High-throughput genotyping of high-homology mutant mouse strains by next-generation sequencing

Abstract

Genotyping of knockout alleles in mice is commonly performed by end-point PCR or gene-specific/universal cassette qPCR. Both have advantages and limitations in terms of assay design and interpretation of results. As an alternative method for high-throughput genotyping, we investigated next generation sequencing (NGS) of PCR amplicons, with a focus on CRISPR-mediated exon deletions where antibiotic selection markers are not present. By multiplexing the wild type and mutant-specific PCR reactions, the genotype can be called by the relative sequence counts of each product. The system is highly scalable and can be applied to a variety of different allele types, including those produced by the International Mouse Phenotyping Consortium and associated projects. One potential challenge with any assay design is locating unique areas of the genome, especially when working with gene families or regions of high homology. These can result in misleading or ambiguous genotypes for either qPCR or end-point assays. Here, we show that genotyping by NGS can negate these issues by simple, automated filtering of undesired sequences. Analysis and genotype calls can also be fully automated, using FASTQ or FASTA input files and an in-house Perl script and SQL database.

Keywords: CRISPR; Genotyping; Mouse; Mutant; NGS; QC.

Fig. 1

Common knockout designs. End-point PCR assays for EUCOMM/KOMP-CSD mutants are designed to amplify either the wild type or mutant alleles and fail if the allele is not present. Exon deletions follow this same premise, although the mutant assay may amplify a larger size if the exon is sufficiently small. For SNP detection, a single assay spans the region of interest.

Fig. 2

NGS genotyping strategy for CRISPR-derived exon deletions. A) In PCR 1, tailed primers containing a linker sequence are amplified in a WT and mutant-specific multiplex reaction. A second round of PCR is then performed using additional tailed primers containing flow cell adaptors and an 8 bp barcode index sequence at each end. B) By using plate-specific and well-specific barcodes, many plates can be multiplexed into one sequencing run. After the sequencing by synthesis is complete, the MiSeq then deconvolutes the clusters and assigns the results to FASTQ files per sample, based on the barcode sequence.

Fig. 3

Data analysis pipeline. Mouse ID and strain information is retrieved from the MySQL database by processing the FASTQ filename. Sequences are compared to a database of genotype-specific motifs, any matches reported, and the genotype called on the ratio. FASTQ files can be pre-filtered on the quality score if required.

Fig. 4

Genotyping results per mouse strain using next generation sequencing. Ratios are calculated by comparing the number of wild type motif sequence hits against mutant motif hits. The shaded area represents the ideal range for heterozygous genotypes.