Progress towards completing the mutant mouse null resource

Abstract

The generation of a comprehensive catalog of null alleles covering all protein-coding genes is the goal of the International Mouse Phenotyping Consortium. Over the past 20 years, significant progress has been made towards achieving this goal through the combined efforts of many large-scale programs that built an embryonic stem cell resource to generate knockout mice and more recently employed CRISPR/Cas9-based mutagenesis to delete critical regions predicted to result in frameshift mutations, thus, ablating gene function. The IMPC initiative builds on prior and ongoing work by individual research groups creating gene knockouts in the mouse. Here, we analyze the collective efforts focusing on the combined null allele resource resulting from strains developed by the research community and large-scale production programs. Based upon this pooled analysis, we examine the remaining fraction of protein-coding genes focusing on clearly defined mouse-human orthologs as the highest priority for completing the mutant mouse null resource. In summary, we find that there are less than 3400 mouse-human orthologs remaining in the genome without a targeted null allele that can be further prioritized to achieve our overall goal of the complete functional annotation of the protein-coding portion of a mammalian genome.

Fig. 1

Timeline highlighting major milestones enabling complete functional annotation of the protein-coding fraction of the mouse genome. Technological advances that made possible the large-scale generation of mutant mouse resources are shown in black. Multi-institutional collaborative programs (e.g., European Mouse Disease Clinic; EUMODIC, and Knockout Mouse Phenotyping Program; KOMP²) implemented these advancements to perform high-throughput animal model production and systematic broad-based phenotyping. This collective work has grown to include additional international sites that have been centralized under the International Mouse Phenotyping Consortium (IMPC) to coordinate animal production, phenotyping, and data dissemination

Fig. 2

Summary of mouse null allele production and progress towards completing a resource encompassing mouse–human orthologs. a Overview of animal production using either ES-cell based technology or CRISPR/Cas9 within independent research groups representing the Community effort or as part of the International Mouse Phenotyping Consortium. Dates for unique alleles generated by the Community were determined using earliest citation for allele transmission and dates for IMPC alleles were based upon record of successful confirmation of germline line transmission (GLT). b Histogram for null allele counts per gene shows that most genes have 1–3 unique null alleles while another ~ 2000 have 5 or more null alleles. c Contributions of different technology employed by the Community and IMPC to generate the null allele. d Venn diagram comparing all mouse genes with a null allele to the total number of mouse genes with a high-confidence human ortholog

Fig. 3

Analysis of gene family representation and constraint for the 3381 mouse–human orthologs that currently lack a null allele in mouse. a The set of remaining genes is specifically enriched for large gene families and the RIKEN cDNA collections. Olfactory receptors and zinc finger proteins are the most highly represented gene families. b Classification of the remaining genes using human orthologs to assess cell essentiality based upon CRISPR/Cas9 screens in cancer cell lines and functional constraint using probability of loss-of-intolerance (pLI) scores that range from 0 to 1 with values closer to 1 associated with higher level of constraint

Fig. 4

Potential human disease relevance for genes without a mouse null allele. a Human orthologs were used to query the Orphanet database (https://www.orpha.net/) and the Centers for Mendelian Genomics (CMG) gene lists (http://mendelian.org/phenotypes-genes). CMG genes are classified into Tier 1 and Tier 2 based upon the supporting level of evidence. Tier 1 genes have the highest-level of confidence with multiple levels of supporting evidence and Tier 2 genes are strong candidates but do not meet stringency criteria set for Tier 1. Currently, ~ 10% of the remaining mouse genes are related to human disease genes. b IMPC progress towards making knockouts for genes relevant to human disease. Of these 358 genes, 252 genes have either been previously attempted and failed (Inactive/aborted) or are currently assigned or in progress through the IMPC production and phenotyping pipeline

Fig. 5

GO term enrichment and pathway analysis on the set of 3381 genes classified into cell essential and nonessential categories. GO terms for a Biological process, b Cellular component, and c Molecular function highlight a role for essential genes in core biological processes including transcription, splicing, and translation, and a role for nonessential genes in metabolic processes and mitochondrial function. d KEGG pathway analysis supports the distinction between cell essential and nonessential genes and associated biological function. Enrichment values were determined using WebGestalt with FDR < 0.5

Fig. 6

Top 10 noncoding RNA sequence features annotated in the mouse genome. The number of known long noncoding RNA’s (lncRNAs) currently exceeds the number of protein-coding genes and the vast majority remain to be studied in depth