Thymidylate synthase (TYMS) is a target of the most widely used chemotherapeutic agents against gastrointestinal malignancies, the fluoropyrimidine-based therapy. TYMS expression levels have been identified as predictive biomarkers for 5-fluoruracil (FU) response in colorectal cancer, but their clinical utility remains controversial. The complexity of fluoropyrimidine response must require more mechanisms that currently have not been completely elucidated. In this context, microRNAs (miRNA) may play a role in modulating chemosensitivity. By carrying out an in silico analysis coupled to experimental validation, we detected that miR-192 and miR-215 target TYMS expression in colorectal cancer cell lines. However, downregulation of TYMS by these miRNAs does not sensitize colorectal cancer cell lines to FU treatment. The overexpression of miR-192/215 significantly reduces cell proliferation by targeting cell cycle progression. This effect was partially associated with p53 status, because reduction of cell proliferation and cell cycle arrest was associated with p21 and p27 induction. The decrease of S-phase cells by these miRNAs mitigates the effects of S phase–specific drugs and suggests that other mechanisms different from TYMS overexpression are essential to direct FU resistance. Finally, ectopic expression of miR-192/215 might have stronger impact to predict FU response than TYMS inhibition. Prospective studies to elucidate the role of these miRNAs as predictive biomarkers to FU are necessary. Mol Cancer Ther; 9(8); 2265–75. ©2010 AACR.

This article is featured in Highlights of This Issue, p. 2155

Today, one of the most important aims of oncology research is the identification of new predictive and prognostic biomarkers. Recent works have shown the value of molecular biomarkers to select individuals for various targeted therapeutics, including cetuximab in colon cancer (1). However, equivalent tools to select patients that most likely respond to commonly used chemotherapeutic drugs are lacking.

Nowadays, fluoropyrimidine-based treatments remain a major component of many standard regimens for numerous cancer types and baseline drugs in many experimental regimens with novel agents (2). Despite the introduction of 5-fluorouracil (FU), one of the most important pyrimidine antagonists, in the 1960s and its use in a wide variety of cancer types, at present we still lack useful response predictors to this drug.

FU, an analog of uracil, is an anticancer prodrug that, after administration, is intracellularly converted into 5-fluoro-2-deoxyuridin monophosphate (FdUMP), fluorodeoxyuridine triphosphate (FdUTP), and fluorouridine triphosphate (FUTP). The main toxic effects seem to be mediated by thymidylate synthase (TYMS) inhibition through the formation of an extremely stable ternary complex among FdUMP, TYMS, and the cofactor, 5,10-methylene-tetrahydrofolate (CH2FH4; refs. 3, 4). Moreover, the incorporation of FdUTP and FUTP into DNA and RNA significantly contributes to FU cytotoxicity. Several works have previously shown that the mechanism of action of FU may differ depending on the dose and schedule of administration (5, 6). Continuous infusion results in FU metabolism mainly to FdUMP, which then acts as a tightly binding inhibitor of TYMS and inhibits DNA synthesis (4). On the other hand, bolus treatment favors drug metabolism to FUTP and its incorporation into RNA that interferes with normal pre-rRNA processing (7, 8).

TYMS has been historically considered an important molecular biomarker able to predict the response to FU (913). However, the high variability of TYMS expression in colorectal cancer tissues and different methodologies used in previous works did not provide definitive conclusions to recommend the routine use of TYMS expression as a predictive molecular marker to FU treatment (14). One of the most studied mechanisms of TYMS expression variability are functional polymorphisms in the TYMS gene (15). Moreover, it is now well established that regulatory events at the level of gene amplification (16, 17), transcription (1820), posttranscription, translation, and posttranslation (21) are also involved in TYMS expression control (22). One example of TYMS posttranscriptional regulation is the rapid increase in TYMS enzyme activity described in various in vitro (2325), in vivo, and clinical model systems following short-term exposure to fluoropyrimidines (26, 27).

Based on this evidence, we hypothesized that one mechanism of TYMS posttranscriptional regulation could be mediated by microRNAs (miRNA), a new class of small noncoding RNAs involved in posttranscriptional gene regulation. Many studies have shown the important role of miRNAs as prognostic biomarkers in cancer. Moreover, new findings suggest that miRNAs could be associated with chemosensitivity in tumor cells. Therefore, the potential role of miRNAs in pharmacogenomics is currently an emerging avenue of investigation. Recent works have found that dihydrofolate reductase, a key enzyme in folate metabolism, is regulated at the translational level by miR-24 (28) and miR-192 (29). The authors showed that the alteration in this mechanism of regulation leads to methotrexate resistance.

We took a systematic approach to identify which miRNAs may target TYMS protein. We showed that miR-192/miR-215 downregulate TYMS expression. However, miR-192/miR-215 are able to modulate cancer sensibility to FU but, given the pleiotropic mechanism of action of miRNAs, mainly by targeting cell cycle progression. Finally, miR-192/miR-215 might be potential candidate biomarkers to predict FU response.

Cell culture and reagents

The human colorectal cancer cell lines RKO, LoVo, DLD1, and SW620 were obtained from the American Type Culture Collection. The DLD1 and SW620 cell lines harbor p53 mutations (mut-p53): a C>T mutation resulting in Ser→Phe at codon 241 and a G>A mutation in codon 273 resulting in an Arg→His substitution, respectively. On the contrary, RKO and LoVo cell lines have a wild-type p53 (wt-p53). In our study p53 mutation status was confirmed by sequencing. All cell lines were maintained in RPMI 1640 (GIBCO-BRL) medium supplemented with 1% penicillin-streptomycin (GIBCO-BRL) and 10% fetal bovine serum (GIBCO-BRL). FU was purchased from Sigma-Aldrich.

Dual luciferase reporter assays

For luciferase target assay, the full 3′ untranslated region (UTR) of the TYMS gene was cloned in the pLR-SV40 vector (Promega). The construct sequence was confirmed by sequencing. HEK293T cells were cotransfected with the reporter constructs (pLR-3′UTR-TYMS), pre-miR-338-5p, pre-miR-192, pre-miR-215, pre-miR-192-mutated, or a negative control (all purchased by Ambion), and pGL3-Promoter vector (Promega), by means of calcium chloride treatment. After 48 hours, cells were harvested and lysed, and ratios between renilla and firefly luciferase activity were estimated using the Dual Luciferase Assay Kit (Promega), according to the manufacturer's instructions. Data were presented as mean value of renilla/firefly luciferase ratios, obtained from at least three independent experiments.

Transfections of miRNA and small interfering RNA specific to TYMS

Pre-miR-192 (50 pmol), pre-miR-215 (50 pmol), and a scrambled control (50 pmol) were transfected into RKO (wt-p53), LoVo (wt-p53), DLD1 (mut-p53), and SW620 (mut-p53) colorectal cancer cells grown in 6-well dishes (plated at 2 × 105 cells per well 24 hours before transfection). Small interfering RNA (siRNA) specific to TYMS (On-Target plus SMART pool L-004717-00-0005) was purchased from Dharmacon. Transfection was done with Lipofectamine 2000 (Invitrogen). Transfection efficiency (>95%) was confirmed using the Silencer FAM-labeled negative control (Ambion).

RNA isolation and quantitative reverse transcription-PCR analysis of miRNA

Total RNA was isolated from cell lines by using TRIzol reagent (Invitrogen) according to the manufacturer's instructions at 24 or 48 hours after transfection. cDNA synthesis was carried out with the High-Capacity cDNA Synthesis kit (Applied Biosystems) using 10 ng total RNA as template. The miRNA sequence-specific quantitative reverse transcriptase-PCR (qRT-PCR) primers for miR-192, miR-215, and endogenous control RNU6B were purchased from Applied Biosystems. qRT-PCR analysis was carried out using Applied Biosystems 7500 Real-time PCR System. The PCR Master Mix containing TaqMan 2× Universal PCR Master Mix (No Amperase UNG), 10× TaqMan assay, and reverse transcription products in 20 μL volume were processed as follows: 95°C for 10 minutes and then 95°C for 15 seconds, 60°C for 60 seconds for up to 40 cycles. Signal was collected at the end point of every cycle. Gene expression was considered as the ΔCT values of the specific miRNAs after normalizing with the internal control RNU6B and relative quantification values were plotted.

qRT-PCR analysis of mRNA expression

cDNA was synthesized with the High-Capacity cDNA Synthesis kit (Applied Biosystems) using 2 μg total RNA as the template and random primers. Real-time qRT-PCR analysis was done on the experimental mRNAs. The PCR primers and probes for TYMS (Hs00926279_g1) and the internal control gene B2M (Hs 99999907_m1) were purchased from Applied Biosystems. qRT-PCR was done on an ABI 7500HT instrument under the following conditions: 50°C for 2 minutes of reverse transcription, 95°C for 10 minutes, 95°C for 15 seconds, 60°C for 1 minute for up to 40 cycles.

Western immunoblot analysis

At 48 hours after transfection with pre-miR-192, pre-miR-215, TYMS-siRNA, or nonspecific control miRNA, the cells were scraped and lysed in radioimmunoprecipitation assay buffer (Sigma). Equal amounts of proteins were resolved by 10% SDS-PAGE gels and transferred to nitrocellulose membranes (Bio-Rad). The membranes were then blocked with 5% nonfat-dry milk in TBS-0.5% Tween-20 at room temperature for 1 hour. The primary antibodies used for the analysis were mouse anti-TYMS (TS106, 1:400, Santa Cruz Biotechnology), mouse anti-p53 (DO-1, 1:1,000, Santa Cruz), mouse anti-p21 (DCS60, 1:2,000, Santa Cruz Biotechnology), rabbit anti-p27 (2552, 1:1,000, Cell Signaling), mouse anti-CDK6 (1:1,000, Santa Cruz Biotechnology), rabbit anti-CDK2 (1:1,000, Santa Cruz Biotechnology), rabbit anti-CCND1 (1:1,000, Santa Cruz Biotechnology), mouse anti-CCNE (14591A, 1:1,000, Pharmingen), rabbit anti-Phospho-Rb (Ser807-811, 1:1,000, Cell Signaling), mouse anti-actin (SLA20, 1:5,000, Calbiochem), and mouse anti-β-tubulin (TU-02, 1:1,000, Santa Cruz Biotechnology). Horseradish peroxidase–conjugated antibodies against mouse (1:10,000; Sigma) or against rabbit (1:10,000; Sigma) were used as the secondary antibodies. Protein bands were visualized with a chemiluminescence detection system using the Western Lightning Chemiluminescence (PerkinElmer).

FU treatment and cytotoxicity assay

Twenty-four hours after transfection with a nonspecific control miRNA, pre-miR-192, pre-miR-215, or siRNA against TYMS, all cell lines were plated in 96-well plates at 5 × 103 cells per well in triplicate. After 24 hours, an additional 100 μL of medium containing FU were added, and cells were cultured for an additional 72 hours. The concentration of FU varied in the range of 0.5 × 10-6 mol/L to 30 × 10-6 mol/L. After 72 hours, cell viability was measured with a CellTiter 96 Aqueous assay kit (Promega) according to the manufacturer's instructions. The results were expressed as a percentage of cell survival on the basis of the difference with the optical density in the control (no drug exposure). To study the modulation of TYMS expression after FU treatment, RKO, LoVo, DLD1, and SW620 cells were seeded in 6-well plates at a density of 2 × 105 per well and then incubated with or without FU (ranging from 0.5 × 10-6 mol/L to 50 × 10-6 mol/L). The cells were collected at 24 hours after incubation, and total RNA and protein were extracted. The same procedure was done for RKO, LoVo, DLD1, and SW620 after transfection with pre-miR-192, pre-miR-215, nonspecific control miRNA, or siRNA against TYMS.

Cell proliferation analysis

RKO, LoVo, DLD1, and SW620 colorectal cancer cells were plated in 96-well plates in triplicate at 1 × 103 cells per well after transfection with pre-miR-192, pre-miR-215, siRNA-TYMS, or nonspecific control miRNA. Cells were cultured for 3, 4, 6, 7, and 8 days posttransfection. Cell viability was measured with a CellTiter 96 Aqueous assay kit (Promega) according to the manufacturer's instructions.

Cell cycle analysis

Colorectal cancer cell lines were transfected with miR-192 precursor, miR-215 precursor, siRNA-TYMS, and the nonspecific control miRNA as described above. Cells were treated with 100 ng/well of bromodeoxyuridine (BrdU; BD Biosciences) for 2 hours before harvesting by trypsinization at 24 hours posttransfection, and BrdU incorporation was analyzed by flow cytometry using the FITC BrdU Flow Kit (Pharmingen). Samples were analyzed on a FACScalibur (BD Biosciences). Modfit LT (ver. 3.1) software was used to analyze the distribution of cells in the different cell cycle stages.

Statistical analysis

All experiments were repeated at least three times. Statistical significance was evaluated by Student's t test (two-tailed) comparison between two groups. The statistical analysis was done with the GraphPad Prism software and with the SPSS 13. Differences were considered statistically significant at P < 0.05.

In silico and experimental identification of TYMS miRNA binding sites

We initially explored the mRNA and protein levels of TYMS in a panel of colon cancer cells. We found that the levels of TYMS mRNA in colon cancer cells did not correlate which TYMS protein expression levels (Supplementary Fig. S1A and B), suggesting that a posttranscriptional regulation of TYMS could be feasible. Thus, we hypothesized that miRNAs could be likely involved in its translational regulation.

To further explore this possibility, we carried out an in silico analysis of those miRNAs that could regulate TYMS expression. Based on the structural analysis of the 3′-UTR of the TYMS gene and using different miRNA predicting target programs (TargetScan, miRBASE, and miRANDA), we identified several miRNAs that potentially interact with the 3′-UTR of the TYMS mRNA. To avoid overprediction, results coming from different databases were intersected and only those miRNAs predicted to target TYMS expression by at least two of the algorithms were considered. Results are shown in Supplementary Table S1.

We first focused on miR-338-5p that was predicted to regulate TYMS by the three algorithms used (Supplementary Table S1). To show a direct interaction between miR-338-5p and TYMS mRNA, we inserted the full-3′-UTR of TYMS mRNA downstream a renilla luciferase reporter cDNA. The HEK293T cells were transfected with this vector together with a pre-miR-338-5p or a control pre-miR. As shown in Fig. 1A, the measurement of renilla activity relative to firefly luciferase did not show a significant repression by miR-338-5p, suggesting that miR-338-5p does not interact with the 3′-UTR of the TYMS mRNA.

Figure 1.

Identification of miR-192 and miR-215 as regulators of TYMS expression. A, HEK293T cell line was transfected with pre-miR338-5p, pre-miR-192, pre-miR-215, pre-miR-192-mutated, or scrambled control; the renilla reporter vector-3′UTR-TYMS and pGL3-Promoter vector were used as normalization control. The relative renilla/luciferase activity was plotted. Results are represented as mean ± SD for three independent experiments. B, expression analysis of TYMS mRNA by real-time-PCR in RKO, LoVo, DLD1, and SW620 cells transfected with a nonspecific miR control (scrambled), pre-miR-192, pre-miR-215, or TYMS-siRNA. Data are expressed as 2-ΔΔCt values obtained by normalization using β2-microglobulin as endogenous control and cells transfected with a nonspecific miR control as calibrator. C, Western-blot and densitometric analysis of TYMS protein expression in RKO, LoVo, DLD1, and SW620 cells transfected with a nonspecific miR control, pre-miR-192, pre-miR-215, or TYMS-siRNA; β-tubulin was used as protein loading control.

Figure 1.

Identification of miR-192 and miR-215 as regulators of TYMS expression. A, HEK293T cell line was transfected with pre-miR338-5p, pre-miR-192, pre-miR-215, pre-miR-192-mutated, or scrambled control; the renilla reporter vector-3′UTR-TYMS and pGL3-Promoter vector were used as normalization control. The relative renilla/luciferase activity was plotted. Results are represented as mean ± SD for three independent experiments. B, expression analysis of TYMS mRNA by real-time-PCR in RKO, LoVo, DLD1, and SW620 cells transfected with a nonspecific miR control (scrambled), pre-miR-192, pre-miR-215, or TYMS-siRNA. Data are expressed as 2-ΔΔCt values obtained by normalization using β2-microglobulin as endogenous control and cells transfected with a nonspecific miR control as calibrator. C, Western-blot and densitometric analysis of TYMS protein expression in RKO, LoVo, DLD1, and SW620 cells transfected with a nonspecific miR control, pre-miR-192, pre-miR-215, or TYMS-siRNA; β-tubulin was used as protein loading control.

Close modal

Next, we tried miR-192 and miR-215, and we used again the full-3′-UTR of the TYMS mRNA containing the putative binding site for these miRNAs (Supplementary Table S1). In this situation, the measurement of renilla activity relative to firefly luciferase activity showed a significant repression by miR-192 and miR-215 (Fig. 1A). Moreover, to validate the direct regulation of TYMS due to miR-192/215, we showed that the precursor of miR-192 mutated in the “seed region” was unable to bind the predictive 3′-UTR region of TYMS (Fig. 1A). These data indicate that both miR-192 and miR-215 interact with the 3′-UTR of the TYMS mRNA in a specific manner.

Translational regulation of TYMS by miR-192 and miR-215

To experimentally confirm that the expression of TYMS was regulated by miR-192/miR-215 in colorectal cancer, the precursors of miR-192 or miR-215 were transfected into RKO, LoVo, DLD1, and SW620 colorectal cancer cells and the expression of the TYMS protein was evaluated by Western immunoblot analysis. To verify the efficiency of transfection, the overexpression of miR-192 and miR-215 was confirmed by qRT-PCR analysis using RNAU6B for normalization (Supplementary Fig. S2). A nonspecific miRNA was used as a negative control and a siRNA for TYMS as positive control. As shown in Fig. 1B, the ectopic expression of miR-192 and miR-215 did not induce an important reduction in TYMS mRNA expression. By contrast, miR-192 and miR-215 clearly decreased TYMS protein expression (Fig. 1C) in all cell lines analyzed. Moreover, we observed a trend to an inverse correlation between miR-192/miR-215 expression levels and TYMS protein in colorectal cancer cell lines (Supplementary Fig. S1C and D).

These results confirm, in agreement with our initial hypothesis, that a posttranscriptional regulation of TYMS occurs and miR-192 and miR-215 are involved in this regulation.

miR-192 and miR-215 do not sensitize colorectal cancer cells to FU treatment

Given that TYMS is the main target for FU treatment, we hypothesized that the increased expression of miR-192 and/or miR-215 might contribute to FU sensitivity by reducing TYMS protein expression. Thus, we used the FU treatment schedule that could be compared with continuous infusion and that has been mainly associated with TYMS inhibition. We tested this hypothesis by an in vitro chemosensitivity assay 48 hours posttransfection with the miRNAs. As shown in Supplementary Fig. S3, overexpression of miR-192 or miR-215 did not change the sensitivity of colorectal cancer cell lines to FU treatment. In contrast, chemosensitivity of RKO, LoVo, and SW620 cells lines treated with the siRNA against TYMS was increased compared with the corresponding scrambled control cells but not in DLD1 cells. The IC50 value for FU was significantly lower in RKO, LoVo, and SW620 transfected with siRNA-TYMS than in the corresponding control cells (1.08 μmol/L versus 0.5 μmol/L for RKO, 0.93 μmol/L versus 0.39 μmol/L for LoVo, and 1.63 μmol/L versus 0.62 μmol/L for SW620).

Although miR-192/215 regulate TYMS at posttranscriptional level, these miRNAs are unable to sensitize colorectal cancer cells to FU. Thus, we investigated which other additional mechanisms mediated by these miRNAs could influence the response to FU. It is known that the active metabolite of FU, FdUMP, forms a ternary complex with TYMS and its reduced-folate cofactor. We explored the effect of miR192/215 on the formation of this ternary complex. We exposed colorectal cancer cell lines to different concentrations of FU for 24 hours. The treatment of the RKO and LoVo cell lines led to the formation of the inhibited ternary complex at 38.5 kDa and continued expression of the free active enzyme at 36 kDa (Fig. 2A). Treatment with 10 umol/L FU for 24 hours of siRNA-TYMS transfected cell lines completely abolished the formation of the inhibited ternary complex and the free active enzyme (Fig. 2B). We also detected an important reduction in the expression of both TYMS forms (ternary complex and free protein) after FU treatment of miR-192/215–transfected cell lines (Fig. 2B). These data suggest that miR-192/215 were also able to reduce the inhibited ternary complex of TYMS, although it was not enough to increase FU cytotoxicity.

Figure 2.

Modulation of TYMS expression after FU treatment: Regulation by miR-192/215 transfected cells. A, RKO and LoVo cells were treated with the indicated concentrations of FU. TYMS expression was measured by Western blot and is represented by free TYMS (36 kDa) and ternary complex of TYMS, 5,10-methylenetetrahydrofolate, and FdUMP (38.5 kDa). β-Tubulin was used as a protein loading control. B, Western immunoblot analysis of free TYMS (36 kDa) and ternary complex of TYMS, 5,10-methylenetetrahydrofolate, and FdUMP (38.5 kDa) in colorectal cancer cell lines transfected with pre-miR-192, pre-miR-215, TYMS-siRNA, or negative control and treated with 10 umol/L of FU for 24 hours. β-Tubulin was used as a protein loading control.

Figure 2.

Modulation of TYMS expression after FU treatment: Regulation by miR-192/215 transfected cells. A, RKO and LoVo cells were treated with the indicated concentrations of FU. TYMS expression was measured by Western blot and is represented by free TYMS (36 kDa) and ternary complex of TYMS, 5,10-methylenetetrahydrofolate, and FdUMP (38.5 kDa). β-Tubulin was used as a protein loading control. B, Western immunoblot analysis of free TYMS (36 kDa) and ternary complex of TYMS, 5,10-methylenetetrahydrofolate, and FdUMP (38.5 kDa) in colorectal cancer cell lines transfected with pre-miR-192, pre-miR-215, TYMS-siRNA, or negative control and treated with 10 umol/L of FU for 24 hours. β-Tubulin was used as a protein loading control.

Close modal

We finally assumed that miR-192/215 did not increase FU sensitivity due to additional mechanisms mediated by these miRNAs or to the fact that translation of the TYMS protein was not completely inhibited. To explore these hypotheses we used two different approaches. We first increased the time of miR-192 transfection before FU treatment. When we did a transfection with pre-miR-192 for 96 hours, we observed that the overexpression of miR-192 did not sensitize cells to FU but, on the contrary, increased resistance to FU (Fig. 3). In the second approach, we carried out a sequential transfection with pre-miR-192 for 48 hours and subsequent siRNA-TYMS inhibition. In this case, we observed that, although a completely TYMS inhibition was obtained, the overexpression of miR-192 previous to the silencing by siRNA-TYMS was able to reduce the chemosensitivity induced by siRNA-TYMS in RKO and LoVo cell lines. These results show that the ectopic expression of miR-192 decreased FU chemosensitivity, despite complete abrogation of TYMS expression. In this context, we supposed that additional events induced by knockdown effect in proteins targeted by miR-192 are involved in cellular response to FU, independently of TYMS inhibition.

Figure 3.

Chemosensitivity to FU of RKO and LoVo colorectal cancer cells transfected with a nonspecific miR control, pre-miR-192 for 96 hours, TYMS-siRNA, or sequential cotransfection of pre-miR-192 and TYMS-siRNA was measured using a MTS assay kit. Transfected cells (5,000 cells/100 uL of medium per well) were seeded in triplicate in 96-well flat-bottom plates. After 24 hours, an additional 100 uL of medium containing FU were added, and the cells were cultured for an additional 72 hours. After 72 hours, cells were treated with 20 uL of CellTiter 96 Aqueous assay kit and after 2 hours the optical density of the solution was measured at 495 nm. The results were expressed as the percentage of cell survival on the basis of the difference between the optical density of untreated cells and treated cells. Each point, mean of three different experiments.

Figure 3.

Chemosensitivity to FU of RKO and LoVo colorectal cancer cells transfected with a nonspecific miR control, pre-miR-192 for 96 hours, TYMS-siRNA, or sequential cotransfection of pre-miR-192 and TYMS-siRNA was measured using a MTS assay kit. Transfected cells (5,000 cells/100 uL of medium per well) were seeded in triplicate in 96-well flat-bottom plates. After 24 hours, an additional 100 uL of medium containing FU were added, and the cells were cultured for an additional 72 hours. After 72 hours, cells were treated with 20 uL of CellTiter 96 Aqueous assay kit and after 2 hours the optical density of the solution was measured at 495 nm. The results were expressed as the percentage of cell survival on the basis of the difference between the optical density of untreated cells and treated cells. Each point, mean of three different experiments.

Close modal

Overexpression of miR-192 and miR-215 reduces cell proliferation in colorectal cancer cell lines

While doing the experiments described above, we appreciated that transfection with miR-192 and miR-215 induced important morphologic changes in the colorectal cancer cell lines studied. In response to these miRNAs, the cells acquired a round shape. In contrast, cells transfected with siRNA-TYMS displayed a similar morphology to control cells (Supplementary Fig. S4).

To assess additional cellular functions of miR-192 and miR-215, we evaluated the effect of both miRNA on cellular proliferation in colorectal cancer cell lines. Our results showed that the overexpression of miR-192 and miR-215 can reduce cellular proliferation in RKO (wt-p53, Fig. 4A), LoVo (wt-p53, Fig. 4B), and DLD1 (mut-p53, Fig. 4C) with less effect on SW620 (mut-p53, Fig. 4D). These results show the effect of miR-192 and miR-215 on the inhibition of cellular proliferation in a manner partially but not completely p53-dependent pathway.

Figure 4.

Effect of miR-192, miR-215, and siRNA-TYMS on colorectal cancer cell line proliferation. Each cell type was transfected with pre-miR-192 (50 nmol/L), pre-miR-215 (50 nmol/L), TYMS-siRNA (50 nmol/L), or negative control (50 nmol/L), and the number of cells was determined by the CellTiter 96 Aqueous assay kit after 3, 4, 6, 7, and 8 days posttransfection.

Figure 4.

Effect of miR-192, miR-215, and siRNA-TYMS on colorectal cancer cell line proliferation. Each cell type was transfected with pre-miR-192 (50 nmol/L), pre-miR-215 (50 nmol/L), TYMS-siRNA (50 nmol/L), or negative control (50 nmol/L), and the number of cells was determined by the CellTiter 96 Aqueous assay kit after 3, 4, 6, 7, and 8 days posttransfection.

Close modal

Effect of miR-192 and miR-215 on cell cycle control

To determine whether the effect of miR-192 and miR-215 on cell proliferation was related to cell cycle control, we analyzed the effect of both miRNAs on cell cycle by flow cytometry.

We pulsed miR-192, miR-215, and siRNA-TYMS transfected cells with the thymidine analogue BrdU to assay the DNA synthesis defects. Our results show that miR-192 and miR-215 significantly increased the number of cells in G0-G1 phase and reduced S-phase cells in LoVo (wt-p53) and RKO (wt-p53) cell lines (Fig. 5A and B). However, in the DLD1 (mut-p53) cell line we did not observe any change in the number of cells in S-phase but we detected an increase of cells in G2-M. In contrast, no effect was observed in the SW-620 cell line (Fig. 5B). These data are in agreement with our previous results with the proliferation assays. On the other hand, we noted that the transfection with the siRNA-TYMS induced a strong arrest of all analyzed cells in the S-phase except for that in DLD-1 cell lines (Fig. 5B). Interestingly, this is the only cell line in which transfection with the siRNA against TYMS did not increase sensitivity to FU (Supplementary Fig. S3). These data strongly suggest that the sensitivity of colorectal cancer cell lines to FU treatment depends on the cell cycle status rather than on TYMS expression. In particular the S-phase of cycle seems to be the major determinant. In this regard, we analyzed BrdU incorporation in the sequential transfection approach (miR-192 and siRNA-TYMS) and we detected an arrest in G1 phase that prevented progression into the S phase and consequently induced a reduction of cells in S phase (Fig. 5C).

Figure 5.

miR-192 and miR-215 cause accumulation of cells at G0-G1 and G2-M cell cycle checkpoints. A, miR-192 induces a defect in the G1-S phase transition and siRNA-TYMS arrests cells in S-phase. Control-transfected (left), miR-192–transfected (center), and siRNA-TYMS–transfected (right) LoVo colon cancer cells were analyzed for 5-BrdU incorporation using flow cytometry. B, representation of the percentages of cells in G0-G1, S, and G2M by Modfit analysis of data from each cell line transfected with pre-miR-192, pre-miR-215, TYMS-siRNA, or negative control. C, flow cytometry analysis of the cell cycle distribution of the RKO cell line following siRNA-TYMS transfection and sequential transfection with pre-miR-192 and siRNA-TYMS.

Figure 5.

miR-192 and miR-215 cause accumulation of cells at G0-G1 and G2-M cell cycle checkpoints. A, miR-192 induces a defect in the G1-S phase transition and siRNA-TYMS arrests cells in S-phase. Control-transfected (left), miR-192–transfected (center), and siRNA-TYMS–transfected (right) LoVo colon cancer cells were analyzed for 5-BrdU incorporation using flow cytometry. B, representation of the percentages of cells in G0-G1, S, and G2M by Modfit analysis of data from each cell line transfected with pre-miR-192, pre-miR-215, TYMS-siRNA, or negative control. C, flow cytometry analysis of the cell cycle distribution of the RKO cell line following siRNA-TYMS transfection and sequential transfection with pre-miR-192 and siRNA-TYMS.

Close modal

Finally, to investigate how these miRNAs affect the cell cycle, we analyzed several key cell cycle–related genes (p53, p21, CCND1, CCNE, CDK4, CDK6, CDK2, and Rb-P). Western blot analysis (Fig. 6A) showed a marked upregulation of p53, p21, and p27 protein expression after ectopic expression of miR-192 and miR-215 in wt-p53 colorectal cancer cell lines (RKO and LoVo) but not in mut-p53 DLD1 and SW620 cell lines (Fig. 6B), confirming the link between these miRNAs and p53. As a consequence of p21 and p27 induction, we detected a reduction of hyper-phosphorylation of Rb in wt-p53 cells (RKO and LoVo) but not in mut-p53 cell lines (DLD1 and SW620). Moreover, in LoVo and RKO cells we observed an increased expression of CCND1 and a reduction of CDK6 protein expression without changes in CDK2 or CCNE. On the other hand, in the p53-mutated cell line DLD1 we only detected a significant reduction in the expression of CCNE and an accumulation of CCND1. In contrast, siRNA against TYMS did not change the expression of the proteins involved in cell cycle control.

Figure 6.

Western immunoblot analysis of cell cycle control genes. A, Western immunoblot analysis of p53, p21, CDK4, CDK6, CCND1, CDK2, CCNE, and Phospho-Rb in wild-type p53 colorectal cancer cell lines (RKO and LoVo) 48 hours after transfection with pre-miR-192, pre-miR-215, TYMS-siRNA, or negative control. α-Actin and β-tubulin were used as protein loading controls. B, Western immunoblot analysis of p53, p21, CDK4, CDK6, CCND1, CDK2, CCNE, and Phospho-Rb in mutated p53 colorectal cancer cell lines (SW620 and DLD1) transfected with pre-miR-192, pre-miR-215, TYMS-siRNA, or negative control, and treated with 10 umol/L of FU for 24 hours. α-Actin and β-tubulin were used as protein loading controls.

Figure 6.

Western immunoblot analysis of cell cycle control genes. A, Western immunoblot analysis of p53, p21, CDK4, CDK6, CCND1, CDK2, CCNE, and Phospho-Rb in wild-type p53 colorectal cancer cell lines (RKO and LoVo) 48 hours after transfection with pre-miR-192, pre-miR-215, TYMS-siRNA, or negative control. α-Actin and β-tubulin were used as protein loading controls. B, Western immunoblot analysis of p53, p21, CDK4, CDK6, CCND1, CDK2, CCNE, and Phospho-Rb in mutated p53 colorectal cancer cell lines (SW620 and DLD1) transfected with pre-miR-192, pre-miR-215, TYMS-siRNA, or negative control, and treated with 10 umol/L of FU for 24 hours. α-Actin and β-tubulin were used as protein loading controls.

Close modal

Regulation by miRNAs is obviously not as simple as one miRNA-one mRNA target gene, and knowledge on the gene regulation network by miRNA is just beginning. Typically, more than one miRNA regulates one mRNA, and one miRNA may have several target genes, reflecting a cooperative translational control and target multiplicity. In the present report, we show that miR-192/215 downregulate TYMS expression posttranscriptionally but due to their pleiotropic effects these miRNAs increase FU resistance in colorectal cancer cell lines. Moreover, we showed that miR-192/miR-215 prevent progression into the S phase and are involved in cell cycle control at multiple levels and subsequently avoid FU cytotoxicity.

It is known that TYMS plays an essential role in catalyzing the reductive methylation of deoxyuridilate (dUMP) to thymidilate (dTMP), which provides the sole intracellular de novo source of dTMP. Due to its biological role, TYMS is critically involved in normal function of the cell cycle and consequently represents an attractive target for anticancer chemotherapy agents such as FU. Although the reason for TYMS expression variability is still unclear, the regulation of the transcriptional and translational activity of this gene likely plays an important role in cell chemosensitivity.

We suppose that miRNAs might be involved in TYMS posttranscriptional regulation. Thus, first by in silico and in vitro expression analysis, we found that miR-192/215 could be potentially implicated in TYMS regulation. We experimentally confirmed that miR-192/miR-215 target TYMS at the 3′-UTR and therefore they might have a key role in the posttranscriptional regulation of TYMS expression. We reasoned that these miRNAs targeting TYMS could have a relevant effect on FU sensitivity. However, we showed that the reduction of TYMS protein expression was not associated with an increase of FU sensitivity. On the contrary, miR-192 overexpression increased the resistance to FU. Our results highlight the role of miR192/215 as key determinants of chemosensitivity to FU but mainly due to their pleiotropic effect rather than their direct effect upon TYMS expression.

The functional analysis of miR-192/miR-215 revealed that they could induce the accumulation of p53 and its target gene products p21 and p27, and that they suppressed proliferation in a partially but not completely p53-dependent pathway. This observation is supported by our data, showing a decreased growth rate due to miR-192/miR-215 overexpression in the p53-mutated DLD1 cell line. Moreover, we showed that the decrease in cell growth in this mutated-p53 cell line might be explained by cell cycle arrest in the G2-M phase. On the contrary, in wild-type p53 cell lines (RKO and LoVo) transfected with miR-192/215 we detected an important cell cycle arrest in the G0-G1 phase. In RKO and LoVo cell lines, p21 and p27 induction could inhibit CDK activity and consequently these cells would be unable to phosphorylate pRb, one of the most important cell cycle checkpoints. In DLD1-p53-mutated cell line, we detected a decrease in CCNE protein expression. Therefore, the decrease in cell growth after miR-192/miR-215 overexpression seems to be due to a general regulation of cell cycle genes. In this sense, recent reports showed that miR-192/miR-215 halt cell cycle progression by coordinately targeting transcripts that play critical roles in mediating the G1-S and G2-M checkpoints (30, 31). In addition, to regulate cell cycle–related genes directly, miR-192/miR-215 could also induce arrest by targeting genes that consequently activate the p53-p21 pathway. In cell line models, inactivation of the p21 gene leads to cell cycle arrest failure and apoptosis after DNA-damaging chemotherapy; conversely, overexpression of p21 may lead to the impossibility of the cell to activate S-phase and prevent treatment-induced DNA damage. These data suggest that induction of p21 and/or p27 by miR-192/miR-215 could represent a marker of resistance to chemotherapy. In this sense, recent clinical trials have shown that p21 overexpression (32) and low proliferative index are both tumor characteristics associated with reduced response to FU treatment (33).

Recent works provide new limits to TYMS as a predictive determinant for fluoropyrimidine sensitivity (34), and to date the role of TYMS as a putative predictive and prognostic biomarker in gastrointestinal cancer remains controversial (3538). In fact, some authors have proposed the proliferation marker Ki-67 as a major determinant of FU clinical response (39). In agreement with these reports, our results highlight the limited role of TYMS expression as predictor marker to FU response and other TYMS-independent events must be involved in FU resistance.

The functional activity of miR-192/215 within colorectal cancer cell lines suggests a potential role of these miRNAs as tumor suppressors. In agreement with this hypothesis, recent studies have shown that miR-192/miR-215 expression levels were decreased in clinical colon cancer specimens compared with adjacent normal tissues of the same patients (30), indicating that reduced levels of these miRNAs in tumors may be contributing to the malignant phenotype. Moreover, the gene expression profile induced by miR-192/miR-215 is able to control the cell cycle progression (40). It has previously been shown that miR-192 target dihydrofolate reductase expression, a key enzyme of folate metabolism. We have shown that miR-192/215 also regulate TYMS protein expression. We could hypothesize that these targets of miR-192/215, directly involved in DNA synthesis, may affect p53 activity and ultimately alter cell cycle progression.

In conclusion, we provide direct evidence that miR-192 and miR-215 influence sensitivity to FU treatment. Other variables than TYMS expression, such as the cell cycle status, could be considered as major factors that modulate the sensitivity of colorectal cancer cells to FU exposure. Our findings suggest that these miRNAs may be useful FU predictive biomarkers; however, additional studies in the clinical setting are clearly warranted.

Finally, the therapeutic modulation of these miRNAs might be worth exploring in an attempt to improve drug efficacy by overcoming mechanisms of resistance. However, it is mandatory to previously know the complexity of miRNA regulation, because multiple miRNAs may affect multiple genes in a combinatorial fashion, with a fine-tuning depending on the state of the cell. In the future this knowledge should be used to design clinical studies with the aim of delivering more effective treatments to selected patients.

No potential conflicts of interest were disclosed.

We thank Dr. A. Patiño and Dr. X. Agirre for their critical reading and discussion of the paper, and Ines Lopez and Marisol Gonzalez for their excellent technical assistance.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1
Karapetis
CS
,
Khambata-Ford
S
,
Jonker
DJ
, et al
. 
K-ras mutations and benefit from cetuximab in advanced colorectal cancer
.
N Engl J Med
2008
;
359
:
1757
65
.
2
Meyerhardt
JA
,
Mayer
RJ
. 
Systemic therapy for colorectal cancer
.
N Engl J Med
2005
;
352
:
476
87
.
3
Pestalozzi
BC
,
McGinn
CJ
,
Kinsella
TJ
, et al
. 
Increased thymidylate synthase protein levels are principally associated with proliferation but not cell cycle phase in asynchronous human cancer cells
.
Br J Cancer
1995
;
71
:
1151
7
.
4
Sommer
H
,
Santi
DV
. 
Purification and amino acid analysis of an active site peptide from thymidylate synthetase containing covalently bound 5-fluoro-2′-deoxyuridylate and methylenetetrahydrofolate
.
Biochem Biophys Res Commun
1974
;
57
:
689
95
.
5
Diasio
RB
,
Harris
BE
. 
Clinical pharmacology of 5-fluorouracil
.
Clin Pharmacokinet
1989
;
16
:
215
37
.
6
Aschele
C
,
Sobrero
A
,
Faderan
MA
,
Bertino
JR
. 
Novel mechanism(s) of resistance to 5-fluorouracil in human colon cancer (HCT-8) sublines following exposure to two different clinically relevant dose schedules
.
Cancer Res
1992
;
52
:
1855
64
.
7
Sobrero
AF
,
Aschele
C
,
Bertino
JR
. 
Fluorouracil in colorectal cancer - a tale of two drugs: implications for biochemical modulation
.
J Clin Oncol
1997
;
15
:
368
81
.
8
Pritchard
DM
,
Watson
AJ
,
Potten
CS
,
Jackman
AL
,
Hickman
JA
. 
Inhibition by uridine but not thymidine of p53-dependent intestinal apoptosis initiated by 5-fluorouracil: evidence for the involvement of RNA perturbation
.
Proc Natl Acad Sci U S A
1997
;
94
:
1795
9
.
9
Johnston
PG
,
Fisher
ER
,
Rockette
HE
, et al
. 
The role of thymidylate synthase expression in prognosis and outcome of adjuvant chemotherapy in patients with rectal cancer
.
J Clin Oncol
1994
;
12
:
2640
7
.
10
Boku
N
,
Chin
K
,
Hosokawa
K
, et al
. 
Biological markers as a predictor for response and prognosis of unresectable gastric cancer patients treated with 5-fluorouracil and cis-platinum
.
Clin Cancer Res
1998
;
4
:
1469
74
.
11
Pestalozzi
BC
,
Peterson
HF
,
Gelber
RD
, et al
. 
Prognostic importance of thymidylate synthase expression in early breast cancer
.
J Clin Oncol
1997
;
15
:
1923
31
.
12
Leichman
CG
,
Lenz
HJ
,
Leichman
L
, et al
. 
Quantitation of intratumoral thymidylate synthase expression predicts for disseminated colorectal cancer response and resistance to protracted-infusion fluorouracil and weekly leucovorin
.
J Clin Oncol
1997
;
15
:
3223
9
.
13
Kwon
HC
,
Roh
MS
,
Oh
SY
, et al
. 
Prognostic value of expression of ERCC1, thymidylate synthase, and glutathione S-transferase P1 for 5-fluorouracil/oxaliplatin chemotherapy in advanced gastric cancer
.
Ann Oncol
2007
;
18
:
504
9
.
14
Locker
GY
,
Hamilton
S
,
Harris
J
, et al
. 
ASCO 2006 update of recommendations for the use of tumor markers in gastrointestinal cancer
.
J Clin Oncol
2006
;
24
:
5313
27
.
15
Horie
N
,
Aiba
H
,
Oguro
K
,
Hojo
H
,
Takeishi
K
. 
Functional analysis and DNA polymorphism of the tandemly repeated sequences in the 5′-terminal regulatory region of the human gene for thymidylate synthase
.
Cell Struct Funct
1995
;
20
:
191
7
.
16
Berger
SH
,
Jenh
CH
,
Johnson
LF
,
Berger
FG
. 
Thymidylate synthase overproduction and gene amplification in fluorodeoxyuridine-resistant human cells
.
Mol Pharmacol
1985
;
28
:
461
7
.
17
Clark
JL
,
Berger
SH
,
Mittelman
A
,
Berger
FG
. 
Thymidylate synthase gene amplification in a colon tumor resistant to fluoropyrimidine chemotherapy
.
Cancer Treat Rev
1987
;
71
:
261
5
.
18
Dong
S
,
Lester
L
,
Johnson
LF
. 
Transcriptional control elements and complex initiation pattern of the TATA-less bidirectional human thymidylate synthase promoter
.
J Cell Biochem
2000
;
77
:
50
64
.
19
Navalgund
LG
,
Rossana
C
,
Muench
AJ
,
Johnson
LF
. 
Cell cycle regulation of thymidylate synthetase gene expression in cultured mouse fibroblasts
.
J Biol Chem
1980
;
255
:
7386
90
.
20
Jenh
CH
,
Geyer
PK
,
Johnson
LF
. 
Control of thymidylate synthase mRNA content and gene transcription in an overproducing mouse cell line
.
Mol Cell Biol
1985
;
5
:
2527
32
.
21
Kitchens
ME
,
Forsthoefel
AM
,
Rafique
Z
,
Spencer
HT
,
Berger
FG
. 
Ligand-mediated induction of thymidylate synthase occurs by enzyme stabilization. Implications for autoregulation of translation
.
J Biol Chem
1999
;
274
:
12544
7
.
22
Jenh
CH
,
Rao
LG
,
Johnson
LF
. 
Regulation of thymidylate synthase enzyme synthesis in 5-fluorodeoxyuridine-resistant mouse fibroblasts during the transition from the resting to growing state
.
J Cell Physiol
1985
;
122
:
149
54
.
23
Chu
E
,
Koeller
DM
,
Johnston
PG
,
Zinn
S
,
Allegra
CJ
. 
Regulation of thymidylate synthase in human colon cancer cells treated with 5-fluorouracil and interferon-γ
.
Mol Pharmacol
1993
;
43
:
527
33
.
24
Chu
E
,
Zinn
S
,
Boarman
D
,
Allegra
CJ
. 
Interaction of γ interferon and 5-fluorouracil in the H630 human colon carcinoma cell line
.
Cancer Res
1990
;
50
:
5834
40
.
25
Welsh
SJ
,
Titley
J
,
Brunton
L
, et al
. 
Comparison of thymidylate synthase (TS) protein up-regulation after exposure to TS inhibitors in normal and tumor cell lines and tissues
.
Clin Cancer Res
2000
;
6
:
2538
46
.
26
Spears
CP
,
Shahinian
AH
,
Moran
RG
,
Heidelberger
C
,
Corbett
TH
. 
In vivo kinetics of thymidylate synthetase inhibition of 5-fluorouracil-sensitive and -resistant murine colon adenocarcinomas
.
Cancer Res
1982
;
42
:
450
6
.
27
Keyomarsi
K
,
Samet
J
,
Molnar
G
,
Pardee
AB
. 
The thymidylate synthase inhibitor, ICI D1694, overcomes translational detainment of the enzyme
.
J Biol Chem
1993
;
268
:
15142
9
.
28
Mishra
PJ
,
Humeniuk
R
,
Mishra
PJ
,
Longo-Sorbello
GS
,
Banerjee
D
,
Bertino
JR
. 
A miR-24 microRNA binding-site polymorphism in dihydrofolate reductase gene leads to methotrexate resistance
.
Proc Natl Acad Sci U S A
2007
;
104
:
13513
8
.
29
Song
B
,
Wang
Y
,
Kudo
K
,
Gavin
EJ
,
Xi
Y
,
Ju
J
. 
miR-192 Regulates dihydrofolate reductase and cellular proliferation through the p53-microRNA circuit
.
Clin Cancer Res
2008
;
14
:
8080
6
.
30
Braun
CJ
,
Zhang
X
,
Savelyeva
I
, et al
. 
p53-Responsive micrornas 192 and 215 are capable of inducing cell cycle arrest
.
Cancer Res
2008
;
68
:
10094
104
.
31
Georges
SA
,
Biery
MC
,
Kim
SY
, et al
. 
Coordinated regulation of cell cycle transcripts by p53-Inducible microRNAs, miR-192 and miR-215
.
Cancer Res
2008
;
68
:
10105
12
.
32
Rau
B
,
Sturm
I
,
Lage
H
, et al
. 
Dynamic expression profile of p21WAF1/CIP1 and Ki-67 predicts survival in rectal carcinoma treated with preoperative radiochemotherapy
.
J Clin Oncol
2003
;
21
:
3391
401
.
33
Cascinu
S
,
Graziano
F
,
Valentini
M
, et al
. 
Vascular endothelial growth factor expression, S-phase fraction and thymidylate synthase quantitation in node-positive colon cancer: relationships with tumor recurrence and resistance to adjuvant chemotherapy
.
Ann Oncol
2001
;
12
:
239
44
.
34
Brody
JR
,
Hucl
T
,
Costantino
CL
, et al
. 
Limits to thymidylate synthase and TP53 genes as predictive determinants for fluoropyrimidine sensitivity and further evidence for RNA-based toxicity as a major influence
.
Cancer Res
2009
;
69
:
984
91
.
35
Johnston
PG
. 
Prognostic markers of local relapse in rectal cancer: are we any further forward?
J Clin Oncol
2006
;
24
:
4049
50
.
36
Popat
S
,
Matakidou
A
,
Houlston
RS
. 
Thymidylate synthase expression and prognosis in colorectal cancer: a systematic review and meta-analysis
.
J Clin Oncol
2004
;
22
:
529
36
.
37
Johnston
PG
,
Benson
AB
 III
,
Catalano
P
,
Rao
MS
,
O'Dwyer
PJ
,
Allegra
CJ
. 
Thymidylate synthase protein expression in primary colorectal cancer: lack of correlation with outcome and response to fluorouracil in metastatic disease sites
.
J Clin Oncol
2003
;
21
:
815
9
.
38
Edler
D
,
Hallström
M
,
Anders
Jakobsen
,
Johnston
PG
,
Magnusson
I
,
Ragnhammar
P
,
Blomgren
H
. 
Thymidylate synthase expression: an independent prognostic factor for local recurrence, distant metastasis, disease-free and overall survival in rectal cancer
.
Clin Cancer Res
2000
;
6
:
1378
84
.
39
Garrity
MM
,
Burgart
LJ
,
Mahoney
MR
, et al
. 
Prognostic value of proliferation, apoptosis, defective DNA mismatch repair, and p53 overexpression in patients with resected Dukes' B2 or C colon cancer: a North Central Cancer Treatment Group Study
.
J Clin Oncol
2004
;
22
:
1572
82
.
40
Wang
W
,
Cassidy
J
,
O'Brien
V
,
Ryan
KM
,
Collie-Duguid
E
. 
Mechanistic and predictive profiling of 5-fluorouracil resistance in human cancer cells
.
Cancer Res
2004
;
64
:
8167
76
.