Genomic Networks of Hybrid Sterility

<div><p>Hybrid dysfunction, a common feature of reproductive barriers between species, is often caused by negative epistasis between loci (“Dobzhansky-Muller incompatibilities”). The nature and complexity of hybrid incompatibilities remain poorly understood because identifying interacting loci that affect complex phenotypes is difficult. With subspecies in the early stages of speciation, an array of genetic tools, and detailed knowledge of reproductive biology, house mice (<i>Mus musculus</i>) provide a model system for dissecting hybrid incompatibilities. Male hybrids between <i>M. musculus</i> subspecies often show reduced fertility. Previous studies identified loci and several X chromosome-autosome interactions that contribute to sterility. To characterize the genetic basis of hybrid sterility in detail, we used a systems genetics approach, integrating mapping of gene expression traits with sterility phenotypes and QTL. We measured genome-wide testis expression in 305 male F₂s from a cross between wild-derived inbred strains of <i>M. musculus musculus</i> and <i>M. m. domesticus</i>. We identified several thousand <i>cis</i>- and <i>trans</i>-acting QTL contributing to expression variation (eQTL). Many <i>trans</i> eQTL cluster into eleven ‘hotspots,’ seven of which co-localize with QTL for sterility phenotypes identified in the cross. The number and clustering of <i>trans</i> eQTL—but not <i>cis</i> eQTL—were substantially lower when mapping was restricted to a ‘fertile’ subset of mice, providing evidence that <i>trans</i> eQTL hotspots are related to sterility. Functional annotation of transcripts with eQTL provides insights into the biological processes disrupted by sterility loci and guides prioritization of candidate genes. Using a conditional mapping approach, we identified eQTL dependent on interactions between loci, revealing a complex system of epistasis. Our results illuminate established patterns, including the role of the X chromosome in hybrid sterility. The integrated mapping approach we employed is applicable in a broad range of organisms and we advocate for widespread adoption of a network-centered approach in speciation genetics.</p>