Molecular pathology of prostate cancer: the key to ...


Published on

  • Be the first to comment

  • Be the first to like this

No Downloads
Total Views
On Slideshare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

Molecular pathology of prostate cancer: the key to ...

  1. 1. Endocrine-Related Cancer (2004) 11 477–488 Molecular pathology of prostate cancer: the key to identifying new biomarkers of disease Ruth Foley, Donal Hollywood and Mark Lawler Department of Haematology and Oncology, Institute of Molecular Medicine, St James’ Hospital, James’ St, Dublin 8, Ireland (Requests for offprints should be addressed to M Lawler; Email: Abstract Microarray technology has recently accelerated the study of the molecular events involved in prostate cancer, offering the prospect of more precise prognosis and new therapeutic strategies. This review summarises current knowledge of the molecular pathology of prostate cancer. The expression and function of numerous genes have been shown to be altered in prostate cancer. Many of these genes are involved in cell cycle regulation, steroid hormone metabolism or regulation of gene expression. The mechanisms by which androgen independence arises are discussed, including cross-activation, gene amplification and point mutations of the androgen receptor. Analysis of changes in the levels of expression of large numbers of genes during prostate cancer progression have provided a better understanding of the basis of the disease, yielding new molecular markers, such as hepsin, with potential use in diagnosis and prognosis. Endocrine-Related Cancer (2004) 11 477–488 Introduction methods, may contribute much towards these goals (Dhanasekaran et al. 2001, Stamey et al. 2001, LaTulippe Prostate cancer represents one in ten cases of cancer in et al. 2002, Rhodes et al. 2002). The molecular events that men, and 200 000 deaths per year, worldwide (Parkin et al. occur in the development of androgen independence have 2001). Because the disease is more common among older also been clarified (Feldman & Feldman 2001), particu- men, its incidence is expected to increase as the population larly by a number of reports of ligand-independent cross- ages (Kirby et al. 2001). Prostate tumours, although activation of the androgen receptor (AR) (Culig et al. slower growing than most tumours, vary widely in their 1994, Craft et al. 1999, Sadar 1999, Godoy-Tundidor et aggressiveness (Chodak et al. 1994). The challenges of al. 2002, Ueda et al. 2002). deciding which patients require immediate treatment, and of developing new therapies, particularly for hormone- refractory tumours, should be eased by a rapidly increasing body of knowledge of the molecular biology Genes involved in prostate cancer of prostate cancer. development Numerous molecular abnormalities have been Not surprisingly, many genes associated with cell cycle described, including chromosomal loss or gain (reviewed regulation and cell proliferation have been implicated. by Roylance et al. 1997), gene amplification, mutations Mutations of p53 are common in prostate cancer, leading to increases or decreases in gene expression, and particularly in advanced disease (Meyers et al. 1998). mutations leading to changes in function of the protein. Aberrant nuclear accumulation of p53 protein is a The challenge is to use this information to develop better negative prognostic factor for disease-free survival (Lei- diagnostic and prognostic indicators and to identify new bovich et al. 2000, Quinn et al. 2000). Furthermore, targets for therapeutic intervention. Several recent large- patients with low-level accumulation of p53 had a high scale studies of gene expression in prostate cancer, using probability of worse prognosis if p53-positive nuclei were cDNA microarrays, tissue microarrays and other located near each other (Quinn et al. 2000). p53 and its Endocrine-Related Cancer (2004) 11 477–488 DOI:10.1677/erc.1.00699 1351-0088/04/011–477 # 2004 Society for Endocrinology Printed in Great Britain
  2. 2. Foley et al.: Molecular pathology of prostate cancer downstream effector, p21, have been implicated in the Prostate growth is normally regulated by activation of development of androgen independence (see section the AR. On binding dihydrotestosterone or similar steroid ‘Molecular biology of androgen independence’). hormones, the nuclear AR dissociates from heat shock The tumour suppressor gene, PTEN, is located on proteins and binds to specific promoters to stimulate chromosome 10q23 in a region in which loss of hetero- transcription. Mutations and alterations in expression of zygosity frequently occurs in prostate cancer (Cairns et al. the AR and related proteins are clearly important in 1997), and encodes a phosphatase that negatively prostate cancer. regulates cell cycle progression. Loss of PTEN at DNA The possible significance, for prostate cancer devel- and protein level has been reported in 25–33% of opment, of two microsatellite repeat polymorphisms advanced prostate tumours (Cairns et al. 1997, McMe- (CAG and GGN) in the AR has been investigated. namin et al. 1999), and at varying frequencies in clinically Shorter alleles of the CAG microsatellite repeat poly- localised prostate cancer (Cairns et al. 1997, Halvorsen et morphism have been linked to risk of advanced prostate al. 2003). Loss of PTEN has been associated with cancer (Ingles et al. 1997) and diagnosis at a younger age increased Gleason score (McMenamin et al. 1999) and (Bratt et al. 1999, Santos et al. 2003), although not to risk of clinical recurrence (Halvorsen et al. 2003). The overall risk of prostate cancer (Ingles et al. 1997, Stanford protein kinase Akt is inhibited by PTEN, and a significant et al. 1997). Conflicting evidence exists as to whether downstream result of this is downregulation of the shorter alleles of the GGN repeat confer a greater risk of expression of genes that are under the transcriptional prostate cancer (Stanford et al. 1997, Correa-Cerro et al. control of the androgen receptor (Nan et al. 2003). 1999). The c-met proto-oncogene encodes the hepatocyte The role of the AR in androgen-independent prostate growth factor/scatter factor receptor. Its expression cancer will be discussed in more detail later. In the steroid increases with Gleason score (Pisters et al. 1995), and 5a-reductase gene, which activates testosterone by con- more than 90% of metastatic specimens were positive for verting it to dihydrotestosterone, an amino acid substitu- c-met, in contrast to 50% of the primary samples tion conferring greater enzyme activity was associated examined in the same study (Knudsen et al. 2002). with a slightly greater risk of prostate cancer (Makridakis Interestingly, Met expression was particularly high in et al. 1997). Allele frequencies at a polyA microsatellite in bone metastases and also occurs in normal bone stroma, the gene encoding the receptor for vitamin D, another raising the question of whether it is partly responsible for steroid hormone that influences growth of prostate cancer the tendency of prostate tumours to metastasise to bone cells (Zhuang & Burnstein 1998), differed between sites (Knudsen et al. 2002). prostate cancer patients and controls (Ingles et al. 1997). The significant role played by the HER2/neu (c-erbB2) Twenty per cent of controls were homozygous for short oncogene in breast cancer led to speculation that it may alleles (17 repeats), compared with none of 26 patients also be important in prostate cancer. However, it is with advanced disease (Ingles et al. 1997). expressed in the latter disease at much lower levels, similar A number of genes involved in cell adhesion are to HER2/neu amplification-negative breast cancer, and in involved in the control of prostate cancer growth. E- only a minority of patients (Jorda et al. 2002). Although cadherin was shown to have lower activity in prostate immunohistochemistry failed to detect HER2/neu in 72 tumours than in normal matched tissue (Rashid et al. samples from locally recurrent or metastatic prostate 2001). The PC-3 human prostate cancer cell line is tumours (Savanainen et al. 2002), an animal model has deficient in a-catenin, but its restoration reduces the suggested a possible role for HER2/neu in androgen proliferation rate (Ewing et al. 1995). Increased concen- independence disease (see section ‘Molecular biology of trations of the transmembrane glycoprotein KAI1 sup- androgen independence’). press the development of metastasis in a rat model (Dong Telomerase reactivation was also detected in 14 of 24 et al. 1995). cases of prostate cancer, but not in 12 controls (Meid et al. Glutathione-S-transferase, part of a pathway protect- 2001), and telomerase mRNA levels were greater in ing cells from oxidative damage, shows a strong associa- malignant than in non-malignant prostate tissue (de Kok tion between inactivation by hypermethylation and et al. 2002). Another study found that E2F4, a transcrip- prostate cancer development (Goessl et al. 2001). The tion factor involved in cell proliferation, was over- oestrogen receptor a is also frequently inactivated by expressed at both mRNA and protein levels in methylation in prostate tumours, but not in benign malignant radical prostatectomy specimens (Waghray et prostate tissue from the same patients (Sasaki et al. 2002). al. 2001). Other proliferation-related genes have recently Table 1 summarises some of the molecular aberrations been implicated by microarray studies (see below). observed to date in prostate cancer. 478
  3. 3. Endocrine-Related Cancer (2004) 11 477–488 Table 1 Genes involved in prostate cancer development. Observation Gene Reference Higher levels in cancer than in normal prostate telomerase Meid et al. 2001 E2F4 Waghray et al. 2001 activated MAPK Gioeli et al. 1999 DD3 PCA3 de Kok et al. 2002 PIM1 Dhanasekaran et al. 2001 hepsin Rhodes et al. 2002 AMACR Xu et al. 2002b Lower levels in cancer than in normal prostate PTEN Cairns et al. 1997 E-cadherin Rashid et al. 2001 IGFBP-3 Rhodes et al. 2002 Higher levels in metastases than in primary tumour c-met Knudsen et al. 2002 cyclin E1 LaTulippe et al. 2002 MTA-1 Dhanasekaran et al. 2001 EZH2 Varambally et al. 2002 Overexpression inhibits metastasis occurrence KAI1 Dong et al. 1995 Higher levels in androgen-independent than in bcl-2 McDonnell et al. 1992 androgen-dependent disease p21WAF-1/CIP1 Baretton et al. 1999 fibronectin Stubbs et al. 1999 IGFBP-2 Bubendorf et al. 1999b insulin receptor Bubendorf et al. 1999b Lower levels in androgen-independent than in BTG-1 Chang et al. 1997 androgen-dependent disease oncostatin-M Bubendorf et al. 1999b Gene amplification in advanced disease androgen receptor Visakorpi et al. 1995 c-myc Bubendorf et al. 1999a Inactivated by hypermethylation glutathione-S-transferase Goessl et al. 2001 oestrogen receptor a Sasaki et al. 2002 Disease-associated mutations p53 Meyers et al. 1998 androgen receptor Taplin et al. 1999 Disease-associated polymorphisms androgen receptor Stanford et al. 1997 steroid 5a-reductase Makridakis et al. 1997 vitamin D receptor Ingles et al. 1997 Prognostic use demonstrated for expression changes PSA Small Roach 2002 p53 Quinn et al. 2000 PIM1 Dhanasekaran et al. 2001 hepsin Dhanasekaran et al. 2001 EZH2/E-cadherin Rhodes et al. 2003 IGFBP, insulin-like growth factor binding protein. Molecular biology of androgen Most androgen-independent prostate tumours express independence the AR (van der Kwast et al. 1991). It is activated in conditions of androgen deprivation by at least three Many prostate tumours, although initially androgen- mechanisms: overexpression as a result of gene amplifica- dependent, become androgen-independent and refractory tion, cross-activation by agents other than androgens, and to hormone withdrawal therapy, presenting a significant mutations that alter its specificity. problem in the treatment of the disease. It is not clear Gene amplification is probably the most common whether androgen-independent prostate tumours derive mutational mechanism for the AR. This event is very rare from pre-existing androgen-independent prostate cells or, (0–1%) in untreated prostate tumours, but occurs in 22– more probably, from prostate cells that acquire the ability 30% of tumours that recur after endocrine treatment to proliferate in the absence of androgen. This ability is (Visakorpi et al. 1995, Koivisto et al. 1997, Bubendorf et acquired by both AR-dependent and AR-independent al. 1999a). AR gene amplification was associated with a mechanisms (Fig. 1). positive response to combined androgen blockade after 479
  4. 4. Foley et al.: Molecular pathology of prostate cancer Figure 1 Molecular mechanisms involved in androgen independence. Evidence has been found for a number of possible mechanisms. AD, androgen-dependent; AI, androgen-independent; AR, androgen receptor; IGF-I, insulin-like growth factor I; IL-6, interleukin-6; TSGs, tumour suppressor genes. failure of endocrine monotherapy in a prospective study At least five non-androgen agents have been found to of 92 patients (Palmberg et al. 2000), although this has not cross-activate the AR in the absence of androgen, by at been confirmed by other studies. Tumours that developed least three distinct signalling pathways. Culig and co- androgen independence for reasons other than increased workers (1994) showed that insulin-like growth factor AR levels were less influenced by even lower androgen (IGF)-I activates AR in the human prostate cancer cell concentrations. Interestingly, one patient with high levels line DU145, although this did not occur in LNCaP cells of AR gene amplification before hormonal therapy had a (Ueda et al. 2002). The difference may be attributable to complete but temporary response to treatment, suggesting variations between the signalling pathways in these cell that this alteration is not always sufficient for androgen- lines. Forskolin (a plant lipid that stimulates cAMP) independent behaviour (Edwards et al. 2001). activates AR through the protein kinase A pathway 480
  5. 5. Endocrine-Related Cancer (2004) 11 477–488 (Sadar 1999). Interleukin-6 (Ueda et al. 2002) and a methodological differences, but may also reflect genetic related cytokine, oncostatin-M (Godoy-Tundidor et al. heterogeneity. One point mutation in the hormone- 2002) both activate AR in a ligand-independent manner, binding domain, found in the LNCaP human prostate but through different pathways. Oncostatin-M may have cancer cell line, confers sensitivity to hormones other than an important role despite reduced levels of expression in androgens, including anti-androgens (Veldscholte et al. androgen-independent cells in a xenograft model (Buben- 1990). This mutation was reported to occur in five of 16 dorf et al. 1999b), as it also modulates the response of the patients after combined androgen blockade with the anti- AR to hydroxyflutamide, causing this anti-androgen to androgen, flutamide (Taplin et al. 1999), but in none of 36 act as an agonist (Godoy-Tundidor et al. 2002). Such hormone-refractory tumours tested by two other authors interactions may explain why, in certain patients, pros- (Visakorpi et al. 1995, Koivisto et al. 1997). tate-specific antigen (PSA) concentrations decrease after Alternatively, in certain tumours the AR pathway cessation of unsuccessful anti-androgen treatment. The may be bypassed altogether, by other mechanisms that transcription factor HER2/neu activates the AR indepen- regulate the balance of proliferation and apoptosis. dently of androgen (Craft et al. 1999, Yeh et al. 1999), and Protein concentrations of the anti-apoptotic proto-onco- increases its response to low concentrations of androgen gene, bcl-2, have been correlated with progression to (Yeh et al. 1999). HER2/neu was overexpressed in androgen-independent status in human tumours, and bcl- androgen-independent derivatives of an androgen-depen- 2 mRNA was upregulated by testosterone withdrawal in dent human prostate cancer xenograft studied by Craft rats (McDonnell et al. 1992). Bcl-2 stable transfection and coworkers (1999). Such alterations in gene expression enabled the androgen-dependent prostate cancer cell line, during endocrine treatment may allow the AR to be re- LNCaP, to survive androgen depletion in vitro and in vivo activated by concentrations of androgen that were (Raffo et al. 1995) and protected the same cell line from previously too low. apoptosis induced by melanoma differentiation associated Several groups have found evidence that the mitogen- gene-7 (mda-7)/IL-24 (Lebedeva et al. 2003). In contrast, activated protein kinase (MAPK) pathway is involved in overexpression of bcl-2 did not influence the viability of cross-activation of the AR (Yeh et al. 1999, Ueda et al. androgen-independent prostate cancer cells treated with 2002, Franco et al. 2003). Activation of the MAPK mda-7/IL-24, which selectively induces apoptosis in a enzymes ERK1 and ERK2 was detected in 70% of range of malignant cell types (Lebedeva et al. 2003). prostate tumours with Gleason scores 8–10, and was c-myc mRNA levels were greater in prostate tumours highly significantly associated with increasing Gleason than in benign prostate tissue (Fleming et al. 1986); score (Gioeli et al. 1999). The MAPK pathway, among furthermore, 11% of 62 metastatic prostate cancer speci- others, is activated by the Ras protein after its stimulation mens, but none of 223 primary tumours, showed c-myc by a range of growth factors. The introduction of gene amplification (Bubendorf et al. 1999a). However, activated mutant forms of the H-Ras oncogene to another study of 130 untreated primary tumours detected androgen-dependent prostate cancer cells conferred c-myc gene amplification in 20% of cases, and found that hypersensitivity to low concentrations of androgen, it was significantly associated with higher Gleason score androgen-independent growth in an animal model, and and risk of disease progression (Sato et al. 1999). The constitutive activity of MAPK enzymes (Bakin et al. variation in reported frequency of c-myc gene amplifica- 2003). Mutations that activate the ras family of oncogenes tion in primary tumours is possibly attributable to the frequently occur in some types of human cancer, but selection of high-grade tumours in the last of these appear to be rare in prostate cancer (Carter et al. 1990, studies. Gumerlock et al. 1991), indicating that other factors are p21WAF-1/CIP1, an effector of the p53 protein, negatively responsible for the frequent MAPK enzyme activation in regulates the cell cycle, and its overexpression inhibits the prostate tumours. However, a study of Japanese patients proliferation of androgen-dependent and androgen- with prostate cancer found a greater frequency of ras independent prostate cancer cells in vitro and tumorigeni- mutations, suggesting that the Ras pathway may have a city in vivo (Gotoh et al. 2003). However, it is expressed more significant role in this low-risk population than in more frequently in androgen-independent than in andro- Western countries (Anwar et al. 1992). gen-dependent prostate cancer (Baretton et al. 1999, Fizazi The clinical significance of AR mutations remains to et al. 2002). In a mouse model, androgen-dependent be clarified. Estimates of the frequency of such mutations prostate tumours lost p21WAF-1/CIP1 expression after in prostate cancer vary, both before (Tilley et al. 1996, castration, but it was restored in androgen-independent Marcelli et al. 2000) and after (Taplin et al. 1999, Wallen tumours that subsequently relapsed (Fizazi et al. 2002). In et al. 1999) hormonal treatment, ranging from 6% to addition, survival for patients with high p21WAF-1/CIP1 44%. This variation may partially be explained by levels was significantly shorter (Baretton et al. 1999). These 481
  6. 6. Foley et al.: Molecular pathology of prostate cancer findings suggest that p21WAF-1/CIP1 may exert a growth- to PSA staining for differentiating between malignant and stimulatory effect by a paracrine pathway (Fizazi et al. benign prostate tissue (Darson et al. 1997). 2002) or that the G1/S cell cycle checkpoint is aberrantly DD3PCA3, a gene identified in a differential display regulated in some androgen-independent prostate study, has shown potential for use in the diagnosis of tumours (Baretton et al. 1999). prostate cancer (Schalken et al. 2003). Analysis of p53 status may also influence responsiveness to DD3PCA3 as a prostate cancer marker in urine samples androgens. In vitro experiments demonstrated that four yielded a negative predictive value of 90% (Hessels et al. p53 mutations common in prostate cancer each enabled 2003). The gene is transcribed to an apparently untrans- the androgen-dependent prostate cancer cell line LNCaP lated RNA of unknown function (Bussemakers et al. to grow in an androgen-independent manner (Nesslinger 1999). DD3PCA3 mRNA levels, measured by quantitative et al. 2003). PCR (of exons 1–4), were 34-fold greater in malignant Several studies have aimed to identify genes differen- prostate than in normal prostate or benign prostatic tially expressed in androgen-dependent and androgen- hyperplasia, and very low or absent in non-prostate independent prostate cancer. Fibronectin, a gene involved tissues (de Kok et al. 2002). Evidence has since been in cell adhesion, is expressed at higher levels in androgen- published that an alternative DD3PCA3 transcript lacking independent prostate cells (Stubbs et al. 1999). A novel exon 4 is expressed in several non-prostate tissues putative cell cycle regulator and tumour suppressor, BTG- (Gandini et al. 2003). Other potential markers such as 1, is expressed at reduced levels in an androgen- PIM1 and hepsin have been identified by microarray independent prostate cancer cell line, compared with its experiments and are discussed later. androgen-dependent parental cell line (Chang et al. 1997). A cDNA microarray study on the androgen-depen- dent human prostate cancer xenograft, CWR22, and its Microarray studies of prostate cancer gene androgen-independent derivatives showed that IGF bind- expression ing protein (IGFBP)-2, insulin receptor and IGF-II were all overexpressed in androgen-independent conditions Microarray technology (Fig. 2) has provided a powerful (Bubendorf et al. 1999b). IGFBP-2 overexpression was tool for large-scale studies of changes in gene expression. validated by a tissue microarray method (Bubendorf et al. This technique involves the spotting of up to several 1999b). It is worth noting, as mentioned previously, that thousand specific cDNAs or oligonucleotides on a single IGF-I can activate the AR in the absence of androgen slide or chip in a known arrangement, and hybridising the (Culig et al. 1994). In addition, prostate cancer risk has slide to the samples. There are a number of approaches to been correlated to IGF-I concentrations in plasma (Chan microarray analysis. In one approach, the samples are et al. 1998), and IGFBP-3 concentrations are lower in prepared by extracting mRNA from the cells or tissues of malignant prostate tissue than in benign tissue (Rhodes et interest (such as prostate cancer and normal prostate), al. 2002). labelling them with different fluorescent dyes, and pooling them. After unbound sample has been removed, fluores- cence at the two relevant wavelengths is measured at each spot on the slide. Analysis of the resulting data reveals the Prognostic markers in prostate cancer levels of expression of the genes in the tissues tested. Prostate-specific antigen (PSA) is well established as the Appropriate and representative sample selection is of biochemical marker of choice in prostate cancer diagnosis crucial importance, and differences in gene expression and prognosis (Small Roach 2002). However, its revealed by microarrays must be validated by other sensitivity and specificity are limited (Kirby et al. 2001), techniques for measuring mRNA and protein concentra- and other markers have been investigated. Recent large- tions. cDNA or oligonucleotide arrays can be comple- scale gene expression analyses have identified possible mented by tissue microarrays, in which large numbers of alternatives, including DD3 and hepsin. tissue sections are arrayed on a single slide and challenged Genes expressed in prostate tumours but showing with an antibody of interest (Dhanasekaran et al. 2001). weak or absent expression in normal prostate include As with other methods, in microarray studies also, prostate-specific membrane antigen (PSMA) (Silver et al. genes involved in cell proliferation are clearly important 1997), prostate mucin antigen (Beckett et al. 1991), and (Stamey et al. 2001, LaTulippe et al. 2002). c-myc mRNA PTI-1 (Shen et al. 1995). Genes expressed at higher levels levels were greater in prostate tumours than in normal in malignant than in normal prostate include human prostate tissue (Dhanasekaran et al. 2001), as found kallikrein 2 (hK2) (Darson et al. 1997), PSGR (Xu et al. previously (Fleming et al. 1986). The proto-oncogene 2000a), and PCGEM1 (Srikantan et al. 2000). Immuno- PIM1 was upregulated in prostate cancer compared with histochemical staining for hK2 was found to be superior normal prostate (Dhanasekaran et al. 2001). However, 482
  7. 7. Endocrine-Related Cancer (2004) 11 477–488 Figure 2 From microarrays to new diagnostic and therapeutic targets. Microarrays allow the analysis of patterns of expression of large numbers of genes in tissues of interest, such as advanced prostate cancer and localised prostate cancer. Visual representations of data show spots of different colours and intensities for genes expressed in either or both tissues at differing levels. 483
  8. 8. Foley et al.: Molecular pathology of prostate cancer among patients, lower PIM1 levels were strongly asso- Microarray experiments have also identified potential ciated with an increased risk of relapse after radical new biochemical markers for prostate cancer. Hepsin, a prostatectomy (Dhanasekaran et al. 2001). The three transmembrane serine protease, was expressed more genes MYBL2, cyclin E and CDC2, which are involved in highly in tumour than in normal prostate in all four stimulating mitosis through the same pathway, are all studies in this meta-analysis (Rhodes et al. 2002). Its upregulated in metastatic prostate cancer compared with overexpression was validated by RT-PCR (Welsh et al. non-recurrent primary tumours (LaTulippe et al. 2002). 2001) and tissue microarray (Dhanasekaran et al. 2001). Dhanasekaran and coworkers (2001) identified a number Stamey and coworkers (2001) investigated differences of genes, including metastasis-associated-1 (MTA-1), that between benign prostatic hyperplasia and Gleason grade showed changes in expression profile in metastatic 4/5 prostate cancer using cDNA arrays and found that, of almost 7000 genes tested, hepsin showed most upregula- prostate cancer when compared with localised cancer. A tion in cancer. The greatest levels of hepsin were found in cDNA microarray study showed that IGFBP-2 was high-grade prostatic intraepithelial neoplasia (PIN), and overexpressed in androgen-independent prostate cancer levels were reduced in metastatic samples compared with compared with androgen-dependent disease (Bubendorf localised disease samples (Dhanasekaran et al. 2001). et al. 1999b). The role of IGF-I in androgen independence Tumours in which hepsin was absent or present in low is discussed above. levels were more likely to relapse after radical prostatect- Varambally et al. (2002) used microarrays to char- omy, although the association of reduced PIM1 with acterise benign prostate, clinically localised prostate relapse was still stronger (Dhanasekaran et al. 2001). cancer and metastatic prostate cancer and, of 55 genes upregulated in metastatic disease compared with localised disease, EZH2 (enhancer of zeste homologue 2) showed Conclusion the greatest magnitude of overexpression. This gene encodes a homologue of a Drosophila transcriptional This review represents a summary of the rapidly increas- repressor gene. Its overexpression in the prostate cell line ing knowledge of the molecular pathology of prostate RWPE resulted in repression of a wide range of genes, at cancer. Alterations in expression or function of numerous least some of which appear to function in prostate cancer genes involved in key aspects of carcinogenesis, including proliferation, as small interfering RNA (siRNA) down- cell cycle regulation, steroid hormone metabolism and regulation of EZH2 in vitro resulted in growth inhibition regulation of gene expression, are associated with disease of prostate cancer cells (Varambally et al. 2002). Subse- progression. Metastatic (Dhanasekaran et al. 2001, quently, prostate tumours that were both EZH2-positive LaTulippe et al. 2002) and androgen-independent (Buben- dorf et al. 1999b) prostate cancer have been specifically and E-cadherin-negative were shown to be significantly examined. Androgen independence can be acquired by a more likely to recur (Rhodes et al. 2003). number of molecular mechanisms, including cross-activa- Fatty acid synthase (Dhanasekaran et al. 2001) and tion by agents such as IGF-I (Culig et al. 1994). Potential another fatty acid metabolism enzyme, human a-methyl- new molecular markers such as hepsin (Rhodes et al. 2002) acyl-CoA racemase (AMACR) (Xu et al. 2000b), were have been identified. These and other identified and novel also shown to be overexpressed in prostate tumours, and genes implicated in the disease merit more investigation of were significantly overexpressed in a meta-analysis of four their clinical relevance. The emergence of microarray studies (Rhodes et al. 2002). Tissue microarrays showed technology allowing simultaneous analysis of the expres- that, as was the case with hepsin, overexpression of sion of thousands of genes, and hundreds of tissue AMACR was greater in localised tumours than in specimens, should facilitate rapid progress in identifying metastases, possibly as a result of dedifferentiation key molecular players in the development of prostate (Kuefer et al. 2002). cancer. Microarray experiments have revealed some trends that would be difficult to identify by other techniques. A meta-analysis by Rhodes and coworkers (2002) of four References microarray studies of prostate cancer showed that biosynthesis of polyamines and of purines was consis- Anwar K, Nakakuki K, Shiraishi T, Naiki H, Yatani R Inuzuka M 1992 Presence of ras oncogene mutations and tently dysregulated in prostate cancer. Almost 25% of the human papillomavirus DNA in human prostate carcinomas. genes found by LaTulippe and coworkers (2002) to be Cancer Research 52 5991–5996. most overexpressed or underexpressed during the pro- Bakin RE, Gioeli D, Sikes RA, Bissonette EA Weber MJ 2003 gression of prostate cancer were associated with regula- Constitutive activation of the Ras/mitogen-activated protein tion of gene expression. kinase signalling pathway promotes androgen hypersensitivity 484
  9. 9. Endocrine-Related Cancer (2004) 11 477–488 in LNCaP prostate cancer cells. Cancer Research 63 prostate cancer in a French-German population. European 1981–1989. Journal of Human Genetics 7 357–362. Baretton GB, Klenk U, Diebold J, Schmeller N Lohrs U 1999 Craft N, Shostak Y, Carey M Sawyers CL 1999 A mechanism Proliferation- and apoptosis-associated factors in advanced for hormone-independent prostate cancer through prostatic carcinomas before and after androgen deprivation modulation of androgen receptor signaling by the HER-2/neu therapy: prognostic significance of p21/WAF1/CIP1 tyrosine kinase. Nature Medicine 5 280–285. expression. British Journal of Cancer 80 546–555. Culig Z, Hobisch A, Cronauer MV, Radmayr C, Trapman J, Beckett ML, Lipford GB, Haley CL, Schellhammer PF Wright Hittmair A, Bartsch G Klocker H 1994 Androgen receptor GL Jr 1991 Monoclonal antibody PD41 recognizes an antigen activation in prostatic tumor cell lines by insulin-like growth restricted to prostate adenocarcinomas. Cancer Research 51 factor-I, keratinocyte growth factor, and epidermal growth 1326–1333. factor. Cancer Research 54 5474–5478. Bratt O, Borg A, Kristoffersson U, Lundgren R, Zhang QX Darson MF, Pacelli A, Roche P, Rittenhouse HG, Wolfert RL, Olsson H 1999 CAG repeat length in the androgen receptor Young CY, Klee GG, Tindall DJ Bostwick DG 1997 gene is related to age at diagnosis of prostate cancer and Human glandular kallikrein 2 (hK2) expression in prostatic response to endocrine therapy, but not to prostate cancer risk. intraepithelial neoplasia and adenocarcinoma: a novel British Journal of Cancer 81 672–676. prostate cancer marker. Urology 49 857–862. Bubendorf L, Kononen J, Koivisto P, Schraml P, Moch H, Dhanasekaran SM, Barrette TR, Ghosh D, Shah R, Varambally Gasser TC, Willi N, Mihatsch MJ, Sauter G Kallioniemi S, Kurachi K, Pienta KJ, Rubin MA Chinnaiyan AM 2001 OP 1999a Survey of gene amplifications during prostate Delineation of prognostic biomarkers in prostate cancer. cancer progression by high-throughout fluorescence in situ Nature 412 822–826. hybridization on tissue microarrays. Cancer Research 59 Dong JT, Lamb PW, Rinker-Schaeffer CW, Vukanovic J, 803–806. Ichikawa T, Isaacs JT Barrett JC 1995 KAI1, a metastasis Bubendorf L, Kolmer M, Kononen J, Koivisto P, Mousses S, suppressor gene for prostate cancer on human chromosome 11p11.2. Science 268 884–886. Chen Y, Mahlamaki E, Schraml P, Moch H, Willi N, Edwards J, Krishna NS, Mukherjee R, Watters AD, Underwood Elkahloun AG, Pretlow TG, Gasser TC, Mihatsch MJ, Sauter MA Bartlett JM 2001 Amplification of the androgen G Kallioniemi OP 1999b Hormone therapy failure in receptor may not explain the development of androgen- human prostate cancer: analysis by complementary DNA and independent prostate cancer. BJU International 88 633–637. tissue microarrays. Journal of the National Cancer Institute 91 Ewing CM, Ru N, Morton RA, Robinson JC, Wheelock MJ, 1758–1764. Johnson KR, Barrett JC Isaacs WB 1995 Chromosome 5 Bussemakers MJ, van Bokhoven A, Verhaegh GW, Smit FP, suppresses tumorigenicity of PC3 prostate cancer cells: Karthaus HF, Schalken JA, Debruyne FM, Ru N Isaacs correlation with re-expression of alpha-catenin and WB 1999 DD3: a new prostate-specific gene, highly restoration of E-cadherin function. Cancer Research 55 overexpressed in prostate cancer. Cancer Research 59 4813–4817. 5975–5979. Feldman BJ Feldman D 2001 The development of androgen- Cairns P, Okami K, Halachmi S, Halachmi N, Esteller M, independent prostate cancer. Nature Reviews Cancer 1 34–45. Herman JG, Jen J, Isaacs WB, Bova GS Sidransky D 1997 Fizazi K, Martinez LA, Sikes CR, Johnston DA, Stephens LC, Frequent inactivation of PTEN/MMAC1 in primary prostate McDonnell TJ, Logothetis CJ, Trapman J, Pisters LL, cancer. Cancer Research 57 4997–5000. Ordonez NG, Troncoso P Navone NM 2002 The Carter BS, Epstein JI Isaacs WB 1990 ras gene mutations in association of p21(WAF-1/CIP1) with progression to human prostate cancer. Cancer Research 50 6830–6832. androgen-independent prostate cancer. Clinical Cancer Chan JM, Stampfer MJ, Giovannucci E, Gann PH, Ma J, Research 8 775–781. Wilkinson P, Hennekens CH Pollak M 1998 Plasma Fleming WH, Hamel A, MacDonald R, Ramsey E, Pettigrew insulin-like growth factor-I and prostate cancer risk: a NM, Johnston B, Dodd JG Matusik RJ 1986 Expression of prospective study. Science 279 563–566. the c-myc protooncogene in human prostatic carcinoma and Chang GT, Blok LJ, Steenbeek M, Veldscholte J, van Weerden benign prostatic hyperplasia. Cancer Research 46 1535–1538. WM, van Steenbrugge GJ Brinkmann AO 1997 Franco OE, Onishi T, Yamakawa K, Arima K, Yanagawa M, Differentially expressed genes in androgen-dependent and Sugimura Y Kawamura J 2003 Mitogen-activated protein -independent prostate carcinomas. Cancer Research 57 kinase pathway is involved in androgen-independent PSA 4075–4081. gene expression in LNCaP cells. Prostate 56 319–325. Chodak GW, Thisted RA, Gerber GS, Johansson JE, Adolfsson Gandini O, Luci L, Stigliano A, Lucera R, Di Silverio F, Toscano J, Jones GW, Chisholm GD, Moskovitz B, Livne PM V Cardillo MR 2003 Is DD3 a new prostate-specific gene? Warner J 1994 Results of conservative management of Anticancer Research 23 305–308. clinically localized prostate cancer. New England Journal of Gioeli D, Mandell JW, Petroni GR, Frierson HF Jr Weber MJ Medicine 330 242–248. 1999 Activation of mitogen-activated protein kinase Correa-Cerro L, Wohr G, Haussler J, Berthon P, Drelon E, associated with prostate cancer progression. Cancer Research Mangin P, Fournier G, Cussenot O, Kraus P, Just W, Paiss T, 59 279–284. Cantu JM Vogel W 1999 (CAG)nCAA and GGN repeats Godoy-Tundidor S, Hobisch A, Pfeil K, Bartsch G Culig Z in the human androgen receptor gene are not associated with 2002 Acquisition of agonistic properties of nonsteroidal 485
  10. 10. Foley et al.: Molecular pathology of prostate cancer antiandrogens after treatment with oncostatin M in prostate programs associated with metastatic disease. Cancer Research cancer cells. Clinical Cancer Research 8 2356–2361. 62 4499–4506. Goessl C, Muller M, Heicappell R, Krause H, Straub B, Schrader Lebedeva IV, Sarkar D, Su ZZ, Kitada S, Dent P, Stein CA, M Miller K 2001 DNA-based detection of prostate cancer Reed JC Fisher PB 2003 Bcl-2 and Bcl-x(L) differentially in urine after prostatic massage. Urology 58 335–338. protect human prostate cancer cells from induction of Gotoh A, Shirakawa T, Wada Y, Fujisawa M, Okada H, apoptosis by melanoma differentiation associated gene-7, Kamidono S Hamada K 2003 The growth inhibitory effect mda-7/IL-24. Oncogene 22 8758–8773. of p21 adenovirus on androgen-dependent and -independent Leibovich BC, Cheng L, Weaver AL, Myers RP Bostwick DG human prostate cancer cells. BJU International 92 314–318. 2000 Outcome prediction with p53 immunostaining after Gumerlock PH, Poonamallee UR, Meyers FJ deVere White radical prostatectomy in patients with locally advanced RW 1991 Activated ras alleles in human carcinoma of the prostate cancer. Journal of Urology 163 1756–1760. prostate are rare. Cancer Research 51 1632–1637. McDonnell TJ, Troncoso P, Brisbay SM, Logothetis C, Chung Halvorsen OJ, Haukaas SA Akslen LA 2003 Combined loss of LW, Hsieh JT, Tu SM Campbell ML 1992 Expression of PTEN and p27 expression is associated with tumor cell the protooncogene bcl-2 in the prostate and its association proliferation by Ki-67 and increased risk of recurrent disease with emergence of androgen-independent prostate cancer. in localized prostate cancer. Clinical Cancer Research 9 Cancer Research 52 6940–6944. 1474–1479. McMenamin ME, Soung P, Perera S, Kaplan I, Loda M Hessels D, Klein Gunnewiek JM, van Oort I, Karthaus HF, van Sellers WR 1999 Loss of PTEN expression in paraffin- Leenders GJ, van Balken B, Kiemeney LA, Witjes JA embedded primary prostate cancer correlates with high Schalken JA 2003 DD3(PCA3)-based molecular urine Gleason score and advanced stage. Cancer Research 59 analysis for the diagnosis of prostate cancer. European 4291–4296. Urology 44 8–16. Makridakis N, Ross RK, Pike MC, Chang L, Stanczyk FZ, Ingles SA, Ross RK, Yu MC, Irvine RA, La Pera G, Haile RW Kolonel LN, Shi CY, Yu MC, Henderson BE Reichardt JK Coetzee GA 1997 Association of prostate cancer risk with 1997 A prevalent missense substitution that modulates genetic polymorphisms in vitamin D receptor and androgen activity of prostatic steroid 5 alpha-reductase. Cancer receptor. Journal of the National Cancer Institute 89 166–170. Research 57 1020–1022. Jorda M, Morales A, Ghorab Z, Fernandez G, Nadji M Block Marcelli M, Ittmann M, Mariani S, Sutherland R, Nigam R, N 2002 Her2 expression in prostatic cancer: a comparison Murthy L, Zhao Y, DiConcini D, Puxeddu E, Esen A, with mammary carcinoma. Journal of Urology 168 1412–1414. Eastham J, Weigel NL Lamb DJ 2000 Androgen receptor Kirby RS, Christmas TJ Brawer MK 2001 Prostate Cancer, mutations in prostate cancer. Cancer Research 60 944–949. edn 2. London: Mosby. Meid FH, Gygi CM, Leisinger HJ, Bosman FT Benhattar J Knudsen BS, Gmyrek GA, Inra J, Scherr DS, Vaughan ED, 2001 The use of telomerase activity for the detection of Nanus DM, Kattan MW, Gerald WL Vande Woude GF prostatic cancer cells after prostatic massage. Journal of 2002 High expression of the Met receptor in prostate cancer Urology 165 1802–1805. metastasis to bone. Urology 60 1113–1117. Meyers FJ, Gumerlock PH, Chi SG, Borchers H, Deitch AD Koivisto P, Kononen J, Palmberg C, Tammela T, Hyytinen E, deVere White RW 1998 Very frequent p53 mutations in Isola J, Trapman J, Cleutjens K, Noordzij A, Visakorpi T metastatic prostate carcinoma and in matched primary Kallioniemi OP 1997 Androgen receptor gene amplification: a tumors. Cancer 83 2534–2539. possible molecular mechanism for androgen deprivation Nan B, Snabboon T, Unni E, X-J Yuan, Whang YE Marcelli therapy failure in prostate cancer. Cancer Research 57 M 2003 The PTEN tumor suppressor is a negative modulator 314–319. of androgen receptor transcriptional activity. Journal of de Kok JB, Verhaegh GW, Roelofs RW, Hessels D, Kiemeney Molecular Endocrinology 31 169–183. LA, Aalders TW, Swinkels DW Schalken JA 2002 DD3 Nesslinger NJ, Shi XB deVere White RW 2003 Androgen- (PCA3), a very sensitive and specific marker to detect prostate independent growth of LNCaP prostate cancer cells is tumors. Cancer Research 62 2695–2698. mediated by gain-of-function mutant p53. Cancer Research 63 Kuefer R, Varambally S, Zhou M, Lucas PC, Loeffler M, Wolter 2228–2233. H, Mattfeldt T, Hautmann RE, Gschwend JE, Barrette TR, Palmberg C, Koivisto P, Kakkola L, Tammela TL, Kallioniemi Dunn RL, Chinnaiyan AM Rubin MA 2002 Alpha- OP Visakorpi T 2000 Androgen receptor gene amplification methylacyl-CoA racemase: expression levels of this novel at primary progression predicts response to combined cancer biomarker depend on tumor differentiation. American androgen blockade as second line therapy for advanced Journal of Pathology 161 841–848. prostate cancer. Journal of Urology 164 1992–1995. van der Kwast TH, Schalken J, Ruizeveld de Winter JA, van Parkin DM, Bray FI Devesa SS 2001 Cancer burden in the Vroonhoven CC, Mulder E, Boersma W Trapman J 1991 year 2000. The global picture. European Journal of Cancer 37 Androgen receptors in endocrine-therapy-resistant human S4–S66. prostate cancer. International Journal of Cancer 48 189–193. Pisters LL, Troncoso P, Zhau HE, Li W, von Eschenbach AC LaTulippe E, Satagopan J, Smith A, Scher H, Scardino P, Reuter Chung LW 1995 c-met proto-oncogene expression in benign V Gerald WL 2002 Comprehensive gene expression and malignant human prostate tissues. Journal of Urology 154 analysis of prostate cancer reveals distinct transcriptional 293–298. 486
  11. 11. Endocrine-Related Cancer (2004) 11 477–488 Quinn DI, Henshall SM, Head DR, Golovsky D, Wilson JD, Small EJ Roach M 3rd 2002 Prostate-specific antigen in Brenner PC, Turner JJ, Delprado W, Finlayson JF, Stricker prostate cancer: a case study in the development of a tumor PD, Grygiel JJ Sutherland RL 2000 Prognostic significance marker to monitor recurrence and assess response. Seminars of p53 nuclear accumulation in localized prostate cancer in Oncology 29 264–273. treated with radical prostatectomy. Cancer Research 60 Srikantan V, Zou Z, Petrovics G, Xu L, Augustus M, Davis L, 1585–1594. Livezey JR, Connell T, Sesterhenn IA, Yoshino K, Buzard Raffo AJ, Perlman H, Chen MW, Day ML, Streitman JS GS, Mostofi FK, McLeod DG, Moul JW Srivastava S Buttyan R 1995 Overexpression of bcl-2 protects prostate 2000 PCGEM1, a prostate-specific gene, is overexpressed cancer cells from apoptosis in vitro and confers resistance in prostate cancer. PNAS 97 12216–12221. to androgen depletion in vivo. Cancer Research 55 Stamey TA, Warrington JA, Caldwell MC, Chen Z, Fan Z, 4438–4445. Mahadevappa M, McNeal JE, Nolley R Zhang Z 2001 Rashid MG, Sanda MG, Vallorosi CJ, Rios-Doria J, Rubin MA Molecular genetic profiling of Gleason grade 4/5 prostate Day ML 2001 Posttranslational truncation and inactivation cancers compared with benign prostatic hyperplasia. Journal of human E-cadherin distinguishes prostate cancer from of Urology 166 2171–2177. matched normal prostate. Cancer Research 61 489–492. Stanford JL, Just JJ, Gibbs M, Wicklund KG, Neal CL, Rhodes DR, Barrette TR, Rubin MA, Ghosh D Chinnaiyan Blumenstein BA Ostrander EA 1997 Polymorphic repeats AM 2002 Meta-analysis of microarrays: interstudy validation in the androgen receptor gene: molecular markers of prostate of gene expression profiles reveals pathway dysregulation in cancer risk. Cancer Research 57 1194–1198. prostate cancer. Cancer Research 62 4427–4433. Stubbs AP, Abel PD, Golding M, Bhangal G, Wang Q, Waxman Rhodes DR, Sanda MG, Otte AP, Chinnaiyan AM Rubin MA J, Stamp GW Lalani EN 1999 Differentially expressed 2003 Multiplex biomarker approach for determining risk of genes in hormone refractory prostate cancer: association with prostate-specific antigen-defined recurrence of prostate chromosomal regions involved with genetic aberrations. cancer. Journal of the National Cancer Institute 95 661–668. American Journal of Pathology 154 1335–1343. Roylance R, Spurr N Sheer D 1997 The genetic analysis of Taplin ME, Bubley GJ, Ko YJ, Small EJ, Upton M, prostate carcinoma. Seminars in Cancer Biology 8 37–44. Rajeshkumar B Balk SP 1999 Selection for androgen Sadar MD 1999 Androgen-independent induction of prostate- receptor mutations in prostate cancers treated with androgen specific antigen gene expression via cross-talk between antagonist. Cancer Research 59 2511–2515. the androgen receptor and protein kinase A signal Tilley WD, Buchanan G, Hickey TE Bentel JM 1996 transduction pathways. Journal of Biological Chemistry 274 Mutations in the androgen receptor gene are associated with 7777–7783. progression of human prostate cancer to androgen Santos ML, Sarkis AS, Nishimoto IN Nagai MA 2003 independence. Clinical Cancer Research 2 277–285. Androgen receptor CAG repeat polymorphism in prostate Ueda T, Bruchovsky N Sadar MD 2002 Activation of the cancer from a Brazilian population. Cancer Detection and androgen receptor N-terminal domain by interleukin-6 via Prevention 27 321–326. MAPK and STAT3 signal transduction pathways. Journal of Sasaki M, Tanaka Y, Perinchery G, Dharia A, Kotcherguina I, Biological Chemistry 277 7076–7085. Fujimoto S Dahiya R 2002 Methylation and inactivation of Varambally S, Dhanasekaran SM, Zhou M, Barrette TR, estrogen, progesterone, and androgen receptors in prostate Kumar-Sinha C, Sanda MG, Ghosh D, Pienta KJ, Sewalt cancer. Journal of the National Cancer Institute 94 384–390. RG, Otte AP, Rubin MA Chinnaiyan AM 2002 The Sato K, Qian J, Slezak JM, Lieber MM, Bostwick DG, polycomb group protein EZH2 is involved in progression of Bergstralh EJ Jenkins RB 1999 Clinical significance of prostate cancer. Nature 419 624–629. alterations of chromosome 8 in high-grade, advanced, Veldscholte J, Ris-Stalpers C, Kuiper GG, Jenster G, Berrevoets nonmetastatic prostate carcinoma. Journal of the National C, Claassen E, van Rooij HC, Trapman J, Brinkmann AO Cancer Institute 91 1574–1580. Mulder E 1990 A mutation in the ligand binding domain of Savanainen KJ, Saramaki OR, Linja MJ, Bratt O, Tammela TL, the androgen receptor of human LNCaP cells affects steroid Isola JJ Visakorpi T 2002 Expression and gene copy binding characteristics and response to anti-androgens. number analysis of ERBB2 oncogene in prostate cancer. Biochemical and Biophysical Research Communications 173 American Journal of Pathology 160 339–345. 534–540. Schalken JA, Hessels D Verhaegh G 2003 New targets for Visakorpi T, Hyytinen E, Koivisto P, Tanner M, Keinanen R, therapy in prostate cancer: differential display code 3 Palmberg C, Palotie A, Tammela T, Isola J Kallioniemi OP (DD3(PCA3)), a highly prostate cancer-specific gene. Urology 1995 In vivo amplification of the androgen receptor gene and 62 S34–S43. progression of human prostate cancer. Nature Genetics 9 Shen R, Su ZZ, Olsson CA Fisher PB 1995 Identification of the 401–406. human prostatic carcinoma oncogene PTI-1 by rapid Waghray A, Schober M, Feroze F, Yao F, Virgin J Chen YQ expression cloning and differential RNA display. PNAS 92 2001 Identification of differentially expressed genes by serial 6778–6782. analysis of gene expression in human prostate cancer. Cancer Silver DA, Pellicer I, Fair WR, Heston WD Cordon-Cardo C Research 61 4283–4286. 1997 Prostate-specific membrane antigen expression in Wallen MJ, Linja M, Kaartinen K, Schleutker J Visakorpi T normal and malignant human tissues. Clinical Cancer 1999 Androgen receptor gene mutations in hormone- Research 3 81–85. refractory prostate cancer. Journal of Pathology 189 559–563. 487
  12. 12. Foley et al.: Molecular pathology of prostate cancer Welsh JB, Sapinoso LM, Su AI, Kern SG, Wang-Rodriguez J, Identification of differentially expressed genes in human Moskaluk CA, Frierson HF Jr Hampton GM 2001 prostate cancer using subtraction and microarray. Cancer Analysis of gene expression identifies candidate markers and Research 60 1677–1682. pharmacological targets in prostate cancer. Cancer Research Yeh S, Lin HK, Kang HY, Thin TH, Lin MF Chang C 1999 61 5974–5978. From HER2/Neu signal cascade to androgen receptor and its Xu LL, Stackhouse BG, Florence K, Zhang W, Shanmugam N, coactivators: a novel pathway by induction of androgen Sesterhenn IA, Zou Z, Srikantan V, Augustus M, Roschke V, target genes through MAP kinase in prostate cancer cells. Carter K, McLeod DG, Moul JW, Soppett D Srivastava S PNAS 96 5458–5463. 2000a PSGR, a novel prostate-specific gene with homology to Zhuang SH Burnstein KL 1998 Antiproliferative effect of a G protein-coupled receptor, is overexpressed in prostate 1alpha,25-dihydroxyvitamin D3 in human prostate cancer cell cancer. Cancer Research 60 6568–6572. line LNCaP involves reduction of cyclin-dependent kinase 2 Xu J, Stolk JA, Zhang X, Silva SJ, Houghton RL, Matsumura activity and persistent G1 accumulation. Endocrinology 139 M, Vedvick TS, Leslie KB, Badaro R Reed SG 2000b 1197–1207. 488