Differences in gene expression between high and low tolerance rainbow trout (Oncorhynchus mykiss) to acute thermal stress
Data files
Jan 13, 2025 version files 48.44 MB
-
README.md
28.47 KB
-
Supplementary_Files_1-12-PONE-D-24-01355-DRYAD-submission.zip
24.48 MB
-
Supplementary_Files-journal.pone.0312694.zip
23.93 MB
Abstract
Understanding the mechanisms that underlie the adaptive response of ectotherms to rising temperatures is key to mitigating the effects of climate change. We assessed the molecular and physiological processes that differentiate between rainbow trout (Oncorhynchus mykiss) with high and low tolerance to acute thermal stress. To achieve our goal, we used a critical thermal maximum trial in two strains of rainbow trout to elicit loss of equilibrium responses to identify high and low-tolerance fish. We then compared the hepatic transcriptome profiles of high and low-tolerance fish relative to untreated controls common to both strains to uncover patterns of differential gene expression and to gain a broad perspective on the interacting gene pathways and functional processes involved. We observed some of the classic responses to increased temperature (e.g., induction of heat shock proteins) but these responses were not the defining factors that differentiated high and low-tolerance fish. Instead, high-tolerance fish appeared to suppress growth-related functions, enhance certain autophagy components, better regulate neurodegenerative processes, and enhance stress-related protein synthesis, specifically spliceosomal complex activities, mRNA regulation, and protein processing through post-translational processes, relative to low-tolerance fish. In contrast, low-tolerance fish had higher transcript diversity and demonstrated elevated developmental, cytoskeletal, and morphogenic, as well as lipid and carbohydrate metabolic processes, relative to high-tolerance fish. Our results suggest that high-tolerance fish engaged in processes that supported the prevention of further damage by enhancing repair pathways, whereas low-tolerance fish were more focused on replacing damaged cells and their structures.
README: Differences in gene expression between high and low tolerance rainbow trout to acute thermal stress challenge
https://doi.org/10.5061/dryad.6wwpzgn74
Description of the data and file structure
The Supplementary Files provided in these folders complement the experimental findings presented in summarized in the following publication: Turner LA, Easton AA, Ferguson MM, and Danzmann RG (2025). Differences in gene expression between high and low tolerance rainbow trout (Oncorhynchus mykiss) to acute thermal stress. PLOS ONE 20(1): e0312694 (https://doi.org/10.1371/journal.pone.0312694). The EXCEL versions of the Supplementary Files are also available directly from the PLoS ONE website and at zenodo using the following URL: https://doi.org/10.5281/zenodo.14619012
The main publication should be referred to regarding details on procedures and methods but a summary synopsis is given as follows: Fish from two different strains of rainbow trout (i.e., the Alma and the Lyndon strain) were subjected to a fairly acute increase in thermal temperature from 12C to around 27C over several time course intervals up to plateau temperature of 28C. The time when any given fish exhibited a loss of equilibrium (LOE) was recorded at which point they were removed from the experiment and euthanized. The experiments were conducted in separate years for each rainbow trout strain as the rearing and husbandry of multiple family crosses for each strain (~ 28 families/strain) precluded our ability to conduct the experiments in one year. The thermal trials were conducted under similar conditions in both years with similar densities of fish used in each replicate tank (8 replicates in total). The juvenile fish that were tested in both strains were approximately 9 months of age. The median times from when the first fish reached LOE across both strains was ~ 4.5 hours, and these were designated as Low Tolerance (L) fish. The median times of the fish with the highest LOE values across both strains was ~ 15.4 hours. These fish were designated as High Tolerance (H) fish. A single fish was also sampled from each of the experimental replicate tanks when the water temperature was 12C (prior to temperature increases) and these fish represented the Control (C) fish.
We prepared mRNA extracts (Qiagen RNeasy Mini Kits) for downstream RNAseq library preparation from small sections of whole liver extracts preserved in RNAlater® that contained blood, vascular, connective, nervous, and cytoskeletal support tissue contained within the liver as well as hepatocyte cells. The RNA preparations from the experimental fish were submitted to Genome Quebec for library preparation and paired-end (100bp) RNA sequencing using the Illumina NovaSeq 6000 Sequencing system. Illumina *.fastq file reads were returned to the University of Guelph with initial barcode trimming performed. Subsequent downstream rechecks and read preparations were performed using the Qiagen CLC Genomics Workbench pipeline (see publication for further details). Reads were aligned to an assembled rainbow trout reference library based on the NCBI OmykA_1.1 genome release (2020/09/23). This represented the genome assembly from 32 chromosomes and 190 scaffolds and had a total size greater than 2, 293 million bp, representing 57,398 genes.
Our goals were directed at discovering important genes and physiological processes that confer increased resistance to acute thermal stress in a coldwater-adapted fish species. Such research is of both basic scientific and commercial interest given that rainbow trout are a major coldwater aquaculture species worldwide. Increasingly, however, aquaculture practices have been challenged due to global warming conditions that can lead to prolonged and fluctuating temperature elevations at rearing sites.
Our approach was to use the more complete genomic knowledge of vertebrate physiological processes based on previous research focused primarily on humans, mice, and rats as a template to guide the interpretation of enhanced or repressed gene expression levels that rainbow trout may exhibit in response to thermal stress. While it is true that a fish is not a human, it is also true that many physiological processes are shared among vertebrates and therefore consideration of the interaction of genomic responses based upon the more complete understanding of human/mouse/rat responses may provide insights to understanding fish physiological responses. To address this goal we used the QIAGEN Ingenuity Pathway Analysis (IPA) knowledge base platform. This reference software contains many modules that provide information on upstream regulators, canonical pathways, and physiological functions, based on the expression levels of a set of genes that are queried by the user. It is directed towards the understanding of disease states such as cancer but also provides detailed information on all physiological processes. To implement usage of the software we first needed to convert all of the trout genes with significant expression differences into their respective human gene orthologs. This was accomplished by using two hierarchical approaches. First, the amino acid protein sequences corresponding to trout genes with significant expression differences were queried using EggNOG mapper software (http://eggnog5.embl.de) which provided information on Ensembl gene IDs (either mouse/rat/human) for best matches to the trout genes. Unassigned genes using this method were reinvestigated using the NCBI Protein Blast (i.e., BlastP) portal (https://blast.ncbi.nlm.nih.gov/Blast.cgi) to attempt to assign the trout genes to their human orthologs (see publication for details). ‘Best match’ human or mouse Ensembl accession IDs and NCBI ‘best match’ BlastP IDs that were assigned to the trout genes were automatically converted into the HuGO Gene Nomenclature Committee (HGNC) gene IDs by the IPA software when we uploaded our final human/mouse orthology protein ID lists into the IPA software suite. It should be noted that the most recent version of the EggNOG mapper software (referenced above) also automatically provides the preferred HGNC gene IDs (if designated), but when we analyzed our datasets for this study, this was not a feature of the standard output.
We were also interested in comparing gene ontology (GO) term differences between the three main treatment groups in the study (H = High tolerance fish; L = Low tolerance fish; and C = Control fish). The EggNOG mapper software provided the GO terms for many of the protein accessions we searched, and also provided PFAM motifs for other genes where GO terms were not assigned. We used these PFAM motif accessions to search the PFAM reference library and associated GODM database for additional listings of GO terms associated with trout genes (see the publication for further information).
One of the challenges in adopting the human/mouse/rat model to the rainbow trout genes is that the mammalian genome is largely represented by single copy genes, and as such the IPA software defaults to only accepting a single expression level variant for each gene. Fish, and in particular salmonid fishes which include rainbow trout, have undergone additional rounds of genome duplication in their ancestral history compared to mammals. As such, it was not uncommon to observe multiple duplicate copies of the trout genes that corresponded to single orthologous human genes. The IPA software ignored most duplicate copies of these trout genes and automatically only selected the single gene copy for that given gene that had the highest expression level. To address this issue of how duplicates could modify the findings from the IPA experimental results we devised a software program that could take into account the presence of all the duplicate gene copies and examine the effect of their cumulative counts on a given physiological function (see the publication for further details). Knowledge of how different genes ‘affect’, ‘increase’, or ‘decrease’ a given physiological function was obtained from the IPA Knowledge Base. These reference gene listings were in turn used to assign a ‘Gene-Function Match Score’ for different physiological functions based on all the genes in the dataset. The results of these analyses are detailed in Supplementary File 11.
In the following sections, a brief description of what is contained in each of the Supplementary Files is provided. The files are mostly provided as a single file *.csv format files, within a folder representing different EXCEL files that are referenced as Supplementary Material in the publication. For certain Supplementary Files such as #12, however, the EXCEL file is presented intact given that it contains gene summaries and formatting that are essential to understanding the content. Supplementary File 12 is not a data summary file, but rather a gene summary/interpretation reference to help the reader understand some of the main physiological processes discovered. The last worksheet in this file (comparison between H vs L datasets (i.e., High tolerance vs. Low tolerance fish) also provides information on significant genes detected at the FDR P < 0.10 level and is therefore also provided as a *.csv file. All other gene analyses presented in the supplementary files show significant genes detected at FDR P < 0.01 level. This is primarily due to the fact that currently, the IPA software suite can only manage data analysis of up to 8000 gene entries. This limit was exceeded in some of the group comparisons when significance thresholds were set at FDR P < 0.05 or lower.
Summary of Supplementary File Contents:
It should be noted that most of the folders contain a *-NOTES.csv file that gives a more complete description of the different worksheets in the file and what is contained in different columns and a more complete description of abbreviations that may be contained in a file or represents column headings.
General abbreviations are used in the title of files.
Most of the files present data derived from pairwise comparisons between the three experimental groups used in the study, such that the following abbreviations apply:
H = High thermal tolerance group
L = Low thermal tolerance group
C = Control group
Therefore, as an example, a file name that includes a … ‘HvC’ in the title would be a comparison between the High thermal tolerance and the Control group.
The results were obtained from two different rainbow trout strains (Alma and Lyndon) and they are expressed as Alma and Lyndon/Lyn in the title headings. However, for the Gene Ontology analyses, groupings of genes shared between the strains were also analyzed. The worksheet results from the shared gene analyses are indicated using an ‘AL’ descriptor in the title of the file to represent that the genes were common to both the Alma & Lyndon strains.
Other abbreviations for the Gene Ontology findings (Supplementary File 5) are as follows:
BP = Biological Process
CC = Cellular Component
MF = Molecular Function
Supplementary File 1: This file has 1 worksheet that provides basic statistics on the data on Illumina sequencing reads obtained from the Genome Quebec facility such as ‘Total Reads’ per sample, etc.
Supplementary File 2: This folder contains 6 worksheets and provides information on the significant genes (FDR P < 0.01) and their respective gene expression differences (Log2-fold change) in the Alma strain. Worksheets a – c detail the findings from the comparisons between the 3 pairwise experimental datasets (i.e., Low tolerance vs Control; High tolerance vs Control, and High tolerance vs Low tolerance). Worksheets d – e provide information on the IPA HGNC assignments for the 3 pairwise experimental groupings examined. These tables list the human/mouse EggNOG and BlastP protein accession orthology assignments (column C), the IPA HGNC orthology assignments (column E), and the respective rainbow trout gene ID (column F) matching these genes.
Supplementary File 3: This folder contains 6 worksheets and provides information on the significant genes (FDR P < 0.01) and their respective gene expression differences (Log2-fold change) in the *Lyndon * strain. Worksheets a – c detail the findings from the comparisons between the 3 pairwise experimental datasets (i.e., Low tolerance vs Control; High tolerance vs Control, and High tolerance vs Low tolerance). Worksheets d – e provide information on the IPA HGNC assignments for the 3 pairwise experimental groupings examined. These tables list the human/mouse EggNOG and BlastP protein accession orthology assignments (column C), the IPA HGNC orthology assignments (column E), and the respective rainbow trout gene ID (column F) matching these genes.
Supplementary File 4: This folder contains 18 worksheets plus a NOTES file that provides additional details. The data contained in the worksheets is a compilation of significant genes that are shared between the two strains as well as genes that are unique to either the Alma or Lyndon strains for each pairwise experimental grouping. These listings provide details separately for each experimental group. For example, there are 3 files that provide information on the Low tolerance genes shared, and unique to either Alma/Lyndon in the Low tolerance vs Control comparison, as well as 3 files detailing the gene findings in the Control group for this comparison. Therefore, each of the pairwise comparisons is represented by 6 different files.
NOTE: The reader should ignore the direction of the gene expression levels shown in worksheets a-c in Supplementary Files 2 and 3 and all worksheets in Supplementary File 4 as these represent the raw output values obtained from the Qiagen CLC Genomics software suite. The software arbitrarily assigns expression values according to the first gene encountered in the datasets, and therefore, these values are often reversed according to how they may be finally represented/portrayed in any given analysis. We were most interested in assessing effects in High tolerance fish, and secondarily Low tolerance fish and therefore, for all pairwise comparisons we reassigned expression levels as being positive whenever either of these two groups was involved in a pairwise comparison, except for the H vs L comparison. For example, in the H vs C comparison, High-tolerance fish were reassigned a +ve expression levels, as were Low tolerance fish in the L vs C comparison. However, in the H vs L pairwise comparison, the Low tolerance fish were assigned a –ve expression level. In the raw genes reported in the Alma dataset, all of these expression levels are reversed as to how they were finally assigned for the analysis, whereas in the raw Lyndon gene set reports the expression levels are reversed in the H vs C and L vs C comparisons. The final column in the worksheets of Supplementary Files 2 and 3 provides the correct group assignments for these genes.
Supplementary File 5: This folder contains 23 worksheets, the original EXCEL file, plus a NOTES file that provides additional details. The data relates to the findings from the pairwise Gene Ontology searches conducted (see the original publication for details). The original EXCEL file is included as it highlights significant differences in color between the two experimental groups being considered. Worksheets a – i detail the results of the Heterogeneity G-tests testing for significant differences in gene counts between the experimental groups being compared for various PIR GO-slim terms identified. The final column in these worksheets indicates which experimental group has a significantly higher count of genes corresponding to the GO term listed in the second to last column. As we were most interested in functions related to High-tolerance fish survival, we investigated the terms related to this experimental grouping. Worksheets j & k provide a listing of GO terms that are shared in H fish obtained from the H vs L and the H vs C datasets, as well as those terms that are unique to H in either dataset comparison. Worksheets l & m provide information on the Biological Process (BP) terms that are shared and unique using the H vs L dataset as a reference (worksheet l), and then similarly, using the H vs C dataset as a reference (worksheet m). Worksheets n & o provide information on the Cellular Component (CC) terms that are shared and unique using the H vs L dataset as a reference (worksheet n), and then similarly, using the H vs C dataset as a reference (worksheet o). Worksheets p & q provide information on the Molecular Function (MF) terms that are shared and unique using the H vs L dataset as a reference (worksheet p), and then similarly, using the H vs C dataset as a reference (worksheet q). Worksheet r gives an assessment of which terms appear to be either enhanced, greatly enhanced, suppressed, or greatly suppressed in High tolerance fish examined in the H vs L comparison group. To make this evaluation, the relative ranking of these terms in the H vs C was used (see NOTES for further explanation, and also the descriptions provided in Supplementary File 9). Worksheets s – w provide information on the raw data used to generate the findings shown in worksheets a – r. Worksheets s & t show the raw GO assignments to the significant rainbow trout genes obtained from EggNOG and PFAM/GODM searches, respectively. Worksheets u – w show all the gene assignments to the various PIR GO-slim categories for the three pairwise experimental group comparisons. These were the raw input files used for the Heterogeneity G-test evaluations.
Note: within some files, the headings r1 and r2 appear. These refer to the ranking positions of GO terms in the two groups being compared within each comparison group. For example, in a comparison between H vs L fish, the term ranked #1 in High tolerance fish would have the greatest proportion of count differences in the H fish for that term, and similarly, the term ranked #1 in Low tolerance fish would have the greatest proportion of count differences across all terms in Low tolerance fish. The r1 ranks would pertain to the experimental groups listed in the previous column (i.e., r1 rankings relate to the groups listed in column C and they pertain to the groups listed in column B). Similarly, r2 ranks (column E) pertain to the experimental groups referenced in column D.
Supplementary File 6: This folder contains 15 worksheets. It is divided into 3 sets of 5 files that provide information on the IPA HGNC gene assignments pertaining to the H vs C, L vs C, and H vs L pairwise experimental results. Using the findings from H vs C as an example, worksheets a & c provide information on the complete set of genes pertaining to the Alma and Lyndon strain respectively. Worksheets b & d provide information on the duplicate genes detected in the Alma and Lyndon strains, respectively, while worksheet e provides information about the duplicate genes may be mismatched (i.e., both positive and negative expression levels), and which of these genes may be mismatches in expression between the strains. The other 10 worksheets follow a similar pattern for reporting the findings from the L vs C and H vs L comparisons.
Supplementary File 7: This folder contains 13 worksheets plus a NOTES file that provides additional details. These worksheets summarize some of the main findings from the IPA analysis. Worksheets a – d represent findings from the H vs C comparison, and similarly worksheets e – h, and i – l list the findings from the L vs C and H vs L comparisons, respectively. In each set, the first two worksheets show the findings for significant and neutral canonical pathways, respectively. The third and fourth worksheets show the results for all the significant upstream regulators that are ranked by either their raw expression levels or their IPA-assigned z-score values. The final worksheet m gives a listing of all observed upstream regulators and their association expression levels (Panel set A). The IPA software suite can also identify other possible upstream regulators based on observed expression levels and gene interactions ascertained within the dataset. This set of predicted upstream regulators is provided in Panel B in the final worksheet. These predicted upstream regulators do not have significant expression level differences at the FDR P < 0.01 and are therefore shown as NA. In panel A, upstream regulators predicted to be significantly activated or inhibited (z-score ≥ 2.0 or ≤ -2.0, respectively) are indicated and if there is a mismatch between these predicated states and their observed expression levels, this is also indicated.
Supplementary File 8: Only the main EXCEL file is provided in this folder and shows the Graphical summaries produced by the IPA software that portray some of the major genes and pathways/functions detected following the analysis of each pairwise grouping. Separate graphical summaries are provided for each strain of rainbow trout. If each of these figures genes/pathways shown in orange are enhanced in the target treatment (i.e., H in the H vs C and H vs L comparison and L in the L vs C comparison), whereas those shown in blue are repressed.
Supplementary File 9: This folder contains one worksheet and a PROCEDURES description. It provides information on which IPA canonical pathways appear to be either enhanced, greatly enhanced, suppressed, or greatly suppressed in High-tolerance fish examined in the H vs L comparison group. To make this evaluation, the relative ranking of these canonical pathways in the H vs C was used (see PROCEDURES for further explanation, and also the descriptions provided for worksheet r in Supplementary File 5 which uses a similar approach for the GO terms analysis comparisons of H vs L to H vs C fish).
Supplementary File 10: This folder contains 7 worksheets plus a NOTES file that provides additional details. The information presented in these files relates to the top 10 ranked IPA ranked upstream regulators identified in both High-tolerance and Low-tolerance fish from the H vs L experimental comparison. The top-ranked regulators were ranked according to both their average expression levels (Alma & Lyndon), and average z-score ranks (Alma & Lyndon). The top regulators activated in High tolerance are interpreted to be associated with increased survival during thermal stress, whereas those regulators activated in Low tolerance fish are interpreted to suppress survival in High tolerance fish. Worksheets in this folder detail physiological functions that are associated with various top-ranked upstream regulators identified via IPA software, and provide information on whether the gene in question ‘affects’, ‘decreases’, or ‘increases’ the particular physiological function listed. Function lists are provided in both an expanded description format (expd) and in shortened description (shrt) format. The final worksheet g sorts the dataset by all the top upstream regulators in the High tolerance fish followed by those in the Low tolerance fish. The list is then sorted by the Upstream Regulator IDs followed by short descriptions of the functions.
Supplementary File 11: This folder contains 7 worksheets plus a NOTES file that provides additional details. These worksheets detail the findings from the IPA Functional Effects analyses of significant terms identified in H vs L comparison study. The functional categories/terms are listed in column A of worksheets a – d, and significant terms (z-score ≥ 2.0 or ≤ -2.0) are referenced in column D as being Increased or Decreased, respectively. These assignments are detailed in worksheets a & b for the Alma and Lyndon strains, respectively. The Gene Function Match of upstream regulators and all rainbow trout genes including those with duplicate copies were then reassessed for their predicted effects on these significant functions, and these findings are shown in worksheets c & d, respectively. For these analyses functional traits with a z-score ≥ 1.6 or ≤ -1.6 were considered. The ranked functions for high tolerance fish are from highest z-score to lowest z-score value. Similarly, for low tolerance fish the functions shown are ranked from lowest z-score value to highest. Column J which indicates the ‘Group with INCREASED function’ would represent a H assignment for all terms from High tolerance group, and an L assignment for all terms from the Low tolerance group based upon the initial IPA analysis. What is shown in column J in worksheets c & d is the superimposition of the re-assignment of these terms to treatment groups based upon using the ‘Gene_vs_Function_Match’ program (see publication for details). If the reassignment is in disagreement with the original IPA group assignment, then this is designated with an ‘OP’ symbol in the final column of worksheets c & d. The Gene_vs_Function_Match program assumes that the data is entered in a 2 x 3 Contingency table format without missing data. If cells with missing data are detected the program reverts to a sample proportions test to assess differences and results from such tests are reported in worksheet e. The gene list used for the upstream regulator analysis is shown as a gene list in the –NOTES worksheet, whereas the complete list of all genes and their duplicates from the H vs L comparison is given in worksheet f.
Supplementary File 12: This Supplementary File is presented as an EXCEL file which details the findings on genes associated with major physiological processes that are either enhanced or repressed in high-tolerance fish. The data is presented in columns C and D of all worksheets (except the last one) and is arrayed as two columns showing the number of copies of the gene indicated with significant expression in High tolerance fish (first column), followed by any genes with significant expression in Low tolerance fish (second column). Each row gives a count of the genes significant at P = 0.10, P = 0.05, and P = 0.01 level from top-to-bottom in each cell. In some gene summaries an ‘x’ will be shown beside the P = 0.10 and P = 0.05 labels. This would indicate that a duplicate copy with a higher significance level is present in that grouping. For example, if only one duplicate with a high significance level of P ≤ 0.01 is detected then that copy is also obviously significant at the P = 0.05 and P = 0.10 level as well, so these categories are not relevant (= x). If, however, there is one copy with a significance level of P ≤ 0.10 detected, and also one copy with a P ≤ 0.01 significance level then these two duplicates would be listed as 1, x, 1 (top-bottom) for that group. Column C presents the findings from the Alma strain, while column D depicts results from the Lyndon strain. Column E gives a color coding of the effect of that gene in either the High or Low-tolerance groups. The pattern of Red – pink – purple – light blue – dark blue indicates the genes are predominately expressed in a qualitative range from largely/exclusively High tolerance (red) to largely/exclusively Low tolerance (dark blue). A shading of dark purple would indicate an equal mixture of duplicates with both positive and negative expression levels across both strains. If pink or light blue, or light purple shading is portrayed it most often represents the fact that the above conditions were only observed in one of the two strains being studied. The final worksheet shows a complete listing of genes significant in both the Alma and Lyndon strains at the FDR P ≤ 0.10 level. This worksheet is also provided as a *.csv formatted data file.
Files and variables
File: Supplementary_Files-journal.pone.0312694.zip
Description: See the descriptions of the various supplementary files and their content as described above.
Code/software
A description of all the software used is provided in the referenced publication in Turner et al. (2025). PLoS ONE 20(1):e0312694.
Access information
NA
Methods
A description of the Procedures used is available at: https://www.protocols.io/private/48F2FF00AB4811EE819C0A58A9FEAC02
Synopsis as follows:
Fish from 27 families (derived from a half-sib step chain cross design) from the Lyndon strain, and similarly, fish from 28 families from the Alma strain of rainbow trout were reared at 8.5°C at the Ontario Aquaculture Research Centre (OARC). Fish were maintained at OARC until around 38 weeks post-fertilization before being transported to the Hagen Aqualab at the University of Guelph for temperature-controlled thermal trials. Fish were acclimated to 12°C for 12 days prior to the initiation of the trial, and food was withheld 72 hours prior to the start of the trial.
At around 40 weeks post-fertilization, an average of 880 fish from each strain were tested for their acute thermal tolerance in 8 replicate tanks within the experimental room in Hagen Aqualab. The acute thermal trials lasted for 18 hours where the fish was subjected to stages of thermal increase. The first stage of the trial involved a ramping of temperature from 12 to 24°C at a rate of ~3.4°C/hour in the Lyndon trial, and ~2.7°C/hour in the Alma trial. These differences were not intentional but resulted from accidental differences in water flow rates between the two trials. After the ramping interval, water was maintained at the target temperature for three hours. Following the three-hour stasis period, the temperature was increased by 1°C over a one-hour interval and maintained for a further three hours at the new level. This stepwise procedure was repeated until the temperature reached 28°C. When each fish lost equilibrium (see description in Procedures document), it was euthanized, and the time from the onset of the thermal trial to that at which it experienced loss of equilibrium (LOE) was recorded. Very few fish (< 1%) maintained equilibrium to the 27-28°C plateau. The sequential rank order of the fish’s time to LOE was used to assign a low tolerance versus high tolerance status to the rainbow trout.
Whole liver including connective, nervous, and vascular tissue was dissected immediately from low- and high-tolerance fish (N = 6 fish/group), as well as those of six control fish sampled from the 12°C environment and were stored in RNAlater (Invitrogen, Burlington, ON, Canada). mRNA was isolated using the Qiagen RNAeasy Mini Kit (QIAGEN, Toronto, ON, Canada.
An Illumina NovaSeq 6000 RNA-Seq analysis (100 bp paired-end reads) was conducted by Génome Québec (GQ; Montreal, QC, Canada), and we processed the subsequent *.fastq files provided by them using Qiagen CLC Genomics Workbench (RRID:SCR_011853). The *.fastq reads were aligned to the first release version of the current reference rainbow trout OmykA_1.1 genome build (GCF_013265732.2) available from the NCBI genome database resources (RRID:SCR_002474). Gene expression comparisons were made between high and low-tolerance rainbow trout to control fish, as well as between high and low-tolerance fish.
Greater details on the methods various analysis steps and findings from the study are provided in Turner et al. (2024). Reference to be provided upon publication.