Identification of the deoxyribonucleic acid-binding site of a regulatory protein involved in prostate-specific and androgen receptor-dependent gene expression.

In this study, we used a 5'-flanking region (-426/+28) of the rat prostatic probasin (rPB) gene shown to be sufficient to direct prostate-specific expression in transgenic mice in identifying the exact DNA-binding site of a putative prostate-specific transcription factor. Chloramphenicol acetyl transferase (CAT)-reporter gene analyses revealed that the construct pCAT PB -244/+52 was equally well induced by androgens in both prostatic LNCaP and nonprostatic COS-1, MCF-7, HEC-1, and HEP-1 cell lines, indicating that although the probasin gene region -244/+52 was important for androgen regulation, it was not regulated in a prostate-specific manner. Further studies suggested that the region -278/-240 was most crucial for prostate-specific expression. The sequence -426/-279 could be considered a silencer area, especially in nonprostatic cells. In deoxyribonuclease I footprinting, a protected 12-bp region was found between the nucleotides -251 and -240 only with nuclear extracts of prostatic origin. Deletion of this area decreased androgen induction significantly (P < 0.05) in transient transfections of prostatic cells compared with the wild-type reporter construct. Glucocorticoids were incapable of increasing the induction of the pCAT PB -278/+52 reporter construct compared with that of pCAT PB -244/+52 in the prostatic cell line LNCaP, suggesting that the putative prostate-specific protein acts as an inducer only when androgen and androgen receptor are present.

Steroid-involved transcriptional regulation of human genes encoding prostatic acid phosphatase, prostate-specific antigen, and prostate-specific glandular kallikrein.

We have compared the steroid regulation of human genes encoding prostatic acid phosphatase (hPAP), prostate-specific antigen (hPSA), and prostate-specific glandular kallikrein (hK2) at the level of transcription. Reporter constructs of hPAP promoter covering the region -734/+467 were functional in both prostatic (LNCaP and PC-3) and nonprostatic (CV-1) cell lines in transient transfections. hPAP -231/+50 with eight identified transcription factor-binding sites showed the highest, and hPAP -734/+467 showed the lowest transcriptional activity in CV-1 cells. The hPAP promoter could not be induced with androgen, glucocorticoid, or progesterone, contrary to the hPSA (-620/+40) and hK2 (-493/+27) promoters in PC-3 cells cotransfected with the respective steroid receptor expression vector. Therefore, steroids cannot directly regulate hPAP gene expression via receptor binding to steroid response elements at -178 and +336, which have been shown to have androgen receptor-binding ability in vitro. Glucocorticoid was the most powerful activator of the hPSA construct at 10-nM steroid concentrations. On the contrary, glucocorticoid stimulation of the transcriptional activity of the hK2 construct was the weakest among the tested steroids. The results indicate that the steroid response elements in the proximal promoters of hPSA and hK2 genes are not androgen specific, offering the molecular basis for the expression of these genes outside the prostate in tissues containing steroid receptors.

17 beta-hydroxysteroid dehydrogenases and cancers.

17 beta-Hydroxysteroid dehydrogenases (17HSDs) catalyze the interconversions between active 17 beta-hydroxysteroids and less-active 17-ketosteroids thereby affecting the availability of biologically active estrogens and androgens in a variety of tissues. The enzymes have different enzymatic properties and characteristic cell-specific expression patterns, suggesting differential physiological functions for the enzymes. Epidemiological and endocrine evidence indicate that estrogens play a key role in the etiology of breast cancer while androgens are involved in mechanisms controlling the growth of prostatic cells, both normal and malignant. Recently, we have developed, using LNCaP prostate cancer cell lines, a cell model to study the progression of prostate cancer. In the model LNCaP cells are transformed in culture condition to more aggressive cells, able to grow in suspension cultures. Our results suggest that substantial changes in androgen and estrogen metabolism occur in the cells during the process. These changes lead to increased production of active estrogens during transformation of the cells. Data from studies of breast cell lines and tissues suggest that the oxidative 17HSD type 2 may predominate in human non-malignant breast epithelial cells, while the reductive 17HSD type 1 activity prevails in malignant cells. Deprivation of an estrogen response by using specific 17HSD type 1 inhibitors is a tempting approach to treat estrogen-dependent breast cancer. Our recent studies demonstrate that in addition to sex hormone target tissues, estrogens may be important in the development of cancer in some other tissues previously not considered as estrogen target tissues such as colon. Our data show that the abundant expression of 17HSD type 2 present in normal colonic mucosa is significantly decreased during colon cancer development.

Expression profiling of PC-3 cell line variants and comparison of MIC-1 transcript levels in benign and malignant prostate.

BACKGROUND: The mechanisms behind prostate cancer progression are largely unknown, but macrophage inhibitory cytokine 1 (MIC-1) has been suggested to be involved in tumour dissemination in vivo due to its reductive effect on cell adhesion. MATERIALS AND METHODS: We used two PC-3 prostate cancer epithelial cell line variants as tools to screen for gene expression differences during prostate cancer progression by cDNA microarray analysis. Selected genes were further analysed by Northern blot analysis using mRNA isolated from prostatic cell lines and tissues. MIC-1 expression was studied by in situ hybridization in archival patient specimens containing benign and malignant prostatic tissue. RESULTS: Gene expression of human collagen type VI, basement membrane heparan sulphate proteoglycan, integrin alpha 1, and fibronectin I were remarkably decreased in suspension-adapted PC-3 (saPC-3) cells, indicating a gene expression profile of reduced cell adhesion. Asparagine synthetase, serine protease 1, stanniocalcin homologue, NAD-dependent methylene tetrahydrate dehydrogenase cyclohydrolase (NMDMC), fortilin, and MIC-1 were overexpressed in saPC-3 cells. In prostate, the MIC-1 gene was mainly expressed in cancer tissue. However, MIC-1 transcripts were detected in benign tissue areas, especially in specimens containing prostate cancer with Gleason sum scores of 5-8. A significant inverse correlation (Spearman's rho correlation coefficient -0.928**) was observed between the ratio of cancerous to benign MIC-1 expression levels and Gleason scores. CONCLUSIONS: Differential expression of the MIC-1 gene occurs during prostate cancer progression. The transcript level of the MIC-1 gene in histologically benign tissue seems to approach that in paired cancer tissue concomitant with an increasing Gleason score.

Differentially expressed genes in two LNCaP prostate cancer cell lines reflecting changes during prostate cancer progression.

Prostate cancer tends to become transformed to androgen-independent disease over time when treated by androgen-deprivation therapy. We used two variants of the human prostate cancer cell line LNCaP to study gene expression differences during prostate cancer progression to androgen-independent disease. Production of prostate-specific antigen was regarded as a marker of androgen-dependence and loss of prostate-specific antigen was regarded as a marker of androgen-independence. mRNA from both cell lines was used for cDNA microarray screening. Differential expression of several genes was confirmed by Northern blotting. Monoamine oxidase A, an Expressed Sequence Tag (EST) similar to rat P044, and EST AA412049 were highly overexpressed in androgen-dependent LNCaP cells. Tissue-type plasminogen activator, interferon-inducible protein p78 (MxB), an EST similar to galectin-1, follistatin, fatty acid-binding protein 5, EST AA609749, annexin I, the interferon-inducible gene 1-8U, and phospholipase D1 were highly overexpressed in androgen-independent LNCaP cells. All studied genes had low or no expression in PC-3 cells. The EST similar to rat P044, the EST similar to galectin-1, follistatin, annexin I, and the interferon-inducible gene 1-8U were also expressed in benign prostatic hyperplasia tissue. The Y-linked ribosomal protein S4, Mat-8, and EST AA307912 were highly expressed in benign prostatic hyperplasia tissue. Additionally, both confirmation of differential expression in Northern blots and in situ hybridization were carried out for monoamine oxidase A, the EST similar to rat P044, the EST similar to galectin-1, fatty acid-binding protein 5, and the interferon-inducible gene 1-8U. We identified several potential prostate cancer markers, indicating that the method used is a useful tool for the screening of cancer markers, but other methods, such as in situ hybridization, are needed to further investigate the observations.

Differential androgen regulation of rat prostatic acid phosphatase transcripts.

Steady-state levels of the mRNAs encoding rat prostatic acid phosphatase (rPAP) were measured from the ventral prostate of rats after castration and testosterone replacement. The longest and the most abundant mRNA molecule (4.9 kb) of the rPAP gene was quite resistant to the hormonal status of the prostate, whereas its medium-sized transcript (2.3 kb) and the shortest one (1.5 kb) were up-regulated by androgens. For comparison, we also analyzed the prostatic mRNA levels of rat lysosomal acid phosphatase (rLAP), whose gene expression is not considered to be hormonally dependent. Only modest variation was observed in the rLAP transcript after androgen withdrawal and replacement. Corresponding protein level investigations are analogous with our regulation studies, particularly in the case of 1.5-kb rPAP mRNA.

Human prostate-specific glandular kallikrein is expressed as an active and an inactive protein.

A polymorphism in the human prostate-specific glandular kallikrein (hKLK2) gene was described by direct sequencing (by PCR) of genomic DNAs isolated from prostatic cancer tissue, benign prostatic hyperplasia tissue, and blood leukocyte specimens. Results showed two forms of human prostate-specific glandular kallikrein protein (hK2), a consequence of a change from C to T at base 792 in the hK2 coding region. Producing the two forms as recombinant proteins in insect cells demonstrated that Arg226-hK2 (CC genotype) is an active protein and Trp226-hK2 (TT genotype) is inactive. Polymorphism studies of 36 patients with prostatic diseases identified only 1 with the TT genotype. The same kind of polymorphism was not detected in the human prostate-specific antigen (hKLK3) gene. Arg226-hK2 possessed only trypsin-like enzyme activity, whereas recombinant human prostate-specific antigen (hPSA) had only chymotrypsin-like activity. Monoclonal and polyclonal antibodies raised against hPSA purified from seminal plasma detected both active and inactive hK2. Thus, because inactive as well as stable hK2 protein may be present, a lack of trypsin-like activity in hPSA standards is not enough to confirm that the materials are free of hK2 contamination.

Expression of transmembrane serine protease TMPRSS2 in mouse and human tissues.

The purpose of this study was to clarify the expression of TMPRSS2 in mice during development and to compare the tissue distribution of the transcripts in adult mouse and human tissues. Mouse TMPRSS2 cDNA was cloned; the predicted amino acid sequence contains 490 residues sharing 81.4% similarity with human TMPRSS2. According to northern blots, mouse TMPRSS2 is expressed mainly in the prostate and kidney, while human TMPRSS2 is expressed in the prostate, colon, stomach, and salivary gland. In situ hybridization analyses of mouse embryos and adult tissues revealed that TMPRSS2 was expressed in the epithelia of the gastrointestinal, urogenital, and respiratory tracts. Expression was very selective and constant after the gene was turned on during development. Expression of TMPRSS2 was localized in the luminal epithelial cells of the mouse and human prostate. The information presented here will be useful in further studies regarding the function and physiological significance of TMPRSS2.

Rat acid phosphatase: overexpression of active, secreted enzyme by recombinant baculovirus-infected insect cells, molecular properties, and crystallization.

Rat prostatic acid phosphatase (rPAP; orthophosphoric-monoester phosphohydrolase (acid optimum), EC 3.1.3.2) was expressed in the baculovirus expression vector system. Recombinant protein was secreted into the medium at a high yield by infected insect cells, which were cultured at high density in a 30-liter bioreactor allowing high oxygen content for rapidly growing cells. About 20% of the cell protein produced was rPAP. Partial sequence determination of the N terminus of the purified recombinant secreted protein revealed identity to the native secreted protein, showing that the signal peptide is recognized and properly cleaved in insect cells. The enzyme was purified by using L-(+)-tartrate affinity chromatography. The purified protein had a high specific activity of 2620 mumol.min-1.mg-1 with p-nitrophenyl phosphate at the substrate, and it also showed phosphotyrosine phosphatase activity. The molecular mass of the recombinant rPAP was 155 kDa. Two subunits of 46 kDa and 48 kDa could be detected in SDS/PAGE, but only one subunit of 41 kDa was present after digestion with N-glycosidase. The active enzyme is a trimer of subunits differing only in glycosylation. When recombinant rPAP was crystallized with polyethylene glycol 6000 as the precipitant, the crystals were trigonal (space group P3(1)21) with cell dimensions a = 89.4 A and c = 152.0 A. The observed diffraction pattern extends to a resolution of at least 3 A.

The TMPRSS2 gene encoding transmembrane serine protease is overexpressed in a majority of prostate cancer patients: detection of mutated TMPRSS2 form in a case of aggressive disease.

The serine protease TMPRSS2 gene expression was studied by in situ hybridization using benign prostatic hyperplasia and prostate cancer tissue samples from 32 patients. Expression of TMPRSS2 gene was higher in cancer cells than that in benign cells in 84% of the specimens containing both benign and malignant tissues. The TMPRSS2 mRNA level was significantly increased in poorly differentiated (p = 0.014, n = 7) and untreated (p = 0.022, n = 13) primary prostate adenocarcinomas compared to benign tissues. In addition, androgen-deprivation therapy significantly decreased the expression of TMPRSS2 in benign prostate tissue (p = 0.07), which is in accordance with the androgen-inducible expression of the gene. The gene copy number of TMPRSS2, analyzed by competitively differential PCR, was duplicated in the malignant cells of about 38% of the prostate cancer patients analyzed. Thus, the increase in the gene copy number is probably not the primary reason for the detected overexpression of the TMPRSS2 gene. Mutations in the TMPRSS2 gene were screened using DNA isolated from paraffin-embedded prostate cancer tissues from 9 patients with aggressive prostate cancer and from 9 patients with nonaggressive disease. Thirteen exons covering the coding region were checked using enzymatic mutation detection and direct sequencing. One patient with aggressive prostate cancer carried a deletion and a stop codon in exon 11, leading to inactivation of the serine protease domain in TMPRSS2.

Comparison of human prostate specific glandular kallikrein 2 and prostate specific antigen gene expression in prostate with gene amplification and overexpression of prostate specific glandular kallikrein 2 in tumor tissue.

BACKGROUND: There is a need for specific markers that can indicate the different stages of prostate carcinoma. There is ongoing speculation concerning the value of prostate specific glandular kallikrein (hK2) in this regard. METHODS: The expression levels of both hK2 and human prostate specific antigen (hPSA) were compared at the mRNA and protein levels by using in situ hybridization and immunohistochemistry techniques in tissue specimens from patients with benign prostatic hyperplasia and malignant prostate carcinoma. The respective gene copy numbers were analyzed by a competitively differential polymerase chain reaction-based method. RESULTS: In situ hybridization revealed that hK2 was expressed at significantly higher levels in malignant prostate tissue compared with benign prostate tissue (P < 0.0005), whereas hPSA expression levels were the reverse (P = 0.06). In benign tissue, the mean level of hK2 mRNA was 82% of the respective value of hPSA (P < 0.003), whereas, in tumor tissue, the mean hK2 expression level was 21% higher than that of hPSA (P < 0.01). The results at the protein level supported the mRNA findings: hPSA expression was lower in malignant tissues compared with benign tissues (17 of 20 specimens), whereas an increase in hK2 expression was detected in 17 of 19 specimens. The authors report that the hK2 gene (hKLK2) was amplified in prostate carcinoma tissue, whereas the hPSA gene was not. There was a correlation between hPSA and hK2 mRNA levels in both benign tissue (correlation coefficient [r] = 0.735; P < 0.01) and malignant tissue (r = 0.767; P < 0.01). CONCLUSIONS: Gene amplification of hKLK2 may be one of the factors leading to higher expression of hK2 in prostate carcinoma. The correlation between hK2 and hPSA expression levels indicates coordinated expression of the genes in both normal and abnormal prostate gland. The results suggest the potential value of hK2 in the diagnosis of prostate carcinoma through mRNA analyses and gene amplification.

Prostatic expression of human 5alpha-reductase type 2 during finasteride therapy: a randomized, double-blind, placebo-controlled study.

This study was aimed at exploring the effect of finasteride, a non-steroidal competitive inhibitor of the enzyme 5alpha-reductase, on 5alpha-reductase type 2 at the mRNA level in human prostate, using an in situ hybridization technique. After randomization, 10 men with benign prostatic hyperplasia (BPH) received oral finasteride (5 mg daily) and five men with BPH received placebo daily. Careful clinical examination was carried out and 2 biopsy samples were taken transrectally before the treatment and after 3, 6, and 12 months of treatment. In situ hybridization was carried out and expression of 5alpha-reductase type 2 mRNA was measured. The results showed that finasteride treatment had no permanent effect on expression of 5alpha-reductase type 2 in prostatic epithelium, compared with placebo treatment. Expression varied during treatment, but there was no clear tendency in this expression. The signal was localized in the epithelial cells. We conclude that finasteride treatment had no clear effect on human 5alpha-reductase type 2 expression in the prostate.

Several genes encoding ribosomal proteins are over-expressed in prostate-cancer cell lines: confirmation of L7a and L37 over-expression in prostate-cancer tissue samples.

A cDNA library specific for mRNA over-expressed in prostate cancer was generated by subtractive hybridization of transcripts originating from prostatic hyperplasia and cancer tissues. cDNA encoding ribosomal proteins L4, L5, L7a, L23a, L30, L37, S14 and S18 was found to be present among 100 analyzed clones. Levels of ribosomal mRNA were significantly higher at least in one of the prostate-cancer cell lines, LNCaP, DU-145 and PC-3, than in hyperplastic tissue, as determined by slot-blot hybridization. Furthermore, L23a- and S14-transcript levels were significantly elevated in PC-3 cells as compared with those in the normal prostate epithelial cell line PrEC. Generally, dramatic changes in the mRNA content of the ribosomal proteins were not detected, the most evident over-expression being that of L37 mRNA, which was 3.4 times more abundant in LNCaP cells than in hyperplastic prostate tissue. The over-expression of L7a and L37 mRNA was confirmed in prostate-cancer tissue samples by in situ hybridization. Elevated cancer-related expression of L4 and L30 has not been reported, but levels of the other ribosomal proteins are known to be increased in several types of cancers. These results therefore suggest that prostate cancer is comparable with other types of cancers, in that a larger pool of some ribosomal proteins is gained during the transformation process, by an unknown mechanism.

Site-directed mutagenesis of prostatic acid phosphatase. Catalytically important aspartic acid 258, substrate specificity, and oligomerization.

At the active site of rat prostatic acid phosphatase (rPAP), residue Asp258 is a suitable candidate to act as an acid/base catalyst during phosphoester hydrolysis. It was changed to Asn, Ser, and Ala by site-directed mutagenesis. All these mutants were inactive, indicating that Asp258 may act as a proton donor in catalysis. Tyr123 and Arg127 residues, located at the entrance of the active site surface in rPAP, are likely to be responsible for the substrate specificity of the enzyme. The corresponding residues in lysosomal acid phosphatase (LAP) are Lys and Gly. In order to clarify the roles of the Tyr123 and Arg127 residues, lysosomal type rPAP mutants (Y123K, R127G and Y123K,R127G) were generated. Sensitivity of Y123K,R127G to tartrate inhibition was similar to that observed in the case of LAP, indicating that these residues might be responsible for differences in substrate specificity between the enzymes of prostatic and lysosomal origin. However, unlike human LAP, the lysosomal type mutants hydrolyzed the suggested PAP-specific substrates, phosphocreatine and phosphocholine, showing that Tyr123 and Arg127 are not the only residues contributing to the substrate specificity of rPAP. The residues Trp106 and His112 appeared to be important in the dimerization of rPAP. Oligomerization mutants (W106E, H112D and W106E,H112D) existed in a monomeric form without catalytic activity or a tartrate binding ability.

Structural comparison of human and rat prostate-specific acid phosphatase genes and their promoters: identification of putative androgen response elements.

Structural comparison of human and rat prostate-specific acid phosphatase (hPAP and rPAP) genes indicate that the exon number is different between these species. The hPAP gene contains 10 exons, whereas the rPAP gene was 11 exons. However, exons 2-9 of the genes are identical in size. The 5' regions of the two genes show 71% identity in the most homologous region +1 to +340. The 5' untranslated regions of the human and rat genes are 50 and 49 nucleotides long, respectively. An Alu sequence is present upstream from the proximal promoter of the hPAP gene. Five putative androgen response elements altogether were localized in both the human and rat gene, one of which is conserved in location and sequence between the two genes. Two of these elements in both genes, the conserved one in the proximal promoter region and another one in intron 1, were shown to bind androgen receptor efficiently in vitro.