Cell Cycle and Biochemical Effects of PD 0183812
2001; Elsevier BV; Volume: 276; Issue: 20 Linguagem: Inglês
10.1074/jbc.m008867200
ISSN1083-351X
AutoresDavid W. Fry, David C. Bedford, Patricia H. Harvey, Alexandra Fritsch, Paul R. Keller, Zhipei Wu, Ellen M. Dobrusin, Wilbur R. Leopold, Ali Fattaey, Michelle D. Garrett,
Tópico(s)Advanced Breast Cancer Therapies
ResumoProgression through the G1 phase of the cell cycle requires phosphorylation of the retinoblastoma gene product (pRb) by the cyclin D-dependent kinases CDK4 and CDK6, whose activity can specifically be blocked by the CDK inhibitor p16INK4A. Misregulation of the pRb/cyclin D/p16INK4A pathway is one of the most common events in human cancer and has lead to the suggestion that inhibition of cyclin D-dependent kinase activity may have therapeutic value as an anticancer treatment. Through screening of a chemical library, we initially identified the [2,3-d]pyridopyrimidines as inhibitors of CDK4. Chemical modification resulted in the identification of PD 0183812 as a potent and highly selective inhibitor of both CDK4 and CDK6 kinase activity, which is competitive with ATP. Flow cytometry experiments showed that of the cell lines tested, only those expressing pRb demonstrated a G1 arrest when treated with PD 0183812. This arrest correlated in terms of incubation time and potency with a loss of pRb phosphorylation and a block in proliferation, which was reversible. These results suggest a potential use of this chemical class of compounds as therapeutic agents in the treatment of tumors with functional pRb, possessing cell cycle aberrations in other members of the pRb/cyclin D/p16INK4A pathway. Progression through the G1 phase of the cell cycle requires phosphorylation of the retinoblastoma gene product (pRb) by the cyclin D-dependent kinases CDK4 and CDK6, whose activity can specifically be blocked by the CDK inhibitor p16INK4A. Misregulation of the pRb/cyclin D/p16INK4A pathway is one of the most common events in human cancer and has lead to the suggestion that inhibition of cyclin D-dependent kinase activity may have therapeutic value as an anticancer treatment. Through screening of a chemical library, we initially identified the [2,3-d]pyridopyrimidines as inhibitors of CDK4. Chemical modification resulted in the identification of PD 0183812 as a potent and highly selective inhibitor of both CDK4 and CDK6 kinase activity, which is competitive with ATP. Flow cytometry experiments showed that of the cell lines tested, only those expressing pRb demonstrated a G1 arrest when treated with PD 0183812. This arrest correlated in terms of incubation time and potency with a loss of pRb phosphorylation and a block in proliferation, which was reversible. These results suggest a potential use of this chemical class of compounds as therapeutic agents in the treatment of tumors with functional pRb, possessing cell cycle aberrations in other members of the pRb/cyclin D/p16INK4A pathway. cyclin-dependent kinase phosphate-buffered saline fetal bovine serum platelet-derived growth factor receptor fibroblast growth factor receptor epidermal growth factor receptor p60src protein kinase C retinoblastoma gene product fluorescence-activated cell sorter The cyclin-dependent kinase (CDK)1 family of serine/threonine protein kinases are the cornerstone of the cell division cycle. Full CDK activity requires association with a cyclin regulatory subunit, and it is the distinct combinations of CDK-cyclin complexes that control progression through the different phases of the cell cycle (1Morgan D.O. Annu. Rev. Cell Dev. Biol. 1997; 13: 261-291Crossref PubMed Scopus (1769) Google Scholar). During early to mid G1, when the cell is responsive to mitogenic stimuli, it is CDK4 and CDK6 in association with the D-type cyclins that control advancement through the cell cycle, by their phosphorylation of the tumor suppressor protein, pRb (2Kato J. Front. Biosci. 1999; 4: 787-792Crossref PubMed Google Scholar). The cyclin D-dependent kinases also act as integrators of extracellular stimuli. For example, induction of cyclin D1 transcription can be activated by the ras signaling pathway (3Albanese C. Johnson J. Watanabe G. Eklund N. Vu D. Arnold A. Pestell R.G. J. Biol. Chem. 1995; 270: 23589-23597Abstract Full Text Full Text PDF PubMed Scopus (759) Google Scholar, 4Gille H. Downward J. J. Biol. Chem. 1999; 274: 22033-22040Abstract Full Text Full Text PDF PubMed Scopus (369) Google Scholar), whereas degradation of cyclin D1 protein is stimulated by the serine/threonine kinase GSK-3β, which is itself regulated by the phosphatidylinositol 3-kinase and Wnt signaling pathways (5Diehl J.A. Cheng M. Roussel M.F. Sherr C.J. Genes Dev. 1998; 12: 3499-3511Crossref PubMed Scopus (1837) Google Scholar, 6Rimerman R.A. Gellert-Randleman A. Diehl J.A. J. Biol. Chem. 2000; 275: 14736-14742Abstract Full Text Full Text PDF PubMed Scopus (66) Google Scholar). The importance of the cyclin D-dependent kinases and pRb in controlling transit through G1 is further reflected by the fact this pathway is a major target of genetic alteration in cancer. This is exemplified by the RB gene itself, which is found mutated in a number of tumor types, and cyclin D1, which is overexpressed in a variety of tumors (7Sellers W.R. Kaelin Jr., W.G. J. Clin. Oncol. 1997; 15: 3301-3312Crossref PubMed Scopus (231) Google Scholar, 8Donnellan R. Chetty R. Mol. Pathol. 1998; 51: 1-7Crossref PubMed Scopus (303) Google Scholar). Another major player in this pathway is the tumor suppressor gene INK4A/CDKN2Aencoding the p16INK4A protein, which specifically inhibits the kinase activity of the two cyclin D-dependent kinases CDK4 and CDK6 and is found genetically altered in multiple tumor types (9Sharpless N.E. DePinho R. Curr. Opin. Genet. Dev. 1999; 9: 22-30Crossref PubMed Scopus (448) Google Scholar). In a number of cases, the pattern of RB mutation appears to inversely correlate with mutations ofINK4A/CDKN2A, suggesting that a single defect of the pRb/cyclin D/p16INK4A pathway is enough for tumor development (10Hall M. Peters G. Adv. Cancer Res. 1996; 68: 67-108Crossref PubMed Google Scholar). The intensity with which this pathway appears to be misregulated in cancer has lead to the suggestion that in situations where cyclin D-dependent kinase activity may be up-regulated such as by mutation of INK4A/CDKN2A or overexpression of cyclin D1, inhibiting the kinase activity of both CDK4 and CDK6 may be an effective anticancer treatment. Evidence for the validity of such an approach has been provided in a number of ways. Reintroduction of INK4A/CDKN2A into tumor cells where this gene is defective will inhibit cell growth in culture and tumor growth in xenograft studies, whereas a peptide mimic of p16INK4A has demonstrated antiproliferative properties in a number of cell lines (11Jin X. Nguyen D. Zhang W.W. Kyritsis A.P. Roth J.A. Cancer Res. 1995; 55: 3250-3253PubMed Google Scholar, 12Rocco J.W. Li D. Liggett Jr., W.H. Duan L. Saunders Jr., J.K Sidransky D. O' Malley Jr., B.W. Clin. Cancer Res. 1998; 4: 1697-1704PubMed Google Scholar, 13Craig C. Kim M. Ohri E. Wersto R. Katayose D. Li Z. Choi Y.H. Mudahar B. Srivastava S. Seth P. Cowan K. Oncogene. 1998; 16: 265-272Crossref PubMed Scopus (118) Google Scholar, 14Fahraeus R. Lain S. Ball K.L. Lane D.P. Oncogene. 1998; 16: 587-596Crossref PubMed Scopus (80) Google Scholar). Interestingly, reintroduction of p16INK4A activity into tumor lines only blocks proliferation in cells where pRb is active (15Lukas J. Parry D. Aagaard L. Mann D.J. Bartkova J. Strauss M. Peters G. Bartek J. Nature. 1995; 375: 503-506Crossref PubMed Scopus (864) Google Scholar, 16Medema R.H. Herrera R.E. Lam F. Weinberg R.A. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 6289-6293Crossref PubMed Scopus (407) Google Scholar). Thus, inhibition of CDK4 and CDK6 kinase activity may only be a useful anticancer treatment in tumors where pRb is functional, but other members of the pathway such as p16INK4A or possibly cyclin D1 have aberrant functions. Further support for this model has come from studies using antisense cDNA technology and an antibody to block cyclin D1 function, which showed that loss of cyclin D1 action only blocks proliferation in cells with active pRb (17Tam S.W. Theodoras A.M. Shay J.W. Draetta G.F. Pagano M. Oncogene. 1994; 9: 2663-2674PubMed Google Scholar, 18Lukas J. Bartkova J. Rohde M. Strauss M. Bartek J. Mol. Cell. Biol. 1995; 15: 2600-2611Crossref PubMed Scopus (339) Google Scholar). Expression of antisense cDNA has also been used to show that loss of cyclin D1 function will also block pRb phosphorylation and inhibit the growth and tumorigenicity of colon cancer cells (19Arber N. Doki Y. Han E.K. Sgambato A. Zhou P. Kim N.H. Delohery T. Klein M.G. Holt P.R. Weinstein I.B. Cancer Res. 1997; 57: 1569-1574PubMed Google Scholar). These data provide a strong rationale for inhibition of cyclin D-dependent kinase activity as an approach to cancer chemotherapy in tumor cells where pRb is functional, but the activity of other members of the pRb/cyclin D/p16INK4A pathway has been altered. Screening efforts have recognized the [2,3-d]pyridopyrimidines as inhibitors of CDK4 (20Barvian M. Boschelli D.H. Cossrow J. Dobrusin E. Fattaey A. Fritsch A. Fry D. Harvey P. Keller P. Garrett M. La F. Leopold W. McNamara D. Quin M. Trumpp-Kallmeyer S. Toogood P. Wu Z. Zhang E. J. Med. Chem. 2000; 43: 4606-4616Crossref PubMed Scopus (147) Google Scholar). Here we describe the identification of one such compound PD 0183812 as a potent and specific inhibitor in vitro of the cyclin D-dependent kinases CDK4 and CDK6 (Fig. 1). This study characterizes the biochemical properties of PD 0183812 and provides evidence that the biological effects of this compound are consistent with cyclin D-dependent kinase inhibition. CDK assays for IC50determinations and kinetic evaluation were performed in 96-well filter plates (Millipore, MADVN6550). All CDK-cyclin kinase complexes were expressed in insect cells through baculovirus infection and purified as described previously (21Booher R.N. Holman P.S. Fattaey A. J. Biol. Chem. 1997; 272: 22300-22306Abstract Full Text Full Text PDF PubMed Scopus (220) Google Scholar). The substrate for the assays was a fragment (amino acids 792–928) of pRb fused to GST (GST·RB-Cterm, Ref. 22Meyerson M. Harlow E. Mol. Cell. Biol. 1994; 14: 2077-2086Crossref PubMed Scopus (730) Google Scholar). The total volume for each well was 0.1 ml containing a final concentration of 20 mm Tris-HCl, pH 7.4, 50 mmNaCl, 1 mm dithiothreitol, 10 mmMgCl2, 25 μm ATP (for CDK4-cyclin D1, CDK6-cyclin D2, and CDK6-cyclin D3) or 12 μm ATP (for CDK2-cyclin E, CDK2-cyclin A, and CDC2-cyclin B) containing 0.25 μCi of [γ-32P]ATP, 20 ng of enzyme, 1 μg of GST·RB-Cterm, and appropriate dilutions of inhibitor. All components except the [γ-32P]ATP were added to the wells, and the plate was placed on a plate mixer for 2 min. The reaction was then started by adding the [γ-32P]ATP, and the plate was incubated at 25 °C for 15 min. The reaction was terminated by addition of 0.1 ml of 20% trichloroacetic acid, and the plate was kept at 4 °C for at least 1 h to allow the substrate to precipitate. The wells were then washed five times with 0.2 ml of 10% trichloroacetic acid, and radioactive incorporation was determined with a β plate counter (Wallac Inc., Gaithersburg, MD). Kinase assays for PDGFr, FGFr, EGFr, SRC, and PKC kinases were performed as described previously (23Fry D.W. Nelson J.M. Slintak V. Keller P.R. Rewcastle G.W. Denny W.A. Zhou H. Bridges A.J. Biochem. Pharmacol. 1997; 54: 877-887Crossref PubMed Scopus (73) Google Scholar, 24Fry D.W. Kraker A.J. Connors R.C. Elliott W.L. Nelson J.M. Showalter H.D. Leopold W.R. Anticancer Drug Design. 1994; 9: 331-351PubMed Google Scholar). All cell lines were maintained at 37 °C, 5% CO2 in Dulbecco's modified Eagle's medium containing 10% fetal bovine serum (Life Technologies, Inc). For cell counts, cells were trypsinized and counted using a Coulter Z2 particle count and size analyzer. Cells were seeded at 2 × 104 per well in a 96-well Cytostar T plate (Amersham Pharmacia Biotech) and incubated overnight. Varying concentrations of PD 0183812 were added to the wells and incubated for 24 h at 37 °C. 0.1 μCi of [14C]thymidine was added to each well, and incorporation of the radiolabel was allowed to proceed for 48 h. Incorporated radioactivity was determined with a β plate counter. Cells were washed and harvested using phosphate-buffered saline (PBS) containing 5 mm EDTA. Cells were then washed in PBS containing 1% FBS (1% FBS/PBS), fixed in 85% ethanol and stored at 4 °C for at least 16 h and up to 5 days before staining with propidium iodide (PI). For PI staining, cells were washed in 1% FBS/PBS prior to incubation at 37 °C for 30 min in the same solution containing 40 μg/ml PI (Molecular Probes) and 250 μg/ml RNase A (Roche Diagnostics Ltd.). Data were collected using a Coulter EPICS Elite ESP equipped with a Spectraphysics argon-ion laser and analyzed using the WinMDI program, version 2.8. Results represent a minimum of 15,000 cells assayed for each sample. Cells were trypsinized and 200 μl of each sample was counted in 19.8 ml of PBS using a Coulter Z2 particle count and size analyzer. The rest of each sample was washed once with ice-cold PBS and lysed on ice for 30 min, in lysis buffer (50 mm HEPES, pH 7.4, 250 mm NaCl, 0.1% Nonidet P-40) containing 1 mmDTT, 1 mm EDTA, 1 mm NaF, 10 mmβ-glycerophosphate, 0.1 mm sodium orthovanadate, and a complete mini EDTA-free protease inhibitor mixture tablet per 10 ml of lysis buffer (Roche Diagnostics Corp). The volume of lysis buffer used was 100 μl per 1 × 106 cells. Following lysis, samples were clarified by centrifugation at 18,000 × gfor 10 min at 4 °C, and protein concentrations were determined using the Bradford assay (25Bradford M.M. Anal. Biochem. 1976; 72: 248-254Crossref PubMed Scopus (211946) Google Scholar). Proteins were resolved by SDS-polyacrylamide gel electrophoresis using 6% polyacrylamide gels (10 μl of cell lysate per gel lane) and electroblotted onto ImmobilonTM-P membranes (Millipore, Bedford, MA). Membranes were blocked overnight at 4 °C in 50 mm Tris-HCl, pH 8.0, 150 mm NaCl, and 0.1% Tween 20 containing 3% casein and then incubated in the same buffer with primary antibodies for 3 h at room temperature followed by a 1-h incubation with a horseradish peroxidase-conjugated goat anti-rabbit secondary antibody (1:10,000 dilution, Bio-Rad). Immunodetection was carried out using enhanced chemiluminescence (ECL, Amersham Pharmacia Biotech). For Western analysis, pRb was detected using the Rb C-terminal control antibody from Cell Signaling Technology at a dilution of 1:2000. The phosphorylation status of pRb at serine 780 was detected with the phospho-Rb (Ser-780) antibody (at a dilution of 1:4000) also from Cell Signaling Technology. We have recently identified the [2,3-d]pyridopyrimidines as inhibitors of CDK4 (20Barvian M. Boschelli D.H. Cossrow J. Dobrusin E. Fattaey A. Fritsch A. Fry D. Harvey P. Keller P. Garrett M. La F. Leopold W. McNamara D. Quin M. Trumpp-Kallmeyer S. Toogood P. Wu Z. Zhang E. J. Med. Chem. 2000; 43: 4606-4616Crossref PubMed Scopus (147) Google Scholar). From this series we identified PD 0183812 as an extremely potent inhibitor of CDK4-cyclin D1 with an IC50 value of 0.008 μm (Table I). Similar IC50 values of 0.0071 and 0.013 μm were also measured for inhibition of CDK6-cyclin D2 and CDK6-cyclin D3 respectively (Table I). In contrast, the non-cyclin D-dependent kinases CDK2 and CDC2 exhibited considerably higher IC50 values for inhibition by PD 0183812 (Table I). Specifically, selectivity ratios of >500-, 26-, and 20-fold were obtained for CDC2-cyclin B, CDK2-cyclin A, and CDK2-cyclin E, respectively versus CDK4-cyclin D1. PD 0183812 had very little inhibitory activity against other protein kinases including PDGFr, FGFr, EGFr, SRC, and PKC (Table I). Kinetic analysis shows that PD 0183812 exhibited pure competitive inhibition with respect to ATP as indicated by the double reciprocal plot of ATP concentrationversus catalytic activity of CDK4-cyclin D1 (Fig.2). Thus, PD 0183812 is a potent and selective ATP competitive inhibitor of the cyclin D-dependent kinases CDK4 and CDK6.Table IPD 0183812 is a selective inhibitor of cyclin D-dependent kinasesKinasePD 0183812 (IC50)μmCDK4/D10.008CDK6/D20.0071CDK6/D30.013CDK2/E0.165CDK2/A0.2095CDC2/B>40PDGFr14FGFr8.6EGFr>50SRC22PKC>50Assays were carried out as described under "Experimental Procedures." Open table in a new tab Assays were carried out as described under "Experimental Procedures." As outlined in the Introduction, the major known cellular target of the cyclin D-dependent kinases is pRb and phosphorylation of this protein is required for progression through G1. Therefore a compound that can inhibit the cellular activity of both CDK4 and CDK6 in tumor cells expressing functional pRb, but where another member of the pRb/cyclin D/p16INK4A pathway is altered would be predicted to invoke a G1arrest. In tumor cells in which pRb is not present or not functional and where the pRb/cyclin D/p16INK4A pathway is not required for G1 progression, treatment with such an inhibitor would not be expected to cause an arrest in G1. To test the ability of PD 0183812 to behave in this manner, we treated three human carcinoma cell lines that express pRb but not p16INK4A, (MDA-MB-453, breast carcinoma; H1299, lung carcinoma; U251, glioblastoma; data not shown; Refs. 26Tam S.W. Shay J.W. Pagano M. Cancer Res. 1994; 54: 5816-5820PubMed Google Scholar, 27Kataoka M. Wiehle S. Spitz F. Schumacher G. Roth J.A. Cristiano R.J. Oncogene. 2000; 19: 1589-1595Crossref PubMed Scopus (56) Google Scholar, 28Kubo A. Nakagawa K. Varma R.K. Conrad N.K. Cheng J.Q. Lee W.C. Testa J.R. Johnson B.E. Kaye F.J. Kelley M.J. Clin. Cancer Res. 1999; 5: 4279-4286PubMed Google Scholar) and three cell lines that do not express pRb, but do express p16INK4A (MDA-MB-468, breast carcinoma; H2O09, lung carcinoma; SF539, glioblastoma; data not shown; Refs. 26Tam S.W. Shay J.W. Pagano M. Cancer Res. 1994; 54: 5816-5820PubMed Google Scholar, 28Kubo A. Nakagawa K. Varma R.K. Conrad N.K. Cheng J.Q. Lee W.C. Testa J.R. Johnson B.E. Kaye F.J. Kelley M.J. Clin. Cancer Res. 1999; 5: 4279-4286PubMed Google Scholar, 29Otterson G.A. Kratzke R.A. Coxon A. Kim Y.W. Kaye F.J. Oncogene. 1994; 9: 3375-33788PubMed Google Scholar), for 24 h with 0.1–1.5 μmPD 0183812. In all three cell lines that express pRb, a significant increase in the G1 fraction and a concomitant decrease in the S and G2/M fractions was observed when cells were treated with between 0.2 and 0.8 μm PD 0183812 (Fig.3). At higher concentrations, a significant increase in G2/M was seen in the H1299 and U251 tumor lines. No increase in G1 was noted between 0.2 and 0.8 μm for the cell lines not expressing pRb (Fig. 3). However a strong increase in G2/M starting at 0.8 μm for all three of these cell lines was noted. From these results, we conclude that in the tumor lines tested, PD 0183812 induces a pRb-specific G1 arrest indicative of cyclin D-dependent kinase inhibition. However, high concentrations of the drug cause a pRb-independent increase in G2/M. We next went on to address whether the pRb-specific G1arrest caused by PD 0183812 correlated with a block in DNA synthesis as measured by thymidine incorporation. To do this we selected the two breast cancer cell lines MDA-MB-453 and MDA-MB-468. Each cell line was exposed to varying concentrations of PD 0183812 for 24 h, at which point the amount of thymidine incorporated by all the samples over the next 48 h in the presence of PD 0183812 was measured (Fig.4). From this graph, the IC50of PD 0183812 for inhibition of DNA synthesis was calculated and found to be 0.32 μm for the MDA-MB-453 cell line, which expresses pRb, and >3.0 μm for the MDA-MB-468 cell line, which does not. For the MDA-MB-453 cell line, this is similar to the concentration range of the compound (0.2–0.8 μm), which causes a G1 arrest (Fig. 4). For the MDA-MB-468 cell line, treatment of the cells with <1.0 μm of PD 0183812 had no significant effect on DNA synthesis. Thus treatment of tumor cell lines with PD 0183812 causes a pRb-specific G1 arrest, which correlates with a block in DNA synthesis. Because we had shown that PD 0183812 is a cyclin D-dependent kinase selective inhibitor in vitroand causes a pRb-specific G1 arrest in human tumor cells, we went on to investigate whether this compound could also block thein vivo phosphorylation of pRb, the cellular substrate of the cyclin D-dependent kinases. For this experiment we again used the human breast carcinoma cell line MDA-MB-453. Cells were treated with 0.1–0.9 μm PD 0183812. Untreated cells or cells treated with the drug vehicle, Me2SO, were included as controls. Mimosine, a plant amino acid often used to synchronize cells in G1, was incorporated into the experiment as a drug that is known to cause a G1 arrest (30Lalande M. Exp. Cell Res. 1990; 186: 332-339Crossref PubMed Scopus (178) Google Scholar). Expression and phosphorylation of pRb was monitored by Western blot analysis using two antibodies, one for detection of pRb expression and a second for detection of phosphorylation at serine 780 of pRb (P-Ser-780Rb). Phosphorylation at serine 780 on pRb was investigated as it has been reported that CDK4 can specifically phosphorylate this sitein vitro and that serine 780 is phosphorylated in mid G1 when both CDK4 and CDK6 but not CDK2 are active (22Meyerson M. Harlow E. Mol. Cell. Biol. 1994; 14: 2077-2086Crossref PubMed Scopus (730) Google Scholar, 31Kitagawa M. Higashi H. Jung H.K. Suzuki-Takahashi I. Ikeda M. Tamai K. Kato J. Segawa K. Yoshida E. Nishimura S. Taya Y. EMBO J. 1996; 15: 7060-7069Crossref PubMed Scopus (528) Google Scholar,32Matsushime H. Quelle D.E. Shurtleff S.A. Shibuya M. Sherr C.J. Kato J.Y. Mol. Cell. Biol. 1994; 14: 2066-2076Crossref PubMed Scopus (1023) Google Scholar). In addition, it has been shown that immunoprecipitates of the transcription factor E2F1, which associates with pRb, contain no serine 780-phosphorylated pRb, indicating that phosphorylation of this site may be involved in disruption of the E2F1·pRb complex (31Kitagawa M. Higashi H. Jung H.K. Suzuki-Takahashi I. Ikeda M. Tamai K. Kato J. Segawa K. Yoshida E. Nishimura S. Taya Y. EMBO J. 1996; 15: 7060-7069Crossref PubMed Scopus (528) Google Scholar). FACS analysis was also carried out on a portion of each sample so that the level of pRb phosphorylation could be compared side by side with the fraction of cells in the G1 phase of the cell cycle. From these studies, we found that whereas the untreated or Me2SO-treated cells possessed mainly hyperphosphorylated pRb, with strong phosphorylation at serine 780, samples treated with 0.3–0.9 μm PD 0183812 showed an increase in the G1 fraction of the cells (peaking at 0.5 μm), which correlated with loss of serine 780 phosphorylation on pRb (Fig.5). We also noted that there was no significant increase in the sub-G1 population of cells, which is often used as an indicator of apoptosis (33Ormerod M, G. Collins M, K. Rodriguez-Tarduchy G. Robertson D. J. Immunol. Methods. 1992; 153: 57-65Crossref PubMed Scopus (187) Google Scholar). Interestingly, the cells treated with mimosine showed an increase in their G1 population equivalent to treatment with 0.3 μm PD 0183812, but there was no significant loss of pRb phosphorylation. Together, these data suggest that treatment of pRb-expressing tumor cells with PD 0183812 causes a G1arrest, which coincides with loss of pRb phosphorylation at serine 780, indicating inhibition of cyclin D-dependent kinase activity. To investigate the long term effects of PD 0183812 on cell proliferation, we treated the MDA-MB-453 cell line with either Me2SO (vehicle control) or 0.5 μm PD 0183812, a concentration of the drug, which we have shown will cause a G1 block and a loss of pRb phosphorylation (Fig. 5). We then took samples for cell counts, and examination of pRb phosphorylation by Western blot at 0–5 days after drug addition. It should be noted that medium with either drug or Me2SO was only added once to the cells (on day 0) and was not replaced during the experiment. Looking at the cell counts between 1 and 5 days after treatment, it was clear that there was a significant block in cell proliferation (Fig.6 A), which correlated with loss of pRb phosphorylation (Fig. 6 B). It should be noted that the control cells ceased to grow after day 4, which may have been because of their confluency and exhaustion of nutrients in the cell medium. Thus, treatment of pRb expressing tumor cells with PD 0183812 causes a cytostatic block in proliferation which is maintained for at least 5 days and correlates with a loss of pRb phosphorylation at serine 780 indicative of cyclin D-dependent kinase inhibition. The final question addressed in this study is whether the block in tumor cell proliferation induced by PD 0183812 is reversible. For this experiment the MDA-MB-453 cells were exposed to either Me2SO (vehicle control) or 0.5 μm PD 0183812 for 24 h. After 24 h (day 1) the medium was removed, and the cells were washed. Fresh medium containing Me2SO was added to the Me2SO-treated cells. Fresh medium containing 0.5 μm PD 0183812 was added to half the PD 0183812-treated cells, whereas the other half were treated with medium containing Me2SO. All medium was changed at day 3 to try and counteract nutrient exhaustion, especially where cells were proliferating for the length of the experiment. Samples were then taken for cell counts and detection by Western blot of pRb expression and phosphorylation at serine 780. It is clear from the cell counts (Fig.7 A), that PD 0183812 causes a sustained block in proliferation. This result is reflected in the Western blot data, which show that phosphorylation at serine 780 of pRb is low in the cells treated with PD 0183812 for the first 5 days of the experiment, compared with the Me2SO-treated cells (Fig.7 B). By day 6, the Me2SO-treated cells have stopped growing, and pRb phosphorylation at serine 780 is low. As in the previous experiment, this may be because of the confluency of the cells and nutrient exhaustion of the medium. If however, the compound is removed on day 1, 24 h after addition, it appears that the cells recover and resume proliferation (Fig. 7 A). It must be noted that there does appear to be a lag in resuming proliferation because cell counts on days 2–4 were still low. In comparison, the Western blot data from this experiment (Fig. 7 B) show that if the compound is removed on day 1 the amount of pRb phosphorylation on serine 780 on days 2–5 returns to the level seen for the vehicle control-treated cells. Therefore PD 0183812 causes a block in proliferation, which is reversible when the compound is removed 24 h after addition and is accompanied by a return of phosphorylation on serine 780 of pRb. Within the last 5 years, enormous progress has been made in deciphering the mechanism by which mammalian cells regulate and progress through the cell cycle. Clearly CDKs have emerged as key regulators of cell cycle progression, and in fact movement either proceeds or stops based on the catalytic activity of these enzymes. CDKs have also been some of the more intensely studied enzymes in terms of potential targets in cancer chemotherapy. The cyclin D-dependent kinases have been of special interest because of the exceptionally high percentage of tumors that exhibit abnormalities in the pRb/cyclin D/p16INK4A pathway. In particular the fact that tumors that have functional pRb often harbor an alteration in another member of this pathway, such as a defect in the cyclin D-dependent kinase inhibitor p16INK4A, has led to the suggestion that inhibition of cyclin D-dependent kinase activity in these tumors may have therapeutic value. We therefore set out to identify inhibitors of cyclin D-dependent kinase activity by initially screening for inhibitors of CDK4. Through this effort we recognized the [2,3-d]pyridopyrimidines as inhibitors of this enzyme (20Barvian M. Boschelli D.H. Cossrow J. Dobrusin E. Fattaey A. Fritsch A. Fry D. Harvey P. Keller P. Garrett M. La F. Leopold W. McNamara D. Quin M. Trumpp-Kallmeyer S. Toogood P. Wu Z. Zhang E. J. Med. Chem. 2000; 43: 4606-4616Crossref PubMed Scopus (147) Google Scholar). Subsequent chemical modification lead to the identification of PD 0183812 as a compound with unprecedented potency in vitroagainst both the cyclin D-dependent kinases CDK4 and CDK6. Investigation into the cellular behavior of PD 0183812 revealed that this compound gave a pRb dependent G1 cell cycle arrest. This result is indicative of cyclin D-dependent kinase inhibition, as this activity has previously been shown to be required for pRb-dependent G1 progression (15Lukas J. Parry D. Aagaard L. Mann D.J. Bartkova J. Strauss M. Peters G. Bartek J. Nature. 1995; 375: 503-506Crossref PubMed Scopus (864) Google Scholar, 16Medema R.H. Herrera R.E. Lam F. Weinberg R.A. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 6289-6293Crossref PubMed Scopus (407) Google Scholar). The concentration of PD 0183812 required to give this pRb-specific G1 arrest (0.5 μm in MDA-MB-453 cells) also caused a loss of phosphorylation on pRb at serine 780, a site previously reported to be targeted by CDK4 for phosphorylation and correlated with its IC50 for inhibition of DNA synthesis (31Kitagawa M. Higashi H. Jung H.K. Suzuki-Takahashi I. Ikeda M. Tamai K. Kato J. Segawa K. Yoshida E. Nishimura S. Taya Y. EMBO J. 1996; 15: 7060-7069Crossref PubMed Scopus (528) Google Scholar). Together these data suggest that the cellular target of PD 0183812, which gives a pRb-specific G1 arrest in human tumor cells, is the cyclin D-dependent kinases. At higher concentrations however, PD 0183812 caused a significant increase in the G2/M population of cells in a non-pRb-dependent manner. There are a number of reasons why this may have occurred. One option is that at high concentrations, PD 0183812 may be inhibiting other kinases (including other CDKs) in the cell that are rate-limiting at G2/M but not G1. A second possibility is that a cyclin D-dependent kinase (but pRb-independent) step may be rate-limiting in G2/M at these concentrations of the drug. There is some evidence that CDK4 may play a role in the G2/M transition and so for the moment, this possibility cannot be ruled out (34Gabrielli B, G. Sarcevic B. Sinnamon J. Walker G. Castellano M. Wang X.Q. Ellem K.A.,. J. Biol. Chem. 1999; 274: 13961-13969Abstract Full Text Full Text PDF PubMed Scopus (63) Google Scholar). Finally we come to the effect of PD 0183812 on cellular proliferation. From the studies presented here it is clear that one treatment of 0.5 μm PD 0183812 is sufficient to produce a G1arrest, loss of phosphorylation on pRb at serine 780, and significantly inhibit growth for at least 5 days. This is an extremely interesting result as there has been some discussion as to whether a catalytic inhibitor of CDK4 would be sufficient to arrest cycling cells (35Shapiro G.I. Harper J.W. J. Clin. Invest. 1999; 104: 1645-1653Crossref PubMed Scopus (352) Google Scholar). One reason for this, was the discovery that expression of p16INK4A, but not a dominant-negative form of CDK4 will cause a cell cycle arrest (36Jiang H. Chou H.S. Zhu L. Mol. Cell. Biol. 1998; 18: 5284-5290Crossref PubMed Google Scholar). This may be because of the fact that p16INK4A expression results in redistribution of the CDK inhibitor p27KIP1 within the cell and inhibition of CDK2 as well as CDK4 (36Jiang H. Chou H.S. Zhu L. Mol. Cell. Biol. 1998; 18: 5284-5290Crossref PubMed Google Scholar). With PD 0183812 we also found that removal from the cells after 24 h allowed rephosphorylation of pRb and an eventual resumption of proliferation. Thus at this stage the effects of PD 0183812 are reversible. It is interesting to note that expression of p16INK4A is associated with replicative senescence, and it will be of interest to investigate the long term effects of PD 0183812 on tumor cells (37Huschtscha L, I. Reddel R.R. Carcinogenesis. 1999; 20: 921-926Crossref PubMed Scopus (95) Google Scholar). In recent years a number of CDK inhibitors have been disclosed, which have a wide variety of potencies and specificities (38Fry D.W. Garrett M.D. Curr. Opn. Oncol. Endocr. Met. Invest. Drugs. 2000; 2: 40-59Google Scholar). Many of these existing compounds such as the purines, olomoucine and roscovitine, and the non-purines, butyrolactone and kenpaullone, show considerable activity against CDK2 and CDC2, but little activity against CDK4 (39Meijer L. Kim S.H.,. Methods Enzymol. 1997; 283: 113-128Crossref PubMed Scopus (130) Google Scholar, 40Zaharevitz D.W. Gussio R. Leost M. Senderowicz A.M. Lahusen T. Kunick C. Meijer L. Sausville E.A. Cancer Res. 1999; 59: 2566-2569PubMed Google Scholar). The flavone flavopiridol also known as L86-8275 will inhibit CDK4 but with a similar potency to CDK2 and CDC2, and there is building evidence that it may have other cellular targets including aldehyde dehydrogenase, glycogen phosphorylase, and duplex DNA (39Meijer L. Kim S.H.,. Methods Enzymol. 1997; 283: 113-128Crossref PubMed Scopus (130) Google Scholar, 41Schnier J.B. Kaur G. Kaiser A. Stinson S.F. Sausville E.A. Gardner J. Nishi K Bradbury E.M. Senderowicz A.M. FEBS Lett. 1999; 454: 100-104Crossref PubMed Scopus (53) Google Scholar, 42Oikonomakos N.G. Schnier J.B. Zographos S.E. Skamnaki V.T. Tsitsanou K.E. Johnson L.N. J. Biol. Chem. 2000; 275: 34566-34573Abstract Full Text Full Text PDF PubMed Scopus (123) Google Scholar, 43Bible K.C. Bible Jr, R.H. Kottke T.J. Svingen P.A. Xu K. Pang Y.P. Hajdu E. Kaufmann S.H. Cancer Res. 2000; 60: 2419-2428PubMed Google Scholar). Thus potent and selective inhibitors of the cyclin D-dependent kinases CDK4 and CDK6 have been elusive. Recently however, there have been several reports of compounds that do inhibit CDK4-cyclin D1 in vitro and show selective growth inhibition of tumor cells with functional pRb or defective p16INK4A (28Kubo A. Nakagawa K. Varma R.K. Conrad N.K. Cheng J.Q. Lee W.C. Testa J.R. Johnson B.E. Kaye F.J. Kelley M.J. Clin. Cancer Res. 1999; 5: 4279-4286PubMed Google Scholar, 44Kent L.L. Hull-Campbell N.E. Lau T. Wu J.C. Thompson S, A. Nori M. Biochem. Biophys. Res. Commun. 1999; 260: 768-774Crossref PubMed Scopus (47) Google Scholar, 45Soni R. Muller L. Furet P. Schoepfer J. Stephan C. Zumstein-Mecker S. Fretz H. Chandhuri B. Biochem. Biophys. Res. Commun. 2000; 275: 877-884Crossref PubMed Scopus (152) Google Scholar) These reports therefore reinforce the gathering evidence that the identification of pharmacological inhibitors of cyclin D-dependent kinases could allow the development of a novel therapeutic approach to the treatment of patients with tumors where the retinoblastoma gene is still functional but other members of the pRb/cyclin D/p16INK4A pathway are defective. With PD 0183812, we have identified the most potent CDK4-cyclin D1 inhibitor to date. We have also shown that PD 0183812 can inhibit CDK6-cyclin D2 and CDK6-cyclin D3 with a similar potency to CDK4-cyclin D1. Therapeutically this is an important point because tumors that, for example, have functional pRb but defective p16INK4A and express CDK6 as well as CDK4 along with one or more of the D-type cyclins, may require inhibition of both these CDKs for therapeutic value. PD 0183812 would be capable of this whereas in contrast the recently discovered CDK4-cyclin D1 inhibitor fascaplysin would not, because it specifically targets this form of the cyclin D-dependent kinases (45Soni R. Muller L. Furet P. Schoepfer J. Stephan C. Zumstein-Mecker S. Fretz H. Chandhuri B. Biochem. Biophys. Res. Commun. 2000; 275: 877-884Crossref PubMed Scopus (152) Google Scholar). Thus the identification of PD 0183812 as a small molecule inhibitor of CDK4 and CDK6 provides us with an important tool to investigate the cellular processes in which the cyclin D-dependent kinases participate and a chemical lead in the development of a chemotherapeutic agent for the treatment of genetically characterized tumors. We would like to thank everyone who has worked on the CDK inhibitor project at both Onyx Pharmaceuticals and at Pfizer Global Research and Development, Ann Arbor Laboratories. We would also like to thank J. Titley and M. Quinn for their assistance with the FACS analysis.
Referência(s)