Immunofluorescence and immunogold electron microscopy were employed to examine the assembly of connexins (Cx) 33, 37, and 43 into testis cell gap junctions in mature and postnatal rats. Cx37 was localized by immunofluorescence to the endothelia of blood vessels in both mature and immature testes and was not further characterized. Only Cx43 assembled into Leydig cell gap junctions, but Cx43 also co-assembled with Cx33 in some Sertoli-Sertoli gap junction plaques within and near Sertoli occluding junctions and on adluminal surfaces Assembly of Sertoli gap junctions appeared to be regulated according to the stage of the seminiferous epithelium since Cx33 (and Cx43) immunoreactivity was strong in Sertoli cells from stages II-VII but weak in stages IX-XIV. During postnatal maturation, assembly of Cx33 into gap junctions was regulated independently of Cx43 assembly. Cx43 was present on Sertoli cells of all tubules from postnatal Day 5 through Day 28. In contrast, Cx33 was not apparent on Sertoli cell surfaces until Day 15 and gradually accumulated in all tubules through Day 28. Between postnatal Days 38 and 43, the immunoreactivities of Cx33 and Cx43 became weak in Sertoli cells containing step 9-14 elongated spermatids. Thus, connexin abundance and gap junction composition in Sertoli cells is regulated during testis maturation and the cycle of the seminiferous epithelium.

Dye coupling experiments were performed to determine whether the gap junctions connecting Sertoli cells with other Sertoli cells and different germ cell stages in rats showed functional variations. Chop loading of adult rat seminiferous tubules was conducted using fluorescent dextran controls and a variety of low-molecular-weight tracers (lucifer yellow, biotin-X-cadaverine, biotin cadaverine, and neurobiotin) to evaluate dye coupling in situ, and scrape loading was used to study dye coupling in Sertoli-germ cell cocultures established using prepuberal rats. Sertoli-Sertoli coupling is relatively short range and nonselective in situ, whereas coupling between Sertoli cells and chains of spermatogonia is strongly selective for the positively charged biotin tracers relative to negatively charged lucifer yellow. Coupling between Sertoli cells and spermatogonia was also asymmetric; lucifer yellow in germ cells never diffused into Sertoli cells, and biotinylated tracers only weakly diffused from spermatogonia to Sertoli cells. Asymmetric coupling would facilitate the concentration in germ cells of molecules diffusing through junctions from Sertoli cells. Dye coupling between Sertoli cells and adluminal germ cells was too weak to detect by fluorescence microscopy, suggesting that the junctional communication between these cells may be functionally different from that between Sertoli and basal germ cells. The results show that there are multiple routes of gap junction communication in rat seminiferous tubules that differ in permeability properties and show alternative gating states. Functional diversity of gap junctions may permit regulated communication among the many interacting Sertoli cells and germ cell stages in the seminiferous epithelium.

Immunocytochemical data demonstrate that the distribution of gap junction connexin43 (Cx43) in rodent testes is dependent on cell type, testis maturation, and stage of the mature seminiferous epithelium. Western blotting and indirect immunofluorescence microscopy using anti-peptide antisera to Cx43 revealed abundant Cx43 in rat and mouse testes and mouse TM3 and TM4 cells. Cx43 mRNA was detected in rat testes and mouse TM4 cells by Northern blot analysis. Cx43 was localized by immunogold electron microscopy to gap junctions on Sertoli cells and Leydig cells. A punctate distribution of Cx43 was observed on peritubular cell surfaces following indirect immunofluorescence of detergent-permeabilized tubule segments. In cryosections from testes of immature (to 30 days) rats, and mature rats and mice, Leydig cells showed a punctate surface distribution of Cx43 following indirect immunofluorescence. A diffuse cytoplasmic fluorescence was also seen in spermatocytes and spermatogonia. Cx43 staining associated with Sertoli cells was age- and stage-dependent. Over 90% of the tubules in immature tests (22-30 days) contained Cx43 in the region of Sertoli-Sertoli occluding junctions and in the adluminal compartment. In mature rat testes, however, Cx43 immunostaining was detected in only 60% of 1195 tubule sections where it was abundant proximal to the Sertoli cell occluding junctions. All strongly stained tubules were from stages I-VIII, while negatively stained tubules were at stages IX-XIV. Cx43 immunostaining in mature mouse testes was also stage-dependent with all positive tubules at stages VI-VIII. In contrast to Cx43, Cx26 and Cx32 were detected by immunofluorescence only in the apical regions of the seminiferous epithelia in 90% of tubules from mature rats. Consistent with the observed Cx43 immunostaining, octanol-sensitive in situ dye-coupling was observed between Leydig cells, between peritubular cells and between Sertoli cells, suggesting the occurrence of functional gap junctions in these cell types. These observations provide evidence for extensive gap junction-mediated communication between a variety of testis cell types important to the support of spermatogenesis.