Functional modelling of human infertility-associated genetic variants using C. elegans and mice
Infertility affects one in six couples worldwide and has a significant genetic component. Genome sequencing efforts have identified increasing numbers of potentially pathogenic variants, including in conserved meiotic genes. However, the lack of functional validation remains a major barrier to establishing variant pathogenicity and clinical diagnoses. Here, we establish an experimental pipeline integrating C. elegans and mouse in vivo models with in silico protein structure predictions to functionally validate potentially pathogenic infertility-associated variants. Using CRISPR-Cas9 we generated worm and mouse models of a homozygous missense variant (P638L) in MSH4, a mismatch repair protein essential for meiotic recombination, identified in a patient with non-obstructive azoospermia (NOA). Homozygous msh-4P546L worms exhibited severe meiotic defects including an absence of chiasmata, accumulation of DNA damage, and loss of crossover markers. Comparable defects were observed in spermatocytes of homozygous Msh4P660L/P660L mice, while Msh4P660L/P660L ovaries lacked oocyte-containing follicles, demonstrating that this variant causes infertility in both sexes. Together, these findings support a model in which the MSH4 P638L substitution disrupts the conversion of early recombination events into crossovers, leading to NOA in men and primary ovarian insufficiency in women. Further worm modelling of human MSH4 genetic variants identified exclusively in heterozygosity (G649R) or also in homozygosity (Y589C) demonstrates the potential of worm models to predict the pathogenicity of missense variants in meiotic proteins found in the human population. We propose that this integrated worm-mouse pipeline provides a powerful framework for accelerating the genetic diagnosis of infertility and informing the clinical management of affected individuals.