Spermiation is the final step of spermatogenesis and culminates in the disengagement (release) of elongated spermatids from Sertoli cells into the seminiferous tubule lumen. Spermiation failure, wherein spermatids are retained by Sertoli cells instead of releasing, occurs after hormone suppression. The mechanisms involved in spermatid disengagement and retention are not well understood. We previously showed that β₁-integrin is associated with spermatids until the point of disengagement, but the ectoplasmic specialisation junction (ES) is not. The aims of this paper are to further characterise the complex that is present immediately prior to spermatid disengagement by identifying the α-integrin form dimerised with β₁-integrin, localising focal adhesion kinase (FAK) and determining if microtubules are involved. Adult Sprague–Dawley rats received testosterone and oestradiol implants and an FSH antibody for 7 days to suppress testicular testosterone and FSH and induce spermiation failure. Control rats were treated with saline. Immunohistochemical analysis showed that α₆-integrin and a phosphorylated form of FAK (FAK-Tyr³⁹⁷) are present between late spermatids and Sertoli cells after ES removal, until the point of disengagement, and both proteins remain associated with retained spermatids after spermiation failure induced by hormone suppression. Using dual-label immunofluorescence, tubulins (and thus microtubules) were observed to co-localise with ES, but were neither associated with elongated spermatids just prior to release nor with retained spermatids following hormone suppression. These results suggest that microtubules are not involved in the final release of spermatids from Sertoli cells. We conclude that spermatid release during spermiation is mediated by a ‘disengagement complex’ containing α₆β₁-integrin and phospho-FAK, the function of which can be affected by gonadotrophin suppression.

Oestrogen is a metabolite of testosterone, but its role in spermatogenesis is ill-defined. Oestrogen may exert its effects on spermatogenesis, as oestrogen receptor (ER)-β has been localised to both germ and somatic cells. This study sought to establish whether the restoration of early germ cell numbers in spermatogenesis by high-dose exogenous testosterone was influenced by its metabolite, oestrogen. The ER antagonist (ICI 182780) was administered, at a dose known to impair oestrogen action in the male reproductive tract, during testosterone treatment of gonadotrophin-releasing hormone (GnRH)-immunised rats, and germ cell numbers were determined. GnRH-immunised adult Sprague–Dawley rats (n=7–8 per group) received two doses of testosterone, either as a Silastic implant (24 cm (T24 cm)) or an injectable ester for 10 days alone or in combination with ICI 182780 (2 mg/kg, s.c. injection daily). Control rats received vehicle alone. Testes were perfusion-fixed and germ cells were quantified by the optical disector technique.

GnRH-immunisation reduced (P<0.001) both type A/ intermediate spermatogonial and type B spermatogonial/ preleptotene spermatocyte number (56% of control) and leptotene/zygotene spermatocyte number (63% of control). Pachytene spermatocyte and round spermatids were reduced to 12% and l% (P<0.01) of control respectively. Testosterone treatment did not increase type A/intermediate spermatogonial number compared with GnRH-immunised controls over the 10-day study period. Treatment with testosterone-esters increased type B spermatogonial/preleptotene spermatocytes and leptotene/zygotene spermatocyte numbers (both being ~83% of control, P<0.05), while T24 cm treatment did not significantly increase their numbers (~73% of control) compared with GnRH-immunised controls. Both treatments increased pachytene spermatocyte and round spermatid numbers to 55% and 8% of control respectively. Co-administration of ICI 182780 had no effect on any of these germ cell numbers. We conclude that oestrogen action plays no role in the short-term restoration of spermatogenesis by testosterone in the GnRH-immunised rat.

Inhibins are expressed in the adrenal cortex, but little is known of their binding or role in the adrenal. The aims of the present study were, first, to establish whether a mouse adrenocortical (AC) cell line expresses inhibins/activins and bone morphogenetic proteins (BMP), along with proteins required for inhibin to antagonise activin and BMP actions and, secondly, to characterise and compare inhibin binding sites and proteins in the rat adrenal gland and AC cells. AC cells were found to: (1) express mRNA for multiple BMPs (BMP-2, -3, -4, -6, -8a), growth/differentiation factors (GDF-1, -3, -5, -9), Lefty A and B, and the inhibin α, β_A and β_B subunits (2) secrete inhibin A and inhibin B and (3) express mRNA encoding the inhibin co-receptor, betaglycan, along with activin and BMP type I (ALK2–7) and type II (ActRII, ActRIIB, BMPRII) receptors, and binding proteins (follistatin, BAMBI, gremlin). When applied to sections of rat adrenal glands, [¹²⁵I]inhibin A specifically bound to cells of the adrenal cortex, mainly in the zona reticularis. Scatchard analyses of in vitro [¹²⁵I]inhibin A binding to dispersed rat adrenal cells and AC cells revealed sites of high affinity (K_d(1) of 0.18 and 0.15 nM, respectively) and low affinity (K_d(2) of 2.6 and 1.3 nM, respectively. Competition for [¹²⁵I]inhibin A binding by activin A or B (30 nM) was negligible, whereas BMP-2, -6 and -7 competed for between 21 and 33% of specific inhibin A binding (IC₅₀ between 0.2 and 0.3 nM). Inhibin B crossreaction with inhibin A binding sites was < 8%. Multiple binding protein complexes (molecular weight ranging from 35 to > 220 kDa) were affinity labelled by [¹²⁵I]inhibin A on both the primary rat adrenal and AC cells. The species of > 220 kDa were shown by immunoprecipitation to include betaglycan, the species of 105 kDa is consistent in size with type II receptors for activin/BMP, and that of 62 kDa co-migrates with the inhibin-follistatin complex.

In summary, the results show that inhibin A binds selectively and with both high and low affinity to AC cells via multiple binding proteins, including a single betaglycan-like species. The results support the role of glycosylated betaglycan in the high affinity binding of inhibin A, but provide consistent evidence from two independent sources of adrenal cells that inhibin A interacts with several membrane proteins in addition to those currently understood to mediate the anti-activin/BMP actions of inhibin.