Stromal and epithelial cells from rat ventral prostate during androgen deprivation and estrogen treatment--regulation of transcription

To identify the functional capacities of prostatic tissue, the expression of steroid hormone receptors, growth factors, oncogenes and particular enzymes was studied at the RNA level in isolated stromal and epithelial cells of rat ventral prostate (RVP) under different hormonal conditions (androgen deprivation, estrogen treatment). Slot blot and Northern blot analyses of isolated RNA resulted in characteristic changes: In the control prostate, androgen receptor (AR) mRNA was high in epithelium of intact prostate, but low in stroma. Its level was increased after castration in the epithelium during the initial 24 hours, whereas an only slight increase occurred in stroma after one week castration. The AR signal was not altered by estradiol treatment in epithelium and stroma. Conversely, the estrogen receptor (ER) mRNA, predominant in stroma and very low in epithelium, decreased after castration in stroma and epithelium within 24 hours and was absent one week later. After estrogen treatment the ER signal increased considerably in stroma. mRNA of both basic fibroblast growth factor (bFGF) and transforming growth factor beta (TGF-beta) were exclusively found in stroma. Both were low in controls and responded in a different way to castration and estrogen treatment within 24 hours. bFGF was increased in estrogen-treatment animals, while TGF-beta was induced by castration. Shortly after castration (2 hours) v-fos expression increased and reached a maximum after 6 hours, but was no more detectable after 12 hours in epithelium.(ABSTRACT TRUNCATED AT 250 WORDS)

Functional properties of isolated stroma and epithelium from rat ventral prostate during androgen deprivation and estrogen treatment

To identify the functional activities of prostatic stroma under different hormonal conditions, isolated stroma and epithelium from rat ventral prostate (RVP, intact or one week castrated or estrogen-treated), were studied in metabolic labeling experiments. Using a semiquantitative stereological procedure, the relative proportion of the epithelial and stromal compartment was determined in situ. In addition, the distribution of the androgen receptor was visualized by in situ hybridization and by immunocytochemistry. In castrated animals protein biosynthesis of the stroma and epithelium exceeded the control value by a factor 7 and 5, respectively. In estrogen-treated animals protein biosynthesis was reduced, reaching only between one tenth and one fifth of the control value. The amount of stroma obtained from these animals was very low. These results were confirmed by stereological findings and indicate a differential regulation of prostatic stroma and epithelium after estrogen challenge and androgen deprivation. Estrogen receptor was induced in epithelium and stroma in estrogenized animals whereas the androgen receptor was reduced in experimental specimens. During estrogenization the biosynthetic activity of both stroma and epithelium is depressed, while estrogen responsivity of the epithelium in terms of estrogen receptor expression is increased. Androgen withdrawal results in active transformation of the gland through increased stromal biosynthetic activity and epithelial regression.

Immunological characterization and activity of transglutaminases in human normal and malignant prostate and in prostate cancer cell lines

Using biochemical assays, we compared enzyme activities with the immunoreactivity of antibodies against rat seminal transglutaminase (TGase), human erythrocyte TGase and guinea pig liver TGase in human normal prostate, primary prostatic carcinomas and prostatic carcinoma cell lines. Glandular cells of the epithelium were only exceptionally positive with the antibody against (rat) secretory TGase. Using the antibodies against tissue-type TGase, most immunoreactive cells were found in the basal cell layer of prostatic epithelium as well as in stroma (fibroblasts, endothelial cells), whereas immunoreactive glandular cells were sparse. In the case of benign prostatic hyperplasia, few, irregularly distributed secretory cells along with a small number of stromal cells were also immunoreactive with the tissue-type TGase antibody. In dedifferentiated carcinomas, immunoreactive cells were nearly completely absent. Of the prostate cancer cell lines, the LNCaP line showed neither TGase enzyme activity nor immunoreactivity, whereas the PC-3 cell line displayed significant enzyme activity and immunoreactivity. No hormone-dependent changes in either enzyme activity or immunoreactivity were recorded after in vitro treatment of the respective cell lines with estrogens, androgens and antiandrogens. As there is no correlation between androgen deprivation and TGase expression in nonmalignant and malignant human prostatic epithelial cells, TGase activity more likely indicates cellular lesions and consecutive repair mechanisms.

Tissue-type transglutaminase expression in the Dunning tumor

Transglutaminases with different functions and tissue distribution patterns can be distinguished by specific antibodies and by inhibition of enzyme activity in the presence of guanosine triphosphate (GTP). The most common form is the so-called tissue-type transglutaminase that is apparently involved in membrane stabilization processes, e.g. during apoptosis, and can be inhibited by incubation with GTP at low calcium concentrations. A secretory transglutaminase that cannot be inhibited by GTP is synthesized in an androgen-dependent manner in the dorsal prostate of the rat, the site suggested to represent the origin of the Dunning tumor used as an experimental model in prostate cancer research. Here we studied the expression of transglutaminases in different Dunning tumor lines--mainly in the highly differentiated H subline--and characterized the enzyme both biochemically and immunocytochemically. A very high enzyme activity was found only in the less well differentiated HI-F tumor line. Immunohistochemical reactions and Western blot analysis showed that there is no secretory transglutaminase present in any of the Dunning tumor lines studied. Transglutaminase activity of the Dunning tumor results from the so-called tissue-type enzyme that is non-organ specific. The absence of a secretory form of transglutaminase does not support the contention of a prostatic origin of the Dunning tumor.

Arguments against the prostatic origin of the R-3327 Dunning H tumor

The Dunning tumor, originally described as a carcinoma of the rat dorsal prostate, has for long been used as an experimental model of prostatic cancer. We have recently presented a number of morphological findings that are incompatible with the prostatic origin of the H-subline of the Dunning tumor. In this paper, biochemical and immunohistochemical markers of rat prostate and mammary gland are studied in the R-3327 Dunning H tumor. Pieces of the H tumor were inoculated in male or lactating female rats. The electrophoretic protein pattern of Dunning tumor extracts was more similar to that of the mammary gland than the dorsolateral prostate. Proteins selectively appearing after metabolic labeling in Dunning tumors grown in lactating rats corresponded to labeled proteins in mammary glands from the same animals. Secretory proteins typical of the lateral prostate (SVS II) and dorsal prostate (transglutaminase) could not be detected immunohistochemically in the Dunning tumor. Western blot studies of tumor extracts and slot blot analysis of RNA preparations from the tumor confirmed the absence of SVS II and prostate specific transglutaminase from the Dunning tumor. On the other hand, the presence of mammary gland proteins such as milk fat globule membrane proteins, lactoperoxidase and lactalbumin were detected in the Dunning tumor by immunohistochemistry and Western blotting, but were absent from the dorsolateral prostate. Transferrin-mRNA, expressed in the male urogenital tract and also in the liver and other tissues, was detected in the mammary gland and Dunning tumor, but not in the dorsolateral prostate. The absence of mammary gland secretory beta-casein in the Dunning tumor was related to the elevated Ha-ras oncogene expression in the tumor, previously reported to suppress casein expression. The findings clearly demonstrate that the prostate cannot be the origin of the Dunning tumor, presently being used in prostatic cancer research. The designation prostatic adenocarcinoma for this tumor is therefore invalid. Furthermore, the data support our view that mammary gland might be the origin of the Dunning tumor, although the derivation from the bulbourethral or the parotid glands cannot strictly be excluded.