6.1. Promoter activity and hormonal regulation of the hPAP gene

Promoter constructs of hPAP covering the entire gene region of -734/+467 were functional in both prostatic and nonprostatic cell lines, but these constructs could not be induced with androgen, glucocorticoid, or progesterone in transfection assays (II). There was no androgen response evident in the case of the hPAP -1652/+43 construct tested in LNCaP and PC-3 cells. Previously Virkkunen et al. (1994) reported that the SREs of the hPAP gene at -178 and +336 and the corresponding ones in rat PAP are able to bind AR in vitro. Furthermore, putative SREs of the human and rat PAP genes at -1576 and -1612, respectively, have weak in vitro AR-binding capacities. Therefore our transfection results indicate that androgens cannot directly regulate hPAP gene expression via receptor binding to those SREs. On the other hand, androgens down-regulate hPAP mRNA in LNCaP cells under similar culture conditions (Henttu et al. 1992, Lin et al. 1994). In nuclei run-on experiments, it seems that androgens stimulate the transcription of hPAP mRNA in low density cells, while, the transcription of hPAP mRNA is suppressed by androgen in high density cells (Zelivianski et al. 1998). These results suggest that androgens could regulate hPAP expression at the level of transcription. One possibility for the contradiction in phenomenon is that the distal parts of the hPAP gene contain additional SRE(s) that might cooperate with the SREs in the proximal areas and be essential for optimal androgen action. There are several examples of genes where the androgen effects cannot be demonstrated maximally without the interaction of several SREs, as in the case of the probasin gene (Rennie et al. 1993, Kasper et al. 1994). Another possibility is that the hPAP gene needs another transcription factor(s) along with AR to enhance transcription and define prostate-specific expression. The response element(s) for this factor(s) could be far upstream or somewhere else in the gene (II). Zelivianski et al. (2000) studied the cis-regulatory elements of the hPAP promoter covering the region from -2899 to +87. Two regions of transcriptional suppression were identified and located at -2899/-2583 and -2583/-1305, whereas the fragment of -1305/-779 had a transcriptional activation function. It has not been reported if these different 5’-deletion constructs could respond to androgen stimulation.

Contrary to hPAP promoter constructs, hPSA and hK2 promoters showed a weak androgen induction in LNCaP cells, but a strong induction in PC-3 or CV-1 cells. This indicates that the hPAP gene might be differently regulated by androgen compared to the hPSA or hK2 gene (II). In fact, androgen stimulation of the proliferation of prostate epithelial cells coincides with a decrease in hPAP and an increase in hPSA mRNA (Henttu et al. 1992, Lin et al. 2000). The difference in the effects of androgens on hPAP and hPSA expression could be due to the observation that the cellular form of hPAP functions as a negative regulator by dephosphorylating c-ErbB-2/neu oncoprotein, resulting in the down-regulation of cell proliferation (Lin & Meng 1996, Lin et al. 1998, Meng & Lin 1998), while PSA may activate an insulin-like growth factor (IGF) by hydrolyzing its binding protein (IGF-BP), leading to the stimulation of cell growth (Cohen et al. 1992, 1994). Androgen down-regulation of hPAP could, in part, be via the PKC signal transduction pathway (Henttu & Vihko 1996). TPA, a ligand of PKC, and DHT have similar effects on PKC activity as well as the hPAP expression in LNCaP cells (Lin et al. 2000).

Androgen up-regulation of the PSA gene could be due to a direct interaction of the androgen/AR complex with the functional ARE in the promoter of the gene (Cleutjens et al. 1996, 1997a). The different androgen response of the PSA or hK2 promoter in LNCaP and PC-3 might be due to the different AR coregulators in these two cell lines. Fujimoto et al. (2001) examined the expression level of some of the AR coactivators in LNCaP, PC-3, DU-145 cancer cell lines, benign prostatic tissue sample, and prostate cancer tissue specimens. In the cell lines, coactivator SRC1 was expressed ubiquitously at almost equal amounts, whereas other coactivators, such as ARA55, ARA54, TIF2, and RAC3, displayed cell line-specific expression.

The promoter activities of PSA and hK2 could also be induced by glucocorticoid and progesterone, indicating that the SREs in the promoters of the PSA and hK2 genes are not androgen-specific, as in the case of the rat probasin gene (Claessens et al. 1996). The androgen-specific regulation of the PB gene can be explained by the presence of an ARE in its promoter, specifically recognized by AR but not GR (Kasper et al. 1999).