Spermatogenesis is a cell differentiation process that ensures the production of fertilizing sperm, which ultimately fuse with an egg to form a zygote. Normal spermatogenesis relies on Sertoli cells, which preserve cell junctions while providing nutrients for mitosis and meiosis of male germ cells. Several genes regulate normal spermatogenesis, some of which are not exclusively expressed in the testis and control multiple physiological processes in an organism. Loss-of-function mutations in some of these genes result in spermatogenesis and sperm functionality defects, potentially leading to the insurgence of rare genetic disorders. To identify genetic intersections between spermatogenesis and rare diseases, we screened public archives of human genetic conditions available on the Genetic and Rare Diseases Information Center (GARD), the Online Mendelian Inheritance in Man (OMIM), and the Clinical Variant (ClinVar), and after an extensive literature search, we identified 22 distinct genes associated with 21 rare genetic conditions and defective spermatogenesis or sperm function. These protein-coding genes regulate Sertoli cell development and function during spermatogenesis, checkpoint signaling pathways at meiosis, cellular organization and shape definition during spermiogenesis, sperm motility, and capacitation at fertilization. A number of these genes regulate folliculogenesis and oogenesis as well. For each gene, we review the genotype–phenotype association together with associative or causative polymorphisms in humans, and provide a description of the shared molecular mechanisms that regulate gametogenesis and fertilization obtained in transgenic animal models.

STUDY QUESTION Were Neanderthals and Denisovans (referred here also as extinct hominidae) carrying deleterious variants in genes regulating reproduction? SUMMARY ANSWER The majority of extinct hominidae analyzed here, presented a considerable number of deleterious variants per individual in proteins regulating different aspects of reproduction, including gonad and uterine function, and gametogenesis. WHAT IS KNOWN ALREADY Neanderthals, Denisovans and extant humans were interfertile and hybridized while occupying geographically overlapping areas in Europe and Asia. This is evidenced by the small archaic genome component (average ∼2%) present in non-African extant humans. STUDY DESIGN, SIZE, DURATION The genome of eight extinct hominidae, together with five human genome databases, plus 44 mothers and 48 fathers (fertile controls), were screened to look for deleterious variants in 1734 protein-coding genes regulating reproduction. PARTICIPANTS/MATERIALS, SETTING, METHODS Ancient DNA from six Neanderthals and two Denisovans dated between ∼82 000 and 43 000 calibrated years was retrieved from the public European Nucleotide Archive. The hominins analyzed include Altai, Vindija 33.15, 33.19, 33.25 and 33.26, El Sidron 1253, Denisova 3 and 11. Their DNA was analyzed using the CLC Genomics Workbench 12, by mapping overlapping paired-end reads (Illumina, FASTQ files) to the human genome assembly GRCh37 (hg19) (Vindija 33.19, 33.25, 33.26, Denisova 3 and Denisova 11) or by analyzing BAM files (Altai, El Sidron 1253 and Vindija 33.15) (human genome reference, GRCh37 (hg19)). Non-synonymous reproductive variants were classified as deleterious or tolerated (PolyPhen-2 and SIFT analyses) and were compared to deleterious variants obtained from extant human genome databases (Genome Aggregation Database (GnomAD), 1000 Genomes, the Haplotype Map (HapMap), Single Nucleotide Polymorphism Database (dbSNPs)) across different populations. A genetic intersection between extant or extinct DNA variants and other genetic disorders was evaluated by annotating the obtained variants with the Clinical Variant (ClinVar) database. MAIN RESULTS AND THE ROLE OF CHANCE Among the eight extinct hominidae analyzed, a total of 9650 non-synonymous variants (only coverage ≥20 reads included; frameshift mutations were excluded) in 1734 reproductive protein-coding genes were found, 24% of which were classified as deleterious. The majority (73%) of the deleterious alleles present in extant humans that are shared between extant humans and extinct hominidae were found to be rare (<1%) in extant human populations. A set of 8044 variants were found uniquely in extinct hominidae. At the single-gene level, no extinct individual was found to be homozygous for deleterious variants in genes necessary for gamete recognition and fusion, and no higher chance of embryo-lethality (calculated by Mendelian Genetics) was found upon simulated mating between extant human and extinct hominidae compared to extant human-extant human. However, three of the eight extinct hominidae were found to be homozygous for 48–69 deleterious variants in 55 genes controlling ovarian and uterine functions, or oogenesis (AKAP1, BUB1B, CCDC141, CDC73, DUSP6, ESR1, ESR2, PATL2, PSMC3IP, SEMA3A, WT1 and WNT4). Moreover, we report the distribution of nine Neanderthal variants in genes associated with a human fertility phenotype found in extant human populations, one of which has been associated with polycystic ovarian syndrome and primary congenital glaucoma. LIMITATIONS, REASONS FOR CAUTION While analyzing archaic DNA, stringent filtering criteria were adopted to screen for deleterious variants in Neanderthals and Denisovans, which could result in missing a number of variants. Such restraints preserve the potential for detection of additional deleterious variants in reproductive proteins in extinct hominidae. WIDER IMPLICATIONS OF THE FINDINGS This study provides a comprehensive overview of putatively deleterious variants in extant human populations and extinct individuals occurring in 1734 protein-coding genes controlling reproduction and provides the fundaments for future functional studies of extinct variants in human reproduction. STUDY FUNDING/COMPETING INTEREST(S) This study was supported by the Department of Biological Science and by the Office of Research and Sponsored Programs at the University of Tulsa (Faculty Research Grant and Faculty Research Summer Fellowship) to M.A. and the University of Tulsa, Tulsa Undergraduate Research Challenge (TURC) program to E.L.; no conflict of interest to declare. TRIAL REGISTRATION NUMBER N/A