Genomic analysis of a novel spontaneous albino C57BL/6N mouse strain.

We report an albino C57BL/6N mouse strain carrying a spontaneous mutation in the tyrosinase gene (C57BL/6N-Tyr(cWTSI)). Deep whole genome sequencing of founder mice revealed very little divergence from C57BL/6NJ and C57BL/6N (Taconic). This coisogenic strain will be of great utility for the International Mouse Phenotyping Consortium (IMPC), which uses the EUCOMM/KOMP targeted C57BL/6N ES cell resource, and other investigators wishing to work on a defined C57BL/6N background.

Members of the International Mouse Phenotyping Consortium (IMPC) (Brown and Moore, 2012) use mutant embryonic stem (ES) cells from the NIH’s Knockout Mouse Project (KOMP), the European Conditional Mouse Mutagenesis (EUCOMM) program, Regeneron, and The Canadian NorComm programme (Bradley ; Skarnes ) to generate mouse strains for high-throughput phenotyping, and for distribution to the research community. The majority of the ES cell resources used by the IMPC are made in JM8-derived cell lines, which were originally derived from C57BL/6N embryos (Pettitt ). To achieve the goal of producing and phenotyping mutant strains on an inbred genetic background, chimeras are bred to C57BL/6N (black) females. When an albino host blastocyst is used for chimera production this prohibits the assessment of germline transmission (GLT) using the coat color of G1 mice, and therefore all progeny of chimera matings must be biopsied and genotyped, with consequent ethical and cost/labor concerns. Although agouti or non-agouti host blastocysts can be used to aid subsequent color selection in certain combinations, their utility in high-throughput production is limited as they must be paired correctly with the agouti status of the injected ESC clone (Pettitt ).

During routine colony expansion a strain generated by the Sanger Mouse Genetics Project (MGP) (Colony ID: MDDH, EUCOMM ES cell clone EPD0176_3_A10 which carries the Stard7 allele) produced albino mice from the first intercross of germline animals. A new colony of these mice was established (Colony ID: MWTH) from offspring selected to be wildtype for the Stard7 allele so the causal mutation could be identified.

RESULTS

PCR analysis of the tyrosinase gene (the most likely candidate locus) failed to amplify exons 1 and 2 from albino mouse DNA. Further analysis of a 24 kb PCR tiling path using PCRTiler (Gervais ) revealed a 14.3 Kb deletion 940 bp 5′ of exon 1, including the minimal promoter region (Klüppel ), extending 3.75 kb 3′ of exon 2; in addition, a 2 bp TA insertion was found at the breakpoint (Fig 1). We refer to this new allele as Tyr and thus our albino strain is C57BL/6N-Tyr (Fig 2). Analysis of DNA from the ES cell clone EPD0176_3_A10 revealed that the Tyr mutation originated in this line, possibly during tissue culture. The mutation was not present in a random selection of 96 other EUCOMM/KOMP ES cell clones which suggests it arose late during the isolation of the Stard7 targeted ES cell line, and is not found in the parental JM8N4 clone.

FIG 1

a: The TyrcWTSI deletion. The deletion spans from 940 bp 5′ to exon 1 from 3570 bp 3′ of exon 2 and completely removes exons 1 and 2. b: The sequence breakpoint of the TyrcWTSI deletion.

FIG 2

A C57BL/6N-TyrcWTSI mouse (left) compared to a black C57BL/6N mouse (right). Image taken at 4 weeks of age.

To further characterize the C57BL/6N-Tyr line, we generated deep (∼45×) whole genome sequence of two albino founders; the genome of these mice being composed of C57BL/6N-Taconic and C57BL/6N ES cell-derived DNA. This analysis revealed only 549 homozygous private single nucleotide variants (SNV) in both sequenced mice that were not present in the reference C57BL/6J genome (Church ; Waterston ), C57BL/6NJ, or in the genomes of 17 other laboratory mouse strains (Keane ; Wong ). Six hundred and forty six homozygous private indels, and 21 private structural variants (Table 1, Supporting Information Table 1) were also detected. We attempted to validate all 58 homozygous SNVs and 39 homozygous indel calls on Chr 7, which are linked to the Tyr deletion, using the Sequenom MassARRAY® platform. A total of 41/43 SNVs and 7/7 homozygous indels were validated, with the remainder of sites failing in the oligo design phase due to the repetitive nature of the flanking sequence (Supporting Information Table 2). An estimated false-discovery rate of 1.3% for SNVs was derived by the validation of a further 155/157 homozygous SNVs randomly selected from the rest of the genome. Some heterozygous variants were also detected (Table 1), however, the high number of false positive SNV calls (51/131) and indel calls (11/17) indicate that the majority of the heterozygous calls are likely due to artefacts resulting from read mapping to the reference genome. To assess the origin of the variants present on chromosome 7 we genotyped DNA from the EPD0176_3_A10 ES cell clone, the parental JM8N4 line, and C57BL/6N-Taconic, and compared these genotypes to calls made in C57BL/6N-Tyr (Supporting Information Table 2). This analysis revealed 31 variants (SNVs and indels) on chromosome 7 of C57BL/6N-Tyr that are shared with both the EPD0176_3_A10 ES cell clone and the JM8N4 parental cell line, but not C57BL/6N-Taconic. Two of the SNVs originating from the JM8N4 parental cell line were predicted to cause amino acid changes (K/M at AA 107 of Sult2a5 and G/E at AA 455 of Zfp74). A comparison of C57BL/6N-Taconic and C57BL/6N-Tyr revealed that only 6 SNVs and one indel on chromosome 7 originate from C57BL/6N-Taconic. We were able to validate 8 SNVs which are unique to C57BL/6N-Tyr. In addition to the Tyr deletion, only one other validated structural variant unique to C57BL/6N-Tyr was localized to chromosome 7; a 233bp homozygous insertion that fell within an intron of Leng8. Details of the variants detected by the initial genome sequencing and subsequent validation can be found in Supporting Information Data S1-S3, and Supporting Information Table 3.

Table 1

Comparison of the Genomes of the C57BL/6N-Tyr Founders and 18 Laboratory Mouse Strains

	Total number of C57BL/6N-TyrcWTSI calls against the reference genomea	Total number of C57BL/6N-TyrcWTSI calls shared with C57BL6/NJb	Total number of C57BL/6N-TyrcWTSI calls not found in the mouse genomes strains or FVB/NJ
SNVs Hom	10,897	10,261	549
Indels Hom	16,200	14,073	646
SNVs Het	36,497	31,623	3,965
Indels Het	64,323	31,132	6,309

Includes only sites where SNV or indel positions were found in both C57BL/6N-Tyr sequenced founder mice.

Includes SNPs or indels in C57BL6/NJ from the mouse genomes project release9 and calls from C57BL6/NJ resequenced with 100bp Illumina paired end reads; indels were compared within a 25bp window.

Because variants not linked to chromosome 7 can be easily segregated away, further rounds of backcrossing to C57BL/6N-Taconic were performed prior to cryopreservation of the C57BL/6N-Tyr strain. To validate the C57BL/6N-Tyr line, Tyr animals were test bred to chimeras from a new EUCOMM strain yielding black mice carrying the targeted ES derived allele while all albino offspring were wildtype (data not shown). Thus using the C57BL/6N-Tyr strain germline transmission can be scored using coat color with only black chimera progeny requiring biopsying and genotyping.

DISCUSSION

Although a large proportion of EUCOMM/KOMP clones carry the reconstructed agouti locus (Pettitt ) which can be used to select for potential GLT in certain conditions, the Sanger MGP uses albino C57BL/6-Tyrc-Brd host blastocysts for microinjection (Liu ) which reduces its potential usefulness (this is also true for BALB/c blastocysts). Because of the dominant effect of agouti, the resulting breeding to nonagouti wildtype mice means that all mice may still have to be genotyped in order to detect GLT as a small percentage of nonagouti mice may also carry the targeted mutation. To test this we analyzed 181 Sanger MGP colonies produced from JM8 agouti cell lines; 25 colonies (13%) produced only black mice at G1 and would therefore have been missed using only agouti color selection. In addition, from a total of 1,351 G1 het mice detected by PCR, 712 (52%) had a black coat color; thus limiting the detection to just agouti mice may also have a significant effect on colony expansion rates in a high-throughput project where a rapid turnaround of GLT to phenotyping is at a premium.

Following backcrossing the C57BL/6N-Tyr strain is almost identical to C57BL/6N-Taconic carrying just two variants that fall into coding sequence, 25 other single nucleotide variants on chromosome 7, and just one SV other than the Tyr deletion. This strain will be of great utility for the International Mouse Phenotyping Consortium, which aims to generate and phenotype knockouts for all mouse protein-coding genes over the next decade. The spontaneous nature of the Tyr mutation has advantages in that no foreign vector or loxP sequences are retained within the genome. The new colony is freely available to the research community.

METHODS

Mouse Production

The care and use of all mice in this study were in accordance with the UK Home Office regulations, UK Animals (Scientific Procedures) Act of 1986 and were approved by the Wellcome Trust Sanger Institute Ethical Review Committee. Embryonic stem cell (ESC) clone EPD0176_3_A10 from the EUCOMM resource was used to produce Stard7 mutant mice through standard blastocyst injection and chimera breeding techniques (Nagy ). C57BL/6-Tyrc-Brd albino host blastocysts were used due to the ease of harvesting sufficient numbers and potential GLT of C57BL/6 over other strains (Schuster-Gossler ). Chimeras were bred to C57BL/6NTac females (Taconic) and offspring were genotyped by a qPCR assay that counts the number of neomycin phosphotransferase cassettes. NeoF GGTGGAGAGGCTATTCGGC; NeoR GAACACGGCGGCATCAG; NeoM1 TGGGCACAACAGACAATCGGCTG FAM. Cycling conditions are: x1 95°C 20 sec; x35 95°C 10 sec; 60°C 30 sec. G1 heterozygotes were backcrossed to C57BL/6NTac USA (Taconic) for one generation before heterozygous offspring were inter-crossed to obtain homozygous Stard7 mice. Two albino mice (male and female) arising from this inter-cross and not carrying the Stard7 allele were used to provide genomic DNA for the full genome sequencing experiments.

Tiling Path Construction

A tiling path of 26 amplicons covering ∼24 kb with minimal overlap was designed using PCRTiler software. Reactions were performed on a selection of albino and wildtype control mice. Amplification conditions were 94°C 5 min, followed by 35 cycles of 94°C 30 sec, 58°C 30 sec, 72°C 1 min 30 sec, with a final extension of 72°C 5 min.

Sequence Analysis

Illumina reads generated from whole genome sequencing were aligned to the mouse reference genome NCBIM37 with BWA version 0.5.9-r16 (Li and Durbin, 2009) and realignment around known indels was performed with the SAMtools calmd function [version 0.1.18-r572; (Li )]. SNPs and indel discovery was performed with the SAMtools mpileup function and calling was performed with the BCFtools view function (Li, 2011). The vcf-annotate function in VCFtools package was used to filter the SNP and indel calls. To predict the functional consequences of SNP and indels we used the Variant Effect Predictor from Ensembl (McLaren ), and queried against Ensembl release 66 gene models. We compared the SNPs and indels from the albino mouse genomes to calls from the C57BL6/NJ genome, which was also sequenced with 100 bp read pairs, and to the SNPs and indels from the Mouse Genomes Project (Keane ) and the FVB/NJ genome (Wong ). Structural variants were called as described previously (Keane ), and compared to those identified in C57BL/6NJ, the Mouse Genomes Project, and FVB/NJ. Genotyping of SNPs and indels was performed using the Sequenom Mass iPLEX Gold Assay (Gabriel ).

Sequence data is available from the European Nucleotide Archive (ENA) under accession number ERP001554.

Genotyping Protocol for C57BL/6N-TyrcWTSI

The following primers were used to detect the Tyr deletion in heterozygotes. Tyr_MGP_F: GCTTCTTCATCCTGCTGGTC, Tyr_MGP_R: AAGCAGAGCAGGCTGACATT. Amplification conditions are 94°C 5 min, followed by 35 cycles of 94°C 30 sec, 58°C 30 sec, 72°C 45 sec, with a final extension of 72°C 5 min. A 169 bp product is observed in mutated animals.