Oncogene (2014), 1–11 © 2014 Macmillan Publishers Limited All rights reserved 0950-9232/14 www.nature.com/onc
ORIGINAL ARTICLE
miR-34a-5p suppresses colorectal cancer metastasis and predicts recurrence in patients with stage II/III colorectal cancer J Gao1,2,5, N Li1,5, Y Dong2,3, S Li1, L Xu2, X Li2, Y Li1, Z Li4, SS Ng3, JJ Sung2, L Shen1 and J Yu2 Although surgery remains the mainstay of curative treatment for colorectal cancer (CRC), many patients still have high chance to experience disease relapse. It is therefore imperative to identify prognostic markers that can help predict the clinical outcomes of CRC. Aberrant microRNA expression holds great potential as diagnostic and prognostic biomarker for CRC. Here we aimed to investigate clinical potential of miR-34a-5p as a prognostic marker for CRC recurrence and its functional significance. First, we validated that miR-34a-5p was downregulated in CRC tumour tissues (P o 0.05). The expression level of tissue miR-34a-5p was then evaluated in two independent cohorts of 268 CRC patients. miR-34a-5p expression was positively correlated with disease-free survival in two independent cohorts (cohort I: n = 205, P o0.001; cohort II: n = 63, P = 0.006). Moreover, the expression of miR-34a-5p was an independent prognostic factor for CRC recurrence by multivariate analysis (P o 0.001 for cohort I, P = 0.007 for cohort II). Ectopic expression of miR-34a-5p in p53 wild-type colon cancer cell HCT116 significantly inhibited cell growth, migration, invasion and metastasis. miR-34a-5p induced cell apoptosis, cell cycle arrest at G1 phase and p53 transcription activity in HCT116 cells, but not in the HCT116 p53 knockout (p53− / −) cells. miR-34a-5p significantly suppressed the HCT116 growth in vivo, whereas it showed no effect on the HCT116 p53− / − xenograft, indicating that the growth-inhibiting effect by miR-34a-5p was dependent on p53. In addition, the expression level of miR-34a-5p in patients with p53-positive expression was higher than that in patients with p53negative expression (P o 0.01). In conclusion, miR-34a-5p inhibits recurrence of CRC through inhibiting cell growth, migration and invasion, inducing cell apoptosis and cell cycle arrest in a p53-dependent manner. Oncogene advance online publication, 3 November 2014; doi:10.1038/onc.2014.348
INTRODUCTION Colorectal cancer (CRC) is potentially curable by surgery. Radical surgery is still the major method to treat CRC.1 However, recurrence following radical surgery is the leading factor to imply poor prognosis.2 Although adjuvant chemotherapy can benefit patients with stage III disease, its role in patients with stage II CRC remains controversial due to the lack of data showing a definite benefit of chemotherapy in this group of patients as a whole. It is crucial to identify reliable predictive biomarkers that can guide personalized chemotherapy. The molecular aetiology of CRC has facilitated the identification of promising biomarkers.3,4 However, despite significant methodological progress, suitable biomarkers have not been identified yet for guiding the adjuvant treatment for stage II CRC cancer. There is increasing evidence indicating that microRNAs (miRNAs) are critically important in the development of CRC.5 Aberrant miRNA expression holds great potential as diagnostic and prognostic biomarker for CRC.6 We have previously reported
that tumour-associated miRNAs can be detected in the plasma or stool samples from patients suffering from CRCs as novel noninvasive biomarkers for the detection of this disease.7–9 Some miRNAs have been demonstrated to be associated with the recurrence of human cancers, including miR-375/miR-142-5p in gastric cancer, miR-21/miR-155 in non-small cell lung cancer and miR-29a-5p in liver cancer.10–12 miR-34a-5p (MIMAT0000255) has been reported to be a direct transcriptional target of p53 and downregulated in several tumours.13,14 miR-34a-5p has been reported to inhibit cell invasion and migration in in vitro experiments,15–18 which suggested that miR-34a-5p might have some role in inhibiting tumour recurrence. However, the functional role of miR-34a-5p in CRC and its potential utility for the prediction of CRC recurrence are still unknown. In this study, we investigated the clinical impact of miR-34a-5p as a recurrence biomarker for stage II and stage III CRC patients. Moreover, the biological functions and the possible molecular basis of miR-34a5p in CRC were characterized in vitro and in vivo.
1 Department of Gastrointestinal Oncology, Key laboratory of Carcinogenesis and Translational Research (Ministry of Education), Peking University Cancer Hospital & Institute, Beijing, China; 2Institute of Digestive Disease, Department of Medicine and Therapeutics, State Key Laboratory of Digestive Disease, CUHK Shenzhen Research Institute, Li Ka Shing Institute of Health Sciences, The Chinese University of Hong Kong, Shatin, Hong Kong; 3Department of Surgery, The Chinese University of Hong Kong, Shatin, Hong Kong and 4Department of Pathology, Key laboratory of Carcinogenesis and Translational Research (Ministry of Education), Peking University Cancer Hospital & Institute, Beijing, China. Correspondence: Professor J Yu, Institute of Digestive Disease, Department of Medicine and Therapeutics, State Key Laboratory of Digestive Disease, CUHK Shenzhen Research Institute, Li Ka Shing Institute of Health Sciences, Prince of Wales Hospital, The Chinese University of Hong Kong, Shatin, New Territories, Hong Kong or Professor L Shen, Department of Gastrointestinal Oncology, Key laboratory of Carcinogenesis and Translational Research (Ministry of Education), Peking University Cancer Hospital & Institute, Fu-Cheng Road 52, Hai-Dian District, Beijing 100142, China. E-mail: [email protected] or [email protected] 5 These authors contributed equally to this work. Received 21 June 2014; revised 24 August 2014; accepted 29 August 2014
miR-34a-5p predicts recurrence of CRC J Gao et al
2
RESULTS The clinicopathological characteristics of CRC patients The characteristics of 268 CRC patients related to the recurrence status are shown in Table 1. There were no significant differences in age, gender, depth of penetration, differentiation and adjuvant chemotherapy between patients with or without recurrence, except for lymph node metastasis in cohort I and cohort II, tumour site in cohort II and lymph-vascular invasion in cohort I. miR-34a-5p was downregulated in primary CRC tissues Quantitative PCR was conducted to detect miR-34a-5p expression in 10 CRC tumours and matched adjacent non-tumour tissues. Among 10 paired CRC tissues and adjacent non-tumour tissues, 9 tumours showed lower miR-34a-5p expression when compared with matched non-tumour tissues (P o0.05; Figure 1a). Moreover, p53 levels were positively correlated with miR-34a-5p expression in the tumour specimen (r2 = 0.9678, P o 0.0001; Supplementary Figure S1A). miR-34a-5p was downregulated in CRC patients with recurrence We examined the expression of miR-34a-5p in patients with or without recurrence. The RNA quality and normalizer RNU6B level are comparable between recurrence or non-recurrence groups. The expression level of miR-34a-5p in tumour tissue of patients with recurrence was significantly lower than that in patients without recurrence (P o0.001; Supplementary Figure S1B). Table 1.
miR-34a-5p expression as an independent predictor for CRC recurrence We further evaluated the correlation of miR-34a-5p expression with CRC clinicopathological features. Receiver-operating characteristic curve of miR-34a-5p has area under the curve value of 0.779 (Figure 1b). A cut-off value (0.3072866 calculated by 2− ΔΔCt method) that maximizes the sum of sensitivity and specificity was selected. miR-34a-5p had a sensitivity of 81.9% and a specificity of 72% in cohort I. miR-34a-5p expression higher than the cut-off level was defined as high expression. According to the cut-off level from receiver-operating characteristic curve (Figure 1b), 103 CRC patients showed high miR-34a-5p expression and 102 CRC patients showed low miR-34a-5p expression in cohort I, and 32 patients showed high miR-34a-5p expression and 31 patients showed low miR-34a-5p expression in cohort II. miR-34a-5p was not associated with clinicopathological features including sex, age, primary tumour site, depth of penetration, lymph node metastasis or differentiation (data not shown). The 3-year recurrence rate in patients with miR-34a-5p low expression was significantly higher than that in patients with miR-34a-5p high expression (P o 0.001 for cohort I, Figure 1c1). The prognostic potential of miR-34a-5p was further analysed in stage II (n = 91) and III (n = 114) of cohort I CRC patients separately. Kaplan–Meier analysis suggests that miR-34a-5p is a strong independent predictor for recurrence in both stage II (P = 0.001) and stage III (P o 0.001) patients (Figure 1d). Among patients with recurrence in cohort I, the median disease-free survival in patients with miR-34a-5p low expression (13.9 months, 95% confidence
Clinicalpathological characteristics of patients in cohort I and II
Characteristics
Recurrence group
Gender Male Female
P-value
Cohort I (n = 205) Non-recurrence group
Recurrence group
No.
%
No.
%
67 33
67.0 33.0
64 41
61.0 39.0
Age (years) Median (range)
59 (28–87)
0.368
P-value
Cohort II (n = 63) Non-recurrence group
No.
%
No.
%
21 12
63.6 36.4
18 12
60.0 40.0
60 (30–76)
52 (28–87)
0.767
44 (42–74)
Tumour site Colon Rectum
48 52
48.0 52.0
48 57
45.7 54.3
0.743
10 23
30.3 69.7
19 11
63.3 36.7
0.009
Depth of penetration T3 T4
31 69
31.0 69.0
35 70
33.3 66.7
0.721
14 19
42.4 57.6
11 19
36.7 63.3
0.641
Lymph node N0 N1 N2
30 31 39
30.0 31.0 39.0
44 42 19
41.9 40.0 18.1
0.004
13 8 12
39.4 24.2 36.4
18 10 2
60.0 33.3 6.7
0.018
Differentiation Moderate/high Low
73 27
73.0 27.0
81 24
77.1 22.9
0.493
24 9
72.7 27.3
22 8
73.3 26.7
0.957
Lymph-vascular Invasion Yes No
31 69
31.0 69.0
18 87
17.1 82.9
0.020
7 26
21.2 78.8
6 24
20.0 80.0
0.905
Adjuvant chemotherapy Yes No
75 25
75.0 35.0
78 27
74.3 25.7
0.906
25 8
75.8 24.2
24 6
80.0 20.0
0.686
Oncogene (2014), 1 – 11
© 2014 Macmillan Publishers Limited
miR-34a-5p predicts recurrence of CRC J Gao et al
3
Figure 1. miR-34a-5p was frequently downregulated in primary CRC tissues, and Kaplan–Meier curves of disease-free survival in patients. (a) miR-34a-5p was significantly downregulated in CRC tumour tissues compared with their matched adjacent non-tumour tissues by quantitative PCR. The miR-34a-5p level was normalized to the internal control RNU6B. (b) The receiver-operating characteristic curve of miR-34a-5p expression. The sensitivity and specificity were 81.9 and 72%, separately, at the cut-off level of 0.3072866 (2− ΔΔCt). (c) Patients with miR-34a-5p low expression had higher risk of recurrence than patients with miR-34a-5p high expression in cohort I (c1) and cohort II (c2). (d) Low level of miR-34a-5p is significantly associated with recurrence in both stage II (P = 0.001) and stage III (Po0.001) patients. © 2014 Macmillan Publishers Limited
Oncogene (2014), 1 – 11
miR-34a-5p predicts recurrence of CRC J Gao et al
4 interval = 12.33–15.47) was shorter than that in patients with miR-34a-5p high expression (20.13 months, 95% confidence interval = 16.96–23.30; P o 0.05; Supplementary Figure S1C). We next validated our result in an independent cohort of 63 stage II/III CRC patients. In consistent with the survival result of cohort I, the patients in cohort II with low expression of miR-34a-5p have a high risk of disease relapse (P = 0.006, Figure 1c2). Multivariate Cox regression analysis confirmed that miR-34a-5p expression was an independent predictor for recurrence (hazard ratio = 3.819, 95% confidence interval = 2.438–5.983, Po 0.001 for cohort I; hazard ratio = 2.973, 95% confidence interval = 1.339– 6.602, P = 0.007 for cohort II; Table 2).
Ectopic expression of miR-34a-5p inhibited the growth of colon cancer cells Downregulation of miR-34a-5p in CRC tissues suggested that miR-34a-5p might act as a tumour suppressor. miR-34a-5p is a direct transcript target of p53. Reciprocally, miR-34a-5p enhances p53 level and activates p53 signalling pathway through positive feedback loops.19 To assess the biological function of miR-34a-5p and whether its tumour suppressive capacity is dependent on p53 status, we investigated the effect of ectopic expression of miR-34a-5p on the growth in p53 wild-type HCT116 colon cancer cell line and the HCT116 p53 − / − cell line in which both alleles of p53 were inactivated by homologous recombination. Ectopic expression of miR-34a-5p was confirmed by quantitative PCR (Figure 2a). Cell viability assay indicated that ectopic expression of miR-34a-5p could inhibit the growth of HCT116 cells (Po 0.01), but had no effect on HCT116 p53− / − cells (Figure 2b), which was further validated by colony formation assay. For HCT116 cells, the colonies formed by miR-34a-5p ectopic expression were significantly fewer than those formed by miR-Ctrl (P o0.05). However, no difference was observed in HCT116 p53− / − cells (Figure 2c).
miR-34a-5p suppressed the colon cancer cell growth in vivo HCT116 and HCT116 p53− / − cells transfected with miR-34a-5p or miR-Ctrl were injected subcutaneously into nude mice to establish xenografts in vivo. Our results demonstrated that, in contrast to HCT116 cells transfected with miR-Ctrl, the growth of tumours formed by HCT116 cells transfected with miR-34a-5p was significantly suppressed (P o0.05), but no difference was observed in tumours established by HCT116 p53− / − cells transfected with miR-34a-5p or miR-Ctrl (Figure 2d), indicating that the growth-inhibiting effect by miR-34a-5p was dependent on p53. Table 2.
miR-34a-5p attenuated migration, invasion and metastasis ability of HCT116 cells in vitro and in vivo miR-34a-5p was downregulated in CRC patients with recurrence, indicating that miR-34a-5p might have a role in inhibiting cell migration and invasion. We assessed the potential role of miR-34a5p in colon cancer cell migration and invasion by wound healing assay and Matrigel invasion assay. The results showed that, compared with control cells, miR-34a-5p could significantly inhibit the migration (P o0.001; Figure 3a) and invasion (P o0.05; Figure 3b) ability in HCT116 cells, but not in HCT116 p53− / − cells. HCT116 cells with stable expression of miR-34a-5p or control were injected intravenously through the tail vein into nude mice. The results indicated that, compared with the pMIR34a-5p stable expression group (no metastatic focus was found in seven mice), the number of metastatic foci in the lung tissues of control group was increased (numerous metastatic foci in one mouse, two metastatic foci in two mice, no metastatic foci in four mice; P = 0.0507; Figure 3c), indicating the inhibiting effect of metastasis by miR-34a-5p. We investigated the key regulators to be involved in the process of cell growth, migration and invasion in several malignancies. It was reported that silent information regulator 1 (SIRT1), c-Myc, NOTCH1, matrix metalloproteinase 1 (MMP1) and BCL2 were miR-34a-5p targets.20–23 We found that miR-34a-5p inhibited the expression of SIRT1, c-Myc, NOTCH1 and MMP1, but not BCL2 in HCT116 cells (Figure 3d). The inhibition effects were more prominent in p53 wild-type HCT116 compared with p53− / − HCT116 cells (Supplementary Figure S2). Moreover, Krüppel-like factor 4 (KLF4) was reported to act as an oncogene in colon cancer and a direct target of miR-34a-5p in liver cancer.24,25 We attached the predicted KLF4 3’UTR-binding sites as well as its mutant form to firefly luciferase reporter (Figure 4a). HCT116 cells transfected with miR-34a-5p repressed wild-type KLF4–3’UTR reporter activity. On the other hand, miR-34a-5p showed no inhibition effect on the mutant KLF4-3’UTR reporter activity (Figure 4b). Ectopic expression of miR-34a-5p induces KLF4 mRNA degradation and suppresses the KLF4 protein level (Figures 4c and d). These results indicate that miR-34 also targeted KLF4 in CRC cells. Taken together, our results add further weight for the tumour suppressive function of miR-34a-5p in CRC. miR-34a-5p induced cell cycle arrest at G1 phase by activation of p53/p21 pathway To explore the potential mechanisms responsible for the inhibitory effect of miR-34a-5p in CRC cells, cell cycle analysis was conducted. The percentage of cells accumulated in G1 phase was significantly increased in HCT116 cells after transfection of miR-34a-5p (65.3%) compared with control (49.8%; P o0.05; Figure 5a). Concomitant with cell cycle arrest at G1 phase, the
Multivariate Cox regression analysis of potential predictor for recurrence
Variates
Age (⩾50/ o50) Sex (male/female) Tumour site (colon/rectum) Depth of penetration (T4/T3) Lymph node (N2/N1/N0) Differentiation (poor/good) Lymph-vascular invasion (yes/no) Adjuvant chemotherapy (no/yes) miR-34a-5p expression (low/high)
Cohort I
Cohort II
HR
95% CI
P-value
HR
95% CI
P-value
1.161 1.106 1.159 1.133 1.384 1.215 1.102 1.047 3.819
0.742–1.815 0.729–1.679 0.783–1.716 0.742–1.731 1.068–1.793 0.781–1.890 0.697–1.743 0.666–1.646 2.438–5.983
0.514 0.635 0.461 0.563 0.014 0.388 0.676 0.843 o 0.001
0.905 1.082 0.425 1.203 2.120 1.125 1.923 1.147 2.973
0.417–1.961 0.485–2.416 0.177–1.019 0.526–2.748 1.321–3.403 0.490–2.577 0.724–5.102 0.680–1.534 1.339–6.602
0.799 0.847 0.055 0.661 0.002 0.782 0.190 0.099 0.007
Abbreviations: CI, confidence interval; HR, hazard ratio.
Oncogene (2014), 1 – 11
© 2014 Macmillan Publishers Limited
miR-34a-5p predicts recurrence of CRC J Gao et al
5
Figure 2. Ectopic expression of miR-34a-5p inhibited the growth of HCT116 cells in vitro and in vivo. (a) Ectopic expression of miR-34a-5p in HCT116 and HCT116 p53 − / − cell lines was confirmed by quantitative PCR. (b) miR-34a-5p inhibited cell viability in HCT116 cells, but not in HCT116 p53− / − cells. (c) Colony formation assay confirmed the inhibitory effect of miR-34a-5p on HCT116 cells, but not on HCT116 p53− / − cells (n = 3, mean ± s.d.). (d) Tumour growth curve of xenografts derived from HCT116 and HCT116 p53− / − cells transfected with miR-34a-5p or miR-Ctrl. Data are mean ± s.d. (n = 5/group), arrows indicated the time of miRNA intratumoural injection.
expressions of p53 and its downstream proteins p21 and p27 were upregulated, with cyclin-dependent kinase (CDK4) and cyclin D1 downregulated (Figure 5c). However, the effect of miR-34a-5p on G1 cell cycle regulation was not observed in HCT116 p53− / − cells (Figure 5b), and the protein expression of cell cycle related regulators was not changed by miR-34a-5p in HCT116 p53− / − cells (Figure 5c).
miR-34a-5p induced apoptosis in HCT116 cells by activation of caspase-dependent pathway To determine whether the decrease in colon cell growth by miR-34a-5p was due to induction of apoptosis, the cellular apoptotic rate was appraised using annexin V-allophycocyanin (V-APC) and 7-amino-actinomycin (7-AAD) staining by flow cytometry. As shown in Figures 5d and e, compared with control, the proportion of early apoptotic cells in HCT116 cell line was significantly increased after transfection of miR-34a-5p (P o 0.05; Figure 5d), this was not found in HCT116 p53− / − cells (Figure 5e). Induction of apoptosis was further confirmed by the analysis of the expression of apoptosis-related proteins by western blot. As shown in Figure 5f, the protein level of the active forms of caspase-9, caspase-7, nuclear poly (ADP-ribose) polymerase (PARP), Bak and Bax was enhanced, but the anti-apoptosis gene survivin was inhibited in the miR-34a-5ptransfected HCT116 cells compared with the control cells, but not in HCT116 p53− / − cells. © 2014 Macmillan Publishers Limited
miR-34a-5p increased p53 luciferase reporter activity in HCT116 cells From the above results, we found that miR-34a-5p had a strong tumour suppressive ability for colon cancer in the presence of p53, and the inhibitory effect of miR-34a-5p abolished when p53 was absent. To further elucidate the interaction of miR-34a-5p with p53, p53 luciferase reporter was transfected together with miR-34a-5p or miR-Ctrl in HCT116 and HCT116 p53− / − cells. As expected, p53 luciferase activity was significantly increased after transfection of miR-34a-5p in HCT116 cells (P o 0.001), but not in HCT116 p53− / − cells (Figure 6a). However, a direct interplay between miR-34a-5p and p53 was not demonstrated due to the reported evidence that p53 directly regulates miR-34a-5p and miR-34a-5p indirectly regulates p53.
miR-34a-5p expression was positively associated with p53 expression in primary CRC tumour tissues Among 205 CRC cases in cohort I, 151 patients had p53-positive expression and 54 patients had p53-negative expression as detected by immunohistochemistry (Figure 6b). The expression level of miR-34a-5p in patients with p53-positive expression was higher than that in patients with p53-negative expression (P o 0.01; Figure 6c). Moreover, miR-34a-5p could predict recurrence better in patients with p53-positive expression (P o 0.001) than in patients with p53-negative expression (P o 0.05; Figure 6d). Oncogene (2014), 1 – 11
miR-34a-5p predicts recurrence of CRC J Gao et al
6
Figure 3. Ectopic expression of miR-34a-5p inhibited the migration and invasion of HCT116 cells in vitro and in vivo. (a) Ectopic expression of miR-34a-5p significantly inhibited cell migration in HCT116 cells by wound healing assay. The degree of wound sealing was calculated according to the distance of two flanks of the wound (n = 3, mean ± s.d.). (b) Compared with control, miR-34a-5p attenuated cell invasion ability in HCT116 cells by counting cell numbers passed through the Matrigel membrane (n = 3, mean ± s.d.). (c) miR-34a-5p suppressed tumour metastasis in vivo. (c1) HCT116 cell expressed miR-34a-5p and control plasmid respectably. (c2) miR-34a-5p expression level was confirmed by quantitative PCR. (c3) Representative images of the lung tissues from nude mice, black arrow indicated the macroscopic metastatic foci. (c4) Representative hematoxylin and eosin-stained lung sections containing metastatic foci. (d) Several important tumorgenic genes were detected by western blot analysis.
DISCUSSION In this study, we demonstrated that miR-34a-5p was downregulated in primary CRC tumour tissues. Promoter hypermethylation of miR-34a-5p was reported to be one of the mechanisms resulting in downregulation of miR-34a-5p expression.26 In keeping with our finding, downregulation of miR-34a-5p has been reported in other solid cancers, including gastric cancer, liver cancer and non-small cell lung cancer.15–18 However, the clinical impact of miR-34a-5p dysregulation in human cancer remains to be investigated. We investigated the association between miR-34a-5p expression and disease recurrence in CRC. Among our two cohorts of stage II and III primary CRC tumours, miR-34a-5p expression level in patients with recurrence was significantly lower than that in patients without recurrence (P o 0.001; Supplementary Figure S1B). Besides, miR-34a-5p low expression was related to poor disease-free survival and high 3-year recurrence rate by Oncogene (2014), 1 – 11
Kaplan–Meier curves and log-rank test (Figure 1c). These results imply that the miR-34a-5p downregulation may specifically predict the most aggressive and fatal types of CRC in cases with stage II/III disease. Indeed, multivariate Cox regression analysis indicated that miR-34a-5p was a potential predictor for the recurrence of patients with stage II/III CRC (Table 2). In supporting this, recent studies have demonstrated that several other miRNAs such as miR-21, miR-362-3p and miR-93 were associated with recurrence of CRC.27–29 Their results are complementary to ours and suggest that dysregulated miRNAs in tumours could predict recurrence of CRC. It was reported that miR-34a-5p was a direct transcript target of p53. Reciprocally, miR-34a enhances p53 level and activates p53 signalling pathway through the following mechanisms: (1) miR-34a-5p/MDM4/p53 loop. miR-34a-5p represses its direct target MDM4, leading to stabilization of p53 and sustainable expression of miR-34a-5p;30 (2) miR-34a-5p/SIRT1/p53 loop. miR-34a-5p inhibits its direct target SIRT1, leading to an increase © 2014 Macmillan Publishers Limited
miR-34a-5p predicts recurrence of CRC J Gao et al
7
Figure 4. KLF4 is a direct target of miR-34a-5p in colon cancer. (a) Human KLF4 3’UTR-binding site for miR-34a-5p. The miR-34a-5p wild-type binding sequence or its mutated form was inserted into C-terminal of the luciferase gene to generate pMIR-KLF4-3’UTR or pMIR-KLF4-mut3’UTR, respectively. (b) miR-34a-5p targeted the wild-type but not the mutant 3′UTR of KLF4. The data are means ± s.d. (c) Ectopic expression of miR-34a-5p downregulated KLF4 mRNA expression in HCT116 cells as determined by quantitative reverse transcription–PCR. (d) miR-34a-5p decreased KLF4 protein level in HCT116 detected by western blot.
in acetylated p53. Acetylated p53 at amino acid K382 abrogates p53 ubiquitination by MDM2 and therefore stabilizes p53 protein.19,31,32 These data collectively suggest that miR-34a-5p indirectly upregulates p53 through suppressing its direct targets MDM4 and SIRT1. Biological functions of miR-34a-5p were therefore analysed in HCT116 and HCT116 p53− / − cells in this study. Our results showed that ectopic expression of miR-34a-5p in HCT116 cells could inhibit cell proliferation, decrease colony formation, suppress cell mobility and invasion and inhibit tumorigenesis in nude mice. However, these effects were not observed in HCT116 p53− / − cells. Moreover, miR-34a-5p could increase p53 luciferase activity in HCT116 cells (P o0.001), but not in HCT116 p53− / − cells (Figure 6a). These results suggested that the tumour inhibitory effect of miR-34a-5p in CRC was dependent on p53. The positive correlation between miR-34a-5p and p53 expression in primary CRC patients further confirmed the importance of p53 in involving in the functional suppression of CRC by miR-34a-5p, although further validation is warranted in a large-scale samples with p53 mutant status. As shown in Figures 6c and d, miR-34a-5p level in patients with p53-positive expression was higher than those with p53-negative expression. miR-34-5p expression is correlated to disease-free survival in either p53-positive or -negative tumour patients. Cancer stem cells are a population of heterogeneous tumours that are crucial for tumour initiation and recurrence. Previous studies suggested that © 2014 Macmillan Publishers Limited
miR-34a-5p inhibited colon cancer stem cell self-renewal and promoted daughter cell differentiation by targeting NOTCH pathway.22 However, the cell lines we used in the current study are all differentiated cells and therefore miR-34a-5p showed a limited effect on the p53− / − cells in a short-term observation period. These may explain the discrepancy between in vitro assay and clinical outcome results. Several target genes regulated by miR-34a-5p were reported to be involved in the process of cell growth, migration and invasion in several malignancies, such as miR-34a-5p target SIRT1, c-Myc, NOTCH1, MMP1, KLF4 and BCL2.19–25 We demonstrated that miR-34a-5p inhibited protein expression of oncogenic genes SIRT1, c-Myc, NOTCH1 and KLF4 in colon cancer cells (Figure 3d). The inhibition effects are more prominent in p53 wild-type HCT116 compared with p53− / − HCT116 cells due to the miR-34a5p/MDM4/p53 loop and miR-34a-5p/SIRT1/p53 loop. MMP1 involved in cancer invasion and metastasis by degrading the extracellular matrix. MMP1 lacks potential miR-34a-5p-binding sequence; previous studies suggest that wild-type p53 represses MMP1 transcription through regulation of its promoter activity.33 Our data demonstrated that miR-34a suppressed MMP1 at both mRNA level and protein level in a p53-dependent manner, suggesting an indirectly regulation of MMP1 by miR-34a-5p. This mechanism could, at least, partially explain the p53-dependent miR-34a-5p repression of cell migration and invasion. Thus, the Oncogene (2014), 1 – 11
miR-34a-5p predicts recurrence of CRC J Gao et al
8
Figure 5. miR-34a-5p induced cell cycle arrest at G1 phase and early apoptosis in HCT116 cells. Ectopic expression of miR-34a-5p induced cell cycle arrest at G1 phase only in HCT116 cells (a), but not in HCT116 p53− / − cells (b; n = 3, mean ± s.d.). (c) The expressions of p53 and its downstream proteins p21 and p27 were upregulated, and CDK4/cyclin D1 were downregulated after transfection of miR-34a-5p in HCT116 cells, not in HCT116 p53− / − cells. (d) Early apoptosis of HCT116 cells was increased by miR-34a-5p. (e) No difference in cell apoptosis was found in HCT116 p53− / − cells transfected with miR-34a-5p or miR-Ctrl (n = 3, mean ± s.d.). (f) Effect of miR-34a-5p on expressions of several apoptosis-related proteins detected by western blot analysis.
tumour suppressive effect of miR-34a-5p at least in part mediated by the inhibition of these important tumorigenic factors in CRC as summarized in Figure 6e. To identify the possible mechanisms involved in the inhibitory effect of miR-34a-5p, cell cycle and cell apoptosis were conducted. In keeping with the tumour suppressive role, fluorescenceactivated cell sorting analysis of the effects of miR-34a-5p on the cell cycle in colon cancer cells revealed a concomitant increase of cells in G0/G1, which lead to the inhibition of cell proliferation. On the basis of the immunoblot analysis of negative cell cycle regulators, G0/G1 arrest by miR-34a-5p was most likely associated with the induction of p53, p21 and p27 (Figure 5c). It is well known that p53 can induce cell cycle arrest through transcriptional upregulation of the CDK inhibitor p21 and p27, a major player in G1 arrest.34,35 Cyclin D1–CDK4 complex accumulation is of great importance for G0/G1 cell cycle progression. In this regard, miR-34a-5p causes G1 cell cycle arrest in colon cancer that was at least by a mechanism involving upregulation of protein expression Oncogene (2014), 1 – 11
of p53, p21 and p27, and downregulation of cyclin D1 and CDK4, which was summarized in Figure 6e. In addition to inhibition of cell proliferation, the growthinhibitory effect of miR-34a-5p was also related to induction of apoptosis. We observed that induction of miR-34a-5p-mediated apoptosis occurs by the modulation of caspase apoptosis pathway. The protein expression of the downstream apoptosis executors caspase-9, caspase-7, caspase-3 and PARP was upregulated. Activation of caspase-9 processes other effector caspase members, including caspase-3 and caspase-7 to initiate a caspase cascade, which further initiates the proteolytic cleavage of the nuclear enzyme PARP and causes apoptosis. Pro-apoptotic Bax and Bak have been implicated in the regulation of p53-dependent apoptosis.36 Survivin, an inhibitor of apoptosis protein, is highly expressed in most cancers and associated with increased tumorigenesis. Taken together, as described in Figure 6e, the induction of apoptosis by miR-34a-5p in CRC is mediated by © 2014 Macmillan Publishers Limited
miR-34a-5p predicts recurrence of CRC J Gao et al
9
Figure 6. miR-34a-5p increased p53 luciferase reporter activity, and the proposed scheme for inhibitory effect of miR-34a-5p. (a) Luciferase activity assay demonstrated that miR-34a-5p increased p53 luciferase activity by 1.5-fold in HCT116 cells, not in HCT116 p53− / − cells (n = 3, mean ± s.d.). (b) Representative results of p53 staining by immunohistochemistry (magnification: × 200). (c) miR-34a expression in patients with p53-positive expression was higher than that in patients with p53-negative expression. (d) miR-34a-5p predicted recurrence better in patients with p53-positive expression (Po 0.001) than in patients with p53-negative expression (P o0.05). (e) Proposed scheme of molecular basis for inhibitory effect of miR-34a-5p in CRC according to our results.
induction of intrinsic apoptotic caspase cascade, pro-apoptosis Bax and Bak and inhibition of anti-apoptotic gene survivin. Our study is a retrospective study. Further prospective studies are warranted for validation of the prognostic power of the candidate miRNA in CRC. In conclusion, we found for the first time that miR-34a-5p downregulation could predict high risk of recurrence in CRC patients after radical surgery, which may be used as a potential recurrence biomarker for the recurrence of the stage II/III CRC patients. miR-34a-5p acts as a tumour suppressor in CRC in a p53dependent manner. MATERIALS AND METHODS Patients and sample collection This study included two cohorts of CRC patients receiving radical surgery in two departments of Peking University Cancer Hospital from 2002 to 2010 (cohort I: n = 205; cohort II: n = 63). All patients were pathologically confirmed as stage II and stage III CRC (the Union for International Cancer Control staging system) and had snap-frozen surgical tumour or nontumour tissues stored at − 80 °C before any therapy. The clinicopathological characteristics of all patients with disease recurrence (n = 133, local recurrence or distant metastasis) or without disease recurrence (n = 135) within 3 years after surgery are shown in Table 1. Patients with stage II and III CRC, and with clinicopathological characteristics and follow-up information available, were included. We excluded patients if they had any previous chemoradiotherapy treatment, metachronous or synchronous cancers. The clinical data of patients were retrospectively obtained from their medical records and the last follow-up was in January 2012. Patients who developed any local recurrence or distant metastasis diagnosed by computerized tomography scan within 3 years after surgery are defined as patients with recurrence. All patients had given written © 2014 Macmillan Publishers Limited
informed consent for their tissues to be used in the future research. This study was approved by the Ethics Committee of Peking University Cancer Hospital.
Colon cancer cell lines HCT116 was obtained from the American Type Culture Collection (Manassas, VA, USA), and HCT116 p53− / − cell line was kindly provided by Professor Bert Vogelstein (Johns Hopkins University). Cells were cultured in McCoy’s 5A modified medium (Invitrogen, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum (Gibco BRL, Invitrogen) and incubated in a humidified 37 °C incubator supplemented with 5% CO2.
RNA extraction and miRNA expression analyses Hematoxylin and eosin staining was performed for each surgical sample. Tumour tissues samples with the cancer cells 460% were eligible for RNA isolation and real-time PCR analysis. Total RNA was freshly extracted from tissues or cell pellets using Trizol reagent according to the manufacturer's instructions (Invitrogen). The RNA samples with OD260/OD280 ratio ranged between 1.9 and 2.0 were considered as good quality. The reverse transcription and quantitative PCR for miR-34a-5p and endogenous control RNU6B were conducted by using TaqMan MicroRNA Assays (Applied Biosystems, Foster City, CA, USA). The relative expression of miR-34a-5p was calculated using the comparative Ct method.
Cell viability assay Cells were transfected with miR-34a-5p (MC11030; Life Technologies, Carlsbad, CA, USA) or negative control miRNA (miR-Ctrl, 4464058; Life Technologies) using lipofectamine 2000 (Invitrogen). After 48 h of transfection, cell viability was determined by the 3-(4,5-dimethylthiazol2-yl)-5-(3-carboxymethoxyphen-yl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS) assay (Promega, Madison, WI, USA) according to the manufacturer’s Oncogene (2014), 1 – 11
miR-34a-5p predicts recurrence of CRC J Gao et al
10 instructions. The absorbance was measured at 490 nm using a spectrophotometer once a day for consecutive 5 days. The MTS assay read was calculated relative to day 1.
Colony formation assay Cells were transfected with miR-34a-5p or miR-Ctrl using lipofectamine 2000 (Invitrogen). After 48 h of transfection, cells were subcultured in a sixwell plate (1000 cells/well) for about 10 days. Colonies were stained with 5% crystal violet solution and counted. All experiments were performed in triplicate wells.
In vivo metastasis assay miR-34a-5p expressing plasmid (PMIRH34a-5pPA-1) and control plasmid PMIRH000PA-1 were purchased from System Biosciences (San Francisco, CA, USA). Stable miR-34a-5p-expressing cell line and control cell line were created through lentivirus infection (Lentivector Expression Systems, System Biosciences). Female Balb/c nude mice (5 weeks old) were used (seven mice per group). Stable miR-34a-5p-expressing or control HCT116 cells (1 × 106) were injected intravenously through the tail vein into each nude mouse. All mice were killed 6 weeks after injection. The lungs from each mouse were excised and embedded in paraffin for hematoxylin and eosin staining. All animal experiments were performed in accordance with the animal experimental guidelines of Chinese University of Hong Kong.
Cell migration assay After cells were transfected with miR-34a-5p or miR-Ctrl for 48 h, cell migration was performed using wound healing assay. The degree of wound sealing was calculated according to the distance of two flanks of the wound once a day for 3 days.
Cell invasion assay Cells were transfected with miR-34a-5p or miR-Ctrl for 48 h, and cell invasion was performed by Matrigel invasion assay (BD Biosciences, Erembodegem, Belgium). Cells in the upper compartment of the chamber were suspended in serum-free medium, and the lower chamber contained medium supplemented with 20% fetal bovine serum. After 24 h incubation, cells that passed through the matrigel membrane were fixed and stained with crystal violet, followed by counting from five random microscopic fields.
Cell cycle assay After transfection with miR-34a-5p or miR-Ctrl for 48 h, cells were collected and fixed in 70% cold ethanol for at least 12 h at 4 °C. Cells were stained with 50 μg/ml propidium iodide (BD Biosciences) at room temperature for 30 min in dark, and cell cycle was performed by FACS Calibur system (BD Biosciences) and analysed by ModFit 3.0 software (Verity Software House, Topsham, ME, USA).
Annexin V apoptosis assay Cell apoptosis was conducted by staining with APC and 7-AAD (BD Biosciences) for 15 min at room temperature in dark, followed by flow cytometry within 1 h (BD Biosciences). Cell apoptosis was analysed by WinMDI 2.9 software (The Scripps Research Institute, La Jolla, CA, USA).
Western blot Total protein was extracted from cell pellets using CytoBuster Protein Extraction Reagent (Merck Millipore, Darmstadt, Germany). Protein concentration was measured by the DC protein assay method of Bradford (Bio-Rad, Hercules, CA, USA), and 10–20 mg of protein from each sample was separated on 12% SDS–polyacrylamide gel electrophoresis. After transfer, the nitrocellulose membrane (GE Healthcare, Piscataway, NJ, USA) was incubated with primary antibody at 4 °C overnight and secondary antibody at room temperature for 1 h (antibody list is shown in Supplementary Table S1). Proteins were visualized using ECL Plus Western Blotting Detection Reagents (GE Healthcare).
In vivo tumorigenicity HCT116 and HCT116 p53− / − cells (2 × 106) transfected with miR-34a-5p or miR-Ctrl were suspended in 0.1 ml phosphate buffered saline, and injected subcutaneously into the dorsal left flank of 5-week-old female Balb/c nude mice (five mice per group). miRNAs were prepared by pre-incubating 0.3 nmole miRNA with 2.5 μl lipofectamine 2000 (Invitrogen) for 15 min and injections were made in a final volume of 100 μl in McCoy's 5 A medium (Sigma-Aldrich, St Louis, MO, USA) per mouse. The injection was repeated every 3 days and consisted of three consecutive injections.37 Tumour was measured twice every week from the first injection and tumour volume was calculated by the formula V = L × W2 × 1/2 (V, volume; L, length; W, width of tumour). All animal experiments were performed in accordance with the animal experimental guidelines of Chinese University of Hong Kong. Oncogene (2014), 1 – 11
Vector construction and Luciferase reporter assay The potential miR-34a-5p-binding sites were predicted by miRanda (www. microRNA.org). Sequence of 56 bp segment with the wild-type or mutant seed region of KLF4 was synthesized and cloned into pMIR-REPORT luciferase vector (Applied Biosystems; Life Technologies). The mutant KLF4 3′UTR sequence was prepared by deleting nine nucleotides in the seed region. Cells (1 × 105/well) transiently transfected with miR-34a-5p or miRCtrl (at 20 nM final concentration) were seeded in 24-well plates. pMIRREPORT vector (195 ng/well) and pRL-TK vector (5 ng/well) were cotransfected using lipofectamine 2000 (Invitrogen). Cells were collected 48 h post transfection and luciferase activities were analysed by the dualluciferase reporter assay system (Promega). Cells in 24-well plates were co-transfected with p53 reporter plasmid (contains 13 copies of the wild-type p53 response elements) and miR-34a5p or miR-Ctrl using lipofectamine 2000 (Invitrogen). After 48 h, luciferase activity was analysed by dual-luciferase reporter assay system according to the manufacturer’s instructions (Promega).
Statistical analysis Statistical analysis was performed using SPSS 18.0 software (SPSS Inc., Chicago, IL, USA). The χ2-test was used to analyse the relationships between clinicopathological characteristics and miR-34a-5p expression. The Kaplan–Meier survival curve and log-rank test were used to describe disease-free survival to miR-34a-5p expression. Multivariate Cox regression model was employed to analyse predictive variety. The difference in miR-34a-5p expression between tumour and non-tumour tissues or patients with or without recurrence was compared by Mann–Whitney Utest. Repeated measures analysis of variance and χ2-test were used to compare the difference in cell or tumour growth and the numbers of metastatic foci between two groups of cells or mice. A cut-off value was selected using receiver-operating characteristic curves for reference, based on high sensitivity and specificity. Po0.05 was considered statistically significant.
CONFLICT OF INTEREST The authors declare no conflict of interest.
ACKNOWLEDGEMENTS This work was supported by National Natural Science Foundation of China (No. 81301853, 81072034), National High Technology Research and Development Program (No. 2012AA 02A 504, 2012AA 02A 506), Shenzhen Technology and Innovation Project Fund, Shenzhen (JSGG20130412171021059), China 863 program (2012AA02A506), Shenzhen Municipal Science and Technology R & D fund (JCYJ20120619152326450) and Shenzhen Virtual University Park Support Scheme to CUHK Shenzhen Research Institute.
REFERENCES 1 Jemal A, Bray F, Center MM, Ferlay J, Ward E, Forman D. Global cancer statistics. CA Cancer J Clin 2011; 61: 69–90. 2 Mirakhorli M, Rahman SA, Abdullah S, Vakili M, Rozafzon R, Khoshzaban A. Multidrug resistance protein 2 genetic polymorphism and colorectal cancer recurrence in patients receiving adjuvant FOLFOX-4 chemotherapy. Mol Med Rep 2013; 7: 613–617. 3 Baranwal S, Alahari SK. miRNA control of tumor cell invasion and metastasis. Int J Cancer 2010; 126: 1283–1290.
© 2014 Macmillan Publishers Limited
miR-34a-5p predicts recurrence of CRC J Gao et al
11 4 Wang Z, Li Y, Kong D, Ahmad A, Banerjee S, Sarkar FH. Cross-talk between miRNA and Notch signaling pathways in tumor development and progression. Cancer Lett 2010; 292: 141–148. 5 Ma Y, Li W, Wang H. Roles of miRNA in the initiation and development of colorectal carcinoma. Curr Pharm Des 2013; 19: 1253–1261. 6 Dong Y, Wu WKK, Wu CW, Sung JJY, Yu J, Ng SSM. MicroRNA dysregulation in colorectal cancer: a clinical perspective. Br J Cancer 2011; 104: 893–898. 7 Ng EKO, Chong WWS, Jin H, Lam EKY, Shin VY, Yu J et al. Differential expression of microRNAs in plasma of patients with colorectal cancer: a potential marker for colorectal cancer screening. Gut 2009; 58: 1375–1381. 8 Wu CW, Ng SSM, Dong YJ, Ng SC, Leung WW, Lee CW et al. Detection of miR-92a and miR-21 in stool samples as potential screening biomarkers for colorectal cancer and polyps. Gut 2012; 61: 739–745. 9 Wu CW, Ng SC, Dong Y, Tian L, Ng SSM, Leung WW et al. Identification of microRNA-135b in Stool as a Potential Noninvasive Biomarker for Colorectal Cancer and Adenoma. Clin Cancer Res 2014; 20: 2994–3002. 10 Zhang X, Yan Z, Zhang J, Gong L, Li W, Cui J et al. Combination of hsa-miR-375 and hsa-miR-142-5p as a predictor for recurrence risk in gastric cancer patients following surgical resection. Ann Oncol 2011; 22: 2257–2266. 11 Yang M, Shen H, Qiu C, Ni Y, Wang L, Dong W et al. High expression of miR-21 and miR-155 predicts recurrence and unfavourable survival in non-small cell lung cancer. Eur J Cancer 2013; 49: 604–615. 12 Zhu HT, Dong QZ, Sheng YY, Wei JW, Wang G, Zhou HJ et al. MicroRNA-29a-5p is a novel predictor for early recurrence of hepatitis B virus-related hepatocellular carcinoma after surgical resection. PLoS One 2012; 7: e52393. 13 Hermeking H. p53 enters the microRNA world. Cancer Cell 2007; 12: 414–418. 14 Yamakuchi M, Lowenstein CJ. MiR-34, SIRT1 and p53: the feedback loop. Cell Cycle 2009; 8: 712–715. 15 Wu J, Wu G, Lv L, Ren Y-F, Zhang X-J, Xue Y-F et al. MicroRNA-34a inhibits migration and invasion of colon cancer cells via targeting to Fra-1. Carcinogenesis 2012; 33: 519–528. 16 Ji Q, Hao X, Meng Y, Zhang M, Desano J, Fan D et al. Restoration of tumor suppressor miR-34 inhibits human p53-mutant gastric cancer tumorspheres. BMC Cancer 2008; 8: 266. 17 Li N, Fu H, Tie Y, Hu Z, Kong W, Wu Y et al. miR-34a inhibits migration and invasion by down-regulation of c-Met expression in human hepatocellular carcinoma cells. Cancer Lett 2009; 275: 44–53. 18 Gallardo E, Navarro A, Viñolas N, Marrades RM, Diaz T, Gel B et al. miR-34a as a prognostic marker of relapse in surgically resected non-small-cell lung cancer. Carcinogenesis 2009; 30: 1903–1909. 19 Yamakuchi M, Ferlito M, Lowenstein CJ. miR-34a repression of SIRT1 regulates apoptosis. Proc Natl Acad Sci USA 2008; 105: 13421–13426. 20 Li L, Yuan L, Luo J, Gao J, Guo J, Xie X. MiR-34a inhibits proliferation and migration of breast cancer through down-regulation of Bcl-2 and SIRT1. Clin Exp Med 2013; 13: 109–117.
21 Yamamura S, Saini S, Majid S, Hirata H, Ueno K, Deng G et al. MicroRNA-34a modulates c-Myc transcriptional complexes to suppress malignancy in human prostate cancer cells. PLoS One 2012; 7: e29722. 22 Bu P, Chen KY, Chen JH, Wang L, Walters J, Shin YJ et al. A microRNA miR-34a-regulated bimodal switch targets Notch in colon cancer stem cells. Cell Stem Cell 2013; 12: 602–615. 23 Chang TC, Wentzel EA, Kent OA, Ramachandran K, Mullendore M, Lee KH et al. Transactivation of miR-34a by p53 broadly influences gene expression and promotes apoptosis. Mol Cell 2007; 26: 745–752. 24 Leng Z, Tao K, Xia Q, Tan J, Yue Z, Chen J et al. Krüppel-like factor 4 acts as an oncogene in colon cancer stem cell-enriched spheroid cells. PLoS One 2013; 8: e56082. 25 Chen Q, Li L, Tu Y, Zheng LL, Liu W, Zuo XY et al. MiR-34a regulates apoptosis in liver cells by targeting the KLF4 gene. Cell Mol Biol Lett 2014; 19: 52–64. 26 Nair VS, Maeda LS, Ioannidis JPA. Clinical outcome prediction by microRNAs in human cancer: a systematic review. J Natl Cancer Inst 2012; 104: 528–540. 27 Christensen LL, Tobiasen H, Holm A, Schepeler T, Ostenfeld MS, Thorsen K et al. MiRNA-362-3p induces cell cycle arrest through targeting of E2F1, USF2 and PTPN1 and is associated with recurrence of colorectal cancer. Int J Cancer 2013; 133: 67–78. 28 Yang IP, Tsai HL, Hou MF, Chen KC, Tsai PC, Huang SW et al. MicroRNA-93 inhibits tumor growth and early relapse of human colorectal cancer by affecting genes involved in the cell cycle. Carcinogenesis 2012; 33: 1522–1530. 29 Kjaer-Frifeldt S, Hansen TF, Nielsen BS, Joergensen S, Lindebjerg J, Soerensen FB et al. The prognostic importance of miR-21 in stage II colon cancer: a populationbased study. Br J Cancer 2012; 107: 1169–1174. 30 Mandke P, Wyatt N, Fraser J, Bates B, Berberich SJ, Markey MP. MicroRNA-34a modulates MDM4 expression via a target site in the open reading frame. PLoS One 2012; 7: e42034. 31 Li L, Wang L, Li L, Wang Z, Ho Y, McDonald T et al. Activation of p53 by SIRT1 inhibition enhances elimination of CML leukemia stem cells in combination with imatinib. Cancer Cell 2012; 21: 266–281. 32 Li M, Luo J, Brooks CL, Gu W. Acetylation of p53 inhibits its ubiquitination by Mdm2. J Biol Chem 2002; 277: 50607–50611. 33 Sun Y, Wenger L, Rutter JL, Brinckerhoff CE, Cheung HS. p53 down-regulates human matrix metalloproteinase-1 (collagenase-1) gene expression. J Biol Chem 1999; 274: 11535–11540. 34 Tokino T, Nakamura Y. The role of p53-target genes in human cancer. Crit Rev Oncol Hematol 2000; 33: 1–6. 35 Mitrea DM, Yoon M-K, Ou L, Kriwacki RW. Disorder-function relationships for the cell cycle regulatory proteins p21 and p27. Biol Chem 2012; 393: 259–274. 36 Tobiume K. Involvement of Bcl-2 family proteins in p53-induced apoptosis. J Nippon Med Sch 2005; 72: 192–193. 37 Dong Y, Zhao J, Wu CW, Zhang L, Liu X, Kang W et al. Tumor suppressor functions of miR-133a in colorectal cancer. Mol Cancer Res 2013; 11: 1051–1060.
Supplementary Information accompanies this paper on the Oncogene website (http://www.nature.com/onc)
© 2014 Macmillan Publishers Limited
Oncogene (2014), 1 – 11
ORIGINAL ARTICLE
miR-34a-5p suppresses colorectal cancer metastasis and predicts recurrence in patients with stage II/III colorectal cancer J Gao1,2,5, N Li1,5, Y Dong2,3, S Li1, L Xu2, X Li2, Y Li1, Z Li4, SS Ng3, JJ Sung2, L Shen1 and J Yu2 Although surgery remains the mainstay of curative treatment for colorectal cancer (CRC), many patients still have high chance to experience disease relapse. It is therefore imperative to identify prognostic markers that can help predict the clinical outcomes of CRC. Aberrant microRNA expression holds great potential as diagnostic and prognostic biomarker for CRC. Here we aimed to investigate clinical potential of miR-34a-5p as a prognostic marker for CRC recurrence and its functional significance. First, we validated that miR-34a-5p was downregulated in CRC tumour tissues (P o 0.05). The expression level of tissue miR-34a-5p was then evaluated in two independent cohorts of 268 CRC patients. miR-34a-5p expression was positively correlated with disease-free survival in two independent cohorts (cohort I: n = 205, P o0.001; cohort II: n = 63, P = 0.006). Moreover, the expression of miR-34a-5p was an independent prognostic factor for CRC recurrence by multivariate analysis (P o 0.001 for cohort I, P = 0.007 for cohort II). Ectopic expression of miR-34a-5p in p53 wild-type colon cancer cell HCT116 significantly inhibited cell growth, migration, invasion and metastasis. miR-34a-5p induced cell apoptosis, cell cycle arrest at G1 phase and p53 transcription activity in HCT116 cells, but not in the HCT116 p53 knockout (p53− / −) cells. miR-34a-5p significantly suppressed the HCT116 growth in vivo, whereas it showed no effect on the HCT116 p53− / − xenograft, indicating that the growth-inhibiting effect by miR-34a-5p was dependent on p53. In addition, the expression level of miR-34a-5p in patients with p53-positive expression was higher than that in patients with p53negative expression (P o 0.01). In conclusion, miR-34a-5p inhibits recurrence of CRC through inhibiting cell growth, migration and invasion, inducing cell apoptosis and cell cycle arrest in a p53-dependent manner. Oncogene advance online publication, 3 November 2014; doi:10.1038/onc.2014.348
INTRODUCTION Colorectal cancer (CRC) is potentially curable by surgery. Radical surgery is still the major method to treat CRC.1 However, recurrence following radical surgery is the leading factor to imply poor prognosis.2 Although adjuvant chemotherapy can benefit patients with stage III disease, its role in patients with stage II CRC remains controversial due to the lack of data showing a definite benefit of chemotherapy in this group of patients as a whole. It is crucial to identify reliable predictive biomarkers that can guide personalized chemotherapy. The molecular aetiology of CRC has facilitated the identification of promising biomarkers.3,4 However, despite significant methodological progress, suitable biomarkers have not been identified yet for guiding the adjuvant treatment for stage II CRC cancer. There is increasing evidence indicating that microRNAs (miRNAs) are critically important in the development of CRC.5 Aberrant miRNA expression holds great potential as diagnostic and prognostic biomarker for CRC.6 We have previously reported
that tumour-associated miRNAs can be detected in the plasma or stool samples from patients suffering from CRCs as novel noninvasive biomarkers for the detection of this disease.7–9 Some miRNAs have been demonstrated to be associated with the recurrence of human cancers, including miR-375/miR-142-5p in gastric cancer, miR-21/miR-155 in non-small cell lung cancer and miR-29a-5p in liver cancer.10–12 miR-34a-5p (MIMAT0000255) has been reported to be a direct transcriptional target of p53 and downregulated in several tumours.13,14 miR-34a-5p has been reported to inhibit cell invasion and migration in in vitro experiments,15–18 which suggested that miR-34a-5p might have some role in inhibiting tumour recurrence. However, the functional role of miR-34a-5p in CRC and its potential utility for the prediction of CRC recurrence are still unknown. In this study, we investigated the clinical impact of miR-34a-5p as a recurrence biomarker for stage II and stage III CRC patients. Moreover, the biological functions and the possible molecular basis of miR-34a5p in CRC were characterized in vitro and in vivo.
1 Department of Gastrointestinal Oncology, Key laboratory of Carcinogenesis and Translational Research (Ministry of Education), Peking University Cancer Hospital & Institute, Beijing, China; 2Institute of Digestive Disease, Department of Medicine and Therapeutics, State Key Laboratory of Digestive Disease, CUHK Shenzhen Research Institute, Li Ka Shing Institute of Health Sciences, The Chinese University of Hong Kong, Shatin, Hong Kong; 3Department of Surgery, The Chinese University of Hong Kong, Shatin, Hong Kong and 4Department of Pathology, Key laboratory of Carcinogenesis and Translational Research (Ministry of Education), Peking University Cancer Hospital & Institute, Beijing, China. Correspondence: Professor J Yu, Institute of Digestive Disease, Department of Medicine and Therapeutics, State Key Laboratory of Digestive Disease, CUHK Shenzhen Research Institute, Li Ka Shing Institute of Health Sciences, Prince of Wales Hospital, The Chinese University of Hong Kong, Shatin, New Territories, Hong Kong or Professor L Shen, Department of Gastrointestinal Oncology, Key laboratory of Carcinogenesis and Translational Research (Ministry of Education), Peking University Cancer Hospital & Institute, Fu-Cheng Road 52, Hai-Dian District, Beijing 100142, China. E-mail: [email protected] or [email protected] 5 These authors contributed equally to this work. Received 21 June 2014; revised 24 August 2014; accepted 29 August 2014
miR-34a-5p predicts recurrence of CRC J Gao et al
2
RESULTS The clinicopathological characteristics of CRC patients The characteristics of 268 CRC patients related to the recurrence status are shown in Table 1. There were no significant differences in age, gender, depth of penetration, differentiation and adjuvant chemotherapy between patients with or without recurrence, except for lymph node metastasis in cohort I and cohort II, tumour site in cohort II and lymph-vascular invasion in cohort I. miR-34a-5p was downregulated in primary CRC tissues Quantitative PCR was conducted to detect miR-34a-5p expression in 10 CRC tumours and matched adjacent non-tumour tissues. Among 10 paired CRC tissues and adjacent non-tumour tissues, 9 tumours showed lower miR-34a-5p expression when compared with matched non-tumour tissues (P o0.05; Figure 1a). Moreover, p53 levels were positively correlated with miR-34a-5p expression in the tumour specimen (r2 = 0.9678, P o 0.0001; Supplementary Figure S1A). miR-34a-5p was downregulated in CRC patients with recurrence We examined the expression of miR-34a-5p in patients with or without recurrence. The RNA quality and normalizer RNU6B level are comparable between recurrence or non-recurrence groups. The expression level of miR-34a-5p in tumour tissue of patients with recurrence was significantly lower than that in patients without recurrence (P o0.001; Supplementary Figure S1B). Table 1.
miR-34a-5p expression as an independent predictor for CRC recurrence We further evaluated the correlation of miR-34a-5p expression with CRC clinicopathological features. Receiver-operating characteristic curve of miR-34a-5p has area under the curve value of 0.779 (Figure 1b). A cut-off value (0.3072866 calculated by 2− ΔΔCt method) that maximizes the sum of sensitivity and specificity was selected. miR-34a-5p had a sensitivity of 81.9% and a specificity of 72% in cohort I. miR-34a-5p expression higher than the cut-off level was defined as high expression. According to the cut-off level from receiver-operating characteristic curve (Figure 1b), 103 CRC patients showed high miR-34a-5p expression and 102 CRC patients showed low miR-34a-5p expression in cohort I, and 32 patients showed high miR-34a-5p expression and 31 patients showed low miR-34a-5p expression in cohort II. miR-34a-5p was not associated with clinicopathological features including sex, age, primary tumour site, depth of penetration, lymph node metastasis or differentiation (data not shown). The 3-year recurrence rate in patients with miR-34a-5p low expression was significantly higher than that in patients with miR-34a-5p high expression (P o 0.001 for cohort I, Figure 1c1). The prognostic potential of miR-34a-5p was further analysed in stage II (n = 91) and III (n = 114) of cohort I CRC patients separately. Kaplan–Meier analysis suggests that miR-34a-5p is a strong independent predictor for recurrence in both stage II (P = 0.001) and stage III (P o 0.001) patients (Figure 1d). Among patients with recurrence in cohort I, the median disease-free survival in patients with miR-34a-5p low expression (13.9 months, 95% confidence
Clinicalpathological characteristics of patients in cohort I and II
Characteristics
Recurrence group
Gender Male Female
P-value
Cohort I (n = 205) Non-recurrence group
Recurrence group
No.
%
No.
%
67 33
67.0 33.0
64 41
61.0 39.0
Age (years) Median (range)
59 (28–87)
0.368
P-value
Cohort II (n = 63) Non-recurrence group
No.
%
No.
%
21 12
63.6 36.4
18 12
60.0 40.0
60 (30–76)
52 (28–87)
0.767
44 (42–74)
Tumour site Colon Rectum
48 52
48.0 52.0
48 57
45.7 54.3
0.743
10 23
30.3 69.7
19 11
63.3 36.7
0.009
Depth of penetration T3 T4
31 69
31.0 69.0
35 70
33.3 66.7
0.721
14 19
42.4 57.6
11 19
36.7 63.3
0.641
Lymph node N0 N1 N2
30 31 39
30.0 31.0 39.0
44 42 19
41.9 40.0 18.1
0.004
13 8 12
39.4 24.2 36.4
18 10 2
60.0 33.3 6.7
0.018
Differentiation Moderate/high Low
73 27
73.0 27.0
81 24
77.1 22.9
0.493
24 9
72.7 27.3
22 8
73.3 26.7
0.957
Lymph-vascular Invasion Yes No
31 69
31.0 69.0
18 87
17.1 82.9
0.020
7 26
21.2 78.8
6 24
20.0 80.0
0.905
Adjuvant chemotherapy Yes No
75 25
75.0 35.0
78 27
74.3 25.7
0.906
25 8
75.8 24.2
24 6
80.0 20.0
0.686
Oncogene (2014), 1 – 11
© 2014 Macmillan Publishers Limited
miR-34a-5p predicts recurrence of CRC J Gao et al
3
Figure 1. miR-34a-5p was frequently downregulated in primary CRC tissues, and Kaplan–Meier curves of disease-free survival in patients. (a) miR-34a-5p was significantly downregulated in CRC tumour tissues compared with their matched adjacent non-tumour tissues by quantitative PCR. The miR-34a-5p level was normalized to the internal control RNU6B. (b) The receiver-operating characteristic curve of miR-34a-5p expression. The sensitivity and specificity were 81.9 and 72%, separately, at the cut-off level of 0.3072866 (2− ΔΔCt). (c) Patients with miR-34a-5p low expression had higher risk of recurrence than patients with miR-34a-5p high expression in cohort I (c1) and cohort II (c2). (d) Low level of miR-34a-5p is significantly associated with recurrence in both stage II (P = 0.001) and stage III (Po0.001) patients. © 2014 Macmillan Publishers Limited
Oncogene (2014), 1 – 11
miR-34a-5p predicts recurrence of CRC J Gao et al
4 interval = 12.33–15.47) was shorter than that in patients with miR-34a-5p high expression (20.13 months, 95% confidence interval = 16.96–23.30; P o 0.05; Supplementary Figure S1C). We next validated our result in an independent cohort of 63 stage II/III CRC patients. In consistent with the survival result of cohort I, the patients in cohort II with low expression of miR-34a-5p have a high risk of disease relapse (P = 0.006, Figure 1c2). Multivariate Cox regression analysis confirmed that miR-34a-5p expression was an independent predictor for recurrence (hazard ratio = 3.819, 95% confidence interval = 2.438–5.983, Po 0.001 for cohort I; hazard ratio = 2.973, 95% confidence interval = 1.339– 6.602, P = 0.007 for cohort II; Table 2).
Ectopic expression of miR-34a-5p inhibited the growth of colon cancer cells Downregulation of miR-34a-5p in CRC tissues suggested that miR-34a-5p might act as a tumour suppressor. miR-34a-5p is a direct transcript target of p53. Reciprocally, miR-34a-5p enhances p53 level and activates p53 signalling pathway through positive feedback loops.19 To assess the biological function of miR-34a-5p and whether its tumour suppressive capacity is dependent on p53 status, we investigated the effect of ectopic expression of miR-34a-5p on the growth in p53 wild-type HCT116 colon cancer cell line and the HCT116 p53 − / − cell line in which both alleles of p53 were inactivated by homologous recombination. Ectopic expression of miR-34a-5p was confirmed by quantitative PCR (Figure 2a). Cell viability assay indicated that ectopic expression of miR-34a-5p could inhibit the growth of HCT116 cells (Po 0.01), but had no effect on HCT116 p53− / − cells (Figure 2b), which was further validated by colony formation assay. For HCT116 cells, the colonies formed by miR-34a-5p ectopic expression were significantly fewer than those formed by miR-Ctrl (P o0.05). However, no difference was observed in HCT116 p53− / − cells (Figure 2c).
miR-34a-5p suppressed the colon cancer cell growth in vivo HCT116 and HCT116 p53− / − cells transfected with miR-34a-5p or miR-Ctrl were injected subcutaneously into nude mice to establish xenografts in vivo. Our results demonstrated that, in contrast to HCT116 cells transfected with miR-Ctrl, the growth of tumours formed by HCT116 cells transfected with miR-34a-5p was significantly suppressed (P o0.05), but no difference was observed in tumours established by HCT116 p53− / − cells transfected with miR-34a-5p or miR-Ctrl (Figure 2d), indicating that the growth-inhibiting effect by miR-34a-5p was dependent on p53. Table 2.
miR-34a-5p attenuated migration, invasion and metastasis ability of HCT116 cells in vitro and in vivo miR-34a-5p was downregulated in CRC patients with recurrence, indicating that miR-34a-5p might have a role in inhibiting cell migration and invasion. We assessed the potential role of miR-34a5p in colon cancer cell migration and invasion by wound healing assay and Matrigel invasion assay. The results showed that, compared with control cells, miR-34a-5p could significantly inhibit the migration (P o0.001; Figure 3a) and invasion (P o0.05; Figure 3b) ability in HCT116 cells, but not in HCT116 p53− / − cells. HCT116 cells with stable expression of miR-34a-5p or control were injected intravenously through the tail vein into nude mice. The results indicated that, compared with the pMIR34a-5p stable expression group (no metastatic focus was found in seven mice), the number of metastatic foci in the lung tissues of control group was increased (numerous metastatic foci in one mouse, two metastatic foci in two mice, no metastatic foci in four mice; P = 0.0507; Figure 3c), indicating the inhibiting effect of metastasis by miR-34a-5p. We investigated the key regulators to be involved in the process of cell growth, migration and invasion in several malignancies. It was reported that silent information regulator 1 (SIRT1), c-Myc, NOTCH1, matrix metalloproteinase 1 (MMP1) and BCL2 were miR-34a-5p targets.20–23 We found that miR-34a-5p inhibited the expression of SIRT1, c-Myc, NOTCH1 and MMP1, but not BCL2 in HCT116 cells (Figure 3d). The inhibition effects were more prominent in p53 wild-type HCT116 compared with p53− / − HCT116 cells (Supplementary Figure S2). Moreover, Krüppel-like factor 4 (KLF4) was reported to act as an oncogene in colon cancer and a direct target of miR-34a-5p in liver cancer.24,25 We attached the predicted KLF4 3’UTR-binding sites as well as its mutant form to firefly luciferase reporter (Figure 4a). HCT116 cells transfected with miR-34a-5p repressed wild-type KLF4–3’UTR reporter activity. On the other hand, miR-34a-5p showed no inhibition effect on the mutant KLF4-3’UTR reporter activity (Figure 4b). Ectopic expression of miR-34a-5p induces KLF4 mRNA degradation and suppresses the KLF4 protein level (Figures 4c and d). These results indicate that miR-34 also targeted KLF4 in CRC cells. Taken together, our results add further weight for the tumour suppressive function of miR-34a-5p in CRC. miR-34a-5p induced cell cycle arrest at G1 phase by activation of p53/p21 pathway To explore the potential mechanisms responsible for the inhibitory effect of miR-34a-5p in CRC cells, cell cycle analysis was conducted. The percentage of cells accumulated in G1 phase was significantly increased in HCT116 cells after transfection of miR-34a-5p (65.3%) compared with control (49.8%; P o0.05; Figure 5a). Concomitant with cell cycle arrest at G1 phase, the
Multivariate Cox regression analysis of potential predictor for recurrence
Variates
Age (⩾50/ o50) Sex (male/female) Tumour site (colon/rectum) Depth of penetration (T4/T3) Lymph node (N2/N1/N0) Differentiation (poor/good) Lymph-vascular invasion (yes/no) Adjuvant chemotherapy (no/yes) miR-34a-5p expression (low/high)
Cohort I
Cohort II
HR
95% CI
P-value
HR
95% CI
P-value
1.161 1.106 1.159 1.133 1.384 1.215 1.102 1.047 3.819
0.742–1.815 0.729–1.679 0.783–1.716 0.742–1.731 1.068–1.793 0.781–1.890 0.697–1.743 0.666–1.646 2.438–5.983
0.514 0.635 0.461 0.563 0.014 0.388 0.676 0.843 o 0.001
0.905 1.082 0.425 1.203 2.120 1.125 1.923 1.147 2.973
0.417–1.961 0.485–2.416 0.177–1.019 0.526–2.748 1.321–3.403 0.490–2.577 0.724–5.102 0.680–1.534 1.339–6.602
0.799 0.847 0.055 0.661 0.002 0.782 0.190 0.099 0.007
Abbreviations: CI, confidence interval; HR, hazard ratio.
Oncogene (2014), 1 – 11
© 2014 Macmillan Publishers Limited
miR-34a-5p predicts recurrence of CRC J Gao et al
5
Figure 2. Ectopic expression of miR-34a-5p inhibited the growth of HCT116 cells in vitro and in vivo. (a) Ectopic expression of miR-34a-5p in HCT116 and HCT116 p53 − / − cell lines was confirmed by quantitative PCR. (b) miR-34a-5p inhibited cell viability in HCT116 cells, but not in HCT116 p53− / − cells. (c) Colony formation assay confirmed the inhibitory effect of miR-34a-5p on HCT116 cells, but not on HCT116 p53− / − cells (n = 3, mean ± s.d.). (d) Tumour growth curve of xenografts derived from HCT116 and HCT116 p53− / − cells transfected with miR-34a-5p or miR-Ctrl. Data are mean ± s.d. (n = 5/group), arrows indicated the time of miRNA intratumoural injection.
expressions of p53 and its downstream proteins p21 and p27 were upregulated, with cyclin-dependent kinase (CDK4) and cyclin D1 downregulated (Figure 5c). However, the effect of miR-34a-5p on G1 cell cycle regulation was not observed in HCT116 p53− / − cells (Figure 5b), and the protein expression of cell cycle related regulators was not changed by miR-34a-5p in HCT116 p53− / − cells (Figure 5c).
miR-34a-5p induced apoptosis in HCT116 cells by activation of caspase-dependent pathway To determine whether the decrease in colon cell growth by miR-34a-5p was due to induction of apoptosis, the cellular apoptotic rate was appraised using annexin V-allophycocyanin (V-APC) and 7-amino-actinomycin (7-AAD) staining by flow cytometry. As shown in Figures 5d and e, compared with control, the proportion of early apoptotic cells in HCT116 cell line was significantly increased after transfection of miR-34a-5p (P o 0.05; Figure 5d), this was not found in HCT116 p53− / − cells (Figure 5e). Induction of apoptosis was further confirmed by the analysis of the expression of apoptosis-related proteins by western blot. As shown in Figure 5f, the protein level of the active forms of caspase-9, caspase-7, nuclear poly (ADP-ribose) polymerase (PARP), Bak and Bax was enhanced, but the anti-apoptosis gene survivin was inhibited in the miR-34a-5ptransfected HCT116 cells compared with the control cells, but not in HCT116 p53− / − cells. © 2014 Macmillan Publishers Limited
miR-34a-5p increased p53 luciferase reporter activity in HCT116 cells From the above results, we found that miR-34a-5p had a strong tumour suppressive ability for colon cancer in the presence of p53, and the inhibitory effect of miR-34a-5p abolished when p53 was absent. To further elucidate the interaction of miR-34a-5p with p53, p53 luciferase reporter was transfected together with miR-34a-5p or miR-Ctrl in HCT116 and HCT116 p53− / − cells. As expected, p53 luciferase activity was significantly increased after transfection of miR-34a-5p in HCT116 cells (P o 0.001), but not in HCT116 p53− / − cells (Figure 6a). However, a direct interplay between miR-34a-5p and p53 was not demonstrated due to the reported evidence that p53 directly regulates miR-34a-5p and miR-34a-5p indirectly regulates p53.
miR-34a-5p expression was positively associated with p53 expression in primary CRC tumour tissues Among 205 CRC cases in cohort I, 151 patients had p53-positive expression and 54 patients had p53-negative expression as detected by immunohistochemistry (Figure 6b). The expression level of miR-34a-5p in patients with p53-positive expression was higher than that in patients with p53-negative expression (P o 0.01; Figure 6c). Moreover, miR-34a-5p could predict recurrence better in patients with p53-positive expression (P o 0.001) than in patients with p53-negative expression (P o 0.05; Figure 6d). Oncogene (2014), 1 – 11
miR-34a-5p predicts recurrence of CRC J Gao et al
6
Figure 3. Ectopic expression of miR-34a-5p inhibited the migration and invasion of HCT116 cells in vitro and in vivo. (a) Ectopic expression of miR-34a-5p significantly inhibited cell migration in HCT116 cells by wound healing assay. The degree of wound sealing was calculated according to the distance of two flanks of the wound (n = 3, mean ± s.d.). (b) Compared with control, miR-34a-5p attenuated cell invasion ability in HCT116 cells by counting cell numbers passed through the Matrigel membrane (n = 3, mean ± s.d.). (c) miR-34a-5p suppressed tumour metastasis in vivo. (c1) HCT116 cell expressed miR-34a-5p and control plasmid respectably. (c2) miR-34a-5p expression level was confirmed by quantitative PCR. (c3) Representative images of the lung tissues from nude mice, black arrow indicated the macroscopic metastatic foci. (c4) Representative hematoxylin and eosin-stained lung sections containing metastatic foci. (d) Several important tumorgenic genes were detected by western blot analysis.
DISCUSSION In this study, we demonstrated that miR-34a-5p was downregulated in primary CRC tumour tissues. Promoter hypermethylation of miR-34a-5p was reported to be one of the mechanisms resulting in downregulation of miR-34a-5p expression.26 In keeping with our finding, downregulation of miR-34a-5p has been reported in other solid cancers, including gastric cancer, liver cancer and non-small cell lung cancer.15–18 However, the clinical impact of miR-34a-5p dysregulation in human cancer remains to be investigated. We investigated the association between miR-34a-5p expression and disease recurrence in CRC. Among our two cohorts of stage II and III primary CRC tumours, miR-34a-5p expression level in patients with recurrence was significantly lower than that in patients without recurrence (P o 0.001; Supplementary Figure S1B). Besides, miR-34a-5p low expression was related to poor disease-free survival and high 3-year recurrence rate by Oncogene (2014), 1 – 11
Kaplan–Meier curves and log-rank test (Figure 1c). These results imply that the miR-34a-5p downregulation may specifically predict the most aggressive and fatal types of CRC in cases with stage II/III disease. Indeed, multivariate Cox regression analysis indicated that miR-34a-5p was a potential predictor for the recurrence of patients with stage II/III CRC (Table 2). In supporting this, recent studies have demonstrated that several other miRNAs such as miR-21, miR-362-3p and miR-93 were associated with recurrence of CRC.27–29 Their results are complementary to ours and suggest that dysregulated miRNAs in tumours could predict recurrence of CRC. It was reported that miR-34a-5p was a direct transcript target of p53. Reciprocally, miR-34a enhances p53 level and activates p53 signalling pathway through the following mechanisms: (1) miR-34a-5p/MDM4/p53 loop. miR-34a-5p represses its direct target MDM4, leading to stabilization of p53 and sustainable expression of miR-34a-5p;30 (2) miR-34a-5p/SIRT1/p53 loop. miR-34a-5p inhibits its direct target SIRT1, leading to an increase © 2014 Macmillan Publishers Limited
miR-34a-5p predicts recurrence of CRC J Gao et al
7
Figure 4. KLF4 is a direct target of miR-34a-5p in colon cancer. (a) Human KLF4 3’UTR-binding site for miR-34a-5p. The miR-34a-5p wild-type binding sequence or its mutated form was inserted into C-terminal of the luciferase gene to generate pMIR-KLF4-3’UTR or pMIR-KLF4-mut3’UTR, respectively. (b) miR-34a-5p targeted the wild-type but not the mutant 3′UTR of KLF4. The data are means ± s.d. (c) Ectopic expression of miR-34a-5p downregulated KLF4 mRNA expression in HCT116 cells as determined by quantitative reverse transcription–PCR. (d) miR-34a-5p decreased KLF4 protein level in HCT116 detected by western blot.
in acetylated p53. Acetylated p53 at amino acid K382 abrogates p53 ubiquitination by MDM2 and therefore stabilizes p53 protein.19,31,32 These data collectively suggest that miR-34a-5p indirectly upregulates p53 through suppressing its direct targets MDM4 and SIRT1. Biological functions of miR-34a-5p were therefore analysed in HCT116 and HCT116 p53− / − cells in this study. Our results showed that ectopic expression of miR-34a-5p in HCT116 cells could inhibit cell proliferation, decrease colony formation, suppress cell mobility and invasion and inhibit tumorigenesis in nude mice. However, these effects were not observed in HCT116 p53− / − cells. Moreover, miR-34a-5p could increase p53 luciferase activity in HCT116 cells (P o0.001), but not in HCT116 p53− / − cells (Figure 6a). These results suggested that the tumour inhibitory effect of miR-34a-5p in CRC was dependent on p53. The positive correlation between miR-34a-5p and p53 expression in primary CRC patients further confirmed the importance of p53 in involving in the functional suppression of CRC by miR-34a-5p, although further validation is warranted in a large-scale samples with p53 mutant status. As shown in Figures 6c and d, miR-34a-5p level in patients with p53-positive expression was higher than those with p53-negative expression. miR-34-5p expression is correlated to disease-free survival in either p53-positive or -negative tumour patients. Cancer stem cells are a population of heterogeneous tumours that are crucial for tumour initiation and recurrence. Previous studies suggested that © 2014 Macmillan Publishers Limited
miR-34a-5p inhibited colon cancer stem cell self-renewal and promoted daughter cell differentiation by targeting NOTCH pathway.22 However, the cell lines we used in the current study are all differentiated cells and therefore miR-34a-5p showed a limited effect on the p53− / − cells in a short-term observation period. These may explain the discrepancy between in vitro assay and clinical outcome results. Several target genes regulated by miR-34a-5p were reported to be involved in the process of cell growth, migration and invasion in several malignancies, such as miR-34a-5p target SIRT1, c-Myc, NOTCH1, MMP1, KLF4 and BCL2.19–25 We demonstrated that miR-34a-5p inhibited protein expression of oncogenic genes SIRT1, c-Myc, NOTCH1 and KLF4 in colon cancer cells (Figure 3d). The inhibition effects are more prominent in p53 wild-type HCT116 compared with p53− / − HCT116 cells due to the miR-34a5p/MDM4/p53 loop and miR-34a-5p/SIRT1/p53 loop. MMP1 involved in cancer invasion and metastasis by degrading the extracellular matrix. MMP1 lacks potential miR-34a-5p-binding sequence; previous studies suggest that wild-type p53 represses MMP1 transcription through regulation of its promoter activity.33 Our data demonstrated that miR-34a suppressed MMP1 at both mRNA level and protein level in a p53-dependent manner, suggesting an indirectly regulation of MMP1 by miR-34a-5p. This mechanism could, at least, partially explain the p53-dependent miR-34a-5p repression of cell migration and invasion. Thus, the Oncogene (2014), 1 – 11
miR-34a-5p predicts recurrence of CRC J Gao et al
8
Figure 5. miR-34a-5p induced cell cycle arrest at G1 phase and early apoptosis in HCT116 cells. Ectopic expression of miR-34a-5p induced cell cycle arrest at G1 phase only in HCT116 cells (a), but not in HCT116 p53− / − cells (b; n = 3, mean ± s.d.). (c) The expressions of p53 and its downstream proteins p21 and p27 were upregulated, and CDK4/cyclin D1 were downregulated after transfection of miR-34a-5p in HCT116 cells, not in HCT116 p53− / − cells. (d) Early apoptosis of HCT116 cells was increased by miR-34a-5p. (e) No difference in cell apoptosis was found in HCT116 p53− / − cells transfected with miR-34a-5p or miR-Ctrl (n = 3, mean ± s.d.). (f) Effect of miR-34a-5p on expressions of several apoptosis-related proteins detected by western blot analysis.
tumour suppressive effect of miR-34a-5p at least in part mediated by the inhibition of these important tumorigenic factors in CRC as summarized in Figure 6e. To identify the possible mechanisms involved in the inhibitory effect of miR-34a-5p, cell cycle and cell apoptosis were conducted. In keeping with the tumour suppressive role, fluorescenceactivated cell sorting analysis of the effects of miR-34a-5p on the cell cycle in colon cancer cells revealed a concomitant increase of cells in G0/G1, which lead to the inhibition of cell proliferation. On the basis of the immunoblot analysis of negative cell cycle regulators, G0/G1 arrest by miR-34a-5p was most likely associated with the induction of p53, p21 and p27 (Figure 5c). It is well known that p53 can induce cell cycle arrest through transcriptional upregulation of the CDK inhibitor p21 and p27, a major player in G1 arrest.34,35 Cyclin D1–CDK4 complex accumulation is of great importance for G0/G1 cell cycle progression. In this regard, miR-34a-5p causes G1 cell cycle arrest in colon cancer that was at least by a mechanism involving upregulation of protein expression Oncogene (2014), 1 – 11
of p53, p21 and p27, and downregulation of cyclin D1 and CDK4, which was summarized in Figure 6e. In addition to inhibition of cell proliferation, the growthinhibitory effect of miR-34a-5p was also related to induction of apoptosis. We observed that induction of miR-34a-5p-mediated apoptosis occurs by the modulation of caspase apoptosis pathway. The protein expression of the downstream apoptosis executors caspase-9, caspase-7, caspase-3 and PARP was upregulated. Activation of caspase-9 processes other effector caspase members, including caspase-3 and caspase-7 to initiate a caspase cascade, which further initiates the proteolytic cleavage of the nuclear enzyme PARP and causes apoptosis. Pro-apoptotic Bax and Bak have been implicated in the regulation of p53-dependent apoptosis.36 Survivin, an inhibitor of apoptosis protein, is highly expressed in most cancers and associated with increased tumorigenesis. Taken together, as described in Figure 6e, the induction of apoptosis by miR-34a-5p in CRC is mediated by © 2014 Macmillan Publishers Limited
miR-34a-5p predicts recurrence of CRC J Gao et al
9
Figure 6. miR-34a-5p increased p53 luciferase reporter activity, and the proposed scheme for inhibitory effect of miR-34a-5p. (a) Luciferase activity assay demonstrated that miR-34a-5p increased p53 luciferase activity by 1.5-fold in HCT116 cells, not in HCT116 p53− / − cells (n = 3, mean ± s.d.). (b) Representative results of p53 staining by immunohistochemistry (magnification: × 200). (c) miR-34a expression in patients with p53-positive expression was higher than that in patients with p53-negative expression. (d) miR-34a-5p predicted recurrence better in patients with p53-positive expression (Po 0.001) than in patients with p53-negative expression (P o0.05). (e) Proposed scheme of molecular basis for inhibitory effect of miR-34a-5p in CRC according to our results.
induction of intrinsic apoptotic caspase cascade, pro-apoptosis Bax and Bak and inhibition of anti-apoptotic gene survivin. Our study is a retrospective study. Further prospective studies are warranted for validation of the prognostic power of the candidate miRNA in CRC. In conclusion, we found for the first time that miR-34a-5p downregulation could predict high risk of recurrence in CRC patients after radical surgery, which may be used as a potential recurrence biomarker for the recurrence of the stage II/III CRC patients. miR-34a-5p acts as a tumour suppressor in CRC in a p53dependent manner. MATERIALS AND METHODS Patients and sample collection This study included two cohorts of CRC patients receiving radical surgery in two departments of Peking University Cancer Hospital from 2002 to 2010 (cohort I: n = 205; cohort II: n = 63). All patients were pathologically confirmed as stage II and stage III CRC (the Union for International Cancer Control staging system) and had snap-frozen surgical tumour or nontumour tissues stored at − 80 °C before any therapy. The clinicopathological characteristics of all patients with disease recurrence (n = 133, local recurrence or distant metastasis) or without disease recurrence (n = 135) within 3 years after surgery are shown in Table 1. Patients with stage II and III CRC, and with clinicopathological characteristics and follow-up information available, were included. We excluded patients if they had any previous chemoradiotherapy treatment, metachronous or synchronous cancers. The clinical data of patients were retrospectively obtained from their medical records and the last follow-up was in January 2012. Patients who developed any local recurrence or distant metastasis diagnosed by computerized tomography scan within 3 years after surgery are defined as patients with recurrence. All patients had given written © 2014 Macmillan Publishers Limited
informed consent for their tissues to be used in the future research. This study was approved by the Ethics Committee of Peking University Cancer Hospital.
Colon cancer cell lines HCT116 was obtained from the American Type Culture Collection (Manassas, VA, USA), and HCT116 p53− / − cell line was kindly provided by Professor Bert Vogelstein (Johns Hopkins University). Cells were cultured in McCoy’s 5A modified medium (Invitrogen, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum (Gibco BRL, Invitrogen) and incubated in a humidified 37 °C incubator supplemented with 5% CO2.
RNA extraction and miRNA expression analyses Hematoxylin and eosin staining was performed for each surgical sample. Tumour tissues samples with the cancer cells 460% were eligible for RNA isolation and real-time PCR analysis. Total RNA was freshly extracted from tissues or cell pellets using Trizol reagent according to the manufacturer's instructions (Invitrogen). The RNA samples with OD260/OD280 ratio ranged between 1.9 and 2.0 were considered as good quality. The reverse transcription and quantitative PCR for miR-34a-5p and endogenous control RNU6B were conducted by using TaqMan MicroRNA Assays (Applied Biosystems, Foster City, CA, USA). The relative expression of miR-34a-5p was calculated using the comparative Ct method.
Cell viability assay Cells were transfected with miR-34a-5p (MC11030; Life Technologies, Carlsbad, CA, USA) or negative control miRNA (miR-Ctrl, 4464058; Life Technologies) using lipofectamine 2000 (Invitrogen). After 48 h of transfection, cell viability was determined by the 3-(4,5-dimethylthiazol2-yl)-5-(3-carboxymethoxyphen-yl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS) assay (Promega, Madison, WI, USA) according to the manufacturer’s Oncogene (2014), 1 – 11
miR-34a-5p predicts recurrence of CRC J Gao et al
10 instructions. The absorbance was measured at 490 nm using a spectrophotometer once a day for consecutive 5 days. The MTS assay read was calculated relative to day 1.
Colony formation assay Cells were transfected with miR-34a-5p or miR-Ctrl using lipofectamine 2000 (Invitrogen). After 48 h of transfection, cells were subcultured in a sixwell plate (1000 cells/well) for about 10 days. Colonies were stained with 5% crystal violet solution and counted. All experiments were performed in triplicate wells.
In vivo metastasis assay miR-34a-5p expressing plasmid (PMIRH34a-5pPA-1) and control plasmid PMIRH000PA-1 were purchased from System Biosciences (San Francisco, CA, USA). Stable miR-34a-5p-expressing cell line and control cell line were created through lentivirus infection (Lentivector Expression Systems, System Biosciences). Female Balb/c nude mice (5 weeks old) were used (seven mice per group). Stable miR-34a-5p-expressing or control HCT116 cells (1 × 106) were injected intravenously through the tail vein into each nude mouse. All mice were killed 6 weeks after injection. The lungs from each mouse were excised and embedded in paraffin for hematoxylin and eosin staining. All animal experiments were performed in accordance with the animal experimental guidelines of Chinese University of Hong Kong.
Cell migration assay After cells were transfected with miR-34a-5p or miR-Ctrl for 48 h, cell migration was performed using wound healing assay. The degree of wound sealing was calculated according to the distance of two flanks of the wound once a day for 3 days.
Cell invasion assay Cells were transfected with miR-34a-5p or miR-Ctrl for 48 h, and cell invasion was performed by Matrigel invasion assay (BD Biosciences, Erembodegem, Belgium). Cells in the upper compartment of the chamber were suspended in serum-free medium, and the lower chamber contained medium supplemented with 20% fetal bovine serum. After 24 h incubation, cells that passed through the matrigel membrane were fixed and stained with crystal violet, followed by counting from five random microscopic fields.
Cell cycle assay After transfection with miR-34a-5p or miR-Ctrl for 48 h, cells were collected and fixed in 70% cold ethanol for at least 12 h at 4 °C. Cells were stained with 50 μg/ml propidium iodide (BD Biosciences) at room temperature for 30 min in dark, and cell cycle was performed by FACS Calibur system (BD Biosciences) and analysed by ModFit 3.0 software (Verity Software House, Topsham, ME, USA).
Annexin V apoptosis assay Cell apoptosis was conducted by staining with APC and 7-AAD (BD Biosciences) for 15 min at room temperature in dark, followed by flow cytometry within 1 h (BD Biosciences). Cell apoptosis was analysed by WinMDI 2.9 software (The Scripps Research Institute, La Jolla, CA, USA).
Western blot Total protein was extracted from cell pellets using CytoBuster Protein Extraction Reagent (Merck Millipore, Darmstadt, Germany). Protein concentration was measured by the DC protein assay method of Bradford (Bio-Rad, Hercules, CA, USA), and 10–20 mg of protein from each sample was separated on 12% SDS–polyacrylamide gel electrophoresis. After transfer, the nitrocellulose membrane (GE Healthcare, Piscataway, NJ, USA) was incubated with primary antibody at 4 °C overnight and secondary antibody at room temperature for 1 h (antibody list is shown in Supplementary Table S1). Proteins were visualized using ECL Plus Western Blotting Detection Reagents (GE Healthcare).
In vivo tumorigenicity HCT116 and HCT116 p53− / − cells (2 × 106) transfected with miR-34a-5p or miR-Ctrl were suspended in 0.1 ml phosphate buffered saline, and injected subcutaneously into the dorsal left flank of 5-week-old female Balb/c nude mice (five mice per group). miRNAs were prepared by pre-incubating 0.3 nmole miRNA with 2.5 μl lipofectamine 2000 (Invitrogen) for 15 min and injections were made in a final volume of 100 μl in McCoy's 5 A medium (Sigma-Aldrich, St Louis, MO, USA) per mouse. The injection was repeated every 3 days and consisted of three consecutive injections.37 Tumour was measured twice every week from the first injection and tumour volume was calculated by the formula V = L × W2 × 1/2 (V, volume; L, length; W, width of tumour). All animal experiments were performed in accordance with the animal experimental guidelines of Chinese University of Hong Kong. Oncogene (2014), 1 – 11
Vector construction and Luciferase reporter assay The potential miR-34a-5p-binding sites were predicted by miRanda (www. microRNA.org). Sequence of 56 bp segment with the wild-type or mutant seed region of KLF4 was synthesized and cloned into pMIR-REPORT luciferase vector (Applied Biosystems; Life Technologies). The mutant KLF4 3′UTR sequence was prepared by deleting nine nucleotides in the seed region. Cells (1 × 105/well) transiently transfected with miR-34a-5p or miRCtrl (at 20 nM final concentration) were seeded in 24-well plates. pMIRREPORT vector (195 ng/well) and pRL-TK vector (5 ng/well) were cotransfected using lipofectamine 2000 (Invitrogen). Cells were collected 48 h post transfection and luciferase activities were analysed by the dualluciferase reporter assay system (Promega). Cells in 24-well plates were co-transfected with p53 reporter plasmid (contains 13 copies of the wild-type p53 response elements) and miR-34a5p or miR-Ctrl using lipofectamine 2000 (Invitrogen). After 48 h, luciferase activity was analysed by dual-luciferase reporter assay system according to the manufacturer’s instructions (Promega).
Statistical analysis Statistical analysis was performed using SPSS 18.0 software (SPSS Inc., Chicago, IL, USA). The χ2-test was used to analyse the relationships between clinicopathological characteristics and miR-34a-5p expression. The Kaplan–Meier survival curve and log-rank test were used to describe disease-free survival to miR-34a-5p expression. Multivariate Cox regression model was employed to analyse predictive variety. The difference in miR-34a-5p expression between tumour and non-tumour tissues or patients with or without recurrence was compared by Mann–Whitney Utest. Repeated measures analysis of variance and χ2-test were used to compare the difference in cell or tumour growth and the numbers of metastatic foci between two groups of cells or mice. A cut-off value was selected using receiver-operating characteristic curves for reference, based on high sensitivity and specificity. Po0.05 was considered statistically significant.
CONFLICT OF INTEREST The authors declare no conflict of interest.
ACKNOWLEDGEMENTS This work was supported by National Natural Science Foundation of China (No. 81301853, 81072034), National High Technology Research and Development Program (No. 2012AA 02A 504, 2012AA 02A 506), Shenzhen Technology and Innovation Project Fund, Shenzhen (JSGG20130412171021059), China 863 program (2012AA02A506), Shenzhen Municipal Science and Technology R & D fund (JCYJ20120619152326450) and Shenzhen Virtual University Park Support Scheme to CUHK Shenzhen Research Institute.
REFERENCES 1 Jemal A, Bray F, Center MM, Ferlay J, Ward E, Forman D. Global cancer statistics. CA Cancer J Clin 2011; 61: 69–90. 2 Mirakhorli M, Rahman SA, Abdullah S, Vakili M, Rozafzon R, Khoshzaban A. Multidrug resistance protein 2 genetic polymorphism and colorectal cancer recurrence in patients receiving adjuvant FOLFOX-4 chemotherapy. Mol Med Rep 2013; 7: 613–617. 3 Baranwal S, Alahari SK. miRNA control of tumor cell invasion and metastasis. Int J Cancer 2010; 126: 1283–1290.
© 2014 Macmillan Publishers Limited
miR-34a-5p predicts recurrence of CRC J Gao et al
11 4 Wang Z, Li Y, Kong D, Ahmad A, Banerjee S, Sarkar FH. Cross-talk between miRNA and Notch signaling pathways in tumor development and progression. Cancer Lett 2010; 292: 141–148. 5 Ma Y, Li W, Wang H. Roles of miRNA in the initiation and development of colorectal carcinoma. Curr Pharm Des 2013; 19: 1253–1261. 6 Dong Y, Wu WKK, Wu CW, Sung JJY, Yu J, Ng SSM. MicroRNA dysregulation in colorectal cancer: a clinical perspective. Br J Cancer 2011; 104: 893–898. 7 Ng EKO, Chong WWS, Jin H, Lam EKY, Shin VY, Yu J et al. Differential expression of microRNAs in plasma of patients with colorectal cancer: a potential marker for colorectal cancer screening. Gut 2009; 58: 1375–1381. 8 Wu CW, Ng SSM, Dong YJ, Ng SC, Leung WW, Lee CW et al. Detection of miR-92a and miR-21 in stool samples as potential screening biomarkers for colorectal cancer and polyps. Gut 2012; 61: 739–745. 9 Wu CW, Ng SC, Dong Y, Tian L, Ng SSM, Leung WW et al. Identification of microRNA-135b in Stool as a Potential Noninvasive Biomarker for Colorectal Cancer and Adenoma. Clin Cancer Res 2014; 20: 2994–3002. 10 Zhang X, Yan Z, Zhang J, Gong L, Li W, Cui J et al. Combination of hsa-miR-375 and hsa-miR-142-5p as a predictor for recurrence risk in gastric cancer patients following surgical resection. Ann Oncol 2011; 22: 2257–2266. 11 Yang M, Shen H, Qiu C, Ni Y, Wang L, Dong W et al. High expression of miR-21 and miR-155 predicts recurrence and unfavourable survival in non-small cell lung cancer. Eur J Cancer 2013; 49: 604–615. 12 Zhu HT, Dong QZ, Sheng YY, Wei JW, Wang G, Zhou HJ et al. MicroRNA-29a-5p is a novel predictor for early recurrence of hepatitis B virus-related hepatocellular carcinoma after surgical resection. PLoS One 2012; 7: e52393. 13 Hermeking H. p53 enters the microRNA world. Cancer Cell 2007; 12: 414–418. 14 Yamakuchi M, Lowenstein CJ. MiR-34, SIRT1 and p53: the feedback loop. Cell Cycle 2009; 8: 712–715. 15 Wu J, Wu G, Lv L, Ren Y-F, Zhang X-J, Xue Y-F et al. MicroRNA-34a inhibits migration and invasion of colon cancer cells via targeting to Fra-1. Carcinogenesis 2012; 33: 519–528. 16 Ji Q, Hao X, Meng Y, Zhang M, Desano J, Fan D et al. Restoration of tumor suppressor miR-34 inhibits human p53-mutant gastric cancer tumorspheres. BMC Cancer 2008; 8: 266. 17 Li N, Fu H, Tie Y, Hu Z, Kong W, Wu Y et al. miR-34a inhibits migration and invasion by down-regulation of c-Met expression in human hepatocellular carcinoma cells. Cancer Lett 2009; 275: 44–53. 18 Gallardo E, Navarro A, Viñolas N, Marrades RM, Diaz T, Gel B et al. miR-34a as a prognostic marker of relapse in surgically resected non-small-cell lung cancer. Carcinogenesis 2009; 30: 1903–1909. 19 Yamakuchi M, Ferlito M, Lowenstein CJ. miR-34a repression of SIRT1 regulates apoptosis. Proc Natl Acad Sci USA 2008; 105: 13421–13426. 20 Li L, Yuan L, Luo J, Gao J, Guo J, Xie X. MiR-34a inhibits proliferation and migration of breast cancer through down-regulation of Bcl-2 and SIRT1. Clin Exp Med 2013; 13: 109–117.
21 Yamamura S, Saini S, Majid S, Hirata H, Ueno K, Deng G et al. MicroRNA-34a modulates c-Myc transcriptional complexes to suppress malignancy in human prostate cancer cells. PLoS One 2012; 7: e29722. 22 Bu P, Chen KY, Chen JH, Wang L, Walters J, Shin YJ et al. A microRNA miR-34a-regulated bimodal switch targets Notch in colon cancer stem cells. Cell Stem Cell 2013; 12: 602–615. 23 Chang TC, Wentzel EA, Kent OA, Ramachandran K, Mullendore M, Lee KH et al. Transactivation of miR-34a by p53 broadly influences gene expression and promotes apoptosis. Mol Cell 2007; 26: 745–752. 24 Leng Z, Tao K, Xia Q, Tan J, Yue Z, Chen J et al. Krüppel-like factor 4 acts as an oncogene in colon cancer stem cell-enriched spheroid cells. PLoS One 2013; 8: e56082. 25 Chen Q, Li L, Tu Y, Zheng LL, Liu W, Zuo XY et al. MiR-34a regulates apoptosis in liver cells by targeting the KLF4 gene. Cell Mol Biol Lett 2014; 19: 52–64. 26 Nair VS, Maeda LS, Ioannidis JPA. Clinical outcome prediction by microRNAs in human cancer: a systematic review. J Natl Cancer Inst 2012; 104: 528–540. 27 Christensen LL, Tobiasen H, Holm A, Schepeler T, Ostenfeld MS, Thorsen K et al. MiRNA-362-3p induces cell cycle arrest through targeting of E2F1, USF2 and PTPN1 and is associated with recurrence of colorectal cancer. Int J Cancer 2013; 133: 67–78. 28 Yang IP, Tsai HL, Hou MF, Chen KC, Tsai PC, Huang SW et al. MicroRNA-93 inhibits tumor growth and early relapse of human colorectal cancer by affecting genes involved in the cell cycle. Carcinogenesis 2012; 33: 1522–1530. 29 Kjaer-Frifeldt S, Hansen TF, Nielsen BS, Joergensen S, Lindebjerg J, Soerensen FB et al. The prognostic importance of miR-21 in stage II colon cancer: a populationbased study. Br J Cancer 2012; 107: 1169–1174. 30 Mandke P, Wyatt N, Fraser J, Bates B, Berberich SJ, Markey MP. MicroRNA-34a modulates MDM4 expression via a target site in the open reading frame. PLoS One 2012; 7: e42034. 31 Li L, Wang L, Li L, Wang Z, Ho Y, McDonald T et al. Activation of p53 by SIRT1 inhibition enhances elimination of CML leukemia stem cells in combination with imatinib. Cancer Cell 2012; 21: 266–281. 32 Li M, Luo J, Brooks CL, Gu W. Acetylation of p53 inhibits its ubiquitination by Mdm2. J Biol Chem 2002; 277: 50607–50611. 33 Sun Y, Wenger L, Rutter JL, Brinckerhoff CE, Cheung HS. p53 down-regulates human matrix metalloproteinase-1 (collagenase-1) gene expression. J Biol Chem 1999; 274: 11535–11540. 34 Tokino T, Nakamura Y. The role of p53-target genes in human cancer. Crit Rev Oncol Hematol 2000; 33: 1–6. 35 Mitrea DM, Yoon M-K, Ou L, Kriwacki RW. Disorder-function relationships for the cell cycle regulatory proteins p21 and p27. Biol Chem 2012; 393: 259–274. 36 Tobiume K. Involvement of Bcl-2 family proteins in p53-induced apoptosis. J Nippon Med Sch 2005; 72: 192–193. 37 Dong Y, Zhao J, Wu CW, Zhang L, Liu X, Kang W et al. Tumor suppressor functions of miR-133a in colorectal cancer. Mol Cancer Res 2013; 11: 1051–1060.
Supplementary Information accompanies this paper on the Oncogene website (http://www.nature.com/onc)
© 2014 Macmillan Publishers Limited
Oncogene (2014), 1 – 11
