Molecular mechanisms associated with Se-allylselenocysteine regulation of cell proliferation and apoptosis

Cancer Letters 162 (2001) 167±173

www.elsevier.com/locate/canlet

Molecular mechanisms associated with Se-allylselenocysteine regulation of cell proliferation and apoptosis Weiqin Jiang a, Zongjian Zhu a, Howard E. Ganther b, Clement Ip c, Henry J. Thompson a,* a

Center for Nutrition in the Prevention of Disease, AMC Cancer Research Center, Lakewood, CO 80214, USA b Department of Nutritional Sciences, University of Wisconsin, Madison, WI 53706, USA c Department of Experimental Pathology, Roswell Park Cancer Institute, Buffalo, NY 14263, USA Received 1 June 2000; received in revised form 5 October 2000; accepted 10 October 2000

Abstract Se-allylselenocysteine (ASC) has been shown to inhibit mammary carcinogenesis in vivo and cell growth in vitro. However, little is known about the molecular events that account for these effects. The goal of the present study was to use a mouse hyperplastic mammary epithelial cell line, TM12, to investigate the underlying mechanism(s) associated with ASC regulation of cell proliferation and apoptosis. Cells were treated with 50 mM ASC and assessed after 3, 6 and 12 h of exposure. A signi®cant inhibition of cell proliferation, as measured by BrdU incorporation into DNA, was observed within 3 h of ASC treatment. This inhibitory effect was slightly magni®ed at the later time points. The induction of apoptosis was also rapid, and progressed from a 1.3-fold increase at 3 h to a 4.4-fold increase at 12 h. Consistent with these cellular events, the levels of phosphorylated Rb protein were greatly reduced at all times points. The other accompanying changes included increases in P53, P21 and P27. Collectively, the results demonstrate for the ®rst time that ASC is able to cause an immediate response in the expression of cell cycle regulatory proteins that favor an arrest in proliferation and an augmentation in apoptosis. q 2001 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Selenium; Alkylselenocysteine; Apoptosis; Cell growth

1. Introduction Many selenium compounds have been shown to inhibit and/or retard cell growth in vitro and tumorigenesis in a variety of experimental animal models in vivo, and a number of mechanisms have been implicated in accounting for these effects [1±5]. Of the selenium compounds studied to date, one that shows particular promise is Se-allylselenocysteine (ASC). ASC is an amino acid prodrug that undergoes clea* Corresponding author. AMC Cancer Research Center, 1600 Pierce Street, Denver, CO 80214, USA. Tel.: 11-303-239-3463; fax: 11-303-239-3443. E-mail address: [email protected] (H.J. Thompson).

vage at the beta position by cysteine conjugate beta lyase and related enzymes [6]. This compound was developed as an extension of our earlier work [7]; our goal was to determine if improved cancer inhibitory activity could be achieved by varying the aliphatic side chain in ASC derivatives. In the initial report of its effect, ASC at 2 ppm Se in the diet, caused an 86% inhibition in a chemically induced mammary tumor model in rats [8]. This magnitude of cancer protection was signi®cantly greater than that observed with other selenium compounds at the same dose level. Studies of the biological effects of ASC are limited, although we have recently reported that ASC inhibits cell growth in vitro and induces apopto-

0304-3835/01/$ - see front matter q 2001 Elsevier Science Ireland Ltd. All rights reserved. PII: S 0304-383 5(00)00647-9

168

W. Jiang et al. / Cancer Letters 162 (2001) 167±173

sis in mammary epithelial cell lines that have either wild-type or mutant P53 activity [9]. It is the potential of ASC as a new chemopreventive agent, and the lack of understanding of its biological effects, that provided the impetus for the work described here. The goal of these experiments was to investigate the molecular responses to ASC that regulate the processes of cell proliferation and apoptosis. To do this, we elected to study the effects of ASC in a hyperplastic mammary epithelial cell line (TM12) since selenium has been reported to inhibit the progression of mammary hyperplasias in vivo [10]. This cell line has been shown to be responsive to selenium and has a wild type P53 gene, which we considered important given our interest in factors affecting both cell cycle progression and apoptosis. Based on our previous work [11,12] as well as that of Medina and co-workers [13±15] using other selenium compounds, we hypothesized that ASC would exert an acute effect in vitro by reducing the rate of cell proliferation and increasing the rate of apoptosis. This hypothesis was tested by examining BrdU incorporation into DNA and apoptotic DNA fragmentation at 3, 6 and 12 h following exposure to ASC. In additional companion studies, the time-dependent effects of ASC on the expression of proteins involved in cell cycle regulation and apoptosis were then investigated.

2. Materials and methods 2.1. Chemicals The following materials were purchased from commercial sources: Dulbecco's modi®ed Eagle's medium and F-12 medium (Sigma, St. Louis, MO), adult bovine serum (Gemini Bioproducts, Calabasas, CA), insulin and epidermal growth factor (Intergen, Purchase, NY), gentamicin reagent solution (Gibco BRL, Grand Island, NY), Cell Proliferation Assay Kit (Oncogene Research Products, Cambridge, MA), Triton X-100 (Sigma, St. Louis, MO), OliGreen (Molecular Probes, Eugene, OR), agarose (Gibco BRL, Grand Island, NY). Anti-P53 and anti-Cip/P21 antibodies; rabbit anti-mouse immunoglobulin- and goat anti-rabbit immunoglobulin-horseradish peroxidase-conjugated secondary antibodies, and antimouse beta-actin, (used as a lane-loading control),

were purchased from Santa Cruz Corp. (Santa Cruz, CA). Anti-Kip1/P27, anti-cyclin D1 and anti-CDK4 antibodies were from Neomarkers, Inc. (Fremont, CA). Anti-Rb antibody was from PharMingen/Transduction Laboratories (San Diego, CA). The ECL detection system was from Amersham Corp. (Arlington Heights, IL). The d,l form of ASC was synthesized from d,l-selenocystine as described previously [8]. 2.2. Cell culture The mouse mammary hyperplastic epithelial cell line TM12 was obtained from the laboratory of Daniel Medina [10,16]. Cells were grown at 378C in a humidi®ed incubator containing 5% CO2 in Dulbecco's modi®ed Eagle's medium and F-12 medium (1:1 DMEM/F-12) containing 2% adult bovine serum, 10 mg/ml insulin, 5 ng/ml epidermal growth factor, and 5 mg/ml gentamicin. Selenium content of complete medium was 6 £ 10 28 M which is considered adequate for supporting the growth of cells in culture [17]. 2.3. Cell proliferation analysis by BrdU labeling Brie¯y, 96-well plates were seeded at uniform density; 24 h after plating, cells were exposed to ASC at 50 mM of Se for 3, 6 and 12 h. Cells were labeled with BrdU during the last 3-h of exposure. The cells were ®xed and incorporation of BrdU was detected by immunoreaction using mouse anti-BrdU antibody and goat anti-mouse IgG horseradish peroxidase conjugate. After substrate solution was added to each well, the amount of BrdU incorporated was determined by measuring absorbance at dual wavelengths 450±540 nm using a spectrophotometric Thermomax Microplate Reader (Molecular Devices, Sunyvale, CA). 2.4. Apoptosis counting TM12 cells were cultured in 12-well plates and exposed to ASC at 50 mM of Se for 3, 6 and 12 h. Cells undergoing apoptosis in this system generally detach from the culture dish. Therefore, apoptosis was assessed by harvesting all cells, both ¯oating and attached. Apoptosis was determined morphologically by ¯uorescent microscopy after labeling with

W. Jiang et al. / Cancer Letters 162 (2001) 167±173

acridine orange and ethidium bromide as described by Duke and Cohen [18].

169

densities yielded similar protein signals (data not shown). 2.6. Statistical Analyses

2.5. Expression of cell cycle regulatory molecules by western blotting In assessing the expression of cell cycle regulatory molecules, only attached cells were studied. TM12 cells were cultured in DMEM/F12 medium with 5 ng/ml EGF, 10 mg/ml insulin, 2% adult bovine serum and 5 mg/ml gentamycin under standard culture conditions. Cultures reaching 70±80% con¯uency were treated with 50 mM ASC for 3, 6 and 12 h. The medium was aspirated at the end of these treatments, and the monolayer of cells was quickly washed two times with cold PBS. A 0.3 ml aliquot of lysis buffer (10 mM Tris±HCl, pH 7.4, 150 mM NaCl, 1% Triton X-100, 1 mM EDTA, 1 mM EGTA, 0.2 mM sodium vanadate, 0.2 mM phenylmethylsulfonyl ¯uoride, 0.5% NP-40, and 0.2 U/ml aprotonin) was then added per plate. After bathing in lysis buffer for 15 min on ice, the cells were scraped from the plate, the mixture of buffer and cells was transferred to microfuge tubes and left in ice for an additional 15 min. The lysate was collected by centrifugation for 15 min in a table top centrifuge at 48C, and protein concentration in the clear supernant was determined by the Bio-Rad protein assay (Hercules, CA). For Western blotting of cell cycle regulatory molecules, 40±100 mg of protein lysate per sample was denatured with SDS-PAGE sample buffer (63 mM Tris±HCl, pH 6.8, 10% glycerol, 2% SDS, 0.0025% bromophenol blue, and 5% 2-mercaptoenthanal), subjected to SDS-PAGE on 8 or 12% gel, and the protein bands blotted onto a membrane. Total protein levels of P53, Cip1/P21, Kip1/P27, cyclin D1, CDK4, and Rb were determined using speci®c primary antibody, followed by treatment with the appropriate peroxidase-conjugated secondary antibody and visualized by the ECL detection system. Signals were quantitated by scanning the ®lm with ScanJet (Hewlett Packard, Palo Alto, CA), and the intensity of the bands was analyzed using the `Image-Pro Plus' software (Media Cybernetics, Silver Spring, MD). To ensure that control cells (not exposed to ASC) gave reproducible baseline protein levels, we carefully maintained cell cultures in a semi-con¯uent and logarithmically growing state. Multiple controls at varying

Differences in cell proliferation and apoptotic cell death in response to ASC treatment were evaluated by analysis of variance (ANOVA) [19]. Post hoc comparisons among treatment conditions were made using the Bonferroni multiple-range test [19]. 3. Results Our previous work showed that ASC reduced, in a dose (25±200 mM) and time (24±72 h) dependent manner, the number of adherent cells in culture [9]. However, we did not determine the relative contribution of cell proliferation versus apoptosis to this outcome. Moreover, we did not assess the acute effect of ASC at time points earlier than 24 h. To investigate these questions more thoroughly, TM12 cells were exposed to only one concentration of ASC at 50 mM and the responses were examined after 3, 6 or 12 h. This dose was chosen because it did not induce necrosis, and its effect on cell number was gradual and progressive. The effects of ASC on cell proliferation and apoptotic cell death are shown in Table 1. During the time frame of the experiment, the semi-con¯uent, logarithmically growing cultures of both untreated and ASC-treated cells incorporated BrdU; however, Table 1 Effects of ASC on cell proliferation and apoptotic cell death a Cell proliferation

Apoptotic cell death

Absorbance

%

Control 3h 6h 12 h a

0.16 ^ 0.01 0.25 ^ 0.01 0.47 ^ 0.03

ASC a a a

0.12 ^ 0.01 0.14 ^ 0.01 0.29 ^ 0.01

Control b b b

6.2 ^ 0.5 5.3 ^ 0.5 5.7 ^ 0.3

ASC a a a

7.8 ^ 0.3 11.7 ^ 0.6 24.8 ^ 2.0

b b b

TM12 cells were exposed to 50 mM ASC. All experiments were repeated three times. In each experiment, eight replicates for cell proliferation or six replicates for apoptotic cell death at each time point were analyzed. The results of a representative experiment are presented. Data are expressed as mean ^ SE. The data were analyzed by analysis of variance between control and treated cultures at each time point. Values with different superscripts are statistically different, P , 0:05.

170

W. Jiang et al. / Cancer Letters 162 (2001) 167±173

exposure of TM12 cells to 50 mM ASC signi®cantly suppressed the rate of BrdU incorporation, and by implication the rate of cell proliferation, as early as 3 h of treatment (77% of untreated control, P , 0:05). The inhibitory effect appeared to plateau by 6 h (58% of untreated control, P , 0:01), and remained stable at this diminished level after 12 h (62% of untreated control, P , 0:01). ASC treatment also induced an elevation of apoptotic cell death at all three time points. In untreated cells, the basal level of apoptosis was approximately 6%. The proportion of apoptotic cells increased slightly by 1.3-fold (P , 0:05) within 3 h, and continued to escalate by 2.2-fold (P , 0:01) and 4.4-fold (P , 0:01) after 6 and 12 h of ASC exposure, respectively. Thus, it appears that ASC decreased cell proliferation as well as increased apoptotsis; however, the magnitude of the effect on apop-

tosis was comparatively greater than the effect on proliferation. We next evaluated the expression of the phosphorylated Rb protein since we had reason to believe that ASC might arrest cell cycle progression at the G1-S checkpoint based on data from other related selenium compounds [13±15]. A signi®cant reduction in the amount of phosphorylated Rb protein was seen at all time points in ASC treated cultures, although the decrease was especially marked in the 6- and 12-h cultures (Fig. 1 and Table 2). The decrease in phosphorylated Rb protein was consistent with the reduced levels of BrdU incorporation reported in Table 1. Since a loss of DNA integrity on exposure of cells to ASC for longer periods of time has been reported, we next investigated the effects of ASC on cellular levels of P53 and the CKI that it regulates, P21. As shown in Fig. 1 and Table 2, the decrease in Rb phosphorylation status was accompanied by an increase in wild-type P53, which began at 3 h (1.5-fold) and plateaued between 6 and 12 h (2.4-fold). Similar to P53, the increase in P21 also peaked at 12 h (3.8-fold). We note that caution is warranted in attributing biological signi®cance to changes in protein levels of less than 2-fold as determined by Western analysis; however, the fact that the levels of both proteins increased, and that the increase in P53 preceded the increase in P21, strengthen the likelihood that the detected changes re¯ect cellular responses to ASC. Cellular levels of another CKI, P27, are thought to re¯ect whether cellular conditions are favorable for proliferation [20], and levels of this protein have been reported to be elevated in vivo by other selenium Table 2 Relative changes in cell cycle regulatory protein expression in TM12 cells treated with ASC a

Fig. 1. Digital images of Western blot analyses of cell cycle regulatory proteins in untreated TM12 cells or cells treated with 50 mM Se-allylselenocysteine (ASC) for 3, 6, or 12 h. Beta actin was used as an indicator for control of lane loading. These data are representative of three separate determinations. See Table 2 for quantitative determinations of protein level changes. ppRb, phosphorylated retinoblastoma.

PpRb CDK4 Cyclin D1 P53 P21 P27

3h

6h

12 h

0.13 ^ 0.02 0.29 ^ 0.04 0.93 ^ 0.14 1.53 ^ 0.33 0.88 ^ 0.12 5.26 ^ 0.55

0.04 ^ 0.01 0.19 ^ 0.04 1.03 ^ 0.14 2.30 ^ 0.83 1.90 ^ 0.09 4.91 ^ 0.81

0.05 ^ 0.01 0.52 ^ 0.01 1.70 ^ 0.18 2.43 ^ 0.40 3.83 ^ 0.12 2.77 ^ 0.71

a Cells were exposed to 50 mM ASC. The results are expressed as fold change relative to the basal value from untreated control. Data from three separate experiments were quanti®ed and are presented as mean ^ SE. PpRb, phosphorylated retinoblastoma.

W. Jiang et al. / Cancer Letters 162 (2001) 167±173

compounds [5]. Therefore, levels of P27 were evaluated and found to be signi®cantly elevated at all time points. This observation was consistent with the reduced rates of cell proliferation and decreased levels of Rb phosphorylation reported in Tables 1 and 2. Because the effects of ASC on levels of phosphorylated Rb, and CKIs P21 and P27 were rapid, and given that changes in these CKIs have been shown to in¯uence levels of cyclins and their cyclin dependent kinase partners, we investigated the effect of ASC on total protein levels of cyclin D1 and CDK4. The effects of ASC on cyclin D1 were modest and did not attain a magnitude to which we attribute biological signi®cance based on Western blot analysis (Fig. 1 and Table 2). ASC exposure resulted in lower levels of CDK4; a reduction of 71, 81 and 48% at 3, 6 and 12 h of exposure (Fig. 1 and Table 2) was observed. Similarly, total protein levels of cyclin E were unaffected in ASC exposed cells; whereas, levels of CDK2 were reduced (data not shown). 4. Discussion While the anticancer activity of selenium has been recognized for a long time, there is limited knowledge of the molecular mechanism(s) underlying this effect. Over the past decade, cell culture studies have provided important clues about speci®c molecular pathways responsive to selenium. The collective evidence from a number of laboratories implicates selenium as a modulator of the cell cycle, leading to a block in cell proliferation and/or an increase in apoptosis. ASC is the newest in a series of Se-alkyl selenoamino acids that have been reported to have chemopreventive activity [2,5,8]. As reported in a recent paper, the potency of this compound appears to surpass that of other structurally related forms of selenium [8]. This attribute prompted us to evaluate more fully its cellular and molecular effects. As shown in Table 1, treatment with ASC resulted in a rapid inhibition of cell proliferation and a progressive increase in apoptosis between 3 and 12 h of culture. Previously, we had observed a reduction in cell number by ASC, but we did not know if this was solely due to the induction of apoptosis, or if inhibition of cell proliferation was

171

also involved [9]. The present data clarify this point. The ®nding that apoptosis occurs early upon exposure to ASC is provocative, because it may give further clues about the mechanisms that underlie death induction. In particular, the rapidity and magnitude of the apoptotic response indicate the feasibility of determining: (1) whether an intracellular or extracellular death signaling pathway is activated by ASC, and (2) whether initiator caspases 2, 8 or 9 mediate death induction. We speculate that caspase 9 will be induced by ASC; if this is correct, then it will be critical to determine if procaspase activation involves a perturbation of a cell survival signaling pathway or a signaling pathway that is involved in mediating cellular responses to stress. Given that previous reports show that various forms of selenium block the entry of cells into the S phase of the cell cycle [13±15], we postulated that ASC would down regulate the phosphorylation of the Rb protein. This hypothesis was based on substantial evidence of the importance of Rb phosphorylation in cell cycle progression, both in the initiation of G1 and in maintaining the cellular machinery necessary for progression through S phase [21,22]. As shown in Table 2 and Fig. 1, treatment with ASC induced a rapid and profound reduction in the amount of phosphorylated Rb protein as determined by Western blot analysis. To our knowledge, this is the ®rst report that phosphorylated Rb is suppressed by a selenium compound. To address the question of why cells treated with ASC had low levels of phosphorylated Rb, we considered the fact that many forms of selenium have been reported to induce cellular stress, and that ASC can induce a modest reduction in DNA integrity following longer periods of exposure in this cell line [9]. This type of cellular response is associated with the activation of pathways leading to cell cycle arrest as well as apoptosis. Prominent among the pathways that mediate cellular responses to stress is that regulated by the P53 gene [23±25]. As shown in Table 2, a modest, time dependent increase in cellular levels of P53 was observed in response to ASC. Given that the magnitude of the increase in P53 protein was borderline relative to what we consider biologically signi®cant based on Western blot analysis, we also examined cellular levels of P21, a member of Cip/Kip family of CKIs, that increased levels of P53 protein is known to upregulate [25]. In parallel with

172

W. Jiang et al. / Cancer Letters 162 (2001) 167±173

the pattern of P53 expression, levels of P21 increased signi®cantly in a time dependent manner, consistent with a cause and effect relationship between P53 and P21. The changes in these regulatory molecules also are consistent with the time-frame of the changes in BrdU incorporation and apoptosis (Table 1). These data imply that inhibition of cell proliferation may precede apoptosis, and this possibility is further supported by the magnitude and timeframe of the decrease in Rb phosphorylation relative to the increases in levels of P53 protein. Nonetheless, clari®cation of the relationships among growth arrest, induction of apoptosis, and regulation of P53 will require further investigation. Given that the levels of one member of the Cip/Kip family of CKIs were elevated in response to ASC, we examined the levels of a second member of this family, P27, which selenium has been reported to modulate and that has not been reported to be associated with cellular responses to stress [20]. Levels of P27 protein are generally low in proliferating cells, but rise in response to antimitogenic conditions [5]. Depending on the amount of P27 present, this CKI has been reported to enhance or inhibit the activity of cyclin D-CDK4 complexes and to consistently inhibit the activity of cyclin E-CDK2 complexes; inhibition of both CDKs would result in low cellular levels of phosphorylated Rb protein [21,26]. We observed signi®cant elevations in cellular levels of P27 at all time points, a ®nding consistent with both the reduced levels of BrdU incorporation and low levels of phosphorylated Rb. Based on these ®ndings, we predicted that levels of cyclin D1 would be reduced whereas levels of CDK4 would be unaffected. The opposite pattern of change was observed, i.e. levels of cyclin D1 varied less than 2-fold, a magnitude of change that we do not consider biologically signi®cant; whereas, a signi®cant reduction in CDK4 protein was noted. When levels of cyclin E and CDK2 were measured, a similar pattern of expression was observed. We currently have no explanation for these patterns of change. However, we judge that future studies need to be carried out using synchronized cells in order to improve the speci®city and sensitivity of these molecular analyses. In conclusion, our data indicate that the effects of ASC on cell proliferation and apoptosis are rapid and that multiple regulatory pathways are affected by

treatment with this compound. Upstream events involved in the phosphorylation of Rb are potential targets of ASC. We speculate that members of the Cip/Kip family of cyclin dependent kinase inhibitors are likely to play a signi®cant role. Whether the factors inducing increased cellular levels of these proteins are associated with speci®c effects of ASC on signal transduction pathways or are mediated by mechanisms associated with cellular stress remains to be determined. Another issue that requires investigation is why such high levels of ASC, 50 mM, are needed to achieve in vitro effects while much lower levels of ASC are required in vivo. We speculate that this may be a consequence of the level of activity of beta lyase and related enzymes in this cell line. This topic is the subject of ongoing studies. Acknowledgements This work was supported by grant No. CA45164 from the National Cancer Institute. References [1] H.J. Thompson, A. Wilson, J. Lu, M. Singh, C. Jiang, P. Upadhyaya, K.E.L. Bayoumy, C. Ip, Comparison of the effects of an organic and an inorganic form of selenium on a mammary carcinoma cell line, Carcinogenesis 15 (1994) 183± 186. [2] H.E. Ganther, Selenium metabolism, selenoproteins and mechanisms of cancer prevention: complexities with thioredoxin reductase, Carcinogenesis 20 (1999) 1657±1666. [3] C. Ip, Lessons from basic research in selenium and cancer prevention, J. Nutr. 128 (1998) 1845±1854. [4] D. Medina, D.G. Morrison, Current ideas on selenium as a chemopreventive agent, Pathol. Immunopathol. Res. 7 (1988) 187±199. [5] C. Ip, H.J. Thompson, H.E. Ganther, Selenium modulation of cell proliferation and cell cycle biomarkers in normal and premalignant cells of the rat mammary gland, Cancer Epidemiol. Biomarkers Prev. 9 (2000) 49±54. [6] I. Andreadou, W.M. Menge, J.N. Commandeur, E.A. Worthington, N.P. Vermeulen, Synthesis of novel Se-substituted selenocysteine derivatives as potential kidney selective prodrugs of biologically active selenol compounds: evaluation of kinetics of beta-elimination reactions in rat renal cytosol, J. Med. Chem. 39 (1996) 2040±2046. [7] C. Ip, H.E. Ganther, Comparison of selenium and sulfur analogs in cancer prevention, Carcinogenesis 13 (1992) 1167±1170. [8] C. Ip, Z. Zhu, H.J. Thompson, D. Lisk, H.E. Ganther, Chemo-

W. Jiang et al. / Cancer Letters 162 (2001) 167±173

[9]

[10]

[11]

[12]

[13]

[14] [15]

prevention of mammary cancer with Se-allylselenocysteine and other selenoamino acids in the rat, Anticancer Res. 19 (1999) 2875±2880. Z. Zhu, W. Jiang, H.E. Ganther, C. Ip, H.J. Thompson, In vitro effects of Se-allylselnocysteine and Se-propylselenocysteine on cell growth. DNA integrity and apoptosis, Biochem. Pharmacol. 60 (2000) 1467±1473. D. Medina, L.C. Stephens, P.J. Bonilla, C.A. Hollmann, D. Schwahn, C. Kuperwasser, D.J. Jerry, J.S. Butel, R.E. Meyn, Radiation-induced tumorigenesis in preneoplastic mouse mammary glands in vivo: signi®cance of P53 status and apoptosis, Mol. Carcinog. 22 (1998) 199±207. J. Lu, C. Jiang, M. Kaeck, H. Ganther, S. Vadhanavikit, C. Ip, H. Thompson, Dissociation of the genotoxic and growth inhibitory effects of selenium, Biochem. Pharmacol. 50 (1995) 213±219. J. Lu, M. Kaeck, C. Jiang, A.C. Wilson, H.J. Thompson, Selenite induction of DNA strand breaks and apoptosis in mouse leukemic L1210 cells, Biochem. Pharmacol. 47 (1994) 1531± 1535. R. Sinha, T.K. Said, D. Medina, Organic and inorganic selenium compounds inhibit mouse mammary cell growth in vitro by different cellular pathways, Cancer Lett. 107 (1996) 277± 284. R. Sinha, D. Medina, Inhibition of cdk2 kinase activity by methylselenocysteine in synchronized mouse mammary epithelial tumor cells, Carcinogenesis 18 (1997) 1541±1547. R. Sinha, S.C. Kiley, J.X. Lu, H.J. Thompson, R. Moraes, S. Jaken, D. Medina, Effects of methylselenocysteine on PKC activity, cdk2 phosphorylation and gadd gene expression in synchronized mouse mammary epithelial tumor cells, Cancer Lett. 146 (1999) 135±145.

173

[16] D. Medina, F.S. Kittrell, Y.J. Liu, M. Schwartz, Morphological and functional properties of TM preneoplastic mammary outgrowths, Cancer Res. 53 (1993) 663±667. [17] D. Medina, C.J. Oborn, Differential effects of selenium on the growth of mouse mammary cells in vitro, Cancer Lett. 13 (1981) 333±344. [18] R.C. Duke, J.J. Cohen, Morphological and biochemical assays of apoptosis, in: J.E. Coligan, A.M. Kruisbreak (Eds.), Current Protocols in Immunology, John Wiley & Sons, New York, 1992, pp. 3.17.1±3.17.16. [19] G.J.W. Snedecor, W.G. Cochran, Statistical Methods, Iowa State University Press, Ames, IA, 1967, 135±317. [20] A. Sgambato, A. Cittadini, B. Faraglia, I.B. Weinstein, Multiple functions of P27(Kip1) and its alterations in tumor cells: a review, J. Cell Physiol. 183 (2000) 18±27. [21] C.J. Sherr, The Pezcoller lecture: cancer cell cycles revisited, Cancer Res. 60 (2000) 3689±3695. [22] C.J. Sherr, J.M. Roberts, C.D.K. inhibitors:, positive and negative regulators of G1-phase progression, Genes Dev. 13 (1999) 1501±1512. [23] N.D. Lakin, S.P. Jackson, Regulation of P53 in response to DNA damage, Oncogene 18 (1999) 7644±7655. [24] L.D. Attardi, T. Jacks, The role of P53 in tumour suppression: lessons from mouse models, Cell. Mol. Life Sci. 55 (1999) 48± 63. [25] Y. Xu, E.M. Yang, J. Brugarolas, T. Jacks, D. Baltimore, Involvement of P53 and P21 in cellular defects and tumorigenesis in Atm- /- mice, Mol. Cell Biol. 18 (1998) 4385±4390. [26] X. Zhang, W. Wharton, M. Donovan, D. Coppola, R. Croxton, W.D. Cress, W.J. Pledger, Density-dependent growth inhibition of ®broblasts ectopically expressing P27(kip1), Mol. Biol. Cell 11 (2000) 2117±2130.