Relationship between androgen receptor binding activity in human prostate cancer and clinical response to endocrine therapy.

The authors investigated the ability of androgen receptor binding in prostate cancer tissue to predict the response of prostate cancer patients to endocrine therapy. The clinical response of 37 previously untreated patients with various grades and stages of prostate cancer was correlated with androgen receptor binding and detailed histologic data obtained before treatment. All patients underwent cold-punch transurethral resection of the prostate and received endocrine therapy. The association between time to progression and cytosolic androgen binding was not significant. However, the associations of time to progression to nuclear binding and to total androgen binding were significant (P = 0.029 and 0.038, respectively). The authors found no association between clinical stage and time to progression, but did find an association between time to progression and pathologic grade (P = 0.003); grade 4 lesions were the least responsive to hormone therapy. When grade 4 lesions were excluded (N = 3), binding levels were still predictive of progression independently of grade and stage. The authors conclude that nuclear receptor binding activity in localized and metastatic prostate cancer tissue is predictive of response to hormonal manipulation.

The effects of androgen on the transcription of specific genes in guinea pig seminal vesicle epithelium.

Steroid hormones have been shown to have highly differential effects on the expression of abundant cell-specific protein genes in a multitude of model tissues. In rat seminal vesicle, for example, DNA clones representing two major secretory protein genes have been used to show that both of the genes are differentially regulated by androgen. In this paper, we have examined the effects of androgen on the transcription of two major secretory protein genes in guinea pig seminal vesicle epithelium. Nuclear run-off experiments were used to show that castration of the adult resulted in a 3-fold decrease in total transcription activity. Surprisingly, the decrease in total transcriptional activity was not reflected in a differential decrease in the transcriptional activity of the two major secretory protein genes. When the effects of castration on the transcriptional activity of the major secretory protein genes were compared to the effects on other genes, it was found that the transcriptional activity of each gene examined was decreased by the same magnitude as the major secretory protein genes. Similarly, the transcriptional activity of every gene examined increased by the same magnitude as the major secretory protein genes during hormone repletion of the castrated adult. Thus, in contrast to the differential effects of steroids on the transcription of abundant cell-specific proteins in many other steroid-dependent tissues, the transcription of major secretory proteins in guinea pig seminal vesicle epithelium appears to be regulated in parallel with many other genes. The generalized effects of androgen on transcriptional activity could account for the generalized effects of androgen on seminal vesicle epithelial cell structure and function.

The growth of individual seminal vesicle epithelial cells and their proliferation.

Histologically seminal vesicle epithelium (SVE) of the intact adult guinea pig is a discrete and segregated monolayer of highly specialized tall columnar cells. The epithelial layer is so sharply demarcated from its attached stroma (primarily smooth muscle), that blunt dissection alone is sufficient to separate epithelium from muscle. After castration the epithelial cells decrease in both size and number so that by the fifth day, the surviving cells are greatly involuted structurally and comprise only about 12% of the original numerical population normally present in one seminal vesicle. Injected testosterone leads to restructuring of individual cells followed by cell replenishment. The major goal of this effort was to elaborate upon the processes of individual cell growth and cell replenishment during restoration of the tissue to normal cell size and number. The two separate processes were studied using light and electron microscopy, [3H]thymidine incorporation, and Northern blots with labeled histone gene probes. By approximately 48 hr of hormone repletion, parenchymal cell size had returned to normal as the result of a dramatic anabolic process of individual cell growth while cell number remained unchanged. During the subsequent 48-hr period of hormone repletion, the cell population was restored to normal as cell replenishment became the predominant process. Microscopic analysis at intervals throughout the 96-hr period failed to disclose any mitotic events to account for cell replenishment even when Colcemid had been administered. Nor could the increase in cell numbers be correlated with a great increase in [3H]thymidine incorporation or in histone mRNA synthesis. Thus, we could provide no evidence that mitotic division of the parenchymal cells themselves is responsible for cell replenishment. During the 24- to 48-hr interval of hormone repletion, electron microscopic examination disclosed the presence of small epithelial cells lying in a basal position. Some of these cells were seen to insert themselves between the basal regions of parenchymal cells and to expand from the basement membrane into the parenchyma. Possible origins of the cells which replenish the tissue are discussed.

The cDNA cloning of a 55-kilodalton protein from guinea pig seminal vesicle. Evidence that the protein is the precursor of a 25-kilodalton basic secretory protein.

Guinea pig seminal vesicle epithelium synthesizes and secretes large amounts of four secretory proteins including a basic protein designated SVP-1. The latter migrates as a protein of 25 kDa by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. When total cellular RNA from seminal vesicle epithelium was translated in vitro by a rabbit reticulocyte system, SVP-1 was not identifiable as a product. In contrast, a major product of the reticulocyte system was a 55-kDa protein which paradoxically could not be identified as one of the four known secretory proteins or as an abundant tissue protein. We isolated a cDNA clone which corresponded to the mRNA which coded for the 55-kDa protein. The mRNA was 1800 bases and of very high abundance; only the transcripts corresponding to the major secretory proteins SVP-3 and SVP-4 were as conspicuous. From these observations, we decided to test whether the 55-kDa protein was related to SVP-1. The 55-kDa protein was found to share properties of SVP-1, including an alkaline pI and selection by antibody against SVP-1. After proteolytic digestion of the 55-kDa protein, one fragment co-migrated with authentic SVP-1. Furthermore, detectable fragments of the 55-kDa protein smaller than SVP-1 all co-migrated with fragments generated from the proteolytic digestion of SVP-1. Finally, we show that the 55-kDa protein could be processed in vitro by seminal vesicle lumenal extracts to yield a 30-kDa protein and a protein the size of native SVP-1. The best explanation of our results is that the 55-kDa protein represents the primary translation product and precursor of SVP-1. Our proteolytic mapping and in vitro processing studies are consistent with the idea that the 55-kDa protein is a tandem repeat of two SVP-1 molecules. We used the cDNA clone to study the expression of the corresponding gene in different states of androgen depletion and repletion.

Androgen receptor binding activity in human prostate cancer.

Androgen binding (cytosol and nucleus) was measured in tissue obtained from 223 untreated patients with proven prostate cancer (199 primary tumor, 24 malignant lymph nodes), 19 patients with hormone refractory cancer, and 46 patients with benign prostatic hyperplasia (BPH). The mean binding in both the cytosol and nucleus was significantly higher for patients with cancer than for those with BPH. Binding appeared to correlate with tumor stage. Androgen binding in malignant nodes can differ from that in the primary tissue and can vary from node to node in the same patient. Results obtained from an assay using a single saturating concentration of R1881 correlated well with those calculated from a full six-point Scatchard analysis when an adequate amount (500 mg) of tissue was available. However, binding results obtained from a single-point analysis performed on needle biopsy specimens (about 50 mg) obtained before complete surgical removal of the prostate correlated poorly with those derived from a full six-point analysis performed on tissue (500-1000 mg) removed from the center of the malignancy. Androgen binding in nuclear extracts of histologically benign tissue adjacent to the malignancy was significantly higher than in nuclear extracts of BPH tissue. Cytosolic androgen binding in tissue removed from patients who were refractory to hormonal therapy was higher than in tissue from untreated cancer patients. The binding of estradiol by extracts of benign and malignant prostate tissue was low or absent and, thus, did not appear to be a significant phenomenon.

Quantification of liver and kidney phosphofructokinase by radioimmunoassay in fed, starved and alloxan-diabetic rats.

A newly developed specific radioimmunoassay was used to quantify phosphofructokinase protein directly and independently of assayable activity in liver and kidney cytosol of normal fed, starved and alloxan-diabetic rats. In the fed state, liver phosphofructokinase concentration was 0.096 microM and the kidney enzyme was 0.086 microM (mumol/kg of tissue). In the starved state (24h), liver and kidney phosphofructokinase concentrations decreased by 30%. Prolonged starvation up to 72h did not further decrease enzyme concentration. In liver, total enzyme content during starvation declined by more than 50%, secondary also to a decrease in liver weight. In the alloxan-diabetic rats, there was a 22% decrease in enzyme protein concentration in liver and kidney. Total enzyme content per liver actually decreased much more (46%), because diabetes also resulted in a decrease in liver size. In conjunction with assayable activity measurements, the results of the radioimmunoassay allowed us to calculate the apparent specific activity of the enzyme. The specific activity of the kidney enzyme was 2-3 times that of the liver. Little or no change in specific activity of the liver or kidney enzyme occurred as a result of starvation or chemically induced diabetes. Tissue enzyme concentrations of phosphofructokinase unequivocally reconcile the ultimate results of changing rates of synthesis and degradation and are useful data in the design of spectrophotometric, kinetic, aggregation-disaggregation and other studies.

Expression of a secretory protein gene during androgen-induced cell growth.

Guinea pig seminal vesicle epithelium is an androgen-dependent tissue composed of tall columnar cells which synthesize four major secretory proteins designated SVP-1-4. Castration promptly causes the seminal vesicle epithelium cells to undergo major morphological changes. By the fifth day after castration, cell number decreases to 12% of normal, and the surviving cells are greatly involuted structurally. An injection of testosterone was sufficient to stimulate individual cell growth to the normal tall columnar configuration by 48 h without increasing cell number. Cell growth following testosterone was preceded by a 4-fold increase between 0-12 h in total cellular RNA signifying a large increase in rRNAs. To aid in the analysis of gene expression during androgen repletion, we identified a cDNA plasmid clone representing the 3'-end of the mRNA of SVP-4. The clone pAT 76 was identified by hybrid-select translation and characterization of the product of in vitro translation. The steady-state levels of SVP-4 mRNA in blots of total cellular RNA increased between 0-24 h; this appeared to be due mainly to a general nonselective increase in all or most of the abundant mRNAs. We demonstrated by in vitro translation studies that the corresponding transcript was still present during the 5th and 36th days of castration without having undergone a selective loss in abundance and during hormone repletion without undergoing a selective increase. Testosterone exerted an early and more conspicuous effect on rRNA synthesis. Whether direct or indirect this may be a key event underlying the ultimate action of testosterone, namely maintenance of cell structure.

Nuclear protein matrix of seminal vesicle epithelium.

Nuclear protein matrix was isolated from guinea pig seminal vesicle epithelium and liver. The two matrices were similar in fine structure as seen by transmission electron microscopy, in protein electropherograms, and in percent composition relative to protein, DNA, and RNA. Scanning electron microscopy was used to examine intact seminal vesicle nuclei, nuclei after treatment with Triton X-100 and DNAse I, and purified nuclear matrix. The matrix surface presented a 'porous' appearance by both scanning and transmission electron microscopy. The matrices of liver and seminal vesicle epithelium (SVE) and the intact nuclei of SVE were assayed for specific binding of free synthetic androgen, 17 alpha-methyltrienolone (R1881). Saturable specific binding was demonstrable for seminal vesicle matrix but not for liver matrix. Maximal binding of androgen occurred at a concentration of approximately 12 nM and was demonstrated to be 1.34 +/- 0.22 pmol of R1881 per mg of seminal vesicle matrix protein; the Kd was approximately 8 nM. The binding of labeled R1881 to matrix could be inhibited with low concentrations of unlabeled androgens, but not with estrogens or other steroids. Our data indicate that the binding of androgen to matrix could account for at least 21% of the binding to intact nuclei.

The concentration of P-enolpyruvate carboxykinase protein in murine tissues in diabetes of chemical and genetic origin.

P-enolpyruvate carboxykinase protein was measured by radioimmunoassay in liver, kidney, and adipose tissue extracts from alloxan- and streptozotocin-diabetic rats, in liver extracts from C57BL/KsJ-db+/db+ "diabetic" mice and in liver extracts from normal mice subjected to different dietary or hormonal states. The radioimmunoassay method measured tissue enzyme concentration (nanomoles/g) and total organ enzyme content (nanomoles/liver) independently of assayable activity (units/g). The "apparent" specific activity (units/nmol) was calculated from the maximum velocity and enzyme concentration data. Extracts of rat liver mitochondria and of skeletal muscle cytosol were also analyzed for P-enolpyruvate carboxykinase by the radioimmunoassay. In "chemical" diabetes, P-enolpyruvate carboxykinase by the radioimmunoassay. In "chemical" diabetes, P-enolpyruvate carboxykinase concentration increased approximately 3-fold over the fed, normal liver value of 0.89 microM, approximately 1.6-fold over the normal kidney value of 1.9 microM and approximately 2.9-fold over the normal adipose tissue value of 0.030 microM. Chemical diabetes caused the specific activity to decrease from 0.38 to approximately 0.27 units/nmol in liver and from 0.48 to approximately 0.32 units/nmol in kidney. Insulin replacement by in vivo injection not only promptly lowered the abnormally high enzyme concentration in both tissues but, paradoxically, decreased the apparent specific activities further to approximately 0.16 in liver and to 0.23 in kidney. In the db+/db+ diabetic mouse the liver enzyme increased from 0.22 microM at 5 weeks of age to 0.44 microM at 18 weeks when serum glucagon concentration is known to be highest and the pancreatic beta cells to be depleted of insulin. While the enzyme protein concentration increased 2-fold in the 18-week-old db+/db+ mouse, total liver enzyme content had actually increased 5-fold due to liver enlargement. In fasted normal mice, glucagon-treated normal mice, and alloxan/streptozotocin-treated mice, the concentration of liver enzyme increased significantly compared to the values in fed control mice. Pharmacological doses of dexamethasone did not induce the mouse enzyme. Rat liver mitochondria contained only trace quantities of immunoassayable enzyme which can be explained by contamination with cytosolic proteins. Rat skeletal muscle also contained only insignificant quantities of the enzyme or perhaps another cross-reacting, immunoassayable protein. The data obtained in diabetes before and after treatment show that complex mechanisms of control by insulin and glucagon do operate to regulate P-enolpyruvate carboxykinase in liver, kidney, and adipose tissue.

The development and application of a radioimmunoassay for rat phosphoenolpyruvate carboxykinase.

We employed a newly developed radioimmunoassay to quantitate P-enolpyruvate carboxykinase protein directly in liver and kidney of intact rats. The radioimmunoassay was dependent upon the presence of sodium dodecyl sulfate in the competitive binding reaction mixture. In conjunction with assayable activity measurements, the radioimmunoassay results also made it possible to assess the average specific activity of the enzyme. In the fed state, liver P-enolpyruvate carboxykinase was 0.89 microM (micromoles/kg of tissue) and the kidney enzyme was 1.90 microM. Following a 48-h fast, the enzyme in liver increased to 2.83 microM and that in kidney to 6.83 microM. Although liver enzyme concentration increased 3-fold, total liver enzymic activity was increased only about 30%. The discrepancy was due to the decrease in liver weight which occurred during fasting. The kidneys, which do not lose weight during fasting, had a 65% elevation of total organ activity. Total organ enzyme activity was predominantly dependent on changes in enzyme mass. In the fed state, the kidney/liver enzyme mass ratio was 0.44, in the fasted state it was 0.68. Variation in enzyme concentration, enzyme half-life, and response to tryptophan and chronic triamcinolone administration all gave evidence for organ-specific regulation.

Insulin and dexamethasone effects on liver fructose bisphosphatase.

Insuling perfusion of isolated diabetic rabbit liver and glucocorticoid perfusion of isolated normal rabbit liver resulted in an increased accumulation of immunoreactive fructose bisphosphatase (nmol/g). The insulin effect was maximal at 30 min with concentration (nmol/g) and specific activity (units/nmol) returning to normal within 2 h after insulin removal. Similarly, dexamethasone perfusion increased enzyme concentration and decreased its specific activity; however, the effects were maximal at 60-120 min. The results parallel those previously demonstrated in whole animals and establish that both hormones influence the enzyme by direct effects on the liver.

Nuclear envelope of the seminal-vesicle epithelium.

The nuclear envelope of seminal-vesicle epithelium was isolated by a procedure involving enzymic digestion with deoxyribonuclease I, sonication in the presence of 0.34 M-sodium citrate, and centrifugation through sucrose density gradients. The mass of envelope DNA was only 0.8% of that of envelope protein, and by transmission electron microscopy the envelope was 98-99% pure. We showed that the envelope possess a protein kinase activity which is uninfluenced by cyclic nucleotides. Both lysine-rich histone and dephosphophosvitin as substrates gave a greater specific activity than did envelope protein itself. Optimum requirements with respect to Na+, Mg2+, pH and ATP were established for each substrate, and the influence of other factors on enzyme activity was investigated. Data, obtained mainly with the use of lysine-rich histone, are presented which indicate that nuclear envelope from intact and 96 h-castrated guinea pigs may have equal protein kinase activities and, in separate experiments, equal phosphoprotein phosphatase activities. Clarification of these initial observations must await identification of the natural substrates or the envelope's phosphorylation-dephosphorylation reactions.

Dexamethasone effects on liver pyruvate kinase.

Dexamethasone in the medium perfusing isolated rabbit livers caused a fast-acting and reversible effect on liver pyruvate kinase. The effect was to lower the assayable V activity (units/g tissue) without changing the concentration (nmol/g enzyme protein). In effect, glucocorticoid lowered the specific activity (units/nmol of enzyme) by direct action on liver. The effect on liver pyruvate kinase is mediated by a relatively stable alteration; 30 min after perfusate (with steroid) was replaced by perfusate (without steroid), the effect remained strongly evident.