EPN: A NOVEL EPITHELIAL CELL LINE DERIVED FROM HUMAN PROSTATE TISSUE

In Vitro Cell. Dev. Biol.~Anima138:165-172, March 2002 �9 2002 Society for In Vitro Biology 1071-2690/01 $10.00+0.00

EPN: A NOVEL EPITHELIAL CELL LINE DERIVED FROM HUMAN PROSTATE TISSUE

ANTONIO A. SINISI, J PAOLO CHIEFFI, DANIELA PASQUALI, ANNAMARIA KISSLINGER, STEFANIA STAIBANO, ANTONIO BELLASTELLA, AND DONATELLA TRAMONTANO 2

Dipartimento di lnternistica Clinica e Sperimentale, Seconda Universit~t di Napoli, Italy (A. A. S., D. P., A. B.), and Dipartimento di Biologia e Patologia Molecolare e Cellulare "Luigi Califano," and Centro di Endocrinologia ed Oncologia Sperimentale del CNR (P. C., A. K.), Dipartimento di Scienze Biomorfologiche e Funzionali (S. S.), and Facoltdt di Seienze MM.FF.NN, Universitd del Sannio (D. T.),

Benevento, Italy

(Received 13 November 2001; accepted 11 December 2001)

SUMMARY

This work reports the isolation and characterization of a line of human, nontransformed and differentiated prostate epithelial cells (EPN) in continuous culture. Primary cultures of epithelial prostate cells were set up using normal tissue isolated from a prostate sample collected after radical prostatectomy for cancer. After 70 passages, EPN cells did not undergo "Hayflikc crisis" and were free of fibroblast contamination and were thus subcloned and characterized. EPN cells in culture, as prostate epithelial ceils in vivo, express high-molecular weight cytokeratin and Pyk2, whereas they do not express desmin. EPN cells are nontransformed because they do not form colonies in semisolid medium and do not form tumors once injected into nude mice. EPN cells express the functional androgen receptor, which can mediate the mitogenic activity of testosterone. Finally, clonal production of the prostate-specific antigen could be detected in EPN cells. The availability of a line of epithelial nontransformed prostate cell in culture will be useful in investigating the complex process regulating normal prostate physiology as well as the development and progression of prostate tumors.

Key words: normal prostate tissue-differentiated prostate markers; cell line; cultures in continuous culture.

INTRODUCTION

Prostate cancer is the most frequent malignancy in the male pop- ulation of the western countries and the second cause of death in men after cardiovascular diseases. In spite of considerable efforts in the search for the appropriate therapy for prostate cancer, the appearance within the tumor of the androgen-insensitive cells oblit- erates the good result obtained by the hormonal therapy. In fact, prostate cancer begins as an androgen-dependent tumor that un- dergoes clinical regression in response to pharmacological or sur- gical strategies that are able to reduce testosterone concentration. Despite these treatments, the cancer in most cases regrows as an androgen- or hormone-independent tumor (Talpin et al., 1995; Vi- sakorpi et al., 1995; Craft et al., 1999).

Development and progression of prostate cancer is a multistep process involving genetic damage to the epithelial cells, such as the activation of oncogenes and the inactivation of onco-suppressor genes, and also changes in the relation between stroma-epithelial- neuroendocrine components both at the cellular level and at the signal transduction level. In this view, hormone independence, which plays a pivotal role in the progression of prostate cancer, is the final resuh and possibly the "no return" point of such a mul- tistep process (Wong and Wang, 2000). Several hypotheses have

1 To whom con'espondence should be addressed (AAS) at Endocrinologia Seconda Universith di Napoli, Via Pansini 5 Napoli 80131, Italy. E-mail: [email protected]

2 To whom correspondence should be addressed (DT) at Facolta' di Scienze MM.FENN., Universith del Sannio, Benevento, Italy. E-mail: tramontano@ unisannio.it

been postulated about the molecular changes causing prostate cell homlone independence (i.e., the androgen receptor itself would me- diate androgen-independent progression); nevertheless, the main picture remains substantially obscure (Talpin et al., 1995; Visakorpi et al., 1995; Craft et al., 1999). It is generally accepted that a major obstacle to advancing our understanding of prostate gland patho- physiology is the lack of an appropriate model system in vitro, al- though several prostate cell lines of the transformed phenotype (Stone et al., 1978; Kaingn et al., 1979; Horoszewicz et al., 1983; Webber et al., 1996; Abate-Shen and Shen, 2000), or variously immortalized by means of viral infection or oncogene transfection (Cussenot et al., 1991; Krill et al., 1997; Webber et al., 1997a, 1997b; Fry et al., 2000; Lang et al., 2000), are currently available, and a variety of isolation and culture techniques have been devel- oped to obtain prostate cells in primary culture (Boudon et al., 1996; Sinisi et al., 1997). Although many aspects of prostate biology and pathophysiology remain poorly understood, it is clear that the functioning of the prostate gland depends on a very complex system of interactions among diverse cell types as well as diverse signals and their pathways. Thus, the perfect model system reproducing in vitro the in vivo situation is not conceivable. An epithelial cell line lacking the stromal and neuroendocrine components, as well as the immune response and vascularization, will represent only a small piece of the complete prostate puzzle. Although far from being a perfect reproduction of the in vivo situation, cells in vitro are unique and irreplaceable tools to approach and answer basic biological questions. Thus, it is important to develop a variety of systems and, once the limitations and the advantages of each are recognized and

165

166 SINISI ET AL.

accepted, to ask a specific question to each appropriate model sys-

tem. With this in mind, we report the isolation and characterization

of a line of human, nontransformed prostate epithelial cells (EPN)

derived from the normal prostate tissue adapted to grow indefinitely

in continuous culture, without undergoing immortalization through

virus or oncogene transformation, which maintained markers of

prostate differentiation.

MATERIALS AND METHODS

Primary cultures. Human prostate tissue was collected from a patient un- dergoing radical prostatectomy for prostate cancer. Because the tumor was very small, at)er the prostatectomy, it was possible to easily isolate the ap- parently normal prostate tissue fi'om the cancer. Two wedge-shaped speci- mens of the fresh prostate were removed from both the normal tissue and the cancer, and they under,vent pathological examination to confirm the prostate origin, the diagnosis, and the absence of other diseases. The apparently nor- real specimen was tbund to be normal from the histopathological standpoint. Primary cultures were established according to the methods previously de- scribed (Cussenot et al., 1991; Krill et al., 1997; Sinisi et al., 1997). Briefly, the prostate tissue was minced into small pieces, and a single-cell suspension was obtained by overnight collagenase digestion (Collagenase IV, GIBCO- BRL, Life Technologies, Paisley, Scotland, U.K.), 10 mg/ml at 37 ~ C. After a series of eentriiugations to eliminate the stromal component, as reported elsewhere (Krill et al., 1997), cells were seeded overnight in Roswell Park Memorial Institute (RPMI) medium supplemented with 10% fetal bovine se- rum to favor cell attachment. The medium was then removed and replaced with keratinocyte-serum free medimn (SFM) (GIBCO-BRL, Milan, Italy) sup- plemented with bovine pituitary extract (10 mg/ml), epidermal growth factor (EGF) (10 ng/ml), cholera toxin (10 ng/ml), 5% fetal calf serum (FCS) (GIB- CO-BRL, Life Technologies, Paisley, Scotland, U.K.), and antibiotics (fungi- zon and penicillin-streptomycin) to avoid fibroblast overgrowth. Once conflu- ence was reached, the cells were subcuhured by the ethylenediaminetetraa- cetic acid (EDTA)-t~psin treatment. All cuhures were performed at 37 ~ C in a humidified 5% CO2 atmosphere.

Af)er 70 passages, EPN cells did not undergo "Hayflikc crisis" (Hayflick and Moorhead, 1961) and were free of fibroblast contamination and were thus subjected to subcloning. One thousand cells were seeded in 100-mm culture dishes, isolated cells were marked, and within a wk colonies started to ap- pear. After 20 d, single colonies were big enough to be individually trypsin- ized with the aid of cloning cylinders, and 10 clones were isolated and grown separately.

For studies on ceil replication, 1 x 10 5 cells were seeded in 60-mm culture dishes in various experimental conditions, as reported in the legend to the figures. At appropriate intervals, triplicate dishes were trypsinized per experimental point, and cell number was determined by counting cell sus- pension in a Neubauer hemocytometer. The values reported represent the mean + standard deviation of three independent samples per experimental point.

In vivo and in vitro analysis of transformed phenotype. Soft Agar colony assay was performed as described by Macpherson and Montagnier (1964). Briefly, EPN cells and NIH-3T3 cells, carrying RET-MEN2A (Santoro et al., 1995), were suspended in a culture medium supplemented with 0.5% noble agar at densities of 104, 2.5 • 104, and 5 • 104. This solution was overlaid onto 7 ml of the culture medium containing 0.5% noble agar in 60-mm dishes. Colony formation was monitored for 3 wk.

The ability of EPN cells to induce tumors was tested according to the standard procedure by using nude mice. All animal studies have been con- ducted in accord with the highest standards of hmnane animal care, as rec- ommended by the international guidelines on the use of experimental ani- mals. Briefly, 10" EPN cells and ARO cells, a cell line derived from highly malignant thyroid tumor (Basolo et ah, 1999), were inoculated subcutaneously into five athymic nude mice. Tmnor occmTence was evaluated 14 d after inoculation and monitored at least once a wk for 4 wk thereafter.

lmmunoprecipitation and Western blot analysis. List of antibodies used were anti-androgen receptor polyclonal antibody raised in rabbit, #sc-815, anti- p53 monoelonal antibody raised in mouse, #sc-263 (Santa Cruz Biotechnol- ogy Inc., Santa Cruz, CA), anti-Pyk2 antibodies and anti-phosphotyrosine raised in rabbit #72, which were kindly provided by Dr. J. Schellissinger and

Dr. I. Dikie (Levet ah, 1995; Tokiwa et al., 1996), and anti-cytokeratin monoclonal antibody raised in mouse, 34BE12 (DAKO Corp., Califbmia).

hnmunopl~cipitation and immunoblotting experiments were performed ac- cording to standard procedures. Briefly, cells were harvested in a lysis bufter (50 nrM N-2-hydroxyethylpiperazine-N'-2-ethanesulionic acid [HEPES], 150 mM NaCI, 1 mM EDTA, 1 nrM ethyleneglycol-bis[aminoethylether]-tetraa- cetic acid, 10% glycerol, 1% Triton-X-100, 1 nrM phenyhnethylsultonyl fluo- ride, 1 mg aprotinin, 0.5 mM sodium orthovanadate, 20 mM sodium pyro- phosphate). After 30 min on ice, the lysates were clarified by 10 rain of centrifugation at 14,000 • g. Protein concentration in the supernatant was estimated by the modified Bradford assay (Bio-Rad, Munich, Germany).

For the immunoprecipitation procedure, 3 mg of protein A-Sephm'ose (Amersham Pharmacia Biotech, Little Chalfimt, UK) was incubated with 5 txl of the given antibodies for 1 h at 4 ~ C and was then washed three times with HNTG (20 mM HEPES, 150 mM NaC1, 0.1% Triton-X-lO0, 5% glycerol). Five hundred micrograms of the total protein lysate was incubated with the antibodies-protein A mix for 2 h at 4 ~ C. Immunopi~cipitates were washed four times in HNTG, and the pellet was boiled in Laemmli's buffer for 5 min before electrophoresis on a 7% sodium dodecyl sulfate-polyacrylanfide gel, under reducing condition. After electrophoresis, proteins were transferred to a nitrocellulose membrane (Immobilion Millipore Corporation, Bedford, MA); complete transfer was assessed using prestained protein standards (Bio-Rad, Melville, NY).

For Western blots, 40 tLg of total protein lysates was separated on a 10% sodium dodecyl sulfate-polyamTlamide gel electrophm~sis. After electix~pho- resis, the proteins were transferred to a nitrocellulose membrane (hmuobilion Mitlipore Corporation); complete transfer was assessed, as described earlier. For immunoblotting, membranes were blocked in Tris-buffered saline (TBS) (10 mM Tris, pH 7.4, 150 mM NaC1), containing 5% bovine serum albumin, for 1 h at room temperature. Incubation with the primary antibody was carried out for 1 h at room temperature followed by three washes in T-TBS tbr 5 min each (10 n~/Tris, pH 7.4, 150 n~/NaCI, Triton 0.05%) and three washes in TBS for 5 min each (10 mM Tris, pH 7.4, 150 mM NaC1). Horse radish peroxides-conjugated secondary antibody (1:2000) was then added for 45 rain followed by three washes in T-TBS for 5 min each and three washes in TBS for 5 min each. The reactions were detected with the enhanced chemi- luminescence (ECL) system (Amersham Life Science, Little Chalfont, UK).

Immunohistochemistry. EPN cells were grown on glass cover slip, washed three times in cold phosphate-buffered saline, and fixed in 10% buft~red formalin. For immunohistochemical studies, the classical avidin-biotin com- plex (ABC) technique was used. The endogenous peroxide was quenched by the incubation of cells in 0.1% sodimn azide with 0.3% hydrogen peroxide for 30 rain at room temperature, whereas the nonspecific bindings were blocked by the incubation with nonimmune serum (1% Tris-bovine albumin for 15 rain at room temperature).

Incubation with the anti-prostate-specific antigen (PSA) antibody (#sc- 7316, mouse monoclonal, Santa Cruz Biotechnolog7 Inc.) at a dilution of 1:100 was performed overnight. As negative control, normal mouse serum was used instead of the primary antibody. The standard ABC procedure was carried out thereafter. The peroxide activity was developed with the use of a filtered solution of 5 nag of 3,3'-diaminobenzidine tetrahydrochloride (dis- solved in 10 ml of 0.05 M Tris buffer, pH 7.6) and 0.03% H202. For nuclear counterstaining, Mayer's hematoxilin was employed, and the slides were mounted with a synthetic medium. Only cells with a brown precipitate in the cytoplasm were considered positive for PSA.

RESULTS

Growth parameters. After ahnost 2 yr, a cell culture derived from

human normal prostate tissue (EPN) maintained an epithelial mor-

phology; cells appear polygonal and rather regular (Fig. 1), com-

pletely devoid of fibroblast contamination. Thus, growth parameters,

tumorigenicity, and expression of prostate-specific functions were

investigated in EPN cells. Dependence of proliferation on serum

addition is an important parameter because it indicates whether the

cells required the presence of the serum supplement in the culture

medium or vice versa, whether they produce enough growth factors

to become serum independent fbr their growth, the latter being a

common trait of cancer cells in culture. The effect of serum con-

EPITHELIAL PROSTATE CELL LINE 167

F~6. 1. Microphotograph of EPN cells in standard growing condition. Magnification: •

centration on the plating efficiency of EPN cells was tested by seed- ing EPN ceils in the presence or in the absence of serum. After 24 h, ceils were nTpsinized and counted (Fig. 2). The absence of serum slightly decreased the attachment of EPN ceils, which was 90% in the presence and 80% in the absence of fetal bovine serum. In the presence of serum, the EPN ceils proliferate with an estimated dou- bling time of roughly 27 h (Fig. 2). Once the confluence was reached, EPN cells cease to divide, remaining contact inhibited. On the other hand, in the absence of serum the EPN cells are unable to proliferate. EPN cells form colonies starting from a single cell, and several separate clones have been isolated and grown as in- dependent lines.

To test anchorage dependence, a single-cell suspension of EPN cells was seeded in a semisolid medium. As shown in Fig. 3 (panel a), EPN cells were unable to form colonies in a semisolid media, whereas NIH-3T3 cells carrying the RET-MEN2A oncogene in the same experimental condition were able to form large colonies within the same time period (Fig. 3, panel b).

The ability of EPN cells to form tumors was tested upon injection into nude mice. After 45 d, not even one of the five animals injected with EPN cells presented outgrowth of tumors at the site of injec- tion, whereas those injected with ARO cells, a line of cells in cul- ture derived from a human thyroid carcinoma, presented tumors at the site of injection after 2 wk. Finally, the presence of the onco- suppressor gene product p53 wild type was analyzed. As shown in Fig. 4, p53 wild type was detected by immunoblot analysis in total cell lysates of EPN parental cell line and EPN clones (Fig. 4)

Biochemical analysis of prostate-specific and epithelial-specific markers. To ensure the epithelial derivation of EPN cells, the ex- pression of cytokeratins was studied. Immunoblot analysis of total cell lysates from EPN wild type and from several EPN clones showed the presence of 68-, 58-, 56.5-, and 50-kDa isoforms cm ~ responding to keratins numbered 1, 5, 10, and 14, respectively, according to the Moll catalog (Moll et al., 1982). The presence of this panel of cytokeratins indicates that EPN cells are derived from the basal component of the prostate epithelium (Fig. 5). The same samples were negative for the presence of desmin, which was un-

1000

100

1,r

1 I i i f ~ i

0 2 4 6 8 10 12

FIG. 2. Growth curve: 100,000 cells were seeded in Coon's modified Ham Fl2 medium in the absence of serum (0) or in the presence of 3% fetal calf serum (A). After 24 h, in half of the cells seeded in the presence of sermn, the culture medimn was replaced with sermn-free mediuln, and cells were counted at different periods of time. The values reported represent the mean + SD of three independent samples per experimental point.

168 SINISI ET AL.

FIG. 3. Microphotograph of EPN cells (panel a) and as control, NIH-3T3-MEN2A cell line (panel b) after 15 d of culture in a semisolid medium. Magnification: x150.

FIG. 4. Western blot detection of the p53 protein: 50 Ixg/lane of total protein extract from EPN cells; EPN clones 1, 3, 4, 5, 6, 7, 8, and normal kidney tissue as positive control were imrnunoblotted with anti-p53 antibody.

EPITHELIAL PROSTATE CELL LINE 169

FIG. 5. Western blot detection of cytokeratin isoforms: 50 lAg/lane of total protein extract from EPN cells; El'IN clones I, 3, 4, O, O, 7, 8, and normal kidney tissue as positive control were immunoblotted with anti-cytokeratin antibody.

FIG. 6. Immunoprecipitations and immunoblots with antibody against P-Tyr (a) and Pyk2 (b) in EPN cells stimulated with tumor necrosis factor-ix (10 min) and KC1 (10 rain).

detectable by immunoblot analysis with antidesmin-specific anti- bodies (data not shown).

Pyk2, a cytoplasmic tyrosine kinase related to the focal ad- hesion kinase, is expressed in normal human prostate epithelia but not in the stromal component of the prostate and in high- grade prostate tumors (Stanzione et al., 2001). EPN cells express Pyk2 tyrosine kinase, which can be activated by canonic stimuli like cytochines and changes in osmolarity (Tokiwa et al., 1996). As shown in Fig. 6, the treatment of EPN cells with tumor ne-

crosis factor-a or KC1 results in Pyk2 tyrosine phosphorylation,

within 10 min. The expression of PSA in EPN cells was investigated by immu-

nohistochemical analysis with specific anti-PSA antibodies. Consis- tent with the basal origin of EPN cells, PSA expression appears to be restricted to gland-like buds in a monolayer of PSA-negative cells (van Leenders et al., 2000) (data not shown).

EPN cells express androgen receptors, which are detected in total cell lysates from EPN cells and EPN clones by immunoprecipitation

1 7 0 SINISI ET AL.

Ptc. 7. (a) Immunopreeipitation: 500 Ixg of the total lysate from EPN cells, EPN clones 1, 3, 5, and 7, and Bosc cells, as positive controls, were subjected to immunoprecipitation with the anti-androgen receptor (AR) antibody followed by immunoblot with the same antibody. A single specific band was observed in all the cells tested. (b) Growth curve: 100,000 cells were seeded in Coon's modified Ham F12 medium supplemented with 3% charcoal-stripped serum (0); 3% charcoal-stripped serum and dihydmtestosterone of 10 7 M (A); and 3% fetal calf serum (11). After 24 h, in half of the cells seeded in the presence of serum, the culture medium was replaced with serum-free medium, and cells were coumed at different periods of time. The values reported represent the mean -+ SD of three independent samples per experimenlal point.

with antibodies against androgen receptors followed by immunoblot with the same antibodies. A band of the appropriate molecular weight, consistent with that of the androgen receptor (AR), was pre- sent in all the cells tested as well as in the Bosc cells (Pear et al.,

1993), a line stably transfected with the AR (Fig. 7a). To ensure that the androgen receptors present in the EPN cells were biolog-

ically active, the mitogenic activity of dihydrotestosterone (DHT) was tested. Cells were seeded in the presence of 3% charcoal- str ipped serum alone or along with DHT 10 _7 M and with 3% FCS.

As shown in Fig. 7b, DHT can sustain the proliferation of EPN cells, compensating almost completely for the absence of the lipid component in the charcoal-stripped serum. These data suggest that

EPITHELIAL PROSTATE CELL LINE 171

in the EPN cells, serum-induced proliferation mainly depends on the presence of steroids in FCS at physiological concentrations.

DISCUSSION

in spite of advances in culture techniques, adaptation in vitro re- mains a very complex and unpredictable phenomenon, and also the best culture conditions cannot guaranty 100% success from a primary culture to a cell line (Roger and Dumont, 1987; Curcio et al., 1994). Cells in the primary culture undergo a sort of aging process, which usually ends with the loss of the culture. During the culture adaptation process, cells accumulate mutations, some of which influence the ex- pression of well-known biochemical events or specific differentiated functions that can be easily monitored and spotted. The loss of PSA by the prostate cell within the first wk of growth in culture is an example of this (Sadar et al., 1999). The majority of the variations, however, cannot be monitored because they are totally stochastic and unpredictable, as is the time of their onset and their imquence on the behavior of each individual primrose culture. In addition, especially in the case of a human specimen, intrinsic variations within the tissue dono~ are important elements that can variously im%ence the success of a primary culture with respect to others. Very seldom, cells adapt spontaneously to indefinite growth in vitro, and when it happens, the cells are "frozen" forever in a peculiar instant of their life. If the ad- aptation process does not profoundly disturb the identity of a cell and does not induce transformation, then this cell may become a good model system in vitro. We are aware that a model system that perfectly mimics in vitro the situation in vivo is not conceivable. A pure epi- thelial cell line lacking the stromal and the neuroendocrine component as well as the control of the immune system is far from being a perfect reproduction of the in vivo situation. Nevertheless, cell model systems in vitro are unique and itTeplaceable tools to respond to the yet un- answered questions of the pathophysiology of the prostate.

EPN cells apparently went through a "soft" process of sponta- neous adaptation to indefinite growth in culture, so they maintained in vitro important traits of prostate identity. The prostate gland con- sists of three main histotypes with a connnon endodermal origin: the basal cell, which proliferates and expresses cytokeratins 5 and 14 but not PSA or prostate acid phosphatase (PAP); the luminal cells, which express PSA, PAP, androgen receptm, and cytokeratins 8 and 18 and are unable to divide; and the neuroendocrine cells (Abrahamsson, 1999). EPN cells, derived from the human normal prostate tissue, after more than 2 yr in culture maintained the ability of proliferating over the limits of 50 passages, suggesting that EPN cells overcame the critical crisis point and possibly adapted to grow indefinitely in culture. The EPN cells express cytokeratins 5 and 14 but not desmin; thus, according to the criteria proposed by Web- bet (Webber et al., 1996), they are of prostatic epithelial derivation (Sherwood et al., 1990; Turhan et al., 1998). The epithelial origin of EPN cells is strengthened by the observation that EPN cells express an active form of the cytoplasm tyrosine kinase Pyk2. In fact, Pyk2 is expressed in the normal human prostate epithelia but not in the stromal component of the prostate (Stanzione et al., 2001). The expression of Pyk2 could represent a potential new criterion for investigating the derivation and the tumorigenicity of prostate cells in culture.

In the prostate, most of the proliferating cells are seen in the basal layer, where subsets of basal cells expressing both basal cy- tokeratins and PSA have been detected, and represent an inter-

mediate stage of differentiation between the basal and secretory cells. The high degree of phenotypie plasticity of the basal cells is the basis of "the stem cell model" proposed by Bonkhoff in 1996 (Bonkhoff and Remberger, 1996; Bonkhoff, 1998), in which the bas- al cells are proliferating cells independent of but sensitive to an- drogen stimulation, and committed basal cells give rise to androgen- independent neuroendocrine cells, androgen-responsive basal cells, and androgen-dependent secretory cells (Bonkhoff, 1998; Hudson et al., 2000). EPN cells express both cytokeratins, of the basal cells, and functional androgen receptors and PSA, characteristic of the secretm~y epithelium. Thus, the EPN cell line apparently represents an intermediate step in the differentiation process from stem-un- differentiated to secretory-differentiated prostate cells. Two hypoth- eses can be suggested to explain the EPN cells phenotype: (1) EPN cells derive from the transiently proliferating or amplifying cell pop- ulations (Isaacs and Coffey, 1989) present within the original human prostate specimen; and (2) the prostate cells that acquired in culture an intermediate phenotype gained a selective advantage with re- spect to the other phenotypes present in the primary culture.

As the secretory- prostate epithelium, the EPN cell expresses a biologically active androgen receptm, which can mediate the mito- genic effect of testosterone on these cells. Regulation of prostate growth is a very complex phenomenon not yet fully understood and is controlled by several hormones and growth factors, acting through diverse postreceptor pathways. Although testosterone is the key fac- tor for the development of the gland, several lines of evidence sug- gest that growth factors, like EGF, insulin, and insulin-like binding protein, possibly produced locally by the stromal component of the prostate, acting through a stromal and acinar cell paracrine loop, play a role in the regulation of prostate cell proliferation (Kooistra et al., 1995; Culig et al., 1996; Lee et al., 1997; Russell et al., 1998; Sherwood et al., 1998; Balbay et al., 1999; Udayakumar et al., 1999). EPN cells express biologically active estrogen receptors c~ and [3 as well as receptors tbr growth factors like EGF (Chieffi et al., in prep.); thns, these cells will prove useful in investigating the complex network of signals and signaling pathways that interact to

regulate the prostate physiology. Finally, during the adaptation to indefinite growth in culture,

EPN cells did not acquire classical markers of transformation. In fact, they are contact inhibited, anchorage dependent, and unable to form tumors once injected into nude mice. A nontransformed differentiated prostate epithelial cell in vitro, like the EPN cells, will enrich the scenario of prostate model systems in vitro and will prove useful in a wide range of studies in cellular genetics, gene regulation, pharmacology, and aging. This will certainly lead to a more detailed understanding of the first step and the development of prostate tumors and in turn to the possibility of designing a new approach to the therapy of prostate diseases.

ACKNOWLEDGMENTS

We thank Dr. Domenico Fulgione and Mario Berardone for their skillful tech- nical artwork assistance. This work was supported by ~'ants from the Associazione Italiana Ricerca sul Czncro and from MURST/CNR "Biotecnologia" to DT, and from MURST (PRIN MM06108877) to AB.

REFERENCES

Abate-Shen, C.; Shen, M. Molecular genetics of prostate cancer. Genes Dev. 14:2410-2434; 2000.

172 SINISI ET AL.

Abrahamsson, P. A. Neuroendocrine cells in tumour growth of the prostate. Endocr. Relat. Cancer 6:503-519; 1999.

Balbay, M. D.; Pettaway, C. A.; Kuniyasu, H., et al. Highly metastatic human prostate cancer growing within the prostate of athymic mice overex- presses vascular endothelial growth factor. Clin. Cancer Res. 5:783-789; 1999.

Basolo, F.; Fiore, L.; Fusco, A., et at. Potentiation of the malignant phenotype of the undifferentiated ARO thyroid cell line by insertion of the bcl- 2 gene. Int. J. Cancer 81:956-962; 1999.

Bonkhoff, H. Neuroendocrine cells in benign and malignant prostate tissue: morphogenesis, proliferation, and androgen receptor status. Prostate 8:18-22; 1998.

Bonkhoff, H.; Remberger, K. Differentiation pathways and histogenetic as- pects of normal and abnormal perostatic growth: a stem cell model. Prostate 28:98-106; 1996.

Boudon, C.; Rodier, G.; Lechevallier, E., et al. Secretion of insulin-like growth factors and their binding proteins by human normal and hyperplastic prostatic cells in primary culture. J. Clin. Endocrinol. Metab. 81:612~617; 1996.

Craft, N.; Shostak, Y.; Carey, M.; Sawyers, C. L. A mechanism for hormone independent prostate cancer through modulation of androgen receptor signaling by the Her-2/Neu tyrosine kinase. Nat. Med. 5:280-285; 1999.

Culig, Z.; Hobisch, A.; Cronauer, M. V., et al. Regulation of prostatic growth and function by peptide growth factors. Prostate 28:393-405; 1996.

Curcio, F.; Ambesi-Impiombato, E S.; Perrella, G.; Coon, H. G. Long-term culture and functional characterization of follicular cells from adult normal human thyroids. Proe. Natl. Acad. Sei. USA 91:9004-9008; 1994.

Cussenot, O.; Berthon, E; Berger, R., et al. Immortalization of adult normal prostatic cells by liposomes containing T-SV40 gene. J. Urol. 146:881-886; 1991.

Fry, R M.; Hudson, D. L.; O'Hare, M. J.; Masters, J. R. W. Comparison of marker protein expression in benign hyperplasia in vivo and in vitro. Br. J. Urol. 85:504-513; 2000.

Hayfliek, L.; Moorhead, R S. The serial cultivation of human diploid cell strains. Exp. Cell Res. 25:561 585; 1961.

Horoszewiez, J. S.; Leong, S. S.; Kawinski, E., et al. LNCaP model of human prostatic carcinoma. Cancer Res. 43:1809-1818; 1983.

Hudson, D.; O'Hare, M.; Watt, E M.; Masters, J. R. W. Proliferative etero- geneity in the human prostate: evidence for epithelial stem cells. Lab. Invest. 80:1243-1250; 2000.

Isaaes, J. T.; Coffey, D. S. Etiology and disease process of benign prostatic hyperplasia. Prostate 2:33-50; 1989.

Kaingn, M. E.; Narayan, K. S.; Ohnuki, Y., et at. Establishment and char- acterization of a human prostatic cancer cell line (PC3). Investig. Urol. 17:16-23; 1979.

Kooistra, A.; Konig, J.; Keizer, D. M., et al. Inhibition of prostatic epithelial cell proliferation by a factor secreted specifically by prostatic stromal cells. Prostate 26:123-132; 1995.

Krill, D.; Shuman, M.; Thompson, M. T., et at. A simple method for isolation and culture of epithelial and stromal cells from benign and neoplastic prostates. Urology 49:981-988; 1997.

Lang, S. H.; Stower, M.; Maitland, N. J. In vitro modeling of epithelial and stromal interactions in non-malignant and malignant prostates. Br. J. Cancer 82:990-997; 2000.

Lee, C.; Kozlowsky, J. M.; Greyhack, J. T. Intrinsic and extrinsic factors controlling benign prostatic growth. Prostate 31:131-138; 1997.

Lev, S.; Moreno, H.; Martinez, R., et al. Protein tyrosine kinase Pyk2 involved in Ca+-induced regulation of ion channel and MAP kinase function. Nature 376:737-744; 1995.

Macpherson, I.; Montagniel, I. Agar suspension culture for the selective assay of cell transformed by polyoma virus. Virology 23:291-294; 1964.

Moll, R.; Werner, W. F.; Schiller, D. L., et al. The catalog of human cytoker- atins: patterns of expression in normal, epithelial, tunmrs and cul- tured cells. Cell 31:11-24; 1982.

Pear, W. S.; Nolan, G. R; Scott, M. L.; Baltimore, D. Production of high-titer helpe~free retroviruses by transient transfection. Proc. Natl. Acad. Sci. USA 90:8392.8396; 1993.

Roger, P. P.; Dumont, J. E. Thyrotropin is a potent growth factor for normal human thyroid cells in primary culture. Biochem. Biophys. Res. Com- mun. 149:707-711; 1987.

Russell, P. J.; Bennett, S.; Stricker, P. Growth factor involvement in progres- sion of prostate cancer. Clin. Chem. 44:705-723; 1998.

Sadar, M. D.; Hussain, M.; Bruchovsky, N. Prostate cancer: molecular biology of early progression to androgen independence. Endocr. Relat. Cancer 6:487-502; 1999.

Santoro, M.; Carlomagno, E; Romano, A., et al. Germ-line mutations of MEN2A e MEN2B activated RET as a dominant transforming gene by different molecular mechanisms. Science 267:381-383; 1995.

Sherwood, E. R.; Berg, L. A.; Mitchell, N. J., et al. Differential cytokeratin expression in normal, hyperplastic and malignant epithelial cells from human prostate. J. Urol. 143:167-171; 1990.

Sherwood, E. R.; Van Dongen, J. L.; Wood, C. G., et al. Epidermal growth factor receptor activation in androgen-independent but not androgen- stimulated growth of human prostatic carcinoma cells. Br. J. Cancer 77:855~361; 1998.

Sinisi, A. A.; Bellastella, A.; Prezioso, D., et at. Different expression patterns of somatostatin receptor subtypes in cultured epithelial cells frmn human normal prostate and prostate cancer. J. Clin. Endocrinol. Me- tab. 82:2566-2569; 1997.

Stanzione, R.; Picascia, A.; Chieffi, P., et al. Variations of proline-rich kinase Pyk2 expression correlate with prostate cancer progression. Lab. In- vest. 81:51-59; 2001.

Stone, K. R.; Mickey, D. D.; Wunderli, H.; Mickey, G. H.; Paulson, D. E Isolation of human prostate carcinoma cell line (DU145). Int. J. Can- cer 21:274--281; 1978.

Talpin, M. E.; Bubtey, G. J.; Shuster, T. D., et al. Mutation of the androgen- receptor gene in metastatic androgen-independent prostate cancer. N. Engl. J. Med. 332:1393-1398; 1995.

Tokiwa, G.; Dikic, I.; Lev, S.; Schlessinger, J. Activation of Pyk2 by stress signals and coupling with JNK signaling pathway. Science 273:792-794; 1996.

Turhan, O. I.; Aydin, N. E.; Sariynce, O.; Ozkan, S. Cells in various benign and malignant conditions of the human prostate express different an- tigenic phenotypes. Int. Urol. Nephrol. 30:731-744; 1998.

Udayakumar, T. S.; Jeyaraj, D. A.; Rajalakshmi, M.; Sharma, R. S. Cuhure of prostate epithelial cells of the rhesus monkey on extracellular ma- trix substrate: influence of steroids and insulin-like growth factors. J. Endocrinol. 162:443-450; 1999.

van Leenders, G.; Dijkman, H.; Hulsbergen-van de Kaa, C., et at. Denmn- stration of intermediate cells during human prostate epithelial differ- entiation in situ and in vitro using triple-staining confocal scanning microscopy. Lab. Invest. 80:1251-1258; 2000.

Visakorpi, T.; Hyytinen, E.; Koivisto, P., et al. In vivo amplification of the androgen receptor gene and progression of human prostate cancer. Nat. Genet. 9:401-406; 1995.

Webber, M. M.; Bello, D.; Quader, S. hnmortalized and tumorigenic adult prostatic epithelial cell line: characteristics and applications. Part I. Cell markers and immortalized non tumorigenic cell line. Prostate 29:386-394; 1996.

Webber, M. M.; Bello, D.; Quader, S. Immortalized and tumorigenic adult prostatic epithelial cell line: characteristics and applications. Part II. Tumorigenic cell line. Prostate 30:N~64; 1997a.

Webber, M. M.; Bello, D.; Quader, S. Inunortalized and tumorigenic adult prostatic epithelial cell line: characteristics and applications. Part III. Oncogenes, suppressor genes, and applications. Prostate 30:136-142; 1997b.

Wong, Y. C.; Wang, Y. Z. Growth factors and epithelial-strmnal interactions in prostate cancer development. Int. Rev. Cytol. 199:65-116; 2000.