Inbred or Outbred? Genetic Diversity in Laboratory Rodent Colonies

Abstract

Nonmodel rodents are widely used as subjects for both basic and applied biological research, but the genetic diversity of the study individuals is rarely quantified. University-housed colonies tend to be small and subject to founder effects and genetic drift; so they may be highly inbred or show substantial genetic divergence from other colonies, even those derived from the same source. Disregard for the levels of genetic diversity in an animal colony may result in a failure to replicate results if a different colony is used to repeat an experiment, as different colonies may have fixed alternative variants. Here we use high throughput sequencing to demonstrate genetic divergence in three isolated colonies of Mongolian gerbil (Meriones unguiculatus) even though they were all established recently from the same source. We also show that genetic diversity in allegedly “outbred” colonies of nonmodel rodents (gerbils, hamsters, house mice, deer mice, and rats) varies considerably from nearly no segregating diversity to very high levels of polymorphism. We conclude that genetic divergence in isolated colonies may play an important role in the “replication crisis.” In a more positive light, divergent rodent colonies represent an opportunity to leverage genetically distinct individuals in genetic crossing experiments. In sum, awareness of the genetic diversity of an animal colony is paramount as it allows researchers to properly replicate experiments and also to capitalize on other genetically distinct individuals to explore the genetic basis of a trait.

A high-density genetic map and molecular sex-typing assay for gerbils

“…Our analysis is a first step towards tying genomic scaffolds to chromosomes, and will hopefully pave the way for future work aimed at both improving the assembly of the Mongolian gerbil genome, and facilitating cross-species comparisons, particularly for the reconstruction of patterns of chromosomal evolution In order to measure recombination frequency and build the map it is easiest to use markers that have variants alternatively homozygous in the parents. To maximise the number of these markers available for mapping we used a cross between the two most divergent strains of gerbil locally available: Edinburgh and Sheffield (Brekke et al 2018).…”

“…All experiments were reviewed and approved by the Bangor University Animal Welfare and Ethical Review Board. A female from the Edinburgh strain and a male from the genetically distinct Sheffield strain (characterised in Brekke et al (2018)) were used as the parents of an F2 mapping panel. Four male and four female F1 individuals were crossed in pairs and produced 137 F2 offspring.…”

“…DNA was shipped to LGC Genomics (Queens Road, Teddington, TW11 0LY) for genotyping with a Genotyping-by-Sequencing (GBS) approach (Elshire et al 2011) using the restriction enzyme MslI (recognition sequence: CAYNN^NNRTG). Overall genetic diversity is low between the Bangor and Edinburgh strains (Brekke et al 2018) and so to identify sufficient variants we sequenced the 150bp paired-end libraries three times with two different size-selection regimes on three lanes of Illumina NextSeq 500 v2. The first two sequencing efforts were size-selected for reads 200-300 bases long while the final was selected for 300-380 bases.…”

“…We ran 'populations' without the population map but included the flags -p 1, --min_maf 0.01, --write_random_snp, and --vcf. Much segregating genetic diversity is shared between the two parental gerbil strains (Brekke et al 2018) and so we identified 3,751 SNPs where the parents were homozygous for alternative alleles to use for building the genetic map. We also identified 7,063 SNPs that were heterozygous in one parent but not the other as some of these can be placed on the genetic map depending on which variant was inherited by each F1.…”

“…We aligned the raw GBS reads to the reference stacks (Supplemental File 3) with bwa (Li and Durbin 2009) and calculated the depth of coverage for each individual with samtools ). By parsing the depth of coverage by sex we identified sex-linked markers as described in (Brekke et al 2018) based on the logic that markers on the Y have coverage in males but not females, scaffolds on the X have twice the coverage in females as in males, and autosomal markers have approximately equal coverage in males and females. To calculate the standardised average coverage for males and females, we divided all read counts for each individual by the sequencing effort of that individual, multiplied by 1,000,000, and then took the mean of all males and all females.…”