Role of canine basal cells in postnatal prostatic development, induction of hyperplasia, and sex hormone-stimulated growth; and the ductal origin of carcinoma.

BACKGROUND: The canine prostate has often been proposed as a model for abnormal growth of the human gland. Hyperplasia of the prostate is common in aging men and has been estimated to be present in 100% of old intact dogs. While prostatic carcinoma is common in older men, it appears to be rare in dogs and unlike the disease in humans, it occurs with relatively high frequency in castrated animals. Since basal cells are thought to be key participants in normal and abnormal growth of the human gland, we used immunohistochemistry to investigate the role that they may play in canine prostatic development, the evolution of hyperplasia and carcinoma, and the effects of sex hormones on these cells. METHODS: Prostate specimens were obtained at autopsy from seven sexually immature dogs, autopsy and biopsy samples from 14 sexually mature intact animals, from four castrates, and from19 dogs with prostatic carcinoma. In addition, we also studied the prostates from two intact dogs treated with 5alpha-dihydrotestosterone (DHT) for 6 months and two castrated dogs that were subsequently treated with 5alpha-androstane-3alpha diol and estradiol-17alpha, as well as specimens from two sexually ablated animals given DHT for 2 weeks. All specimens were immunostained for high molecular weight cytokeratin (HMC), pancytokeratin, androgen receptor (AR), and the proliferative marker KI-67. RESULTS: We find that basal cells are the major proliferative cell type in the neonatal and adult canine prostate and that the expression of HMC staining, which defines these cells, may be regulated by androgens. In the adult gland, ductal basal cells formed a contiguous layer, whereas those lining acini were discontinuous. Populations of both basal cell types were variably AR positive, but while HMC immunostaining was abolished in acinar cells following long-term castration, staining remained in ductal cell counterparts. Paralleling the histological development of hyperplasia, the acinar basal cell population increased with age and were the major cell type that expressed KI-67. In contrast, ductal basal cell populations did not expand in the prostates of older dogs and were seldom positively stained for KI-67. The numbers of HMC and KI-67-stained acinar basal cells were dramatically increased in the prostates of intact dogs treated with DHT when compared with glands of untreated controls. This was not the case with ductal basal cells. Androgens given alone or together with estrogen to castrated dogs induced widespread HMC and KI-67 immunostaining in both populations of basal cells. In addition, our results indicate that the majority of canine prostatic carcinomas likely arise exclusively from ductal epithelium. Only one of the 19 cases of carcinoma contained cells that expressed AR, which suggests that androgens may not be required for the initiation or progression of these cancers. CONCLUSIONS: Our findings indicate that two biologically distinct populations of basal cells may exist in the canine prostate. In this regard, the age-related expansion of proliferating acinar basal cell populations, probably mediated by sex steroids, is a key factor in the pathogenesis of canine prostatic hyperplasia. Additionally, we find that prostatic carcinoma in the dog likely arises from ductal cells. Taken together, these findings may indicate that canine acinar basal cells and ductal epithelium have separate susceptibilities to factors that promote hyperplastic or neoplastic development.

Role of canine basal cells in prostatic post natal development, induction of hyperplasia, sex hormone-stimulated growth; and the ductal origin of carcinoma.

BACKGROUND: The canine prostate has often been proposed as a model for abnormal growth of the human gland. Hyperplasia of the prostate is common in aging men and has been estimated to be present in 100% of old intact dogs. While prostatic carcinoma is common in older men it appears to be rare in dogs and unlike the disease in humans it occurs with relatively high frequency in castrated animals. Since basal cells are thought to be key participants in normal and abnormal growth of the human gland, we used immunohistochemistry to investigate the role that they may play in canine prostatic development, the evolution of hyperplasia and carcinoma, and the effects of sex hormones on these cells. METHODS: Prostate specimens were obtained at autopsy from seven sexually immature dogs, autopsy and biopsy samples from 14 sexually mature intact animals, from four castrates, and from 19 dogs with prostatic carcinoma. In addition, we also studied the prostates from two intact dogs treated with 5 alpha-dihydrotestosterone (DHT) for 6 months and two castrated dogs that were subsequently treated with 5 alpha-androstane-3 alpha diol and estradiol-17 alpha as well as specimens from two sexually ablated animals given DHT for 2 weeks. All specimens were immunostained for high molecular weight cytokeratin (HMC), Pancytokeratin, androgen receptor (AR), and the proliferative marker KI-67. RESULTS: We find that basal cells are the major proliferative cell type in the neonatal and adult canine prostate and that the expression of HMC staining, which defines these cells, may be regulated by androgens. In the adult gland, ductal basal cells formed a contiguous layer whereas those lining acini were discontinuous. Populations of both basal cell types were variably AR positive but while HMC immunostaining was abolished in acinar cells following long-term castration, staining remained in ductal cell counterparts. Paralleling the histological development of hyperplasia, the acinar basal cell population increased with age and were the major cell type that expressed KI-67. In contrast, ductal basal cell populations did not expand in the prostates of older dogs and were seldom positively stained for KI-67. The numbers of HMC and KI-67-stained acinar basal cells were dramatically increased in the prostates of intact dogs treated with DHT when compared with glands of untreated controls. This was not the case with ductal basal cells. Androgens given alone or together with estrogen to castrated dogs induced widespread HMC and KI-67 immunostaining in both populations of basal cells. In addition, our results indicate that the majority of canine prostatic carcinomas likely arise exclusively from ductal epithelium. Only one of the 19 cases of carcinoma contained cells that expressed AR which suggests that androgens may not be required for the initiation or progression of these cancers. CONCLUSIONS: Our findings indicate that two biologically distinct populations of basal cells may exist in the canine prostate. In this regard the age-related expansion of proliferating acinar basal cell populations, probably mediated by sex steroids, is a key factor in the pathogenesis of canine prostatic hyperplasia. Additionally we find that prostatic carcinoma in the dog likely arises from ductal cells. Taken together these findings may indicate that canine acinar basal cells and ductal epithelium have separate susceptibilities to factors that promote hyperplastic or neoplastic development. Prostate 47:149-163, 2001.

Prolactin receptor expression in the developing human prostate and in hyperplastic, dysplastic, and neoplastic lesions.

In situ hybridization and immunohistochemistry were used to localize and compare the expression of the long form of the human prolactin receptor in fetal, prepubertal, and adult prostate. Results were then compared with hyperplastic, dysplastic, and neoplastic lesions. Both receptor message and protein were predominately localized in epithelial cells of the fetal, neonatal, prepubertal, and normal adult prostate. In hyperplastic lesions the expression of the receptor was unchanged with respect to normal epithelial cells. Irrespective of grade, markedly enhanced expression of the receptor was evident in dysplastic lesions. In lower Gleason grade carcinomas the intensity of receptor signal at the message and protein levels approximated that found in normal prostatic epithelium. However, in foci within higher grade cancers, receptor expression appeared diminished. Results from our study suggest that prolactin action plays a role in the development and maintenance of the human prostate and may also participate in early neoplastic transformation of the gland. Diminution of receptor expression in high grade neoplasms could reflect the emergence of a population of cells that are no longer responsive to the peptide hormone.

Androgen receptor expression in prostatic dysplasia (prostatic intraepithelial neoplasia) in the human prostate: an immunohistochemical and in situ hybridization study.

To evaluate the role of androgens in the pathogenesis of prostatic dysplasia, we compared the localization of androgen receptor (AR) in proliferative and nonproliferative cells in normal and dysplastic acini. Basal cells, the only proliferating cells identified in normal acini, contained AR mRNA but lacked an immunodetectable receptor. Both AR mRNA and immunodetectable receptor were present, however, in secretory and stromal cells. Androgen receptor localization in dysplastic lesions was identical to normal but here the proliferative marker Ki-67 was found in both basal and secretory cells. Our findings suggest that androgens do not directly initiate the division of basal cells, the putative precursors of secretory cells. Instead, the hormone may act through its fully translated receptor to mainly mediate the differentiation of secretory cells. The presence of both AR and Ki-67 in dysplastic secretory cells may indicate an abnormal direct androgen-mediated proliferation in this compartment. This is consistent with previous evidence that secretory cell differentiation is impaired in dysplasia.

Induction of atypical hyperplasia, apoptosis, and type II estrogen-binding sites in the ventral prostates of Noble rats treated with testosterone and pharmacologic doses of estradiol-17 beta.

BACKGROUND: We have previously shown that combined administration of testosterone (T) and a low dose of estradiol 17 beta (T+LDE2) for 16 weeks induces an atypical proliferative lesion, termed dysplasia, in the dorsolateral prostates of intact Noble rats (1, 2). The lesion was accompanied by increases in the levels of a moderate affinity, high capacity, estrogen-binding site (type II sites) found exclusively in dorsolateral prostates of these animals (1, 3). In contrast, a proliferative response and type II sites were not observed in the ventral prostates (VP) of the same rats treated with this hormonal regimen. In the current study, rats were treated with a higher dose of E2 (4 x LDE2) but the same dose of T (T+HDE2) for 16 weeks. Our aims were to determine how the VP would respond to the T+HDE2 treatment. EXPERIMENTAL DESIGN: Intact Noble rats were treated with T+HDE2 for 16 weeks. Prostatic tissues were removed for histology, electronmicroscopy, and type II site measurements. Proliferating cells were identified by the histochemical detection of proliferating cell nuclear antigen and colcemid-arrested mitotic figures. Apoptotic cells were recognized by their characteristic histologic and ultrastructural features and by in situ detection of nuclear DNA fragmentation. Data were compared with results previously obtained from VP of rats treated with T+LDE2. RESULTS: The VP of T+HDE2-treated animals contained focal atypical hyperplasia and wide-spread apoptosis. Proliferating cell nuclear Ag-positive-stained epithelial cells and mitotic figures were only present in foci of atypical hyperplasia. Total DNA content of the VP was significantly increased, but the tissue wet weight was not augmented. Nuclear type II sites, never observed in untreated or T+LDE2-treated rats, were detected in the VP of the majority of T+HDE2-treated animals. CONCLUSIONS: The administration of a high dose of E2 with T produced a unique lesion in the VP, characterized by simultaneous occurrence of apoptosis and proliferation. The synergy between androgens and estrogens, via type II site induction, likely produces the proliferative response. On the other hand, inhibition of intracellular androgen activation pathways, leading to reduction in cell survival factors, may be the cause for the apoptotic development. Our model, thus, provides a unique opportunity to further study the balance/switch between cell proliferation and apoptosis that is often disturbed during cancer development.

Early alterations in ras protooncogene mRNA expression in testosterone and estradiol-17 beta induced prostatic dysplasia of noble rats.

BACKGROUND: The simultaneous treatment of intact Noble rats with testosterone and estradiol-17 beta for 16 weeks consistently induces intraductal dysplasia exclusively in the dorsolateral lobe (DLP) of the prostate. The lesion closely resembles human prostatic dysplasia and is considered to be a preneoplastic alteration, since invasive carcinoma frequently develop after long-term treatment of rats with both steroids. In our current study, we investigated steady-state ras transcript expression at the earliest recognized stages of sex steroid-induced dysplasia in the DLP. Our interest in studying ras expression in these evolving lesions stems from the pivotal role this family of genes are thought to play in the regulation of cell division and differentiation as well as in the genesis of a variety of human and animal neoplasms. EXPERIMENTAL DESIGN: Northern blotting and in situ hybridization were used to study ras protooncogene mRNA expression in the DLPs of NBL rats harboring sex steroid-induced ductal dysplasia and to compare findings with those from prostates of castrated and castrated androgen-treated animals. Since the prostate is an androgen-dependent gland, alterations in ras expression were compared with changes in the transcript levels of two androgen-responsive genes that encode for a prostatic secretory protein, seminal vesicle secretion protein II, and the androgen receptor. RESULTS: Similar to the situation for androgen receptor expression, orchiectomy initially enhanced levels of both H- and K-ras transcripts, whereas T administration to castrates was found to return the values to levels found in intact rats. Sixteen weeks of T and E2 administration to intact rats caused levels of H-ras mRNA and a 2.4 kb K-ras transcript to rise by 50 and 60%, respectively in the DLPs with dysplasia when compared with counterpart lobes from untreated control animals. In situ hybridization revealed markedly enhanced H-ras expression in some dysplastic DLP foci and no changes in histologically normal ducts and acini. CONCLUSIONS: Taken together, results from our studies suggest that the enhanced focal expression of ras protooncogenes may participate in early aberrant proliferation of prostatic ductal cells of the DLP. Early alterations of ras expression in dysplastic lesions may therefore be a key contributing event in the multistage development of prostate cancer in this animal model.

Immunohistochemical and in situ hybridization studies of androgen receptor expression in a transplantable androgen-independent prostatic carcinoma line (AIT) of Noble rats.

BACKGROUND: In a previous study, we had shown that testosterone (T) administration to castrated Noble (NBL) rats, bearing an androgen-independent prostatic carcinoma line (AIT), caused a dramatic increase in high-affinity nuclear-androgen binding sites in the neoplasms. This increase in nuclear androgen receptors (AR) was accompanied by a transient doubling of the mitotic index in the tumors. EXPERIMENTAL DESIGN: In our current study we used immunohistochemistry and in situ hybridization to investigate the effects of androgen withdrawal and replacement on AR expression at the molecular/cellular level in the AIT. Results from immunohistochemical studies of AR expression in the AIT were compared with those from the prostates of intact, castrated, and castrated T-treated rats. RESULTS: Immunopositive staining for AR was found only in the nuclei of prostatic cells from the glands of intact, castrated or castrated T-treated animals. A few immunopositive tumor cells were present in AITs carried in untreated castrated hosts. In all instances, the reaction product was found in the cytoplasm, but it was also present in the nuclei of some tumor cells. Five days of T administration to castrated AIT-bearing rats caused a dramatic increase in immunopositive tumor cells. Nuclear staining was observed in all positive cells, but the reaction product was also present as well in the cytoplasm of some tumor cells. No differences in AR mRNA expression was detected by in situ hybridization studies of the AITs from castrated and castrated T-treated rats. CONCLUSIONS: The differences in localization of AR between normal prostate and carcinoma cells may reflect alterations in DNA binding domains of the AR protein that occurred with neoplastic transformation. Our in situ findings suggest that unlike the normal prostate, where AR mRNA levels are autoregulated by androgen, AIT cells constitutively express these transcripts. Taken together, our findings suggest that the T-mediated increases in nuclear AR in the AIT, detected previously by binding assay and now by immunohistochemistry, are likely the result of post-transcriptional modifications in the receptor protein.

Comparison of 35S- and digoxigenin-labeled RNA and oligonucleotide probes for in situ hybridization. Expression of mRNA of the seminal vesicle secretion protein II and androgen receptor genes in the rat prostate.

The sensitivity of radiolabeled and digoxigenin-labeled RNA probes and synthetic oligonucleotide probes for the detection of seminal vesicle secretion protein II (SVS II) and androgen receptor (AR) mRNA was compared by in situ hybridization in paraformaldehyde-fixed cryostat sections of the rat prostate. Both genes are expressed in different amounts in the various prostatic lobes and contiguous glands. SVS II or AR RNA probes were either labeled with digoxigenin-11-UTP or [35S]UTP by in vitro transcription. A synthetic SVS II oligonucleotide probe was 3' end-labeled (tailed) with either digoxigenin-11-dUTP or [35S]dATP. Hybridized 35S-labeled probes were detected by autoradiography and digoxigenin-labeled probes by immunohistochemistry using alkaline phosphatase conjugated anti-digoxigenin antibody or gold-labeled antibody followed by protein A-gold and silver enhancement. Digoxigenin-labeled probes provided the same degree of sensitivity as their 35S-labeled counterparts for the detection by in situ hybridization of weakly and strongly expressed mRNA. Using both labeling methods, the SVS II RNA probes were more sensitive than the oligonucleotide probes and background labelling of the 35S-labeled oligonucleotide probe was high. The digoxigenin method produced less background with all probe types, hybridization signals showed higher resolution and results were obtained faster than with radiolabeled probes. The immunogold silver enhancement system provided the fastest detection of digoxigenin-labeled probes with a sensitivity and resolution similar to that provided by alkaline phosphatase anti-digoxigenin immunohistochemistry. It is concluded that digoxigenin probe labeling and detection provides a sensitive, reliable, and efficient alternative to radiolabeled probes for in situ hybridization of mRNA.

Androgen-supported estrogen-enhanced epithelial proliferation in the prostates of intact Noble rats.

Using a stathmokinetic in vivo metaphase-arrest technique, we studied cell proliferation and histological changes in the ventral (VP) and dorsolateral (DLP) prostate lobes of intact Noble (Nb) rats following a 16 week treatment with testosterone (T) or 5 alpha-dihydrotestosterone (DHT) administered separately or in combination with various estrogens. The combined treatment of rats with T and either estradiol-17 beta, estradiol-17 alpha, or moxestrol induced florid dysplasia and markedly elevated the mitotic index (MI) in affected regions of the DLP. In contrast, joint DHT and estrogen treatment caused only mild proliferative lesions in this lobe. The separate administration of either androgens or estrogens suppressed epithelial proliferation in both the VP and DLP, but they differed in their histological effects on these tissues. Thus DHT or T alone maintained the morphological integrity of VP and DLP, whereas E2-17 beta or moxestrol caused massive atrophy of both lobes. Although dysplastic foci were randomly scattered throughout the DLP, the most dramatic lesions occurred in periurethral ducts. With the exception of joint T and E2-17 alpha treatment, which induced proliferative alterations in the VP, dysplasia was always restricted to the DLP of all animals receiving both androgens and estrogens. Concomitant comparative stathmokinetic studies of the prostates of T-treated castrates suggest that protracted androgen-supported estrogen stimulation of the DLP is necessary to overcome factors that normally limit cell proliferation.

Biochemical alterations in sex hormone-induced hyperplasia and dysplasia of the dorsolateral prostates of Noble rats.

Simultaneous implantation of intact Noble (Nb) rats with testosterone and 17 beta-estradiol (E2)-filled silastic capsules for 16 weeks caused atypical hyperplasia (dysplasia) and striking enlargement exclusively in the dorsolateral prostates (DLPs) of all animals. The dysplastic lesion may be preneoplastic since long-term administration of these steroids to Nb rats is known to induce a high incidence of adenocarcinoma in the DLP. Treatment of rats with nonaromatizable 5 alpha-dihydrotestosterone (DHT) for 16 weeks caused enlargement but not dysplasia, implicating estrogen as a key factor in the genesis of the proliferative lesion. Compared with controls, the testosterone plus E2 treatment caused a 2.5-fold increase in nuclear type II estrogen binding sites which were confined to the DLP. Neither treatment significantly altered androgen content or levels of androgen receptor in the ventral prostate or DLP. Organ cultures of enlarged DLP containing foci of dysplasia metabolized more [3H]DHT than control tissue, which resulted in increased formation of the 5 alpha-androstane-3 beta, 17 beta-diol (3 beta-androstanediol) metabolite by these explants. Because 3 beta-androstanediol has previously been shown to displace [3H]E2 from cytosolic type I estrogen binding sites, the dysplasia may be caused by hyperstimulation of the DLP by the hormones and their normal metabolites produced in abnormal amounts.

Testosterone-mediated increase in 5 alpha-dihydrotestosterone content, nuclear androgen receptor levels, and cell division in an androgen-independent prostate carcinoma of Noble rats.

An androgen-independent, transplantable prostate carcinoma line (AIT), originally derived from the dorsolateral prostate (DLP) of Noble rat, was implanted into orchiectomized Noble rats and its response to androgen stimulation was studied and compared to that of the regenerating DLP tissue in sexually ablated rats. AIT tumors carried in castrated hosts displayed a high basal level of proliferative activity (mitotic index (MI), 15.0 +/- 0.5) while DLP tissue in untreated castrates exhibited no proliferative activity. Following androgen stimulation by testosterone capsule implantation into host rats, the AIT responded with a marked increase in cell proliferation; MI values doubled to 30.0 +/- 2.9 on Day 5 following androgen stimulation. This androgen-induced increase in MI values was coincident with elevations in nuclear androgen receptor (20-fold increase) and 5 alpha-dihydrotestosterone content (3-fold increase) in the tumor. However, by Day 10 following androgen treatment, indices of cell proliferation in the AIT declined to pre-androgen-stimulated levels (MI, 14.8 +/- 1.9) despite the continued elevations in nuclear androgen receptor and tissue 5 alpha-dihydrotestosterone contents. Parallel changes in MI were also observed in the normal regenerating DLP following androgen stimulation. MI values in this tissue increased from nondetectable levels to 38.1 +/- 4.7 on Day 5 but declined to relatively low levels (4.5 +/- 0.9) by Day 10 following androgen replacement. Taken together these findings led us to conclude that the AIT carried in castrates is capable of responding to testosterone in a manner similar to that observed for androgen-stimulated DLP of sexually ablated rats. Thus, in both the neoplastic and regenerating tissues, the initial response to androgen is characterized by a marked enhancement of cell proliferation which was correlated with an increase in androgen receptor and 5 alpha-dihydrotestosterone content. However, like its tissue of origin, the AIT possesses mechanisms which act to limit androgen-induced cell division despite continued elevations in key parameters of androgen activation.

Multiple phenotypes of prostatic glandular cells in castrated dogs after individual or combined treatment with androgen and estrogen. Morphometric, ultrastructural, and cytochemical distinctions.

To demonstrate a potential for multidirectional differentiation in mature prostatic epithelium, 17 beta-estradiol 17-cyclopentylpropionate (ECP) and 5 alpha-androstane-3 alpha, 17 beta-diol dipropionate (3 alpha-diol DP) were administered individually and in combination to castrated dogs. Quantitative ultrastructural and cytochemical methods were used to distinguish phenotypes of glandular cells in the various hormonal environments. Castration-induced glandular cell regression was accompanied by an increased nuclear to cytoplasmic ratio; by enhanced keratin positivity, expressed as dispersed immunolabeled tonofilaments; and by an absence of peanut agglutinin (PNA) binding sites on luminal membranes. Administration of ECP resulted in squamous metaplasia as well as hypertrophy of the glandular epithelium. The hypertrophied estrogen-modified glandular (EMG) cells were characterized by a new population of small (0.29 micron in diameter) secretory granules, bundles of tonofilaments, and PNA-positive luminal membranes. Treatment of castrated dogs with 3 alpha-diol DP produced a greater epithelial hypertrophy than ECP. These cells were characterized by larger (0.49 micron in diameter) secretory granules, dispersed tonofilaments, and no detectable PNA receptors. Joint administration of ECP and 3 alpha-diol DP caused a florid response including squamous metaplasia and hypertrophy of the glandular epithelium which was associated with the emergence of a novel phenotype in androgen-estrogen modified glandular (A-EMG) cells. In A-EMG cells, secretory granules were similar in size to those found in 3 alpha-diol DP-dominated epithelium whereas tonofilaments often appeared in bundles and luminal membranes were PNA positive, i.e., features found in EMG cells. Our results indicate that atrophic canine prostatic glandular cells possess pluripotentiality of response to sex hormones.

Characterization of a human, sex steroid-responsive transitional cell carcinoma maintained as a tumor line (R198) in athymic nude mice.

We have established a transplantable tumor line, R198, derived from a papillary (transitional cell) carcinoma of the human urinary bladder. In nude mice, the tumor line exhibits properties attributable to both prostatic and transitional epithelia. In tumor-bearing animals given androgens, the neoplasm has a rapid growth rate, possesses low levels of measurable androgen receptors, produces tartrate-inhibitable acid phosphatase, and forms well-encapsulated, cystic tumors composed of transitional, glandular, and squamous cells. The administration of estrogens or transplantation of the tumor into female mice causes regression of the tumor. In a small percentage of the transplants placed into females or estrogenized animals, selection occurs for tumor cells which can grow under these conditions. The resulting tumors are infiltrating scirrhous carcinomas that closely resemble squamous cell carcinomas of the urinary bladder. These neoplasms grow slowly and do not possess androgen receptors or secretory material. They are composed of a homogeneous population of squamous cells which are locally invasive. The paradox of a bladder tumor with some prostatic characteristics may be explained by the fact that the tumor was derived from the trigone region of the bladder, which embryologically is formed by an admixture of tissue from the wolffian duct and the urogenital sinus. Some trigone-derived neoplasms have characteristics of both bladder and prostate. We hypothesize that sex steroid-sensitive R198, with characteristics of both bladder transitional cells and prostatic epithelia, is a tumor which phenotypically expresses the embryological origins of these tissues. As such, the tumor line will serve as a useful model for studying sex steroid-responsive cells of the urogenital epithelium.

Cell surface modifications in the epithelium of rat ventral prostate during adaptation to in vitro conditions: an ultrastructural study.

Sequential changes in epithelial cells of collagenase-dissociated rat ventral prostate were studied by thin-section and freeze-fracture electron microscopy. Epithelial cells did not attach to the substrate for 48 h. Pelleted cells obtained 1, 24, and 48 h after dissociation were assigned to three categories depending on morphology and cellular associations. (a) Solitary epithelial cells degenerated as determined by extensive vacuolization in the cytoplasm and aggregation of intramembranous particles (IMP). (b) Epithelial clusters consisted of a homogeneous population of well-maintained, closely packed cells. Aggregation of IMP was minimal. Tight junctions that formed between cells at the periphery of the clusters appeared normal and provided an effective permeability barrier demonstrated by the exclusion of ruthenium red tracer. (c) Tissue fragments were comprised of varying combinations of epithelial, endothelial, and smooth muscle cells as well as fibroblasts and erythrocytes. Maintenance of tissue fragments was variable. Plasma membranes often displayed aggregated IMP and proliferated tight junctional strands. An effective permeability barrier was absent. After the 48 h "latent period," epithelial cells in the clusters lost interdependence, disassociated from one another, and attached to the substrate. These isolated cells, which did not display aggregated IMP, retained the ability to form an effective permeability barrier upon reaching confluency. During the first 48 h, epithelial cells did not tolerate solitary existence, yet as participants in clusters they were well maintained. After this interval, they no longer required interactions with neighbors in order to survive. These results indicate that under our experimental conditions, an adaptation period is required by prostatic epithelial cells. The enhanced quality of maintenance associated with epithelial clusters suggests that control over the internal microenvironment, provided by a tight junctional barrier, may be important during the initial period of adaptation in vitro.

Histochemical studies of epithelial cell glycoconjugates in atrophic, metaplastic, hyperplastic, and neoplastic canine prostate.

The nature and distribution of lectin receptors were studied in normal, atrophic, metaplastic, hyperplastic, and neoplastic epithelium of canine prostate. Results were compared with prostatic epithelium of castrated dogs treated for 2 weeks with estradiol-17 beta 17-cyclopentylpropionate, 5 alpha-adrostane-3 alpha,17 beta-diol dipropionate, or 5 alpha-dihydrotestosterone. Eight biotinylated lectins were used as histochemical probes and avidin-biotin-peroxidase complex served as the visualant. Receptors for Ulex europaeus agglutinin-I were present in atrophic prostatic epithelium. Receptors for U. europaeus agglutinin-I, wheat, germ agglutinin, Dolichos biflorus agglutinin, and soybean agglutinin were present in epithelium that had undergone squamous metaplasia. Binding of peanut agglutinin receptors was present, to a limited extent, in squamous epithelium and was increased after they were unmasked (sialic acid residues cleaved with neuraminidase). In glandular cells of normal canine prostate and in benign prostatic hyperplasia, receptor sites were stained with Ricinus communis agglutinin-I, Concanavalia ensiformis agglutinin wheat germ agglutinin, and U. europaeus agglutinin-I. The basal cells in these tissues did not bind lectins. Prostatic carcinoma cells demonstrated receptors for wheat germ agglutinin and U. europaeus agglutinin-I. Responding and atrophic acini were present in prostates of castrated dogs treated with estradiol-17 beta 17-cyclopentylpropionate. Glandular cells of atrophic acini exhibited lectin receptor profiles similar to counterparts in castrated-untreated dogs. However, glandular cells responding to estrogen exhibited staining of free and cryptic peanut agglutinin receptor sites. Glandular cells of castrated dogs treated with 5 alpha-androstane 3 alpha,17 beta-diol dipropionate and 5 alpha-dihydrotestosterone have a pattern of lectin receptors similar to that found in normal and hyperplastic epithelium. Our studies show significant differences in lectin-binding patterns in the epithelium of atrophic, metaplastic, hyperplastic, and neoplastic canine prostate. They also demonstrate that the species of carbohydrate residues present in the glandular cells can be modified with sex hormones.

Ultrastructural and biochemical expressions of divergent differentiation in prostates of castrated dogs treated with estrogen and androgen.

To investigate the role of estrogen and androgen in prostatic differentiation and induction of epithelial hyperplasia, we studied ultrastructural and biochemical responses to estradiol-17 beta 17-cyclopentylpropionate (ECP) and 5 alpha-androstane-3 alpha, 17 beta-diol dipropionate (3 alpha-diol DP) in glands of castrated dogs. The hormones were injected individually or in combination. Organ cultures, incubated with 1.7 microM radioisotope-labeled testosterone in serum-free Trowell T8 medium, were used to compare capacities of key transforming enzymes in hormone-modified glands. High-affinity binding of labeled 8.5 nM estradiol-17 beta and 5 alpha-dihydrotestosterone (5 alpha-DHT) to 0.4 M KCl-extractable explant protein was also determined. Treatment with 1 mg. of ECP per week for 2 weeks produced basal cell mitosis and early squamous metaplasia. The glandular epithelium hypertrophied but was not repopulated. When compared with radiotestosterone disposition by explanted prostate from untreated castrates, increased formation and egress of 17-oxo C19O2 steroids, predominantly 4-androstene-3,17-dione, occurred at the expense of 5 alpha-reduced 17 beta-hydroxy C19O2-steroids and hydroxylated metabolites. Administration of 2 x 50 mg. of 3 alpha-diol DP per week for 2 weeks also induced basal cell proliferation. The glandular epithelium was repopulated, and atrophic glandular cells were partially restored. This treatment increased accumulation of radiotestosterone-derived 5 alpha-reduced C19O2-metabolites and C19O3-steroids in the explants. Joint administration of ECP and 3 alpha-diol DP yielded proliferating squamous and glandular cells within the same acinus. Each type of proliferating cell was identified by specific cytologic markers. Chromosomes were observed with tonofilament bundles in squamous cells and with secretory granules in glandular cells. However, most glandular cells were not dividing. They were characterized by co-existing tonofilament bundles and secretory granules. The dual hormone administration increased radiotestosterone metabolism. The separate effect of each hormone was notable since estrogen increased the ratio of 17-oxo C19O2 to 5 alpha-reduced 17 beta-hydroxy C19O2-metabolites, whereas androgen restored both terminal hydroxylations and high-affinity binding of 5 alpha-DHT. The levels of saturable binding of estradiol-17 beta were high but variable in explants of each treatment group. We conclude that estrogen and androgen act cooperatively and synergistically on basal cells of regressed canine prostate to induce divergently differentiated epithelial cells. Together with stromal components, these glandular and squamous cells express distinctive pathways of androgen disposition.

Membrane differentiation in the Golgi apparatus of mammalian urinary bladder epithelium.

The endomembrane system in superficial and intermediate epithelial cells of mammalian urinary bladder was studied by cytochemistry, thin-section and freeze-fracture electron microscopy to determine the sites where special forms of membrane differentiation first appear. Glutaraldehyde-resistant NADH-ferricyanide reductase, distinctive 11-12 nm intramembrane particles (IMP), and asymmetry of membrane leaflets served as markers of membrane maturation. The three markers were specifically associated with the maturing face of Golgi apparatus and were absent from the remainder of the endomembrane system. Activity of this enzyme was associated with the lateral regions of the maturing face, fusiform vesicles, and the plasmalemma. Asymmetric unit membrane (AUM) plaques were not observed in the Golgi apparatus per se but were present in immature fusiform vesicles that had not detached from the maturing face. When freeze-fracture replicas and thin sections were compared, randomly arranged 11-12 nm IMP first appeared in maturing face membranes that were adjacent to clusters of "free" polyribosomes in the Golgi apparatus region. The proximity of these polyribosomes suggests that they may be related to the coincident appearance of the 11-12 nm IMP in the maturing face membrane. Our observations support the hypothesis that membranes undergo differentiation during "flow" through compartments of the endomembrane system. The lateral regions of the maturing face of the Golgi apparatus appear to be a critical location for the morphogenesis of plasma membranes in urinary bladder.

Ultrastructural changes in surface topography, glycocalyx and plasma membrane interior of tumor cells during exocytosis of mucus.

Plasma membrane changes, which occur during exocytosis, were studied by electron microscopy in six specimens of primary mucous-producing adenocarcinoma of human urinary bladder. The luminal surfaces of non-secretory neoplastic cells display numerous microvilli, which have a thick ruthenium red-positive glycocalyx. Neoplastic secretory cells have a smooth luminal surface with relatively few microvilli. The glycocalyces of these cells are thin, especially at locations where the plasmalemma and underlying mucous granule membrane are in close apposition. The fused membranes often bulge into the lumen forming distinct protuberances. In freeze-fracture replicas the protuberances appear devoid of 7-8 nm intramembrane particles. Our results suggest that there are differences between non-secretory and secretory neoplastic cells that are expressed in surface topography, glycocalyx and internal membrane structure. Secretion-induced changes observed in plasma membranes of the neoplastic cells closely resemble changes that are known to occur in the plasmalemma of normal secretory cells.

Estrogen-mediated exocytosis in the glandular epithelium of prostates in castrated and hypophysectomized dogs.

Glandular cells in the prostate of the intact, adult dog contain numerous, large secretory granules that are released by exocytosis. Following hypophysectomy or castration, the glandular epithelium atrophies and the secretory granules degenerate and eventually disappear. Pharmacologic doses of estradiol-17 beta 17-cyclopentylpropionate cause the regressed glandular cells to synthesize a new population of smaller granules that are also released by exocytosis, even though estrogen is known to inhibit fluid secretion by the canine prostate. Thus, the mechanisms involved in prostatic synthesis and exocytosis of secretory granules are independent of those regulating fluid secretion and are operative in the absence of androgen or pituitary hormones.

Heterogeneous distribution of filipin-sterol complexes in nuclear membranes.

Filipin, a sterol-specific polyene antibiotic, has been shown by electron microscopy to form complexes in membranes of mouse urinary bladder cells. Following instillation of a glutaraldehyde-filipin-dimethylsulfoxide solution into the bladder lumen, filipin-cholesterol complexes appear as membrane corrugations in thin sections and as 20-25 nm protuberances and depressions on PF and EF faces in freeze-fracture replicas. The complexes are observed in plasmalemma, Golgi membrane, rough endoplasmic reticulum and nuclear membrane of five different cell types (urothelial, endothelial, mesothelial, smooth muscle and fibroblasts). In the present report, we direct particular attention to the localization of numerous filipin-cholesterol complexes present in the nuclear envelopes of these cells. Our results suggest that enrichment of cell membranes with cholesterol occurs at an earlier stage in the flow-differentiation process than previously suspected. In addition, the unequal distribution of complexes in favor of the outer nuclear membrane suggests that it has a higher cholesterol content than the inner membrane.