Oncogene (1998) 16, 1913 ± 1920  1998 Stockton Press All rights reserved 0950 ± 9232/98 $12.00 http://www.stockton-press.co.uk/onc Increased cell growth and tumorigenicity in human prostate LNCaP cells by overexpression to cyclin D1

Yian Chen, Luis A Martinez, Margaret LaCava, Lezlee Coghlan and Claudio J Conti

The University of Texas MD Anderson Cancer Center, Science Park-Research Division, Smithville, Texas 78957, USA

Deregulated expression of cyclin D1 has been found in embryo ®broblasts (Jiang et al., 1993a; Quelle et al., several types of human tumors. In order to investigate 1993; Resnitzky et al., 1994). Induction of cyclin D1 in factors involved in human prostate cancer progression, the human breast cancer cell line, T-47D, also increases we studied the e€ects of cyclin D1 overexpression on the rate of transition from G1 to S phase and the human prostate cancer cell proliferation and tumorigeni- induced cyclin D1 is sucient to allow cells to city by transfecting LNCaP cells with a retroviral vector complete the cell cycle even if they are arrested in containing human cyclin D1 cDNA. When compared to early G1 phase after growth-factor deprivation the parental and control-vector transfected LNCaP cells, (Musgrove et al., 1994). Microinjection of antibodies these cyclin D1-transfected cells had more cells in S- or antisense plasmids into cells in mid-G1 inhibits phase and lower growth factor requirements. Further- cyclin D1 and prevents the cells from entering S phase more, these cells grew more in androgen-free medium. (Baldin et al., 1993; Quelle et al., 1993; Lukas et al., We also detected higher levels of Rb phosphorylation and 1994a,b; Tam et al., 1994; Zhou et al., 1995). This E2F-1 protein levels in LNCaP/cyclin D1 cells than that indicates that cyclin D1 limits the rate of progression in the parental and vector control cells in medium with or through G1 phase. Deregulated expression of cyclin D1 without androgen. Cyclin D1 transfected clones formed has been found in several types of tumors. In some tumors more rapidly than control and parental cells. parathyroid adenomas, cyclin D1 expression is elevated These tumors were refractory to the androgen-ablation by a translocation that juxtaposes the cyclin D1 treatment by castration, whereas tumors from parental parathyroid hormone genes at 11p15 (Rosenberg et and vector-control LNCaP cells regressed within 4 weeks al., 1991). Ampli®cation of the human 11q13 region after castration. These results suggest that overexpres- and elevated cyclin D1 transcript levels have been sion of cyclin D1 changes the growth properties, reported in a subset of head and neck squamous cell increases tumorigenicity and decreases the requirement carcinomas, esophagus, breast, bladder and liver for androgen stimulation in LNCaP cells both in vitro carcinomas, and B-cell lymphomas. (Lammie et al., and in vivo. 1991; Buckley et al., 1993; Jiang et al., 1992, 1993b; Schuuring et al., 1992; Zhang et al., 1993). Thus, cyclin Keywords: prostate cancer cell lines; androgen; cyclin D1 is probably involved in tumor development. D1; gene transfection; cell proliferation; tumorigenicity In androgen-induced rat prostate cell proliferation, cyclin D1 is greatly induced when prostate cells pass from G0 into early G1 phase. The changes in the cyclin D1 levels correlate with the proliferation rate after Introduction androgen stimulation in castrated rats (Furuya et al., 1995; Chen et al., 1996). This suggests that cyclin D1 Cyclin D1 is a 36-Kda protein that belongs to a family plays an important role in prostate cell proliferation. of cell-cycle regulators called cyclins (Xiong et al., We also examined the expression of cyclin D1 in three 1991, 1992a). It functions in the G1 phase of cell cycle human prostate cancer cell lines (PC3, DU145 and and activates the kinase activities of G1 cyclin- LNCaP). The growth of PC3 and DU145 cells is dependent kinases (for review, see Sherr, 1993, 1994). androgen independent, and that of LNCaP is androgen The active cyclin/cdk complexes phosphorylate pro- dependent. We found that cyclin D1 was constitutively teins which are important in cell-cycle control. One expressed in PC3 and DU145 cells grown with or putative substrate of these cyclin D1-activated kinases without serum but cyclin D1 expression was signifi- is the retinoblastoma protein (Rb). The hypopho- cantly lower in LNCaP cells grown without serum sphorylated form of Rb prevents cells from entering (Chen et al., 1995). So far it is still not clear whether the S phase. When Rb is hyperphosphorylated by this constitutive high expression of cyclin D1 in kinases (e.g., cdks), it can no longer prevent the cells androgen-independent cells is related to their non- from entering S phase (for review, see Weinberg, 1991). responsiveness to androgen ablation in prostate cancer. The activities of cyclin/cdk can be inhibited by several To further understand the role of cyclin D1 in the negative cell-cycle regulatory proteins (for review, see progression of prostate cancer, we overexpressed Sherr and Roberts, 1995). cyclin D1 in the LNCaP cell line by using a retroviral Overexpression of cyclin D1 shortens the G1 phase, vector containing human cyclin D1 cDNA (Jiang et decreases cell size and increases tumorigenicity in rat al., 1993a). The LNCaP cell line was derived from a lymph node of a patient with metastatic stage D prostate cancer (Horoszewicz et al., 1980, 1983). Correspondence: CJ Conti This work was supported by American Cancer Society CN143 LNCaP cells are well di€erentiated and have the Received 30 April 1995; revised 10 November 1997; accepted 10 characteristics of prostate cells; they produce prostate- November 1997 speci®c antigen (Montgomery et al., 1992), prostate- Cyclin D1 overexpression in LNCaP cells YChenet al 1914 speci®c membrane antigen (Israeli et al., 1994) and with the cyclin D1 expression vector as LNCaP/cyclin androgen receptor (Veldscholte et al., 1990). Although D1 cells (LN/cyl D1-A and LN/cyl D1-B) and with the there are mutations in androgen receptor of LNCaP vector control as LNCaP/pl (LN/pl-A). Northern blot cells, these cells are androgen responsive in cell culture analysis showed that when these cells were grown in (Sonnenschein et al., 1989) and androgen dependent RPMI 1640 medium supplemented with 10% newborn when they form tumors in male nude mice calf serum (NCS), LNCaP/cyclin D1 cells expressed (Horoszewicz et al., 1980, 1983; Lim et al., 1993). more cyclin D1 mRNA than that of the parental LNCaP cells can become more tumorigenic when they LNCaP and LNCaP/pl cells (Figure 2a). To further are mixed with matrigel (Pretlow et al., 1991; Lim et con®rm that each clone was transfected, we also al., 1993) or bone ®broblasts (Wu et al., 1994) and performed Northern blot analysis with a neo gene some tumors become androgen independent in nude fragment as a probe. Figure 2b shows that a 4.8 kb mice (Wu et al., 1994). Due to these characteristics, transcript was detected in LNCaP/cyclin D1 cells and a we chose this cell line to study the relationship 3.7 kb transcript was detected in LNCaP/pl cells (LN/ between cyclin D1 overexpression and androgen pl-A). There were no neo transcripts detected in independence in the progression of prostate cancer LNCaP cells. cells. In doing so, we examined the e€ects of androgen on the expression of cyclin D1 in LNCaP cells. We investigated the e€ects of cyclin D1 overexpression on LNCaP cell-cycle progression, cell proliferation, androgen responsiveness and involvement of the LNCaP other cell-cycle regulators in cell culture. Finally, we studied the tumorigenicity of various LNCaP deriva- DHT(nM) 0 0.3 1 10 30 tives and their response to androgen ablation in male nude mice. cyclin D1 —

Results

Induction of cyclin D1 mRNA expression in LNCaP cells by androgen 7S RNA — Previous in vivo studies show that cyclin D1 expression increases during androgen-induced rat prostate cell proliferation (Furuya et al., 1995; Chen et al., 1996). To determine the relationship between androgen stimulation and cyclin D1 expression in LNCaP cells Figure 1 Induction of cyclin D1 by DHT in LNCaP cells. in vitro, we starved the cells in serum-free medium for 2 LNCaP cells were starved in serum-free medium for 2 days. Then days and then stimulated them with 0.3, 1, 10, 30 nM 0, 0.3, 1, 10, and 30 nM of DHT were added to the medium for 1 of dihydrotestosterone (DHT) for 1 day according to day. About 10 mg of total mRNA from each treatment was the protocol from Dr LWK Chung's lab (The electrophoresed through a 1% formaldehyde-denaturing agarose gel and then transferred to a nylon membrane. The blot was University of Virginia Health Science Center, Charlot- hybridized with an a-32P-dCTP labeled Prad-1 probe. Rehybridi- tesville, Virginia). Northern blot analysis showed that zation to an a-32P-dCTP labeled 7S RNA probe was used to when LNCaP cells were cultured in serum-free normalize RNA loading medium, cyclin D1 mRNA was barely detectable. When DHT was added to the serum-free medium, induction of cyclin D1 mRNA was detected with the ab highest induction level at 1 and 10 nM of DHT (Figure 1).

Overexpression of cyclin D1 in LNCaP cells by a LN/pl-A LN/pl-A LNCaP LNCaP retrovirus cyclin D1 expression vector LN/cyl D1-A LN/cyl D1-B LN/cyl D1-A LN/cyl D1-B The human cyclin D1 expression plasmid, pMV7plCCND1, (Jiang et al., 1993a) and the control cyclin D1 — — Neo vector, pMV7pl, (Kirshmeier et al., 1988) were transferred into the GP+envAm12 amphytropic packaging cell line by electroporation. These two 7S RNA — — 7S RNA plasmids carried the neomycin phosphotransferase (neo) gene which was used as a selection marker. Clones of transfected GP+envAm12 cells were selected Figure 2 Overexpression of cyclin D1 mRNA in LNCaP/cyclin with 400 mg/ml of neomycin and expanded. We D1 transfected cells. About 10 mg of total RNA from LNCaP, harvested the viral supernatant from transfected LNCaP/pl and LNCaP/cyclin D1 cells was electrophoresed GP+envAM12 cells and infected LNCaP cells by a through a 1% formaldehyde denaturing agarose gel and then transferred to a nylon membrane. The blot was hybridized with method previously described (Jiang et al., 1993a). Each an a-32P-dCTP labeled Prad-1 probe (a)oraneo probe (b). colony that grew after G418 selection was picked and Rehybridization to an a-32P-dCTP labeled 7S RNA probe was expanded. We designated the LNCaP cells infected used to normalize RNA loading Cyclin D1 overexpression in LNCaP cells YChenet al 1915 ted with 10% charcoal-stripped FBS (cFBS) and E€ect of cyclin D1 overexpression on LNCaP growth in various concentrations of androgen (0, 0.001 n and vitro M 0.1 nM of DHT). The concentration of steroid LNCaP cells proliferate well in RPMI 1640 medium hormones (including androgen) is extremely low in supplemented with 5 ± 10% serum, and respond to this medium because of the special charcoal treatment androgen (DHT) in the culture medium (Horoszewiez (Breen and Leake, 1987). When 10% cFBS plus 0 or et al., 1983). To characterize the LNCaP/cyclin D1 cell 0.001 nM of DHT were added into the medium, growth properties in vitro, we analysed cell growth in LNCaP and LNCaP/pl cells grew slowly, as did the medium supplemented with 5% and 10% NCS. The LNCaP/cyclin D1 cells, but to a less extent. When the LNCaP/cyclin D1 cells and their nuclei were smaller cells were fed with the medium supplemented with 10% than those of the parental and vector-control cells but cFBS plus 0.1 nM of DHT, all of these cells grew faster were spindle shaped and loosely attached to the surface (Figure 4). These results indicated that overexpression of cell-culture ¯asks like the parental cells. When the of cyclin D1 in LNCaP cells partially reduced the cell's cells were grown in medium with 10% NCS, ¯ow requirement for androgen stimulation in vitro. cytometry analysis showed that more LN/cyclin D1 To understand the mechanism of the di€erent cells than LNCaP and LN/pl cells were in S phase androgen requirements among the LNCaP and its during the exponential phase of cell growth (Table 1). derivative cell lines, we further studied the involvement LNCaP/cyclin D1 cells doubled faster than LNCaP of cell cycle-related proteins in the regulation of these and LNCaP/pl cells (average 37 h versus 51 h cell proliferation in vitro. It is known that cyclin D1 respectively) and had greater cell saturation densities can form complexes and activate cdk4 in mid to late (Table 1, Figure 3). The higher saturation densities also G1 phase (Xiong et al., 1992a,b; Kato et al., 1993). indicate that the LNCaP/cyclin D1 cells were smaller Once cdk4 is activated, it can further phosphorylate than the parental and control cells (Table 1). When the downstream substrates such as the Rb protein in G1 cells were grown in medium with 5% NCS, LNCaP/ phase (Sherr, 1993, 1994). Our in vitro kinase assay, cyclin D1 cells also grew faster than LNCaP and using pRb as the substrate, showed that LNCaP and LNCaP/pl cells (data not shown). Thus, LNCaP/cyclin its derivative cells had high cdk4 enzyme activities in D1 cells had a growth advantage over LNCaP and medium with 10% FBS and their activities reduced LNCaP/pl cells in the RPMI medium supplemented dramatically in medium with 10% cFBS. However, with 10% or 5% serum. LNCaP/cyclin D1 cells had higher cdk4 kinase activity To examine the androgen dependence of LNCaP/ than LNCaP/pl cells in both types of media (Figure cyclin D1 cells, we analysed the growth rates of 5A). Immunoprecipitation of the protein lysates with LNCaP/cyclin D1, LNCaP/pl and LNCaP cells cyclin D1 antibody and Western blot analysis with cultured in RPMI phenol-red free medium supplemen- cyclin D1 and cdk4 antibodies showed that there were more cyclin D1 protein and cyclin D1/cdk4 complexes detected when the cells were grown in medium with Table 1 Comparison of cell growth properties and cell cycle 10% FBS, and were less when the cells were grown in a parameters medium with 10% cFBS. However, the levels of cyclin LN/cyl LN/cyl D1 protein and cyclin D1/cdk4 complex formed in LNCaP LN/pl-A D1-A D1-B LNCaP/cyclin D1 were always higher than that in Doubling time 50.5 52.0 36.4 37.5 LNCaP and LNCaP/pl cells under both conditions (hours) (Figure 5B). The reduction in the cyclin D1/cdk4 Saturation density 156105 386105 606105 846105 Cells in G1 (%) 69.3+5 71.9+4.7 59.9+4.1 51.0+3.5 complex levels may be related to the decrease of cyclin Cells in S (%) 22.4+3.1 19.1+1.7 27.5+3.7 31.9+1.4 D1, but not cdk4 protein levels, after androgen Cells in G2 (%) 8.3+2.4 9.0+2.0 12.6+1.9 17.1+0.6 deprivation in the cell culture medium (Figure 5C). aCells were grown in medium with 10% NCS, and the cell cycle We further investigated the status of other important parameters were examined during exponential phase of cell growth

Figure 4 E€ect of cyclin D1 overexpression on androgen Figure 3 Growth of LNCaP/cyclin D1 and control cells in dependence of LNCaP cells. Cells (26105) were seeded in a culture medium with 10% NCS. LNCaP, LNCaP/pl and LNCaP/ T25 ¯ask with medium with 10% NCS. After 1 day, these cells cyclin D1 cells (56105 each) were seeded in a T25 ¯ask with were washed twice with PBS and switched to RPMI phenol red- medium containing 10% NCS on day 0. The cells were counted free medium supplemented with 10% cFBS and 0, 0.001 or 2 ± 3 days 0.1 nM of DHT. The cells were counted after 8 days Cyclin D1 overexpression in LNCaP cells YChenet al 1916 cell-cycle regulators, such as cyclin E, cdk2, Rb and protein in late G1 phase. Figure 6 shows that after E2F-1 proteins (Sherr, 1993, 1994; Sherr and Roberts, androgen deprivation in cell culture, the protein levels 1995; Weinberg, 1991). Cyclin E form complexes with of cyclin E, cdk2 (detected by Western blot analysis) and activate cdk2 which can also phosphorylate Rb and cdk2 kinase activity (detected by in vitro kinase assay using pRb as a substrate) were signi®cantly reduced in LNCaP and its derivative cells. Rb is another very important protein which inhibits cells 10% FBS 10% cFBS from going through the G1/S checkpoint (Weinberg, 1991; Hinds and Weinberg, 1994). The hypopho- sphorylated forms of Rb can bind and suppress the function of E2F-1, which is a transcription factor for the genes required for DNA synthesis in S phase LN/pl-A LN/pl-A LN/cyl D1-A LN/cyl D1-A LNCaP LNCaP (Nevins, 1992). Hyperphosphorylation of Rb relieves this inhibition by releasing E2F-1 (Chellappan et al., pRb A α 1991; Hinds and Weinberg, 1994). Western blot ( -cdk4) analysis showed that LNCaP/cyclin D1 cells had higher levels of hyperphosphorylated Rb as well as E2F-1 protein than parental and vector-control cells, cyclin D1 even if these levels were reduced when the cells were B grown in medium without androgen (10% cFBS). These di€erences were more clear with cells grown in cdk4 medium without androgen (Figure 7).

Cyclin D1 overexpression enhanced the tumorigenicity of LNCaP cells in male nude mice cyclin D1 LNCaP cells have very weak tumorigenicity in nude C mice. They must be injected with Matrigel (Pretlow et al., 1991; Lim et al., 1993) or bone ®broblast cells (Wu cdk4 et al., 1994) to form tumors in male nude mice. The growth of these tumors is mostly androgen dependent Figure 5 Function and expression of cyclin D1 protein in since regression of the tumors is observed after LNCaP and its derivative cells before and after androgen ablation in culture. The enzyme activity of cdk4 was analysed by a kinase castration (Lim et al., 1993; Wu et al., 1994). assay using pRb as the substrate. Protein lysates were However, some tumors can lose androgen dependence immunoprecipitated with anti-cdk4 andibody and (A). Protein and grow again after castration (Wu et al., 1994). To lysates were also immunoprecipitated with anti-cyclin D1 anti- investigate the e€ects of cyclin D1 overexpression on body and Western blotted with anti-cyclin D1 and cdk4 the tumorigenicity of LNCaP cells, we injected 26106 antibodies (B). Cyclin D1 and cdk4 expression in the cells treated with or without androgan in cell culture medium were cells/site in 0.25 ml of phosphate-bu€ered saline (PBS) analysed by Western blot analysis (C) with Matrigel (1 : 1) subcutaneously into intact male nude mice. The earliest visible tumor appeared 2 weeks after injection of LN/cyl D1-A, LN/cyl D1-B and LN/ cyl D1-C cells (which also expressed a high level of 10% FBS 10% cFBS cyclin D1, data not shown); 6 ± 7 weeks after injection of LNCaP cells and 6 weeks after injection of LNCaP/ pl (LN/pl-A cells) (Figure 8). By 4 ± 5 weeks after cell LNCaP LN/pl-A LN/cyl D1-A LNCaP LN/pl-A LN/cyl D1-A 10% FBS 10% cFBS cyclin E

A

cdk2 LNCaP LN/pl-A LN/cyl D1-A LNCaP LN/pl-A LN/cyl D1-A

Rb

pRb B E2F-1 (α-cdk2)

Figure 6 Expression and function of cyclin E and cdk2 protein Figure 7 Rb phosphorylation and E2F-1 expression levels in in LNCaP and its derivative cells before and after androgen LNCaP and its derivative cells before and after androgen ablation ablation in culture. Expression of cyclin E and cdk2 protein in the in culture. Rb phosphorylation and E2F-1 expression levels in cells treated with or without androgen in cell culture medium LNCaP and its derivative cells cultured in medium with and were examined by Western blot analysis (A). The enzyme without androgen were analysed by Western blot analysis. activities of cdk2 in the cells under both conditions were Polyclonal anti-Rb and anti-E2F-1 antibodies were used in the analysed by kinase assay using pRb as the substrate (B) analysis Cyclin D1 overexpression in LNCaP cells YChenet al 1917 injection, 58.3% of the mice injected with LN/cyl D1-B tumors from LNCaP cells grew 5 weeks later after cells, 53.3% of the mice with LN/cyl D1-C cells, 28.0% castration. Thus, cyclin D1 overexpression in LNCaP of the mice with LN/pl-A cells and 23.0% of the mice cells enhances tumorigenicity of the cells and inhibits with LNCaP cells had tumors (Table 2). The tumor regression after castration. tumorigenicity data from LN/cyl D1-A cells was not included in Table 2 because their tumors were already over 500 mm3 by 5 ± 6 weeks after injection. If we did Discussion not castrate or sacri®ce the mice, the tumors would grow too big that mice were in poor health. Thus, the In this study, we investigated the e€ects of cyclin D1 tumors from LNCaP/cyclin D1 cells grew better than overexpression on cell proliferation and tumorigenicity the parental and control cells in vivo. We also injected in the LNCaP cell line. Compared with parental and LNCaP, LN/pl and LNCaP/cyclin D1 cells into female vector-control cells, cyclin D1-transfected clones were and castrated male nude mice with Matrigel or intact smaller, had a larger proportion of cells in S phase and male nude mice without Matrigel. No tumors formed grew faster. They were less dependent on androgen in during a 12-week observation period. Therefore, culture, developed tumors earlier and were refractory without Matrigel and androgen, overexpression of to androgen-ablation treatment in nude mice. cyclin D1 alone could not initiate the LNCaP tumor The pMV7plCCND1 plasmid used in this study was formation in nude mice. developed in Dr IB Weinstein's lab (Jiang et al., To examine whether overexpression of cyclin D1 in 1993a). Rodent embryo ®broblasts cells transfected LNCaP leads to resistance to hormone-ablation in vivo, with this plasmid overexpress cyclin D1, are smaller we castrated the mice when they developed tumors of and have an increased saturation density. They have about 200 ± 300 mm3 and determined tumor size weekly more cells in S phase but the same doubling time as the thereafter. Figure 9 shows that the tumors derived parental and vector-control cells. These cells form from LNCaP and LNCaP/pl cells gradually regressed tumors in nude mice but vector-control cells do not within 4 weeks after castration. On the other hand, (Jiang et al., 1993a). The results obtained in our study tumors from LNCaP/cyclin D1 cells did not regress but are consistent with those of Weinstein's group except kept growing after castration. Consistent with other that our LNCaP/cyclin D1 cells have a faster doubling reports (Lim et al., 1993; Wu et al., 1994), some of the time than the parental and vector control cells. This variation may be due to the di€erences between the cell lines (normal rodent embryo ®broblasts cells versus human malignant prostate cells). The growth of LNCaP cells is androgen dependent since when cultured in an androgen-free medium (10% cFBS), their growth is greatly inhibited (Sonnenschein et al., 1989). In this study, we showed that androgen ablation had a smaller inhibitory e€ect on LNCaP/cyclin D1 cells than the LNCaP and LNCaP/pl cells. Thus, constitutive overexpression of cyclin D1 in LNCaP led to more cells being in S phase and decreased the dependence of the cells on physiological stimuli from androgen. In the G1 phase of cell cycle, cyclin D1 functions by activating the kinase activities of cdk4 and cdk6. The activated kinases can further phosphorylate Rb at mid to late G1 phase (Sherr, 1993, 1994). Our studies showed that when cells were grown in medium with Figure 8 In¯uence of cyclin D1 overexpression on tumorigenicity of LNCaP cells in male nude mice. LNCaP, LNCaP/pl and LNCaP/cyclin D1 cells (26106 cells/site) were mixed in 0.25 ml of PBS with Matrigel (1 : 1) and injected subcutaneously into intact male nude mice. The mice were examined weekly after injection. The incidence of tumors from each cell line was the percentage of mice with tumors

Table 2 E€ect of cyclin D1 on the tumorigenicity of LNCaP cells LN/cyl LN/cyl LNCaP LN/pl-A D1-B D1-C Number of nude 13 7 12 10 mice Tumor incidence 23.0 28.0 58.3 53.3 (%)a Mean tumor size 24 125 251 554 Figure 9 E€ects of cyclin D1 overexpression in LNCaP/cyclin (mm3)b D1 on the refraction of androgen ablation in male nude mice. aThe tumor incidence shows the percentage of mice that developed When the tumors formed from LNCaP and its derivative cells tumors within 8 weeks after being injected with 26106 cells. bThe were about 200 ± 300 mm3, the mice were castrated by bilateral mean tumor size was determined 8 weeks after cell injection. The orchiectomy under metofane anesthesia. The tumors were formula for calculating each tumor size was: L6W6H60.5236 measured weekly for castration Cyclin D1 overexpression in LNCaP cells YChenet al 1918 10% FBS, LNCaP/cyclin D1 cells had higher cdk4 Thus, overexpression of cyclin D1 enhances prostate kinase activity and higher levels of Rb phosphorylation tumor growth in male nude mice and reduces tumor than LNCaP and LNCaP/pl cells. Thus, these results dependence on androgen in vivo. Interestingly, LNCaP/ were consistent with the function of cyclin D1 cyclin D1 cells showed higher resistance to androgen previously described. E2F-1 is another important ablation in vivo than in vitro. One of the reasons may protein in cell cycle regulation because it regulates be that tumor formation and growth in the nude mice the expression of genes required for DNA synthesis in are a more complicated process than cell growth in S phase, including dehydrofolate reductase, DNA vitro, which is a di€erent growth environment for cells. polymerase a, thymidine kinase and cdc2 (reviewed The degree of tumorigenicity is not only determined by by Nevins 1992). The transcription of E2F-1 is also the ability of cell proliferation but also by the known to be activated by itself (Johnson et al., 1994). interaction between host and tumor cells, as well as The function of E2F-1 is regulated by the members of other events, such as angiogenesis and genetic the Rb protein. Hypophosphorylated Rb binds to E2F- instability during tumor formation and growth. Thus, 1 and inhibit E2F-1 from activating gene transcription. the mechanisms that lead to androgen resistance of the Phosphorylation of Rb relieves this inhibition by tumors formed from LNCaP/cyclin D1 cells are very releasing E2F-1 which leads to activation of this complicated and will be further studied. We also found protein and drives cells into S phase (Chellappan et that cyclin D1-overexpressing LNCaP cells, just like al., 1991; Hinds and Weinberg, 1994; Johnson, 1995). LNCaP cells, could not form tumors in male nude mice Our studies showed that when cells were grown in without Matrigel in female nor in castrated male nude medium with 10% FBS, LNCaP/cyclin D1 cells had mice with Matrigel. These results suggest that over- higher expression levels of E2F-1 than LNCaP and expression of cyclin D1 alone is not sucient for LNCaP/pl cells. This observation is consistent with initiating LNCaP tumor formation in nude mice. Other previous reports from other labs stating that cyclin D1/ factors such as androgen, components in the Matrigel cdk4 induces E2F-1 indirectly by phosphorylation of (including extracellular matrix proteins and growth Rb resulting in the release of E2F-1 which then factors such as ®broblast growth factors, tissue activates its own transcription (Schulze et al., 1994; plasminogen activators, etc.) and other biological Oswald et al., 1994; Vairo et al., 1995; Johnson, 1995; changes are required to initiate tumor formation. Shao and Robbins, 1995). Many studies have shown that resistance to androgen Although cyclin D1 is overexpressed in LNCaP/ ablation in advanced prostate cancer may be the result cyclin D1 cells, cell proliferation was greatly inhibited of genetic alterations and deregulation of gene when these cells were grown in medium supplemented expression. Several groups have reported that muta- with 10% cFBS, as that of the LNCaP and vector- tions in androgen receptors or reduction in its control cells, but to a less degree. The medium expression may be related to androgen resistance in supplemented with 10% cFBS contains most of the advanced prostate cancer (Tilly et al., 1994; Ruizeveld growth factors and hormones except steroid hormones de Winter et al., 1994). Alterations in tumor suppressor which includes androgen. Thus, androgen is an genes, for example, p53 mutation are also associated important mitogen for the LNCaP, vector-control as with hormone resistance in advanced prostate cancer well as its cyclin D1 overexpressing cells. The kinase (Navone et al., 1993; Voeller et al., 1994; Kallakury et assay and Western blot analysis performed in this al., 1994). In addition, the Rb gene is mutated in 20 ± study also indicate that androgen plays a role in the 50% of advanced prostate cancers, while mutation is regulation of the cell-cycle machinery. The expression rare in early tumors (Isaacs et al., 1994). In this study, levels of cyclin D1, cyclin E, cdk2, E2F-1, phosphor- we observed that androgen has an important role in the ylation of Rb protein and cyclin D1/cdk4 complex regulation of the cell-cycle machinery including induc- formation, as well as kinase activities of cdk4 and cdk2 tion of cyclin D1 leading to phosphorylation of Rb and were reduced in all of these cells after androgen release of E2F-1 transcription factor in the LNCaP deprivation from the medium. Hence, these changes cells. Thus, changes at any points in this pathway may may be a result of the androgen deprivation and part lead to the inactivation of Rb which represents a critical of the factors leading to inhibition of cell proliferation. event during tumorigenesis of prostate. The results from In LNCaP/cyclin D1 cells, there was a less reduction of our study suggest that overexpression of cyclin D1 in cyclin D1, E2F-1 protein expression, Rb phosphoryla- LNCaP cells cooperates with other biological altera- tion levels and cdk4 kinase activity. This is related to tions in the cells and facilitates the cell growth, tumor the constitutively higher expression levels of cyclin D1 formation and resistance to androgen ablation both in which may lead to less growth inhibition of these cells vitro and in vivo. in androgen deprived environment. LNCaP cells are not tumorigenic when they are injected alone into nude mice. When they are injected Materials and methods with Matrigel (Pretlow et al., 1991; Lim et al., 1993) or bone ®broblast cells (Wu et al., 1994), they form Cell culture tumors in male mice. LNCaP cells are androgen dependent not only in vitro but also in vivo as their The human prostate cancer cell line LNCaP (American tumors regress after castration (Lim et al., 1993; Wu et Type Culture Collection, Rockville, MD) was cultured in RPMI 1640 medium (GIBCO, Gaithersburg, MD) supple- al., 1994). In our study, LNCaP/cyclin D1 cells mixed mented with NCS or FBS (Hyclone, Logan, Utah). The with Matrigel formed tumors in male nude mice earlier amphytropic retrovirus packaging cell line, GP+envAm12, than the parental and vector-control cells did. After a gift from Dr Yacob Ben-David and Dr J Filmus castration, these tumors did not regress, whereas the (Sunnybrook Health Science Center, Toronto, Canada), tumors from parental and vector-control cells did. was cultured in Dulbecco's modi®ed Eagle's medium Cyclin D1 overexpression in LNCaP cells YChenet al 1919 (DMEM) with high glucose, and 10% NCS. Both cell lines Biotechnology, Inc. (Santa Cruz, CA). A secondary were also supplemented with 2 mM of glutamine and 50 mg/ antibody conjugated with horseradish peroxidase (Amer- ml of gentamicin and incubated in a humidi®ed atmosphere sham Corp., Arlington Heights, IL) was used and the with 5% CO2. To examine the dependence of LNCaP and ®nal detection was performed by using an ECL kit its derivatives on androgen, we cultured the cells in RPMI (Amersham Corp.). For the kinase assay, we followed the 1640 phenol red-free medium (Sigma Chemical Co., St. protocol from Dr CJ Sherr's Lab (St. Jude Children's Louis, MO) supplemented with 10% FBS or 10% cFBS Research Hospital, Memphis, TN) (Matsushime et al., with various concentrations of DHT (Sigma Chemical 1994). The frozen protein lysates were thawed at 48Cand Co.). The dextran and activated charcoal used in the 500 mg of protein lysates from each sample were preparation of cFBS were obtained from Sigma Chemical immunoprecipitated with either cdk4 or cdk2 antibodies Co. (Santa Cruz, CA). pRb, a GST fusion protein produced in E. coli (Santa Cruz, CA), was used as substrate in the kinase assay. The phosphorylation of pRb was labeled by Cyclin D1 retroviral expression plasmid transfection g-32P-ATP. For quanti®cation of cyclin D1/cdk4 com- The cyclin D1 expression plasmid pMV7plCCND1 (Jiang plexes formed in the cells, protein lysates were et al., 1993a) and the control vector pMV7pl (Kirshmeir, immunoprecipitated with cyclin D1 antibody and 1988) were gifts from Dr IB Weinstein (Columbia Western blotted with cdk4 antibodies. University, New York, NY). The plasmid pMV7plCCND1 contains a 1.1 kb human cyclin D1 In vitro cell growth analysis cDNA and the 5' Moloney murine leukemia virus long terminal repeat (MoMuLV LTR) as the cyclin D1 To characterize cell growth in vitro, we performed ¯ow promoter. A neo gene was included as a selection marker cytometric analysis as well as studies of cell growth curves (Jiang et al., 1993a). We transfected pMV7plCCND1 or and cell population doubling time. For the ¯ow pMV7pl into GP+envAM12 cells by electroporation. cytometric analysis, exponentially growing cells in Forty-eight hours later, cells were trypsinized and replated medium with 10% NCS were trypsinized, washed with in DMEM with 10% NCS and 400 mg/ml of G418. G418- cold PBS and ®xed in 75% ethanol. Propidium iodide resistant clones were selected and expanded. To transfect with 2 mg/ml RNase was used to stain the cells. DNA pMV7plCCND1 or pMV7pl into LNCaP cells, we content and the percentage of cells in various phases of harvested the viral supernatant of the G418-resistant the cell cycle were determined by measuring ¯uorescent GP+envAm12 cells and infected LNCaP cells by a intensity with a Coulter Elite Flow cytometer. To examine method previously described (Jiang et al., 1993a). Forty- cell growth, 56105 cells were plated in each of three eight hours after transfection, 400 mg/ml of G418 was 25 mm2 ¯asks in DMEM with 10%, 5% NCS or 26105 added to the RPMI medium with 10% NCS. Individual cells in RPMI 1640 phenol red-free medium supplemented colonies were picked and grown in RPMI medium with with 10% cFBS with 0, 0.001, 0.1 nM of DHT. Cells were 10% NCS and 100 mg/ml of G418 for further analysis. counted every 2 ± 3 days with a hematocytometer. The cells were fed twice a week. Northern blot analysis Tumorigenicity assay in nude mice Total RNA was isolated from cells by using TRI-reagent (Molecular Research Center, Inc., Cincinnati, OH). Ten Male nude mice were purchased from Frederick Cancer micrograms of total cellular RNA from each sample was Research and Development Center (Frederick, MD) and electrophoresed through a 1.5% agarose gel containing 8% kept aseptically in an athymic animal room. The pMV7pl- formaldehyde. After electrophoresis, the RNA was and pMV7plCCND1-transfected LNCaP cells and parental transferred to a nylone membrane by the capillary LNCaP cells were trypsinized, washed with PBS, resus- method, and then cross-linked to the membrane with pended in PBS and mixed with an equal volume of ultraviolet light. Prad-1 and neo probes were labeled with Matrigel (Collaborative Biomedical Products/Becton Dick- a-32P-dCTP by using a random primed DNA labeling kit inson, Bedford, MA). Then 26106 cells in 0.25 ml of the (Boehringer Mannheim Corporation, Indianapolis IN). mixture were injected subcutaneously into nude mice by a The Prad-1 probe, a human cyclin D1 probe, was kindly method previously described (Lim et al., 1993). The mice provided by Dr A Arnold (Massachusetts General underwent bilateral orchiectomy under metofane anesthesia Hospital, Boston, MA). The hybridization was performed (Pitman-Moore, Washington Crossing, NJ) when the at 688C with Quickhyb solution (Strategene, La Jolla, CA). tumors were 200 ± 300 mm3. All the mice were examined The next day, each membrane was washed twice for 15 min once a week, and each tumor's length (L), width (W) and each time at room temperature with 26SSC and 0.1% thickness (H) were measured. Tumor volume was SDS and once for 30 min at 688C with 0.16SSC and 0.1% calculated by the formula L6W6H60.5236 (Janek et SDS and then exposed to Kodak X-AR5 ®lm for 1 day at al., 1975). 7708C. Rehybridization with a-32P-dCTP-labeled 7S RNA was used to normalize the amount of RNA loaded in each line. Acknowledgements We would like to thank Dr IB Weinstein for providing the pMV7plCCND and pMV7pl plasmids; Dr Yacob Ben- Western blot analysis and Rb kinase assay David and Dr J Filmus for providing GP+envAM12 cells; Protein was extracted from cells and kept frozen at Dr A Arnold for providing the Prad-1 (cyclin D1)probe; 7808C by methods described previously (Chen, 1995). Dr LWK Chung for sharing their protocol of DHT One hundred micrograms of protein lysate from each supplementation in cell culture and Dr CJ Sherr for sample was electrophoresed through a 10% SDS- providing their protocol for kinase assays. Dale Wiss and polyacrylamide gel, transferred to a nitrocellulose Robert B Whalin for their extensive assistance with the membrane by wet transfer and blotted with various animal surgery and care; Dennis Walker for ¯ow cyto- antibodies. Polyclonal anti-cyclin D, cdk4, cyclin E, cdk2, metric analysis; Judy G Ing for artwork and Dr Maureen Rb and E2F-1 antibodies were from Santa Cruz Goode for editing the paper. Cyclin D1 overexpression in LNCaP cells YChenet al 1920 References

Baldin V, Lukas J, Marcote MJ, Pagano M and Draetta G. Musgrove EA, Lee CSL, Buckley MF and Sutherland RL. (1993). Genes & Dev., 7, 812 ± 821. (1994). Proc. Natl. Acad. Sci. USA, 91, 8022 ± 8026. Breen B and Leake RE. (ed). (1987). Steroid Hormones: A Navone NM, Troncoso P, Piters LL, Goodrow TL, Palmer Practical Approach. IRL Press: London. JL, Nichols WW, von Eschenbach AC and Conti CJ. Buckley MF, Sweeney KJE, Hamilton JA, Sini RL, Manning (1993). J. Natl. Cancer Inst., 85, 1657 ± 1669. DL, Nicholson RI, deFazio A, Watts CKW, Musgrove Nevins JR. (1992). Science, 258, 424 ± 429. EA and Sutherland RL. (1993). Oncogene, 8, 2127 ± 2133. Oswald F, Lovec H, MoÈ roÈ y T and Lipp M. (1994). Oncogene, Chellappan SP, Hiebert S, Mudryj M, Horowitz JM and 9, 2029 ± 2036. Nevins JR. (1991). Cell, 65, 1053 ± 1061. Pretlow TG, Delmoro CM, Dilley GG, Spadafora CG and Chen Y, Navone NM and Conti CJ. (1995). Urol. Oncol., 1, Prelow TP. (1991). Cancer Res., 51, 3814 ± 3817. 101 ± 108. Quelle DE, Ashmun RA, Shurtle€ SA, Kato J-Y, Bar-Sagi Chen Y, Robles AI, Martinez LA, Liu F, Gimenez-Conti IB D, Roussel MF and Sherr CJ. (1993). Genes & Dev., 7, and Conti CJ. (1996). Cell Growth Di€er., 7, 1571 ± 1578. 1559 ± 1571. Furuya Y, Lin XS, Walsh JC, Nelson WG and Isaacs JT. Resnitzky D, Gossen M, Bujard H and Reed S. (1994). Mol. (1995). Endocrinology, 136, 1898 ± 1906. Cell. Biol., 14, 1669 ± 1679. Hinds PW and Weinberg RA. (1994). Curr. Opin. Genet. Rosenberg CL, Kim HG, Shows TB, Kronenberg HM and Dev., 4, 135 ± 141. Arnold A. (1991). Oncogene, 6, 449 ± 453. Horoszewicz JS, Leong SS, Chu TM, Wajsman ZL, Ruizeveld de Winter JA, Jannssen PJ, Sleddens HM, Friedman M, Papsidero L, Kim U, Chai LS, Kakati S, Verleiun-Mooijman MC, Trapman J, Brinkmann AO, Arya SK and Sandberg AA. (1980). Prog. Clin. Biol. Res., Santerse AB, Schroder FH and van der Kwast TH. (1994). 37, 115 ± 132. Am. J. Pathol., 144, 735 ± 746. Horoszewicz JS, Leong SS, Kawinski E, Karr JP, Rosenthal Schulze A, Zerfa K, Spitkovsky D, Henglein B and Jansen- H, Chu TM, Mirand EA and Murphy GP. (1983). Cancer DuÈ rr P. (1994). Oncogene, 9, 3475 ± 3482. Res., 43, 1809 ± 1818. Schuuring E, Verhoeven E, Mooi WJ and Michalides RJ. Israeli RS, Powell CT, Corr JG, Fair WR and Heston WDW. (1992). Oncogene, 7, 355 ± 361. (1994). Cancer Res., 54, 1807 ± 1811. Shao A and Robbins PD. (1995). Oncogene, 10, 221 ± 228. Issacs WB, Bova GS, Morton RA, Bussemakers MJ, Brooks Sherr CJ and Roberts JM. (1995). Genes & Dev., 9, 1149 ± JD and Ewing CM. (1994). Cold Spring Harb. Symp. 1163. Quant. Biol., 59, 653 ± 659. Sherr CJ. (1993). Cell, 73, 1059 ± 1065. Janek P, Briand P and Hartman NR. (1975). Cancer Res., 35, Sherr CJ. (1994). Cell, 79, 551 ± 555. 3698 ± 3704. Sonnenschein C, Olea N, Pasanen E and Soto AM. (1989). Jiang W, Kahn SM, Tomita N, Zhang YJ, Lu SH and Cancer Res., 49, 3474 ± 3481. Weinstein IB. (1992). Cancer Res., 52, 2980 ± 2983. Tam SW, Theodoras AM, Shay JW, Draetta GF and Pagano Jiang W, Kahn SM, Zhou P, Zhang Y-J, Cacace AM, Infante M. (1994). Oncogene, 9, 2663 ± 2674. AS, Doi S, Santella RM and Weinstein IB. (1993a). Tilly WD, Lim-Tio SS, Horsfall DJ, Aspinall JO, Marshall Oncogene, 8, 3447 ± 3457. VR and Skinner JM. (1994). Cancer Res., 54, 4096 ± 4102. Jiang W, Zhang Y-J, Kahn SM, Hollstein MC, Santella RM, Vairo G, Livingston DM and Ginsberg D. (1995). Genes & Lu S-H, Harris CC, Montesano R and Weinstein IB. Dev., 9, 869 ± 881. (1993b). Proc. Natl. Acad. Sci. USA, 90, 9026 ± 9030. Veldscholte J, Ris-Stalpers C, Kuiper GG, Jenster G, Johnson DG. (1995). Oncogene, 11, 1685 ± 1692. Berrevoets C, Claassen E, van Rooij HC, Trapman J, Johnson DG, Ohtani K and Nevins JR. (1994). Genes & Dev., Brinkman AO and Mulder E. (1990). Biochem. Biophys. 8, 1514 ± 1525. Rev. Commun., 173, 534 ± 540. Kallakury BV, Figge J, Ross JS, Fisher HA, Figge HL and Voeller HJ, Sugars LY, Pretlow T and Gelman EP. (1994). J. Jennings TA. (1994). Hum. Pathol., 25, 92 ± 97. Urol., 151, 492 ± 495. Kato J-Y, Matsushime H, Hiebert SW, Ewen ME and Sherr Weinberg RA. (1991). Science, 254, 1138 ± 1146. CJ. (1993). Genes & Dev., 7, 331 ± 342. Wu H-C, Hsieh J-T, Gleave ME, Brown NM, Pathak S and Kirshmeir PT, Housey GM, Johnson MD, Perkins AS and Chung LW. (1994). Int. J. Cancer, 57, 406 ± 412. Weinstein IB. (1988). DNA, 7, 219 ± 225. Xiong Y, Connolly T, Futcher B and Beach D. (1991). Cell, Lammie GA, Fant IV, Smith R, Schuuring E, Brookes S, 65, 691 ± 699. Michalides R, Dickson C, Arnold A & Peters G. (1991). Xiong Y, Menninger J, Beach D and Ward DC. (1992a). Oncogene, 6, 439 ± 644. Genomics, 13, 575 ± 584. Lim DJ, Liu X-L, Sutkowski DM, Braun EJ, Lee C and Xiong Y, Zhang H and Beach D. (1992b). Cell, 71, 505 ± 514. Kozlowski JM. (1993). Prostate, 22, 109 ± 118. Zhang Y-J, Jiang W, Chen CJ, Lee CS, Kahn SM, Sanntwlla Lukas J, Jadayel D, Bartkova J, Nacheva E, Dyer MJ, RM and Weinstein IB. (1993). Biochem. Biophys. Res. Strauss M and Bartek J. (1994a). Oncogene, 9, 2159 ± 2167. Commun., 196, 1010 ± 1016. Lukas J, Pagano M, Staskova Z, Draetta G, Bartek J. Zhou P, Jiang W, Zhang Y-J, Kahn SM, Schieren I, Santella (1994b). Oncogene, 9, 707 ± 718. RM and Weinstein IB. (1995). Oncogene, 11, 571 ± 580. Matsushime H, Quelle DE, Shurtle€ SA, Shibuya M, Sherr CJ and Kato J-Y. (1994). Mol. Cell. Biol., 14, 2066 ± 2076. Montgomery BT, Young CY, Bilhartz DL, Andrews PE, Prescott JL, Thompson NF and Tindall DJ. (1992). Prostate, 21, 63 ± 73.