAC2;LIN28A;KL;ARRB2;PIK3R5;MFHAS1;MYORG;P2RY12;HP SE;VEGFB;RHOG;HCLS1;IRS2;SRC BP 13 GO:001406 8 MYOC;FGFR1;MAZ;PTPN6;VEGFA;SOX9;PRR5L;PIK3R5;NCF1;HC LS1;IRS2;SRC 89 GO cla ss GO clu ster GO term ID Human ortholog gene symbols BP 13 GO:005189 8 PHLPP2;HYAL2;MUL1;NOP53;INPP5K;ARRB2;FLCN;XDH;OTUD3;P TEN;PDCD6 BP 13 GO:190074 5 SPI1;GADD45B;VEGFA;MINK1;MFHAS1;XDH;NCF1;AGER MF 1 GO:000375 6 PDIA5;PDIA6;PDIA4;PDIA3;CRELD2;PDIA2 MF 1 GO:000905 5 CIAPIN1;NDUFS1;CYBA;ME1;P4HA2;ME2;PHGDH;ETFB;COX7A2;C OX7A2L;NCF2;SDHB;AKR1A1;AOC2;FDX1;ETFA;SDHC;XDH;NCF1; NDUFS2;DHRS3;NDUFAF2;CYBB;ETFDH;GLRX;NDUFV2;COX5A;A DH5 MF 1 GO:000443 8 MTMR7;MTMR3;PTEN MF 2 GO:005153 6 CIAPIN1;NDUFS1;POLD1;NTHL1;FECH;ACO2;NUBP1;NDUFS7;SD HB;LIAS;ACO1;MOCS1;ELP3;RSAD2;ISCU;FDX1;NUBPL;XDH;NDU FS2;UQCRFS1;NFU1;ETFDH;RFESD;POLE;NDUFV2;BOLA2;NFS1 MF 2 GO:005153 9 CIAPIN1;NDUFS1;POLD1;NTHL1;ACO2;NUBP1;NDUFS7;SDHB;LIA S;ACO1;MOCS1;ELP3;RSAD2;ISCU;NUBPL;NDUFS2;NFU1;ETFDH; POLE MF 3 GO:000516 4 STAT1;FADD;TRAF6 MF 3 GO:003525 7 CRY1;HIF1A;STAT1;TACC2;CTNNB1;JUP MF 4 GO:004301 5 CEP70;B9D2;TUBGCP6;DIXDC1 MF 4 GO:003506 4 TDRD3;L3MBTL2;FMR1;SPIN1;RBBP5;RRP8;CBX4;CDYL;BPTF;PH F8;SUZ12;JMJD7 MF 4 GO:004362 1 AGA;SYP;SGTA;CTSC;TRIM32;GPSM2;ATXN1;RSAD2;TDG;MDH2; NKX3-1;CRK;TMEM43;MYD88;LDB1 MF 4 GO:001995 5 CD74;IFNGR1;CSF3R;TNFRSF11A;FZD4;CSF1R MF 4 GO:003423 6 CSK;SOX9;PRKAR1B MF 4 GO:003172 0 HBA1;HBE1;HBG1;HBD;HBB MF 5 GO:000368 4 CREBBP;ERCC1;POLB;UNG;XRCC5;APEX1;RPA2;POLK;DDB2;RE V1;XPA;APTX;DCLRE1C;SMC6;KDM4D MF 5 GO:000367 7 RBM5;RECQL;ZFX;ETV1;GTF2IRD1;CRY1;ERCC1;POLR3B;POLA2; YAF2;SARS1;RFC1;SPEN;RORA;POLB;TCF3;MBD3;HLTF;SMARCE 1;BAZ2A;TOP2B;THOC1;XRCC5;RFX3;MEF2C;PGR;NFE2L1;APLP2 ;SRRT;KDM2B;CERS4;KIF4A;IKZF5;POLR2E;HIF1A;APEX1;DNTTIP 1;RPAP1;MYEF2;NUCB1;PDCD5;LHX2;DLX4;SUPT6H;DBX1;TIMEL ESS;CHD4;PPARD;CENPA;POLE4;STAT1;ORC4;EPAS1;CHD5;ASH 1L;TFAP2E;CR2;ARID1A;RPA2;ATF6;DNMT3A;MXI1;TAF12;MTERF 2;TARDBP;POLK;ZNF644;TWIST1;RNF11;BAZ2B;RBPJL;HIF3A;SO X4;GTF3C4;GTF2F1;DTD1;MAX;STAT5A;KLF2;FEZF1;PARP2;ZNF2 36;TOP2A;MYBBP1A;TERF2;RFXAP;FBH1;DDB2;TFCP2;EMX1;RE V1;SETDB2;TTC5;VEZF1;TBR1;XPA;APTX;DMRT1;NUSAP1;HNRN 90 GO cla ss GO clu ster GO term ID Human ortholog gene symbols PD;AEBP2;TCF12;POLG;ZFHX3;SMAD4;NFIC;DMRTA2;HDGF;CER S2;POGZ;ASXL2;SHPRH;CCDC25;ST18;POLE3;GTF3C5;THRB;KIN ;ZIC1;FEZF2;TGIF2LX;RABGEF1;BATF;SUPV3L1;BRPF1;ZNF618;Z NF362;PKNOX1;SAFB;GMEB1;FUBP1;LHX8;ZNF281;H3- 3A;SMARCAD1;POLR2H;RAD54L2;UNCX;GBX1;HNRNPK;GTF2A1; POLL;MCM7;FAM111A;TERF2IP;KMT2D;STAT3;ZEB2;HIC2;ZBTB26 ;CAMTA1;KLF17;ESRRA;THAP2;AGFG1;ENDOV;LIG4;NR1D2;BANF 1;DRAP1;ERCC4;RMI2;NR2F1;SOX11;DEAF1;TOP3A;JUN;MAF;PU F60;BHLHA15;ZBTB20;RFX7;TSHZ2;POU3F1;POLR1D;H1- 6;TCEA1;NHEJ1;H3- 5;ZNF292;MYT1;LONP1;XRCC2;HES5;QRICH1;ZNF536;TOP1;FOX O6;SP9;ZNF709;PRKDC;C17orf49;SEBOX MF 6 GO:000368 2 CREBBP;SCMH1;NCAPH2;NUCKS1;MBD3;SMARCE1;MLH1;TOP2B ;SMARCA2;CCNT2;TDRD3;FUS;IFT74;L3MBTL2;STAG2;FMR1;LHX 2;KAT2A;EZH1;KMT5B;RNF8;PHF1;KAT2B;CENPA;ASH1L;MLH3;D NMT3A;ENY2;NCAPH;RNF2;TTC21B;RBPJL;KLHDC3;UXT;PARP2; TOP2A;TTC5;CDK9;APTX;DMRT1;HNRNPD;POLG;CBX4;SMAD4;B RD4;ASXL2;DLX1;SETD7;SCML4;RBMX;THRB;CDYL;FEZF2;RNF20 ;AUTS2;PKNOX1;MCM3AP;SAFB;SMARCAD1;BAP1;WDR82;NIPBL; PRKCB;CTNNB1;SIN3A;PHF8;TADA2B;CENPS;NPM1;HES5;TOP1; SAMD13;BRD2;CORT;BAHCC1;PCGF2; MF 6 GO:000803 5 PLTP;LRP8;ABCA1 MF 6 GO:001921 2 TESC;ARPP19;ANP32E;ENSA MF 6 GO:009700 1 PLTP;CERT1;VDAC1