¹Department of Clinical Biochemistry, Faculty of Health Sciences, Ben-Gurion University of the Negev and Soroka Medical Center of Kupat Holim, Beer-Sheva, Israel

²Cancer Research Program, Garvan Institute of Medical Research, St. Vincent's Hospital, Sydney, New South Wales 2010, Australia

Correspondence to: Y Sharoni, Department of Clinical Biochemistry, Faculty of Health Sciences, Ben-Gurion University of the Negev, P.O. Box 653, Beer-Sheva 84105, Israel

Numerous studies have demonstrated the anticancer activity of the tomato carotenoid, lycopene. However, the molecular mechanism of this action remains unknown. Lycopene inhibition of human breast and endometrial cancer cell growth is associated with inhibition of cell cycle progression at the G₁ phase. In this study we determined the lycopene-mediated changes in the cell cycle machinery. Cells synchronized in the G₁ phase by serum deprivation were treated with lycopene or vehicle and restimulated with 5% serum. Lycopene treatment decreased serum-induced phosphorylation of the retinoblastoma protein and related pocket proteins. This effect was associated with reduced cyclin-dependent kinase (cdk4 and cdk2) activities with no alterations in CDK protein levels. Lycopene caused a decrease in cyclin D1 and D3 levels whereas cyclin E levels did not change. The CDK inhibitor p21^Cip1/Waf1 abundance was reduced while p27^Kip1 levels were unaltered in comparison to control cells. Serum stimulation of control cells resulted in reduction in the p27 content in the cyclin E-cdk2 complex and its accumulation in the cyclin D1-cdk4 complex. This change in distribution was largely prevented by lycopene treatment. These results suggest that lycopene inhibits cell cycle progression via reduction of the cyclin D level and retention of p27 in cyclin E-cdk2, thus leading to inhibition of G₁ CDK activities. Oncogene (2001) 20, 3428-3436.

lycopene; breast cancer; endometrial cancer; cell cycle; cyclin D1; retinoblastoma; p27^kip1

Introduction

Numerous epidemiological studies have demonstrated that consumption of vegetables and fruit reduces the risk of breast and other types of human cancers. Recently, there has been a growing interest in the tomato carotenoid, lycopene, as a cancer preventive agent. Epidemiological studies have suggested that lycopene decreases the risk of several types of human malignancies, such as breast (Zhang et al., 1997), prostate (Giovannucci et al., 1995) and lung (Michaud et al., 2000) cancer. Giovannucci reviewed data from clinical and epidemiological studies and found that most of these studies show an inverse association between tomato intake or blood lycopene level and the risk of cancer (Giovannucci, 1999). These epidemiological data are reinforced by studies showing the inhibitory effect of lycopene on tumor cell growth in vitro (Amir et al., 1999; Countryman et al., 1991; Wang and Lin, 1989) and in vivo (Kim et al., 1997; Kobayashi et al., 1996; Nagasawa et al., 1995; Narisawa et al., 1996; Sharoni et al., 1997; Wang et al., 1989).

We have previously demonstrated that lycopene inhibits the growth of mammary, endometrial, lung cancer (Levy et al., 1995) and leukemic (Amir et al., 1999) cells. This inhibition was associated with a delay in cell cycle progression from G₁ to S phase. However, data regarding its effect on cell cycle mechanisms are not available. In MCF-7 mammary cancer cells, the inhibitory effect of lycopene was more pronounced in cells partially synchronized in early G₁ phase by serum starvation followed by restimulation with serum-containing medium, causing re-entry into the cell cycle (Karas et al., 2000). Therefore, this model was used in the current study to investigate the molecular mechanisms involved in lycopene inhibition of G₁ phase progression in human breast (MCF-7, T-47D) and endometrial (ECC-1) cancer cells.

Recently, IGF-I has been suggested as an important risk factor for breast, prostate and colorectal cancer (Hankinson et al., 1998; Ma et al., 1999; Mantzoros et al., 1997). Therefore, the possibility that diet components such as lycopene may decrease the risk of growth factor-related cancer warrant examination. It is well-documented that growth factors affect the cell cycle apparatus primarily during G₁ phase, and that the main components acting as growth factor sensors are the D-type cyclins (Sherr, 1995). Moreover, cyclin D1 is an oncogene that is overexpressed in many breast cancer cell lines as well as in primary tumors (Buckley et al., 1993). Most interestingly, many anticancer agents, including those used for breast cancer therapy, convey their inhibitory effect in G₁ phase primarily by reducing cyclin D1 levels. These include pure antiestrogens (Watts et al., 1994), tamoxifen (Planas Silva and Weinberg, 1997), retinoids (Teixeira and Pratt, 1997; Zhou et al., 1997), progestins (Musgrove et al., 1998), as well as 1.25-dihydroxyvitamin D₃ (Wang et al., 1996) and TGF- (Mazars et al., 1995).

Besides their effects on cyclin D1, the above anticancer agents have been shown to exert additional effects on the cell cycle apparatus. For example, antiestrogens upregulate the CDK inhibitors p21^Cip1/Waf1 and p27^Kip1 (Watts et al., 1995), whereas retinoic acid does not alter p27 levels but reduces p21 protein abundance (Zhou et al., 1997). The inhibitory effect of retinoids is mediated partly by reduction in cdk2 protein levels (Teixeira and Pratt, 1997), while antiestrogens and tamoxifen do not alter kinase protein abundance (Planas Silva and Weinberg, 1997; Watts et al., 1995). Therefore, the aim of this study was to determine whether lycopene treatment affects the G₁ phase regulatory proteins: pocket proteins (pRb, p107 and p130), cyclins D and E, cyclin-dependent kinases (cdk4, cdk2), and CDK inhibitors, p21 and p27.

The present study is the first to show that the molecular mechanism responsible for the inhibitory effect of lycopene on cell cycle in breast and endometrial cancer cells is based on reduction in cyclin D levels, and retention of p27 in cyclin E-cdk2 complexes resulting in decreased cdk4 and cdk2 activities and inhibition of pRb phosphorylation.

Results

Lycopene inhibits cell cycle progression and cell proliferation

To assess the effect of lycopene on cell cycle progression and DNA synthesis, human breast (MCF-7 and T-47D) and endometrial (ECC-1) cancer cells were partially synchronized in the G₁ phase by serum deprivation (0.5% FCS) for 48 h with or without lycopene. This was followed by serum stimulation (5% FCS), which drove the cells from the G₁ into the S phase. As shown in Figure 1, in serum-deprived control MCF-7, T-47D and ECC-1 cells, the S phase population was 16±2%, 3±0.5% and 4±1%, respectively.

A similar distribution was seen in the lycopene-treated cells. Serum stimulation of MCF-7, T-47D and ECC-1 control cells for 24 h resulted in about two-, four- and threefold increase in the S phase population, respectively, whereas the lycopene-treated cells showed only a minimal increase in S phase. Thus lycopene, although not completely arresting cell cycle progression, delayed G1-S transition in all cell lines tested.

In accordance with our previous data (Levy et al., 1995), the [³H]-thymidine incorporation assay also revealed a significant inhibitory effect of lycopene. In this experiment, serum-starved cells showed a low level of incorporation. Twenty-four hours after serum re-addition, a twofold increase in [³H]-thymidine incorporation was observed in control cells whereas only a marginal change was evident in the lycopene-treated cells (not shown).

Effect of lycopene on the pocket proteins: pRb, p130 and p107

Phosphorylation of pRb and its related pocket proteins, p130 and p107, is a key regulatory event during G₁ phase. Thus, our first step was to determine whether their phosphorylation state was altered by lycopene action. In serum-deprived MCF-7 control cells, both the active hypophosphorylated form and the inactive hyperphosphorylated (slower migrating) forms of both pRb and p130 were present in about equal amounts (Figure 2a). Treatment with lycopene during serum starvation (time 0 h) resulted in a decrease in the pRb protein levels and a clear dominance of the hypophosphorylated forms of pRb and p130 (~80% of total protein). Serum addition to vehicle-treated cells, leading to re-entry of cells into the cell cycle, was accompanied by phosphorylation of pRb and p130 beginning at 6 h, and resulting in progressive accumulation of the hyperphosphorylated forms of these proteins. Pocket protein phosphorylation was markedly reduced and delayed in the lycopene-treated cells, occurring 24 h after serum addition.

The lycopene-induced reduction in pRb protein abundance, although unexplained, has been reported in studies carried out with retinoic acid (Teixeira and Pratt, 1997), antiestrogen ICI182780 (Watts et al., 1995), and progestins (Musgrove et al., 1998) in mammary cancer cells (MCF-7 and T-47D). The amount of the S phase-related pocket protein, p107, also decreased in the lycopene-treated MCF-7 cells despite serum stimulation (Figure 2a).

T-47D and ECC-1 cell lines exhibited similar changes in pRb phosphorylation in response to lycopene treatment (Figure 2b), whereas the protein level did not change. These results suggest that reduction in the pRb amount is a rather isolated effect of the carotenoid in MCF-7 cells.

Effects of lycopene on CDKs and cyclins

The decrease in pRb and p130 phosphorylation in the lycopene-treated cells is probably due to decreased CDK activity, and thus the effect of lycopene on kinase activity of the two predominant G₁ phase CDKs (cdk4 and cdk2) was characterized. Cdk4 activity was determined in cdk4 immunoprecipitates using a GST-pRb fusion protein as a substrate. The activity measured in MCF-7 (Figure 3a), ECC-1 and T-47D cells (data not shown) was very low in serum-starved control cells, and increased 4-6 h after serum re-addition, reaching peak values (fivefold increase) after 12 h. In the lycopene-treated cells, the increase in the cdk4 kinase activity was very low 6 h after serum re-addition and its maximal level was about 40% of the control (Figure 3a).

Cdk2 kinase activity measured in cyclin E immunoprecipitates from MCF-7, T-47D and ECC-1 cells using histone H1 as a substrate was low in serum-starved control cells, appeared 12 h following serum addition and further increased after 24 h by six-, three- and fivefold in these cell lines, respectively (Figure 3b). In the lycopene-treated starved cells, this kinase activity was nearly absent and reached only 30-40% of that measured in the corresponding control cells upon serum stimulation.

Lycopene-induced inhibition of serum-stimulated activity of G₁ cyclin-kinase complexes correlated with the observed changes in pocket protein hypophosphorylation. Since this reduction in CDK activity could have resulted from changes in CDK protein levels, we then measured cdk4 and cdk2 levels in control and lycopene-treated cells. However, as shown for MCF-7 cells in Figure 3c, no such changes were detected.

Inhibition of CDK activity could also result from changes in the amount of G₁ cyclins (cyclins D and E), therefore levels of these proteins were measured. In control MCF-7 cells, serum stimulation resulted in a threefold increase in both cyclin D1 and D3 levels starting after 4 h and reaching a maximal level at 12 h for cyclin D1, and 24 h for cyclin D3 (Figure 4a). Lycopene treatment inhibited cyclin D1 and D3 induction to a level of 40±5% as measured in the control cells.

ECC-1 cells treated with vehicle showed a fivefold increase in cyclin D1 levels in response to serum stimulation, whereas in the lycopene-treated cells, only about a 1.5-fold increase was observed (Figure 4a) reaching about 20% of the maximal level measured in control cells. T-47D cells demonstrated the same trend (Figure 4a). Another G₁ cyclin, cyclin E, displayed a small serum-induced increase in the protein level in both lycopene-treated and control MCF-7 cells, with no significant difference between the two (Figure 4b).

Lycopene effects on CDK inhibitors, p21^CIP1 and p27^KIP1

The inhibitory effect of lycopene on cdk4 kinase activity most likely results from the observed reduction in cyclin D protein levels. However, since no change was observed in cyclin E or cdk2 protein levels we explored the possibility that inhibition of cdk2 activity was a consequence of increase in the levels of CDK inhibitors (CDKIs).

p27 levels were not significantly altered in response to serum stimulation in either control or lycopene-treated MCF-7 cells (Figure 4c). In contrast, an increase in p21 levels was observed in response to serum stimulation in both lycopene-treated and non-treated cells. In control cells, p21 levels reached a maximum at 16 h followed by a decrease 24 h after serum stimulation. Interestingly, although a large serum-induced increment was detected in the lycopene-treated cells, substantially lower p21 protein levels were observed as compared with the control cells at all time points (Figure 4c). A similar trend was found in experiments conducted with retinoic acid, where the p21 protein was induced by serum addition but to a much lesser extent than in the control cells (data not shown).

Lycopene effects on the composition of CDK-cyclin complexes

Since lycopene treatment of MCF-7 cells did not increase the total protein abundance of CDKIs, we next set out to determine whether changes in the level and ratio of inhibitors and activators in the CDK-cyclin complexes could account for the inhibition of cdk2 and cdk4 kinase activity. For this purpose, these complexes were co-immunoprecipitated with antibodies against cyclin E and cyclin D1. Cyclin E complexes immunoprecipitated from control cells exhibited a time-dependent decrease in p27 content beginning at 4 h after stimulation, whereas the lycopene-treated cells retained high levels of p27 and thus exhibited a higher inhibitor to cyclin ratio in the complex (Figure 5). Levels of cdk2 were not altered significantly either in the control or the lycopene treated cells. The p21 content in cyclin E complexes was higher in precipitates from control cells, and showed a biphasic response to serum stimulation similar to that seen in whole cell lysates.

Cyclin D1 complexes immunoprecipitated from the serum-stimulated control cells (Figure 6) showed an initial increase (about twofold) in the p27 content followed by its decline after 12 h (Figure 6b). In contrast, only a slight increase in p27 content was observed in the complexes obtained from the lycopene-treated cells (Figure 6b). Interestingly, the protein levels of p21 in the cyclin D1 complex increased in parallel to cyclin D1 induction in control cells, whereas a much smaller induction was detected in the lycopene-treated cells (Figure 6).

In immunoprecipitates obtained with anti-p27 antibody (Figure 7), the amount of the p27 both in the control and in the lycopene-treated cells did not change significantly during starvation and serum stimulation, similar to the findings in whole cell lysates (Figure 4c). Cyclin D1 levels in the p27 precipitates from the control cells increased in response to serum stimulation, a phenomenon that was markedly inhibited in the lycopene-treated cells. Cyclin E levels in complexes from control cells decreased following serum stimulation, whereas lycopene treatment abolished this response (Figure 7).

Taken together, these data further support our hypothesis that lycopene does not increase CDKI protein levels but rather affects their distribution between cyclin-CDK complexes.

Discussion

We have recently shown that the inhibitory effect of lycopene on leukemic (Amir et al., 1999) and breast (Karas et al., 2000) cancer cells is associated with the inhibition of cell cycle progression through G₁ phase. In this study we reveal the molecular changes underlying this action in breast and endometrial cancer cells.

A major effect on the cell cycle apparatus mediated by lycopene was a reduction in cyclin D1 and D3 protein abundance. Cyclin D1 is a known oncogene (Sherr, 1996), and an important key element in cell cycle progression. It is overexpressed in several cancer cell lines and tumors, especially breast cancer (Buckley et al., 1993; Sweeney et al., 1996). Furthermore, overexpression of cyclin D1 mRNA has been reported to distinguish malignant from benign breast cancer tumors (Weinstat Saslow et al., 1995). Thus reduction in cyclin D1 by lycopene treatment may contribute to its proposed action in prevention of breast and prostate cancer (Giovannucci et al., 1995; Zhang et al., 1997).

The reduction in cyclin D1 levels by lycopene treatment may have two major consequences. The first is a direct effect causing reduction in cyclin D-cdk4 complexes resulting in a decrease in cdk4 kinase activity. The second is probably retention of p27 in cyclin E-cdk2 complexes, an indirect effect that leads to the inhibition of cdk2 activity. It has been suggested that it is not necessarily the total amount of inhibitors, but also their distribution between different complexes and the ratio of inhibitors to activators in each complex that determines CDK kinase activity (Planas Silva and Weinberg, 1997; Sherr and Roberts, 1995). Accordingly, we hypothesized that during starvation, when the level of cyclin D-cdk4 complexes is low, the inhibitor p27 molecules reside in both the cyclin E-cdk2 and cyclin D-cdk4 complexes in quantities disabling kinase activity. Following serum stimulation, as a result of an increase in the abundance of cyclin D-cdk4 complexes, the p27 molecules are titrated by these complexes, thus lower amounts reside in the cyclin E-cdk2 complexes. This probably relieves cdk2 from the inhibitor constraint. We have shown that this distribution of the inhibitor among CDK-cyclin complexes is significantly prevented by lycopene treatment. As a result, cdk2 complexes from the lycopene-treated cells contained larger amounts of p27 and had a higher inhibitor to cyclin ratio resulting in reduction of cdk2 activity.

The end target of CDK activation is phosphorylation of the 'pocket proteins', mainly pRb. In this study, we have shown that phosphorylation of pRb and p130 is inhibited (although not abolished) by lycopene in three different cell lines, MCF-7, T-47D, and ECC-1. In addition the protein level of pRb is reduced in MCF-7 cells but not in the other two cell types suggesting that this reduction is probably confined to MCF-7 cells and is not the general rule in the action of lycopene.

Interestingly, the serum-mediated induction of p21 simultaneous to that of cyclin D1 in control cells, was markedly reduced by lycopene treatment, as found both in total cell lysate and in the cyclin-CDK immunoprecipitates. A similar decrease in p21 abundance during inhibition of the cell cycle has been described previously in retinoids studies (Teixeira and Pratt, 1997; Zhou et al., 1997). Several mechanisms could be responsible for this effect, such as inhibition of p53 pathways, or disruption of growth factor signals known to induce p21 protein when cells re-enter the cell cycle (Liu et al., 1996; Macleod et al., 1995). In addition to these mechanisms, Hiyama et al. (1998) have suggested that induction of cyclin D1 is accompanied by p21 induction mediated by the transcription factor E2F1. Therefore, in lycopene-treated cells, the low levels of p21 may result from inhibition of the growth factor pathway, thereby preventing cyclin D and p21 induction.

Of importance was the finding that the cyclin D1-cdk4 complexes from control cells showed an accumulation of p21 accompanying the increase in kinase activity while the complexes from lycopene-treated cells contained much lower p21 levels. This supports the suggestion that p21 promotes complex assembly and is essential for its kinase activity (Cheng et al., 1999; LaBaer et al., 1997; Zhang et al., 1994). These data suggest the possibility that the low levels of p21 may serve as an additional inhibitory mechanism affecting cdk4 activity in lycopene-treated cells. This hypothesis is currently being studied in our laboratory.

Lycopene both resembles and differs from other anti-cancer agents in regard to the targeted components of the cell cycle machinery. For example, lycopene like tamoxifen, pure antiestrogens, retinoids, progestins, TGF- and TNF- reduces cyclin D protein abundance (Jeoung et al., 1995; Mazars et al., 1995; Planas Silva and Weinberg, 1997; Watts et al., 1995). Whereas most of these agents have been shown to increase levels of the CDK inhibitors, p21 and p27, lycopene, like retinoic acid, reduces p21 levels. However, unlike lycopene and the above agents, retinoic acid also decreases cdk2 protein abundance (Teixeira and Pratt, 1997; Zhou et al., 1997). These differential effects on distinct elements of the cell cycle apparatus could produce a complementary effect on cell cycle progression, that may be the basis for a synergistic anticancer activity of lycopene and retinoic acid. Our preliminary results demonstrate such a synergistic inhibition of cell cycle progression of mammary cancer cells.

In conclusion, the data presented in this study suggest that the inhibitory effect of lycopene on cell cycle progression is mediated primarily through the down regulation of cyclin D, this action directly leads to reduction in cdk4 kinase activity. The decrease in cyclin D levels is probably related to retention of p27 in the cyclin E-cdk2 complex resulting in inhibition of cdk2 kinase activity. The above chain of events is responsible for a decrease in pRb phosphorylation and inhibition of G₁/S transition. The suggested molecular mechanism strengthens and complements the data accumulated from epidemiological and animal studies that demonstrated the anticancer activity of lycopene.

Materials and methods

Lycopene purified from tomato extracts (>97%) was a gift from LycoRed Natural Products Industries (Beer Sheva, Israel). Tetrahydrofuran (THF), containing 0.025% butylated hydroxytoluene as an antioxidant, was purchased from SIGMA-Aldrich (Israel). Dulbecco's modified Eagle's medium (DMEM), fetal calf serum (FCS), bovine insulin and Ca²⁺/Mg²⁺-free phosphate buffered saline (PBS) were purchased from Biological Industries (Beth Haemek, Israel). Lycopene was dissolved in THF at a concentration of 2 mM and stored at -20°C. Immediately before the experiment, the stock solution was added to the cell culture medium, as described previously (Amir et al., 1999). The final concentration of lycopene in the medium was measured by spectrophotometer after extraction in 2-propanol and n-hexane:dichloromethane (Levy et al., 1995). All procedures were performed under dim lighting.

Cell culture and cell proliferation assay

MCF-7, human mammary cancer cell line, was purchased from American Type Culture Collection (Rockwell, MD, USA). T-47D, human mammary cancer cell line, were kindly provided by Dr Yafa Keidar (Tel-Aviv University, Israel). ECC-1 human endometrial cancer cells established and kindly provided by Dr Pondichery Satyaswaroop (Hershey Medical Center, Pennsylvania State University, USA) (Tabibzadeh et al., 1990). Cells were grown in DMEM containing penicillin (100 U/ml), streptomycin (0.1 mg/ml), nystatin (12.5 g/ml), insulin 0.6 g/ml, and 5% FCS. Cells were seeded into 75-cm² flasks (1.5´10⁶ cells) and 24-multiwell plates (35 000 cells per well) in medium containing 5% FCS. Two days later, the medium was changed to one containing 0.5% FCS and solubilized lycopene or THF alone (see above). After 48 h of serum starvation (media was replaced daily), cells were reintroduced to serum containing medium with solubilized lycopene or vehicle. THF had no effect on cell growth or the parameters measured in this study (data not shown). Cells harvested from 75 cm² flasks at different time points were used for Western blotting, immunoprecipitation, kinase assays and flow cytometry assays. In parallel, cells grown in 24-multiwell plates were used for the thymidine incorporation assay. Thymidine incorporation was determined as follows: 1.25 Ci/well of [³H]-thymidine (specific radioactivity 5 mCi/mmol) was added for 1 h. Nucleotide incorporation was stopped by adding unlabeled thymidine (0.5 mol). The cells were then trypsinized and collected on a glass-fiber filter using a cell harvester (Inotech, Switzerland). Radioactivity was determined by a radioactive image analyzer (BAS 1000, Fuji, Japan).

Cell cycle analysis

Cells from 75 cm² flasks were trypsinized, collected and washed twice with PBS. Cell pellets were resuspended in 200 l PBS, fixed in 1 ml of 70% ethanol/30% saline and stored at -20°C. Cells were washed twice with PBS followed by incubation for 40 min in 1 ml PBS containing 0.1% Triton X-100 and 30 g of RNAse (DNAse free) at room temperature. Ten g of propidium iodide were added and the suspension was incubated in the dark at room temperature for an additional 15 min. The suspension was then filtered through a 35 m filter and analysed for DNA content. Flow cytometry studies were performed using the FACSCalibur flow cytometer (Becton Dickinson, Mountain View, CA, USA). The per cent of cells in different phases of the cell cycle was determined using a ModFit (Verity Software House) program.

Western blot analysis

Cells were lysed as described previously (Watts et al., 1995) with modifications. Cell monolayers were washed twice in ice-cold PBS and then scraped into ice-cold lysis buffer A (50 mM HEPES, pH 7.5, 150 mM NaCl, 10% (v/v) glycerol, 1% (v/v) Triton X-100, 1.5 mM MgCl₂, 1 mM EGTA, 10 g/ml aprotinin, 10 g/ml leupeptin, 1 M phenylmethylsulfonyl fluoride, 2 mM sodium orthovanadate, 10 mM sodium pyrophosphate, 50 mM NaF, and 0.2 mM DTT). The lysates were incubated for 10 min on ice, and the cellular debris was cleared by centrifugation (13 000 r.p.m., 5 min, 4°C). Protein content of the samples was determined by the Bradford method using a protein assay kit (Bio-Rad). Equal amounts of protein (30-50 g) were separated by SDS-PAGE and then transferred to a PVDF membrane (Gelman, Inc.).

Proteins were visualized using the ECL detection system (Amersham) after incubation overnight at 4°C with the following primary antibodies: cyclin D1 (HD-11), cyclin D3 (H-292), cyclin E (HE12), p27 (C-19), Cdk2(M2), Cdk4(C-22), p21 (C-19), p130 (C-20), p107 (C-18) from Santa Cruz Biotechnology Inc. (Santa Cruz, CA, USA); or pRB (G3-245) from PharMingen (San Diego, CA, USA). Protein abundance was quantitated by densitometry analysis.

Immunoprecipitation

For cdk4 activity assays, cell monolayers were washed twice with PBS, scraped into ice-cold PBS and pelleted by centrifugation (13 000 r.p.m., 5 min). The pellets were frozen in liquid nitrogen and then resuspended in 1 ml of ice-cold lysis buffer B (50 mM HEPES, pH 7.5, 150 mM NaCl, 1 mM EDTA, 2.5 mM EGTA, 0.1% (v/v) Tween 20, 10% (v/v) glycerol, 10 mM -glycerophosphate, 1 mM NaF, 0.1 mM sodium orthovanadate, 10 g/ml leupeptin, 10 g/ml aprotonin, 0.2 mM DTT, and 0.1 mM phenylmethylsulfonyl fluoride). The lysates were placed on ice for 60 min with intermittent vortexing and then centrifuged at 13 000 r.p.m. for 5 min at 4°C. The supernatants were stored at -70°C. Cdk4 complexes were immunoprecipitated from equivalent amounts of protein with rabbit polyclonal anti-human cdk4 antibody conjugated to protein A-Sepharose (Zymed Laboratories), for 3 h at 4°C. Immunoprecipitates were washed four times with ice-cold lysis buffer B and three times with ice-cold 50 mM HEPES pH 7.5, 1 mM DTT. The supernatants were aspirated, and the immunoprecipitates were used for the kinase assay.

To perform the cdk2 kinase activity assay, lysates were prepared with lysis buffer A as described above for Western blotting. Equivalent amounts of lysates were immunoprecipitated with anti-human cyclin E polyclonal antibody (C-19; Santa Cruz Biotechnology) conjugated to protein A-Sepharose, for 3 h at 4°C. The immunoprecipitates were then washed three times with ice-cold lysis buffer A, and then three times with ice-cold 50 mM HEPES, pH 7.5, 1 mM DTT.

Immunoprecipitation of cyclin D1 and cyclin E was performed as described above for cdk2 kinase activity assay, except that the antibodies were chemically cross-linked with dimethylpimilimidate (Sigma) to protein A-sepharose to reduce the background.

Efficiency of imuunoprecipitation was determined by Western analysis of both the precipitate and supernatant. In all assays, three cycles of immunoprecipitation were performed yielding no immunodetectable quantities of the protein of interest in the supernatant.

Cyclin-dependent kinase assays

The kinase reactions were initiated by resuspending the immunoprecipitates in 30 l of kinase buffer (50 mM HEPES, pH 7.5, 1 mM DTT, 2.5 mM EGTA, 10 mM MgCl₂, 20 M ATP, 10 Ci of [-³²P]ATP, 0.1 mM orthovanadate, 1 mM NaF, 10 mM -glycerophosphate) containing either 2 g of pRb (purified protein 769, Santa Cruz) or 10 g of histone H1 (Boehringer Mannheim) as a substrate. After incubation for either 30 min (cdk4), or 15 min (cdk2) at 30°C, the reactions were terminated by the addition of 15 l of 3´SDS sample buffer (187 mM Tris-HCl, pH 6.8, 30% (v/v) glycerol, 6% SDS, 15% (v/v) -mercaptoethanol). The samples were then heated at 95°C for 2 min and separated using 12% SDS-PAGE. The dried gel was exposed to X-ray film. Relative band intensities were quantitated by radioactive image analysis (BAS 1000, Fuji Photo Film Co., Tokyo, Japan).

Acknowledgements

We thank Prof Robert L Sutherland, Dr Elizabeth A Musgrove, Alex Swarbrick and Angela Lei (Garvan Institute of Medical Research, Sydney) for the many helpful discussions and the technical support. We thank Dr Zohar Nir, LycoRed Natural Products Industries, Beer Sheva, Israel for donating purified lycopene. The studies were supported in part by the Israel Science Foundation founded by the Israel Academy of Science and Humanities; by LycoRed Natural Products Industries, Beer-Sheva, Israel; and by the S Daniel Abraham International Center for Health and Nutrition, Ben-Gurion University of the Negev.

Amir H, Karas M, Giat J, Danilenko M, Levy R, Yermiahu T, Levy J, Sharoni Y. (1999). Nutr. Cancer 33, 105-112. MEDLINE

Buckley MF, Sweeney KJ, Hamilton JA, Sini RL, Manning DL, Nicholson RI, deFazio A, Watts CK, Musgrove EA, Sutherland RL. (1993). Oncogene 8, 2127-2133. MEDLINE

Cheng M, Oliver P, Diehl JA, Fero M, Roussel MF, Roberts JM, Sherr CJ. (1999). EMBO J 18, 1571-1583. Article MEDLINE

Countryman C, Bankson D, Collins S, Man B, Lin W. (1991). Clin. Chem. 37, 1056.

Giovannucci E. (1999). J. Natl. Canc. Inst. 91, 317-331.

Giovannucci E, Ascherio A, Rimm EB, Stampfer MJ, Colditz GA, Willett WC. (1995). J. Natl. Canc. Inst. 87, 1767-1776.

Hankinson SE, Willett WC, Colditz GA, Hunter DJ, Michaud DS, Deroo B, Rosner B, Speizer FE, Pollak M. (1998). Lancet 351, 1397-1399. MEDLINE

Hiyama H, Iavarone A, Reeves SA. (1998). Oncogene 16, 1513-1523. MEDLINE

Jeoung DI, Tang B, Sonenberg M. (1995). J. Biol. Chem. 270, 18367-18373. MEDLINE

Karas M, Amir H, Fishman D, Danilenko M, Segal S, Nahum A, Koifmann A, Giat Y, Levy J, Sharoni Y. (2000). Nutr. Cancer 36, 101-111. MEDLINE

Kim DJ, Takasuka N, Kim JM, Sekine K, Ota T, Asamoto M, Murakoshi M, Nishino H, Nir Z, Tsuda H. (1997). Cancer Lett. 120, 15-22. MEDLINE

Kobayashi T, Itjima K, Mitamura T, Torilzuka K, Cyong J, Nagasawa H. (1996). Anticancer Drugs 7, 195-198. MEDLINE

LaBaer J, Garrett MD, Stevenson LF, Slingerland JM, Sandhu C, Chou HS, Fattaey A, Harlow E. (1997). Genes Dev. 11, 847-862. MEDLINE

Levy J, Bosin E, Feldman B, Giat Y, Miinster A, Danilenko M, Sharoni Y. (1995). Nutr. Cancer 24, 257-267. MEDLINE

Liu Y, Martindale JL, Gorospe M, Holbrook NJ. (1996). Cancer Res 56, 31-35. MEDLINE

Ma J, Pollak MN, Giovannucci E, Chan JM, Tao Y, Hennekens CH, Stampfer MJ. (1999). J. Natl. Cancer Inst. 91, 620-625. MEDLINE

Macleod KF, Sherry N, Hannon G, Beach D, Tokino T, Kinzler K, Vogelstein B, Jacks T. (1995). Genes Dev. 9, 935-944. MEDLINE

Mantzoros CS, Tzonou A, Signorello LB, Stampfer M, Trichopoulos D, Adami HO. (1997). Br J Cancer 76, 1115-1118. MEDLINE

Mazars P, Barboule N, Baldin V, Vidal S, Ducommun B, Valette A. (1995). FEBS Lett 362, 295-300. MEDLINE

Michaud DS, Feskanich D, Rimm EB, Colditz GA, Speizer FE, Willett WC, Giovannucci E. (2000). Am. J. Clin. Nutr. 72, 990-997. MEDLINE

Musgrove EA, Swarbrick A, Lee CS, Cornish AL, Sutherland RL. (1998). Mol. Cell. Biol. 18, 1812-1825. MEDLINE

Nagasawa H, Mitamura T, Sakamoto S, Yamamoto K. (1995). Anticancer Res 15, 1173-1178. MEDLINE

Narisawa T, Fukaura Y, Hasebe M, Ito M, Aizawa R, Murakoshi M, Uemura S, Khachik F, Nishino H. (1996). Cancer Lett 107, 137-142. MEDLINE

Planas Silva MD, Weinberg RA. (1997). Mol. Cell. Biol. 17, 4059-4069. MEDLINE

Sharoni Y, Giron E, Rise M, Levy J. (1997). Cancer Detect Prevent 21, 118-123. MEDLINE

Sherr CJ. (1995). Trends Biochem. Sci. 20, 187-190. MEDLINE

Sherr CJ. (1996). Science 274, 1672-1677. Article MEDLINE

Sherr CJ, Roberts JM. (1995). Genes Dev. 9, 1149-1163. MEDLINE

Sweeney KJ, Musgrove EA, Watts CK, Sutherland RL. (1996). Cancer Treat Res. 83, 141-170. MEDLINE

Tabibzadeh S, Kaffka KL, Kilian PL, Satyaswaroop PG. (1990). In Vitro Cell Dev. Biol. 26, 1173-1179. MEDLINE

Teixeira C, Pratt MAC. (1997). Mol. Endocrinol 11, 1191-1202. MEDLINE

Wang C -J, Lin J -K. (1989). Proc. Natl. Sci. Counc. B ROC 13, 176-183.

Wang CJ, Chou MY, Lin JK. (1989). Cancer Letters 48, 135-142. MEDLINE

Wang QM, Jones JB, Studzinski GP. (1996). Cancer Res. 56, 264-267. MEDLINE

Watts CK, Brady A, Sarcevic B, deFazio A, Musgrove EA, Sutherland RL. (1995). Mol. Endocrinol. 9, 1804-1813. MEDLINE

Watts CK, Sweeney KJ, Warlters A, Musgrove EA, Sutherland RL. (1994). Breast Cancer Res. Treat. 31, 95-105. MEDLINE

Weinstat Saslow D, Merino MJ, Manrow RE, Lawrence JA, Bluth RF, Wittenbel KD, Simpson JF, Page DL, Steeg PS. (1995). Nat. Med. 1, 1257-1260. MEDLINE

Zhang H, Hannon GJ, Beach D. (1994). Genes Dev. 8, 1750-1758. MEDLINE

Zhang SM, Tang GW, Russell RM, Mayzel KA, Stampfer MJ, Willett WC, Hunter DJ. (1997). Am. J. Clin. Nutr. 66, 626-632. MEDLINE

Zhou Q, Stetler Stevenson M, Steeg PS. (1997). Oncogene 15, 107-115. MEDLINE

Figure 1 Lycopene inhibition of cell cycle progression. Exponentially growing MCF-7, T47D and ECC-1 cells were synchronized to G₁ phase by serum deprivation (0.5% FCS) in the presence of lycopene (2-3 M) or 0.5% THF (control) for 48 h and then re-stimulated with serum (5%) contained medium (0 h). Cells were harvested and lysed at the indicated times and a sample of the harvested cells was used for fluorescence activated cell sorter (FACS) analysis of DNA content. (a) Cell cycle histogram of a representative experiment conducted in MCF-7 cell line. (b) Graphical presentation of percentage of cells in S phase in MCF-7, T-47D, and ECC-1 cell lines. Data from three separate experiments are shown as the mean±s.e.m

Figure 2 Lycopene effects on the 'pocket proteins': pRb, p130 and p107. (a) Whole MCF-7 cell lysates from the experiment described in Figure 1 were immunoblotted with antibodies against pRb, p130 and p107 at the indicated times after re-stimulation with serum. A representative blot from one of three experiments is shown. The slower mobility (hyperphosphorylated) and the faster mobility (hypophosphorylated) forms for pRb, and three forms of p130 are indicated. (b) Whole cell lysates from T-47D and ECC-1 cells were immunoblotted with anti-pRb antibody

Figure 3 Lycopene effect on CDK protein level and activity. (a) Cdk4 activity. Whole cell lysates obtained from serum-stimulated MCF-7 cells at the indicated time points were immunoprecipitated with antibody against cdk4. The kinase activity assay was performed with GST-pRb as a substrate. (b) Cyclin E-associated cdk2 activity. Whole cell lysates from MCF-7, T-47D and ECC-1 obtained at the indicated time points following serum stimulation were immunoprecipitated with anti-cyclin E antibody. The kinase activity assay was performed with histone H1 as a substrate. Phosphorylated substrates were detected and visualized by radioactive image analysis. (c) Cdk2 and cdk4 protein level. Whole cell lysates from MCF-7 cells obtained at the indicated time points after serum stimulation (see experiment described in Figure 1) were immunoblotted for cdk4 and cdk2

Figure 4 Lycopene effects on protein levels of G₁ cyclins and CDK inhibitors, p21 and p27. MCF-7, T-47D, and ECC-1 cells were treated as described in Figure 1. Whole cell lysates from lycopene-treated and control cells were analysed by Western blotting for expression of: (a) cyclins D1 (MCF-7, T-47D and ECC-1), cyclin D3 (MCF-7). (b) cyclin E (MCF-7) (c) p21, p27, (MCF-7) as described under Material and methods. A representative blot from one of three experiments is shown

Figure 5 Lycopene effects on composition of cyclin E complexes. MCF-7 cells were treated as described in Figure 1. Whole cell lysates from lycopene treated and control cells were subjected to immunoprecipitation with antibodies against cyclin E. (a) Immunoprecipitates were Western blotted sequentially for cyclin E, cdk2, p21 and p27. (b) Graphical presentation of the cyclin E, cdk2, p27 and p21 protein levels in the representative blot above. Levels of each protein are expressed relative to vehicle treated control cells at time 0 h

Figure 6 Lycopene effects on composition of cyclin D1 complexes. MCF-7 cells were treated as described in Figure 1. Whole cell lysates from lycopene treated and control cells were subjected to immunoprecipitation with antibodies against cyclin D1. (a) Immunoprecipitates were Western blotted sequentially for cyclin D1, p21 and p27. (b) Graphical presentation of the cyclin D1, p27 and p21 protein levels in the representative blot above. Levels of each protein are expressed relative to vehicle treated control cells at time 0 h

Figure 7 Lycopene effects on composition of cyclin E and cyclin D1 complexes. MCF-7 cells were treated as described in Figure 1. Whole cell lysates from lycopene-treated and control cells were subjected to immunoprecipitation with p27 antibody and Western blotted (a) sequentially for cyclin E, cyclin D1, and p27. (b) Graphical presentation of the cyclin D1, cyclin E, and p27 protein levels in the representative blot above. Levels of each protein are expressed relative to vehicle treated control cells at time 0 h