Skip to content

Advertisement

  • Research article
  • Open Access

Variable gene transcription underlies phenotypic convergence of hypoxia tolerance in sculpins

BMC Evolutionary Biology201818:163

https://doi.org/10.1186/s12862-018-1275-1

  • Received: 15 November 2017
  • Accepted: 18 October 2018
  • Published:

Abstract

Background

The degree by which mechanisms underlying phenotypic convergence are similar among taxa depends on the number of evolutionary paths available for selection to act upon. Likelihood of convergence will be influenced by an interplay of factors such as genetic architecture, phylogenetic history and population demography. To determine if there is convergence or divergence in mechanisms underlying phenotypic similarity, we assessed whether gene transcription patterns differed among species with similar levels of hypoxia tolerance.

Results

Three species of marine fish from the superfamily Cottoidea (smoothhead sculpin [Artedius lateralis], sailfin sculpin [Nautichthys oculofasciatus] and Pacific staghorn sculpin [Leptocottus armatus]), all of which have previously been shown to share the same level of hypoxia tolerance, were exposed to short-(8 h) and longer-term (72 h) hypoxia and mRNA transcripts were assessed using a custom microarray. We examined hypoxia-induced transcription patterns in metabolic and protein production pathways and found that a high proportion of genes associated with these biological processes showed significant differences among the species. Specifically, the data suggest that the smoothhead sculpin, unlike the sailfin sculpin and the Pacific staghorn sculpin, relied on amino acid degradation rather than glycolysis or fatty acid oxidation to generate ATP during hypoxia exposure. There was also variation across the species in the transcription of genes involved in protein production (e.g. mRNA processing and protein translation), such that it increased in the smoothhead sculpin, decreased in the sailfin sculpin and was variable in the Pacific staghorn sculpin.

Conclusions

Changes in metabolic and protein production pathways are part of the key responses of fishes to exposures to environmental hypoxia. Yet, species with similar overall hypoxia tolerance exhibited different transcriptional responses in these pathways, indicating flexibility and complexity of interactions in the evolution of the mechanisms underlying the hypoxia tolerance phenotype. The variation in the hypoxia-induced transcription of genes across species with similar hypoxia tolerance suggests that similar whole-animal phenotypes can emerge from divergent evolutionary paths that may affect metabolically important functions.

Keywords

  • Convergent evolution
  • Transcriptomics
  • Hypoxia
  • Fish
  • Sculpin

Background

Natural selection has long been implicated as the evolutionary cause of convergent evolution of phenotypic traits. In theory, phenotypic convergence among species may evolve through similar or different genetic changes depending on the number of evolutionary paths available for selection to act upon. For example, genetic constraints, such as pleiotropy or limited available genetic variation may decrease the number of ways in which an animal can evolve in response to an environmental stressor, leading to similar genetic solutions [13]. In contrast, if genetic constraints are minimal and genetic variation is high, there may be a greater number of pathways available for evolution, and different genetic solutions could underlie convergent phenotypes. There has been increased effort in recent years to identify the underlying genetic and molecular mechanisms that yield trait convergence, with the long-term goal of understanding the principles that govern the genetic basis of phenotypic convergence [48].

Genetic convergence has been examined across hierarchical levels in a number of different animal systems [4, 9, 10], revealing the complexity of how similar phenotypes can evolve in response to the same ecological pressure. Parallel genotypic adaptation can involve differing levels of biological organization: from the exact same mutation in the same gene, to different mutations in the same gene, through different genes in the same gene network or pathway [11, 12]. With the advent of high-throughput transcriptomics, similarities in gene expression in pathways and networks across populations or species adapting to similar environmental stressors have been identified [5, 1318]. Focusing on transcriptomics is important because it provides an understanding of the variation in genome regulation in response to the environment, ultimately providing an important link in elucidating genotype-environment interactions, phenotypic plasticity and phenotypic evolution [1921].

Environmental hypoxia (i.e., low O2 levels) is well known to have significant ecological and physiological consequences in animals, particularly in aquatic organisms, which can experience hypoxia frequently due to the low solubility of O2 in water. As a result, there have been multiple, independent origins of hypoxia tolerance among diverse fish lineages [22]. Accumulating evidence suggest that hypoxic survival in fish is linked to the maintenance of energy supply and demand [23], which is most often achieved through a combination of modifications for O2 uptake to sustain aerobic metabolism, increased reliance on O2-independent ATP production (e.g. glycolysis), as well as decreased energy demand through a coordinated suppression of energetically expensive processes such as protein production [2426]. Modifications to these processes can be brought about through differences in gene transcription [27, 28], post-translational modifications [29], as well as physiological [30, 31] and behavioural [32] changes. Although a great deal is known about some of the mechanisms that confer hypoxia tolerance, no study has directly examined whether species with similar levels of hypoxia tolerance show convergence in the underlying mechanisms. Specifically, have species in the wild that are subjected to the same degree of hypoxia evolved these complex phenotypes via convergence at the level of gene transcription in the key hypoxic response pathways?

We exposed three species of marine fish from different clades in the superfamily Cottoidae [33] to hypoxia for up to 72 h and sampled liver for transcriptomics analysis. The smoothhead sculpin (Artedius lateralis), sailfin sculpin (Nautichthys oculofasciatus) and Pacific staghorn sculpin (Leptocottus armatus) all inhabit the high subtidal environment, which is susceptible to periodic decreases in O2, and show the same level of whole-animal hypoxia as indicated by the critical O2 tension of O2 consumption rate (Pcrit), a complex phenotype [34]. Available data on the smoothhead sculpin and the Pacific Staghorn sculpin shows that the species also have similar time to loss of equilibrium when exposed to low O2 levels enforcing the idea that these species have the same level of tolerance to environmental hypoxia [35].

The hypoxia tolerances of the focal species are intermediate between hypoxia tolerant species that live in the tidepool environment and experience large and frequent O2 fluctuations and the hypoxia intolerant species that live in the deeper, O2 stable environments. Using the hypoxia tolerant tidepool sculpin (Oligocottus maculosus) and hypoxia intolerant silverspotted sculpin (Blepsias cirrhosis), we have previously shown that these species have highly divergent patterns of gene transcription in both the induction or repression patterns of the transcriptional response and the timing of the hypoxia-induced transcriptional change [36]. It remains unknown, however, whether species with similar levels of hypoxia tolerance also show divergence in gene transcription patterns in pathways known to impact hypoxic survival, which would suggest that differences in genome regulation can underlie equivalent complex phenotypes such as hypoxia tolerance.

Methods

Experimental animals

Individuals of Artedius lateralis (smoothhead sculpin) and Leptocottus armatus (Pacific staghorn sculpin) were collected in July and August near the Bamfield Marine Sciences Centre (BMSC), Bamfield, British Columbia. Smoothhead sculpins and Pacific Staghorn sculpins were captured at Ross Islets (48°52′24.0”N 125°09′42.0”W) and Bamfield Inlet (48°49′06.1”N 125°08′30.9”W) using seine nets during low tide and baited minnow traps. The two collection sites are prone to decreases in O2 as a result of the ebb and flow of the tides and both sculpin species have been found in the locations during low tide cycle, indicating that the species are exposed to similar periodic environmental hypoxia (personal observations). Sculpins collected at BMSC were transported to The University of British Columbia and a third species, Nautichthys oculofasciatus (sailfin sculpin) was acquired from the Vancouver Aquarium, British Columbia, Canada. The sailfin sculpins were first generation offspring from wild fish caught in the waters around Stanley Park, Vancouver, BC. All three species were held in fully aerated recirculating seawater at 12 °C and 30 ppt salinity for at least 2 months prior to experimentation. Fish were fed daily with prawns and krill up to 24 h before the experiments when feeding ceased. Animal collections were approved by Department of Fisheries and Oceans (License XR 143 2010) and Bamfield Marine Sciences Centre animal care committee (AUP RS-11-17) and followed the appropriate guidelines. All experiments were approved by The University of British Columbia animal care committee (A09–0611).

Critical O2 tension

Critical O2 tension (Pcrit; water PO2 at which an animal no longer maintains routine O2 consumption rate and the rate decreases with a decline in water PO2) is related to hypoxia tolerance in sculpins [35] and Pcrit values for smoothhead and Pacific staghorn sculpins taken from [34] were 35.7 ± 6.9 Torr and 37.4 ± 1.2 Torr, respectively. Following protocols described in detail in [37], we determined that the Pcrit for sailfin sculpin was 35.5 ± 2.3 Torr (n = 6) at 12 °C and 30 ppt salinity. The Pcrit values were not statistically different between species (ANOVA, p = 0.900).

Hypoxia exposure

For each species, individuals were transferred into four separate, 136 l tanks of a custom built recirculating system, the Low Oxygen Control and Monitoring Aquatic Research System (Integrated Aqua Systems, Inc.). Fish were allowed to recover in air-saturated water for 24 h, after which O2 levels were decreased to 23 Torr, a level of hypoxia corresponding to 65% Pcrit of all three species. Liver was sampled from each fish at 3, 8, 24, 48 and 72 h of hypoxia and two normoxic fish were also sampled at each of the hypoxic time points. Fish were removed from the tank and immediately euthanized by exposure to an overdose of benzocaine (125 mg/mL; Sigma-Aldrich). Tissue was dissected, flash frozen in liquid N2 and stored at − 80 °C until analysis.

Microarray experiments

The protocols used for RNA extraction, reverse-transcription of RNA to cDNA, fluorescent labeling of the cDNA and hybridization of the microarrays are described in detail in [36]. Briefly, RNA was extracted from the liver of each species using TRIzol (Invitrogen). Total RNA was then purified using glass-fibre filter columns (Qiagen) and reverse transcribed using MMLV-reverse transcriptase (Epicentre). The cDNA samples were purified using QIAquick PCR purification kits (Qiagen) and each sample was labeled separately with Cy5 and Cy3 dyes. Fluorescently labeled cDNA samples were hybridized to a custom cDNA array from woolly sculpin (Clinocottus analis). A total of 25 samples per species were used: 5 individuals sampled at normoxia and 4 individuals sampled at each hypoxic time-point. Hybridizations were only performed between 2 samples from the same species hybridized on the woolly sculpin array. We used an interwoven loop design [38], where each sample was hybridized to either 2 or 4 arrays with fluor-reversal, creating a separate loop for each species (Additional file 1: Figure S1).

Microarray and statistical analysis

The custom woolly sculpin array is comprised of approximately 9600 cDNA clones (array elements; [36]). A total of 1700 clones were sequenced and the resulting non-redundant contigs were annotated with the gene name only if the matched sequences yielded a hit with an E-value of <1e-4 in SwissProt (for details see [36]). For the purpose of this manuscript, the term ‘clone’ refers to all array elements (sequenced and non-sequenced), while the term ‘gene’ refers only to the sequenced and annotated clones. Sequence divergence between the woolly sculpin used to create the microarray platform and the sculpin species under investigation may result in poor hybridization at certain clones; however, we attempted to minimize the effect of poor hybridization by only considering clones with expression levels > 2 standard deviations above background for any species. Using this criterion, 3294 clones of the possible 9600 clones were analyzed.

Raw data were log2-transformed and lowess normalized using the MAANOVA package [39] in the R programming language [40]. To test for the effects of time in hypoxia, differential expression for each clone across time-points was detected using linear mixed models (‘array’ as a random effect and ‘dye’ and ‘time’ as fixed effects) with 100 permutations and the F-test statistic. False discovery rate [41] in the q value package [42] in R was used to determine significance level (set at q < 0.01). Clones showing a main effect of time in hypoxia in at least one of the species of sculpin (2734 clones in total; 1670 clones for smoothhead sculpin, 1183 clones for sailfin sculpin and 1778 clones for Pacific staghorn sculpin) were first categorized into annotated and non-annotated groups. Annotated genes were assigned to gene ontology biological process categories using searches in EMBL-EBI QuickGO (www.ebi.ac.uk/QuickGo/) and Uniprot (http://www.uniprot.org). Only genes associated with energy metabolism and protein production were further analyzed to determine if the transcription patterns differed among the species.

To determine whether the direction and/or rate of transcriptional response varied among species we compared the slope of the transcription responses for short-term hypoxia exposure (0, 3 h, 8 h) and long-term hypoxia exposure (0, 24 h, 48 h, 72 h). The time-course was partitioned into short- and long-term responses because: 1) daily tidal events are the primary cause of hypoxic exposure in the environments inhabited by these species, resulting in ecologically relevant hypoxic exposure of up to 8 h; 2) shallow coastal environments are increasingly influenced by anthropogenic eutrophication resulting in longer-term, physiologically challenging hypoxia exposures [43]; and 3) our previous work comparing the responses of hypoxia-tolerant and -intolerant sculpins demonstrated a strong effect of time on transcription with differences between the early and the late hypoxia exposure [36]. A maximum-likelihood model was applied to the data to determine the best fit intercept and linear slope for each species over short- and long-term hypoxia, using the subplex method of the optim function, as implemented in the find.mle routine of the diversitree package [44]. For each species pair (smoothhead sculpin vs. sailfin sculpin, smoothhead sculpin vs. Pacific staghorn sculpin, and sailfin sculpin vs. Pacific staghorn sculpin), we then used a likelihood-ratio test to determine whether the slopes were significantly different by comparing the full likelihood model (which allowed both species to have different slopes) to a constrained model with a single slope. If the drop in log-likelihood between the full and the constrained model was > 1.92 (the critical value for a chi-square distribution with 1 degree of freedom and an alpha level of 0.05) we then rejected the hypothesis that the slopes were the same for the two species. From this analysis, we broadly classified the data into 4 categories of transcriptional response (Fig. 1): differences among the species during short-term hypoxia exposure, differences among the species during long-term hypoxia exposure, differences among the species in both the short- and long-term hypoxia exposure, and no differences among the species.
Fig. 1
Fig. 1

Transcript levels of a representative gene for each category of the transcription response. Smoothhead sculpin (black line, square symbol), sailfin sculpin (blue line, diamond symbol) and Pacific staghorn sculpin (red line, inverted triangle symbol) were exposed to 72 h of hypoxia. Opaque lines represent the portion of the time-course that is different among the species, while transparent lines represent the portion of the time-course for which transcription does not differ among the species. Letters represent significant difference in transcript levels between species (A-B represent difference in the short term hypoxia and X-Y represent differences in the long term hypoxia). Full gene names are as follows: fructose-bisphosphate aldolase B (ALDOB), eukaryotic translation initiation factor 4E-1A (EIF4E1A), HIG1 domain family member 1A (HIGD1A) and NADH dehydrogenase 1 alpha subcomplex subunit 1 (NDUFA1)

Wild caught versus lab-reared species

In this study, individuals of the smoothhead sculpin and the Pacific staghorn sculpin were wild-caught, while individuals of the sailfin sculpin were first generation lab-reared. In theory this may complicate interpretation, as the wild-caught species may have been exposed to hypoxia during development, which can have consequences on the adult hypoxic phenotype [45]. However, an examination of the patterns of all transcriptionally-responsive clones using heat maps across the three species produced no clear indication to suggest that the overall transcriptional response of the sailfin sculpin differed to a great degree from the other two species (data not shown). Of course, this does not exclude the possibility that some of the patterns described in this study are a result of differences in developmental plasticity. One could potentially get around the effects of developmental plasticity by examining first generation lab-reared individuals for all species (caveat: wild-caught individuals of many sculpin species are difficult to breed under laboratory conditions), although epigenetic effects can persist across many generations increasing the difficulty of unraveling the genetic mechanisms of survival in hypoxia.

Results and discussion

To determine whether hypoxia tolerance is a result of convergence of traits at the underlying molecular level we examined temporal patterns of gene transcription in response to hypoxia exposure in three species of sculpin from superfamily Cottoidae. Intriguingly, in spite of their similar overall hypoxia tolerance, genes associated with energy metabolism and protein production, the key pathways involved in hypoxia survival, have divergent transcription patterns (59% and 73% of genes respectively) among the three species. This indicates that different genomic solutions were used to achieve the same phenotypic outcome.

Energy metabolism

Transcription of metabolic genes varied among the three sculpin species, suggesting that the species may rely on different metabolic pathways to generate ATP during hypoxia exposure. These transcriptional data strongly indicate that the three sculpin species use different and diverse metabolic pathways for ATP generation during hypoxia exposure (Fig. 2), challenging the classic dogma that hypoxia survival is solely dependent on the Pasteur effect and the up-regulation of glycolytic function for ATP production [22].
Fig. 2
Fig. 2

Schematic representation of genes associated with metabolism. Arrows represent gene transcription changes in response to hypoxia in the smoothhead sculpin (black), sailfin sculpin (blue) and Pacific staghorn sculpin (red). Gray arrows represent genes with similar transcription patterns among the species. Short arrows indicate transcription patterns during short-term hypoxia, long arrows indicate transcription patterns during long-term hypoxia, double arrows indicate transcription patterns during short- and long-term hypoxia, a horizontal line indicates no change from normoxia and the letter ‘v’ represents a variable response. Full names for each gene abbreviation as well as complete transcription profiles of the genes are found in Additional files 2, 4 and 5 (Figures S2, S4-S5).

In the smoothhead sculpin, transcript level changes in response to hypoxia suggest this species relies on enhanced amino acid degradation rather than glycolysis or fatty acid oxidation to generate substrates for the tricarboxylic acid (TCA), and thereby support ATP generation by oxidative phosphorylation (Fig. 2). Transcript levels of genes associated with amino acid degradation and the TCA cycle showed an increase in transcript levels during long-term hypoxia, while transcript levels of genes associated with amino acid synthesis decreased as compared to normoxic levels (Fig. 2). A reduced reliance on glycolysis during hypoxia is also consistent with the observation of a reduction in the transcript levels of glycolytic genes (e.g., fructose-bisphosphate aldolase B [ALDOB], glyceraldehyde-3-phosphate dehydrogenase [GAPDH]) and genes associated with glycogen degradation (glycogen debranching enzyme [GDE]; Fig. 2, Additional file 2: Figure S2). The smoothhead sculpin also had a reduction in transcript levels for genes associated with fatty acid oxidation during short- and long-term hypoxia exposure (e.g., hydroxyacyl-coenzyme A dehydrogenase [HADH]; Fig. 2, Additional file 2: Figure S2).

If gene transcript levels relate to biological function, these data suggest that in addition to the typical hypoxia-induced reliance on glycolysis for ATP generation, prolonged hypoxia exposure in smoothhead sculpins induces a greater reliance on amino acid catabolism as a source of ATP. An increased reliance on O2-dependent ATP production via amino acid catabolism may seem like a surprising response for an organism to have in an O2 deprived environment, but previous studies have shown enhanced amino acid catabolism during hypoxia in at least two other fish species (Tilapia mossambica; [46] and Oncorhynchus mykiss; [47]).

In contrast to the smoothhead sculpin, both the sailfin sculpin and the Pacific staghorn sculpin showed a general increase in the transcription of glycolytic genes during short- and long-term hypoxia exposure (e.g., GAPDH; Fig. 2, Additional file 2: Figure S2). This would suggest an increased reliance on glycolysis during hypoxic exposure, a response that has been well documented in fishes, reptiles and mammals [4851]. Both species also showed a short-term increase in transcript levels of genes associated with fatty acid oxidation (e.g., HADH) followed by a decrease in transcript levels during long-term hypoxia (e.g., medium-chain specific acyl-CoA dehydrogenase [ACADM]; Fig. 2, Additional file 2: Figure S2). The increase or lack of change in transcription of fatty acid oxidation genes during short-term hypoxia exposure followed by a decrease in the transcription of the genes during long-term hypoxia exposure have been noted in the heart of rats [52]. The decline in fatty acid oxidation during hypoxia has been thought to contribute to a decrease in energy expenditure as well as an avoidance of lipotoxicity [53, 54].

The sailfin and Pacific staghorn sculpins showed divergent responses in gene transcription associated with other metabolic processes. Transcription patterns of genes associated with amino acid synthesis and degradation in the sailfin sculpin were similar to those of the smoothhead sculpin, with a decrease in beta-ureidopropionase [UPB1] and an increase in 4-hydroxyphenylpyruvate dioxygenase [HPD] (Fig. 2, Additional file 2: Figure S2). However, this was in contrast to the Pacific staghorn sculpin, which showed transcriptional patterns that suggest both increased amino acid synthesis and degradation (Fig. 2). We also identified a significant increase in genes involved in glycogen degradation in the sailfin sculpin while in the Pacific staghorn sculpin there was shift from glycogen synthesis in the short-term to glycogen degradation during long-term hypoxia exposure.

Protein production, protein localization and protein folding

Protein production is a significant energy sink and inhibition of this process during hypoxia is thought to be an important step in conserving the limited energy stores [2426]. Multiple steps are involved in cellular protein production, including gene transcription, RNA processing and translation (Fig. 3). Differences in the transcriptional patterns of genes associated with these processes suggest that during short and long-term hypoxia exposure the smoothhead sculpin increased protein production through up-regulation of genes associated with translation, while the sailfin sculpin showed a corresponding decrease in these same genes (e.g., eukaryotic translation initiation factor 4 gamma 2 [EIF4G2] and eukaryotic translation initiation factor 4E-1A [EIF4E1A]; Fig. 3, Additional file 3: Figure S3). The Pacific staghorn had a variable response in genes associated with protein production, as highlighted by the transcriptional patterns of both the genes associated with RNA processing (e.g. ribonuclease inhibitor [RNHI] versus RNA-binding motif, single-stranded-interacting protein 2 [RBMS2]) and genes associated with translation (e.g. EIF4G2 versus EIF4E1A; Fig. 3, Additional file 3: Figure S3).
Fig. 3
Fig. 3

Schematic representation of genes associated with protein production, localization and folding. Arrows represent gene transcription changes in response to hypoxia in the smoothhead sculpin (black), sailfin sculpin (blue) and Pacific staghorn sculpin (red). Gray arrows represent genes with similar transcription patterns among the species. Short arrows indicate transcription patterns during short-term hypoxia, long arrows indicate transcription patterns during long-term hypoxia, double arrows indicate transcription patterns during short- and long-term hypoxia, a horizontal line indicates no change from normoxia and the letter ‘v’ represents a variable response. See Additional files 3, 4 and 5 (Figures S3-S5) for full gene names and complete transcription profiles of the genes

To fully examine protein production, it is also important to consider steps such as protein folding, targeting and localization. Although transcription patterns in a large proportion of genes associated with transcription, RNA processing and translation differed among the species, the majority of genes associated with protein localization and protein folding were transcriptionally similar among the species (Fig. 3, Additional file 4: Figure S4, Additional file 5: Figure S5). In particular, all three species showed an increase in transcription of genes associated with protein folding, especially during long-term hypoxia exposure. This is consistent with previous work that shows hypoxia exposure increases protein-folding chaperones in the liver of rats (heat shock protein 70; [55]) and in the Chinese shrimp (heat shock protein 90; Fenneropenaeus chinensis; [56]).

Linking gene transcription to phenotype

From genotype to phenotype, variation in gene transcription is important to consider as it can be heritable and an outcome of both adaptive and neutral evolutionary processes [20, 5759]. However, the relationship between gene transcription and phenotype is complex and causally linking variation in gene transcription to phenotypic variation can be difficult. Gene transcription is but one step of many along the pathway towards building a phenotype such as hypoxia tolerance. For example, hypoxia exposure has a significant impact on the translation efficiency of genes [60], and it is possible that the variation among the species at gene transcription level is accentuated or diminished at subsequent levels of biological organization. Therefore, convergence may be occurring not only at the level of hypoxia tolerance but in the underlying metabolic and protein production pathways among the three species of sculpin. While further work is required to determine if there is convergence at subsequent levels of biological organization, it is clear that there is divergent variation in gene transcription underlying hypoxia tolerance among the three species of sculpin.

Comparisons with a hypoxia tolerant sculpin

Our previous work showed that the transcription patterns in the hypoxia-tolerant tidepool sculpin aligned with the characteristics that have long been attributed to hypoxia-tolerant animals, including an increase in transcript levels of glycolytic genes and a decrease in transcripts of genes associated with energetically expensive processes such as cell proliferation [36]. The three intermediate–tolerant species investigated in this study shared some of the same transcription patterns as the tidepool sculpin in genes considered important for hypoxia tolerance. For example, similar to the tidepool sculpin, the smoothhead sculpin showed a decrease in transcription of genes involved in some of the pathways that require O2, such as fatty acid oxidation, while the sailfin sculpin showed a decrease in protein production, and both the sailfin sculpin and Pacific staghorn sculpin showed an increase in glycolysis (Fig. 2, Fig. 3). Overall, there was significant variation among the three species in transcription patterns for genes that were a priori predicted to be important in hypoxia. The transcriptional patterns thus strongly suggest that the three species of sculpin have evolved similar levels of hypoxia tolerance via different underlying mechanisms.

Short-term versus long-term responses

Time is an important factor to consider when examining changes in gene transcription [19]. In our previous work [36] we showed that time in hypoxia had a strong effect on gene transcription, and that hypoxia-tolerant and -intolerant sculpins showed different temporal patterns of gene transcript levels. The short-term (8 h) hypoxia exposure is an ecologically relevant time-frame for species inhabiting the intertidal, while the longer duration of hypoxia lasting for several days is a physiological challenge that tests the species’ capacities for hypoxic survival. Therefore, it is not surprising that different mechanisms may shape the hypoxic response between the short and long exposures to hypoxia and we predicted that in the current study differences in transcription patterns among the species would occur in both the short- and long-term hypoxia exposure. Indeed, 16% of genes showed differences among species during short-term hypoxia exposure, 33% during long-term hypoxia exposure and 6% showed transcriptional differences across both time scales in pathways associated with energy production and protein production (percentages are based on genes represented in the figures that are expected to be a priori involved in processes related to hypoxia). As an example, transcription of glycolytic genes differed among species in both short- and long-term hypoxia, although most genes differed only in one time frame or the other (Fig. 2). Divergent transcriptional responses as well as strong temporal effects thus influence the underlying mechanisms of hypoxia tolerance in sculpins. However, caution should be noted, as these species were exposed to a single bout of hypoxia whereas in the wild the species encounter hypoxia on a cyclical basis. This may have an impact on the transcriptional patterns and a critical future step is to examine the transcriptional response of metabolic and protein production genes over time in species exposed to tidal-influenced cyclical hypoxia.

Conclusions

We found that three species of sculpin with similar hypoxia tolerance had very different patterns of transcription of genes involved in metabolism and protein production. These biological processes are known to have a significant influence on hypoxic survival and divergent transcription patterns underlying these traits suggest that different genetic changes were selected for during the evolution of intermediate hypoxia tolerance in the three species of sculpin. Our results indicate that there are multiple genomic pathways to achieve the same phenotype, and indicate that similarities in hypoxia tolerance among sculpins is due to phenotypic convergence rather than species relatedness.

Abbreviations

ACADM: 

Medium-chain specific acyl-CoA dehydrogenase

ACYL: 

ATP-citrate synthase

AKR1B1: 

Polyol pathway aldose reductase

ALDH6A1: 

Methylmalonate-semialdehyde dehydrogenase, acylating

ALDOB: 

Fructose-biphosphate aldolase B

ARF5: 

ADP-ribosylation factor 5

ATP: 

Adenosine triphosphate

ATP50: 

ATP synthase subunit O

ATP5H: 

ATP synthase subunit d

ATP5I: 

ATP synthase subunit e

ATPIF1: 

ATPase inhibitor

BHMT: 

Betaine—homocysteine S-methyltransferase 1

CCT6A: 

T-complex protein 1 subunit zeta

CCT7: 

T-complex protein 1 subunit eta

COX14: 

Cytochrome c oxidase assembly protein COX14

CTDSP2: 

Carboxy-terminal domain RNA polymerase II polypeptide A small phosphatase 2

DAO: 

D-amino-acid oxidase

EEF1G: 

Elongation factor 1-gamma

EIF1B: 

Eukaryotic translation initiation factor 1b

EIF4E1A: 

Eukaryotic translation initiation factor 4E-1A

EIF4EBP3: 

Eukaryotic translation initiation factor 4E-binding protein 3

EIF4G2: 

Eukaryotic translation initiation factor 4 gamma 2

EIF5A: 

Eukaryotic translation initiation factor 5A-1

ENO1: 

Alpha-enolase

FKBP1A: 

Peptidyl-prolyl cis-trans isomerase FKBP1A

GAPDH: 

Glyceraldehyde-3-phosphate dehydrogenase

GDE: 

Glycogen debranching enzyme

GLUD1: 

Glutamate dehydrogenase

GPI: 

Glucose-6-phosphate isomerase

GPT: 

Alanine aminotransferase 2

HADH: 

Hydroxyacyl-coenzyme A dehydrogenase

HGD: 

Homogentisate 1,2-dioxygenase

HIGD1A: 

HIG1 domain family member 1A

HPD: 

4-hydroxyphenylpyruvate dioxygenase

HSP90AA1: 

Heat shock protein HSP 90-alpha

HSP90AB1: 

Heat shock protein HSP 90-beta

HSPE1–10 kDa: 

Heat shock protein

IPO7: 

Importin-7

MPC1: 

Mitochondrial pyruvate carrier 1

NDUFA1: 

NADH dehydrogenase 1 alpha subcomplex subunit 1

NDUFS8: 

NADH dehydrogenase iron-sulfur protein 8

NR1D2: 

Nuclear receptor subfamily 1 group D member 2

NSA2: 

Ribosome biogenesis protein NSA2 homolog

Pcrit

Critical O2 tension; water PO2 at which an animal no longer maintains routine O2 consumption rate and the rate decreases with a decline in water PO2

PDIA4: 

Protein disulfide-isomerase A4

PDK: 

Pyruvate dehydrogenase kinase isozyme 4

PGM1: 

Phosphoglucomutase-1

PO2

Partial pressure of oxygen

ppp1r3cb: 

Protein phosphatase 1 regulatory subunit 3C-B

QSOX1: 

Sulfhydryl oxidase 1

RBMS2: 

RNA-binding motif, single-stranded-interacting protein 2

RNHI: 

Ribonuclease inhibitor

RPE: 

Ribulose-phosphate 3-epimerase

RPL10: 

60S ribosomal protein L10

RPL30: 

60S ribosomal protein L30

RPL39: 

60S ribosomal protein L39

RPS21: 

40S ribosomal protein S21

RPS27: 

40S ribosomal protein S27

RPS7: 

40S ribosomal protein S7

SAP18: 

Histone deacetylase complex subunit SAP18

SEC61A: 

Protein transport protein Sec61 subunit alpha

SEC63: 

Translocation protein SEC63 homolog

SPCS2: 

Signal peptidase complex sunbunit 2

SSR3: 

Translocon-associated protein subunit gamma

TCA: 

Tricarboxylic acid

UPB1: 

Beta-ureidopropionase

UQCRH: 

Cytochrome b-c1 complex subunit 6

Declarations

Acknowledgements

We thank B. Speers-Roesch, M. Siegle and D. Groom for their assistance in fish collection, and G. Lau, R. Dhillon, G. Cox and S. Lee for their help in sampling fish. The authors would also like to thank the staff at Bamfield Marine Sciences Centre.

Funding

This work was supported by a Natural Sciences and Engineering Research Council (NSERC) of Canada Discovery grant to J.G.R.; an Alexander Graham Bell Canada Graduate Scholarship (CGS) from NSERC and a Michael Smith Foreign Study Travel Award to M.M. and a Postdoctoral Fellowship from NSERC to A.C.G. This research was also funded by the National Science Foundation under Grant No. 0706908 to M.L.R. The funding bodies had no role in the design of the study, analysis or interpretation of the data.

Availability of data and materials

The datasets supporting the conclusions of this article have been deposited in the ArrayExpress database at EMBL-EBI (www.ebi.ac.uk/arrayexpress) under accession number E-MTAB-6233. The datasets supporting this article have been uploaded as part of the Supplementary Material.

Authors’ contributions

MM and JGR designed the experiment; MM carried out hypoxic exposures; MM and MLR performed the laboratory work; MM and ACG carried out the statistical analyses; MM wrote the manuscript; all authors provided input on the manuscript; and all authors have read and approved the manuscript.

Ethics approval

Animal collections were approved by Department of Fisheries and Oceans (License XR 143 2010) and Bamfield Marine Sciences Centre animal care committee (AUP RS-11-17) and followed the appropriate guidelines. All experiments were approved by The University of British Columbia animal care committee (A09–0611).

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Authors’ Affiliations

(1)
Department of Zoology, University of British Columbia, 6270 University Blvd, Vancouver, BC, V6T 1Z4, Canada
(2)
Department of Biological Sciences, University of Southern California, 3616 Trousdale Parkway, Los Angeles, CA 90089-0371, USA
(3)
Department of Genetics, Cell Biology and Development, University of Minnesota, 321 Church Street, Minneapolis, MN 55455, USA
(4)
Bamfield Marine Sciences Centre, 100 Pachena Dr, Bamfield, BC, V0R 1B0, Canada

References

  1. Chevin L-M, Martin G, Lenormand T. Fisher’s model and the genomics of adaptation: restricted pleiotropy, heterogenous mutation, and parallel evolution. Evolution. 2010;64:3213–31.View ArticleGoogle Scholar
  2. Christin P-A, Weinreich DM, Besnard G. Causes and evolutionary significance of genetic convergence. Trends Genet. 2010;26:400–5.View ArticleGoogle Scholar
  3. Losos JB. Convergence, adaptation, and constraint. Evolution. 2011;65:1827–40.View ArticleGoogle Scholar
  4. Hoekstra HE, Hirschmann RJ, Bundey RA, Insel PA, Crossland JP. A single amino acid mutation contributes to adaptive beach mouse color pattern. Science. 2006;313:101–4.View ArticleGoogle Scholar
  5. Gross JB, Borowsky R, Tabin CJ. A novel role for Mc1r in the parallel evolution of depigmentation in independent populations of the cavefish Astyanax mexicanus. PLoS Genet. 2009;5:e1000326.View ArticleGoogle Scholar
  6. O’Quin KE, Hofmann CM, Hofmann HA, Carleton KL. Parallel evolution of opsin gene expression in African cichlid fishes. Mol Biol Evol. 2010;27:2839–54.View ArticleGoogle Scholar
  7. Rosenblum EB, Rompler H, Schoneberg T, Hoekstra HE. Molecular and functional basis of phenotypic convergence in white lizards at White Sands. Proc Natl Acad Sci U S A. 2010;107:2113–7.View ArticleGoogle Scholar
  8. Conte GL, Arnegard ME, Peichel CL, Schluter D. The probability of genetic parallelism and convergence in natural populations. Proc R Soc B. 2012;279:5039–47.View ArticleGoogle Scholar
  9. Linnen CR, Kingsley EP, Jensen JD, Hoekstra HE. On the origin and spread of an adaptive allele in deer mice. Science. 2009;325:1095–8.View ArticleGoogle Scholar
  10. Natarajan C, Hoffmann FG, Weber RE, Fago A, Witt CC, Storz JF. Predictable convergence in hemoglobin function has unpredictable molecular underpinnings. Science. 2016;354:336–9.View ArticleGoogle Scholar
  11. Manceau M, Domingues VS, Linnen CR, Rosenblum EB, Hoekstra HR. Convergence in pigmentation at multiple levels: mutations. genes and function Philos Trans R Soc B. 2010;365:2439–50.View ArticleGoogle Scholar
  12. Rosenblum EB, Parent CE, Brandt EE. The molecular basis of phenotypic convergence. Annu Rev Ecol Evol Syst. 2014;45:203–26.View ArticleGoogle Scholar
  13. Derome N, Duchesne P, Bernatchez L. Parallelism in gene transcription among sympatric lake whitefish (Coregonus clupeaformis Mitchill) ecotypes. Mol Ecol. 2006;15:1239–49.View ArticleGoogle Scholar
  14. Roberge C, Einum S, Guderley H, Bernatchez L. Rapid parallel evolutionary changes of transcription profiles in farmed Atlantic salmon. Mol Ecol. 2006;15:9–20.View ArticleGoogle Scholar
  15. Fisher MA, Oleksiak MF. Convergence and divergence in gene expression among natural populations exposed to pollution. BMC Genomics. 2007;8:108–18.View ArticleGoogle Scholar
  16. Jeukens J, Bittner D, Knudsen R, Bernatchez L. Candidate genes and adaptive radiation: insights from transcriptional adaptation to the limnetic niche among coregonine fishes (Coregonus spp., Salmonidae). Mol Biol Evol. 2009;26:155–66.View ArticleGoogle Scholar
  17. Whitehead A, Pilcher W, Champlin D, Nacci D. Common mechanism underlies repeated evolution of extreme pollution tolerance. Proc R Soc B. 2011. https://doi.org/10.1098/rspb.2011.0847.View ArticleGoogle Scholar
  18. Ghalambor CK, Hoke KL, Ruell EW, Fischer EK, Reznick DN, Hughes KA. Non-adaptive plasticity potentiates rapid adaptive evolution of gene expression in nature. Nature. 2015;525:372–5.View ArticleGoogle Scholar
  19. Aubin-Horth N, Renn SCP. Genomic reaction norms: using integrative biology to understand molecular mechanisms of phenotypic plasticity. Mol Ecol. 2009;18:3763–80.View ArticleGoogle Scholar
  20. Whitehead A. Comparative genomics in ecological physiology: toward a more nuanced understanding of acclimation and adaptation. J Exp Biol. 2012;215:884–91.View ArticleGoogle Scholar
  21. Alvarez M, Schrey AW, Richards CL. Ten years of transcriptomics in wild populations: what have we learned about their ecology and evolution? Mol Ecol. 2015;24:710–25.View ArticleGoogle Scholar
  22. Hochachka PW, Somero GN. Biochemical adaptation: mechanism and process in physiological evolution. 2nd ed. New York: Oxford University Press; 2002.Google Scholar
  23. Boutilier RG. Mechanisms of cell survival in hypoxia and hypothermia. J Exp Biol. 2001;204:3171–81.PubMedGoogle Scholar
  24. Hochachka PW. Defense strategies against hypoxia and hypothermia. Science. 1986;231:234–41.View ArticleGoogle Scholar
  25. Hochachka PW, Buck LT, Doll CJ, Land SC. Unifying theory of hypoxia tolerance: molecular/metabolic defense and rescue mechanisms for surviving oxygen lack. Proc Natl Acad Sci U S A. 1996;93:9493–8.View ArticleGoogle Scholar
  26. Bickler PE, Buck LT. Hypoxia tolerance in reptiles, amphibians, and fishes: life with variable oxygen availability. Annu Rev Physiol. 2007;69:145–70.View ArticleGoogle Scholar
  27. Gracey AY, Troll JV, Somero GN. Hypoxia-induced gene expression profiling in the euryoxic fish Gillichthys mirabilis. Proc Natl Acad Sci U S A. 2001;98:1993–8.View ArticleGoogle Scholar
  28. Ton C, Stamatiou D, Liew CC. Gene expression profile of zebrafish exposed to hypoxia during development. Physiol Genomics. 2003;13:97–106.View ArticleGoogle Scholar
  29. Koumenis C, Naczki C, Koritzinsky M, Rastani S, Diehl A, Sonenberg N, Koromilas A, Wouters BG. Regulation of protein synthesis by hypoxia via activation of the endoplasmic reticulum kinase PERK and phosphorylation of the translation initiation factor eIF2a. Mol Cell Biol. 2002;22:7405–16.View ArticleGoogle Scholar
  30. Sollid J, De Angelis P, Gundersen K, Nilsson GE. Hypoxia induces adaptive and reversible gross morphological changes in crucian carp gills. J Exp Biol. 2003;206:3667–73.View ArticleGoogle Scholar
  31. Rutjes HA, Nieveen MC, Weber RE, Witte F, van Thillart GEEJM. Multiple strategies of Lake Victoria cichlids to cope with lifelong hypoxia include hemoglobin switching. Am J Physiol Regul Integr Comp Physiol. 2007;293:R1376–83.View ArticleGoogle Scholar
  32. Sloman KA, Wood CM, Scott GR, Wood S, Kajimura M, Johannsson OE, Almeida-Val VMF, Val AL. Tribute to R. G. Boutilier: the effect of size on the physiological and behavioural responses of Oscar, Astronotus ocellatus, to hypoxia. J Exp Biol. 2006;209:1197–205.View ArticleGoogle Scholar
  33. Knope ML. Phylogenetics of the marine sculpins (Teleostei: Cottidae) of the north American Pacific coast. Mol Phyl Evol. 2013;66:341–9.View ArticleGoogle Scholar
  34. Mandic M, Todgham AE, Richards JG. Mechanisms and evolution of hypoxia tolerance in fish. Proc R Soc B. 2009;276:735–44.View ArticleGoogle Scholar
  35. Mandic M, Speers-Roesch B, Richards JG. Hypoxia tolerance in sculpins is associated with high anaerobic enzyme activity in brain but not in liver or muscle. Physiol Biochem Zool. 2013;86:92–105.View ArticleGoogle Scholar
  36. Mandic M, Ramon ML, Gracey AY, Richards JG. Divergent transcriptional patterns are related to differences in hypoxia tolerance between the intertidal and the subtidal sculpin. Mol Ecol. 2014;23:6091–103.View ArticleGoogle Scholar
  37. Henriksson P, Mandic M, Richards JG. The osmorespiratory compromise in sculpins: impaired gas exchange is associated with freshwater tolerance. Physiol Biochem Zool. 2008;81:310–9.View ArticleGoogle Scholar
  38. Kerr MK, Churchill GA. Experimental design for gene expression microarrays. Biostatistics. 2001;2:183–201.View ArticleGoogle Scholar
  39. Wu H, Yang H, Sheppard K, Churchill G, Kerr MK, Cui X. MAANOVA: Tools for analyzing micro array experiments. 2013. R package version 1.30.0. Google Scholar
  40. R Core Team. R: A language and environmental for statistical computing. Vienna: R Foundation for Statistical Computing; 2013. URL http://www.R-project.org/ Google Scholar
  41. Storey JG. A direct approach to false discovery rates. J R Stat Soc B. 2002;64:479–98.View ArticleGoogle Scholar
  42. Dabney A, Storey JD. Q value: q-value estimation for false discovery rate control. 2013. R package version 1.34.0.Google Scholar
  43. Diaz RJ, Breitburg DL. The hypoxic environment. In: Richards JG, Farrell AP, Brauner CJ, editors. Hypoxia. San Diego: Elsevier Academic Press; 2009. p. 1–23.Google Scholar
  44. FitzJohn RG. Diversitree: comparative phylogenetics analyses of diversification in R. Methods Ecol Evol. 2012;3:1084–92.View ArticleGoogle Scholar
  45. Robertson CE, Wright PA, Koblitz L, Bernier NJ. Hypoxia-inducible factor-1 mediates adaptive developmental plasticity of hypoxia tolerance in zebrafish, Danio rerio. Proc R Soc B. 2014;281:1–9.View ArticleGoogle Scholar
  46. Kutty MN. Respiratory quotient and ammonia excretion in Tilapia mossambica. Mar Biol. 1972;16:126–33.Google Scholar
  47. Medale F, Parent JP, Vellas F. Responses to prolonged hypoxia by rainbow trout (Salmo gairdneri) I. free amino acids and protein in plasma, liver and white muscle. Fish Physiol Biochem. 1987;3:183–9.View ArticleGoogle Scholar
  48. Kelly DA, Storey KB. Organ-specific control of glycolysis in anoxic turtles. Am. J. Physiol. Regul. Integr. Comp. Physiol. 1988;255:R774–9.View ArticleGoogle Scholar
  49. Churchill TA, Cheetham KM, Simpkin S, Green CH, Wang LCH, Fuller BJ. Liver metabolism in cold hypoxia: a comparison of energy metabolism and glycolysis in cold-sensitive and cold-resistant mammals. J Comp Physiol B Biochem Syst Environ Physiol. 1994;164:396–404.View ArticleGoogle Scholar
  50. van den Thillart G, van Raaij M. Endogenous fuels; non-invasive versus invasive approaches. In: Hochachka PW, Mommsen TP, editors. Biochemistry and molecular biology of fishes. Amsterdam: Elsevier; 1995. p. 33–63.Google Scholar
  51. Webster KA. Evolution of the coordinate regulation of glycolytic enzyme genes by hypoxia. J Exp Biol. 2003;206:2911–22.View ArticleGoogle Scholar
  52. Sharma S, Taegtmeyer H, Adrogue J, Razeghi P, Sen S, Ngumbela K, Essop MF. Dynamic changes of gene expression in hypoxia-induced right ventricular hypertrophy. Am J Physiol Heart Circ Physiol. 2004;286:H1185–92.View ArticleGoogle Scholar
  53. Dewald O, Sharma S, Adrogue J, Salazar R, Duerr GF, Crapo JD, Entman ML, Taegtmeyer H. Downregulation of peroxisome proliferator-activated receptor-alpha gene expression in a mouse model of ischemic cardiomyopathy is dependent on reactive oxygen species and prevents lipotoxicity. Circulation. 2005;112:407–15.View ArticleGoogle Scholar
  54. Speers-Roesch B, Sandblom E, Lau GY, Farrell AP, Richards JG. Effects of environmental hypoxia on cardiac energy metabolism and performance in tilapia. Am. J. Physiol. Regul. Integr. Comp. Physiol. 2010;298:R104–19.View ArticleGoogle Scholar
  55. Aoe T, Inaba H, Kon S, Imai M, Aono M, Mizuguchi T, Saito T, Nishino T. Heat shock protein 70 messanger RNA reflects the severity of ischemia/hypoxia-reperfusion injury in the perfused rat liver. Crit Care Med. 1997;25:324–9.View ArticleGoogle Scholar
  56. Li F, Luan W, Zhang C, Zhang J, Wang B, Xie Y, Li S, Xiang J. Cloning of cytoplasmic heat shock protein 90 (FcHSP90) from Fenneropenaeus chinensis and its expression response to heat shock and hypoxia. Cell Stress Chaperon. 2009;14:161–72.View ArticleGoogle Scholar
  57. Oleksiak MF, Churchill GA, Crawford DL. Variation in gene expression within and among natural populations. Nature Genet. 2002;32:261–6.View ArticleGoogle Scholar
  58. Whitehead A, Crawford DL. Variation within and among species in gene expression: raw material for evolution. Mol Ecol. 2006;15:1197–211.View ArticleGoogle Scholar
  59. Whitehead A, Crawford DL. Neutral and adaptive variation in gene expression. Proc Natl Acad Sci U S A. 2006;103:5425–30.View ArticleGoogle Scholar
  60. Koritzinsky M, Magagnin MG, van den Beucken T, Seigneuric R, Savelkouls K, Dostie J, Pyronnet S, Kaufman RJ, Weppler SA, Voncken JW, Labmin P, Koumenis C, Sonenberg N, Wouters BG. Gene expression during acute and prolonged hypoxia is regulated by distinct mechanisms of translational control. EMBO J. 2006;25:1114–25.View ArticleGoogle Scholar

Copyright

© The Author(s). 2018

Advertisement