Expression of Twist and the AR are

Expression of Twist1 and the AR are increased by oxidative stress, but the change in the receptor (mRNA/protein) is lost after treatment with siRNAs that target Twist 1 (Shiota et al., 2010). Twist 1 was found to bind to E-boxes, 5′-CANNTG-3′, in the proximal promoter (−442 to +51 bp) and upstream regions (−539 to −974 bp and −1187 to −1589 bp) of the human AR gene. Functional and mutational mapping highlighted the importance of the E-boxes in the proximal promoter (Fig. 2: −168 to −150 and −26 to −8 bp), for Twist 1 stimulation of AR mRNA expression (Shiota et al., 2010). In a further twist, expression of Twist 1 can be indirectly repressed by androgens in prostate cancer Torin 2 by NKX3-1. Androgens strongly upregulate NKX3-1 production in prostate epithelial cells and in tumour cell lines (DePrimo et al., 2002, Prescott et al., 1998), where NKX3-1 binds to the promoter of the Twist 1 gene and robustly represses transcription (Eide et al., 2013). However, it has also been reported that androgens upregulate Twist 1 expression, observed in prostate cancer microarrays (Ngan et al., 2009), and in LNCaP cells after 72 h of treatment with the powerful androgen agonist R1881 (Eide et al., 2013). These apparently contradictory findings appear to be due to the differing timescales used, with an initial downregulation of Twist 1 expression by increased NKX3-1 eventually being overcome by androgen upregulation. Zeb1 (Table 1) is a transcription factor that can act as a repressor or activator of target gene expression, and ChIP analysis has confirmed that it binds directly to an E box in the hAR 5′-UTR, at position +1001 to +1006 bp (Fig. 2) where it induces increased AR transcription in triple negative breast cancer cells and human foreskin cells (Graham et al., 2010, Qiao et al., 2012). Zeb1 is intimately associated with cancer by promoting EMT and metastasis (Zhang et al., 2015) and overexpression correlates with a shorter time of progression to advanced prostate cancer after radiotherapy (Marin-Aguilera et al., 2014). Invasive ductal breast carcinomas have higher stromal expression of Zeb1 than do in situ tumours (Soini et al., 2011). In a feed forward mechanism, expression of the gene for Zeb1 is upregulated by androgens through two AREs (Anose and Sanders, 2011) and use of the androgen antagonist bicalutamide in triple negative breast cancer reduces levels of Zeb1 (Graham et al., 2010). Oestradiol upregulation of Zeb1 expression has also been reported in both human foreskin cells (Qiao et al., 2012) and in the endometrial stroma and myometrium of the mouse and human uterus (Spoelstra et al., 2006).
Conclusions and future perspectives In this review we have focused on the direct transcriptional regulation of the human AR gene in different tissues. However, it is worth noting that RNA interference mechanisms, resulting in DNA methylation of the promoter (Cho et al., 2014) or targeting positive transcriptional regulators of the receptor gene (Nadiminty et al., 2012), have also been found to down-regulate AR expression. Similarly, RNA binding proteins, PCBP1 (polyC-binding protein) (Cloke et al., 2010) and EBP-1 (Erb3-binding protein) (Zhou et al., 2010) have been shown to target the 3′UTR of the AR mRNA regulating RNA stability in endometrial stromal cells and prostate cancer respectively. The diversity of mechanism that have been identified clearly emphasise the importance of tight regulation of AR mRNA and protein in different target tissues. The situation is further complicated in that the most commonly used rodent models may not fully capture the regulation of the human AR gene. The development of more physiologically representative culture approaches offers an exciting alternative approach (Ellem et al., 2014). Notwithstanding the technical difficulties inherent in the use of spheroid cultures or tissue explants, such approaches offer the ability to consider 3D cell interactions and tissue architecture. More tractable, co-cultures will allow for epithelial-stromal interactions and functional contacts with the extra cellular matrix to be investigated in terms of regulation of AR expression and function.