Abstract-Text
For many years the laboratory mouse has remained the quintessential research animal of choice for studying molecular mechanisms of human development disorders. The recent development of CRISPR/Cas based genome editing tools now allow the investigation of any gene or regulatory element in vivo. However, the current phenotyping approaches lack the necessary throughput and resolution for detailed investigations of pleiotropic disorders at the organismal scale. The recent developments in single-cell genomics offer the possibility to overcome these shortfalls and answer central questions of development.
In the current study we set out to establish single cell RNA sequencing as a tool for large scale standardized and comprehensive phenotypic analysis of whole mouse mutant embryos. In a single multiplexed experiment, we applied combinatorial indexing based single cell RNA sequencing to profile 103 whole mouse embryos of 22 different mutants and 4 different wildtype strains at embryonic stage E13.5. Towards evaluating the sensitivity of this technique, the selected mouse mutants in this study range from established multisystem disorders to single enhancer knockouts resulting in different phenotype severities. The resulting Mouse Mutant Cell Atlas (MMCA) consists of over 1.9 million single cell RNA-seq profiles. We developed an analytical framework for molecular phenotyping of cell type and trajectory composition changes, gene expression alterations and developmental phenotypes. Moreover, we identify mutation and strain specific cell type changes, compare phenotyping of gain and loss of function mutants, and characterize deletions of topological associating domain boundaries. Overall, we identified 300 significantly changed cell type proportions from 52 sub trajectories across the 22 mutants compared to the wildtype. Some pleiotropic genes such as Sox9 showed general changes in over 30 sub-trajectories indicating their generally regulatory functions during embryogenesis, while other mutants such as Ttc21b showed very specific changes in single sub-trajectories such as the retina. The deletions of noncoding elements showed milder changes compared to the other mutants, however some specific trajectories still revealed major phenotypes in the limb and the brain. Some trajectories also showed a retardation of development. A unique strength of the experimental framework of the approach allowed combining multiple mutants which enabled the discovery of a shared phenotype between unrelated genotypes, undetected before.
In summary our findings show that whole embryo single cell phenotyping represents a powerful tool to systematically characterize developmental disorders at unprecedented resolution.