Do spermatogonia, including spermatogonial stem cells (SSCs), undergo metabolic changes during prepubertal development? Here, we show that the metabolic phenotype of prepubertal human spermatogonia is distinct from that of adult spermatogonia and that SSC development is characterized by distinct metabolic transitions from oxidative phosphorylation (OXPHOS) to anaerobic metabolism. Maintenance of both mouse and human adult SSCs relies on glycolysis, while embryonic SSC precursors, primordial germ cells (PGCs), exhibit an elevated dependence on OXPHOS. Neonatal porcine SSC precursors reportedly initiate a transition to an adult SSC metabolic phenotype at 2 months of development. However, when and if such a metabolic transition occurs in humans is ambiguous. To address our research questions: (i) we performed a meta-analysis of publicly available and newly generated (current study) single-cell RNA sequencing (scRNA-Seq) datasets in order to establish a roadmap of SSC metabolic development from embryonic stages (embryonic week 6) to adulthood in humans (25 years of age) with a total of ten groups; (ii) in parallel, we analyzed single-cell RNA sequencing datasets of isolated pup (n = 3) and adult (n = 2) murine spermatogonia to determine whether a similar metabolic switch occurs; and (iii) we characterized the mechanisms that regulate these metabolic transitions during SSC maturation by conducting quantitative proteomic analysis using two different ages of prepubertal pig spermatogonia as a model, each with four independently collected cell populations. Single testicular cells collected from 1-year, 2-year and 7-year-old human males and sorted spermatogonia isolated from 6- to 8-day (n = 3) and 4-month (n = 2) old mice were subjected to scRNA-Seq. The human sequences were individually processed and then merged with the publicly available datasets for a meta-analysis using Seurat V4 package. We then performed a pairwise differential gene expression analysis between groups of age, followed by pathways enrichment analysis using gene set enrichment analysis (cutoff of false discovery rate < 0.05). The sequences from mice were subjected to a similar workflow as described for humans. Early (1-week-old) and late (8-week-old) prepubertal pig spermatogonia were analyzed to reveal underlying cellular mechanisms of the metabolic shift using immunohistochemistry, western blot, qRT-PCR, quantitative proteomics, and culture experiments. Human PGCs and prepubertal human spermatogonia show an enrichment of OXPHOS-associated genes, which is downregulated at the onset of puberty (P < 0.0001). Furthermore, we demonstrate that similar metabolic changes between pup and adult spermatogonia are detectable in the mouse (P < 0.0001). In humans, the metabolic transition at puberty is also preceded by a drastic change in SSC shape at 11 years of age (P < 0.0001). Using a pig model, we reveal that this metabolic shift could be regulated by an insulin growth factor-1 dependent signaling pathway via mammalian target of rapamycin and proteasome inhibition. New single-cell RNA sequencing datasets obtained from this study are freely available through NCBI GEO with accession number GSE196819. Human prepubertal tissue samples are scarce, which led to the investigation of a low number of samples per age. Gene enrichment analysis gives only an indication about the functional state of the cells. Due to limited numbers of prepubertal human spermatogonia, porcine spermatogonia were used for further proteomic and in vitro analyses. We show that prepubertal human spermatogonia exhibit high OXHPOS and switch to an adult-like metabolism only after 11 years of age. Prepubescent cancer survivors often suffer from infertility in adulthood. SSC transplantation could provide a powerful tool for the treatment of infertility; however, it requires high cell numbers. This work provides key insight into the dynamic metabolic requirements of human SSCs across development that would be critical in establishing ex vivo systems to support expansion and sustained function of SSCs toward clinical use. This work was funded by the NIH/NICHD R01 HD091068 and NIH/ORIP R01 OD016575 to I.D. K.E.O. was supported by R01 HD100197. S.K.M. was supported by T32 HD087194 and F31 HD101323. The authors declare no conflict of interest.
Read full abstract