ABSTRACT: Meiosis is a specialized form of cell division essential for sexual reproduction. The unique cellular events of meiosis have been extensively studied in model eukaryotes, including mammals. In mammals, the molecular and cellular changes that germ cells must undergo to transition from mitotic spermatogonia to meiotic spermatocytes remain largely undefined. Retinoic acid (RA) has been proposed as the meiosis inducing substance (MIS) required for female and male meiotic initiation. However, evidence for this presumption is based on studies of fetal gonads – meiosis is initiated in fetal ovaries but suppressed in fetal testes. Critically, the requirement for RA in meiosis has never been formally tested because germ cell differentiation and meiotic initiation in the fetal ovary are temporally inseparable. Using the postnatal mouse testis, where differentiation and meiotic initiation are temporally distinct (8.6 days apart), we test the hypothesis that RA is dispensable for initiation and progression through meiosis. Complementary in vivo experimental models were designed to address the hypothesis. In the first, “RA-sufficient” (RA-suf) model, spermatogenesis was synchronized by administration of potent and selective RA synthesis inhibitor; as a result, testes contained only undifferentiated spermatogonia. Then, the inhibitor was discontinued, and mice were given a single dose of exogenous RA to initiate spermatogonial differentiation. Germ cells completed differentiation, entered and completed meiosis, and formed haploid gametes at predicted timepoints. In the second, “RA-deficient” (RA-def) model, mice were treated as above, except the inhibitor was administered continuously. The lack of testicular RA in this model was confirmed in multiple ways, including assessment of RA-responsive gene expression, in vitro serum-free cultures, and dose-response experiments. To our surprise, germ cells in RA-def testes completed differentiation, underwent meiotic DNA replication, completed the landmark cytological processes of meiosis, and formed postmeiotic haploid spermatids, all on the predicted timeline of the RA-suf model. To define the effect of RA-deficiency on gene expression during meiotic initiation, large, highly enriched populations of premeiotic spermatogonia from RA-suf mice, and meiotic preleptotene spermatocytes from both models were FACS-isolated and subjected to bulk RNA-seq. Interestingly, analyses revealed few significant differences in expression of characteristic meiotic genes at the mRNA level. However, genes that were significantly different between RA-suf and RA-def preleptotene spermatocytes were associated with spermiogenesis functions. Taken together, our data reveal that, after initiating spermatogonial differentiation, RA is dispensable for initiation, progression through, and completion of meiosis in the male germline.
