Polymorphic SNPs, short tandem repeats and structural variants are responsible for differential gene expression across C57BL/6 and C57BL/10 substrains

Abstract

C57BL/6J is the most widely used inbred mouse strain and is the basis for the mouse reference genome. In addition to C57BL/6J, several other C57BL/6 and C57BL/10 substrains exist. Previous studies have documented extensive phenotypic and genetic differences among these substrains, which are presumed to be due to the accumulation of new mutations. These differences can be used for genome wide association studies.They can also have unintended consequences for reproducibility when substrain differences are not properly accounted for. In this paper, we performed genomic sequencing and RNA-sequencing in the hippocampus of 9 C57BL/6 and 5 C57BL/10 substrains. We identified 985,329 SNPs, 150,344 Short Tandem Repeats (STR) and 896Structural Variants (SV), out of which 330,178 SNPs and 14,367 STRs differentiated the C57BL/6 and C57BL/10 groups. We found several regions that contained dense polymorphisms. We also identified 578 differentially expressed genes for C57BL/6 substrains and 37 differentially expressed genes for C57BL/10 substrains (FDR < 0.01).We then identified nearby SNPs, STRs and SVs that matched the gene expression patterns. In so doing, we identified SVs in coding regions of Wdfy1, Ide, Fgfbp3 and Btaf1 that explain the expression patterns observed. We replicated several previously reported gene expression differences between substrains (Nnt, Gabra2) as well as many novel gene expression differences (e.g. Kcnc2). Our results illustrate the impact of new mutations on gene expression among these substrains and provides a resource for future mapping studies.

Copy number evolution in simple and complex tandem repeats across the C57BL/6 and C57BL/10 inbred mouse lines

“…These two lines also did not have higher rates of SNPs or small indels than other lines (Mortazavi et al 2020). However, C57BL/6NHsd had the highest number of large-scale deletions found by CNVnator, and C57BL/6NJ had the highest number of duplications called by Lumpy (Mortazavi et al 2020), concordant with the direction of gains and losses of kmers that we found. Figure 4.…”

“…Relevant scripts used to analyze the data are available here: https://github.com/jmf422/BL6_mouse_satellites Mouse sequencing Data used are from (Mortazavi et al 2020). In brief, DNA from 1 male per strain was extracted from spleens.…”

“…First, we manually reproduced the phylogeny in .tre format. For all relationships within B6 and within B10, we used the historical records of known divergence times (Mortazavi et al . 2020).…”

“…First, we manually reproduced the phylogeny in .tre format. For all relationships within B6 and within B10, we used the historical records of known divergence times (Mortazavi et al 2020). Since the history of the lines is recorded by the year of divergence, we used 4.33 generations per year to estimate the number of generations of each branch (Jackson Laboratories https://www.research.uci.edu/forms/docs/iacuc/JAX-breeding-strategies.pdf).…”

“…Almost all cases of phylogenetic signal showed grouping/separation in copy number between the two major clades C57BL/6 and C57BL/10. This may be driven by the detected introgression from a divergent strain of about 5% of the genome likely into C57BL/10 (Mortazavi et al 2020). Of the 7 kmers with a strong phylogenetic signal, 3 represented losses in C57BL/10 and gains in C57BL/6, and 3 represented gains in C57BL/10 and losses in C57BL/6, and one was a gain specific to the C57B/6J clade.…”