Báo cáo khoa học: MicroRNA-143 reduces viability and increases sensitivity to 5-fluorouracil in HCT116 human colorectal cancer cells

MicroRNA-143 reduces viability and increases sensitivity to 5-fluorouracil in HCT116 human colorectal cancer cells Pedro M. Borralho1, Betsy T. Kren2, Rui E. Castro1, Isabel B. Moreira da Silva1, Clifford J. Steer2,3 and Cecı´lia M. P. Rodrigues1

1 Research Institute for Medicines and Pharmaceutical Sciences, Faculty of Pharmacy, University of Lisbon, Portugal 2 Department of Medicine, University of Minnesota Medical School, Minneapolis, MN, USA 3 Department of Genetics, Cell Biology and Development, University of Minnesota Medical School, Minneapolis, MN, USA

Keywords 5-fluorouracil; apoptosis; chemosensitizer; ERK5; miR-143

Correspondence C. M. P. Rodrigues, Research Institute for Medicines and Pharmaceutical Sciences, Faculty of Pharmacy, University of Lisbon, Avenida Professor Gama Pinto, 1649-003 Lisbon, Portugal Fax: +351 21 794 6491 Tel: +351 21 794 6400 E-mail: cmprodrigues@ff.ul.pt

(Received 29 April 2009, revised 2 August 2009, accepted 15 September 2009)

doi:10.1111/j.1742-4658.2009.07383.x

MicroRNAs are aberrantly expressed in cancer; microRNA-143 (miR-143) is down-regulated in colon cancer. HCT116 human colorectal cancer cells were used to investigate the biological role of miR-143. Transient miR- 143 overexpression resulted in an approximate 60% reduction in cell via- bility. In addition, stable miR-143 overexpressing cells were selected with G418 and exposed to 5-fluorouracil. Increased stable expression of miR- 143 was associated with decreased viability and increased cell death after exposure to 5-fluorouracil. These changes were associated with increased nuclear fragmentation and caspase -3, -8 and -9 activities. In addition, extracellular-regulated protein kinase 5, nuclear factor-jB and Bcl-2 pro- tein expression was down-regulated by miR-143, and further reduced by exposure to 5-fluorouracil. In conclusion, miR-143 modulates the expres- sion of key proteins involved in the regulation of cell proliferation, death and chemotherapy response. In addition, miR-143 increases the sensitivity of colon cancer cells to 5-fluorouracil, probably acting through extracellu- lar-regulated protein kinase 5 ⁄ nuclear factor-jB regulated pathways. Col- lectively, the data obtained in the present study suggest anti-proliferative, chemosensitizer and putative pro-apoptotic roles for miR-143 in colon cancer.

Introduction

small noncoding, gene regulate that negatively

mechanism, and further processed by the RNAse III enzyme Dicer, producing miRNA duplexes. One strand of the duplex is incorporated into the RNA-induced silencing complex, whereas the other is usually rapidly degraded. In mammals, miRNA-directed gene silencing occurs primarily via incomplete miRNA binding to 3¢- UTRs of target mRNAs, which represses translation and, in some cases, leads to mRNA degradation [1].

Abbreviations ERK5, extracellular-regulated protein kinase 5; 5-FU, 5-fluorouracil; LDH, lactate dehydrogenase; miRNA, microRNA; miR-143, miRNA-143; miR-145, miRNA-145; MTS, 3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium; NF, nuclear factor; pNA, p-nitroanilide

FEBS Journal 276 (2009) 6689–6700 ª 2009 The Authors Journal compilation ª 2009 FEBS

6689

MicroRNAs (miRNAs) are a recently discovered and single-stranded growing class of RNAs expression. miRNA biogenesis initiates by RNA polymerase II transcription of a primary transcript (pri-miRNA). This pri-miRNA is processed in the nucleus by the RNase III enzyme Drosha, producing pre-miRNAs, comprising hairpin structures of approximately 70 nucleotides. Subsequently, pre-miRNAs are exported to the cytoplasm through an Exportin-5-dependent More than 670 miRNAs have been identified in humans, but mRNA targets and regulatory pathways

P. M. Borralho et al.

miR-143 modulates 5-FU cytotoxicity

expression was

5-FU has been shown to induce apoptosis in colon cancer cell lines [20,21]. The ability of tumour cells to evade apoptosis is an enormous obstacle for effective treatment. Consequently, strategies aiming to over- come tumour cell resistance to chemotherapy and to increase drug efficacy, thereby minimizing toxic effects, are critically important. The molecular mechanisms of 5-FU cytotoxicity have been characterized, especially its ability to incorporate into RNA and DNA and to inhibit thymidylate synthase. 5-FU has recently been shown to modulate miRNA expression in colon cancer cells [22]. However, the relationship between 5-FU and interactions and their rele- miRNAs, their potential vance for drug efficacy have not been extensively stud- ied. Nevertheless, drug function could potentially be improved via modulation of miRNAs that play a role in chemoresistance.

results obtained demonstrate

contribute

have only been explored for a handful of miRNAs. Furthermore, miRNAs are significantly involved in the regulation of a myriad of biological processes, such as differentiation, proliferation and apoptosis [2], which are commonly altered in cancer. It is now apparent that miRNAs are differentially expressed in a wide array of human cancers, including colorectal cancer [3,4]. Nevertheless, the molecular signalling pathways modulated by miRNAs, which play a role in colorectal cancer, as well as cancer in general, are only partially characterized. miRNA-143 and miRNA-145 (miR-143 and miR-145) found reduced in tumours versus matched normal mucosa, both at the adenomatous and cancer stages of colorectal neoplasia, as well as in colon cancer cell lines [3,5]. Furthermore, miR-143 is a putative cancer biomarker that is also down-regulated in B-cell malignancies [6], bladder [7] and cervical cancer [8]. miR-143 targets extracellular- regulated protein kinase 5 (ERK5) during adipocyte differentiation [9] and in DLD-1 colon cancer cells [5]. To date, miR-143 has been demonstrated to directly the mRNA of KRAS [10,11], ERK5 [12], target DNMT3A [13] and ELK1 [14]. Despite growing evi- dence for an anti-proliferative and putative pro-apop- totic role for miR-143, a more detailed understanding of miR-143 biological function is warranted because reduced miR-143 expression may deregulate molecular signalling pathways with direct implications in cancer onset and ⁄ or progression and response to chemo- therapy.

In the present study, we evaluated the role of miR-143 in the response of HCT116 colon cancer cells to 5-FU. The that miR-143 decreased colon cancer cell viability and increased 5-FU sensitivity, suggesting that it may act as a potential chemosensitizer to 5-FU in colon cancer cells. In addition, miR-143 overexpression resulted in the down-regulation of ERK5, NF-jB and Bcl-2 pro- tein expression, which was further reduced by 5-FU. Collectively, our data indicate that reduced miR-143 expression in colon cancer may to unchecked proliferation and decreased sensitivity to 5-FU. Furthermore, increasing miR-143 expression in colon cancer cells may comprise a promising strategy for reducing tumour growth and aggressiveness, at the same time as increasing sensitivity to 5-FU.

Results

miR-143 overexpression reduces colon cancer cell viability

ERK5 is a mitogen-activated protein kinase that is regulated by a wide range of mitogens and cellular stresses. In addition, activated ERK5 is involved in cell survival, differentiation and proliferation. Several ERK5 targets and potential regulators of cell prolifera- tion have been identified, including c-Myc, cyclin D1 and nuclear factor (NF)-kB [15]. ERK5 activation of NF-jB promotes cellular transformation [16] and is a critical factor for G2–M cell cycle progression and timely mitotic entry [17]. NF-jB is a critical transcrip- tion factor involved in the suppression of apoptosis, stimulation of cell growth and the modulation of genes that appear to be important in tumour promotion, angiogenesis and metastasis. NF-jB activation is also associated with increased resistance to chemotherapeu- tic agents [18]. Therefore, strategies aimed at reducing ERK5 and NF-jB signalling may modulate tumour growth and sensitivity to chemotherapeutic agents.

FEBS Journal 276 (2009) 6689–6700 ª 2009 The Authors Journal compilation ª 2009 FEBS

6690

We first evaluated miR-143 expression in colon cancer cells lines by semi-quantitative RT-PCR, using total RNA and miRNA-enriched RNA (providing a higher abundance of < 200 nucleotides RNA molecules per lg RNA). The results obtained confirmed that miR-143 levels were almost undetectable in HCT116, LoVo, SW480 and SW620 colon cancer cells, using up to 500 ng of miRNA-enriched RNA and 30 PCR cycles (data not shown). By contrast, miR-143 expres- sion was readily detected from total heart RNA, using lower RNA input, whithout miRNA-enrichement. This was achieved with as little as 25 ng of total heart RNA, and at a lower PCR cycle number, thus reinforcing the notion that mir-143 is expressed at low 5-Fluorouracil (5-FU) has been the drug of choice for the treatment of colorectal cancer for more than four decades. Its limited efficacy as a single agent for advanced colorectal cancer has been improved by combination with newer chemotherapeutic agents [19].

P. M. Borralho et al.

miR-143 modulates 5-FU cytotoxicity

A

levels in colon cancer cells. U6 was used throughout as normalization control. HCT116 cells were chosen for further experiments because the molecular pathways of 5-FU cytotoxicity have already been explored in this cell line [20].

We next confirmed that mature miR-143 is produced from pCR3-pri-miR-143 vector and that it specifically binds to miR-143 sensor vector. This was achieved by co-transfecting HCT116 cells with pCR3-pri-miR-143, a firefly luciferase miR-143 sensor plasmid (miR-143 sensor) as a reporter for mature miR-143 expression, and with either miR-143 specific inhibitor (anti-miR- 143) or control (anti-miR-control). pRL-SV40 was also co-transfected and used as a normalization control. The results obtained showed that the lower availability of mature miR-143 after anti-miR-143 co-transfection led to increased firefly activity (Fig. 1A).

B

After validating the vectors, we determined the effect of miR-143 transient overexpression in HCT116 colon cancer cells. The results obtained demonstrated that pri-miR-143 overexpression reduced HCT116 cell via- bility by approximately 60%, at 48 h post-transfection (P < 0.05) as evaluated by the 3-(4,5-dimethylthiazol- 2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)- 2H-tetrazolium (MTS) assay (Fig. 1B). In addition, miR-143 overexpression increased cell death as early as 24 h post-transfection (data not shown). pCR3-empty transfection resulted in no significant change in cell viability.

Fig. 1. Mature miR-143 overexpression decreases HCT116 cell viability. Cells were transfected with the indicated plasmids and 50 or 100 nM anti-miR inhibitors, and analyzed at 48 h post-transfec- tion. (A) Cells were lysed and firefly and renilla luciferase activities were determined by the dual luciferase assay. (B) Cell viability was evaluated by MTS metabolism assays. Cells were then lysed and renilla luciferase activity was determined by the dual luciferase assay, for normalization of the MTS metabolism assay. The results are expressed as the mean ± SEM from at least three indepen- dent experiments. (cid:2)P < 0.01 and *P < 0.001 compared to controls; §P < 0.05 compared to pCR3-empty + pRL-SV40.

miR-143 increases 5-FU cytotoxicity

We next determined the effects of 5-FU on colon can- cer cells with endogenous low levels of miR-143. Therefore, we exposed HCT116 colon cancer cells to 8 lm 5-FU, a clinically relevant concentration [23] that has been shown to induce cell death and apoptosis [20]. 5-FU reduced cell viability by approximately 40% and increased general cell death by approximately 60% after 48 h of exposure (P < 0.01). After miR-143 transient overexpression, 5-FU also reduced cell viabil- ity and increased cell death compared to pCR3-empty 5-FU exposed cells (data not shown). This prompted us to create a stable miR-143 overexpression model by transfecting HCT116 cells with pCR3-pri-miR-143 and pCR3-empty and selecting with G418.

miR-143, and RNU6B for normalization. The results obtained showed that miR-143 expression is increased five-fold in HCT116-OV3 cells compared to HCT116 parental cells (P < 0.01) (Fig. 2B).

FEBS Journal 276 (2009) 6689–6700 ª 2009 The Authors Journal compilation ª 2009 FEBS

6691

We also evaluated whether 5-FU directly modulates the expression of mature miR-143. Total RNA was extracted from HCT116, HCT116-OV3 and HCT116- EM1 cells exposed to 8 lm 5-FU for 48 h. Mature On the basis of miR-143 expression levels evaluated by luciferase activity after pGL3-miR-143 sensor and pGL3-control, we selected a miR-143 overexpression clone and an empty vector clone, designated HCT116- OV3 and HCT116-EM1, respectively (Fig. 2A). In addition, we measured miR-143 expression by TaqMan for mature real-time PCR, using specific primers

P. M. Borralho et al.

miR-143 modulates 5-FU cytotoxicity

A

B

C

evaluated by the MTS metabolism and lactate dehy- drogenase (LDH) release assays. The results obtained indicated that mature miR-143 enhanced sensitivity to 5-FU. Indeed, cell viability was reduced and cell death was increased in HCT116-OV3 compared to parental and HCT116-EM1 cells, after 1–10 lm 5-FU exposure for 72 h (P < 0.05) (Fig. 3A). In addition, increased sensitivity to 5-FU was also observed in G418 selected pCR3-pri-miR-143 transfected mixed populations com- pared to pCR3-empty transfected mixed populations after exposure to same doses of 5-FU (Fig. 3B). Furthermore, 5-FU-induced apoptosis was enhanced in HCT116-OV3 cells, resulting in increased caspase-3, -8 and -9 activities and subsequent nuclear fragmentation compared to parental HCT116 and HCT116-EM1 control cells (P < 0.05) (Fig. 4).

Finally, we evaluated miR-143 expression in SW480 cancer cells exposed to 5-FU. SW480 cells were less sensitive to 5-FU than HCT116 cells (Fig. 5A). In addition, 5-FU up-regulated miR-143 expression in both cell lines (Fig. 5B). Nevertheless, miR-143 expres- sion was significantly lower in SW480 cells. The DCt values were 12.10 ± 0.35 in SW480 cells compared to 5.45 ± 0.08 in HCT116 cells (P < 0.001), whereas RNU6B Ct values for control and 5-FU-treated HCT116 and SW480 cells were 24.43 ± 0.22 and 24.42 ± 0.25, respectively (P = 0.954). Collectively, this represents an increase of almost 100-fold in miR-143 expression in HCT116 relative to SW480 cells, thereby reinforcing the potential involvement of miR-143 in 5-FU sensitivity.

Fig. 2. 5-FU increases miR-143 expression in HCT116 cells. (A) miR-143 expression by luciferase activity assays. Cells were co-transfected with either pGL3-miR-143 sensor or pGL3-control and pRL-SV40 and analyzed at 48 h after transfection. Cells were lysed and luciferase activity was evaluated with the dual luciferase assay. miR-143 levels were expressed as the luciferase signal ratio of pGL3-miR-143 sensor to pGL3-control cells. (B, C) Cells were harvested for total RNA extraction after 48 h in culture with and without 8 lM 5-FU exposure. miR-143 expression was evaluated from 1.33 lL cDNA of 10 ng of total RNA RT reactions, using spe- cific primers for miR-143 and RNU6B for normalization. miR-143 expression levels were calculated by the DDCt method, using HCT116 control cells as calibrator. The results are expressed as the mean ± SEM fold-change compared to HCT116 cells from three to six independent experiments. *P < 0.001 and (cid:2)P < 0.05 compared to controls.

miR-143 and 5-FU down-regulate ERK5, NF-jB and Bcl-2 protein expression

miR-143 expression was measured by TaqMan real- time PCR. Interestingly, the results obtained showed that miR-143 increased almost three-fold in parental HCT116 and HCT116-EM1 cells, but by only two-fold in HCT116-OV3 (P < 0.05) (Fig. 2C).

FEBS Journal 276 (2009) 6689–6700 ª 2009 The Authors Journal compilation ª 2009 FEBS

6692

We next evaluated the effect of increased mature miR-143 expression in colon cancer cells exposed to 5-FU. For this purpose, cells were exposed to 1–10 lm 5-FU for 72 h, and cell viability and cell death were ERK5 targeting by miR-143 has been shown in adipo- cytes [9] in addition to DLD-1 colon cancer cells [5]. More recently, miR-143 has been demonstrated to directly target the mRNA of KRAS [10,11], ERK5 [12], DNMT3A [13] and ELK1 [14]. To determine whether miR-143 modulates ERK5 in HCT116 cells as well as its relevance to 5-FU cytotoxicity, we evaluated ERK5 protein expression after transfection of miR-143 precursor molecules (pre-miR-143) and a nonspecific control (pre-miR-control). The results obtained showed that miR-143 reduced ERK5 protein expression in a leading to an approximate dose-dependent manner, 70% reduction at 48 h post-transfection with 80 nm pre-miR-143 (P < 0.001) (Fig. 6A, upper panel). In addition, after transfection with 80 nm pre-miR-143, 5-FU further reduced ERK5 protein expression in a time-dependent manner, reaching a reduction of 50%, 60% and 80% at 24, 48 and 72 h, respectively

P. M. Borralho et al.

miR-143 modulates 5-FU cytotoxicity

A

B

Fig. 3. mir-143 overexpressing cells are more sensitive to 5-FU cytotoxicity. Cells were exposed to either 1–10 lM 5-FU or no addition (con- trol) for 72 h. Mixed populations of miR-143 overexpressing cells (pCR3-pri-miR-143 mix) and respective control cells (pCR3-empty mix) were generated by transfecting HCT116 cells with pCR3-pri-miR-143 or pCR3-empty vector, respectively. Cells were selected and maintained with 1 mgÆmL)1 G418, under the same conditions as those used for single-clone derived HCT116-OV3 and HCT116-EM1 cells. (A, B) Cells with higher miR-143 expression (HCT116-OV3 and pCR3-pri-miR-143 mix) displayed decreased viability and increased cell death after exposure to 5-FU. Cell viability and cell death were evaluated by MTS metabolism and LDH activity assays, respectively. In the far right panels of (A) and (B), LDH normalized to MTS is also plotted for the mean values of each 5-FU concentration used. The results are expressed as the mean ± SEM fold-change compared to controls from at least three independent experiments. *P < 0.01 and (cid:2)P < 0.05 from HCT116-EM1. No signif- icant changes were observed between HCT116 and HCT116-EM1 or between HCT116 and pCR3-empty mix for 1–10 lM 5-FU.

(Fig. 6A,

specific inhibitor

Furthermore, we also found that stable miR-143 overexpressing cells express lower levels of ERK5, NF-jB and Bcl-2 proteins compared to control and parental HCT116 cells. Furthermore, 5-FU potentiated ERK5, NF-jB and Bcl-2 expression knockdown (Fig. 7). These results further indicate that miR-143 is a key player in the regulation of cell proliferation and the response to 5-FU growth inhibition ⁄ cytotoxicity in HCT116 cells, probably by acting through the ERK5 ⁄ NF-jB axis.

Discussion

FEBS Journal 276 (2009) 6689–6700 ª 2009 The Authors Journal compilation ª 2009 FEBS

6693

increasing It is now well-established that miRNAs regulate a plethora of crucial cellular functions, including cell growth, differentiation and apoptosis, which are commonly altered in cancer cells. In the present study, we evaluated the effect of miR-143 overexpression on HCT116 colon cancer cells. HCT116 cells were trans- fected with pCR3-pri-miR-143, and the respective con- trol, pCR3-empty. miR-143 production was assessed by luciferase assays. The results obtained demonstrate (P < 0.05) compared to controls lower panel). In addition, we were also able to regulate ERK5 protein expression by modulating the availabil- ity of mature miR-143 (Fig. 6B). At 72 h after co-transfection of 40 nm pre-miR-143 with 60 nm of its anti-miR-143, ERK5 was increased compared to pre-miR-143 and anti-miR con- trol co-transfection experiments (Fig. 6B, lane 4 versus 5). Moreover, pre-miR-control and anti-miR-143 co-transfection also increased ERK5 protein expression compared to co-transfection of pre-miR-control and anti-miR-control (Fig. 6B, lane 6 versus 7). Exposure to 5-FU resulted in a further reduction of ERK5 pro- tein expression (Fig. 6B, lanes 9–14 versus 1–7). Accordingly, higher miR-143 abundance decreased cell viability (Fig. 6C). Finally, ERK5 expression was knocked-down in HCT116 cells by transfecting 80 nm of specific ERK5 siRNA and then exposing cells to 5-FU. The results obtained show that ERK5 silencing augmented apoptosis compared to mock-transfected cells, whereas 5-FU-induced apoptosis (P < 0.05) (Fig. 6D).

P. M. Borralho et al.

miR-143 modulates 5-FU cytotoxicity

Fig. 4. HCT116-OV3 cells are more sensi- tive to 5-FU-induced apoptosis. Cells were exposed to 8 lM 5-FU and processed for evaluation of nuclear morphology after Hoechst staining at 24 h, and for caspase activity assays at 12 h. The results are expressed as the mean ± SEM fold-change compared to controls from at least three independent experiments. *P < 0.01 and (cid:2)P < 0.05 from HCT116-EM1. No significant changes were observed between HCT116 and HCT116-EM1.

and the respective created stable miR-143 overexpressing control

that increased pri-miR-143 expression reduced cell via- bility in pCR3-pri-miR-143 transfected cells by 60% compared to control cells. These results are in accor- in which pre-miR-143 dance with a previous study, transfection in SW480 and DLD-1 colon cancer cells reduced cell viability in a dose-dependent manner [5].

FEBS Journal 276 (2009) 6689–6700 ª 2009 The Authors Journal compilation ª 2009 FEBS

6694

We were particularly interested in investigating the effect of miR-143 on the cellular response to 5-FU. 5-FU is a well-known apoptosis-inducing drug; indeed, it was shown previously that 5-FU induces apoptosis in HCT116 cells [20]. 5-FU has been used in the clinic for several decades, and its metabolic pathways and cyto- toxic modes of action through the inhibition of thymi- dylate synthase activity and incorporation into RNA and DNA are well known. However, the molecular pathways modulated by 5-FU that lead to cell growth inhibition and cell death induction are not entirely understood, nor are the mechanisms of tumour cell escape from 5-FU cytotoxicity. In addition, little is known about the role of miRNAs in modulating the tumour cell response to chemotherapeutic agents, such as 5-FU. Nevertheless, 5-FU alters the expression of a set of 22 miRNAs in colon cancer cell lines [22]. Further- more, S-1, a fourth-generation 5-FU-based oral drug developed to improve efficacy, also alters the expression levels of certain miRNAs, as demonstrated in tumour tissue from patients undergoing S-1 therapy [24]. To evaluate the effect of miR-143 on 5-FU sensitivity, cells we (HCT116-OV3) cells (HCT116-EM1) by pCR3-pri-miR-143 and pCR3- empty transfection, respectively, and G418 selection. miR-143 expression was evaluated by luciferase assays and also by TaqMan real-time PCR with specific primers for miR-143 and RNU6B. The results obtained showed that miR-143 expression was increased five-fold in HCT116-OV3 cells. Cells were then exposed to different concentrations of 5-FU for 72 h to plot growth inhibi- tion and cell death dose–response curves. 5-FU at a con- centration 1–10 lm was significantly more cytotoxic in HCT116-OV3 cells compared to HCT116-EM1 control or HCT116 parental cells. These effects were already evi- dent 48 h after 5-FU exposure but more intense after 72 h of drug exposure. Furthermore, we also observed higher sensitivity to 5-FU in mixed populations of miR- 143 overexpressing cells compared to controls. Curi- ously, a chemosensitizer role of miR-143 to 5-FU has also been reported in gastric cancer cells. Indeed, trans- fection of pre-miR-143 into MKN-1 cells was shown to reduce the number of viable cells after exposure to 10 lm 5-FU [25]. Collectively, these data indicate that miR-143 is an important modulator of 5-FU sensitivity not only in colon cancer, but also in other types of cancer, particularly those of the gastrointestinal tract.

P. M. Borralho et al.

miR-143 modulates 5-FU cytotoxicity

A

miR-143 expression to a lesser extent in HCT116-OV3 than in HCT116-EM1 and HCT116 cells, which may be attributed to the relatively high levels of miR-143 in HCT116-OV3 cells. Furthermore, up-regulation of miR-143 by 5-FU also occurs in SW480 cells that are less sensitive to 5-FU cytotoxicity. However, the basal expression of miR-143 is significantly lower in SW480 cells than in HCT116, which underlines the relevance of miR-143 expression on 5-FU sensitivity.

B

Fig. 5. SW480 cells are less sensitive to 5-FU than HCT116 cells and express less miR-143. Cells were exposed to either 1–100 lM 5-FU or no addition (control) for 72 h. (A) HCT116 cells displayed increased growth inhibition and cell death after exposure to 5-FU, as evaluated by MTS metabolism and LDH activity assays, respec- tively. (B) Cells were exposed to 8 lM 5-FU for 72 h and miRNA- enriched RNA was extracted with a mirVana(cid:2) PARIS kit. miR-143 expression was evaluated from 4 lL of cDNA of a 50 ng miRNA- enriched RNA RT reaction, using specific primers for miR-143, and RNU6B for normalization. miR-143 expression levels were calcu- lated by the DDCt method, using control cells as calibrator. miR-143 was up-regulated in HCT116 and SW480 cells after 5-FU exposure by approximately three- and two-fold, respectively. RNU6B Ct val- ues for control and 5-FU-treated HCT116 and SW480 cells were 24.43 ± 0.22 and 24.42 ± 0.25, respectively (P = 0.954). The results are expressed as the mean ± SEM percentage of growth inhibition, LDH activity or miR-143 expression from at least three independent experiments. (A) *P < 0.01 and (cid:2)P < 0.05 from SW480 cells; (B) *P < 0.001 compared to HCT116 and (cid:2)P < 0.05 compared to the respective nontreated cells.

The results obtained in the present study are in agree- ment with a recent report showing that p53 enhances the post-transcriptional maturation of miR-143 in response to DNA damage [10]. Using wild-type p53 HCT116 cells exposed to the potent p53 inducer doxorubicin, and despite no significant changes in pri-miR-143 levels, there was increased processing of pri-miR-143 into pre- and mature miR-143 [10]. In addition, this DNA-dam- age-induced up-regulation of pre-miR-143 and mature miR-143 was diminished in p53 null HCT116 cells, thereby suggesting that increased mature miR-143 expression may be a result of increased p53 expression. These data may explain the increased levels of mature miR-143 that we found in the present study after expo- sure of wild-type p53 HCT116 cells to 5-FU because it has previously been shown that 5-FU also strongly increases p53 expression in this cell line [20].

Interestingly, when mutant p53 (R273H) was intro- duced into p53 null HCT116 cells, there was a reduc- tion in pre- and mature miR-143 production, contrasting with constant levels of pri-miR-143 [10], which suggests that mutant p53 hampers miR-143 pro- cessing in a transcription-independent manner. These results are once again in agreement with the results obtained in SW480 cells. In these cells expressing the R273H mutant p53, we observed a dramatic approxi- mately 100-fold decrease in mature miR-143 expression compared to wild-type p53 HCT116 cells. However, in SW480 cells, we also found an increase in mature miR-143 after 5-FU exposure, which may indicate that there is some mature miR-143 production that escapes wild-type p53 control. Nevertheless, the levels of mature miR-143 in SW480 cells treated with 5-FU are much lower than those observed in untreated HCT116 cells. This finding may contribute to or be associated with SW480 lower sensitivity to 5-FU cytotoxicity.

FEBS Journal 276 (2009) 6689–6700 ª 2009 The Authors Journal compilation ª 2009 FEBS

6695

We next selected 8 lm 5-FU to further explore the differential sensitivity of our cell model. We have pre- viously shown that it efficiently triggers apoptosis in HCT116 cells [20]. Furthermore, the results obtained study now show that 8 lm 5-FU in the present induced growth inhibition by 55% in addition to a two-fold increase in cell death. Notably, 5-FU significantly increased caspase activation and nuclear fragmentation in HCT116-OV3 compared to control and parental cells. We have also evaluated miR-143 expression after 48 h of 5-FU exposure and found it to be significantly increased. Curiously, 5-FU increased We cannot conclude whether the basal difference in miR-143 expression between HCT116 and SW480 cells results from loss of transcriptional or post-transcrip- tional p53 modulation. p53 has been shown to increase mature miR-145 expression and bind to a putative response element located upstream of miR-145 [26]. In addition, miR-143 and miR-145 are highly conserved miRNAs in close genomic proximity [14,27], suggesting

P. M. Borralho et al.

miR-143 modulates 5-FU cytotoxicity

A

C

D

B

Fig. 6. miR-143 and 5-FU down-regulate ERK5 protein expression. (A) HCT116 cells were transfected with 40 or 80 nM pre-miR-143 or pre- miR-control and ERK5 protein expression was evaluated at 48 h after transfection (upper panel). HCT116 cells were transfected with 80 nM pre-miR-143 or pre-miR-control. At 9 h after transfection, cells were exposed to 8 lM 5-FU for 24, 48 or 72 h. ERK5 protein expression was evaluated at the indicated time-points (lower panel). (B) HCT116 cells were transfected with 40 nM pre-miR-143 or pre-miR-control, or with additional 60 nM anti-miR-control or anti-miR-143. At 9 h after transfection, cells were exposed to 8 lM 5-FU for 72 h, at which time ERK5 protein expression was evaluated. (C) HCT116 cells were transfected with 80 nM pre-miR-143, pre-miR-control, anti-miR-143 or anti-miR-con- trol. At 9 h after transfection, cells were exposed to 8 lM 5-FU for 72 h, at which time cell viability was determined. The results are expressed as the fold-change of LDH normalized to MTS for pre-miR-143 and anti-miR-143 compared to the respective controls. Black bar, ratio of pre-miR-143 over pre-miR-control; white bar, ratio of anti-miR-143 over anti-miR-control. (D) HCT116 cells were transfected with 80 nM ERK5 siRNA or mock transfected. Twenty-four hours after transfection, cells were exposed to 8 lM 5-FU for 48 h and processed for evaluation of nuclear morphology after Hoechst staining. ERK5 protein expression was also evaluated at the same time. The results are expressed as the mean ± SEM fold-change compared to controls from at least three independent experiments. Representative blots from at least three independent experiments are shown. *P < 0.05 compared to control HCT116 cells (mock) and (cid:2)P < 0.05 compared to 5-FU-trea- ted HCT116 cells (5-FU-treated mock).

the

overexpression

FEBS Journal 276 (2009) 6689–6700 ª 2009 The Authors Journal compilation ª 2009 FEBS

6696

that their expression is linked to a bicistronic primary precursor. Indeed, the use of specific primers for miR-143 and miR-145 resulted in amplicons containing both miR-143 and miR-145, suggesting that similar the regulation of mechanisms are responsibe for miR-143 and miR-145 [14]. Nevertheless, p53 induc- tion by doxorubicin increases transcriptional activation of miR-34a, which is a known p53 transcrip- tional target, and does not increase pri-miR-143 or pri-miR-145 [10]. In addition, the increased maturation of miR-145 after DNA damage was similar to the reported increase in miR-143 maturation in multiple lines. Further studies are required to clarify the cell molecular details of the relationship between p53 and miR-143 and miR-145. miR-143 was initially shown to regulate adipocyte differentiation by targeting ERK5 [9]. Recently, ERK5 [5], KRAS [10,11] and DNMT3A [13] were also shown to be negatively modulated by miR-143 in these genes colon cancer cells. More importantly, have been proven to be direct targets of miR-143 [10–13]. Our data obtained in HCT116 cells are in reports and show that accordance with previous protein reduces ERK5 miR-143 expression. ERK5 is involved in cell proliferation and differentiation [28], as well as the inhibition of endothelial cell apoptosis [29]. In addition, ERK5 has been identified as a pro-survival kinase during mito- sis, whereas ERK5 knockdown by RNA interference induced apoptosis [30]. ERK5 is also a critical factor

P. M. Borralho et al.

miR-143 modulates 5-FU cytotoxicity

protein Bcl-2 associated with

as well as an increased response to 5-FU. In addition, the anti- we also see a strong down-regulation of apoptotic higher miR-143 expression, which may contribute to the this miRNA. Our putative pro-apoptotic role of results are particularly relevant because they suggest that miR-143 is an endogenous cell growth ⁄ prolifera- tion master switch, in which any loss may result in unchecked cellular proliferation via ERK5 ⁄ NF-jB signalling. In turn, this may constitute a key event for colon cancer onset and progression.

miR-143 expression was evaluated in situ in mouse colonic epithelial crypts [27] and found mainly in the cytoplasm, where it was more intense in the ascending more differentiated pre-apoptotic glandular cells than in the immature, proliferating cells at the bottom of crypts. Therefore, the extremely low miR-143 expres- sion levels in cancer and, in particular in colon cancer, may be an important factor contributing to tumour growth and ⁄ or escape from apoptosis. In addition, miR-143 expression has been shown to increase after a-mangostin exposure in human colon cancer DLD-1 cells, resulting in increased apoptosis [31]. Further- more, a-mangostin acted synergistically with low dose 5-FU, increasing DLD-1 growth inhibition [31]. Our data also show that 5-FU increases miR-143 expres- sion, which may potentiate 5-FU sensitivity, suggesting a feed-forward mechanism of action.

compared

shown.

are

insight

Fig. 7. miR-143 and 5-FU down-regulate ERK5, NF-jB and Bcl-2 protein expression. Populations of HCT116 cells, miR-143 over- expressing cells (HCT116-OV3 and pCR3-pri-miR-143 mix) and respective control cells (HCT116-EM1 and pCR3-empty mix) were synchronized to early S-phase by a double thymidine block. Eight- hours after seeding, 2 mM thymidine was added and cells were cultured for 14 h. Cells were then released from the block for 10 h in media lacking thymidine, followed by an additional 14 h of culture in the presence of 2 mM thymidine (second block). Cells were released from the second thymidine block into media with 5-FU or no addition (control) and harvested for protein extraction 72 h later. Representative blots from at least three independent experiments to HCT116; (cid:3)P < 0.05 (cid:2)P < 0.05 compared to 5-FU-treated HCT116 cells; §P < 0.05 compared to the respective untreated cells; *P < 0.05 compared to HCT116-EM1 or pCR3-empty mix cells; #P < 0.05 compared to 5-FU-treated HCT116-EM1 or 5-FU-treated pCR3-empty mix.

to 5-FU. Further

The results obtained in the present study provide into miR-143 regulated pathways additional and their impact on 5-FU sensitivity. Exposure of miR-143 overexpressing cells to 5-FU potentiated a significant decrease in ERK5, NF-jB and Bcl-2 protein expression and resulted in increased 5-FU cytotoxicity. Collectively, our findings emphasize the potential pivotal relevance of miR-143 in the colon cancer envi- ronment and suggest that it has a role as chemosensi- studies are necessary to tizer elucidate the full extent of the molecular signalling pathways and players modulated by miR-143.

Materials and methods

HCT116 cells were grown in DMEM supplemented with 10% fetal bovine serum (Invitrogen, Grand Island, NY, USA), 1% l-glutamine 200 mm (Merck and Co. Inc., Whitehouse Station, NJ, USA) and 1% antibiotic ⁄ antimy- cotic solution (Sigma Chemical Co., St Louis, MO, USA) and maintained at 37 (cid:3)C in a humidified atmosphere of 5% CO2. In selected experiments, HCT116 cells were compared with SW480, LoVo and SW620 human colorectal cancer

Cell culture

FEBS Journal 276 (2009) 6689–6700 ª 2009 The Authors Journal compilation ª 2009 FEBS

6697

in G2–M cell cycle progression and timely mitotic entry because stimulation of ERK5 activated NF-jB, whereas inhibition of NF-jB at G2–M significantly delayed mitotic entry [17]. These results suggest a potential crosstalk between ERK5 and the apoptotic machinery. The results obtained in the present study are in agreement with this notion by demonstrating that miR-143 overexpression leads to a significant reduction of ERK5 and NF-jB protein expression. Because ERK5 directly activates NF-jB to promote cell cycle progression through G2–M, reduced levels of these proteins may account for reduced cell growth

P. M. Borralho et al.

miR-143 modulates 5-FU cytotoxicity

cells. All cell lines were grown under identical conditions. Cells were seeded at 0.75 · 105 cellsÆmL)1 for pre-miR-143 tranfections; 1 · 105 cellsÆmL)1 for morphological assess- ment of apoptosis; and at 2 · 105 cellsÆmL)1 for RNA and protein extraction, cell viability and cell death assays.

Generation of HCT116 cells with stable expression of miR-143

HCT116 cells were selected and maintained with 1 mgÆmL)1 G418 (Invitrogen), after transfection with pCR3-pri-miR- 143 and pCR3-empty vectors. Isolated single clone-derived cell foci were picked up by ring cloning strategies, expanded and propagated in complete media plus G418. In addition, mixed populations of miR-143 overexpressing and control cells were obtained in a similar manner, without the single- cell selection step. miR-143 expression was evaluated by luciferase assay after co-transfection and normalization with pGL3-miR-143 sensor and pRL-SV40. This assay was performed regularly to control cell clone consistency. In addition, TaqMan real-time PCR (Applied Biosystems) was used to confirm the expression of mature miR-143 in cell clones.

Cell synchronization was performed using a double thymi- dine block (early S-phase). Eight-hours after seeding 0.75 · 105 cellsÆmL)1, 2 mm thymidine (Sigma) was added and cells were cultured for 14 h. Cells were then released from block for 10 h in media lacking thymidine, followed by an additional 14 h culture in the presence of 2 mm thymidine (second block). Cells were released from second thymidine block into media with either 5-FU or no addi- tion (control). Cells were harvested for protein extraction at the indicated times.

Cell synchronization

Evaluation of cell death and viability

At the indicated times, general cell death was evaluated by the LDH assay kit (Sigma). LDH activity was evaluated in cell culture media, using a Bio-Rad microplate reader Model 680 (Bio-Rad, Hercules, CA, USA). Cell viability was evaluated with CellTiter96(cid:4) AQueous Non-Radioactive Cell Proliferation Assay (Promega), using MTS inner salt. Finally, cells were processed for luciferase assay and trans- fection efficiency normalization.

5-FU (Sigma) stock solutions of 100 and 8 mm were pre- pared in dimethyl sulfoxide. Twenty-four hours after plat- ing, cells were incubated with either 5-FU or no addition (control) for the indicated times. For the 5-FU dose– response curves, media was removed 24 h after plating and replaced with fresh media containing 5-FU. The final dimethyl sulfoxide concentration was always 0.1%.

5-FU exposure

Luciferase activity

At the indicated times, firefly and renilla luciferase activities were measured using the Dual-Luciferase(cid:4) Reporter Assay System (Promega). Renilla luciferase activity was used as a transfection normalization control.

Transfection of miR-143 vectors, anti-miR-143 inhibitor, pre-miR-143 and ERK5 siRNA

Hoechst labelling of cells was used to detect apoptotic nuclei. Attached cells were fixed with 4% paraformaldehyde in NaCl ⁄ Pi (pH 7.4) and incubated with Hoechst dye 33258 (Sigma Chemical Co.) at 5 lgÆmL)1 in NaCl ⁄ Pi for 3 min. Fluorescent nuclei were categorized according to the con- densation and staining characteristics of chromatin. Three random microscopic fields per sample of (cid:2) 100 nuclei were counted and mean values expressed as the percentage of apoptotic nuclei.

Nuclear morphology

Caspase activity was determined in cytosolic protein extracts after harvesting and homogenization of cells in isolation buf- fer, containing 10 mm Tris-HCl buffer (pH 7.6), 5 mm MgCl2, 1.5 mm potassium acetate, 2 mm dithiothreitol and

HCT116 cells were transiently transfected with miR-143 overexpression vector coding for the miR-143 precursor (pCR3-pri-miR-143) and miR-143 sensor comprising two sequences complementary to mature miR-143 sequence (pGL3-miR-143 sensor) [9]. pRL-SV40 (Promega, Madison, WI, USA) was used for transfection normalization. pGL3- control plasmid (Promega) and pCR3-empty vector were used as negative controls. To further validate the experi- mental model, cells were co-transfected with anti-miR inhibitors by adding either 50 or 100 nm anti-miR-143 or anti-miR-control inhibitors to the vector mixture described above. In addition, HCT116 cells were transfected with pre-miR-143, pre-miR-control, anti-miR-143 and anti-miR- control at a final concentration of 40 or 80 nm. We also co-transfected HCT116 cells with 40 nm pre-miR-143 or pre-miR-control plus 60 nm anti-miR-143 or anti-miR- control, at a final concentration of 100 nm. ERK5 silencing was performed using 80 nm MAPK7 Silencer(cid:4) Select Pre-Designed & Validated siRNA (all from Applied Bio- systems, Foster City, CA, USA). Transfections were performed using Lipofectamine 2000 (Invitrogen), accord- ing to the manufacturer’s instructions.

FEBS Journal 276 (2009) 6689–6700 ª 2009 The Authors Journal compilation ª 2009 FEBS

6698

Caspase activity

P. M. Borralho et al.

miR-143 modulates 5-FU cytotoxicity

Super Signal(cid:2) substrate (Pierce, Rockford, IL, USA). b-actin (Sigma) was used as a loading control. Protein concentrations were determined using the Bio-Rad protein assay kit according to the manufacturer’s instructions.

(Sigma),

All data are expressed as the mean ± SEM from at least three independent experiments. Statistical significance was evaluated using the Student’s t-test. P < 0.05 was consid- ered statistically significant.

protease inhibitor mixture tablets (Complete; Roche Applied Science, Mannheim, Germany). General caspase-3, -8 and -9-like activities were determined by enzymatic cleavage of chromophore p-nitroanilide (pNA) from the substrates N-acetyl-Asp-Glu-Val-Asp-pNA (DEVD-pNA), N-acetyl- Ile-Glu-pro-Asp-pNA (IEPD-pNA) and N-acetyl-Leu-Glu- His-Asp-p-NA (LEHD-pNA) respectively. The proteolytic reaction was carried out in isolation buffer con- taining 50 lg of cytosolic protein and 50 lm specific caspase substrate. The reaction mixtures were incubated at 37 (cid:3)C for 1 h, and the formation of pNA was measured by monitoring A405 using a 96-well plate reader.

Statistical analysis

Acknowledgements

reagent

Total and miRNA-enriched RNA was extracted from cells using the mirVana(cid:2) PARIS kit from Ambion (Austin, TX, USA), recovering either small-RNA containing total RNA (total RNA) or small RNA species of less than 200 nucleo- tides (miRNA-enriched RNA). In addition, total RNA was also extracted using Trizol (Invitrogen). RNA concentration was determined by monitoring A260.

miR-143 expression

The authors thank Dr Christine Esau, ISIS Pharma- ceuticals Inc., for the kind gift of miR-143 overexpres- sion and sensor plasmids. The study was supported by grants FCG 68796 ⁄ 2004 from Fundac¸ a˜ o Calouste Gulbenkian and PTDC ⁄ SAU-GMG ⁄ 099161 ⁄ 2008 from Fundac¸ a˜ o para a Cieˆ ncia e a Tecnologia (FCT), Lisbon, Portugal (to C.M.P.R.); by PhD fellowship SFRH ⁄ BD ⁄ 24165 ⁄ 2005 (to P.M.B.) from FCT; and by postdoctoral fellowship SFRH ⁄ BPD ⁄ 30257 ⁄ 2006 (to R.E.C) from FCT.

References

1 Bartel DP (2004) MicroRNAs: genomics, biogenesis,

mechanism, and function. Cell 116, 281–297.

2 Carthew RW (2006) Gene regulation by microRNAs.

Curr Opin Genet Dev 16, 203–208.

3 Michael MZ, SM OC, van Holst Pellekaan NG, Young

GP & James RJ (2003) Reduced accumulation of specific microRNAs in colorectal neoplasia. Mol Cancer Res 1, 882–891.

For real-time PCR, RT reactions were performed using a TaqMan MicroRNA Reverse Transcription Kit and TaqMan MicroRNA assays for hsa-miR-143 and human RNU6B for normalization to endogenous control. Data were collected with 7000 System Sequence Detection Software, version 1.2.3 (Applied Biosystems). Triplicate reactions were run per sam- ple. The comparative threshold cycle method was used to cal- culate the amplification factor where the threshold cycle (Ct) is defined as the cycle number at which the fluorescence passes the fixed threshold intensity level. miR-143 expression levels in different samples were calculated on the basis of DDCt method. HCT116 or SW480 parental untreated (control) cells were used as calibrator; the n-fold change in miR-143 expres- sion was obtained using the formula: 2(cid:3)DDCt .

4 Volinia S, Calin GA, Liu CG, Ambs S, Cimmino A, Petrocca F, Visone R, Iorio M, Roldo C, Ferracin M et al. (2006) A microRNA expression signature of human solid tumors defines cancer gene targets. Proc Natl Acad Sci USA 103, 2257–2261.

5 Akao Y, Nakagawa Y & Naoe T (2006) MicroRNAs

143 and 145 are possible common onco-microRNAs in human cancers. Oncol Rep 16, 845–850.

6 Akao Y, Nakagawa Y, Kitade Y, Kinoshita T & Naoe T (2007) Downregulation of microRNAs-143 and -145 in B-cell malignancies. Cancer Sci 98, 1914–1920.

7 Lin T, Dong W, Huang J, Pan Q, Fan X, Zhang C &

Huang L (2009) MicroRNA-143 as a tumor suppressor for bladder cancer. J Urol 181, 1372–1380.

8 Lui WO, Pourmand N, Patterson BK & Fire A (2007) Patterns of known and novel small RNAs in human cervical cancer. Cancer Res 67, 6031–6043.

9 Esau C, Kang X, Peralta E, Hanson E, Marcusson EG, Ravichandran LV, Sun Y, Koo S, Perera RJ, Jain R

Steady-state levels of ERK5, NF-jB (p65) and Bcl-2 pro- tein were determined by immunoblot analysis. Briefly, 50 lg of total protein extracts were separated on 8% or 12% SDS-polyacrylamide electrophoresis gels. After elec- trophoretic transfer onto nitrocellulose membranes, immu- noblots were incubated with 15% H2O2 for 15 min at room temperature. After blocking with 5% milk solution, the blots were incubated overnight at 4 (cid:3)C with primary rabbit polyclonal antibody reactive to ERK5 (#3372; Cell Signal- ing, Beverly, MA, USA) and a mouse monoclonal antibody reactive to NF-jB (p65) or to Bcl-2 (#sc-8008 and #sc- 7382, respectively; Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA). Finally, membranes were incubated with secondary anti-mouse ⁄ -rabbit sera conjugated with horse- radish peroxidase (Bio-Rad) for 3 h at room temperature. The membranes were processed for protein detection using

FEBS Journal 276 (2009) 6689–6700 ª 2009 The Authors Journal compilation ª 2009 FEBS

6699

Immunoblotting

P. M. Borralho et al.

miR-143 modulates 5-FU cytotoxicity

21 Longley DB, Allen WL, McDermott U, Wilson TR,

et al. (2004) MicroRNA-143 regulates adipocyte differ- entiation. J Biol Chem 279, 52361–52365.

Latif T, Boyer J, Lynch M & Johnston PG (2004) The roles of thymidylate synthase and p53 in regulating Fas-mediated apoptosis in response to antimetabolites. Clin Cancer Res 10, 3562–3571.

10 Suzuki HI, Yamagata K, Sugimoto K, Iwamoto T, Kato S & Miyazono K (2009) Modulation of microRNA processing by p53. Nature 460, 529–533. 11 Chen X, Guo X, Zhang H, Xiang Y, Chen J, Yin Y, Cai X, Wang K, Wang G, Ba Y et al. (2009) Role of miR-143 targeting KRAS in colorectal tumorigenesis. Oncogene 28, 1385–1392.

22 Rossi L, Bonmassar E & Faraoni I (2007) Modification of miR gene expression pattern in human colon cancer cells following exposure to 5-fluorouracil in vitro. Pharmacol Res 56, 248–253.

23 Petit E, Milano G, Levi F, Thyss A, Bailleul F &

12 Akao Y, Nakagawa Y, Iio A & Naoe T (2009) Role of microRNA-143 in Fas-mediated apoptosis in human T-cell leukemia Jurkat cells. Leuk Res 33, 1530–1538.

13 Ng EK, Tsang WP, Ng SS, Jin HC, Yu J, Li JJ,

Schneider M (1988) Circadian rhythm-varying plasma concentration of 5-fluorouracil during a five-day contin- uous venous infusion at a constant rate in cancer patients. Cancer Res 48, 1676–1679.

24 Nakajima G, Hayashi K, Xi Y, Kudo K, Uchida K,

Rocken C, Ebert MP, Kwok TT & Sung JJ (2009) MicroRNA-143 targets DNA methyltransferases 3A in colorectal cancer. Br J Cancer 101, 699–706.

14 Cordes KR, Sheehy NT, White MP, Berry EC, Morton

Takasaki K, Yamamoto M & Ju J (2006) Non-coding microRNAs hsa-let-7g and hsa-miR-181b are associated with chemoresponse to S-1 in colon cancer. Cancer Genomics Proteomics 3, 317–324.

SU, Muth AN, Lee TH, Miano JM, Ivey KN & Srivastava D (2009) miR-145 and miR-143 regulate smooth muscle cell fate and plasticity. Nature 460, 705–710.

15 Wang X & Tournier C (2006) Regulation of cellular

functions by the ERK5 signalling pathway. Cell Signal 18, 753–760.

16 Pearson G, English JM, White MA & Cobb MH (2001) ERK5 and ERK2 cooperate to regulate NF-jB and cell transformation. J Biol Chem 276, 7927–7931. 17 Cude K, Wang Y, Choi HJ, Hsuan SL, Zhang H,

25 Takagi T, Iio A, Nakagawa Y, Naoe T, Tanigawa N & Akao Y (2009) Decreased expression of microRNA-143 and -145 in human gastric cancers. Oncology 77, 12–21. 26 Sachdeva M, Zhu S, Wu F, Wu H, Walia V, Kumar S, Elble R, Watabe K & Mo YY (2009) p53 represses c-Myc through induction of the tumor suppressor miR-145. Proc Natl Acad Sci USA 106, 3207–3212. 27 Akao Y, Nakagawa Y & Naoe T (2007) MicroRNA- 143 and -145 in colon cancer. DNA Cell Biol 26, 311– 320.

Wang CY & Xia Z (2007) Regulation of the G2-M cell cycle progression by the ERK5-NFjB signaling pathway. J Cell Biol 177, 253–264.

28 Nishimoto S & Nishida E (2006) MAPK signalling: ERK5 versus ERK1 ⁄ 2. EMBO Rep 7, 782–786.

29 Pi X, Yan C & Berk BC (2004) Big mitogen-activated

18 Yu YY, Li Q & Zhu ZG (2005) NF-jB as a molecular

protein kinase (BMK1) ⁄ ERK5 protects endothelial cells from apoptosis. Circ Res 94, 362–369.

target in adjuvant therapy of gastrointestinal carcinomas. Eur J Surg Oncol 31, 386–392.

30 Girio A, Montero JC, Pandiella A & Chatterjee S

(2007) Erk5 is activated and acts as a survival factor in mitosis. Cell Signal 19, 1964–1972.

19 Longley DB, Harkin DP & Johnston PG (2003) 5-fluo- rouracil: mechanisms of action and clinical strategies. Nat Rev Cancer 3, 330–338.

31 Nakagawa Y, Iinuma M, Naoe T, Nozawa Y & Akao

20 Borralho PM, Moreira da Silva IB, Aranha MM, Albu- querque C, Nobre Leitao C, Steer CJ & Rodrigues CM (2007) Inhibition of Fas expression by RNAi modulates 5-fluorouracil-induced apoptosis in HCT116 cells expressing wild-type p53. Biochim Biophys Acta 1772, 40–47.

Y (2007) Characterized mechanism of alpha-mangostin- induced cell death: caspase-independent apoptosis with release of endonuclease-G from mitochondria and increased miR-143 expression in human colorectal cancer DLD-1 cells. Bioorg Med Chem 15, 5620–5628.

FEBS Journal 276 (2009) 6689–6700 ª 2009 The Authors Journal compilation ª 2009 FEBS

6700