DNA hydroxymethylation differences underlie phenotypic divergence of somatic growth in Nile tilapia reared in common garden

Abstract

Phenotypic plasticity of metabolism and growth are essential for adaptation to new environmental conditions, such as those experienced during domestication. Epigenetic regulation plays a key role in this process but the underlying mechanisms are poorly understood, especially in the case of hydroxymethylation. Using reduced representation 5-hydroxymethylcytosine profiling, we compared the liver hydroxymethylomes in full-sib Nile tilapia with distinct growth rates (3.8-fold difference) and demonstrated that DNA hydroxymethylation is strongly associated with phenotypic divergence of somatic growth during the early stages of domestication. The 2677 differentially hydroxymethylated cytosines between fast- and slow-growing fish were enriched within gene bodies (79%), indicating a pertinent role in transcriptional regulation. Moreover, they were found in genes involved in biological processes related to skeletal system and muscle structure development, and there was a positive association between somatic growth and 5hmC levels in genes coding for growth factors, kinases and receptors linked to myogenesis. Single nucleotide polymorphism analysis revealed no genetic differentiation between fast- and slow-growing fish. In addition to unveiling a new link between DNA hydroxymethylation and epigenetic regulation of growth in fish during the initial stages of domestication, this study suggests that epimarkers may be applied in selective breeding programmes for superior phenotypes.

Keywords: DNA hydroxymethylation; Epigenetics; domestication; phenotypic plasticity; somatic growth; teleosts.

Graphical abstract

Figure 1.

Barplots displaying the average weight (A) and length (B) of Nile tilapia females between the large (fast-growing, dark blue) and small phenotypes (slow-growing, green). Asterisks reflect significant differences for both weight and length between fish with distinct growth rates (one-tailed t-test, p-value <0.0001). Data are represented as means ± standard deviations (n = 5).

Figure 2.

Venn diagram showing the percentages of unique and common 5hmC sites. Hydroxymethylated cytosines identified exclusively within the large (i.e., fast-growing) (34%) and small (i.e., slow-growing) (4%) fish are represented in dark blue and light green, respectively, while common hydroxymethylated sites (62%) are shown in dark green. The 5hmC enrichment was obtained from 291,847 CCGG sites with non-zero values among all samples (n = 5).

Figure 3.

Circular representation of the Nile tilapia nuclear genome with each arc corresponding to a linkage group (LG1-LG23). Blue (oriented out) and green (oriented in) peaks represent hyper-DhmCs in the fast- and slow-growing groups, respectively. The enrichment of single DhmCs is based on the average filtered counts (n = 5) and is reflected on the height of each peak.

Figure 4.

Barplot displaying the comparison of 5hmC enrichment per annotated feature (introns, exons, promoters, transcription termination sites -TTS and intergenic regions) between potential RRHP targets (whole genome CCGG sites; dark purple), RRHP raw 5hmC sites (purple), RRHP filtered sites (light purple) and RRHP differentially hydroxymethylated cytosines between fast- and slow-growing phenotypes (very light purple).

Figure 5.

Visual representation of conserved regions of pax7 among four fish species. Highly conserved, intermediately conserved and non-conserved regions of the gene are represented by red, green and white colour, respectively. Numbered black boxes represent the exons of pax7 in Nile tilapia while each dot-plot corresponds to the pairwise alignments (50–100% similarity) between Nile tilapia and the fish species that is designated on the left side. The zoom-in below shows the exact position of the DhmC among fast- and slow-growing Nile tilapia as well as the myogenin TFBS sequence that is highly conserved among these four species.