Molecular and Functional Characterization of Conserved and Divergent Features of Mammalian Spermatogenesis

Abstract

Spermatogenesis is a complex differentiation process which is coordinated spatiotemporally both within and along seminiferous tubules. The cellular heterogeneity of tubule cross-sections has made it challenging to obtain a stage-specific molecular resolution of germ cell development, resulting in reliance on subsets of cell populations that can be isolated via cell surface markers or transgenic lines for further study. These limitations are further compounded by our inability to translate findings from model organisms to humans. As a result, despite the high prevalence of infertility, our understanding of underlying causes and tools for intervention remain limited. To advance both, an improved foundation of knowledge of processes involved in gamete formation is crucial.

To address these challenges, we applied single cell RNA-sequencing from mouse, human, and macaque testes to map a continuous developmental trajectory from spermatogonia to spermatids, defining molecular signatures of testis cell-subtypes without a priori biases. This characterized cellular heterogeneity at unprecedented resolution for both known and previously unknown cell types. By then defining molecularly similar cell types across species, we produced a high-resolution three-species atlas with aligned somatic and germ cell types. This interspecies comparison reveals a similar framework in the genetic programs employed for major biologic processes. In particular, we identified and functionally characterized a new zinc-finger protein, ZCWPW1, that is essential for meiosis, acts at PRDM9-specified recombination hotspots, and may also have recombination-independent functions.

However, we also observed areas of divergence throughout spermatogenesis including (1) shifts in the timing of molecular events and transcription modules, (2) differences in RNA turnover during spermatid differentiation, (3) increased reliance on species-specific genes in meiotic entry and spermiogenesis, and (4) population differences in spermatogonia. We further functionally validated an additional undifferentiated spermatogonia state enriched in primates which is molecularly distinct from previously defined stem cell populations in rodents, suggesting that primates have a unique stem/progenitor pool.

Germ cell differentiation is not solely autonomous, as it also requires extrinsic signals from nearby somatic cells. A greater understanding of somatic cell types and states is important for better defining the molecular pathways important for germ cell differentiation and possibly causes of idiopathic infertility. Our single cell RNA-sequencing identified previously unappreciated diversity in the soma, including unexpected populations. In mice, we identified and characterized an adult reserve somatic progenitor population, which maintains tissue homeostasis throughout aging and aids in regeneration in response to injury. Examination of testis somatic cells across species identified comparable populations in primates, however, there are some molecular and programmatic differences between primates and rodents. These differences include global changes in transcriptome and ligand-receptor pairs between germ and somatic cells. Although parallel pathways are likely employed to regulate similar stages of spermatogenesis, some interactions may have alternative source (secreting) cells, target (receptor) cells, or both. This may provide an explanation for failure of current human spermatogonial stem cell culture methods and cross-species transplantation.

Overall, my studies leverage single-cell RNA-sequencing technologies to characterize molecular and functional heterogeneity among and between mammalian testes cell types, enabling the discovery of new features of spermatogenesis and testis function. Furthermore, these analyses provide a reference for interspecies comparisons of spermatogenesis, identifying molecularly equivalent states between human, macaque, and mouse. Thus, despite the differences in spermatogenesis between these species, mechanistic parallels can be drawn to enable future studies to utilize these complementary models.