No abstract

ZIP9 is a Zn transporter, testosterone receptor, and mediator of signaling events through G-proteins. Despite these pivotal properties, however, its physiological and pathophysiological significance has not yet been comprehensively addressed. Using a cell line that lacks the classical androgen receptor we show that ZIP9-mediated phosphorylation of Erk1/2, CREB, or ATF-1 and expression of claudin-5 and zonula occludens-1 by testosterone can be completely antagonized by bicalutamide (Casodex), an anti-androgen of significant clinical impact. Computational modeling and docking experiments with ZIP9 reveal typical characteristics of ZIP transporters and an extracellular binding site for testosterone capable of accommodating bicalutamide. The presence of this site is verified by our demonstration that the membrane-impermeable testosterone analogue T-BSA-FITC labels the membrane only when ZIP9 is expressed and that this labeling is completely prevented by bicalutamide. The study connects structural features of ZIP9 to its functions and indicates a possible relevance of ZIP9 as a pharmacological target.

Testosterone is known to mediate its effects by two different mechanisms of action. In the so-called "classical" pathway testosterone binds to cytosolic androgen receptors (AR), which essentially function as ligand-activated transcription factors. Once activated, these receptors bind to DNA and activate the expression of target genes. In the "non-classical" pathway, the steroid hormone binds to receptors associated with the plasma membrane and induces signaling cascades mediated through activation of Erk1/2. The precise nature of the membrane-associated AR, however, remains controversial. Although some assume that the membrane and cytosolic AR are identical, others propose that the AR of the membrane is a G-protein-coupled receptor (GPCR). To evaluate these two possibilities we first searched for testosterone-induced signaling cascades in the spermatogenic cell line GC-2. Testosterone was found to cause phosphorylation (activation) of Erk1/2, CREB, and ATF-1, consistent with its non-classical mechanism of action. Silencing of AR expression by means of siRNA did not influence testosterone-induced activation of Erk1/2, CREB, or ATF-1, indicating that this pathway is not activated by the classical cytosolic/nuclear AR. In contrast, when the expression of the G-protein Gnα11 is suppressed, the activation of these signaling molecules is abolished, suggesting that these responses are elicited through a membrane-bound GPCR. The results presented here and the identification of the testosterone-specific GPCR in future investigations will help to reveal and characterize new testosterone-mediated mechanisms associated not only with fertility and reproduction but perhaps also with other physiological processes.

LNCaP cells are derived from a metastatic lesion of human prostate adenocarcinoma. They express the classical androgen receptor (AR) and ZIP9, a Zn transporter that also binds testosterone and mediates signaling by interacting with G-proteins. Our results show that LNCaP cells respond to testosterone by mobilizing their migratory machinery. Their exposure to testosterone triggers the formation of lamellipodia, reorganization of the actin cytoskeleton, phosphorylation of focal adhesion kinase (FAK) at Tyr925 and of paxillin at Tyr118, expression of matrix metalloproteinase 2 (MMP-2), and cell migration. Silencing ZIP9 expression by means of siRNA does not affect the responsiveness of the classical AR to testosterone; however, it prevents all of the testosterone effects described above: formation of lamellipodia cannot be induced, stimulation of FAK or paxillin phosphorylation or MMP-2 expression is prevented, and cell migration does not take place in the absence of ZIP9. The data presented show that testosterone/ZIP9 interactions might have not only physiological but also pathophysiological relevance. The fact that the migratory machinery of a metastatic prostate cancer cell line is activated exclusively through testosterone/ZIP9 and not through testosterone/AR interactions suggests that targeting specific inhibition of testosterone/ZIP9-mediated events might help in developing new therapeutic strategies against androgen-induced progression of prostate cancer.

A gonadotropin-releasing hormone agonist (GnRH-A) implant induces hormonal castration in dogs that is associated with reduced prostate and testes size. We address the molecular events associated with hormonal castration by examining GnRH-A effects on expression and phosphorylation of a number of key signaling proteins. Male beagles were treated for 5 months with a GnRH-A implant, and then surgically castrated at 0, 3, 6, 12, and, 24 weeks after implant removal; untreated animals served as controls. GnRH-A treatment led to activation of c-Raf, Erk1/2, and, p53 in the testes. Phosphorylation of p53 occurred at Ser15, consistent with activation of the c-Raf-Erk1/2-p53 signaling cascade that triggers growth arrest or apoptosis. GnRH-A also suppressed the anti-apoptotic protein Bcl-xL; reduced phosphorylation of the transcription factors CREB and ATF1; and down-regulated expression of StAR and P450scc, proteins involved in steroidogenesis. Although androgen receptor expression was little affected by GnRH-A treatment, levels of ZIP9, a membrane-bound Zn transporter that mediates non-classical signaling of testosterone, were abrogated. All of these effects were reversed within 24 weeks after implant removal. Thus, molecular signatures of implant-dependent hormonal castration include reversible cell cycle arrest and apoptosis, loss of steroidogenesis, and reduced transcriptional activity. Mol. Reprod. Dev. 83: 1092-1101, 2016. © 2016 Wiley Periodicals, Inc.