| Transcriptional regulation of the human prostatic acid phosphatase gene: Tissue-specific and androgen-dependent regulation of the promoter constructs in cell lines and transgenic mice | ||
|---|---|---|
| Prev | Chapter 5. Results | Next |
The 5’-flanking region (-426/+28) of the rat probasin gene, which has been shown to direct prostate-specific gene expression in transgenic mice, was used to identify the exact DNA-binding site of a putative prostate-specific transcription factor. Promoter activity analysis by CAT assay revealed that the construct pCAT PB -244/+52 was equally well induced by androgens both in prostatic LNCaP and nonprostatic COS-1, MCF-1, HEC-1, and HEP-1 cell lines (III). This indicated that although the probasin promoter region -244/+52 was important for androgen regulation, it did not act in a prostate-specific manner. Androgen induction of the promoter was decreased to 50% in LNCaP cells upon the 5’-deletion of the construct from -278 to -244. Since no effect was observed in COS-1 cells, the sequence around -278 to -244 was shown to be crucial for prostate-specific gene expression. DNA-protein interaction study within the region from -310 to -96 by DNase I footprinting exhibited a protected window between the nucleotides -251 and -240 when prostatic LNCaP and PC-3 nuclear extracts were used. The sequence of this area is 5’-GAAAATATGATA-3’. EMSA using rPB -257/-235 as a probe, which covering the 12 bp sequence, displayed a strong shift band also when LNCaP or PC-3 nuclear extracts were used. Mutation of two consecutive nucleotides in the 5’-end, 3’-end, or in middle of that 12 bp sequence abolished the formation of the specific DNA-protein complex. Deletion of the 12 bp from pCAT PB -426/+52 or mutation of those six nucleotides within the 12 bp sequence simultaneously in pCAT PB -426/+52 significantly decreased androgen induction in LNCaP cells compared to the wild-type construct. The pCAT PB -257/+52 construct, which contained the binding site of the putative prostatic regulatory protein and six additional 5’-nucleotides, did not show nearly so good androgen induction as pCAT PB -278/+52. This suggested that at least 20 additional 5’-nucleotides of the binding site were needed for the regulatory protein to confer optimal androgen induction of the probasin promoter. Glucocorticoid could not induce the promoter activity of PB -278/+52, indicating that the putative prostate-specific transcription factor might interact specifically with AR.
The GAAAATATGATA sequence was also found at the first intron (+1144/+1155) of the hPAP gene. EMSA using hPAP +1136/+1164 as a probe showed a strong specific shift band when LNCaP or PC-3 nuclear extracts were used, and an rPB -57/-235 fragment could compete off the shift band (III). These results suggested that the putative prostatic regulatory protein could also play a role in the transcriptional regulation of the hPAP gene.