DDX3Y (DBY), located within AZoospermia Factor a (AZFa) region of the human Y chromosome (Yq11), encodes a conserved DEAD-box RNA helicase expressed only in germ cells and with a putative function at G1–S phase of the cell cycle. Deletion of AZFa results most often in germ cell aplasia, i.e. Sertoli-cell-only syndrome. To investigate the function of DDX3Y during human spermatogenesis, we examined its expression during development and maturation of the testis and in several types of testicular germ cell tumours (TGCTs), including the pre-invasive carcinoma in situ (CIS) precursor cells which are believed to originate from fetal gonocytes. DDX3Y protein expression was analysed during development in different tissues by western blotting. The localization of DDX3Y in normal fetal and prepubertal testis tissue of different ages as well as in a series of distinct TGCT tissue samples (CIS, classical seminoma, spermatocytic seminoma, teratoma and embryonal carcinoma) was performed by immunohistochemistry. Germ cell-specific expression of DDX3Y protein was revealed in fetal prospermatogonia but not in gonocytes and not before the 17th gestational week. After birth, DDX3Y was expressed at first only in the nuclei of Ap spermatogonia, then also in the cytoplasm similarly to that seen after puberty. In CIS cells, DDX3Y was highly expressed and located predominantly in the nuclei. In invasive TGCT, significant DDX3Y expression was found in seminomas of the classical and spermatocytic type, but not in somatically differentiated non-seminomas, consistent with its germ-cell specific function. The fetal germ cell DDX3Y expression suggests a role in early spermatogonial proliferation and implies that, in men with AZFa deletion, germ cell depletion may begin prenatally. The strong expression of DDX3Y in CIS cells, but not in gonocytes, indicates phenotypic plasticity of CIS cells and suggests partial maturation to spermatogonia, likely due to their postpubertal microenvironment.

Genomic AZFb deletions in Yq11 coined “classical” (i.e. length of Y DNA deletion: 6.23 Mb) are associated with meiotic arrest (MA) of patient spermatogenesis, i.e., absence of any postmeiotic germ cells. These AZFb deletions are caused by non-allelic homologous recombination (NAHR) events between identical sequence blocks located in the proximal arm of the P5 palindrome and within P1.2, a 92 kb long sequence block located in the P1 palindrome structure of AZFc in Yq11. This large genomic Y region includes deletion of 6 protein encoding Y genes, EIFA1Y, HSFY, PRY, RBMY1, RPS4Y, SMCY. Additionally, one copy of CDY2 and XKRY located in the proximal P5 palindrome and one copy of BPY1, two copies of DAZ located in the P2 palindrome, and one copy of CDY1 located proximal to P1.2 are included within this AZFb microdeletion. It overlaps thus distally along 2.3 Mb with the proximal part of the genomic AZFc deletion. However, AZFb deletions have been also reported with distinct break sites in the proximal and/or distal AZFb breakpoint intervals on the Y chromosome of infertile men. These so called “non-classical” AZFb deletions are associated with variable testicular pathologies, including meiotic arrest, cryptozoospermia, severe oligozoospermia, or oligoasthenoteratozoospermia (OAT syndrome), respectively. This raised the question whether there are any specific length(s) of the AZFb deletion interval along Yq11 required to cause meiotic arrest of the patient’s spermatogenesis, respectively, whether there is any single AZFb Y gene deletion also able to cause this “classical” AZFb testicular pathology? Review of the literature and more cases with “classical” and “non-classical” AZFb deletions analysed in our lab since the last 20 years suggests that the composition of the genomic Y sequence in AZFb is variable in men with distinct Y haplogroups especially in the distal AZFb region overlapping with the proximal AZFc deletion interval and that its extension can be “polymorphic” in the P3 palindrome. That means this AZFb subinterval can be rearranged or deleted also on the Y chromosome of fertile men. Any AZFb deletion observed in infertile men with azoospermia should therefore be confirmed as “de novo” mutation event, i.e., not present on the Y chromosome of the patient’s father or fertile brother before it is considered as causative agent for man’s infertility. Moreover, its molecular length in Yq11 should be comparable to that of the “classical” AZFb deletion, before meiotic arrest is prognosed as the patient’s testicular pathology.

DDX3X, the functional X homologue of the major AZFa gene, DDX3Y, belongs to the highly conserved PL10-subfamily of DEAD-box RNA helicase genes which are functionally conserved from yeast to man. They are mainly involved in cell cycle control and translation initiation control of gene transcripts with long 5′UTR extensions containing complex secondary structures. Interestingly, in humans both gene copies were found to be expressed at different phases of human spermatogenesis. Whereas DDX3Y transcripts are translated only in premeiotic male germ cells, the DDX3X protein is expressed only in postmeiotic spermatids. In this study, we found that the major class of DDX3X transcripts in human testis become activated first after meiosis and at a specific core promoter not active in somatic tissues and not present upstream of the DDX3Y homologue. Two alternative 5′UTR transcript lengths are subsequently produced by an additional testis-specific 5′UTR splicing event. Both transcripts are mainly processed for polyadenylation in their proximal 3′UTR. A minor transcript class starting at the same male germ line-specific core promoter produces primary transcripts with an extremely long 3′UTR (∼17 kb), which is subsequently spliced at distinct sites resulting in six short 3′UTR splice variants (I–VI). Comparative analyses of the DDX3X transcripts in mouse and primates revealed that this complex pattern of male germ line-specific transcript variants first evolved in primates. Our data thus suggest complex translational control mechanism(s) for the human DDX3X gene locus functioning only in the male germ line and resulting in expression of its protein only in the postmeiotic spermatids.

We aimed to determine whether a functional link with impact on female ovarian reserve exists between FMR1 expression and expression ratios of AKT/mTOR signaling genes in human granulosa cells in vivo, as suggested from prior in vitro data. Three hundred and nine women, who were classified as normal (NOR; n = 225) and poor (POR; n = 84) responders based on their ovarian reserve, were recruited during stimulation for assisted reproductive techniques. Expressions of FMR1 and of key genes of the AKT/mTOR and AKT/FOXO1/3 signaling pathways were comparatively analyzed in their granulosa cells. FMR1 expression in granulosa cells of NOR and POR correlated significantly with AKT1, TSC2, mTOR, and S6K expression. No correlation was found between FMR1 and FOXO1 in all, and FOXO3 expression in POR, patients. AKT1 expression was significantly higher and FOXO1 expression lower in POR samples, whereas AKT1 expression was lower and FOXO1 expression was higher in NOR samples. In human native granulosa cells, FMR1 expression significantly correlated with the expression of key genes of the AKT/mTOR signaling pathway, but not with the FOXO1/3 signaling pathway. Our data point to a functional link between FMR1 expression and expression of the AKT/mTOR signaling pathway genes controlling human follicular maturation.

Translational control is a major level of gene expression regulation in the male germ line. DDX3Y located in the AZFa region of the human Y chromosome encodes a conserved RNA helicase important for translational control at the G1-S phase of the cell cycle. In human, DDX3Y protein is expressed only in premeiotic male germ cells. In primates, DDX3Y evolved a second promoter producing novel testis-specific transcripts. Here, we show primate species-specific use of alternative polyadenylation (APA) sites for these testis-specific DDX3Y transcript variants. They have evolved subsequently in the 3´UTRs of the primates´ DDX3Y transcripts. Whereas a distal APA site (PAS4) is still used for polyadenylation of most DDX3Y testis transcripts in Callithrix jacchus; two proximal APAs (PAS1; PAS2) are used predominantly in Macaca mulatta, in Pan trogloydates and in human. This shift corresponds with a significant increase of DDX3Y protein expression in the macaque testis tissue. In chimpanzee and human, shift to predominant use of the most proximal APA site (PAS1) is associated with translation of these DDX3Y transcripts in only premeiotic male germ cells. We therefore assume evolution of a positive selection process for functional DDX3Y testis transcripts in these primates which increase their stability and translation efficiency to promote its cell cycle balancing function in the human male germ line.