Investigating the function of RBMXL2 in spermatogenesis in humans and using mouse models

Abstract

RBMX, RBMXL2, and RBMY belong to a family of RNA-binding proteins. How important these proteins are in whole animal biology, what their global RNA targets and mechanisms of action are, and to what extent these proteins overlap in function has been poorly understood. Using histology I found that Rbmxl2 knockout phenotypes differed between mouse strains (C57BL/6 and Sv/129). Both strains, however, were infertile, with no sperm found, which indicates the importance of RBMXL2 in spermatogenesis. I performed an iCLIP analysis to investigate RBMXL2-RNA binding, which revealed that RBMXL2 directly binds to exons and introns of protein-coding genes. The strongest RBMXL2 binding was observed in genes associated with spermatogenesis and reproduction, specifically meiosis, such as Meioc and Esco1. Binding was also enriched within ultra-long exons to prevent the selection of cryptic splice sites. I attempted minigene experiments to try and model RBMXL2 function in transfected cells, but these experiments did not replicate RBMXL2 function in vivo. As an alternative I developed a stable cell line “rescue” approach to over-express RBMXL2 and test if this could replace endogenous RBMX. Using this approach, I demonstrated that RBMXL2 is likely to replace RBMX's function in meiosis when the X and Y are inactive. RBMXL2's ability to restore splicing control is dependent on its disordered domain not the RRM. RBMY, which is distantly related to RBMX, also replaces RBMX function in somatic cells. These experiments show that RBMXL2 is crucial in male fertility, but there are differences in how different mouse strains respond to deletion of the Rbmxl2 gene. RBMXL2 binds to RNAs from genes important in spermatogenesis. Despite the divergence between RBMX and RBMXL2 and RBMY these proteins can still substitute for each other suggesting they fulfil the same functions in their respective cell types.