br Duality of interest br Contribution statement br Acknowle

Duality of interest
Contribution statement
Acknowledgments
Introduction Rapid, cell surface-initiated steroid actions have been reported for all major groups of steroid hormones and our understanding of membrane-mediated steroid actions has progressed rapidly over the last two decades. Membrane progestin and estrogen receptors were first cloned and characterized in 2003 and 2005, respectively (Revankar et al., 2005, Thomas et al., 2005, Zhu et al., 2003). These receptors mediate a variety of nonclassical actions in a wide range of vertebrate vorapaxar and tissues (Prossnitz and Barton, 2011, Thomas, 2012). Rapid, membrane-initiated androgen actions have also been reported in vertebrate tissues and cells, but the underlying mechanism and membrane androgen receptor (mAR) mediating many of these responses has remained elusive. In 2014, a novel membrane androgen receptor (mAR), unrelated to the nuclear androgen receptor (nAR), was discovered in Atlantic croaker ovaries. The croaker mAR was identified as ZIP9, a member of the ZIP (SLC39A) zinc transporter family (Berg et al., 2014). Human ZIP9 expressed in nAR-negative breast and prostate cancer cell lines was also shown to display all the characteristics of a mAR (Thomas et al., 2014). Moreover, ZIP9 is the intermediary in testosterone-induced apoptosis in both croaker ovarian follicle cells and human cancer cells (Berg et al., 2014, Thomas et al., 2014). ZIP9 is the first novel protein identified in vertebrates with all the characteristics of a mAR and the only steroid receptor found to date that directly regulates zinc homeostasis by regulating zinc transport. In this short review we briefly discuss membrane-initiated androgen actions in vertebrate tissues and cells, particularly in the ovary, the discovery and characterization of ZIP9 as a mAR, the physiological functions of zinc and ZIP zinc transporters, and features of ZIP9 in other cell lines.
Rapid, cell surface-initiated androgen actions Rapid androgen actions have been reported in a variety of tissues and cell types including the ovary (Braun and Thomas, 2003, Lutz et al., 2003, Machelon et al., 1998), testis (Fix et al., 2004, Lyng et al., 2000), prostate and breast cancer cell lines (Kampa et al., 2005, Lyng et al., 2000, Papadopoulou et al., 2008), and macrophages (Benten et al., 1999). Many of these responses involve G protein activation (Machelon et al., 1998, Benten et al., 1999), and while several were found to require expression of the nAR (Fix et al., 2004, Lutz et al., 2003), several of these androgen-responsive cells such as PC-3 and DU145 prostate cancer cell lines (Lyng et al., 2000, Papadopoulou et al., 2008), and IC-21 macrophages (Benten et al., 1999), are nAR negative. Therefore, there must be an additional way in which androgens can mediate cell surface-mediated actions, such as by acting through a distinct mAR. Interestingly, nonclassical androgen actions have been shown to result in similar response in various models, such as inducing a rapid increase in intracellular calcium fluxes in Sertoli cells, prostate cancer cells (Lyng et al., 2000), granulosa cells (Machelon et al., 1998), and macrophages (Benten et al., 1999). Likewise, androgens have been shown to mediate apoptosis through cell-surface initiated actions in metastatic prostate and breast cancer cell lines (Hatzoglou et al., 2005, Kampa et al., 2005, Papadopoulou et al., 2008), and glial cells (Gatson et al., 2006). These analogous responses mediated through androgens acting on the cell surface suggest that there may be common underlying mechanisms that can mediate rapid, cell surface-mediated androgen actions in vertebrates. In the ovary, androgens are involved in various reproductive processes including follicle growth (Forsgren and Young, 2012, Vendola et al., 1998, Wang et al., 2001), transition of follicles to more advanced stages (Yang and Fortune, 2006), and in atresia and follicle cell apoptosis (Billig et al., 1993, Duda et al., 2012). However, the role of the nAR in mediating these responses was not examined. The decreased follicle growth and high levels of apoptosis and follicle atresia in nAR knockout animals (Hu et al., 2004, Sen and Hammes, 2010, Walters et al., 2007, Walters et al., 2012) suggest that the nAR plays an important role in folliculogenesis, but does not mediate atresia and follicle cell apoptosis. Androgen-induced apoptosis appears to be localized to granulosa cells (Billig et al., 1993, Duda et al., 2012) which are one of the cell models in which androgens have been shown to act in a nonclassical fashion. The potential for androgens to mediate different physiological responses by acting through distinct membrane and nuclear receptors within ovarian follicle cells may explain some of the differing roles of androgens in the ovary.