Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Oncogenic KRAS-associated gene signature defines co-targeting of CDK4/6 and MEK as a viable therapeutic strategy in colorectal cancer

Oncogene volume 36, pages 4975–4986 (2017)Cite this article

Subjects

Abstract

Therapeutic strategies against KRAS mutant colorectal cancers are developed using cell line models, which do not accurately represent the transcriptome driven by oncogenic KRAS in tumors. We sought to identify a KRAS-associated gene signature from colorectal tumors to develop a precise treatment strategy. Integrative analysis of quantitative KRAS mutation detection and matched gene expression profiling in 55 CRC bulk tumors was carried out to define a gene signature enriched in CRC tumors with high KRAS mutation. The KRAS-associated gene signature identified exhibits functional enrichment in cell cycle and mitosis processes, and includes mitotic transcription factor, FOXM1. Combination treatment of CDK4/6 inhibitor Palbociclib and MEK inhibitor PD0325901 was tested in KRAS-mutant, BRAF-mutant CRC, normal colon epithelial lines and xenografts models to determine their efficacy and toxicity and to monitor the changes in the gene signature. Inhibiting CDK4/6, an upstream regulator of FOXM1, and MEK synergistically depleted FOXM1 and KRAS-associated gene signature, suggesting that CDK4/6 and MEK regulate the KRAS gene signature. The combined inhibition of CDK4/6 and MEK elicited a robust therapeutic response in KRAS-dependent and BRAF-mutant CRC, both in vitro and in vivo and this correlated with downregulation of the KRAS-associated gene signature. Our preclinical study demonstrated the efficacy of Palbociclib and PD0325901 combinatorial treatment selectively in KRAS-dependent and BRAF-mutant CRC but not in normal colon epithelial cells. The KRAS-associated gene signature could facilitate the identification of responsive metastatic CRC to this therapeutic strategy in clinical settings.

Download PDF

Introduction

KRAS mutation is found in about 50% of all colorectal cancers (CRC), and it is one of the main and earliest drivers of tumorigenesis in the progression of the disease.1, 2 Despite tremendous efforts in developing KRAS-related therapies, chemotherapy remains the key treatment for KRAS-mutant metastatic CRC, while the treatment of KRAS wild type CRC has advanced with the use of cetuximab, a monoclonal antibody against the epidermal growth factor receptor.3 Even then, recent report has also attributed the emergence of KRAS mutations as the cause of cetuximab resistance, either intrinsic or acquired, rendering previously cetuximab-sensitive cancer cells independent of EGFR signaling and leading to relapse in patients.4 This unmet clinical need highlights an urgent need to develop more robust therapeutic approaches to target KRAS-mutant CRC.

Mutated KRAS constitutively activates multiple downstream pathways, including the MEK/ERK (MAPK-extracellular signal-regulated kinase) and PI3K/AKT signaling pathways.5 Previous studies, however, have showed that the presence of mutated KRAS did not always correlate with high levels of phosphorylated ERK and is therefore an unreliable biomarker to predict therapeutic response to MEK inhibitor treatment.6 Development of treatment strategies through the use of high throughout RNA interference screenings to identify and target synthetic lethal genes specifically in KRAS-mutant cell line model,7 or in isogenic cells with differing KRAS mutation status8 had not yet resulted in any treatment strategies translated into clinical use. Plausible reasons could be the lack of optimal stratification of patients predicted to be treatment-responsive, because of inter/intratumoral heterogeneity9, 10 which cannot be recapitulated in cell line models. There is a need to identify clinically relevant molecular biomarkers in KRAS-driven CRC to define a more precise treatment strategy.

We sought to address this gap by leveraging on pyrosequencing to quantify KRAS mutation and associated gene expression in individual CRC tumors to define molecular determinants of KRAS dependency. We demonstrated that combinatorial treatment of Palbociclib, an recently approved inhibitor targeting cyclin-dependent kinase 4/6 (CDK4/6), and PD0325901, a MEK inhibitor currently in clinical trials, robustly eliminated the KRAS dependency gene signature and effectively targeted KRAS-dependent CRC both in vitro and in vivo.

Results

KRAS-associated gene signature reveals CDK4/6-FOXM1 node as a potential molecular determinant of MEK inhibitor sensitivity

CRC tumors are often heterogeneous, with subclones driven by different oncogenic drivers and this poses an obstacle in identifying a specific oncogene-driven gene expression. Here, to identify a KRAS dependent CRC gene expression, we performed pyrosequencing of KRAS at codon 12 and 13 and Illumina expression array analysis in 55 CRC tumors. On the basis of the percentage of KRAS mutation detected in the reads, we stratified the tumors into KRAS wildtype (<10%), low KRAS mutation (10–40%) and high KRAS mutation tumors (>40%) (Figure 1a and Supplementary Table S1). We propose that in high KRAS mutation tumors, KRAS mutated cells formed the dominant clones in the bulk tumors, whereas KRAS wildtype or low KRAS mutation tumors consist mainly or more KRAS wildtype cells. Supervised gene cluster analysis of these tumors identified 97 genes that were significantly correlated (P-value<0.05) with KRAS mutation and 55 met the cutoff of Pearson correlation of ⩾0.3 or ⩽−0.3 (Supplementary Table S2), including 34 and 22 genes upregulated and downregulated respectively in the high KRAS mutation tumors (Figures 1a and b).

Figure 1
figure 1

Gene signature associated with KRAS mutation identifies CDK4/6-FOXM1 node as a molecular determinant of MEK inhibitor sensitivity. (a) Percentage of KRAS codon 12 and 13 mutated in the total number of reads in 55 colorectal patients tumors detected via pyrosequencing. Tumors were stratified into KRAS wild type (reads containing KRAS mutation <10%), Low KRAS mutation (10–40%), High KRAS mutation (⩾40%). (b) Heatmap showing expression of the 56 genes that were significantly correlated with KRAS mutation rate (cutoff using correlation ⩽−0.3 or ⩾0.3 with P value<0.05). Colored scale bar represents log2 transformed values. (c) No significant differences were observed in the three groups regarding the expression of gene markers for the different cell types in the tumor microenvironment:CDH1, EPCAM and KRT20 for epithelial; CD12, CD19 and PTPRC for leukocytes; CDH5, ENG and VWF for endothelial; and DCN, PDPN and FAP for fibroblasts. (one way ANOVA test, P-value>0.05). Colored scale bar represents log2 transformed values. (d) PD0325901 EC50for colorectal cancer cell lines KRAS mutant and dependent DLD1 and HCT116, BRAF mutant RKO and KRAS independent HCT15 with and without the addition of 1 μM Palbociclib. (e) PD0325901 EC50 for normal epithelial cells with and without 1 μM Palbociclib. Two way ANOVA was used for the statistical analysis of interaction between the EC50 curves with and without the presence of 1 μM Palbociclib. **P<0.01, ****P<0.0001, n.s., not significant

Full size image

Interestingly, the upregulated genes showed significant enrichment in pathways related to cell cycle and mitosis with FOXM1, the master transcription regulator of mitosis,11, 12 present in the signature (Figure 1b and Supplementary Figure S1B). This observation led us to hypothesize that CDK4/6 pathway, being an important regulator of cell cycle transcription factors activator FOXM1 and E2Fs,13, 14, 15, 16 could be a co-driver to the MAPK pathway in promoting the growth of KRAS-mutant CRC. Of note, CRC tumors are rarely presented with downstream RB1 mutations (Supplementary Figure S1C, Supplementary Table S3),17, 18 an indication that inhibiting CDK4/6 activity is expected to elicit a response in CRC tumors.

To exclude the possibility that the differential gene expression seen in the three group was because of different level of stroma content in the tumor sampled, we checked for the expression of markers that were previously used to identify the various stroma cell types19 (Figure 1c). None of the markers were found to be significantly different in any groups (Supplementary Figure S1A), strengthening the relevance of this KRAS gene signature to the CRC tumor cells.

Next, we evaluated if inhibition of CDK4/6 could sensitize KRAS-mutant CRC cell lines to MEK inhibition by comparing the dose response of MEK inhibitor PD0325901 in CRC cell lines in the presence or absence of Palbociclib, a CDK4/6 inhibitor. Addition of CDK4/6 inhibitor markedly sensitized two KRAS-mutant lines, DLD1 and HCT116, to MEK inhibitor treatment, resulting in 7.69- and 11.9-fold reduction of their PD0325901EC50 respectively (Figure 1d). A similar result was seen in RKO, a BRAFV600E-mutant CRC line (5.48-fold reduction; Figure 1d), suggesting that CDK4/6 activity may also be important in BRAF-mutant tumors. In comparison to HCT15, a KRAS-mutant but previously identified as a KRAS-independent cell line7 (validated in Figure 5c), Palbociclib insignificantly sensitized these cells to PD0325901 (1.95-fold reduction in EC50). In contrast, when we examined the effect of this combination treatment on normal colon epithelial cells, we found that Palbociclib co-treatment even reduced the sensitivity of these cells to PD0325901 (Figure 1e). Thus, we conclude that CDK4/6 inhibitor is able to selectively increase the sensitivity of KRAS/BRAF-mutant colon cancer cells to MEK inhibitor but yet reduce the cytotoxic effect of MEK inhibitor in normal colon epithelial cells.

Characterization of Palbociclib and PD0325901 combination effects on oncogenic growth and apoptosis induction in KRAS-dependent and BRAF-mutant CRC

To extend the analysis, we further characterized the effects of this combination treatment on various oncogenic growth properties of CRC. In a time course analysis of cell proliferation up to 9 days, combination of Palbociclib and PD0325901 led to a drastic repression of cell proliferation as compared to single inhibitor treatments in KRAS-dependent (Figure 2a) and BRAF-mutant CRC (Figure 2b) but not in KRAS-independent HCT15 (Figure 2c) and normal colon epithelial cells (Figure 2d). Combinatorial treatment using alternative CDK4/6 and MEK inhibitors, Ribociclib and Trametinib also demonstrated similar further proliferation repression in DLD1 and HCT116 (Supplementary Figure S2A). This highlights again the selective advantage of this combination strategy towards KRAS-dependent/BRAF-mutant CRC but not the normal cells. PD0325901 was used at a much lower concentration in HCT116 and HT29 as they display greater sensitivity to MEK inhibitor (Figure 1d, data not shown for HT29). Further analysis using 2D monolayer clonogenic and 3D anchorage-independent soft agar growth assay showed that the combination Palbociclib and PD0325901 almost eradicated colony formation in both growth conditions in KRAS-dependent and BRAF-mutant CRC lines when compared with single inhibitor treatments (Figure 2e). Moreover, this combination treatment was able to elicit sufficient apoptosis in KRAS-dependent and BRAF-mutant cells but not in KRAS-independent HCT15 or normal non-transformed cell, by assessment of proportions of cells in Sub-G1 phase of Propidium iodide-stained cells and cleaved PARP expression (Figure 2f, Supplementary Figure S2B).

Figure 2
figure 2

Reduction in cell viability in KRAS dependent and BRAF-mutant colorectal cancer treated with combined Palbociclib/PD0325901. (a–d) Cell proliferation assay in colorectal cancer cell lines treated with Palbociclib and PD0325901 either alone or in combination with top up of fresh media and drug on day 5. Cell viability is represented as relative to Day 1 control Cell Titer Glo (CTG) signal. Concentrations used are as follow: (a)DLD1 (1 μM Palb, 2 μM PD); HCT116 (1 μM Palb, 0.1 μM PD). (b) RKO (1 μM Palb, 1 μM PD); HT-29 (1 μM Palb, 0.05 μM PD). (c) HCT15 (1 μM Palb, 1 μM PD). (d) FHs 74 Int and CCD 841 CoN (2 μM Palb, 2 μM PD). (e) Representative images of monolayer foci formation assay of colorectal cancer lines after 12 days of treatment with Palbociclib and PD0325901 either alone or in combination. Bar graphs: Quantification of number of colonies grown in anchorage independent soft agar after 12 days with the same treatment groups. Concentrations used are as stated. (f) Percentage of apoptotic cells after 7 days treatment with the same treatment group as quantified by fluorescence-activated cell sorting (FACS) analysis through assessing sub-G1 DNA. Concentrations used are as follows: DLD1 (2 μM Palb, 2 μM PD); HCT116 (2 μM Palb, 0.1 μM PD); RKO (1 μM Palb, 1 μM PD); HCT15 (1 μM Palb, 5 μM PD); FHs 74 Int (2 μM Palb, 2 μM PD); CCD 841 CoN (2 μM Palb, 2 μM PD). (g) Cell proliferation assay in DLD1 KRAS isogenic lines treated with 2 μM Palbociclib and 2 μM PD0325901 either alone or in combination with same duration as in a–d. Cell viability is represented as CTG signals relative to respective day 5 signals and log2 transformation was performed. All data in the graphs represent mean±s.e.m., n=3. n.s., not significant. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001. P-values calculated with paired T-test.

Full size image

To demonstrate whether combination treatment is dependent on KRAS, we utilized DLD1 KRAS isogenic lines, a KRAS wildtype only (KRAS WT), which displayed reduced growth, and a KRAS-mutant only DLD1 (KRAS G13D) (Supplementary Figure S2B). Although the three cell lines did not show differential response to the Palbociclib/PD0325901 combination in the first 5 days (Supplementary Figure S2C), KRAS mutant DLD1 showed much reduced cell viability by day 7 and 11 as compared to KRAS wildtype DLD1 (Figure 2g). This data provided direct evidence that KRAS dependency confers sensitivity to combined Palbociclib and PD0325901 treatment in CRC.

Inhibition of CDK4/6 and MEK converge to downregulate KRAS-associated gene signature and cell cycle related transcription factors

To identify possible crucial mediators in the combinatorial inhibition of CDK4/6 and MEK, we performed a gene expression analysis after single or combination treatment in DLD1 cells at 24, 48 and 72 h and identified 1805 genes that showed differential expression with a cutoff fold change of at least 2 in any of the treatment conditions (Figure 3a). Out of the 1805 genes, 50 and 157 genes showed synergistic upregulation and downregulation respectively in the combination treatment as compared to the two single treatments, using a synergistic scoring method previously described.20 Pathway analysis of 1805 genes using ingenuity pathway analysis identified FOXM1, MYC and E2F1 among other transcriptional factors whose target genes showed significant enrichments in the combination treatment (Figure 3b, Supplementary Figure S3A). These transcription factors, unlike TBX2, showed incremental decrease in their activation z-score after combination treatment at both 48 and 72 h, suggesting that the activity of these transcriptional factors are likely regulated by both CDK4/6 and MAPK pathways. Thus, upon combinatorial inhibition of these two pathways, these transcription factors were further inactivated. Analysis using the synergistic genes input into ingenuity pathway analysis and a computationally simulated prediction model of CDK4/6 and MEK inhibition also predicted involvement of similar transcription factors (Supplementary Figure S3B–D). Here, we showed that targets of FOXM1 were also further downregulated by the combination treatment (Figure 3c), supporting the involvement of FOXM1 in the treatment.

Figure 3
figure 3

Inhibition of CDK4/6 and MEK converge to downregulate KRAS associated gene signature and important cell cycle related transcription factors. (a) Diagram showing the workflow of transcription factors prediction. (Left) Heatmap display of 1805 genes that expression showed at least 2 fold change in one treatment condition (2 μM Palbociclib, 2 μM PD0325901 or combination) in DLD1. (Right) Heatmap of 207 synergistic genes were identified through the use of synergy scoring. These 2 gene sets were input into Ingenuity Pathway Analysis for upstream regulators prediction. (b) Activation z-score of predicted upstream regulator using log2 transformed expression of genes that met the cutoff of fold change 2 after 48 and 72 h treatment of 2 μM Palbociclib or 2 μM PD0325901 alone or in combination. Activation z-score infers the activation state of predicted transcription factors through experimentally observed gene expression. (c) mRNA expression of known transcriptional targets of FOXM1 after indicated treatment with cooperative downregulation observed in DLD1 and HCT116 but not in KRAS-independent HCT15. (d) Box plot displaying 27 out of 32 genes of the KRAS associated gene mutation significantly further downregulated after combination treatment. (e) Validation of synergistic downregulation of KRAS associated gene signature using qPCR in KRAS-dependent DLD1 and HCT116 and lack of cooperative downregulation in KRAS-independent HCT15 (concentration of inhibitors as used in (c)). Error bars represent s.e.m, n=3. (f) qPCR represented in a scatter plot showing gene expression of PBK, BUB1B, FOXM1B, CDK2, CDCA7, KIF11 and TIMELESS in the KRAS associated gene signature significantly lower in isogenic KRAS wild type DLD1 compared to KRAS G13D mutant DLD1. Relative expression to DLD1 parental was shown. *P<0.05, ****P<0.0001. P-values were calculated with two-tailed t test.

Full size image

Importantly, we showed that combined inhibition of CDK4/6 and MEK pathways robustly inhibit the expression of KRAS-associated gene signature identified in CRC samples. 27/32 genes in the signature were significantly further downregulated after combination treatment as compared to single MEK inhibitor treatment (Figure 3d). Since MEK inhibitor monotherapy has been previously shown to be ineffective, we propose that residual expression of these genes after MEK inhibition was adequate to drive the survival of the cancer cells, and that co-inhibition of CDK4/6 is required to sufficiently abolish the expression of these genes. Furthermore, 40% of the gene signature showed synergistic downregulation by the dual treatment as compared to the single treatment (Supplementary Figure S3E). This downregulation was validated with representative genes in HCT116 but was not observed in KRAS-independent HCT15 (Figure 3e). Thus, the change of KRAS-associated gene expression following the combination treatment corresponds to the degree of cellular response to the combination treatment seen in Figures 2a and c.

Moreover, in DLD1 isogenic KRAS WT and KRAS G13D cells, there was significantly higher level of KRAS-associated genes expressed in KRAS G13D (Figure 3f), further demonstrating the relevance of this gene signature to KRAS. Taken together, these findings suggest that CDK4/6 and MEK pathway are vital pathways driving KRAS-mutant CRC tumorigenesis and inhibition of both pathways is required to sufficiently shut down their transcriptional program and decrease cell viability.

CDK4/6 and MEK inhibition synergistically downregulate FOXM1 in KRAS-dependent CRC

To gain further molecular insights into the combinatorial effect of Palbociclib and PD0325901, we profiled the changes of related signaling pathways in response to the treatments. CDK4/6 inhibitor Palbociclib treatment alone had negligible effect on the RAS-MAPK pathway as seen in the insignificant changes in phosphorylated ERK1/2 and its downstream effector c-MYC, whereas PD0325901 alone was sufficient to abrogate both P-ERK1/2 and c-MYC expression (Figure 4a). This suggests that C-MYC downregulation is not a crucial determinant of the synergistic effect seen in Palbociclib/PD0325901 treatment. In contrast, although Palbociclib alone only led to modest downregulation of P-RB and E2F1, which are also downstream effectors of CDK4/6, combination with PD0325901, strongly downregulated their expression, though this synergistic effect was more prominent in DLD1 compared to HCT116. Consistent with our prediction in Figure 3, FOXM1 expression, another effector of CDK4/6, as well as its downstream target Cyclin B1, also displayed a synergistic downregulation upon combination treatment when compared to single inhibitor treatments in both KRAS-dependent DLD1 and HCT116 (Figure 4a, Supplementary Figure 4A). Other transcription factors such as E2F4 and p-FOXO3a did not have differential expression after combination treatment in DLD1 and HCT116. Similar observation was also obtained in BRAF-mutant CRC cells (Figure 4b). In KRAS-independent HCT15, however, combination treatment did not result in similar synergistic effects and was unable to abolish the expression of P-RB, E2F1, FOXM1 and Cyclin B1 (Figure 4c, Supplementary Figure 4A). These data provided further evidence that CDK4/6 and RAS-MAPK signaling cooperatively regulate the downstream targets of E2F1 and FOXM1, whose effective inhibition is crucial for loss of viability in KRAS-dependent and BRAF-mutant CRC.

Figure 4
figure 4

CDK4/6 and MEK inhibition downregulate FOXM1 in KRAS dependent colorectal cancer but not in KRAS independent colorectal cancer. (a) Western blot (cropped) showing CDK4/6 and MAPK signaling and predicted transcription factors in KRAS dependent DLD1 and HCT116 after 48 h treatment with Palbociclib or PD0325901 alone or in combination. (b) Western blots (cropped) showing expression of predicted transcription factors involved after 48 h treatment in BRAF-mutant RKO and HT29. (c) Western blot (cropped) showing CDK4/6 and MAPK signaling and predicted transcription factors in KRAS independent HCT15 after 48 h treatment with Palbociclib or PD0325901 alone or in combination.

Full size image

Moreover, when we examined the dynamic changes of CDK4/6 and RAS-MAPK pathways in response to single and combination treatments, we observed that single inhibition of CDK4/6 was only able to downregulate RB phosphorylation and E2F1 for 24 h, but after 48 h, the expression of P-RB and E2F1 rebounded (Supplementary Figures S4B–C). In contrast, the combination treatment continued to maintain the downregulation of P-RB, E2F1 as well as FOXM1, indicating that the combination treatment is able to repress the signaling pathways for a longer period, preventing the ‘leakage’ of signals that could continue to drive survival in CRC as seen in the single treatment of either Palbociclib or PD0325901.

FOXM1 depletion with MEK inhibition is sufficient to inhibit growth and downregulate KRAS-associated gene signature in KRAS-dependent CRC

Next, we sought to validate the involvement of proposed transcription factors in KRAS dependency. FOXM1 knockdown reduced cell proliferation and colony formation in soft agar in KRAS-dependent DLD1 and HCT116, but not in KRAS-independent HCT15 (Figures 5a–d). In contrast, E2F1 knockdown did not result in any effect on proliferation (data not shown) and this could be attributed by functional redundancy in the E2F family.21 Furthermore, in DLD1 KRAS isogenic lines, KRAS knockdown resulted in more prominent reduction of cell proliferation in the parental and KRAS G13D lines, which harbor KRAS mutation, as compared to the KRAS wildtype line (Supplementary Figure S5A). To investigate the cooperative effect of FOXM1 depletion and MEK inhibition, we treated FOXM1 knockdown cells with PD0325901. The combination of FOXM1 and MEK inhibition led to a further reduction of proliferation and soft agar colony formation as well as the expression of KRAS-associated genes in both KRAS-dependent DLD1 and HCT116 but not in KRAS-independent HCT15 (Figures 5e–h). FOXM1 expression was also still present after KRAS knockdown in DLD1 and HCT116 (Supplementary Figure S5B), suggesting again that RAS-MEK inhibition alone is insufficient to downregulate FOXM1. Collectively, these results support an important role of FOXM1 in KRAS-dependent cells, and downregulation of FOXM1 expression contributes to loss of cell viability of KRAS-dependent CRC in response to combined CDK4/6 and MEK inhibition.

Figure 5
figure 5

Knockdown of FOXM1 cooperate with MEK inhibition to inhibit growth and downregulate KRAS associated gene signature in KRAS-dependent colorectal cancer. (a) Cell proliferation assay in DLD1, HCT116 and HCT15 treated with 30 nM of siNC or siKRAS. Proliferation is represented relative to respective siNC after normalization to baseline CTG unit on Day 0. (b) Cell proliferation assay in DLD1, HCT116 and HCT15 treated with 30 nM of siNC or siFOXM1. Proliferation is represented relative to respective siNC after normalization to baseline CTG unit on Day 0. (c) Anchorage independent soft agar assay of DLD1, HCT116 and HCT15 after knock down of KRAS and FOXM1 using the same condition as in (a, b). Number of colonies on day 12 was represented relative to respective siNC. (d) qPCR and protein validation of FOXM1 knock down efficiency 72 h after transfection in DLD1, HCT116 and HCT15. (e) Cell proliferation assay (Left) and soft agar assay (Right) of KRAS dependent DLD1 treated with siFOXM1 in the presence or absence of 518 nM PD0325901. Proliferation is represented relative to Day 0 CTG unit. Number of colonies was quantified on day 12. (f) Cell proliferation assay (Left) and soft agar assay (Right) of KRAS dependent HCT116 treated with siFOXM1 in the presence or absence of 126 nM PD0325901. Proliferation is represented relative to Day 0 CTG unit. Number of colonies was quantified on day 12. (g) FOXM1 knockdown and 1 μM PD0325901 treatment in HCT15 showing no effect on cell proliferation (Left) and soft agar colony formation (Right). (h) qPCR expression of representative genes from the KRAS associated gene signature in DLD1, HCT116 and HCT15 72 h after siFOXM1 transfection and in the presence or absence of PD0325901 24 h treatment (concentration of PD0325901 as used in cell proliferation assay in (e–g)). All data in the graphs represent mean±s.e.m., n=3. *P<0.05, **P<0.01, ***P<0.001, n.s., not significant. Student’s t test was used for statistical analysis.

Full size image

Validation of combination therapy in CRC mouse xenograft models

To demonstrate the combinatorial effect of Palbociclib and PD0325901 in vivo, we engrafted NOD/SCID mice with DLD1 and RKO cells and upon tumor formation, the mice were treated with vehicle, 30 mg/kg Palbociclib (50 mg/kg for RKO xenografts), 20–25 mg/kg PD0325901 or a combination of both inhibitors for 14 days. In the DLD1 model, though single treatment did not yield statistically significant effect on tumor growth, the combination treatment led to a significant tumor growth inhibition (Figure 6a). In RKO model, there is a significantly greater reduction in tumor growth in the combination-treated group when compared to the MEK inhibitor treated group (Figure 6b). For both xenograft models, negligible body weight loss was observed in the combination-treated group over the treatment period when compared to vehicle-treated group (Figure 6e, Supplementary Figure 6), though PD032591 given at 25 mg/kg resulted in mortality in a few mice in both single and combination group. In addition, mice treated with combination treatment showed further suppression of CDK4/6 signaling in terms of P-RB and E2F1 downregulation and also reduction of FOXM1, Cyclin B1 and c-MYC expression (Figures 6c and d). Thus, this combination treatment demonstrated more remarkable activity against KRAS-dependent/BRAF-mutant CRC in vivo, consistent with in vitro results.

Figure 6
figure 6

Antitumor efficacy of CDK4/6 and MEK inhibition In Vivo. (a) DLD1 xenografts tumor growth (mm3) in NOD/SCID mice during treatment with vehicle, 30 mg/kg Palbociclib (daily), 20 mg/kg PD0325901 (every 5d with 2d break) or in combination. Treatment initiated when average tumors volume reached ~100 mm3.Vehicle, n=9; Palbociclib-treated, n=10; PD0325901-treated, n=5; Palbociclib and PD0325901 treated, n=6. (b) RKO xenografts tumor volume (mm3) in NOD/SCID mice during treatment with 50 mg/kg Palbociclib (daily) and 20 mg/kg PD0325901 (every 5d with 2d break) or in combination. Treatment initiated when average tumors volume reached ~100 mm3. Vehicle, n=9; Palbociclib-treated, n=10; PD0325901-treated, n=8; Palbociclib and PD0325901 treated, n=7. Mean tumor volume±s.e.m. is shown. *P<0.05; ***P<0.001, n.s., not significant. Statistical analysis is done using unpaired t-test. (c, d) Western blot (cropped) of protein lysates from DLD1 and RKO tumor xenografts at the end of the treatment regimen. Phosphorylation sites are as stated in Figure 4. (e) Body weight of mice at start and end of treatment regimen in DLD1 and RKO xenografts.

Full size image

Discussion

We implemented an integrated approach to the identification of a KRAS-associated gene signature from colorectal tumors using quantitative KRAS mutation pyrosequencing and gene expression analysis in order to define new treatment strategies. Our analysis identified that in CRC tumors with higher proportion of KRAS mutation, CDK4/6 and MAPK co-regulated gene set is highly enriched and targeting this KRAS-associated gene signature with CDK4/6 and MEK inhibitors efficiently inhibited CRC growth and elicited apoptosis in KRAS-dependent and BRAF-mutant CRC. We further demonstrate that downregulation of mitotic transcription factor FOXM1, as well as E2F1 and c-MYC, plays a key role in the synergistic action of combined CDK4/6 and MEK inhibition.

The presence of oncogenic KRAS mutation has excluded metastatic CRC patients from targeted therapies, leaving them with only chemotherapy, or worse no treatment if disease is chemorefractory. Numerous preclinical therapeutic strategies were developed22, 23, 24, 25 but none have been approved for clinical use because of safety issues or lack of objective responses during clinical trials. Palbociclib is approved for breast cancer treatment and PD0325901 had been studied in phase 1/2 clinical trials. Combined treatment of Palbociclib at 50 mg/kg and PD0325901 at 20 mg/kg showed negligible body weight loss in our in vivo model and showed reduced cytotoxicity to normal epithelial colon cells in our in vitro testing, though we did notice the toxicity of PD0325901 at 25 mg/kg. Of note, half the dose of CDK4/6 inhibitor (35 and 50 mg/kg) was administered as compared to the higher dosage of 100 mg/kg used in previous study.26 The therapeutic approach of co-targeting CDK4/6 and RAS-MAPK pathways is currently in phase1/2 clinical trial for mainly KRAS-mutant non-small cell lung cancer patients (NCT02022982), highlighting the high translatability of this treatment strategy to KRAS-mutant CRC and based on our study, a lower dose of CDK4/6 inhibitor could be used to achieve sufficient efficacy, yet limiting toxicity at the same time.

Mechanistically, the efficacy of combination treatment of CDK4/6 and MEK inhibitors may involve the downregulation of several transcription factors such as FOXM1, E2F1 and CMYC which are often overexpressed and have vital roles in cancers.27, 28, 29 We showed that mitotic transcription factor FOXM1 may play a key role in conferring the synergistic inhibition of CRC by the combination treatment. Others have shown that RAS-driven cancers have elevated levels of cell cycle and mitosis activity, which we also showed in the DLD1 KRAS wild type cells displaying slower growth compared to its mutant counterpart, either through its downstream effector MAPK or through altering cancer cell metabolism.30, 31, 32 Our findings regarding FOXM1 dependence in KRAS-driven cancers could be because of FOXM1 role in regulating genes involved in DNA damage repair,33 a process vital in KRAS-mutant cancer cells under replicative stress34, 35 which is a possible explanation for the effectiveness of this combination treatment in KRAS-dependent CRC. Furthermore, E2F1 activity is also reported to limit DNA damage in cancers with high replicative stress and this treatment completely downregulated E2F1 expression and its predicted transcriptional network.36 And in KRAS-mutant pancreatic cancer, incomplete downregulation of CMYC by MEK inhibitor continued to drive survival of cancer cells through phosphorylated RB, supporting that CDK4/6 inhibition is vital in KRAS-driven cancers.37

The identified KRAS-associated gene signature has potential clinical application. Given its high correlation with KRAS dependency, it can serve as a biomarker for KRAS functionality together with the current clinical method for KRAS mutation detection. This signature is a summation of transcriptional output of upstream signaling and by integrating the signaling generated by a myriad of genetic and epigenetic changes in tumor cells, it is more informative than just the KRAS mutation status, considering multiple varied genomic alterations present in different tumors. Activity of RAS-MAPK pathway is not only determined by KRAS or BRAF mutation status but could be also because of alterations in other activators, repressors and negative feedback mechanisms.38 Thus with this gene signature, we can identify not only KRAS-mutant CRC but also CRC with elevated RAS-MAPK signaling, which will also likely be responsive to CDK4/6 and MEK inhibitors treatment. It could also be important to differentiate KRAS-dependent from KRAS-independent cancer even in KRAS-mutant tumors as we and other groups7, 8, 39 have observed this phenomenon which may lead to resistance to treatments requiring KRAS dependence especially with the development of direct mutant KRAS inhibitor.40, 41 This will affect treatment options for this subset of cancer and emphasize the need to stratify CRC into clearer groups to facilitate effective treatment.

Materials and methods

Cell lines and drug treatment

DLD1, HCT116, HCT15, RKO, HT29, FHs74 Int and CCD841 CoN cell lines were obtained from American Type Culture Collection. FHs 74 Int and CCD841 CoN were cultured in DMEM supplemented with 10% FBS and 100 μg/ml penicillin/streptomycin, 30 ng/ml EGF, non-essential amino acid, 10 μg/ml insulin, 1 mM oxaloacetate and 0.5 mM sodium pyruvate and the rest were grown in Dulbecco’s modified Eagles’ medium supplemented with 10% FBS and 100 μg/ml penicillin/streptomycin. KRAS isogenic DLD1 lines were obtained from Horizons Discovery (Cambridge, UK) and grown in RPMI-1640 medium supplemented with 10% FBS and 100 μg/ml penicillin/streptomycin. Palbociclib (Axon1505), PD0325901 (Axon1408), Ribociclib (Axon2273) and Trametinib (Axon1761) for in vitro drug treatment were obtained from Axon Medchem (Groningen, Netherlands).

Cell viability assay, fluorescence activated cell sorting (FACs)

Cells were seeded at optimal seeding density in 96-well plates 24 h before drug treatment. CellTiter-Glo reagent was added and luminescence was measured using GloMax Explorer (Promega, Madison, WI, USA) on stated days. All experimental setups were done in at least triplicates. CellTiter-Glo values obtained were expressed normalized to either day of seeding or drug addition.

Cell cycle analysis was done by DNA content quantification to quantify sub-G1 population, which is reflective of extent of cell death. Briefly, cells treated for 7 days were fixed with 70% ethanol and stained with Propidium iodide (50 μg/ml). Stained cells were analyzed with FACScalibur (BD Biosciences, Singapore) and quantified by CellQuest software (BD Biosciences).

Anchorage independent colony formation assay

Approximately 5000 cells were seeded in complete media containing 0.3% bactoagar, and plated onto pre-coated 0.6% bactoagar plate. Media or media containing drug treatment were added to each well the following day. After 12 days, colonies were stained with iodonitrotetrozolium chloride (Sigma, St Louis, MO, USA), scanned and quantified using GelCount. Average of three replicates was determined.

Immunoblotting

For immunoblotting, cells were trypsinized and washed in PBS before subjected to lysis using radioimmunoprecipitation assay lysis buffer (50 mM Tris–HCl pH7.4, 1 mM EDTA, 150 mM NaCl, 1% Igepal CA630, 0.5% sodium deoxycholate, 1 mM Na2VO4, 20 mM NaF, 1 mM PMSF and complete protease inhibitor (Roche, Mannheim, Germany)) and sonication. To obtain protein lysates from xenograft studies, tumors were lysed in the same buffer with the use of Tissuelyser II (Qiagen, Hilden, Germany) before sonication. Protein concentration was estimated using BioRad Bradford dye with known BSA concentration as standards and measured using TecanXfluor software.

Protein samples (30 ug) were separated by SDS–polyacrylamide gel electrophoresis (SDS–PAGE) gel and transferred onto PVDF membrane (Millipore, Billerica, MA, USA) using Trans-Blot SD Semi-Dry transfer cell (Bio-Rad, Hercules, CA, USA). Membranes were probed with antibodies listed above. Chemiluminescence was detected after exposure to Supersignal WestFemto maximum sensitivity substrate (Thermo Scientific, Rockford, IL, USA) using ChemiDoc MP imaging systems (Biorad, Hercules, CA, USA) with Imagelab software.

Pyrosequencing for quantitative analysis of KRAS mutation

Genomic material extracted from fresh frozen bulk CRC tumors were amplified along KRAS codon 12 and 13 region with KRAS PCR forward primer (5′-AGGCCTGCTGAAAATGACTG-3′) and biotinlyated KRAS PCR reverse primer (5′-[Biotin]CAAGATTTACCTCTATTG-3′) using the Qiagen PyroMark PCR kit according to manufacturer’s protocol.

Biotinlyated PCR products were bound to Streptavidin Sepharose beads, denatured and washed, leaving only biotinlyated single-stranded templates which were incubated with KRAS sequencing primer (5′-TTGTGGTAGTTGGAGC-3′) on PyroMark Q24 plate. Mixture is then incubated at 80 °C for 2 min and cooled to room temperature for primer and template hybridization. Reaction mixtures were ran on PyroMark Q24 using PyroMark Q24 Gold Q24 Reagents Kits and data were analyzed using the PyroMark Q24 software.

Microarray analysis

Illumina gene expression data of human CRC and matched normal controls have been previously described42 and can be found in GEOarchive under accession number GSE10972 and GSE74604. Gene expression for DLD1 treated with Palbociclib and PD0325901 for 24, 48 and 72 h can be found under accession number GSE74604. Total RNA was isolated from cell lines using Trizol (Invitrogen, Carlsbad, CA, USA) and purified with RNeasy Mini Kit (Qiagen). Reverse transcription was performed using an RNA Amplification kit (Ambion, Carlsbad, CA, USA). Microarray hybridization was carried out using Illumina Gene Expression SentrixBeadChip HumanRef-8_V2 and HumanHT12-v4 according to manufacturer’s protocol and data analysis was performed using GeneSpring software from Agilent Technologies as described in Tan et al.43

Gene ontology analysis

Genes with two-fold change in expression in any of the treatment groups, as compared to control were imported into Ingenuity Pathway Analyses software. Activation Z-score for predicted upstream transcription factors were obtained. Same was done for synergistic genes using 72 h timepoint expression values. Results were exported and plotted on GraphPad.

Quantitative RT-PCR analysis

Total RNA was converted to complementary DNA (cDNA) using High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Carlsbad, CA, USA). RT-qPCR was carried out using KAPA SYBR FAST qPCR Kit (Kapa Biosystems, Wilmington, MA, USA) and was amplified and quantified with PRISM 7900 Sequence Detection System (Applied Biosystems). RT-qPCR primer sequences (5′ to 3′) are as followed: ACTB:FP-GCACAGAGCCTCGCCTT, RP-GTTGTCGACGACGAGCG; 18S:FP-CGAACGTCTGCCCTATCAACTT, RP-ACCCGTGGTCACCATGGTA; FOXM1:FP-AACCGCTACTTGACATTGGC, RP-GCAGTGGCTTCATCTTCC; BUB1:FP-ATCTCCCTGGGTAGCTTCGT, RP-CCATCAAGCCCAAGACTGAA; PBK:FP-CAGCTGCCGGGCGTATGTGT, RP-CTCAGTCCAGAGTCTCACCGCCT; CDCA7:FP-GGCTTTTCAGAAAGTGAGGTGC, RP-AACTTCATCGCCACCCTGAG; FOXM1B:FP-AGGTGTTTAAGCAGCAGAAACG, RP-GCTAGCAGCACCTTGGGGGCAA; CDK2:FP-ATCCGCCTGGACACTGAGAC, RP-TTGCAGCCCAGGAGGATTTC; KIF11:FP-CTGCCAGCAAGCTGCTTAAC, RP-CCTGGGAATGGGTCTGCTTT; TIMELESS:FP-ATGACAGGTCTTCCAGTCGC, RP-TGGATGATCTGCTTGCGTGT; BUB1B:FP-GCAAAGGGAAAAAGACAGCA, RP-TGCATCTGTTGAGGAAATGG.

Synergy scoring

Fold change two cut-off was applied on gene expression data from DLD1 to obtain differentially regulated genes after treatment. To identify genes whose expression synergistically responded to treatment with Palbociclib and PD0325901, we used a formula previously described in McMurray et al.20 The formula defines synergistic genes whose expression fits for upregulated genes and for downregulated genes, where ‘a’ represents the expression value of given gene after Palbociclib treatment, ‘b’ represents expression value for the same gene after PD0325901 treatment and ‘d’ represents expression value for this gene after Palbociclib and PD0325901 combination treatment.

siRNA transfection

Transfection was done with Lipofectamine RNAiMAX (Invitrogen) according to manufacturer’s protocol. Both non-targeting control, siNC, and target-specific siRNA were obtained from Integrated DNA Technologies. siKRAS:5′-GACGATACAGCTAATTCAGAA-3′; siFOXM1:5′-GGACCACUUUCCCUACUUU-3′

Antibodies

Antibodies against following proteins were obtained from Cell Signaling Technology (Danvers, MA, USA) and used at indicated dilution: RB(4H1) #9309-(1:2000), P-RB(Ser780) #9307-(1:1000), P-p44/22 MAPK(Thr202/Tyr204) #9101-(1:1000), Total p44/22 MAPK #9102-(1:2000), c-Myc #9402-(1:2000), pFOXO3A(Ser253) #9466-(1:1000) and cleaved parp #9541-(1:1000). Antibodies against following proteins were obtained from Santa Cruz Biotechnologies (Santa Cruz, CA, USA): FOXM1(A11) #sc-271746-(1:500), Cyclin B1(GNS1) #sc-245-(1:500), E2F1(C-20) #sc-193-(1:1000), E2F4(A-20) #sc-1082-(1:1000) and Cyclin D1(HD11) #sc-246-(1:1000). Beta Actin antibody (Sigma-Aldrich) was used at the dilution of 1:200 000.

Animal studies

DLD1 and RKO xenografts were generated via injection of 3 × 106 cells with Matrigel (BD Biosciences #354234) in a ratio of 1:1 in a 50 μl volume into flank of 4–8 week old female NOD/SCID mice obtained from InVivos (Singapore). Mice were randomized to 4 treatment groups once tumors reached an average size of 100 mm3. Palbociclib were administered daily at 35 mg/kg for DLD1 xenografts and at 50 mg/kg for RKO xenografts via oral gavage. PD0325901 were administered daily for 5 days at 20 mg/kg via oral gavage. Vehicle used for Palbociclib was PBS and for PD0325901 was 0.5% hydropropylmethylcellulose, 0.2% Tween-80 and 5% DMSO. Tumor volume was measured by electronic caliper twice a week and calculated with following formula: Length × (Width2) × 0.5. Palbociclib (CT-PD2991) and PD0325901 (CT-PD03) for in vivo use were obtained from ChemieTek (Indianapolis, IN, USA).

Study approval

All animal studies were performed in compliance with protocols approved by Biopolis Institutional Animal Care and Use Committee of Singapore. Human tissue DNA and RNA samples were originally obtained from Singapore Tissue Network and National University of Singapore (NUS) using protocols approved by Institutional Review Board of NUS; informed consent was obtained from each individual who provided the tissues.

Statistical analyses

In-vitro experiments were repeated at least three times and data were reported as mean+s.e.m. Statistical significances were assessed by two-tailed Student’s t-test, one-way or two-way analysis of variance for multiple group comparisons using GraphPad Prism 6 software. P⩽0.05 was considered significant.

Accession codes

Accessions

Gene Expression Omnibus

References

  1. Prior IA, Lewis PD, Mattos C . A comprehensive survey of Ras mutations in cancer. Cancer Res 2012; 72: 2457–2467.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  2. Russo A, Bazan V, Agnese V, Rodolico V, Gebbia N . Prognostic and predictive factors in colorectal cancer: Kirsten Ras in CRC (RASCAL) and TP53CRC collaborative studies. Ann Oncol 2005; 16 (Suppl 4): iv44–iv49.

    Article  PubMed  Google Scholar 

  3. Van Cutsem E, Kohne CH, Hitre E, Zaluski J, Chang Chien CR, Makhson A et al. Cetuximab and chemotherapy as initial treatment for metastatic colorectal cancer. N Engl J Med 2009; 360: 1408–1417.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  4. Misale S, Yaeger R, Hobor S, Scala E, Janakiraman M, Liska D et al. Emergence of KRAS mutations and acquired resistance to anti-EGFR therapy in colorectal cancer. Nature 2012; 486: 532–536.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  5. Mendoza MC, Er EE, Blenis J . The Ras-ERK and PI3K-mTOR pathways: cross-talk and compensation. Trends Biochem Sci 2011; 36: 320–328.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  6. Yeh JJ, Routh ED, Rubinas T, Peacock J, Martin TD, Shen XJ et al. KRAS/BRAF mutation status and ERK1/2 activation as biomarkers for MEK1/2 inhibitor therapy in colorectal cancer. Mol Cancer Ther 2009; 8: 834–843.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  7. Scholl C, Frohling S, Dunn IF, Schinzel AC, Barbie DA, Kim SY et al. Synthetic lethal interaction between oncogenic KRAS dependency and STK33 suppression in human cancer cells. Cell 2009; 137: 821–834.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  8. Luo J, Emanuele MJ, Li D, Creighton CJ, Schlabach MR, Westbrook TF et al. A genome-wide RNAi screen identifies multiple synthetic lethal interactions with the Ras oncogene. Cell 2009; 137: 835–848.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  9. Kosmidou V, Oikonomou E, Vlassi M, Avlonitis S, Katseli A, Tsipras I et al. Tumor heterogeneity revealed by KRAS, BRAF, and PIK3CA pyrosequencing: KRAS and PIK3CA intratumor mutation profile differences and their therapeutic implications. Hum Mutat 2014; 35: 329–340.

    CAS  Article  PubMed  Google Scholar 

  10. Ciardiello F, Normanno N, Maiello E, Martinelli E, Troiani T, Pisconti S et al. Clinical activity of FOLFIRI plus cetuximab according to extended gene mutation status by next-generation sequencing: findings from the CAPRI-GOIM trial. Ann Oncol 2014; 25: 1756–1761.

    CAS  Article  PubMed  Google Scholar 

  11. Laoukili J, Kooistra MR, Bras A, Kauw J, Kerkhoven RM, Morrison A et al. FoxM1 is required for execution of the mitotic programme and chromosome stability. Nat Cell Biol 2005; 7: 126–136.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  12. Fu Z, Malureanu L, Huang J, Wang W, Li H, van Deursen JM et al. Plk1-dependent phosphorylation of FoxM1 regulates a transcriptional programme required for mitotic progression. Nat Cell Biol 2008; 10: 1076–1082.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  13. Malumbres M, Barbacid M . Cell cycle, CDKs and cancer: a changing paradigm. Nat Rev Cancer 2009; 9: 153–166.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  14. Choi YJ, Anders L . Signaling through cyclin D-dependent kinases. Oncogene 2014; 33: 1890–1903.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  15. Anders L, Ke N, Hydbring P, Choi YJ, Widlund HR, Chick JM et al. A systematic screen for CDK4/6 substrates links FOXM1 phosphorylation to senescence suppression in cancer cells. Cancer Cell 2011; 20: 620–634.

    CAS  PubMed  PubMed Central  Google Scholar 

  16. Harbour JW, Dean DC . The Rb/E2F pathway: expanding roles and emerging paradigms. Genes Dev 2000; 14: 2393–2409.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  17. Gao J, Aksoy BA, Dogrusoz U, Dresdner G, Gross B, Sumer SO et al. Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal. Sci Signal 2013; 6: pl1.

    Article  PubMed  PubMed Central  Google Scholar 

  18. Cerami E, Gao J, Dogrusoz U, Gross BE, Sumer SO, Aksoy BA et al. The cBio cancer genomics portal: an open platform for exploring multidimensional cancer genomics data. Cancer Discov 2012; 2: 401–404.

    Article  PubMed  PubMed Central  Google Scholar 

  19. Calon A, Lonardo E, Berenguer-Llergo A, Espinet E, Hernando-Momblona X, Iglesias M et al. Stromal gene expression defines poor-prognosis subtypes in colorectal cancer. Nat Genet 2015; 47: 320–329.

    CAS  Article  PubMed  Google Scholar 

  20. McMurray HR, Sampson ER, Compitello G, Kinsey C, Newman L, Smith B et al. Synergistic response to oncogenic mutations defines gene class critical to cancer phenotype. Nature 2008; 453: 1112–1116.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  21. Wu L, Timmers C, Maiti B, Saavedra HI, Sang L, Chong GT et al. The E2F1-3 transcription factors are essential for cellular proliferation. Nature 2001; 414: 457–462.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  22. Berndt N, Hamilton AD, Sebti SM . Targeting protein prenylation for cancer therapy. Nat Rev Cancer 2011; 11: 775–791.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  23. She QB, Halilovic E, Ye Q, Zhen W, Shirasawa S, Sasazuki T et al. 4E-BP1 is a key effector of the oncogenic activation of the AKT and ERK signaling pathways that integrates their function in tumors. Cancer Cell 2010; 18: 39–51.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  24. Ebi H, Corcoran RB, Singh A, Chen Z, Song Y, Lifshits E et al. Receptor tyrosine kinases exert dominant control over PI3K signaling in human KRAS mutant colorectal cancers. J Clin Invest 2011; 121: 4311–4321.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  25. Corcoran RB, Cheng KA, Hata AN, Faber AC, Ebi H, Coffee EM et al. Synthetic lethal interaction of combined BCL-XL and MEK inhibition promotes tumor regressions in KRAS mutant cancer models. Cancer Cell 2013; 23: 121–128.

    CAS  Article  Google Scholar 

  26. Ziemke EK, Dosch JS, Maust JD, Shettigar A, Sen A, Welling TH et al. Sensitivity of KRAS Mutant Colorectal Cancers to Combination Therapy that Co-Targets MEK and CDK4/6. Clin Cancer Res 2016; 22: 405–414.

    CAS  Article  PubMed  Google Scholar 

  27. Pilarsky C, Wenzig M, Specht T, Saeger HD, Grutzmann R . Identification and validation of commonly overexpressed genes in solid tumors by comparison of microarray data. Neoplasia 2004; 6: 744–750.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  28. Dang CV . MYC on the path to cancer. Cell 2012; 149: 22–35.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  29. Putzer BM, Engelmann D . E2F1 apoptosis counterattacked: evil strikes back. Trends Mol Med 2013; 19: 89–98.

    Article  PubMed  Google Scholar 

  30. Kwong LN, Costello JC, Liu H, Jiang S, Helms TL, Langsdorf AE et al. Oncogenic NRAS signaling differentially regulates survival and proliferation in melanoma. Nat Med 2012; 18: 1503–1510.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  31. Liu X, Jakubowski M, Hunt JL . KRAS gene mutation in colorectal cancer is correlated with increased proliferation and spontaneous apoptosis. Am J Clin Pathol 2011; 135: 245–252.

    CAS  Article  PubMed  Google Scholar 

  32. Bryant KL, Mancias JD, Kimmelman AC, Der CJ . KRAS: feeding pancreatic cancer proliferation. Trends Biochem Sci 2014; 39: 91–100.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  33. Zona S, Bella L, Burton MJ, Nestal de Moraes G, Lam EW . FOXM1: an emerging master regulator of DNA damage response and genotoxic agent resistance. Biochim Biophys Acta 2014; 1839: 1316–1322.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  34. Tan Y, Chen Y, Yu L, Zhu H, Meng X, Huang X et al. Two-fold elevation of expression of FoxM1 transcription factor in mouse embryonic fibroblasts enhances cell cycle checkpoint activity by stimulating p21 and Chk1 transcription. Cell Prolif 2010; 43: 494–504.

    CAS  Article  PubMed  Google Scholar 

  35. Gilad O, Nabet BY, Ragland RL, Schoppy DW, Smith KD, Durham AC et al. Combining ATR suppression with oncogenic Ras synergistically increases genomic instability, causing synthetic lethality or tumorigenesis in a dosage-dependent manner. Cancer Res 2010; 70: 9693–9702.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  36. Bertoli C, Herlihy AE, Pennycook BR, Kriston-Vizi J, de Bruin RA . Sustained E2F-Dependent Transcription Is a Key Mechanism to Prevent Replication-Stress-Induced DNA Damage. Cell Rep 2016; 15: 1412–1422.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  37. Hayes TK, Neel NF, Hu C, Gautam P, Chenard M, Long B et al. Long-Term ERK Inhibition in KRAS-Mutant Pancreatic Cancer Is Associated with MYC Degradation and Senescence-like Growth Suppression. Cancer Cell 2016; 29: 75–89.

    CAS  Article  PubMed  Google Scholar 

  38. Katz M, Amit I, Yarden Y . Regulation of MAPKs by growth factors and receptor tyrosine kinases. Biochim Biophys Acta 2007; 1773: 1161–1176.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  39. Singh A, Greninger P, Rhodes D, Koopman L, Violette S, Bardeesy N et al. A gene expression signature associated with 'K-Ras addiction' reveals regulators of EMT and tumor cell survival. Cancer Cell 2009; 15: 489–500.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  40. Ostrem JM, Peters U, Sos ML, Wells JA, Shokat KM . K-Ras(G12C) inhibitors allosterically control GTP affinity and effector interactions. Nature 2013; 503: 548–551.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  41. Cox AD, Fesik SW, Kimmelman AC, Luo J, Der CJ . Drugging the undruggable RAS: Mission possible? Nat Rev Drug Discov 2014; 13: 828–851.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  42. Jiang X, Tan J, Li J, Kivimae S, Yang X, Zhuang L et al. DACT3 is an epigenetic regulator of Wnt/beta-catenin signaling in colorectal cancer and is a therapeutic target of histone modifications. Cancer Cell 2008; 13: 529–541.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  43. Tan J, Yang X, Zhuang L, Jiang X, Chen W, Lee PL et al. Pharmacologic disruption of Polycomb-repressive complex 2-mediated gene repression selectively induces apoptosis in cancer cells. Genes Dev 2007; 21: 1050–1063.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

This work was supported by the A*STAR of Singapore, International S&T Cooperation Program of China (ISTCP) (No. 2013DFG32990) and National High Technology Research and Development Program (863) of China (No. 2012AA02A520). We thank Puay Leng Lee for technical assistance in western blot analysis.

Author information

Affiliations

  1. Cancer Therapeutics and Stratified Oncology, Genome Institute of Singapore, A*STAR (Agency for Science, Technology and Research), Biopolis, Singapore

    M Pek, S M J M Yatim, Y Chen, M Gong, X Jiang & Q Yu

  2. Graduate School for Integrative Sciences and Engineering, National University of Singapore, Singapore

    M Pek

  3. Department of Colorectal Surgery, The Sixth Affiliated Hospital, Sun Yat-sen University, Guangzhou, Guangdong, China

    Y Chen & X Wu

  4. Guangdong Provincial Key Laboratory of Colorectal and Pelvic Floor Disease, The Sixth Affiliated Hospital, Sun Yat-sen University, Guangzhou, Guangdong, China

    Y Chen & X Wu

  5. Computational Biology, Genome Institute of Singapore, A*STAR (Agency for Science, Technology and Research), Biopolis, Singapore

    J Li

  6. School of Computer Engineering, Nanyang Technological University, 50 Nanyang Avenue, Singapore

    F Zhang & J Zheng

  7. Guangdong Institute of Gastroenterology, Guangzhou, Guangdong, China

    X Wu

  8. Department of Physiology, Yong Loo Lin School of Medicine, National University of Singapore, Singapore

    Q Yu

  9. Cancer Research Institute, Jinan University, Guangzhou, Guangdong, China

    Q Yu

  10. Cancer and Stem Cell Biology, Duke-National University of Singapore, Graduate Medical School Singapore, Singapore

    Q Yu

Authors
  1. M Pek
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. S M J M Yatim
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Y Chen
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. J Li
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. M Gong
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. X Jiang
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. F Zhang
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. J Zheng
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. X Wu
    View author publications

    You can also search for this author in PubMed Google Scholar

  10. Q Yu
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Q Yu.

Ethics declarations

Competing interests

The authors declare no conflict of interest.

Additional information

Supplementary Information accompanies this paper on the Oncogene website

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Pek, M., Yatim, S., Chen, Y. et al. Oncogenic KRAS-associated gene signature defines co-targeting of CDK4/6 and MEK as a viable therapeutic strategy in colorectal cancer. Oncogene 36, 4975–4986 (2017). https://doi.org/10.1038/onc.2017.120

Download citation

  • Received:

  • Revised:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/onc.2017.120

Further reading

Search

Advanced search

Quick links