Chemotherapy for colorectal cancer has become more complicated and diversified with the appearance of molecular-targeting agents. 5-Fluorouracil (5-FU) has been a mainstay of chemotherapy for colorectal cancer, but it is still unknown whether the combining of 5-FU with novel molecular-targeting agents is effective. Thymidylate synthase (TS) is a direct target of 5-FU, and the low TS level has been generally supposed to sensitize 5-FU's efficacy. We therefore hypothesized that RB-reactivating agents could enhance the efficacy of 5-FU, because the RB-reactivating agents could suppress the function of transcription factor E2F of TS gene promoter. We used three RB-reactivating agents, trametinib/GSK1120212 (MEK inhibitor), fenofibrate (PPARα agonist), and LY294002 (PI3K inhibitor), with 5-FU against human colon cancer HT-29 and HCT15 cells. Trametinib induced p15 and p27 expression and reduced cyclin D1 levels in HT-29 cells. Fenofibrate also dephosphorlated ERK1/2 and reduced cyclin D1 levels in HT-29 cells. LY294002 induced p27 expression in HCT15 cells. All three agents caused dephosphorylation of RB protein and G1-phase arrest with a reduction of TS expression. As a consequence, the combination of 5-FU with each of the agents resulted in a significant decrease of colony numbers in HT-29 or HCT15 cells. These results suggest “RB-reactivation therapy” using molecular-targeting agents to be a new strategy for 5-FU-based chemotherapy. In particular, we strongly expect trametinib, which was discovered in Japan and was recently submitted to FDA for approval, to be used together with established regimens for colorectal cancer.
5-Fluorouracil (5-FU) has been widely used for the treatment of advanced colorectal cancers. In recent years, combined treatments based on 5-FU have improved the outcome of patients with advanced colorectal cancers.[1, 2] However, 5-FU has problems with chemosensitivity or toxicity, including considerable bone marrow suppression in some cases.
5-Fluorouracil inhibits thymidylate synthase (TS), which plays a central role in DNA biosynthesis. Thymidylate synthase expression is crucial to the efficacy of 5-FU.[4, 5] Actually, the level of TS expression is inversely correlated with the clinical response to 5-FU-based chemotherapy for colorectal and other cancers.. Therefore, the downregulation of TS expression might be a promising way to enhance the efficacy of 5-FU. On the other hand, TS is a target of E2F1, whose transcriptional ability is repressed by direct binding to hypophosphorylated RB protein. We then hypothesized that clinically used molecular-targeting agents that reactivate RB might be useful to improve the efficacy of 5-FU.
In the present study, we used three RB-reactivating agents; the MEK inhibitor trametinib/GSK1120212 (also called JTP-74057), PPARα agonist fenofibrate, and PI3K inhibitor LY294002. Trametinib is a novel allosteric MEK1/2 inhibitor, originally identified as a p15 inductive compound by us, shown to have broad antitumor effects in various tumor xenograft models.[10, 11] Of note, trametinib improved progression-free and overall survival of patients with B-Raf mutated melanoma in phase 3 clinical trial,[12, 13] and was submitted to FDA for approval. Fenofibrate, is a PPARα agonist widely used in the treatment of hyperlipidemia. Recently, several reports have showed anticancer effects of fenofibrate in various cancer cell lines.[15-20] The third agent, LY294002, is known to reactivate RB function by inhibiting the PI3K-Akt pathway.[21, 22]
We found that these three compounds enhanced the efficacy of 5-FU in human colon cancer cells by downregulating TS expression. This study provides a new treatment strategy for 5-FU-based chemotherapy against colon cancer.
Materials and Methods
Human colon adenocarcinoma HT-29 cells were cultured in DMEM. Human colon adenocarcinoma HCT15 cells were cultured in RPMI medium. Each medium was supplemented with 10% FBS, glutamine (4 mM for DMEM and 2 mM for RPMI), and antibiotics (50 U/mL penicillin and 100 μg/mL streptomycin). Cells were incubated at 37°C in a humidified atmosphere of 5% CO2.
5-FU was obtained from Nacalai Tesque (Kyoto, Japan). Trametinib was kindly provided by GlaxoSmithKline (Brentford, Middlesex, UK). Fenofibrate was obtained from Cayman Chemical (Ann Arbor, MI, USA). LY294002 was obtained from Cell Signaling Technology (Beverly, MA, USA). They were dissolved in DMSO.
Cell viability assay
The number of viable cells was measured by a Cell Counting Kit-8 assay according to the manufacturer's instructions (Dojindo, Kumamoto, Japan). After the incubation of HT-29 cells or HCT15 cells for 72 h with the indicated concentrations of various drugs, kit reagent WST-8 was added to the medium and incubated for 4 h. The absorbance at 450 nm of the samples was measured using a multi-plate reader (DS Pharma Biomedical, Osaka, Japan).
Analyses of cell cycle and apoptosis
Cells were incubated with various drugs for 24 or 72 h, and then harvested by trypsinization. After centrifugation, the cells were suspended in PBS containing 0.1% Triton X-100, 150 μg/mL RNase A (Sigma, St Louis, MO, USA), and 25 μg/mL propidium iodide. The stained cells were analyzed using FACSCalibur (Becton Dickinson, Franklin Lakes, NJ, USA). The data were analyzed using Modifit LT software and Cell Quest software (Becton Dickinson).
Protein isolation and western blot analysis
Cells were lysed with a buffer containing 50 mM Tris-HCl, 1% SDS, 1 mM DTT, 0.43 mM 4-(2-aminoethyl) benzenesulfonyl fluoride hydrocholoride (ABSF), and Phosphatase Inhibitor Cocktail (Nacalai Tesque). The lysate was sonicated and centrifuged at 15 000 g for 20 min at 4°C, and the supernatant was collected. Equal amounts of the protein extract were subjected to SDS-PAGE, and transferred to a PVDF membrane (Millipore, Bedford, MA, USA). The following were used as the primary antibody: rabbit anti-human p15 polyclonal antibody (C-20, Santa Cruz Biotechnology, Santa Cruz, CA, USA), rabbit anti-human p21 polyclonal antibody (C-19, Santa Cruz Biotechnology), rabbit anti-human p27 polyclonal antibody (C-19, Santa Cruz Biotechnology), mouse anti-human cyclin D1 monoclonal antibody (MBL, Nagoya, Japan), mouse anti-human thymidylate synthase monoclonal antibody (Abcam, Cambridge, UK), rabbit anti-human Akt, phospho-Akt (Ser473), p42/44 MAPK, phospho-p42/44 MAPK, phospho-Rb (Ser780), and phospho-Rb (Ser807/811) (Cell Signaling Technology). The signals were detected with a Chemi-Lumi One L (Nacalai Tesque) or an Immobilon Western Chemiluminescent HRP Substrate (Millipore).
RNA isolation and real-time quantitative RT-PCR
Total RNA was isolated from cells treated with agents for 24 h using Sepasol-RNAI(Nacalai Tesque) according to the manufacturer's instructions. Total RNA (10 μg) was reverse transcribed to cDNA in a 20 μL reaction volume, with MMTV-reverse transcriptase (Promega, Madison, WI, USA) and oligo (dT) primers (Toyobo, Osaka, Japan). The reaction mixture was incubated at 37°C for 1 h. An equivalent volume (1 μL) of cDNA solution was used for the quantitative RT-PCR. cDNA was amplified by PCR using TaqMan Probes (Applied Biosystems, Foster, CA, USA) and an ABI 7300 real-time PCR system (Applied Biosystems). The expression of TS mRNA was normalized to that of β2-microgloblin (β2MG) mRNA in the same sample.
Colony formation assay
HT-29 or HCT15 cells were seeded at a density of 200 cells per well in 6-well plates. After incubating for 24 h, cells were treated with 5-FU (0.5 μM) with or without trametinib (10 nM), fenofibrate (50 μM) or LY294002 (50 μM) for 24 h. The medium was then replaced with one containing 5-FU (0.5 μM) only. After incubating for 24 h, the medium was replaced with fresh medium, and the cells were cultured for 7–10 days. The colonies fixed in 10% formalin were stained in crystal violet. The images of stained colonies were obtained using a Keyence fluorescence stereomicroscope attached to a cooled CCD camera (VB-G05, VB-7010/7000, Keyence, Osaka, Japan). The number of colonies in each well was counted with the software BZ-H1C (Keyence).
All data are presented as the mean ± SD. The statistical significance of the differences of means between three or more groups was tested using a one-way anova. That of comparison between two groups was tested using an unpaired Student's t-test.
Trametinib induces G1 arrest by reactivating RB protein
B-Raf mutant cancer cells are known to be sensitive to MEK inhibitors.[23-25] In fact, tramenitib, a selective allosteric inhibitor of MEK1/2,[9-11] showed a subnanomolar IC50 value for 72 h in a Cell Counting Kit-8 assay of human colon cancer HT-29 cells, harboring constitutively active B-Raf mutation (Fig. 1a). We then performed a cell cycle analysis and measured the sub-G1 population to quantify apoptotic cells, using flow cytometry. The treatment with trametinib for 24 h dose-dependently increased the G1 phase with a decrease in the S phase (Fig. 1b,c), and 72 h treatment induced apoptosis in a dose-dependent manner together with G1 arrest (Fig. S1a,b).
We next analyzed the expression of cell-cycle regulatory proteins by western blotting. The treatment with trametinib for 24 h induced the expression of cyclin-dependent kinase (CDK) inhibitors, p15 and p27, and reduced cyclin D1 expression, accompanied by the dephosphorylation of ERK1/2 and RB (Fig. 1d).
Trametinib reduces TS expression
Thymidylate synthase is a transcriptional target of E2F, which is activated by phosphorylation of RB. We assumed that trametinib downregulated TS expression through the RB/E2F pathway. As shown in Figure 2(a), trametinib at 10 nM or more reduced TS protein expression with dephosphorylation of RB protein. Furthermore, we performed quantitative real-time RT-PCR, and found that trametinib similarly decreased TS mRNA levels (Fig. 2b). These results suggest that MEK inhibition causes suppression of TS expression through reactivation of RB protein.
Trametinib enhances the efficacy of 5-FU on HT-29 cells
We treated HT-29 cells with a combination of 5-FU (0.5 μM) and/or trametinib (10 nM), according to the time schedule shown in Figure S2(a), and analyzed cell-cycle and sub-G1 population by flow cytometry. As shown in Figure S2(b,c), the 5-FU treatment caused S phase arrest, and the combined treatment with 5-FU and trametinib synergistically induced apoptosis. We next performed a colony formation assay using HT-29 cells by the same treatment schedule (Fig. 3a) as that in Figure S2(a). Combined treatment with 5-FU (0.5 μM) and trametinib (10 nM) markedly decreased the number of colonies compared with each drug alone (Fig. 3b,c). These results indicate that trametinib can sensitize cells to 5-FU.
Fenofibrate induces G1 arrest by blocking the MAPK cascade
Fenofibrate has been attracting attention as an antitumor drug. Actually, fenofibrate at 50 μM or more markedly inhibited the growth of HT-29 cells (Fig. 4a). Furthermore, the treatment with fenofibrate for 24 h dose-dependently increased the population of G1 phase (Fig. 4b), and also increased the populations of G1 and sub-G1 after 72 h (Fig. S3a,b).
As fenofibrate interfered with the ERK1/2 and Akt signaling pathways in a mouse melanoma cell line, we verified the status of ERK1/2 and Akt. We found that 100 μM fenofibrate after 24 h treatment caused the dephosphorylation of ERK1/2 (Fig. 4c), but not Akt (data not shown). In addition, 100 μM fenofibrate downregulated cyclin D1 expression (Fig. 4c).
Fenofibrate reduces TS expression and enhances the efficacy of 5-FU in HT-29 cells
We next examined the phosphorylation status of RB and the expression of TS, and found that fenofibrate decreased phosphorylated RB protein with downregulation of TS protein in a dose-dependent manner (Fig. 5a). The TS mRNA was also reduced by fenofibrate in a dose-dependent manner (Fig. 5b).
We then performed flow cytometry and a colony formation assay by the same treatment schedule as that by trametinib (Figs S2a and Fig. 3a). The 5-FU treatment caused S phase arrest (Fig. S4a), and the combined treatment with 5-FU and fenofibrate synergistically induced apoptosis (Fig. S4b). Regarding a colony formation assay, the combined treatment with 5-FU and fenofibrate significantly decreased colony numbers of HT-29 cells (Fig. 5c,d).
The PI3K inhibitor LY294002 reactivates RB protein and exhibits a combined effect with 5-FU on HCT15 cells
In human colon cancers, the PI3K-Akt pathway is frequently activated by a PIK3CA mutation or loss of function of PTEN.[26, 27] We then added the PI3K inhibitor LY294002 to human colon cancer HCT15 cells with active-mutated PIK3CA. As shown in Figure 6(a), LY294002 at 12.5 μM or more increased the G1 phase of the cell cycle. LY294002 upregulated p27 expression with dephosphorylation of Akt, and caused dose-dependent dephosphorylation of RB protein and a reduction in TS (Fig. 6b). We next performed flow cytometry and a colony formation assay by the same treatment schedule as that by trametinib (Fig. S2a and Fig. 3a). Whereas 5-FU (0.5 μM) alone had an affect on neither cell cycle nor sub-G1 population (Fig. S5a,b), the combined treatment with 5-FU and LY294002 (50 μM) induced modest apoptosis (Fig. S5b) and decreased colony numbers of HCT15 cells (Fig. 6c).
Cell cycle-dependent chemotherapy, with vincristine, etoposide, and so on, has been reported to show antagonism with G1-arrest inducers.[28-31] However, we previously found that CDK inhibitor SU9516 reactivated RB protein and enhanced susceptibility to 5-FU with a reduction in TS expression. Therefore, we raised the possibility that TS reduction following RB reactivation might increase the efficacy of 5-FU, known to induce DNA damage and cause S arrest and apoptosis. Actually, in the present study, we proved that each of three RB-reactivating agents, trametinib, fenofibrate, and LY294002, reduced TS expression, and the combined treatment with 5-FU resulted in induction of apoptosis and decrease in colony formation of human colon cancer cells.
A novel oral MEK inhibitor, trametinib, was originally discovered by high-throughput screening for a p15 inducer and is now being tested in multiple clinical trials without serious adverse events: i.e. phase 3; melanoma with B-Raf mutation, phase 2; pancreatic cancer, leukemia, non-small cell lung cancer (http://clinicaltrials.gov/ct2/results?term=GSK1120212).[12, 33, 34] We expect this compound to also be effective against colorectal cancers, combined with 5-FU-based chemotherapy. Furthermore, in our previous paper, we also suggested a combined effect of trametinib with oxaliplatin or SN-38 (the active metabolite of irinotecan) in HT-29 cells. Taken together, trametinib might be a good candidate for use with established regimens, FOLFOX and FOLFILI, especially for B-Raf mutated colorectal cancer.
The second compound fenofibrate, a clinically used anti-hyperlipidemia agent, also dephospholyrated ERK1/2 and enhanced sensitivity to 5-FU with a reduction of TS expression. More importantly, the effects were observed at a concentration near that in practical oral administration. Therefore, our data show the utility of fenofibrate as a safe and inexpensive 5-FU sensitizer.
Third, we combined a PI3K inhibitor LY294002 with 5-FU. About 30% of colon cancers have an active mutated PIK3CA,[27, 35] which leads to resistance to MEK inhibitors.[36, 37] Even in such cases, LY294002 decreased phosphorlylated RB protein through the induction of p27, and enhanced the efficacy of 5-FU in HCT-15 cells. Considering that anti-apoptotic proteins, Bcl-xL and MCL1, downstream of Akt, are also related with 5-FU resistance,[38, 39] the use of a PI3K inhibitor with 5-FU may be favorable, when the PI3K-Akt pathway is activated.
As mentioned, “RB reactivation therapy” could be a rational chemotherapeutic strategy to sensitize cells to 5-FU. We speculate that previously reported data on the enhancement of 5-FU sensitivity by HDAC inhibitors,[40-43] mTOR inhibitors,[44-46] and dietary factors, apigenin, quercetin, and epigallocatechin-3-gallate, might be explained by their G1-arresting ability with downregulation of TS. Interestingly, TS was clarified to be an oncogene. Especially in human colon cancer cells, a number of clinical studies have shown a correlation between high TS level and poor prognosis[51-53] or low susceptibility to 5-FU,[54-58] and we consider that colon cancer cells could be highly addictive to oncogene TS. Therefore, reduction of TS by RB-reactivating agents could suppress the oncogenic function of TS as well as enhance 5-FU's efficacy against colon cancer cells.
Another possibility that RB-reactivating agents enhanced the 5-FUs efficacy might be explained as follows. As our treatment schedule indicates in Figures 3(a) or S2(a), when the cells were treated by the combined treatment of 5-FU and RB-reactivating agents, RB-reactivating agents were supposed to cause at least partial G1-phase arrest, together with S-phase arrest by 5-FU on day 2. However, after removing the RB-reactivating agents, the G1-arrested cells were expected to progress into S-phase and were attacked by 5-FU. We therefore postulate the RB-reactivating agents could stimulate the efficacy of 5-FU by the possible mechanisms above. From this standpoint, we have to carefully evaluate a treatment schedule, if we perform a clinical trial using RB-reactivating agents with 5-FU.
In conclusion, we found that three RB-reactivating agents, trametinib, fenofibrate, and LY294002, reduced TS expression and increased sensitivity to 5-FU in human colon cancer cells. These findings suggest that “RB reactivation therapy” using molecular-targeting agents is efficacious in the 5-FU-based treatment of colorectal cancers. In other words, the addition of RB reactivators can abate the side-effects of 5-FU. Other RB reactivators might also be useful as 5-FU sensitizers. Trametinib, currently in clinical trials, is a promising adjuvant of 5-FU for patients with colorectal cancers.
This work was partly supported by the Ministry of Education, Culture, Sports, Science and Technology of Japan. We thank M. Tomosugi, H. Uchiyama, and R. Nakanishi for technical advice.