Extended Survivability of Prostate Cancer Cells in the Absence Of

(CANCER RESEARCH 58. 3466-3479. August 1. 1998] Extended Survivability of Prostate Cancer Cells in the Absence of Trophic Factors: Increased Proliferation, Evasion of Apoptosis, and the Role of Apoptosis Proteins Dean G. Tang,1 Li Li, Dharam P. Chopra, and Arthur T. Porter

Department of Radiation Oncology, Wayne State University, Detroit, Michigan 48202 ID. G. T., A. T. P.Â¡; Karmanos Cancer Institute, Detroit, Michigan 48201 [D. G. T., A. T. P.]; Biomide Corporation, Grosse Pointe Farms. Michigan 48236 [D. G. T., L. L.]; and institute of Chemical Toxicologv, Wavne State University, Detroit, Michigan 48226 ID. P. C.I

ABSTRACT accepted that apoptosis is a dominant feature unless suppressed or overridden by growth/survival factors; therefore, most eukaryotic This project was undertaken to study the survival properties of various cells (with the exception of, probably, blastomeres) will undergo prostate cells, including normal (NHP), BPH (benign prostate hyperplasia), primary carcinoma (PCA), and metastatic prostate cancer cells apoptosis when deprived of appropriate survival factors (1, 2). In (LNCaP, PC3, and Dul45), in the absence of trophic factors. Cell prolif creased cell proliferation, extended cell survival, and diminished eration and cell death were quantitated by enumerating the number of live apoptosis have all been implicated in tumorigenesis. cells using MTS/PMS kit and of dead (apoptotic) cells using 4',6-dia- There has been a close association between apoptosis and prostate midino-2-phenylindole dihydrochloride nuclear staining. These cells dem cancer initiation, progression, metastasis, and response to treatment. onstrated an overall survivability in the order of BPH < NHP < LNCaP The growth and survival of normal prostate epithelial cells depend on < PC3 < PCA < Du 145. Upon growth factor deprivation, NHP/BPH cells androgens; therefore, castration or chemical blocking of androgen rapidly underwent apoptosis, leading to a decreased number of live cells. function leads to prostate atrophy. Prostate cancer is composed of PCA/PC3/Dul45 cells, in contrast, demonstrated an initial phase of ag androgen-dependent and androgen-independent cells. Androgen abla gressive growth during which apoptosis rarely occurred, followed by a "plateau" phase in which cell loss by apoptosis was compensated by cell tion eliminates most androgen-dependent cancer cells by inducing proliferation, followed by a later phase in which apoptosis exceeded the apoptosis but can rarely cure the patients due to the presence of cell proliferation. LNCaP cells demonstrated survival characteristics be androgen-independent cells and emergence of apoptosis-resistant tween those of NHP/BPH and PCA/PC3/Dul4S cells. We concluded that clones (3, 4). Bcl-2 oncoprotein expression has been correlated with the increased survivability in prostate cancer cells results from enhanced the progression and metastasis of prostate cancer cells and the emer cell proliferation as well as decreased apoptosis. The molecular mecha gence of androgen-independent cells, as well as the generation of nisms for evasion of apoptosis in prostate cancer cells were subsequently apoptosis- and therapy-resistant cancer cell clones (5-11). On the investigated. Quantitative Western blotting was used to examine the pro tein expression of PS3 and P21WAIM, Bcl-2 and Bcl-XL (anti-apoptotic other hand, androgen-independent prostate cancer cells still retain proteins), and Bax, Bak, and Bad (proapoptotic proteins). The results appropriate apoptotic machinery and can be induced to undergo revealed that, upon trophic factor withdrawal, NHP and BPH cells up- apoptosis by a wide spectrum of biological and pharmacological regulated wild-type p53 and proapoptotic proteins Bax/Bad/Bak and treatments, including wild-type P53 and P212 (12-14), prostate down-regulated the expression of P21. Furthermore, NHP and BPH cells apoptosis response-^ (15), prevention of adhesion (16), transforming endogenously expressed little or no Bcl-2. In sharp contrast, prostate growth factor ÃŸ(17), tumor necrosis factor a (18), Fas (19), chemo- cancer cells expressed nonfunctional P53 and various amounts of Bcl-2 proteins. Upon deprivation, these cancer cells up-regulated P21 and Bcl-2 therapeutic drugs such as cisplatin, camptothecin, teniposide, and and/or BciXL, lost response to withdrawal-induced up-regulation of Bax/ vincristine (20, 21), novel chemical entities such as Linomide (quin- Bad/Bak or decreased or even completely lost Bax expression and ex oline-3-carboxamide; Ref. 22), ÃŸ-lapachone(23, 24), phenylbutyrate pressed some novel proteins such as P25 and P54/55 complex. These data (25), ribonucleotidase inhibitors (26), diethylstilbestrol (27), ÃŸ-L-(-)- together suggest that prostate cancer cells may use multiple molecular dioxolane-cytidine (28), and retinoid derivatives (29), and signal mechanisms to evade apoptosis, which, together with increased prolifer transduction modifiers such as phorbol esters (30-32) and thapsigar- ation, contribute to extended survivability of prostate cancer cells in the gin (33). We recently reported apoptosis induction by prostate secre absence trophic factors. tory protein 94 (34) and hydroxamic acid compounds (35, 36) in hormone-refractory human prostate cancer cells. INTRODUCTION Little information exists in regard to differential properties of normal and cancerous prostate epithelial cells in terms of sensitivity to Tissue and organ homeostasis is controlled by balanced cell pro apoptosis induction. Furthermore, no systematic studies have been liferation, cell survival, and cell death. Cell proliferation is driven and performed to investigate the interrelationship among the proliferation, regulated by various peptide growth factors and/or cytokines. Cell survival represents the cell's intrinsic life expectancy in the absence of survival, and apoptosis of various types of prostate cells. Recently, a series of normal (NHP2), BPH, and PCA prostate cultures have been survival factors, many of which overlap with the growth factors/ cytokines that drive cell proliferation. Physiological cell death occurs established and characterized (37). In this study, we took advantage of mostly via programmed cell death or apoptosis (1). Apoptosis is a these primary cultures and compared their growth and survival prop genetically controlled suicidal program that takes place, in many erties in the absence of trophic factors with those of the established cases, when the cells run out of their survivability in the absence of malignant prostate cancer cell lines, i.e., LNCaP, PC3, and Dul45. growth/survival (or trophic) factors, although it can also be triggered We further investigated the changes and the potential roles of several by a wide variety of physiological or obnoxious stimuli. It is generally apoptosis proteins in these six different cell types upon growth factor withdrawal. The results show that prostate carcinoma cells demon

Received 2/5/98; accepted 6/2/98. strate a much higher proliferative potential and survive much longer The costs of publication of this article were defrayed in part by the payment of page than normal and BPH cells. charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1To whom requests for reprints should be addressed, at Department of Radiation 2 The abbreviations used are: NHP, normal human prostate; BPH, benign prostate hyperplasia; PCA, primary prostatic carcinoma; DAPI, 4',6-diamidino-2-phenylindole; Oncology, Wayne State University, 407 Life Science Building, Detroit, MI 48202. Phone: (313)577-2184; Fax: (313)577-2375; E-mail: [email protected]. MTT, 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl-tetrazolium bromide; P21, P21WAIM.

3466

MATERIALS AND METHODS form formazan dyes. Viable cells with active mitochondria cleave the tetrazolium ring into a visible dark blue formazan reaction product while dead cells Cell Culture, Antibodies, and Reagents. PC3. Du 145, and LNCaP (FGC- are not stained. The viable cells in each well were counted by bright-field light 10) cells were obtained from American Type Culture Collection and routinely microscopy. cultured in RPMI 1640 supplemented with 10% fetal bovine serum. 2 mM Western Blotting. Western blotting protocols were detailed before (38. L-glutamine. and antibiotics. NHP, BPH. and PCA cells were surgically 39). Whole-cell lysates were prepared by directly scraping control or starved isolated from clinically diagnosed and pathologically confirmed patients and prostate cells into the TNC lysis buffer [10 mM Tris-acetate (pH 8.0), 0.5% characterized as detailed previously (37). All of these primary cells were NP40. and 5 mM CaCU], and solubilized proteins were analyzed on 4-20% cultured in KBM basic medium (Clonetics) supplemented with 5 fig/ml gradient or 12% SDS-PAGE. Isolated proteins were transferred to nitrocellu insulin. 0.5 /^g/ml hydrocortisone. 10 ng/ml epidermal growth factor, 50 /J.g/ml lose and probed with individual primary antibodies as specified in "Results." bovine pituitary extract. 10 ng/ml cholera toxin, and antibiotics (37). Only The secondary antibody used was goat anti-rabbit or -mouse IgG conjugated to cultures of passages 2-5 were used throughout this study. The following horseradish peroxidase. Then membranes were stripped and reprobed with primary antibodies were used in the immunoblotting experiments, (ai anti-p53: different antibodies. In some cases, the membrane was incubated with two a mouse monoclonal anti-human wild-type p53 (clone DO-1; amino acids different primary antibodies. The intensities of individual bands were quanti- 21-25 as the immunogen) from Oncogene Research/Calbiochem and another tated by densitometric scanning (41). and the values were normalized against mouse monoclonal anti-human wild-type p53 (clone DO-1 ; amino acids 37-45 the intensity of actin. as the immunogen) from Santa Cruz Biotechnology: (b) anti-p21: a rabbit Characterization of Bcl-2-related Proteins. The Bcl-2 protein normally polyclonal anti-human peptide antibody (amino acids 146-164 as the immuno runs at M, â€”¿26.000onSDS-PAGE under reducing conditions. Two prominent gen) from Santa Cruz: (c) anti-Bcl-2: a mouse monoclonal anti-human peptide upper bands (M, -27.000 and M, 30.000. respectively) were also detected in Bcl-2 (amino acids 41-54 as the immunogen) from either Dako (clone 124) or this study using a monoclonal anti-Bcl-2 (raised against the 41-54 peptide of Calbiochem (clone 100), and a rabbit polyclonal anti-human Bcl-2 raised human Bcl-2) in both PC3 and/or Dul45 cells (see "Results"). These upper against a peptide corresponding to amino acids 4-21 (sc-492) from Santa bands may represent phosphorylated isoforms of Bcl-2, some novel Bcl-2- Cruz; (d) anti-Bcl-Xs/I : a rabbit polyclonal anti-human peptide antibody related proteins, or simply nonspecific products. These possibilities were tested (amino acids 2-19 as the immunogen) from Santa Cruz; (e) anti-Bax: a rabbit with the following two sets of experiments. In the first. PC3 and Du 145 cells polyclonal anti-human Bax peptide antibody (amino acids 150-165 as the were harvested, and cell lysates were prepared using the TNC lysis buffer, as immunogen) from Calbiochem and a rabbit anti-human Bax peptide antibody described above. Five hundred /ig of PC3 or Du 145 cell lysates were used for (amino acids 43-61 as the immunogen) from PharMingen; (/) anti-Bak: a immunoprecipitation experiments, as described previously (41) with some rabbit polyclonal anti-human Bak (amino acids 82-104 as the immunogen) modifications. Briefly, the cell lysates in 1 ml of TNC lysis were first from Santa Cruz; (#) anti-Bad: a rabbit polyclonal anti-mouse Bad (amino precleared by incubating with 75 fil protein A/G-agarose (Santa Cruz Biotech acids 185-204 as the immunogen) from Santa Cruz; and (/i) anti-actin: a mouse nology) at 4Â°Cfor 1 h. Then 5 /ig of the monoclonal anti-Bcl-2 or isotype- monoclonal anti-actin from ICN. The secondary antibody was either goat matched control monoclonal antibody MOPC-21 (IgGl) were added to the anti-rabbit or anti-mouse IgG (whole molecule) conjugated to horseradish precleared supernatants and incubated at 4Â°Cfor 2 h under rotating conditions. peroxidase (Amersham). ECL reagents were bought from Amcrsham. The At the end. 100 Â¿dof protein A/G-agarose beads were added, and the mixture Western blotting experiment was performed as described previously (38. 39). was further incubated for l h at 4Â°C.The beads with bound antigen-antibody For starvation, various cells cultured in regular media were washed (three complex were washed extensively: TNC lysis buffer (two times). TNC-0.1% times) with growth factor-free RPMI medium followed by culturing cells for SDS (one time). TNC-1 M NaCI (one time), and TNC lysis buffer (one time). different periods of time in RPMI. Finally, the washed beads were reconstituted in 50 Â¡Â¿Iof2X SDS-PAGE Quantification of Live Cells Using the MIS Assays. Cell survival was sample buffer [125 mM Tris-HCI (pH 6.8), 10% ÃŸ-mercaptoethanol. 4% SDS. determined using the CellTiter MTS/PMS cell proliferation kit (Promega 20% glycerol. and 0.001% bromphenol blue]. After boiling. 25 /il of the Corp.), as described (39). Briefly, log-phase growing prostate cells were precipitated sample were loaded onto a 12% SDS-PAGE. and separated cultured in Hat-bottomed 96-well plates (1 X IO4 cell/well) in either RPMI- proteins were transferred to nitrocellulose membrane and then blotted using the 10% PCS medium (for LNCaP. PC3. and Dul45 cells) or in KBM medium monoclonal or the rabbit polyclonal anti-Bcl-2. In the second set of experi supplemented with various growth factors (for NHP. BPH. and PCA cells). ments. 50 /j.g of clarified PC3 or Du 145 cell lysates were incubated with 2 The wells were gently washed with plain media (times three), and then 200 /il units of alkaline phosphatase at 37Â°Cfor 4 h. At the end. 25 /ig of the protein of trophic factor-free RPMI were added to each well to start the deprivation were loaded onto a 12% SDS-PAGE. and separated proteins were used for experiment. The number of live cells on each day thereafter was determined by immunoblotting with the monoclonal anti-Bcl2. as described above. the uptake of MTS/PMS dyes and measured on a 3550-UV microplate reader (Bio-Rad) using the/1.,,,,, absorbance. One set of wells were measured just prior to the media change, and the absorbance readings were used as the baseline RESULTS value (i.e., day 0). The cell number on each day was expressed as the percentage cell number of the day 0. For each day's measurement, five- or Expression of Apoptosis Proteins in Various Prostate Cells. We first examined the expression of various apoptosis-related proteins in six-well repeats were run, and each experiment was performed two to six times (see figure legends). A >100% cell number indicates cell growth during that different types of prostate cells. The results are plotted in Fig. 1. All period. cells except PC3 expressed P53, as revealed by a monoclonal anti Apoptosis Assays by Light Microscopy and DNA Fragmentation. These body. Note that longer exposure of the film also resulted in the p53 methods have been detailed previously (35. 36. 38. 39). band in BPH cells (see also Fig. 7). P53 expressed in the primary cells. Quantitation of Cell Death with DAPI Staining and MTT Assays. i.e., NHP, BPH. and PCA, was shown previously to be wild-type (37). Different types of prostate cells cultured on glass coverslips in the absence of LNCaP cells also express wild-type P53 (12). PC3 cells are P53-null; growth factors for various days were directly stained with DAPI (5 jag/ml) for therefore, no protein could be detected (Fig. 1). Dul45 cells expressed 10 min at 37Â°C.At the end, coverslips were gently rinsed (two times) and the highest amount of P53 protein because the p53 gene in these cells mounted onto slides with a small amount of SlowFade anti-fade mounting has two mutations (codons 223 and 274), resulting in accumulation of media (Molecular Probes, Eugene, OR). The slides were observed under a the mutant protein (12). P21, a P53 target, was expressed in all of the Zeiss fluorescence microscope and pictures were recorded on Tmax-400 cells including PC3 (Fig. 1). There appear to be some differences in black/white films. The cells showing condensed chromatin and/or fragmented terms of the protein amount of p2lWAF-' in different cells (Fig. 1). nuclei were scored as apoptotic. An average of 500-1000 cells were counted When using 25-100 jug of whole-cell lysates, the M, 26,000 Bcl-2 for each slide/day. The results were expressed as the percentage of apoptosis of day 0. protein was undetectable or. in some cases (e.g.. Fig. 3S), only faintly Alternatively, cell growth/death were monitored by MTT (40) assays. In this detectable in NHP cells (Fig. 1). In similar experiments, Bcl-2 was not assay, the MTT dyes are metabolized by mitochondria! dehydrogenases to detected at all in BPH and PCA cells (Fig. 1; see also Figs. 7 and 9) 3467

antibody (data not shown). Western blotting was also performed with three proapoptotic proteins, Bax, Bad, and Bak. All of the prostate cells expressed the M, -22,000 Bad (Figs. 3-7) and Mr -25,000 Bak (Fig. 1). All prostate cells except Dul45 expressed the Mr -21,000 P53 â€”¿ Bax protein, as revealed by an antibody recognizing the amino acids 43-61 of human Bax. The lack of Bax protein expression in Du 145 cells was also confirmed by another polyclonal anti-Bax raised against peptide 150-165 (data not shown). Interestingly, LNCaP and PC3 WAF-l P21 â€¢¿21kd cells appear to express more Bax proteins than the three primary prostate cells (Fig. 1). The monoclonal anti-Bcl-2 used in our studies did not recognize any other Bcl-2 family proteins (Fig. 1 and data not shown). There ,30kd-Bcl-2 Bcl-2 fore, the A/r 27,000 and MT30,000 bands detected by this antibody in 3cl-2 (27 kd) PC3 and/or Du 145 cells may represent phosphorylated Bcl-2 protein x26-kd Bcl-2 (42-46) or a new Bcl-2-related protein, such as BRAG-1, which migrates at M, â€”¿30,000(47),or simply represent nonspecific reaction products. We performed two sets of experiments to differentiate among these possibilities. In the first, Bcl-2 and/or Bcl-2-related Bel-XL â€”¿ â€¢¿30kd proteins were immunoprecipitated, and the precipitates were further subject to immunoblotting with the polyclonal or monoclonal anti- Bcl-2. This type of "double-shot" (i.e., immunoprecipitation followed by immunoblotting) protocol has been widely used to confirm the Bax ~ ^. 21 kd specificity of protein bands identified on Western blot. As shown in Fig. 24, the MT 30,000 band was specifically immunoprecipitated down by the monoclonal anti-Bcl-2 but not by control antibody MOPC-21, suggesting that this protein represents a Bcl-2-related Bak . â€¢¿25kd protein instead of a nonspecific product. Blotting with the polyclonal anti-Bcl-2 antibody produced similar results (data not shown). Note that the status of the Mr 26,000 and Mr 27,000 proteins in this experiment was not very clear due to the proximity of these bands to Actin the immunoglobulin light chain (MTâ€”¿25,000;Fig.2A). The direct use of whole-cell lysates in immunoblotting again identified the Mr 26,000, Mr 27,000, and M, 30,000 (on the order of Mr 26,000 Â» Mr Fig. 1. Expression of apoptosis proteins in prostate cells. Twenty five /ig of whole-cell lysates prepared from regularly cultured primary [i.e., NHP (normal), BPH. and PCA 27,000 > Mr 30,000 in terms of intensity) protein bands in PC3 cells, (primary) cells] or established prostate cancer cells (i.e., LNCaP, PC3. and Dul45) were and Air 26,000 and Mr 30,000 (Mr 30,000 Â» Mr 26,000) protein resolved on 4-20% SDS-PAGE and used for immunoblotting with a monoclonal anti-P53 bands in Du 145 cells (Fig. 24). Only the Mr 26,000 Bcl-2 protein was (amino acids 21-25 as the immunogen). The membrane was sequentially stripped [62.5 IHMTris-HCI (pH 6.5), 2% SDS, and 100 mM 2-ME, l h at 50Â°C]and reprobed with: detected in A431 human epidermoid carcinoma cells (used as a rabbit polyclonal anti-p21v , monoclonal anti-Bcl-2 (amino acids 41-54 as the im control; Fig. 24, Lane 1). In the second set of experiments, whole-cell munogen), polyclonal anti-Bcl-X^s, polyclonal anti-Bax (amino acids 43-61 as the lysates from PC3 and Dul45 cells were treated with alkaline phos- immunogen), polyclonal anti-Bak, polyclonal anti-Bad (not shown), or monoclonal anti- actin (see "Materials and Methods" for details). The same panels of antibodies were used phatase and then immunoblotted with monoclonal anti-Bcl-2. As in all of the subsequent experiments unless specified otherwise. As indicated in "Materials and Methods" and other parts of the text, several different antibodies were sometimes used shown in Fig. 2B, in PC3 cells, phosphatase treatment dramatically for certain molecules for the purpose of characterization. The secondary antibodies used decreased the amount of the Mr 27,000 protein with a simultaneous were either goal anti-rabbit or anti-mouse IgG conjugated to HRP. Left, identities of increase in the quantity of MT 26,000 protein, suggesting that the MT individual proteins; right, corresponding molecular weights. Actin is detected as a Afr 27,000 protein in PC3 cells is phosphorylated Bcl-2. In contrast, the â€”¿45.000protein.The monoclonal anti-Bcl-2 consistently detected several different bands in LNCaP (M, 26,000 Bcl-2), PC3 (M, 26,000, M, 27,000, and M, 30,000 Bcl-2) and level of the Mr 30,000 protein in PC3 cells was hardly affected by the Du 145 cells (M, 30,000 Bcl-2). The M, 27,000 band is labeled as the phosphorylated Bcl-2 phosphatase treatment, which similarly did not affect the mobility of (see Fig. 2). the Mr 30,000 protein in Dul45 cells (Fig. 20). The intensity of the MT 30,000 protein in Du 145 cells appeared to decrease a small amount; however, no increased Mr 26,000 Bcl-2 protein was observed after the in several repeat experiments, even after extended exposure. The M, phosphatase treatment (Fig. 2B). Taken together, these data suggest 26,000 Bcl-2 protein was detected in LNCaP cells (Fig. 1). In PC3 that the Mr 27,000 protein represents the phosphorylated isoform of cells, in addition to the Mr 26,000 Bcl-2, a Mr 27,000 and a Mr 30,000 Bcl-2 protein, and the Mr 30,000 protein may represent a novel protein were also detected (Fig. 1). In Dul45 cells, by contrast, only Bcl-2-related protein, the true molecular identity of which awaits the Mr 30,000 protein was detected (Fig. 1) in addition to the Mr further characterization. 26,000 Bcl-2, which was detected only when higher amounts of Survival of NHP Cells in the Absence of Trophic Factors. As proteins/longer exposure were used (Fig. 2). This pattern of expres early as 24 h after the withdrawal of growth factors, NHP cells sion for Bcl-2 protein was also observed when the membrane was stopped proliferation, as measured by MTS assays (Fig. 3A). There probed with a polyclonal antibody (data not shown). In general, the after, interestingly, there appeared to have a slight "bouncing-back" in Bcl-XL protein was detected as a faint Mr â€”¿30,000protein in all the cell growth 48-72 h after culturing in the absence of trophic prostate cells (Fig. 1, indicated by an arrow). This polyclonal antibody factors (Fig. 3A), as observed consistently in several repeat experi also recognized two prominent upper bands, the identities of which ments. The reason for this observation was not clear. DNA fragmen were unknown and whose intensities varied with different cell types. tation assays (Fig. 3B, inset), DAPI staining (Fig. 3Q, and MTT No Bcl-Xs protein was detected in any cell type with this polyclonal assays (not shown) revealed that NHP cells died through apoptosis. 3468

Anti-Bcl-2 Blotting B

Whole Anti- cell lysates Bcl-2 MOPC I.P Anu-Bcl-2 Blotting PC3 Dul45 Cells pi u + - - + Alk. phos. s kÃ®Ã SriÃD fi U 97.4 â€”¿ 97.4- 66â€”¿ 66- --pn â€”¿Ij;heavy â€¢¿â€” â€”¿Â» -BSA 46 â€”¿ chain 46- 30â€”¿ ^-* â€¢¿Â»â€¢*-ilight 30- ,-30kd-Bcl-2 21.5â€”¿ chain =^Â©-Bcl-2 (27 kd) Bcl-2 (26 kd) 14.3â€”¿ 21.5â€”¿

Fig. 2. Characterization of Bcl-2-related proteins. In A. the M, 30,000 protein represents a specific Bcl-2-related protein. Five hundred /ig whole-cell lysates from either PC3 (Lanes 4 and 6) or Dul45 (Â¡Mnes5 and 7) cells were immunoprecipitated with either monoclonal anti-Bcl-2 (Lanes 4 and 5) or isotype-matched control antibody MOPC-21 (Lanes 6 and 7). The immunoprecipitates were resolved on a 12% SDS-PAOE. and the membrane was probed with the monoclonal anli-Bcl-2. As controls, 5 fig of total cell lysates from A431 (human epidermoid carcinoma; Lane /), 50 ng of whole-cell lysates from Du 145 (Lane 3). or 70 fig of lysates from PC3 cells (Lane 2) were directly used for Western blotting. Left, molecular weight markers. Right. Ihe immunoglobulin heavy (M, ~54,000) and light (M, â€”¿25.000)chainsunder reducing conditions. The M, â€”¿66,000bandin the whole-cell lysates (Lanes 1-3) represents the BSA (the monoclonal anti-Bcl-2 was raised against a peptide conjugated to BSA). The Afr 30,000 Bcl-2 protein was precipitated by anti-Bcl-2 but not by MOPC-21 (arrowhead on the right). B, the M, 27,000 band is the phosphorylated Bcl-2 protein. For details, see "Materials and Methods" and other parts of the text. In untreated PC3 cells (indicated by the "-" sign), the M, 26,000, M, 27,000, and M, 30,000 proteins were identified by the monoclonal anli-Bcl-2, which detected the M, 30.000 and the faint M, 26,000 protein in untreated Du 145 cells. Phosphatase (Alk. phos. ) treatment (indicated by the " + " sign) resulted in the disappearance of the M, 27.000 with simultaneous increase in the amount of M, 26,000 Bcl-2 in PC3 cells, without affecting the mobility of the M, 30.000 band in either PC3 or Dul45 cells.

Increased apoptosis (i.e., positive DAPI labeling) was observed 24 h 3, after which its levels gradually decreased (Figs. 7 and 5ÃŸ).The P21 after growth factor withdrawal. On day 2, â€”¿20%cellswere apoptotic level did not undergo appreciable changes and was not affected by (Fig. 3, B and C, middle panel). By day 7, 60-70% cells were dead P53 up-regulation. No Bcl-2 was detected in these cells (Fig. 7). The (Fig. 3, B and C. right panel). Characterization of cell growth and cell Bcl-XL level was also very low, which showed only a marginal death by several different methods (i.e., MTS assays, DAPI staining, increase (Figs. 7 and 5B). As in NHP cells, Bax and Bad proteins DNA fragmentation assays, and MTT) overall produced consistent persistently increased with time; however, Bak did not show any results; NHP cells do not proliferate and undergo steady apoptotic changes (Figs. 7 and 5B). death upon removal of trophic factors (Fig. 3 and data not shown). Survival of PCA Cells in the Absence of Trophic Factors. PCA Note that the DNA ladder formation in all three primary cells (Fig. 3; cells behaved completely differently with respect to their survival in see also Figs. 6 and 8) was not as clear-cut as in established cell lines the absence of trophic factors. Immediately (i.e., 24 h) after growth (see Figs. 10, 12, and 14). This is not surprising because apoptosis in factor withdrawal, PCA cells proliferated (>2.5 fold increase in cell primary cultures often is not accompanied by distinct DNA ladder number) as determined by MTS assays (Fig. 8). PCA cells did not formation (reviewed in Ref. 4). Western blotting was used to examine demonstrate significant (15-20%) apoptosis until days 5-6, as sup the changes of apoptosis proteins. As shown in Fig. 4, all proteins ported by faintly increased DNA ladder formation (Fig. 8ÃŸ,inset). except P21 increased immediately after deprivation, as determined by MTT assays (data not shown) also confirmed increased cell prolifer densitometric scanning after normalizing the protein levels to that of ation and rare cell death. The total number of live cells did not actin (Fig. 5: see legend for details). The most significant increase was decrease, even by day 5 when increased apoptosis was observed (Fig. observed with Bcl-XL (> 15 fold by day 6; Figs. 4 and SA). The NHP 8, B and C, middle panel), suggesting that continued cell proliferation cells expressed barely detectable levels of Bcl-2, which appeared to compensated for the cell loss. By day 11, nearly the same number of increase marginally by days 3 and 4 (Fig. 4). In contrast, all three cells (i.e., the same number of cells as initially input or 10,000 proapoptotic proteins, i.e., Bax, Bad, and Bak, significantly increased cells/well; see "Materials and Methods") were still alive (Fig. 8, A and with time (Figs. 4 and 5A). P53 also demonstrated a 2-3-fold increase, C, right panel). The most significant changes in apoptosis proteins in whereas its target. P21, did not show any corresponding changes and PCA cells after trophic factor deprivation were the up-regulation of p2]WAF-i and BCI_XL around day 4 (Figs. 9 and 5Q, prior to the whose levels actually declined sharply by day 5 (Figs. 4 and 5/4). Therefore, the P21 expression in NHP cells appears to be p53 inde initiation of apparent DNA fragmentation (Fig. 8S, inset). Like NHP pendent. and BPH cells, PCA cells expressed extremely low levels of Bcl-2, Survival of BPH Cells in the Absence of Trophic Factors. Like which was undetectable (Fig. 9). In sharp contrast to NHP and BPH NHP cells, BPH cells did not grow, upon growth factor withdrawal, as cells, all other proteins, including P53, Bax, Bad, and Bak, were not revealed by MTS assays (Fig. 6). In fact, BPH cells appeared to be up-regulated (Fig. 9). However, interestingly, three prominent protein more sensitive than NHP cells to deprivation because they demon bands, i.e., a MT -25,000 protein and a M, 54,000/55,000 doublet, strated an even faster loss in the number of live cells (compare Figs. were simultaneously detected on day 4 after growth factor withdrawal, 6A and 3/4). For example, â€”¿40%ofBPH cells were apoptotic (Fig. 6, when using the monoclonal anti-P53 for the Western blotting (Fig. 9, B and C, middle) by day 2 compared with â€”¿20%apoptosisin NHP top panel). These two novel entities of proteins were designated P25 cells at the same time (Fig. 3). By day 7, >80% cells were dead (Fig. and P54/55, respectively. 6, B and C, right panel). Western blotting revealed quite different Survival of LNCaP Cells in the Absence of Trophic Factors. patterns of changes in apoptosis proteins than found in NHP cells. The LNCaP cells demonstrated marginal proliferation upon serum starva P53 up-regulation was most dramatic, with â€”¿30-foldincrease by day tion, as revealed by MTS assays (Fig. 10/4). However, the cells did not 3469

proteins except P21 increased (2-10-fold) in expression levels upon loo- serum starvation (Figs. 11 and 5D). Also like in NHP cells, the P21 levels decreased in a time-dependent manner, which was barely de tectable by day 7 (Fig. 11). The only significant difference is that the up-regulation of Bcl-XL in LNCaP cells was not as dramatic as in NHP cells. Interestingly, the anti-P53 detected a lower band in addi tion to the expected M, 53,000 protein. The molecular nature of this lower band is unclear, although it might be a proteolytic product (48). Survival of PC3 Cells in the Absence of Trophic Factors. PC3 20 â€¢¿ cells behaved like PCA cells when deprived of serum growth factors. These cells demonstrated a steady growth, which peaked on day 3 (Fig. \2A). The cell proliferation plateaued on day 4 (Fig. 1X4), when 0 1 2 3 4 5 C increased apoptosis (DAPI labeling and DNA fragmentation) was Time (days post growth factor withdrawal) observed (Fig. 12, B and C, middle panel). This increased cell death clearly negated the cell proliferation; therefore, the net number of live cells did not further increase on day 4 (Fig. \2A). Starting from day 5, B the number of live cells began to decrease (Fig. \2A), suggesting that, by this time, cell death (Fig. 125) began to surpass the cell prolifer ation. Direct quantitation of apoptosis demonstrated that by day 8, â€”¿60%ofthe cells were apoptotic (Fig. 12, B and C, righi panel). This observation was consistent with the MTS measurement showing that i approximately the same number of cells (i.e., 100% cell number) as initially input (i.e., 10,000 cells/well) were still alive (Fig. 12A) by day 8, considering that PC3 cells initially had a >2-fold proliferation (Fig. 12A). Western blotting examination of apoptosis proteins also re vealed similar alterations as observed in PCA cells in that PC3 cells

Time (days post growth factor withdrawal) Days after growth factor withdrawal

0 1

Actin -/

WAF-l P21

Bcl-2 Fig. 3. Survival of NHP cells in the absence of trophic factors. A. quantitation of live cells using the MTS assays. The percentage cell number (compared to DO, which is 100%) represents means derived from three independent experiments; bars, SE. B. quantitation of apoptosis by DAPI staining. Inset, DNA fragmentation. C, micrographs showing DAPI-stained NHP cells starved for 0 (Â¡eft).2 (middle), or 7 (right) days. Apoptotic cells were brightly stained due to chromatin condensation. Note that in the middle and righi panels, some apoptotic cells turned necrotic in culture and thus degraded. Therefore, an overall decreased cell number was noticed. The arrowheads in the left and middle panels illustrate apoptotic NHP cells, whereas the arrowheads in the right panel indicate very few live cells. X100. Bax die immediately (as did NHP and BPH cells) until days 3-4 (Fig. 10ÃŸ).Prominent apoptosis (DAPI labeling and DNA fragmentation) Bak â€”¿Â«.' was observed around day 4 (Fig. 10, B and C, middle panel), yet the total cell number (i.e., the same number as initially input) was main tained up to day 6, after which the cell number went down (Fig. 10A). (22 kd) These observations suggest that the cell loss due to apoptosis between Fig. 4. Alterations in apoptosis proteins in NHP cells after growth factor withdrawal. days 4-6 was partially compensated by cell proliferation. By day 8, Thirty u.g of whole-cell lysates prepared from unstarved NHP cells or from cells cultured â€”¿70%cellswere apoptotic (Fig. 10, A-C, right panel). Immunoblot- in the absence of growth factors for various days were resolved on a 12% SDS-PAGE. The same membrane was sequentially probed with individual or mixed primary antibodies (see ting examination of various apoptosis proteins revealed quite similar Fig. 1). Left, individual molecules. The arrowhead on the right points to the M, â€”¿30.000 patterns of alterations as observed in NHP cells. Specifically, all Bcl-XL. 3470

10.0

Vi 4) a a. vx w 5.0 'â€¢O "3Â£ 06 Ã¤sv 2.5 -

0.0 DU Dl D2 D3 D4 Ãœ5 D6 D7 DO Dl D2 D3 D4 DS D6 Time (days) Time (days)

4- e 'Z^o vt o W o. 3- X s. 4) 4)X 4) 2- â€¢¿^ â€”¿ iÂ» "wâ€”¿ 1-

â€”¿lâ€” â€”¿lâ€” â€”¿lâ€” â€”¿lâ€” DO Dl D2 D3 D4 DS D6 DO Dl D2 D3 D4 DS D6 D7 D8 Time (days) Time (days)

P21 Bcl-2 4- o Bcl-XL VI vt Bak 4l Bad 2 hl o. =- 3- X X v 9) W2: 2H

W "Â«3â€”¿ ai 1-

DO Dl D2 D3 D4 DS D6 D7 DO Dl D2 D3 D4 DS D6 D7 Time (days) Time (days)

Fig. 5. Quantitative presentation of scanning data as determined by densitometry. The intensity of each individual protein band was first determined by scanning (41) and then normalized to the intensity of actin protein band. After normalization, the individual protein levels are plotted as "relative expression" {i.e., fold changes) compared with unstarved cells (i.e.. DO samples), the protein expression levels of which are considered as \.A, NHP cells. B. BPH cells. The values for P53 are: 1, 1.8, 9.7, 30.9, 15.4, 17.5, and 15 from DO to D6, respectively. C, PCA cells. Note that the data for P53 were not plotted because the P53 protein on immunoblot overlapped with the P54/55 doublet. D, LNCaP cells. E. PC3 cells. The values for Bcl-XL are: 1, 1.1, 0.9, 1.0, 1.0, 1.2, and 1.1 from DO to D7. respectively. F. Dul45 cells. Bcl-2 represents the M, 30,000 Bcl-2 protein band.

3471

with the M, 26,000 Bcl-2 protein (Fig. 13). In contrast, the Bcl-XL protein did not change significantly (Figs. 13 and 5Â£). Survival of Dul45 Cells in the Absence of Trophic Factors. Dul45 cells demonstrated the highest survival potential (Fig. 144). Upon serum removal, these cells gradually proliferated, nearly dou bling the cell number by day 3 when DNA fragmentation was first observed (Fig. 14ÃŸ),suggesting that the cell proliferation exceeded cell death by apoptosis. The death rate in starved Du 145 cells dem onstrated a very slow kinetic (Fig. 14ÃŸ),from ~10% on day 3, to -15% on day 6 (Fig. 14C, middle), and to -25% by day 12, at which 1234567 time still about the same number (i.e., 10,000 cells/well initially Time (days post growth factor withdrawal) plated) of cells were alive (Fig. 14C, right panel). A portion (â€”5%) of the Du 145 cells in the culture survived up to 30 days after serum withdrawal (data not shown). Western blotting of apoptosis proteins revealed that the proapoptotic proteins Bak and Bad did not undergo significant changes in their expression levels (Figs. 15 and 5F), as in

Days after serum withdrawal n 3 4 6

P53 l 234567

Time (days post growth factor withdrawal) WAF-l P21

Bcl-2

Bax Fig. 6. Survival of BPH cells in the absence of trophic factors. A, quantitation of live cells using the MTS assays. The percentage cell number (compared to DO. which is 100%) represents means derived from three independent experiments; bars, SE. B, quantitation of apoptosis by DAPI staining. Inset. DNA fragmentation. C micrographs showing DAPI-stained NHP cells starved for 0 (left), 2 (middle), or 7 (right) days. Apoptotic cells were brightly stained because of chromatin condensation. Note that in the middle and right Bak - â€ž¿ panels, some apoptotic cells turned necrotic in culture and thus degraded. Therefore, an overall decreased cell number was noticed. The arrowheads in the left and middle panels illustrate apoptotic BPH cells, whereas the arrowheads in the right panel indicate very few live cells. X100.

Bad lost their response in up-regulating all three proapoptotic proteins upon serum deprivation (Figs. 13 and 5Â£).In fact, the Bax protein actually went down dramatically by day 6 (Figs. 13 and 5Â£).Also similar to PCA cells, the P21 levels significantly increased by day 3 Actin prior to the initiation of DNA fragmentation (Figs. 13 and 5Â£). Another aspect of the similarity between PC3 cells and PCA cells was revealed by the detection of the P54/55 doublet in PC3 cells on day 5 after serum starvation (Fig. 13). Interestingly, the P25 band detected Fig. 7. Alterations in apoptosis proteins in BPH cells following growth factor with drawal. Thirty Â¿tgof whole cell lysates prepared from unstarved BPH cells or from cells in PCA cells was not observed in PC3 cells (Fig. 13). Different from cultured in the absence of growth factors for various days were resolved on a 12% PCA cells, PC3 cells expressed the Mr 26,000 Bcl-2 protein, the levels SDS-PAGE. The same membrane was sequentially probed with individual primary antibodies (see Fig. 1). Left, individual molecules. The arrowheads on the right point to of which significantly increased upon serum starvation (Figs. 13 and the M, -30,000 Bcl-X, and M, -22,000 Bad, respectively. Note the upper band in the 5E). The M, 27,000 Bcl-2 protein levels also elevated simultaneously Bad blot was incompletely deprobed Bak. 3472

A 3fl0 (37). All three types of cells were diploid in karyotype, positive for both keratinocytes and PSA, and negative for ras and p53 mutations 25*' (37). When cultured in defined medium containing various growth

200 factors, these cells demonstrated progressively accelerating growth properties; the population doubling times of NHP, BPH, and PCA lso cells were 51, 37, and 29 h, respectively (37). Here we compared the ,â€ž growth and survival properties of these primary prostate cells with those of three established prostate cancer cell lines in the absence of 50 trophic factors. We first characterized the expression of several apoptosis-related 2 4 6 g 10 12 proteins in different types of prostate cells, from normal to malignant. Time (days post growth factor withdrawal) All cells express P21, Bcl-XL, Bad, and Bak, and all cells except Du 145 express Bax protein. However, primary cells generally express little or no Bcl-2, whereas LNCaP cells express abundant Bcl-2 protein. In PC3 and Du 145 cells, a more complicated pattern of Bcl-2 expression was observed. In addition to the regular Mr 26,000 Bcl-2, PC3 cells also express two minor upper bands that migrate at M, 27,000 and M, 30,000, respectively. In contrast, Du 145 cells

Days post growth factor withdrawal 1

23456789 i ii 0123451 6 Time (days post growth factor withdrawal) P53 P54/55 Actin-/ i

P21 WAF-l

Bcl-2

Fig. 8. Survival of PCA cells in the absence of trophic factors. A, quantitation of live cells using the MTS assays. The percentage cell number (compared to DO. which is 100%) represents means derived from three independent experiments; bars, SE. B, quantitation Bcl-XL â€”¿â€”^ of apoptosis by DAP] staining. Inset, DNA fragmentation. C. DAPI-stained NHP cells starved for 0 (left), 5 (middle), or 11 (right) days. Apoptotic cells were brightly stained due to chromatin condensation. The arrowheads illustrate apoptotic PCA cells. X100.

PCA and PC3 cells. Also different from the PCA/PC3 cells, Du 145 did not express Bax protein, as tested by several antibodies raised Bax against different regions of the molecule (Fig. 15; data not shown). The most salient feature is the expression of the Mr 30,000 Bcl-2 protein in Du 145 cells, the levels of which increased steadily after serum starvation (Figs. 15 and 5F). Other aspects in the changes of Bak apoptosis proteins after serum starvation in Du 145 cells were similar to those observed in PC3 cells. Specifically, P21 protein levels went up (nearly 2-fold), whereas the Bcl-XL underwent a slight increase (Figs. 15 and 5F). Also, like in PC3 cells, the P54/55 doublet was Bad ___. detected in Du 145 cells on day 3 after serum starvation (Fig. 15). Interestingly, the P25 protein band was also detected (Fig. 15).

Fig. 9. Alterations in apoptosis proteins in PCA cells after growth factor withdrawal. DISCUSSION Thirty fig of whole-cell lysates were prepared from unslarved PCA cells or from cells cultured in the absence of growth factors for various days were resolved on a 12% SDS-PAGE. The same membrane was sequentially probed with individual primary Recently, NHP, BPH, and PCA cells were established, and their antibodies (see Fig. 1). Left, individual molecules. The arrowhead on the right points to growth properties and nutritional requirements were characterized the M, -30,000 Bcl-X,. Right, novel P25 and P54/55 proteins. 3473

treatment established the Mr 27,000 band but not the Mr 30,000 band as the phosphorylated Bcl-2 protein. Therefore, the Mr 30,000 Bcl-2 protein identified in PC3 and Du145 cells is different from the Mr 30,000 Bcl-2 in HT29 cells, which is a hyperphosphorylated form of the Mr 26,000 Bcl-2 (46). It is not yet known whether the Mr 30,000 Bcl-2 protein in PC3/Dul45 cells is or is related to the newly reported Mr 30,000 BRAG-1, a Bcl-2-related protein primarily expressed in the brain (47). Both Mr 27,000 and Mr 30,000 Bcl-2 proteins are detected in PC3 and/or Du145 cells without any prior stimulation, suggesting that these proteins are endogenously expressed in both cells. 2468 Time (days post serum withdrawal)

B 10Â° Days after serum withdrawal MO 1 2 3 4 5 (days) I I 01234567 8

4Â»

20 WAF-l P21 2468 10 Time (days post serum withdrawal)

Bcl-2

Bel-XL

Bax ^â€ž.Â»-Â»-^^

Fig. 10. Survival of LNCaP cells in the absence of trophic factors. A, quantitation of live cells using the MTS assays. The percentage cell number (compared to DO, which is 100%) represents means derived from three independent experiments; bars, SE. B, Bak quantitation of apoptosis by DAPI staining. Inset, DNA fragmentation. C, DAPI-stained NHP cells starved for 0 (left), 4 (middle), or 8 (right) days. Apoptotic cells were brightly stained because of chromatin condensation. The arrowheads illustrate apoptotic LNCaP cells. XI00. Bad appear to primarily express the Mr 30,000 band with very low levels of the Mr 26,000 Bcl-2 protein. It is well known that Bcl-2 can undergo phosphorylation on serine residues, resulting in either Mr 27,000 Bcl-2 in multiple cell types such as leukemia, MCF-7, and PC3 cells (42-46) or Mr 30,000 Bcl-2 in HT-29 colon cancer cells (46). The Mr 30,000 Bcl-2 protein has not been reported in either PC3 or Du145 cells, even after treatment with taxol or okadaic acid (42,43, 45). We therefore further characterized the Mr 27,000 and A/r30,000 bands detected in PC3 and/or Du145 cells by the monoclonal anti- Fig. 11. Alterations in apoptosis proteins in LNCaP cells after growth factor with drawal. Thirty (ig of whole-cell lysates prepared from unstarved LNCaP cells or from Bcl-2 antibody. Immunoprecipitation followed by Western blotting cells cultured in the absence of serum for various days were resolved on a 12% first established the Mr 30,000 band in both cell types as a specific SDS-PAGE. The same membrane was sequentially probed with individual primary antibodies (see Fig. 1). Left, individual molecules. The arrowhead on the right points to Bcl-2-related protein instead of nonspecific antibody binding. Subse the Mr -30,000 Bcl-XL. Note that the anti-P53 detected the expected M, 53,000 protein quently, a dephosphorylation experiment with alkaline phosphatase and a lower band, whose identity is presently unknown. 3474

in the absence of trophic factors, than in PC3/PCA/Dul45 cells 300 (summarized in Fig. 16), suggesting that cancerous prostate cells 250 naturally possess a longer life span (or survivability) than NHP/BPH cells. 200 LNCaP cells, initially isolated from the lymph node metastasis of a 150 prostate cancer patient, demonstrate a survivability between that of

100 * NHP/BPH and that of PC3/PCA/Dul45 cells. These cells behave differently from other cancer cells in that they do not proliferate significantly in the absence of trophic factors. However, LNCaP cells also behave differently from NHP/BPH cells in that they do not die 2 4 6 H immediately after serum withdrawal. Instead, these cells keep a rela Time (days after serum withdrawal) tively constant cell number for 5-6 days, although the apoptosis is initiated at approximately day 4. The mechanism(s) for the unique survival characteristics of LNCaP cells are presently unknown. Taken 100 B Days post serum together, the present data suggest that prostate cancer cells generally withdrawal demonstrate longer survivability than their normal counterparts due to 80 6543210 increased proliferative capability and decreased cell death rate. The L co increased cell proliferation results from the production by cancer cells .2 of various growth factors (49), whereas the decreased death rate

Ã¤. â€¢¿< 20" Days after serum withdrawal

2468 10 234 6 7 Time (days after serum withdrawal)

P21WAF-l

Bcl-Xt

Fig. 12. Survival of PC3 cells in the absence of trophic factors. A. quantitation of live Bax cells using the MTS assays. The percentage cell number (compared to DO, which is 100%) represents means derived from six independent experiments; bars, SE. B. quantitation of apoptosis by DAP1 staining. Inset. DNA fragmentation. C, DAPl-stained NHP cells starved for 0 (left). 4 (middle), or 8 (right) days. Apoptotic cells were brightly stained due to chromatin condensation. X 100. Bak â€”¿

Growth and Survival of Various Prostate Cells upon Trophic Factor Withdrawal. NHP and BPH cells do not grow and demon Bad _ Ã„ strate the lowest survival capacity upon trophic factor withdrawal. Apoptosis is observed in these cells as early as 24 h after growth factor withdrawal. In contrast, primary (PCA) and metastatic (PC3 and Dul45) carcinoma cells survive much longer, and these cells keep Actin _. . ^ dividing and proliferating in the absence of trophic factors. These three cancer cell types follow very similar growth/survival patterns upon removal of trophic factors: an initial aggressive growth phase Fig. 13. Alterations in apoptosis proteins in PC3 cells after scrum withdrawal. Thirty (the first 2-3 days) in which apoptosis rarely occurs; a second "coun Hg of whole-cell lysates prepared from unstarved PC3 cells or from cells cultured in the terbalancing" phase in which cell loss by apoptosis (taking place on absence of serum for various days were resolved on a 12% SDS-PAGE. The same membrane was sequentially probed with individual primary antibodies (see Fig. 1). Left, days 3-4) is compensated by sustained cell proliferation; and a third individual molecules. The arrowhead on the right points to the M, -30.000 Bcl-X,. The "death" phase in which apoptosis surpasses the cell growth. Direct Bcl-2 proteins and Ihe P54/55 doublet are also labeled on the riditi. Note that Lanes 2 and 3 (i.e., days I and 2) loaded less protein, as evidenced by the actin band. Thus, the measurement of cell death by DAPI staining and MTT assays (not intensities of these two lanes and all of the rest of the lanes were normali/ed against that shown) has demonstrated a much higher death rate in NHP/BHP cells. of the actin band. 3475

trum of human tumor cells (4). In this study, there appears to have a correlation between the P53 response and apoptotic sensitivity of various prostate cells. NHP, BPH, and LNCaP cells, which are the cell types sensitive to apoptosis induction by trophic factor withdrawal, demonstrated a time-dependent increase in P53 protein expression. BPH cells demonstrated the most robust (~30-fold increase by day 3) and longest up-regulation of P53, and these cells are most sensitive to 3 apoptosis induction. In contrast, PCA cells, although expressing wild- type P53, do not demonstrate up-regulation of P53, and these cells survive much longer than NHP, BPH, and LNCaP cells (Fig. 16). 2468 10 Similarly, PC3 and Dul45 cells, which do not express functional P53, Time (days after serum withdrawal) also possess longer survivability in the absence of trophic factors. Taken together, these observations correlate an up-regulation of func tional P53 with increased sensitivity of prostate cells to apoptosis B 100 Days post serum withdrawal induction by deprivation (Fig. 16). 0123-156 Quite contrasting alterations are also seen with P21 in sensitive

Ml Days after serum withdrawal

40 012345 67

WAF-l 2468 10 P21 - Time (days after serum withdrawal)

P54/55 Bcl-2 30kd-Bcl-2

Bel-XL

Bax Fig. 14. Survival of Du 145 cells in the absence of trophic factors. A. quantitation of live cells using the MTS assays. The percentage cell number (compared to DO. which is 100%) represents means derived from four independent experiments: bars, SE. B, quantitation of apoptosis by DAPI staining. Insel, DNA fragmentation. C, DAPI-stained NHP cells starved for 0 (left}, 6 (middle), or 12 (righi) days. Apoptotic cells were brightly stained because of chromatin condensation. X100. results from evasion of apoptosis. It is of note that PCA, one of the primary cell types used, demonstrates even higher survivability than Bad â€”¿ LNCaP and PC3 cells, which are established cell lines kept in culture for many generations. Therefore, it is unlikely that any of the differ ences observed among these cells may have resulted from tissue culture artifacts. Actin â€¢¿â€¢Â»â€”¿â€”¿â€”¿â€¢Â« Role of p53 and P21 in Prostate Cell Growth/Apoptosis upon Trophic Factor Withdrawal. p53 plays two essential roles in main taining the cellular homeostasis: introducing "checkpoints" to arrest cell cycle; and triggering apoptosis to self-destroy those cells beyond Fig. 15. Alterations in apoptosis proteins in Du 145 cells after serum withdrawal. Thirty repair (50). p53 mediates its biological functions by transcriptionally fig of whole-cell lysates prepared from unstarved Du 145 cells or from cells cultured in the absence of serum for various days were resolved on a 12% SDS-PAGE. The same regulating (either activating or inhibiting) many of its targets, among membrane was sequentially probed with individual primary antibodies (see Fig. 1). Left, which is p21, a CIP/KIP family cyclin-dependent kinase inhibitor. In individual molecules. The arrowhead on the right points to the A/r â€”¿30,000Bcl-X,. The response to genotoxic shocks, cells generally halt cell cycle progres M, 30,000 Bcl-2, P25, and P54/55 doublet are also labeled on the right. Note that the sion by up-regulating P21 protein in a p53-dependent manner. Over- regular Mr 26.000 Bcl-2 protein was barely detectable in this experiment. Note that Lane 2 (i.e., day 1) loaded less protein, as evidenced by actin band. Thus, the intensities of this expression of wild-type p53 induces apoptotic death of a wide spec- lane and all of the rest of the lanes were normalized against that of the actin band. 3476

nullP21:A mutantP21:A Awt â€”¿ p53:iP21:-Bcl-2: p53:AP21:fBcl-2: Bcl-2:ABcl-XL: RiiBcl-2: Bcl-2:ABcl-XL: 1:fBcl-2: â€”¿Bad/Bak: little/noBcl-XLlBax/Bad/Bak:ABad/Bak: little/noBcl-XL:ABax/Bad/Bak:AGrowthlittle/noBcl-XL: A,Bcl-XL:ABax/Bad/Bak:â€”¿Bax:JP25/P54/55: â€”¿Bax: A.Bax/Bad: â€”¿P25/P54/55: deletedP25/P54/55: AGrowth AGrowth yesGrowth yesGrowth yesGrowth Fig. 16. A, the overall survivability of various -Apop -Apop +Apop +++Apop +++Apop +++Apop prostate cells in the absence of trophic factors is +++BPH +++NHP ++i ++Â£PC3~^0wtp53:+â€¢ +â€¢Dul45 presented: BPH < NHP < LNCaP < PC3 < PCA < Dul45. Shown above are changes in apoptosis -^wt ~^0wtp53:AP2LNCaP-^p53: PCA -^p53: proteins. Refer to text for details. â€”¿,nochange; upward arrowhead, increase; downward arrow head, decrease; little/no, little or no expression. B Normal Prostate Epithelium Shown in the middle are the cell growth and apop tosis properties of various prostate cells in the ab androgen-dependence No cell sence of trophic factors, as deduced from preceding growth factor dependence figures. â€”¿,nogrowth. + to +4-+, different levels longtimewild-type population doubling proliferation of apoptosis. BPH and NHP cells do not proliferate >V Decreased upregulation .â€”^ cell number at all and demonstrated fast apoptotic kinetics. LN- P53 P21: no change or decrease â€¢¿Apoptosis CaP cells demonstrated a low level of proliferation litue/no Bcl-2DeprivationDeprivationP53Bax/Bad/Bakupregulalion and an intermediate level of apoptosis. PC3 cells Bcl-2: little/no expression demonstrated very aggressive growth and interme diate levels of apoptosis. PCA and Du 145 cells, in contrast, demonstrated a rapid cell proliferation as well as a much delayed and decreased levels of apoptosis. B, summary of molecular events leading Prostate Cancer Cells to differential survivabilities of normal versus can cerous prostate epithelial cells in the absence of androgen-independence Â£e^No trophic factors. Note thai there exist exceptions to autocrine growth factor production this generalized scheme. For example, PCA cells shorttimenon-functional population doubling possess wild-type P53 and express little/no Bcl-2. P53 P53 upregulalion â€”¿^ survivabili ty highBcl-2low/no P21 upregulation of -^ BaxDeprivationDeprivation Bax/Bad/Bak:decrease .^ apoptosisT^Extended or deletion Bcl-2: upregulation Novel proteins: P25; P54/55 (?)proliferationEvasion

(i.e.. BPH, NHP, and LNCaP) versus resistant (i.e.. PCA, PC3, and calcineurin), or the death-promoting activity of Bad (through Du 145) cells. In the former, upon trophic factor withdrawal, P21 Raf-1; Refs. 55 and 56). The precise mechanism(s) for the anti- either does not change significantly (BPH) or gradually decreases death effects of Bcl-2/Bcl-XL is still unknown and most likely is (NHP and LNCaP) while in the latter; the P21 levels invariably cell type and inducer dependent. Bcl-2 has been shown to be increase either prior to or simultaneously with the appearance of DNA phosphorylated by Raf-1 and some other protein kinases, and this fragmentation. These quite distinct changes in the two groups of cells phosphorylation has been hypothesized to inactivate the cytopro- suggest that: (a) the P21 expression is unrelated to or independent of tective effect of Bcl-2 (42-46). In addition, both Bcl-2/Bcl-XL and the P53 status, which is consistent with data showing that P21 ex Bax have been shown to be transcriptionally regulated by p53 (57, pression can be p53-independent (50, 51); (b) the involvement of P53 58). Increased Bcl-2/Bcl-X, protein levels are generally associated and P21 in apoptosis of prostate cells is independent of each other, as with cytoprotection and extended cell survival. In this study, reported by others in oncogene-mediated fibroblast apoptosis (52); NHP/BPH/PCA cells express little or no Bcl-2, and growth factor and (c) increased P21 expression is protecting PCA/PC3/Dul45 cells withdrawal fails to up-regulate Bcl-2, although all three cell types from apoptosis, as observed in other cell systems (53, 54). The fact express wild-type p53. In contrast, Bcl-2 in LNCaP (wild-type that P21 levels go up before apoptosis suggests a causal rather than a p53) and PC3 cells (p53-null) steadily and significantly increases coincidental relationship between the two phenomena. The increased upon deprivation. More interestingly. Du 145 cells primarily ex P21 could be inducing a cell cycle arrest because the peak level of press the Mr 30,000 Bcl-2. a very stable protein, the levels of which induction of this protein also coincides with the time when cells stop remain persistently elevated throughout the starvation period. With proliferation. It is known that dividing cells are much more prone to respect to Bcl-X,, all cells except PC3 demonstrate a certain apoptosis induction (55); therefore, it is quite likely that the increased degree of up-regulation of this protein in response to starvation. P21 will put a "brake" on the proliferative machinery in prostate Based on these results, it appears that; (a) the up-regulation of cancer cells to avoid "mitotic catastrophe" (55). Bcl-2 and/or Bcl-X, in various prostate cells represents a general Role of Bcl-2 and Bcl-X, in Prostate Cell Growth/Apoptosis defensive response to stresses such as deprivation. The up-regula upon Trophic Factor Withdrawal. Bcl-2 and Bcl-XL are two tion of Bcl-X, may be the major defensive (survival) mechanism in members in the Bcl-2 family that possess anti-apoptosis functions. NHP cells; the significant up-regulation of this protein in NHP Many biological activities have been assigned to these proteins for cells may also explain why these cells are relatively more resistant their death-sparing effects: forming membrane channels; modulat to apoptosis induction than BPH cells. However, the cytoprotective ing mitochondria permeability transition pore; retarding cyto- effect of up-regulated Bcl-XL will be overridden by the simulta chrome c and/or AIF apoptosis-inducing factor release from mito neous up-regulation of Bax, Bad. and Bak: (/?) the up-regulation of chondria; working as antioxidants; preventing calcium fluctuation; Bcl-2 and/or Bcl-X, may not depend on p53: and (c) the differ sequestering proapoptotic proteins (such as Bax. Bak, and Bik); ential up-regulation of Bcl-2 and/or Bcl-X, in various prostate and indirectly modulating caspase activity (through CED-4). cell cells might partially contribute to their differential survival abili cycle (through p53BP-2), NF-AT transcriptional activity (through ties (Fig. 16). 3477

Role of Bax, Bad, and Bak in Prostate Cell Growth/Apoptosis REFERENCES upon Trophic Factor Withdrawal. In contrast to Bcl-2 and Bcl-XL, i. Raff. M. C. Social controls on cell survival and cell death. Nature (Lond.), 356: proapoptotic proteins Bax, Bad, and Bak demonstrate very distinct 397-400, 1992. alterations in their expression levels that highly correlate with the Raff. M. C.. Barres. B. A., Bume, J. F., Coles, H. S.. Ishizaki, Y., and Jacobson, M. D. Programmed cell death and the control of cell survival: lessons from the nervous sensitivity to apoptosis induction in individual cell types (Fig. 16). system. Science (Washington DC), 262: 695-700. 1993. Specifically, upon growth factor withdrawal, all sensitive cells, i.e., Denmead. S. R., Lin, X. S.. and Isaacs. J. T. Role of programmed (apoptotic) cell NHP, BPH, and LNCaP, demonstrate a time-dependent up-regulation death during the progression and therapy for prostate cancer. Prostate, 28: 251-256. of these proteins. In contrast, long-living cells (PCA, PC3, and 1996. Tang. D. G., and Porter, A. T. Target to apoptosis: a hopeful weapon for prostate Du 145) invariably demonstrate a loss of this up-regulation response. cancer. Prostate, 32: 284-293, 1997. More dramatically, PC3 cells demonstrate a gradual decrease, and McDonnell, T. J.. Troncoso, P., Brisby, S. M., Logothetis, C. L., Chung, L. W. K., Hsieh, J. T., Tu. S. M., and Campbell, M. L. Expression of the protooncogene Bcl-2 Du 145, the most resistant cells, demonstrate a complete loss in Bax in thÃ¨prostate and its association with emergence of androgen-independent prostate expression. The Bax/Bad/Bak protein expression appears to be cor cancer. Cancer Res., 52: 6940-6944, 1992. Colombe!, M., Symmans, F., Gil, S., O'Toole, K. M., Choplin, D., Benson, M., related with the p53 functionality; their levels are increased in cells Olsson. C. A.. Korsmeyer. S., and Buttyan. R. Detection of the apoptosis-suppressing with wild-type p53 (NHP, BPH, and LNCaP), and their levels show oncoprotein Bcl-2 in hormone-refractory human prostate cancers. Am. J. Pathol.. 143: no increase or show decrease or loss in cells with no or mutant p53 390-400, 1993. Liu, A. Y., Corey, E., Bladou, F., Lange, P. H., and Vessella, R. L. Prostatic cell (PC3 and Du 145). PCA cells, although expressing functional p53, lineage markers: emergence of Bcl2* cells of human prostate cancer xenograft somehow do not up-regulate this protein in response to deprivation LuCaP 23 following castration. Int. J. Cancer, 65: 85-89. 1996. and therefore resemble PC3 and Du 145 cells in terms of p53 func Fuyuya, Y., Krajewski, S., Epstein. J. !.. Reed, J. C., and Isaacs. J. T. Expression of Bcl-2 and progression of human and rodent prostatic cancers. Clin. Cancer Res.. 2: tionality. 389-398, 1996. Role of Novel Proteins in Prostate Cell Growth/Apoptosis upon McConkey, D. J.. Greene. G.. and Pettaway. C. A. Apoptosis resistance increases with Trophic Factor Withdrawal. In addition to the novel M, 30,000 metastatic potential in cells of the human LNCaP prostate carcinoma line. Cancer Bcl-2 protein detected in Du 145 cells, several other proteins, specif Res., 56: 5594-5599, 1996. 10. Berchem, G. J., Bossier. M., Sugars, L. Y.. Voeller. H. J., Zeitlin. S.. and Gelman. ically P25 and P54/55 doublet, are also detected in PCA, PC3, and E. P. Androgens induce resistance to bcl-2-mediated apoptosis in LNCaP prostate Du 145 cells. These proteins appear together at the same time and can cancer cells. Cancer Res., 55: 735-738, 1995. Raffo, A. J., Perlman, H.. Chen. M-W., Day, M. L., Stritman, J. S., and Buttyan. R. be detected by at least two unrelated monoclonal antibodies, i.e., Overexpression of bcl-2 protects prostate cancer cells from apoptosis in vitro and monoclonal anti-P53 (in PCA cells; Fig. 9) and monoclonal anti-Bcl-2 confers resistance to androgen depletion in vivo. Cancer Res., 55: 4438-4445, 1995. (in PC3 and Du 145 cells; Figs. 13 and 15). Based on these charac 12. Isaacs. W. B., Carter. B. S.. and Ewing, C. M. Wild-type p53 suppresses growth of human prostate cancer cells containing mutant p53 alÃeles.Cancer Res.. 51: 4716- teristics and their apparent molecular weights, these proteins appear to 4720. 1991. be immunoglobulin heavy (M, -55,000) and light (M, -25,000) Eastman, J. A.. Hall, S. J.. Sehgal. Wang, I. J.. Tirarne, T. L., Yang, G.. Connell- chains (see also Fig. 2). Why these "immunoglobulin molecules" are Crowley, L.. Elledge. S. J., Zhang, W-W.. Harper. J-W.. and Thompson, T. C. In vivo gene therapy with p53 or p21 adenovirus for prostate cancer. Cancer Res.. 55: up-regulated is unclear. However, the fact that they are only up- 5151-5155. 1995. regulated in resistant cells (PCA, PC3, and Dul45) and their regula 14. Yang. C., Cirielli, C., Capogrossi, M. C., and Passanti. A. Adenovirus-mediated wild-type p53 expression induces apoptosis and suppresses tumorigenesis of prostatic tion follows a distinct pattern (i.e., around the time when cells initiate tumor cells. Cancer Res., 55: 4210-4213, 1995. apoptotic programs) point to an intriguing possibility that they might 15. Sells, S. F.. Muthukkumar. S., Maddiwar, N.. Johnstone. R., Boghaert. E., Gillis. D., be causally involved in extending the survival of prostate cancer cells. Liu, G.. Nair. P., Monnig. S., Collini. P.. Mattson, M. P., Sukhatme, V. P., Zimmer, S. G., Wood, D. P., Jr., McRoberts, J. W., Shi, Y., and Rangnekar, V. M. Expression Synopsis. The present study uses a simple model to compare the and function of the leucine zipper protein par-4 in apoptosis. Mol. Cell. Biol., 17: growth/survival characteristics of normal versus cancerous prostate 3823-3832. 1997. cells cultured in the absence of trophic factors. Fig. 16 summarizes the Day, M. L., Foster, R. G.. Day. K. C., Zhao, X.. Humphrey, P., Zhang. S. H.. and Dean, D. C. Cell anchorage regulates apoptosis through the retinoblastoma tumor sequential molecular events that occur as these cells progress from a suppressor/E2F pathway. J. Biol. Chem., 272: 8125-8128. 1997. sensitive to a resistant stage with respect to apoptosis induction. NHP 17. Hsing. A. Y., Kadomatsu. K., Bonham. M. J.. and Daneilpour. D. Regulation of and BPH cells express wild-type p53 and little/no Bcl-2. In response apoptosis induced by transforming growth factor-ÃŸl in nontumorigenic and tumori- genie rat prostatic epithelial cell lines. Cancer Res., 56: 5146-5149, 1996. to growth factor withdrawal, these cells stop proliferation and rapidly 18. Sensibar, J. A., Sutkowski. D. M., Raffo, A.. Buttyan, R.. Griswold, M. D., Sylvester. up-regulate p53, which may then up-regulate proapoptotic proteins S. R., Kozlowski. J. M.. and Lee, C. Prevention of cell death induced by tumor Bax/Bad/Bak. Cells rapidly enter an apoptotic phase. In contrast, necrosis factor a in LNCaP cells by Overexpression of sulfated glycoprotein-2 (clusterin). Cancer Res.. 55: 2431-2437, 1995. prostate cancer cells, upon deprivation, still proliferate aggressively 19. Rokhlin, O. W.. Bishop, G. A.. Hostager, B. S., Waldschmidt. T. J.. Sidorenko, S. P.. due to autocrine production of various growth factors. These cells Pavloff. N., Kiefer, M. C., Umanski, S. R., Glover, R. A., and Cohen, M. B. (except PCA) endogenously express much higher levels of Bcl-2. Fas-mediated apoptosis in human prostatic carcinoma cell lines. Cancer Res., 57: 1758-1768. 1997. They express either mutant p53 (PC3 and Dul45) or wild-type p53, 20. Borner. M. M., Myers, C. E., Sartor, O., Sei, Yâ€žToko, T., Trepel, J. B.. and which somehow does not respond to starvation by up-regulating its Schneider. E. Drug-induced apoptosis is not necessarily dependent on macromolec- expression (PCA), leading to the loss of up-regulation or decrease of ular synthesis or proliferation in the p53-negative human prostate cancer cell line PC-3. Cancer Res., 53: 2122-2128, 1995. proapoptotic proteins. Furthermore, cancer cells such as Du 145 even Sinha, B. K.. Yamazaki, H., Eliot, H., Schneider. E.. Borner, M. M.. and O'Connor, completely delete a proapoptotic protein (i.e., Bax). At the same time, P. M. Relationships between proto-oncogene expression and apoptosis induced by these cells up-regulate P21, which could induce cell cycle arrest at an anticancer drugs in human prostate tumor cells. Biochim. Biophys. Acta. 7270: 12-18. 1995. appropriate time (because proliferating cells are generally more sen 22. Vukanovic, J., and Isaacs, J. T. Human prostatic cancer cells are sensitive to sitive to apoptosis induction; Ref. 55), and Bcl-2, which by itself programmed (apoptotic) death induced by the antiangiogenic agent Linomide. Cancer could extend cell survival. Last but not least, the long-living cells Res., 55: 3517-3520, 1995. 23. Li, C. J., Wang, C., and Pardee, A. B. Induction of apoptosis by ÃŸ-lapachone in express some novel immunoglobulin or immunoglobulin-like proteins human prostate cancer cells. Cancer Res., 55: 3712-3715, 1995. that might also help maintain the cell survivability. Cells with inter 24. PlanchÃ³n, S. M., Wuerzberger, S.. Frydman. B., Witiak. D. T., Hutson, P.. Church, D. P.. Wilding. G.. and Boothman, D. A. /3-Lapachone-mediated apoptosis in human mediate life span, i.e., LNCaP, demonstrate a variety of characteristics promyelocytic leukemia (HL-60) and human prostate cancer cells: a p53-independent lying between the above two cell types. Sustained proliferation, to response. Cancer Res., 55: 3706-3711, 1995. gether with multiple molecular mechanisms to evade apoptosis, leads 25. Carducci, M. A., Nelson. J. B.. Chan-Tack, K. M., Ayyagari, S. R.. Sweatt, W. H., Campbell. P. A.. Nelson. W. G., and Simons, J. W. Phenylbutyrate induces apoptosis to significantly extended survivability of prostate cancer cells, com in human prostate cancer and is more potent than phenylacetate. Clin. Cancer Res., 2: pared with their normal counterparts (Fig. 16). 379-387. 1996. 3478

26. Wright. P. S.. Cross-Doersen. D.. Th'ng. J. P. H., Guo, X-W., Crissman, H. A.. GSH: role of 12-lipoxygenase in regulating tumor cell survival. J. Cell Physiol., 172: Bradbury. E. M., Montgomery. L. R.. Thompson, E. Y.. Loudy. D. E., Johnston. J. O.. 155-170, 1997. and Bitoni. A. J. A ribonucleotide reducÃaseinhibitor. MDL 101,731, induces apop- 40. Mo.sman. T. Rapid colorimetrie assay for cellular growth and survival: application to tosis and elevates TRPM-2 mRNA levels in human prostate tumor xenografts. Exp. proliferation and cytotoxicity assays. J. Immunol. Methods. 65: 55-63. 1983. Cell Res., 222.- 54-60, 1996. 41. Tang. D. G., Chen, Y. Q.. Newman. P. J.. Shi, L., Gao, X.. Diglio. C. A., and Honn, 27. Robertson, C. N.. Robertson. K. M.. Padilla, G. M., O'Brien, E. T., Cook, J. M., Kim, K. V. Identification of PECAM-1 in solid tumor cells and its potential involvement C-S., and Fine, R. L. Induction of apoplosis by dielhylstilbestrol in hormone- in tumor cell adhesion to endothelium. J. Biol. Chem., 268: 22882-22894. 1993. insensitive prostate cancer cells. J. Nati. Cancer Inst., X8: 908-917. 1996. 42. Haldar, S., Jena, N., and Croce, C. M. Inactivation of Bcl-2 by phosphorylation. Proc. 28. Grove, K. L., and Cheng, Y-C. Uptake and metabolism of the new anticancer Nati. Acad. Sci. USA. 92: 4507-4511. 1995. compound ÃŸ-L-(-)-dioxolane-cytidine in human prostate carcinoma DU-145 cells. 43. Haldar. S., Chintapalli. J.. and Croce. C. M. Taxol induces hcl-2 phosphorylation and Cancer Res., 56: 4187-4191, 1996. death of prostate cancer cells. Cancer Res., 56: 1253-1255, 1996. 29. Robertson, K. M.. Penland. S. N., Padilla. G. M., Selvan, R. S.. Kim, C-S., Fine. R. L., 44. Chen. C-Y.. and Faller. D. V. Phosphorylation of Bcl-2 protein and association with and Robertson, C. N. Fenretinide. Induction of apoptosis and endogenous transform p2lR" in Ras-induced apoptosis. J. Biol. Chem.. 271: 2376-2379, 1996. ing growth factor ÃŸin PC-3 prostate cancer cells. Cell Growth Differ.. 8: 101-111, 45. Blagoskonny, M. V.. Schulte. T.. Nguyen, P.. Trepel. J.. and Neckers. L. M. Taxol- 1997. induced apoptosis and phosphorylation of Bcl-2 protein involves c-Raf-1 and repre 30. Day, M. L., Zhao, X.. Wu, S., Swanson, P. E.. and Humphrey, P. A. Phorbol sents a novel c-Raf-1 signal transduction pathway. Cancer Res.. 56: 1851-1854. ester-induced apoptosis is accompanied by NGF1-A and c-fos activation in androgen- 1996. sensitive prostate cancer cells. Cell Growth Differ.. 5: 735-741. 1994. 46. Guan. R-J.. Moss. S. F.. Arber. N.. Krajewski, S., Reed. J. C.. and Holt. P. R. 30 Kd 31. Powell, C. T., Brittis. N. J.. Stec, D.. Hug. H.. Heston, W. D. W., and Fair, W. R. phosphorylated form of Bcl-2 protein in human colon. Oncogene. 12: 2605-2609, Persistent membrane translocation of protein kinase Ca during 12-O-tetradecanoyl- 1996. phorbol-13-acetate-induced apoptosis of LNCaP human prostate cancer cells. Cell 47. Das. R.. Reddy. E. P., Chatterjee, D., and Andrews, D. W. Identification of a novel Growth Differ., 7: 419-428, 1996. Bcl-2 related gene, BRAG-1. in human glioma. Oncogene, 12: 947-951. 1996. 32. Rusnak, J. M., and Lazo. J. S. Downregulation of protein kinase C suppresses 48. Kubbutat. M. H. G., and Vousden, K. H. Proteolytic cleavage of human p53 by induction of apoptosis in human prostatic carcinoma cells. Exp. Cell Res., 224: calpain: a potential regulator of protein stability. Mol. Cell. Biol., 17: 460-468. 1997. 189-199, 1996. 49. Culic, Z.. Hobisch, A., Cronauer, M. V.. Radmayr, C., Hittmair, A., Zhang, J.. 33. Huyuya. Y., Lundmo. P.. Short, A. D., Gill, D. L.. and Isaacs, J. T. The role of Thumher. M.. Bartsch. G.. and Klocker. H. Regulation of prostatic growth and calcium, pH, and cell proliferation in the programmed (apoptotic) death of androgen- function by peptide growth factors. Prostate, 28: 392-405, 1996. independent prostatic cancer cells induced by thapsigargin. Cancer Res., 54: 6167- 50. Ko. L. J.. and Prives. C. P53: puzzle and paradigm. Genes Dev.. 10: 1054-1072, 6175, 1994. 1996. 34. Garde. S.. Basrur. V.. Finkelman. M.. Li. L.. Krishan. A.. Ben-Josef. E.. Haddad. M.. 51. Levine, A. p53, the cellular gatekeeper for growth and division. Cell, 88: 323-331, Taylor. J. D., Porter, A. T.. and Tang. D. G. Prostate secretory protein (PSP94) 1997. suppresses the growth of human androgen-independent prostate adenocarcinoma cells 52. Attardi. L. D., Lowe. S. W.. Brugarolas. J., and Jacks. T. Transcriptional activation by in vitro and in \'ivu by inducing apoptosis. Proc. Am. Soc. Clin. Oncol., 76; A1970, p53. but not induction of the p21 gene, is essential for oncogene-mediated apoptosis. 1997. EMBO J.. 15: 3693-3701, 1996. 35. Tang. D. G.. Li, L., Zhu, Z., and Joshi, B. Apoptosis in the absence of cytochrome r 53. Gorospe, M.. Cirielli, C., Wang, X., Seth, P.. Capogrossi, M. C.. and Holbrook, N. J. accumulation in the cytosol. Biochem. Biophys. Res. Commun.. 242: 380-384, 1998. P21Waf-l/Cipl protects against p53-mediated apoptosis of human melanoma cells. 36. Tang, D. G., Li. L., Zhu, Z.. Joshi. B.. Johnson, C. R., Marnett. L. J.. Honn. K. V., Oncogene. 14: 929-935. 1997. Crissman. J. D.. timar. J., and Porter. A. T. BMD188. a novel hydroxamic acid 54. Polyak. K., Waldman. T.. He. T-C., Kinzler, K. W., and Vogelstein, B. Genetic compound, demonstrates potent anti-prostate cancer effects m vitro and in vivo: determinants of p53-induced apoptosis and growth arrest. Genes Dev., 10: 1945- requirements for mitochondria, reactive oxygen species, and proteases. Pathol. Oncol. 1952, 1996. Res., in press, 1998. 55. Reed, J. C. Double identity for proteins of the Bcl-2 family. Nature (Lond.), 387: 37. Chopra, D. P., Grignon, D. J., Joiakim, A., Mathieu, P. A.. Mohamed, A., Sakr, W. A., 773-776, 1997. Powell. I. J.. and Sarkar. F. H. Differential growth factor responses of epithelial cell 56. Hockenbery, D. M. Bcl-2, a novel regulator of cell death. BioEssays, 17: 631-638, cultures derived from normal human prostate, benign prostatic hyperplasia and 1995. primary prostate carcinoma. J. Cell Physiol., 169: 269-280, 1996. 57. Merchant, A. K.. Loney, T. L.. and Maybaum. J. Expression of wild-type p53 38. Tang. D. G., Chen, Y. Q.. and Honn. K. V. Arachidonate lipoxygenases as essential stimulates an increase in both Bax and Bcl-X, protein content in HT29 cells. regulators of cell survival and apoptosis. Proc. Nati. Acad. Sci. USA. 93: 5241-5246, Oncogene. 13: 2631-2637. 1996. 1996. 58. Miyashita, T., Krajewski, S., Krajewski, M., Wang, H. G., Lin, H. K., Liebermann, 39. Tang, D. G.. and Honn, K. V. Apoptosis of W256 carcinosarcoma cells of the D. A.. Hoffman, B.. and Reed, J. C. Tumor suppressor p53 is a regulator of hcl-2 and monocytoid origin induced by NDGA involves lipid peroxidation and depletion of box gene expression in vitro and in vivo. Oncogene. 9: 1799-1805, 1994.

3479

Downloaded from cancerres.aacrjournals.org on September 30, 2021. © 1998 American Association for Cancer Research. Extended Survivability of Prostate Cancer Cells in the Absence of Trophic Factors: Increased Proliferation, Evasion of Apoptosis, and the Role of Apoptosis Proteins

Dean G. Tang, Li Li, Dharam P. Chopra, et al.

Cancer Res 1998;58:3466-3479.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/58/15/3466

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/58/15/3466. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.