Localization of Activin bA-, bB-, and bC-Subunits in Human Prostate and Evidence for Formation of New Activin Heterodimers of bC-Subunit*

Activin ligands are formed by dimerization of activin bAand/or bB-subunits to produce activins A, AB, or B. These ligands are members of the transforming growth factor-b superfamily and act as growth and differentiation factors in many cells and tissues. New additions to this family include activin bC-, bD-, and bE-subunits. The aim of this investigation was to examine the localization of and dimerization among activin subunits; the results demonstrate that activin bC can form dimers with activin bA and bB in vitro, but not with the inhibin a-subunit. Using a specific antibody, activin bC protein was localized to human liver and prostate and colocalized with bAand bB-subunits to specific cell types in benign and malignant prostate tissues. Activin C did not alter DNA synthesis of the prostate tumor cell line, LNCaP, or the liver tumor cell line, HepG2, in vitro when added alone or with activin A. Therefore, the capacity to form novel activin heterodimers (but not inhibin C) resides in the human liver and prostate. Activin A, AB, and B have diverse actions in many tissues, including liver and prostate, but there is no known biological activity for activin C. Thus, the evidence of formation of activin AC or BC heterodimers may have significant implications in the regulation of levels and/or biological activity of other activins in these tissues. (J Clin Endocrinol Metab 85: 4851–4858, 2000) A ARE members of the transforming growth factor-b (TGFb) superfamily of proteins (1), which consist of homoand heterodimers of activin bAor bBsubunits that combine to form activin A (bA-bA), activin AB (bA-bB), and activin B (bB-bB) (2–4). Activins were originally identified in ovarian follicular fluid as regulators of FSH secretion from the pituitary gland. However, activins are regulators of growth and differentiation in a range of tissues and cells, where they act in autocrine or paracrine pathways. Activins (mainly activin A) have been implicated in diverse processes, including mesoderm induction (5), mammalian embryogenesis (6), erythropoiesis (7), nerve cell survival (8), release of insulin (9), and wound repair (10). Newer members of the activin family were identified, and bCand bE-subunits were cloned from mouse (11, 12) and human liver (13). The activin bD-subunit has been cloned from Xenopus, but no mammalian equivalent has yet been described (14). In contrast to activins formed from bAand bB-subunits, those from bC, bD, and bE appear to have limited actions in liver and mesoderm induction (14–16) and are thought to be a subset of related sequences (17). In the liver it was suggested that activin bC was a putative chalone (15, 16) due to the temporal expression of bAand bC-subunit messenger ribonucleic acids (mRNAs) after partial rat hepatectomy, but there is no evidence for biological activity of activin bC in a dimeric or monomeric form in the liver. It is possible that heterodimers of activin bC and bA or bB are produced, and if so, this may affect the level of production of activin A, B, or AB ligands. Furthermore, the putative heterodimers bA-bC (activin AC) and bB-bC (activin BC) may be inactive or have biological activity different from that of activins A, B, or AB. The idea that heterodimers of bC and bA or bB exist would require at least the colocalization of subunit proteins to the same cell/tissue type, but coexpression and localization of subunit proteins have not been possible to date, as no specific antibodies are available, nor are there any data to show that activin bC is able to dimerize with bAor bB-subunits. As we have identified activin bC-subunit mRNA in human prostate (18) and human prostate tumor cell lines LNCaP, DU145, and PC3 (19), the first aim of this investigation was to clone activin bC from the DUI45 human prostate tumor cell line. To determine whether the bC-subunit could form homodimers or heterodimers with activin bAor bBor inhibin a-subunits, we cotransfected the corresponding cDNAs into human kidney 293 cell lines (20). To examine the expression of bC-subunit protein in human prostate and liver, we deReceived December 8, 1999. Revision received June 16, 2000. Accepted September 6, 2000. Address all correspondence and requests for reprints to: Dr. Gail P. Risbridger, Monash Institute of Reproduction and Development, 27–31 Wright Street, Clayton 3168, Victoria, Australia. E-mail: gail.risbridger@ med.monash.edu.au. * This work was supported by a Program Grant from the National Health and Medical Research Council of Australia and by the Medical Research Council and the European Commission [Contract BMH4CT98–9574 (DG12-SSMI); to N.P.G.). 0021-972X/00/$03.00/0 Vol. 85, No. 12 The Journal of Clinical Endocrinology & Metabolism Printed in U.S.A. Copyright © 2000 by The Endocrine Society