Tel: +44-1158466863, Fax: +44-1159513412
CDKI-71, a novel CDK9 inhibitor, is preferentially cytotoxic to cancer cells compared to flavopiridol
Version of Record online: 21 JUN 2011
Copyright © 2011 UICC
International Journal of Cancer
Volume 130, Issue 5, pages 1216–1226, 1 March 2012
How to Cite
Liu, X., Shi, S., Lam, F., Pepper, C., Fischer, P. M. and Wang, S. (2012), CDKI-71, a novel CDK9 inhibitor, is preferentially cytotoxic to cancer cells compared to flavopiridol. Int. J. Cancer, 130: 1216–1226. doi: 10.1002/ijc.26127
- Issue online: 20 DEC 2011
- Version of Record online: 21 JUN 2011
- Accepted manuscript online: 11 APR 2011 07:15AM EST
- Manuscript Accepted: 25 MAR 2011
- Manuscript Received: 29 OCT 2010
- Cancer Research UK. Grant Number: C21568/A8988
- anticancer agents;
- CDK9/cyclin T inhibitor;
- RNA polymerase II;
- antiapoptotic proteins Bcl-2 and Mcl-1;
- p53 independent apoptosis;
- caspase activation;
- chronic lymphocytic leukemia;
- normal B-cells and T-cells
Cancer cells appear to depend heavily on antiapoptotic proteins for survival and so targeted inhibition of these proteins has therapeutic potential. One innovative strategy is to inhibit the cyclin-dependent kinases (CDKs) responsible for the regulation of RNA polymerase II (RNAPII). In our study, we investigated the detailed cellular mechanism of a novel small-molecule CDK inhibitor (CDKI-71) in cancer cell lines, primary leukemia cells, normal B - & T- cells, and embryonic lung fibroblasts and compared the cellular and molecular responses to the clinical CDK inhibitor, flavopiridol. Like flavopiridol, CDKI-71 displayed potent cytotoxicity and caspase-dependent apoptosis induction that were closely associated with the inhibition of RNAPII phosphorylation at serine-2. This was caused by effective targeting of cyclinT–CDK9 and resulted in the downstream inhibition of Mcl-1. No correlation between apoptosis and inhibition of cell-cycle CDKs 1 and 2 was observed. CDKI-71 showed a 10-fold increase in potency in tumor cell lines when compared to MRC-5 human fibroblast cells. Significantly, CDKI-71 also demonstrated potent anti-chronic lymphocytic leukemia activity with minimal toxicity in normal B- and T-cells. In contrast, flavopiridol showed little selectivity between cancer and normal cells. Here, we provide the first cell-based evidence that flavopiridol induces DNA double-strand breaks: a fact which may explain why flavopiridol has such a narrow therapeutic window in preclinical and clinical settings. Taken together, our data provide a rationale for the development of selective CDK inhibitors as therapeutic agents and CDKI-71 represents a promising lead in this context.
Cyclin-dependent kinases (CDKs) can be classified into two major groups based on their roles in the regulation of cell cycle and transcription.1, 2 Members of the first group are cell-cycle regulators. The complexes of cyclin D–CDK4/6 and cyclin E–CDK2 facilitate the transition of G1–S phase by sequentially phosphorylating the retinoblastoma protein (Rb), while cyclin A–CDK1/2 and cyclin B–CDK1 are essential for progression of S phase and transition of G2–M, respectively.3 Multiple CDKs control the cell cycle in mammals and have long been considered essential for normal proliferation, development and homeostasis. However, most CDKs have recently turned out to be dispensable for cell proliferation due to a high level of functional redundancy, promiscuity and compensatory mechanisms.4 This suggests that specific inhibition of CDK2 or CDK4/6 alone may not be an effective therapeutic strategy.
The second group comprises the transcriptional CDKs, particularly cyclin T–CDK9 and cyclin H–CDK7, which are involved in the regulation of RNA transcription. Transcription of RNA polymerase II (RNAPII) is highly regulated and involves a sequence of events leading to phosphorylation of the carboxy-terminal domain (CTD), during which the general transcription factor II (TFIIH) complex, containing CDK7, first phosphorylates the serine-5 residue of the CTD heptapeptide repeats Y1S2P3T4S5P6S7. The positive transcriptional elongation factor b (p-TEFb), consisting of cyclin T–CDK9, then phosphorylates the 5,6-dichlorobenzimidazone-1-β-D-ribofuranoside (DRB)-sensitive inducing factor and the negative elongation factor, followed by serine 2 of the CTD to facilitate transcriptional elongation.2 While CDK7 is also recognized as a CDK-activating kinase (CAK) involved in cell-cycle regulation, CDK9 appears to play no role in cell-cycle regulation.
Over the past decade, the intensive search for pharmacological CDK inhibitors has led to several clinical candidates, and the focus on transcriptional CDKs underlines their antitumor activity. Flavopiridol, the first CDK inhibitor to enter clinical trials, is the most potent CDK9 inhibitor identified to date and has demonstrated marked antitumor activity in chronic lymphocytic leukemia (CLL).2, 5 Although flavopiridol inhibits a number of CDKs and other kinases, the primary mechanism of action of flavopiridol is believed to be through CDK9 inhibition, leading to downregulation of the transcription of antiapoptotic proteins in CLL cells.6, 7
Resistance to apoptosis is a universal hallmark of cancer and is often achieved by upregulating antiapoptotic proteins and prosurvival signalling pathways. Bcl-2 family proteins are key regulators of apoptosis,8, 9 and the expression of Mcl-1 appears to predict the in vitro response to drugs and clinical outcome in a number of human cancers including CLL.10 Unlike other members of the Bcl-2 family, Mcl-1 mRNA and protein have a short half-life. Importantly, selective downregulation of Mcl-1 has been shown to be sufficient to promote apoptosis in myeloma and CLL cancer cells suggesting that continuous expression of Mcl-1 is required for survival of tumor cells.11, 12
Our interest in the development of kinase inhibitors has resulted in a number of clinical and preclinical drug candidates.13–15 We report here the cellular mechanism of a novel preclinical candidate, CDKI-71. We compared the effects of CDKI-71 with flavopiridol to verify its cellular mechanism of action. Using biochemical and cell-based assays, we demonstrated that CDKI-71 inhibited CDK9 activity that resulted in the downregulation of antiapoptotic proteins and the induction of apoptosis. Unlike flavopiridol, however, CDKI-71 did not induce DNA strand breaks and showed preferential toxicity in cancer cells when compared to normal cells.
Material and Methods
Synthesis of 3-(5-cyano-4-(4-methyl-2-(methylamino)thiazol-5-yl)pyrimidin-2-ylamino)benzenesulfonamide (CDKI-71) has previously been described.16 Flavopiridol and cisplatin were purchased from Sigma-Aldrich (Gillingham, UK).
Inhibition of CDKs and other kinases was measured by radiometric assay using the Millipore KinaseProfiler services. Half-maximal inhibition (IC50) values were calculated from 10-point dose–response curves and apparent inhibition constants (Ki) were calculated from the IC50 values and appropriate Km (ATP) values for the kinases in question.17
A2780 and MRC-5 cell lines were purchased from ECACC. All other cell lines were obtained from the cell bank at the Centre for Biomolecular Sciences, University of Nottingham, UK. MRC-5 cells were cultured in essential medium with 10% fetal bovine serum (FBS), 7.5% sodium bicarbonate, 1% 0.1 mM nonessential amino acids, 1% 1 M HEPES (4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid), 1% 200 mM L-glutamine and 1% penicillin. All other cell lines were maintained in RPMI-1640 (Roswell Park Memorial Institute) with 10% FBS.
MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide, Sigma, Gillingham, UK) assays were performed as reported previously.15 Concentrations of compound required to inhibit 50% of cell growth (GI50) were calculated using nonlinear regression analysis.
Activity of caspase 3/7 was measured using the Apo-ONE Homogeneous Caspase-3/7 kit (Promega G7790, Promega, Southampton, UK).
Cell cycle analysis and detection of apoptosis
Cells (4 × 105) were cultured for 48 hr in medium alone or with varying concentrations of inhibitor. Cell cycle status was analyzed using a Beckman Coulter EPICS-XL MCLTM flow cytometer, and data were analyzed using EXPO32TM software. Apoptosis was also confirmed using Fluorescein isothiocyanate (FITC) annexinV/propidium iodide (PI) staining after cells were cultured in medium only or with varying concentrations of inhibitors according to the protocols (BD Bioscience, Oxford, UK). The annexin V/PI-positive apoptotic cells were enumerated using flow cytometry. The percentage of cells undergoing apoptosis was defined as the sum of early apoptosis (annexin V-positive cells) and late apoptosis (annexin V-positive and PI-positive cells). The pan-caspase inhibitor Z-Val-Ala-Asp-(OMe)-CH2F (Z-VAD-fmk, Sigma, Gillingham, UK) was dissolved in <0.1% Dimethyl sulfoxide (DMSO) and used at a concentration of 50 μM.
Detection of apoptosis in primary cells
Freshly isolated primary CLL cells and normal B- and T-cells were cultured in RPMI with 10% fetal calf serum and L-glutamine, penicillin and streptomycin. Cells were maintained at 37°C in an atmosphere containing 95% air and 5% CO2 (v/v). CLL cells (106 per mL) were treated with either CDKI-71 or flavopiridol for 48 hr. Subsequently, cells were labeled with CD19-APC (Caltag) and then resuspended in 200 μL of binding buffer containing 4 μL of annexin V–FITC (Bender Medsystems, Vienna, Austria). Apoptosis was quantified in the CD19+ CLL cells, CD19+ normal B-cells and CD3+ normal T-cells using a FACSCalibur flow cytometer and CellQuest software (BD Biosciences, Oxford, UK). LD50 values were calculated from line-of-best-fit analysis of the sigmoidal dose–response curves.
Western blotting was performed as described.6 Antibodies used were as follows: total RNAP-II (8WG16), phosphorylated RNAP-II Ser-2 and Ser-5 (Covance, Leeds, UK), Bcl-2 and p53 (Dako, Denmark A/S), MDM-2 and β-actin (Sigma-Aldrich), Mcl-1, PARP (Poly (ADP-ribose) polymerase), PP1α, Phospho-PP1α, total Rb (Cell Signalling Technologies, Hitchin, UK), p21 (Santa Cruz Biotechnology, Wembley, UK), Anti-RbT821 (Invitrogen, Paisley, UK), antiphosphohistone H2AX (Millipore, Dundee, UK). Both anti-mouse and anti-rabbit immunoglobulin G (IgG) horseradish peroxidase-conjugated antibodies were obtained from Dako.
The assay was performed using the Trevigen Comet Assay kit under alkaline conditions, following the manufacturer's instructions. Nikon Eclipse TS100 fluorescence microscopy was used to record images and analysis was performed using the Comet Assay IV program. Typically 50 cells were analyzed per slide with at least two slides for each treatment.
All experiments were performed in triplicate and repeated at least twice; representative experiments being selected for figures. Statistical significance of differences for experiments was determined using one-way analysis of variance, with a minimal level of significance at p ≤ 0.05.
CDKI-71 is a potent CDK9 inhibitor and an antiproliferative agent
The CDK inhibitory activity of CDKI-71 and flavopiridol were determined by biochemical assays (Table 1). Both flavopiridol and CDKI-71 showed similar activity against CDKs 9, 7, 1 and 6, with the highest potency being against CDK9 (Ki = 3 and 6 nM for flavopiridol and CDKI-71, respectively). However, CDKI-71 is a more potent CDK2 inhibitor (Ki = 4 nM) than flavopiridol (Ki = 79 nM). The antiproliferative effect of CDKI-71 on a panel of 12 tumor cell lines and two normal human diploid fibroblast cell lines were compared to flavopiridol using 48 hr MTT assay, and half-maximal growth inhibition (GI50) values are summarized in Table 2. MRC-5 and WI-38 cells were chosen because they are well characterized normal diploid cell lines and the most commonly used noncancerous cell lines.18 To investigate cell-type sensitivity to CDKI-71 and flavopiridol, we included HCT-116 colon carcinoma (wild-type and null p53), MCF-7 breast carcinoma (wild-type p53, pRb positive and containing cyclin D-CDK 4/6) and MDA-MB-468 breast carcinoma cells (mutant p53, pRb negative, lacking cyclin D-CDK 4/6).19 Similar sensitivity was observed for cells with different p53, Rb and CDK4/6 status for both compounds. CDKI-71 suppressed tumor cell proliferation with GI50 values ranging from 287 to 768 nM, irrespective of cell type, but was 10-fold less potent than flavopiridol in the same cell lines (GI50 = 21–81 nM). However, MRC-5 and WI-38 cells (not shown) were significantly less sensitive to CDKI-71 with GI50 = 2265–4277 nM (p ≤ 0.01). In contrast, both cell lines were sensitive to flavopiridol (GI50 = 49–77 nM). The time-course assays were performed using HCT-116, A2780 and MRC-5, and as shown in Table 3, treatment for 24 hr was sufficient to achieve the most growth inhibition for both compounds.
CDKI-71 induces apoptosis in a caspase-dependent manner
We further investigated whether the antiproliferative effect of CDKI-71 was a consequence of activation of the cellular apoptosis programme. Induction of caspase 3/7 activity was measured in HCT-116, A2780 or MRC-5 untransformed cells after treatment with either CDKI-71 or flavopiridol following exposure to drug for 24 hr (Figs. 1a and 1b). CDKI-71 significantly induced caspase 3/7 activity in both HCT-116 and A2780 cells at 0.5 μM, but no such activity was detected in MRC-5 cells below 5 μM (Fig. 1a). In contrast, flavopiridol activated caspase 3/7 at 0.1 μM in the cancer cell lines and at 0.2 μM in the MRC-5 cell line (Fig. 1b).
As CDKI-71 has been shown to inhibit CDKs 1, 2 and 9 in biochemical kinase assays, we next investigated its effects on the cell cycle. HCT-116 cells were treated with CDKI-71 (or flavopiridol) for a period of 24 hr (Fig. 2a). Treatment with 2 μM (4 × GI50 by MTT) CDKI-71 was shown to cause accumulation of cells in G2/M of the cell cycle and the same cell-cycle profile was observed with 0.1 μM (∼ 2 × GI50) flavopiridol.
Induction of apoptosis was analyzed by annexinV/PI double staining in HCT-116 cells (Fig. 2b), A2780 cells (data not shown) and MRC-5 cells (Fig. 2c). CDKI-71 induced apoptosis starting from 0.5 μM GI50 from 24 hr onward with maximal effect after 48 hr of treatment, which was consistent with the observed caspase 3/7 activity and growth inhibition. Like CDKI-71, flavopiridol also induced apoptosis in a dose- and time-dependent manner, but starting from 0.1 μM (no effect being observed at 0.05 μM, the concentration causing cytotoxicity by MTT assay). When cells were treated with 2 μM CDKI-71 and 50 μM Z-VAD-fmk, a pan-caspase inhibitor, apoptosis was suppressed suggesting a caspase-dependent mechanism of apoptosis induction for CDKI-71 similar to that observed after exposure to flavopiridol. CDKI-71-induced apoptosis is not the consequence of cell-cycle arrest, because CDKI-71 only caused G2/M arrest at the higher concentrations, that is, ≥2.0 μM (Fig. 2a). MRC-5 cells were less sensitive to CDKI-71 treatment compared to the cancer cell lines, and apoptotic cells were detected on treatment with 5 μM CDKI-71 (Fig. 2c). In contrast, flavopiridol induced apoptosis in MRC-5 cells after exposure to concentrations from 0.2 μM.
The potency and selectivity of CDKI-71 was further tested in patient-derived CLL cells, as well as healthy normal B- and T-cells, by annexin V–FITC apoptosis assay. As shown in Figure 2d, the compound exhibited excellent activity with a LD50 = 0.43 μM against CLL cells (48 hr treatment), while little toxicity was observed in the normal B- and T-cells (LD50 > 500 and > 700 μM, respectively). In contrast, flavopiridol showed little selectivity with LD50 values of 0.34, 0.59 and 0.81 μM with respect to CLL B-cells, normal B- and T-cells, respectively (Fig. 2e).
CDKI-71 inhibits phosphorylation of RNAPII CTD and downregulates the level of Mcl-1
Western blot analysis of HCT-116 cells after treatment with CDKI-71 or flavopiridol for a period of 24 hr showed that phosphorylation at Ser-2 of RNAPII CTD was significantly reduced at 0.5 μM CDKI-71, indicating cellular CDK9 inhibition (Fig. 3a). The level of phosphorylated Ser-5 was not affected unless the cells were treated with ≥2.5 μM CDKI-71, confirming the relative selectivity for CDK9 over CDK7. However, flavopiridol decreased the phosphorylation of both Ser-2 and Ser-5 at 0.2 μM, suggesting similar potency for both CDK7 and CDK9, despite its apparent selectivity for CDK9 over CDK7 in the kinase assays. We further investigated the phosphorylation status of substrates specific for cyclin A–CDK2 and cyclin B–CDK1. Reductions in the levels of phosphorylated pRbT821, and to a lesser extent, CDK1 substrate PP1 (p-PP1α) were caused by CDKI-71 at 1.0 and 2.5 μM, respectively, confirming the lower inhibitory activity against these enzymes compared to that of CDK9. In contrast, flavopiridol suppressed cellular CDK2 and CDK1 at the concentrations of 0.2 μM (Fig. 3a) once again pointing to the relatively less selective inhibitory activity of this drug.
Similar to HCT-116 cells, treatment of A2780 ovarian cancer cells with CDKI-71 at 0.5 μM for 24 hr resulted in a significant reduction in the phosphorylation of Ser-2, but with a minimal effect on the phosphorylation of Ser-5 (Fig. 3b). The same treatment caused downregulation of Mcl-1 antiapoptotic protein, but had little effect on XIAP or Bcl-2. Induction of apoptosis was indicated by PARP cleavage (0.5 μM). Analogous results were obtained with flavopiridol: inhibition of the phosphorylation of Ser-2 and downregulation of Mcl-1 and MDM2 after exposure to 0.2 μM. Flavopiridol also inhibited the phosphorylation of Ser-5 but had little effect on the expression of Bcl-2 at this time point. The levels of p53 and p21 increased at the concentrations causing MDM2 reduction for each compound, although decreased p21 was observed at 0.5 μM flavopiridol. A2780 cells were also treated with CDKI-71 or flavopiridol for a short period of 6 hr (Fig. 3c) with the levels of pSer-2 and Mcl-1 being reduced by the same concentration of CDKI-71 and flavopiridol.
MRC-5 fibroblast cells were treated with CDKI-71, flavopiridol or cisplatin for 24 hr (Fig. 3d). The levels of phosphorylated Ser-2 and Ser-5 were unaffected by CDKI-71 at concentrations <5 μM. This was also true for the levels of Mcl-1, while Bcl-2 was not affected by any of the concentrations tested. Similarly, induction of apoptosis, as evidenced by PARP cleavage and annexin V/PI assays, was only observed with 5 μM CDKI-71. These data confirmed that the cancer cells were more sensitive (10-fold) than the untransformed cells toward CDKI-71. The relative insensitivity of MRC-5 cells to CDKI-71-mediated dephosphorylation of Ser-2 appears to be critical but the reasons for this remain unresolved. In contrast, flavopiridol failed to demonstrate any differential effects; treatment with 0.2 μM flavopiridol resulted in reduced phosphorylation of Ser-2 and Ser-5 as well as a reduction in Mcl-1 expression.
Flavopiridol induces DNA double-strand breaks
Given that the kinase inhibition profile of CDKI-71 and flavopiridol were similar but their cytotoxicity profiles were not, it seemed likely their mechanism of cell killing was different. As flavopiridol showed remarkably little selectivity towards cancer cells when compared to normal cells we suspected that DNA damage may be a consequence of treatment with this drug. Consequently, we investigated the effects of flavopiridol and CDKI-71 on the expression of γ-H2AX (Figs. 4a and 4b). H2AX is a variant of histone H2A, which is required to maintain genomic stability.20 Phosphorylated H2AX, that is, γ-H2AX, has been shown to be a sensitive marker of DNA double strand breaks induced by DNA-damaging agents.21, 22 Western blot analysis of A2780 cells after treatment with ≥0.05 μM flavopiridol or 5 μM cisplatin showed elevated γ-H2AX, but 2.0 μM CDKI-71 had negligible effect (Fig. 4a). In MRC-5 cells γ-H2AX was induced after treatment with flavopiridol at 0.2 μM, while 5 μM CDKI-71 caused no significant differences in the level of γ-H2AX compared to DMSO treatment (Fig. 4b). To further confirm DNA damage in individual cells, and to estimate the damage distribution in a population of cells, we performed alkaline comet assays (Fig. 4c).23, 24 MRC-5 cells were treated with 5 μM of CDKI-71 or 0.2 μM of flavopiridol for a period of 6 hr. Significant numbers of cells with DNA double-strand breaks were observed, as evidenced by the formation of the comet, enlarged tail and increased tail intensity, after treatment with flavopiridol. In contrast, no detectable comet cells were observed after CDKI-71 or control (DMSO vehicle) treatment.
CDK inhibitors are being developed as potential cancer therapeutics based on the promise that they may counteract the unchecked proliferation of cancer cells by targeting the cell-cycle regulatory functions of CDKs. In recent years, the conventional understanding of how CDKs function and their roles in the cell cycle has been challenged.2, 25 The observation that cancer cell lines and some embryonic fibroblasts lacking CDK2 proliferate normally, and that knockout mice are viable,4, 26 suggests that CDK2 performs a nonessential role in cell cycle control. Furthermore, redundancy of CDKs 4 and 6 was also suggested in mammalian cells that enter the cell cycle normally.27 It has been demonstrated that mouse embryos deficient in all of the interphase CDKs develop to midgestation as CDK1 can form complexes with all cyclins and consequently phosphorylates Rb protein. This in turn activates E2F-mediated transcription of proliferation factors.28 Elimination of CDKs 2, 3, 4 and 6 in these investigations showed that expression levels of CDK1, as well as its cyclins, appeared to be unchanged and that catalytic activity of CDK1 was sufficient for cell-cycle progression. However, cyclin B–CDK2 is readily detected in cells deleted for CDK1 and this can facilitate G2/M progression.25 Therefore, these studies suggest that specifically targeting individual cell-cycle CDKs may not be an optimal therapeutic strategy.
Inhibition of transcriptional CDKs has become the preferred anti-cancer strategy because the most sensitive transcripts, those with short half-lives, encode cell cycle regulators, mitotic regulatory kinases and apoptosis regulators such as Mcl-1.29, 30 Although CDKs 1, 2, 7, 8, 9 and 11 have all been implicated in phosphorylation of CTD RNAP-II,31 the most important are CDKs 7 and 9.32 It has been shown that combined depletion or inhibition of CDKs 1, 2 and 9 results in effective induction of cellular apoptosis through both E2F- and RNAPII CTD-mediated effects.25, 33
In our study, we report the first cellular investigation of a novel CDK inhibitor CDKI-71 and compare its effects with flavopiridol, a clinically active CDK inhibitor. CDKI-71 inhibited CDK9 with low nM Ki value and exhibited a broad spectrum of antitumor activity with an average GI50 value of 0.5 μM. Furthermore, it showed no differential cytotoxicity toward cell lines on the basis of p53, p21, p16 or pRb status. However, it was 5–10-fold less potent in normal diploid lung fibroblast MRC-5 cells compared to the tumor cell lines examined. Significantly, CDKI-71 exhibited excellent anti-CLL activity but showed little toxicity toward normal B- and T-cells (Fig. 2d). This was in marked contrast to flavopiridol, which appeared to have little selectivity toward CLL cells when compared to normal B- and T-cells (Fig. 2e). Despite similar Ki values for CDK inhibition, flavopiridol showed ∼ 10-fold greater growth inhibitory potency than CDKI-71 in the same panel of tumor cell lines. However, unlike CDKI-71, flavopiridol proved equally toxic toward MRC-5 cells (Tables 2 and 3). This raises the possibility that another mechanism, in addition to CDK inhibition, might contribute to flavopiridol-induced cytotoxicity.
Both CDKI-71 and flavopiridol exhibited similar cell cycle profiles in HCT-116 cells, inducing cell arrest at G2/M, but the CDKI-71 cell-cycle effects were only observed after treatment with ≥2 μM (∼ 4 × GI50; Fig. 2a), indicating that cell-cycle disturbance may not be the primary contributor to its cytotoxic potency. The fact that 0.5 μM CDKI-71 (GI50) effectively induced apoptosis in cancer cells, pointed to a cell cycle-independent mechanism of cell death induction. CDK9 regulates RNAPII transcriptional activity and plays no role in cell-cycle regulation as evidenced by CDK9 siRNA experiments that showed no effect on cell-cycle distribution.25 Inhibition of RNAPII-mediated transcription by blocking CDK9 activity was confirmed as a primary molecular target for CDKI-71 by Western blot analysis. CDKI-71 was shown to inhibit the RNAPII CTD Ser-2 phosphorylation at 0.5 μM (Figs. 3a and 3b), the concentration used to elicit a cytotoxic GI50 response, as well as to induce apoptosis in HCT-116 and A2780 cells (Figs. 1a and 2b). The same treatment resulted in the downregulation of Mcl-1 and MDM2 and PARP cleavage, but the levels of phosphorylated Ser-5, pRbT821 and PP1α were unaffected; further supporting the hypothesis that cellular CDK9 inhibition, leading to the downregulation of Mcl-1, is the primary mechanism of action of CDKI-71. Similarly, flavopiridol (at 0.2 μM) inhibited RNAPII transcription leading to reduction of Mcl-1 and MDM2 and induction of apoptosis in cancer cells.
Cancer cells often have a high demand for transcription and translation of antiapoptotic proteins to resist programmed cell death. Many of these proteins have short half-lives at both the mRNA and protein levels.34 In our study, we showed that targeting the CTD Ser-2 of RNAP-II by CDK9 inhibition is sufficient to inhibit transcription and translation of the encoded proteins. Mcl-1, but not Bcl-2 protein, was found to be particularly sensitive to both CDKI-71 and flavopiridol treatments. These findings are consistent with previous studies of flavopiridol and another CDK inhibitor, SNS-032, in which Mcl-1 protein decreased in cancer cells after treatment, but the expression of Bcl-2 remained unchanged.6, 7 The specific motifs of mRNA or protein sequences signaling for degradation are responsible for the different responses of Mcl-1 and Bcl-2.35–37 The susceptibility of Mcl-1 to rapid intracellular proteolysis results in its extremely short half life (t1/2 = 0.5–1 hr) compared to Bcl-2, which is a relatively long-lived protein with t1/2 = 10–24 hr. Increased Mcl-1 protein levels have been reported in a number of tumor samples, and as tumor cells appear to be dependent on Mcl-1 for survival, it is an excellent target for cancer therapy.38
Apoptotic cell death can be initiated by an intrinsic pathway involving mitochondrial damage that is regulated by Bcl-2 family proteins and an extrinsic pathway activated by the interaction of death receptors with their ligands. Once triggered by either pathway, apoptosis is executed via a cascade of activated caspases.39 As part of the apoptotic process, caspases undergo proteolytic cleavage resulting in loss of a prodomain and subsequent conformational change to an activated form. Caspases 3/7 are the most crucial effector components of cell death pathways,40 and the cleavage of PARP is a hallmark of caspase 3-mediated apoptosis. We demonstrated that both CDKI-71 and flavopiridol were capable of activating caspase 3/7 while the pan-caspase inhibitor Z-VAD-fmk could prevent their activation. Moreover, during the apoptotic process, mitochondrial cytochrome C is released in response to proapoptotic stimuli and this in turn activates caspase 9, followed by caspases 3, 6 and 7.41 Mcl-1 can prevent mitochondrial outer membrane permeabilization and cytochrome C release;42, 43 downregulation of Mcl-1 by CDKI-71 and flavopiridol may therefore facilitate release of cytochrome C leading to induction of apoptosis.
Induction of p53 was observed on treatment with CDKI-71 at cytotoxic concentrations of 0.5 and 5 μM in A2780 and MRC-5 cells, respectively. Expression of p53 is tightly regulated at the post-translational level by association with its negative regulator MDM2, a short half-life protein which is also reduced by CDKI-71 and flavopiridol (Fig. 3). At the same concentrations, CDKI-71 decreased the level of MDM2 and phosphorylation of Ser-2 RNAPII in A2780 and MRC-5 cells. Upregulation of p21 in a dose-dependent manner was also detected in CDKI-71-treated A2780 cells. This is consistent with the result of a previous study demonstrating that p21 was induced by DRB, the most selective CDK9 inhibitor identified to date.44 The absence of differential sensitivity between p53 wild type, p53 null, and p53 mutant cells in our cytotoxicity assays suggested that p53 is not essential for CDKI-71-induced apoptosis but rather its increase is a molecular consequence of CDK9 inhibition-mediated MDM2 downregulation.
In MRC-5 diploid cells, the mechanisms controlling the cell cycle and apoptosis are fully intact and therefore suitable to identify the targets that are relevant for the cytotoxic mechanism of action of an inhibitor. Although CDKI-71 induced apoptosis in cancer cells at 0.5 μM, MRC-5 cells were insensitive to the cytotoxic effects of CDKI-71 at ≥5 μM (Figs. 1a and 2c). In contrast, MRC-5 cells were sensitive to flavopiridol (Figs. 1b and 2c). Consistent with the cellular cytotoxicity, 5 μM CDKI-71 resulted in the reduction of RNAPII CTD Ser-2 phosphorylation, downregulation of Mcl-1 and PARP cleavage. Our study demonstrates that although the primary dephosphorylation of RNAPII and downregulation of antiapoptotic proteins is consistent across the transformed and untransformed cell lines, the ultimate fate of the cell is governed by its relative dependence on antiapoptotic proteins for survival. The requirement for continuous antiapoptotic protein production in cancer cells can result in them being more sensitive to targeting RNAPII transcription with CDKI-71. In this regard, CDKI-71 was found to be extremely cytotoxic against patient-derived primary CLL B-cells, while it spared normal healthy B- and T-cells (Figs. 2d and 2e).
Despite MRC-5 cells being relatively resistant to CDKI-71 treatment, they showed a marked cytotoxic response to flavopiridol. Flavopiridol displayed potent antiproliferative activity against a panel of tumor cell lines with a mean GI50 value of 0.05 μM, irrespective of cell type. However, the growth inhibitory activity did not correlate with either apoptosis or cellular CDK9 inhibitory potency. Dephosphorylation of RNAPII at Ser-2 and downregulation of Mcl-1 in A2780 and MRC-5 cells were only observed after treatment with ≥0.2 μM (i.e., 4 × GI50) of flavopiridol (Fig. 3). Furthermore, the induction of apoptosis, as detected by caspase 3/7 activation, annexin V/PI-positive cells and PARP cleavage, occurred after treatment with 0.1 μM flavopiridol, suggesting that CDK9 is not the only target for this drug.
To investigate whether DNA may be a target for flavopiridol and to confirm that CDKI-71-induced p53 protein is not a consequence of a cellular DNA damage response, we measured the level of γ-H2AX, expression of which is indicative of DNA double-strand breaks.45 After DNA damage, Ser-139 in the unique carboxy-terminal tail of H2AX is phosphorylated and the level of γ-H2AX increases in a linear fashion with the severity of damage. After treatment of A2780 or MRC-5 cells with various concentrations up to 2 or 5 μM of CDKI-71, respectively, the presence of γ-H2AX remained undetectable (Figs. 4a and 4b). In contrast, flavopiridol elevated levels of γ-H2AX at concentrations above 0.05 μM in A2780 and 0.2 μM in MRC-5 cells; at the same concentrations flavopiridol caused apoptosis in the respective cell lines. Comet assays revealed a significant number of cells with DNA double-strand breaks in the 0.2 μM flavopiridol-treated MRC cells, but little effect was observed on the cells treated with 5 μM CDKI-71. These findings suggest that induction of DNA damage may be a part of the cytotoxic mechanism of action of flavopiridol. These direct genotoxic effects may also explain limited selectivity of flavopiridol in cancer cells when compared to untransformed and normal cells. Importantly, these effects were not seen with CDKI-71.
Flavopiridol has attracted much interest in the past decade, not only because of its novel cellular mechanism but also because it causes apoptosis in cancer cells and shows antitumor efficacy in clinical trials.46 There is evidence that flavopiridol interacts with DNA in vitro.47 The equilibrium dissociation constant of the flavopiridol–DNA complex was found to lie in the range observed for binding of doxorubicin and pyrazoloacridine to DNA. Analysis of the pattern of flavopiridol-induced cytotoxicity in the NCI-tumor cell line panel using the COMPARE algorithm suggested a high correlation with those of intercalating or DNA-damaging agents.47 However, here we show for the first time that flavopiridol induces DNA double-strand breaks in cancer cells and fibroblasts. Another intriguing property of flavopiridol is its higher cellular potency against RNAPII phosphorylation via CDK9 and CDK7 when compared to CDKI-71, despite their similar in vitro kinase activites. This may arise from the ability of flavopiridol to intercalate into the DNA double strands, which may serve as a template to facilitate the association of flavopiridol with the RNAPII CTD and therefore inhibit CDKs 9 and 7. Interestingly, a lack of competition of ATP with flavopiridol's action on p-TEFb has been noted and is thought to be because of the extremely tight binding of flavopiridol to CDK9 compared to other CDKs. For example, in a mixture of kinases, flavopiridol associates preferentially with cyclin T-CDK9 rather than other CDKs.48 However, X-ray crystallography provides evidence that flavopiridol binds to the ATP-binding site of CDK9 in the same way as it does to CDK2.49, 50 The Ki for flavopiridol of 3 nM may be underestimated. In our biochemical assay, it is four orders of magnitude below the Km for ATP, and had a lower concentration of CDK9 been used, a still lower Ki might have been determined. The study of immobilized p-TEFb/flavopiridol complexes suggested that the dissociation constant for this drug is significantly below 1 nM.48 These considerations may explain the discrepancy of the cellular potency in RNAPII phosphorylation between flavopiridol and CDKI-71.
In summary, our study demonstrates that by inhibiting cellular RNAPII transcriptional activityCDKI-71 mediates downregulation of antiapoptotic proteins and thereby renders cells sensitive to apoptosis. The apoptosis is caspase-dependent and p53-independent. Like flavopiridol, CDKI-71 causes significant reductions in the levels of the antiapoptotic protein Mcl-1 and MDM-2 in cancer cells. However, unlike flavopiridol, it does not appear to target DNA and hence is less cytotoxic to normal cells. This presumably contributes to its ability to selectively induce apoptosis in primary cancer cells and cancer cell lines. Based on these findings CDKI-71 has the potential to be developed as an anti-cancer agent with a better therapeutic window compared to flavopiridol.
- 162009., , , . Pyrimidines, triazines and their use as pharmaceutical agents, WO/2009/118567,
- 30Genomic-scale measurement of mRNA turnover and the mechanisms of action of the anti-cancer drug flavopiridol. Genome Biol 2001; 2:research0041.1–0041.11., , , , , , , , , , , , et al.