Baseline mRNA expression differs widely between common laboratory strains of zebrafish

Common strains of wildtype zebrafish (Danio rerio) have unique genomic features including SNPs and CNV, but strain information often goes unreported in the literature. As a result, the confounding effects of interstrain variation makes repetition of studies in zebrafish challenging. Here we analyze hepatic mRNA expression patterns between three common zebrafish strains (AB, Tuebingen (TU), and WIK) using Agilent 4 × 44 K gene expression microarrays to establish baseline mRNA expression across strains and between sexes. We observed wide variation in sex-specific gene expression within AB and WIK strains (141 genes in AB and 67 genes in WIK), but no significant variation between sexes within TU. After partitioning the dataset into male and female subsets, we detected 421 unique mRNA transcripts with statistically significant differential expression; 269 mRNA transcripts varied between males, 212 mRNA transcripts varied between females, and 59 mRNA transcripts varied across the three strains, regardless of sex. It is not surprising that mRNA expression profiles differ between sexes and strains, but it is imperative to characterize the differences. These results highlight the complexity of variation within zebrafish and underscore the value of this model system as a valid representation of normal variation present in other species, including humans.


SupplementaryDataset1.xls
Differentially expressed mRNA transcripts between males and females in AB. File includes probe name, log fold change, average expression, p-value, Bonferroni-corrected p-value, and genomic coordinates (danRer7/Zv9) as reported by Agilent Feature Extract and annotation files. Entrez gene ID, Ensembl ID, gene symbol, and gene name were manually confirmed and harmonized. Gene symbols and names highlighted in pink are differentially expressed in females. Gene symbols and names highlighted in blue are differentially expressed in males.

SupplementaryDataset2.xls
Differentially expressed mRNA transcripts between males and females in WIK. File includes probe name, log fold change, average expression, p-value, Bonferroni-corrected p-value, and genomic coordinates (danRer7/Zv9) as reported by Agilent Feature Extract and annotation files. Entrez gene ID, Ensembl ID, gene symbol, and gene name were manually confirmed and harmonized. Gene symbols and names highlighted in pink are differentially expressed in females. Gene symbols and names highlighted in blue are differentially expressed in males.

SupplementaryDataset3.xls
Differentially expressed mRNA transcripts between sexes regardless of strain. File includes probe name, log fold change, average expression, p-value, Bonferroni-corrected p-value, and genomic coordinates (danRer7/Zv9) as reported by Agilent Feature Extract and annotation files. Entrez gene ID, Ensembl ID, gene symbol, and gene name were manually confirmed and harmonized. Gene symbols and names highlighted in pink are differentially expressed in females. Gene symbols and names highlighted in blue are differentially expressed in males. Log fold change, average expression, p-values, and Bonferronicorrected p-values are averaged between AB and WIK differential expression datasets (SupplementaryDataset1 and SupplementaryDataset2).

SupplementaryDataset4.xls
Differentially expressed mRNA transcripts between AB, TU, and WIK in males. File includes probe name, log fold change, average expression, p-value, Bonferroni-corrected p-value, and genomic coordinates (danRer7/Zv9) as reported by Agilent Feature Extract and annotation files. This file also includes calculated log fold change values for each strain individually by taking the average of the relative log fold change: (AB.TU-AB.WIK)/2 = ABcalc. Calculated strain-specific log fold change values were then centered on zero for each probe (ABcenter). Entrez gene ID, Ensembl ID, gene symbol, and gene name were manually confirmed and harmonized. Several probes annotate to deprecated gene IDs; the few that fall into this category are retained in the file, but identified by strike-through.

SupplementaryDataset5.xls
Differentially expressed mRNA transcripts between AB, TU, and WIK in females. File includes probe name, log fold change, average expression, p-value, Bonferroni-corrected p-value, and genomic coordinates (danRer7/Zv9) as reported by Agilent Feature Extract and annotation files. This file also includes calculated log fold change values for each strain individually by taking the average of the relative log fold change: (AB.TU-AB.WIK)/2 = ABcalc. Calculated strain-specific log fold change values were then centered on zero for each probe (ABcenter). Entrez gene ID, Ensembl ID, gene symbol, and gene name were manually confirmed and harmonized. Several probes annotate to deprecated gene IDs; the few that fall into this category are retained in the file, but identified by strike-through.

SupplementaryDataset6.xls
Differentially expressed mRNA transcripts between strains regardless of sex. File includes probe name, log fold change, average expression, p-value, Bonferroni-corrected p-value, and genomic coordinates (Affymetrix), and WT_liver = Mouse Wild Type Liver (GNF microarray). "Merge" tab combines all circadian-driven genes from sd1-6 tabs with duplicates removed. "Unique" tab lists the gene symbol for the 82 genes described in this dataset.