Recent progress in our understanding of inhibin in the prostate gland.

Inhibin was originally described as the active principle or substance in the testis which was able to prevent the appearance of castration cells in the male rat pituitary gland (McCullagh 1932). It was not until the early 1970s that interest in the identity of the putative feedback regulator of follicle-stimulating hormone (FSH) was renewed. Using castrate male rats, Franchimont (1982) showed that the subcutaneous injection of seminal plasma from normal men or patients with oligospermia significantly inhibited serum FSH; seminal plasma from men with azoospermia did not. The identity of the ‘inhibin’ in seminal plasma was a subject of considerable debate over the following years. Seminal plasma was used as a source for the purification of inhibin by Seth and his colleagues (1984) who isolated ‘prostatic inhibin peptide’ (PIP), a 14 kDa, 98 amino acid non-glycosylated protein. Ramasharma et al. (1984) also identified an ‘inhibin-like peptide’ from seminal plasma which was a 31 amino acid peptide of 5 kDa which suppressed FSH in vitro. Neither of these proteins showed a consistent inhibition of FSH and they are unrelated to dimeric inhibin which was finally isolated from follicular fluid by Robertson et al. in 1986. In the testis, the Sertoli cells are the principal source of inhibin. Dimeric inhibin is composed of an á subunit and one of two â subunits linked by disulphide bonds to form inhibin A or inhibin B. The inhibin â subunits can also form homoor heterodimers: activin A, activin B and activin AB (Burger 1988). In contrast to inhibins, activins stimulate rather than inhibit pituitary FSH (Mason et al. 1985). Following these original discoveries, it has become clear that activins are growth and differentiation factors with roles in a number of cells and tissues, and are members of the transforming growth factor â (TGFâ) superfamily. In terms of cell growth, activins regulate cell proliferation and apoptosis in a number of cells and tissues (Hedger et al. 1989, Mather et al. 1990, Nishihara et al. 1993, Kaipia et al. 1994). Prostate cell growth in normal and disease states is controlled by the balance between the regulation of cell death and proliferation, and this led us to re-examine the presence of inhibin-like proteins in the prostate gland. In 1997, studies from this laboratory showed that the human prostate gland had the capacity to synthesise dimeric inhibins and activins. Using samples collected from men with benign prostatic hyperplasia (BPH), Thomas et al. (1997) reported expression of mRNA for the inhibin á and â subunits; thus the capacity to produce all the known inhibins and activins resides in the human prostate. Interestingly, mRNA for the putative âC subunit which was originally cloned from human liver and believed to be specifically expressed in liver (Hotten et al. 1995), was also detected in the prostate. If the activin-like bioactivity of homoor heterodimers of the âC subunit is confirmed in future studies, then it is possible that novel activins are present in the human prostate. In addition to the á and â subunits, the human prostate expressed activin type II receptor which implies that activins have a local role in the prostate itself. The role of these proteins in the function of the non-malignant prostate is not known, but in vitro actions of activins on human prostate cancer cells were described. Three groups reported that activin A inhibited cell proliferation and/or induced apoptosis in the human prostate tumour cell line LNCaP (Dalkin et al. 1996, Wang et al. 1996, McPherson et al. 1997). This was rather surprising, because the growth inhibitory actions of activin A and B on the LNCaP tumour cells contrasts with the malignant characteristics of these cells; this, therefore, raised some questions. Are the actions of activins observed using LNCaP cells also recorded with other prostate tumour cell lines? Does the expression and localisation of the activin â, or the á inhibin, subunits differ in the non-malignant compared with malignant prostatic epithelium? The growth inhibitory action of activin A on LNCaP cells is not reproduced in the androgen-independent PC3 cell line which is resistant to activin A; an intermediate response to activin A occurs with DU145 cells. Therefore, in vitro, there is a differential response to exogenously added activin A by the three cell lines LNCaP, DU145 and PC3 (McPherson et al. 1997). Interestingly, all these cell lines expressed activin âA and âB subunit mRNA (Batres et al. 1995, Furst et al. 1995, Ying et al. 1995), although no dimeric activin immunoreactivity was detected in the media from these tumour cells in culture, and it is not certain that prostate tumour cells produce detectable levels of endogenous activin ligands (McPherson et al. 1997). What was known about the 1