European Journal of Cancer (2014) xxx, xxx– xxx
Available at www.sciencedirect.com
ScienceDirect journal homepage: www.ejcancer.com
Clinical and biological significance of miR-378a-3p and miR-378a-5p in colorectal cancer Hui Li 1, Sujuan Dai 1, Tiantian Zhen 1, Huijuan Shi, Fenfen Zhang, Yang Yang, Lili Kang, Yingjie Liang, Anjia Han ⇑ Department of Pathology, The First Affiliated Hospital, Sun Yat-Sen University, Guangzhou, China
KEYWORDS MiR-378a-3p MiR-378a-5p Insulin-like growth factor 1 receptor Colorectal cancer
Abstract To investigate miR-378a-3p and miR-378a-5p expression and their relationships with the clinicopathological features of colorectal cancer (CRC). Our results showed that miR-378a-3p and miR-378a-5p expression were dramatically lower in CRC cell lines and tissues than that in adjacent normal colorectal mucosal tissues, respectively. MiR-378a-3p and miR-378a-5p expression were significantly associated with histological differentiation and TNM stage, respectively. CRC patients with low miR-378a-3p and miR-378a-5p expression had a significantly shorter survival time than those patients with high miR-378a-3p and miR-378a-5p expression (p < 0.001, p < 0.001), respectively. Univariate and multivariable Cox regression analysis showed that tumour size, TNM stage, miR-378a-3p expression and miR-378a-5p expression were independent prognostic factors for CRC patients. Ectopic miR-378a-3p or miR-378a-5p expression inhibited cellular proliferation and colony formation, induced apoptosis and G1-phase cell cycle arrest in CRC cells, but had no effect on migration and invasion of CRC cells. Furthermore, miR-378a-3p over-expression or downregulation could inhibit or enhance insulin-like growth factor 1 receptor (IGF1R) expression in CRC cells. There was a significantly negative correlation between IGF1R protein expression and miR-378a-3p expression in CRC tissues. MiR-378a-3p over-expression or down-regulation suppressed or enhanced phosphorylated-ERK1/2 protein level, but had no effect on phosphorylated-Akt protein level. In conclusion, miR-378a-3p and miR-378a-5p expression might play an important role as tumour suppressor gene in the initial stage of carcinogenesis of CRC. Ó 2013 Elsevier Ltd. All rights reserved.
⇑ Corresponding author: Address: Department of Pathology, The First Affiliated Hospital, Sun Yat-Sen University, 58, Zhongshan Road II, Guangzhou 510080, China. Tel./fax: +86 20 87332235. E-mail address: [email protected] (A. Han). 1 Hui Li, Sujuan Dai and Tiantian Zhen contributed equally to this work.
0959-8049/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ejca.2013.12.010
Please cite this article in press as: Li H. et al., Clinical and biological significance of miR-378a-3p and miR-378a-5p in colorectal cancer, Eur J Cancer (2014), http://dx.doi.org/10.1016/j.ejca.2013.12.010
2
H. Li et al. / European Journal of Cancer xxx (2014) xxx–xxx
1. Introduction Colorectal cancer (CRC) is one of the most common malignancies worldwide. The prognosis of patients with CRC remains poor despite the improvement of current treatment modalities including surgical resection, radiotherapy and chemotherapy. The 5-year overall survival rate for CRC patients is just 50–60% mainly because of tumour recurrence and/or metastasis [1]. Thus, it is imperative to find novel biomarkers and develop novel treatment strategies for CRC. MicroRNAs (miRNAs) which are composed of 21–25nucleotides are small, endogenous non-coding RNAs and regulate the expression of a variety of genes based on sequence complementarities with their target mRNA molecules. It has been proposed that miRNAs imperfectly complementary to the binding sites in the 30 untranslated regions (UTR) of their target mRNA repressprotein expression through translational inhibition, whereas those miRNAs perfectly complementary to mRNA targets induce mRNA degradation through the RNA-mediated interference pathway [2]. miRNAs are involved in diverse cellular biological functions, including proliferation, development, differentiation, apoptosis, migration, invasion and angiogenesis. An increasing number of evidence shows that miRNAs contribute to development and progression of a variety of human cancers [3,4]. The transcript of miR-378 via post-transcriptional processing generates two mature miRNAs, namely, miR-378a-3p and miR-378a-5p. An increasing number of evidence shows that miR-378 induces malignant transformation of human breast cancer cell [5], increases cell survival, colony formation, contributes to multiple drug resistance in human malignant glioma cells [6], and promotes migration, invasion and tumour angiogenesis of human lung cancer cell [7]. MiR-378 up-regulation might have an adverse impact on prognosis in acute myeloid leukaemia [8]. Circulation of miR-378 in serum is a potential biomarker for renal cell carcinoma and early detection of gastric cancer [9,10]. However, microarray-based miRNA profiling study reveals that miR-378 is down-regulated in oral squamous carcinoma, basal cell carcinoma of the skin, gastric cancer and CRC [11–14]. Whether miR-378 is an oncogenic or tumour suppressor gene in CRC remains unknown. Moreover, whether miR-378a-3p or miR-378a-5p plays an important role in development and progression of CRC needs clarify. Herein, we investigated the clinical and biological significance of miR-378a3p and miR-378a-5p in human CRC.
mucosal tissues were obtained from our Hospital between January 2004 and December 2005. Twelve pairs of fresh CRC and adjacent normal colorectal mucosal tissues were collected from May 2012 through October 2012. The fresh CRC specimens and paired adjacent normal colorectal tissues from surgery were immediately stored at 4 °C for 24 h in RNA later solution (Sigma, St. Louis, MO), then at –80 °C until RNA extraction. No patients had received chemotherapy and/or radiotherapy before surgery. The histopathology of the disease was determined by two pathologists according to the criteria of the World Health Organisation (2010). Clinical staging was done according to the Union for International Cancer Control classification. For the research purposes of these clinical materials, prior patient’s consents and approval from the Institutional Research Ethics Committee were obtained. Follow-up information was available for all patients. Detailed clinical information about these patients including sex, age, tumour location, histological differentiation, tumour size, TNM stage, T classification, N classification and distant metastasis status, is summarised in Table 1. 2.2. Cell lines and transfection Human CRC cell lines LOVO and SW1116 were maintained in RPMI-1640 medium (Invitrogen, Carlsbad, CA). The CRC cell line HT29 was cultured in Dulbecco’s modified Eagle’s medium (Invitrogen). The CRC cell lines SW480 and SW620 were grown in Leibovitz’s L-15 Medium (Invitrogen, Carlsbad, CA). The CRC cell line HCT116 was maintained in McCoy’s 5A Medium (Invitrogen). All medium was supplemented with 10% (v/v) foetal bovine serum (GIBCO, Carlsbad, CA), 1 antibiotic/antimycotic (100 units/mL streptomycin, 100 units/mL penicillin and 0.25 mg/mL amphotericin B). All cell lines were cultured in a humidified incubator at 37 °C with 5% CO2. 2 105 cells were plated in 6-well plates and transfected with 100 nmol/L miRNAs using LipofectamineÒ RNAiMAX Transfection Reagent (Invitrogen) according to the manufacturer’s protocol. The miR-378a-3p mimic, miR-378a-5p mimic, non-specific control mimic, miR-378a-3p inhibitor, miR-378a-5p inhibitor and non-specific control inhibitor were purchased from Life Technologies (Carlsbad, CA). 2.3. Reliability analysis of miRNA in CRC FFPE tissues by quantitative real time polymerase chain reaction (PCR)
2. Materials and methods 2.1. Clinical samples and patient information Ninety-six pairs of formalin-fixed paraffin-embedded (FFPE) archived CRC and adjacent normal colorectal
Total RNA extracted from 12 pairs of CRC FFPE tissues and frozen tissues using Trizol reagent (Invitrogen) according to the manufacturer’s instructions was used to determine whether miR-378a-3p could be detected reliably in archival specimens by real-time
Please cite this article in press as: Li H. et al., Clinical and biological significance of miR-378a-3p and miR-378a-5p in colorectal cancer, Eur J Cancer (2014), http://dx.doi.org/10.1016/j.ejca.2013.12.010
H. Li et al. / European Journal of Cancer xxx (2014) xxx–xxx
3
Table 1 MiR-378a-3p and miR-378a-5p expression levels and their relationship with clinicopathological features of colorectal cancer. Variable
No. (%)
Expression level of miR-378a-3p
P value
Expression level of miR-378a-5p
P value
Sex Male Female
53(55.21) 43(44.79)
–1.44 ± 1.52 –1.45 ± 1.27
0.965
–1.22 ± 1.49 –1.45 ± 1.49
0.447
Age (years) 663 >63
53(55.21) 43(44.79)
–1.54 ± 1.28 –1.34 ± 1.55
0.496
–1.52 ± 1.44 –1.08 ± 1.52
0.154
Location Colon Rectum
50(52.08) 46(47.92)
–1.49 ± 1.41 –1.40 ± 1.41
0.748
–1.23 ± 1.51 –1.43 ± 1.47
0.520
Histological differentiation Well Moderately & poorly
20(20.83) 76(79.17)
–0.76 ± 1.36 –1.63 ± 1.36
0.013
–0.62 ± 1.40 –1.51 ± 1.46
0.016
TNM stage I–II III–IV
47(48.96) 49(51.04)
–1.14 ± 1.58 –1.74 ± 1.15
0.033
–0.98 ± 1.58 –1.66 ± 1.33
0.025
T classification T1–T2 T3–T4
18(18.75) 78(81.25)
–0.84 ± 1.24 –1.59 ± 1.41
0.032
–0.82 ± 1.35 –1.44 ± 1.50
0.094
N classification N0 N1–N2
45(46.88) 51(53.12)
–1.21 ± 1.61 –1.66 ± 1.17
0.121
–1.06 ± 1.53 –1.56 ± 1.42
0.103
M classification M0 M1
91(94.79) 5(5.21)
–1.48 ± 1.39 –0.85 ± 1.64
0.329
–1.28 ± 1.39 –2.19 ± 2.87
0.182
PCR. Relative to U6 snRNA, the miR-378a-3p expression level was determined using 2–DCT, where DCT = CT (miR-378a-3p in FFPE tissues)–CTU6 or DCT = CT (miR-378a-3p in Frozen tissues)–CTU6. 2–DCT value was not fitted to normality distribution and was subsequently transformed at log 2. Then linear regression was used to investigate the correlation between miRNA expression in CRC FFPE tissues and frozen tissues.
and adjacent normal colon tissues. The quantity of miR-378a-3p and miR-378a-5p in each CRC tissue relative to its paired adjacent normal colon tissue was calculated using the equation [RQ = 2–DDCT, DDCT = (CTmiRNA-CTU6)T – (CTmiRNA-CTU6)N]. The expression level of miR-378a-3p or miR-378a-5p was classified into low expression group and high expression group compared with the mean expression level of miR-378a-3p or miR-378a-5p, respectively.
2.4. Expression of miR-378a-3p and miR-378a-5p by quantitative real time PCR
2.5. Western blot analysis
Reverse transcription was performed using One Step PrimeScript miRNA cDNA Synthesis Kit (TaKaRa Biotechnology Co., Ltd., Dalian, China), and quantitative real time PCR was performed using SYBR Premix Ex Taq II (TaKaRa Biotechnology). RNU6B snRNA was used for sample loading normalisation. The specific forward primer of miR-378a-3p was 50 -CTGGACTTGGAGTCAGAAGGAA-30 (TaKaRa Biotechnology). The specific forward primer of RNU6B was 50 -ACGCAAATTCGTGAAGCGTT-30 . The specific forward primer of miR-378a-5p was 50 -CTGACTCCAGG TCCTGTGTA-30 . Reverse primer for miR-378a-3p, miR-378a-5p and U6B snRNA was Uni-miR qPCR primer (TaKaRa code: D350A). Each sample was analysed in triplicate. Relative expression levels of miR-378a-3p and miR-378a-5p were calculated using 2–DCT in CRC
As described previously [15], 30 lg of total proteins were loaded onto 10% sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS–PAGE) and transferred onto polyvinylidene difluoride (PVDF) membranes (Millipore, Billerico, MA) that were subsequently blocked in 5% non-fat milk in TBST (20 mmol/L Tris, pH 7.6, 137 mmol/L NaCl, 0.1% Tween-20). The membranes were incubated with primary antibodies including rabbit anti-insulin-like growth factor 1 receptor (IGF1R), rabbit anti-phospho-AKT (Ser473), rabbit anti-p44/42MAPK (ERK1/2), rabbit anti-phospho-p44/ 42 MAPK (p-ERK1/2) (Thr202/Tyr204) and rabbit anti-glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (Cell Signalling Technology, Danvers, MA) at 4 °C overnight. After washing, the membranes were incubated with secondary antibody HRP-conjugated goat
Please cite this article in press as: Li H. et al., Clinical and biological significance of miR-378a-3p and miR-378a-5p in colorectal cancer, Eur J Cancer (2014), http://dx.doi.org/10.1016/j.ejca.2013.12.010
4
H. Li et al. / European Journal of Cancer xxx (2014) xxx–xxx
anti-rabbit (Cell Signalling Technology) and visualised by enhanced chemiluminescence detection kit (Millipore). 2.6. Immunohistochemistry staining and evaluation As described previously [16], immunohistochemistry staining for IGF1R was performed on CRC FFPE tissue sections. The sections were deparaffinised, rehydrated in serially graded ethanol and heated in citric buffer (pH 6.0) once for 3 min in a microwave oven for antigen retrieval. They were then washed with distilled water, blocked with 3% hydrogen peroxide and incubated with rabbit anti-IGF1R (Cell Signalling Technology) monoclonal antibody, at dilution of 1:100 at 4 °C for 12 h. The tissues were then stained using EnVision-HRP secondary antibody (Dako, Carpinteria, CA) for 30 min at room temperature, washed with a 0.01 mol/L concentration of phosphate buffered saline (PBS), stained with 0.5% diaminobenzidine and counterstained with Mayer’s haematoxylin, then air dried, and mounted with glycerol gelatin. The degree of IGF1R immunostaining was based on both the proportion of positively stained tumour cells and intensity of staining. The proportion of tumour cells was scored as follows: 0 (25–50% positive tumour cells), 3 (>50–75% positive tumour cells) and 4 (>75% positive tumour cells). Meanwhile, staining intensity was scored as follows: 0 (no staining); 1 (weak staining = light yellow), 2 (moderate staining = yellow brown) and 3 (strong staining = brown). Staining index was calculated as the proportion score staining intensity score. In each case, 10 high-power fields of representative areas were counted. The staining was independently assessed by two researchers who were unaware of the specific proteins being assessed or their putative role in CRC. 2.7. Cell proliferation assay Cells were plated in 96-well plates at 2 103 per well and transfected with miR-378a-3p or miR-378a-5p mimics or inhibitors. Non-specific control miRNA mimic or inhibitor was used as the control group. After 24, 48, 72, 96, 120 and 144 h transfection, cell proliferation was determined by 3-(4,5-dimethyl thiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) assay (CellTiter 96 NonRadioactive Cell Proliferation Assay Kit, Promega Corporation, Madison, WI). The experiment was performed in triplicate. 2.8. Adhesion-dependent colony formation assay After 24 h transfection with miR-378a-3p or miR378a-5p mimics, the cells were plated in 6-well plates 2 102 per well and grown for 2 weeks. Then, the cells
were washed twice with PBS, fixed with 4% paraformaldehyde and stained with 0.5% crystal violet for 15 min. The number of colonies in 10 random view fields was counted under a microscope and the average number of colonies was achieved. The experiment was triplicated independently. 2.9. Analysis of cell cycle and apoptosis 6 105 cells were seeded in 6-well plates and incubated overnight till 50–60% confluence. The cells were transfected with 100 nM miR-378a-3p or miR-378a-5p mimic or miR-378a-3p or miR-378a-5p inhibitor and harvested at 48 h, washed in cold PBS, fixed with 80% ethanol for 8 h at 4 °C, then stained with propidium iodide buffer (50 mg/ml propidium iodide, 0.1% sodium citrate and 0.1% Triton X-100) for 3 h at 4 °C. Non-specific control miRNA mimic or inhibitor was used as control group. 2 104 cells were analysed for cell cycle and apoptosis using a Becton Dickinson FACScan (Becton Dickinson Immunocytometry Systems, San Jose, CA). The percentage of cells in each phase of the cell cycle and apoptotic cells was quantified using Cell Quest software, respectively. This experiment was performed in triplicate. 2.10. Scratch wound migration assay 2 105 LOVO cells were seeded in 6-well plates and grown to 60% confluency in complete medium. The cells were transfected with 100 nM miR-378a-3p mimic or miR-378a-5p mimic, non-specific control miRNA mimic was used as the control group. After 24 h transfection, vertical scratches were then made using a 200 ll plastic filter tip to create a ‘wound’ of approximately 200 lm in diameter. To eliminate dislodged cells, culture medium was removed and wells were washed with PBS. The average distance of migrating cells was determined under an inverted microscope at 0, 12, 24 and 48 h. The experiment was performed in triplicate. 2.11. Transwell invasion assay Transwell invasion assay was carried out using 24-well MILLIcell Hanging Cell Culture Inserts 8 lm membrane coated with matrigel in the upper chamber (MILLIPORE). After 24 h transfection, cells were seeded into the upper chamber at a density of 3 104/ chamber well and incubated at 37 °C for 24 h, followed by removal of cells inside the upper chamber with cotton swabs. Invasive cells on the lower membrane surface were fixed with 95% ethanol for 30 min and stained with 0.2% Crystal Violet solution for 20 min. Five random fields from each invasion assay were counted under an inverted microscope. Three independent experiments were performed and the data were presented as mean ± standard deviation (SD).
Please cite this article in press as: Li H. et al., Clinical and biological significance of miR-378a-3p and miR-378a-5p in colorectal cancer, Eur J Cancer (2014), http://dx.doi.org/10.1016/j.ejca.2013.12.010
H. Li et al. / European Journal of Cancer xxx (2014) xxx–xxx
2.12. Statistical analysis All statistical analyses were performed using SPSS 16.0 statistics software. Student’s t test was used to compare the levels of cellular proliferation, apoptosis, migration, invasion and colony formation between different groups. Chi-square test was used to compare the levels of miR-378a-3p or miR-378a-5p expression and various clinicopathological parameters of CRC patients. The correlation of miR-378a-3p or miR-378a-5p expression level in CRC FFPE tissues and frozen tissues was analysed by Spearman’s correlation coefficients and linear regression analysis. Overall survival time of CRC patients was analysed by Kaplan–Meier analysis. Independent prognostic factor for CRC patients was analysed using Cox regression analysis. The correlation between miR-378a-3p expression and IGF1R expression in CRC was analysed by Spearman’s correlation coefficients. p < 0.05 was set to be statistically significant. 3. Results 3.1. Reliability of miRNA expression detection in CRC FFPE tissues by quantitative real time PCR To determine the reliability of miRNA expression detection in CRC FFPE tissues by quantitative real time PCR, we detected miR-378a-3p expression in 12 pairs of CRC FFPE tissues and frozen tissues by quantitative real time PCR. As shown in Fig. 1A, there was a significantly positive correlation of miR-378a-3p expression between CRC FFPE tissues and frozen tissues (R2 = 0.526, p = 0.008). The mean value of –DCTmiR-378a-3p in CRC FFPE tissues was –6.52. The mean value of –DCTmiR-378a-3p in their respective CRC frozen tissues was –6.17. The result indicated that it was reliable to detect miR-378a-3p expression level in CRC FFPE tissues by quantitative real time PCR. 3.2. MiR-378a-3p and miR-378a-5p expression in CRC cell lines and tissues As shown in Fig. 1B, the expression level of miR-378a3p and miR-378a-5p was lower in six CRC cell lines including HT29, HCT116, SW480, SW1116, SW620 and LOVO than that in normal colorectal mucosal tissue, respectively. Consistently, the expression level of miR378a-3p and miR-378a-5p was lower in 12 samples of fresh CRC tissues than that in their respective adjacent normal colorectal mucosal tissues, respectively (Fig. 1C). Furthermore, our result showed that the expression level of miR-378a-3p (–5.64 ± 1.34) and miR-378a-5p (–9.35 ± 1.58) was significantly lower in 96 samples of CRC FFPE tissues than that (–4.15 ± 0.99; –8.02 ± 1.10) in their respective adjacent normal colorectal mucosal tissues, respectively
5
(p < 0.001 and p < 0.001) (Fig. 1D). Moreover, there was a significantly positive correlation between miR378a-3p and miR-378a-5p expression in 96 samples of adjacent normal colorectal mucosal tissues (R2 = 0.572, p < 0.001) and 96 samples of CRC tissues (R2 = 0.770, p < 0.001), respectively. After Log2 transformation of miR-378a-3p and miR-378a-5p expression level, a significantly positive correlation was also found in 96 samples of CRC tissues (R2 = 0.689, P < 0.001) (Fig. 1E). 3.3. Relationship between miR-378a-3p and miR-378a-5p expression and clinicopathological features of CRC The mean expression level of miR-378a-3p in 96 samples of CRC was –1.45 ± 1.40 (median, –1.56; range, –4.68–3.11). The mean expression level of miR-378a5p in 96 samples of CRC was –1.32 ± 1.49 (median, –1.45; range, –6.88–4.22). Then, CRC patients were divided into two groups according to the mean expression level of miR-378a-3p and miR-378a-5p, respectively. Forty-five cases (46.9%) were low miR-378a-3p expression. Fifty-one cases (53.1%) were high miR378a-3p expression. Forty-one (42.7%) cases were low miR-378a-5p expression and 55 (57.3%) cases were high miR-378a-5p expression. The relationships between miR-378a-3p and miR-378a-5p expression and clinicopathological features of CRC were summarised in Table 1. The expression of miR-378a-3p was significantly associated with histological differentiation (p = 0.013), T classification (p = 0.032) and TNM stage (p = 0.033) of CRC. The expression of miR-378a-5p was significantly associated with histological differentiation (p = 0.016) and TNM stage (p = 0.025) of CRC. Moreover, miR-378a-3p and miR-378a-5p expression was lower in CRC with moderately and poorly-differentiation, high TNM stage (III–IV) and large tumour size (T3–T4) than that in CRC with well-differentiation, low TNM stage (I–II) and small tumour size (T1–T2), respectively (Fig. 2A–C). We also found that the greatest diameter of CRC was significantly larger (6.09 ± 3.17 cm) in low miR-378a-3p expression group than that (4.80 ± 1.88 cm) in high miR-378a-3p expression group (p = 0.019). Likewise, the greatest diameter of CRC was significantly larger (6.29 ± 2.80 cm) in low miR-378a-5p expression group than that (4.88 ± 1.86 cm) in high miR-378a-5p expression group (p = 0.007) (Fig. 2D). However, the expression level of miR-378a-3p and miR-378a-5p was not significantly associated with sex, age, tumour location, N classification and distant metastasis status, respectively (Table 1). 3.4. The relationship between miR-378a-3p and miR378a-5p expression and prognosis of CRC patients Kaplan–Meier survival analyses showed that CRC patients with low miR-378a-3p and miR-378a-5p
Please cite this article in press as: Li H. et al., Clinical and biological significance of miR-378a-3p and miR-378a-5p in colorectal cancer, Eur J Cancer (2014), http://dx.doi.org/10.1016/j.ejca.2013.12.010
6
H. Li et al. / European Journal of Cancer xxx (2014) xxx–xxx
Fig. 1. MiR-378a-3p and miR-378a-5p expression in colorectal cancer (CRC) cell lines and tissues. (A) Comparison of miR-378a-3p expression pattern from matched samples of CRC formalin-fixed paraffin-embedded (FFPE) and frozen tissues. (B) MiR-378a-3p (left) and miR-378a-5p (right) expression in CRC cell lines compared with adjacent normal colon mucosa epithelial cells (Control) by quantitative real time polymerase chain reaction (PCR) analysis. (C) MiR-378a-3p (left) and miR-378a-5p (right) expression in twelve pairs of fresh CRC tissues (T) and matched adjacent normal colorectal mucosa tissues (N). (D) MiR-378a-3p (left) and miR-378a-5p (right) in 96 pairs of CRC FFPE tissues (T) and matched adjacent normal colorectal mucosal tissues (N). The term DCT (DCT = CTmiR-378a-CTU6) was used to describe the expression level of miR-378a3p and miR-378a-5p. Small nuclear RNA U6 served as internal normalised references for miR-378a-3p and miR-378a-5p. X ± SD are represented in the images. (E) The correlation between miR-378a-3p expression and miR-378a-5p expression in 96 samples of adjacent normal colorectal mucosal tissues (left) and CRC tissues (middle and right).
Please cite this article in press as: Li H. et al., Clinical and biological significance of miR-378a-3p and miR-378a-5p in colorectal cancer, Eur J Cancer (2014), http://dx.doi.org/10.1016/j.ejca.2013.12.010
H. Li et al. / European Journal of Cancer xxx (2014) xxx–xxx
7
Fig. 2. The relationship between miR-378a-3p and miR-378a-5p expression and clinicopathological characteristics of colorectal cancer (CRC). (A) The expression level of miR-378a-3p (left) and miR-378a-5p (right) in moderately to poorly differentiated CRC group was lower than that of well differentiated CRC group. (B) The expression level of miR-378a-3p in T3–T4 group was lower than that in T1–T2 group (left), while the expression level of miR-378a-5p was not significantly associated with T classification (right). (C) The expression level of miR-378a-3p (left) and miR-378a-5p (right) in III–IV stage was lower than that of I–II stage. (D) Low expression of miR-378a-3p (left) and miR-378a-5p (right) was significantly associated with large tumour size, respectively. (E) Kaplan–Meier survival curve and log-rank test for CRC patients with different expression levels of miR-378a-3p (left) and miR-378a-5p (right). Please cite this article in press as: Li H. et al., Clinical and biological significance of miR-378a-3p and miR-378a-5p in colorectal cancer, Eur J Cancer (2014), http://dx.doi.org/10.1016/j.ejca.2013.12.010
8
H. Li et al. / European Journal of Cancer xxx (2014) xxx–xxx
expression had a significantly shorter survival time than those patients with high miR-378a-3p and miR-378a-5p expression (p < 0.001, p < 0.001), respectively (Fig. 2E). The 3- and 5-year median survival rate for CRC patients was 56.9% (range, 43.4–70.4%) and 50.8% (range, 37.1–64.5%) in low miR-378a-3p expression group compared with 93.3% (range, 91.9–94.4%) and 86.7% (range, 76.7–96.7%) in high miR-378a-3p expression group, respectively. The 3 and 5-year median survival rate for CRC patients was 58.2% (range, 45.1–71.3%) and 52.5% (range, 39.2–65.8%) in low miR-378a-5p expression group compared with 95.1% (range, 94.1–96.0%) and 87.8% (range, 77.8–97.8%) in high miR-378a-5p expression group, respectively. Univariate Cox regression analysis showed that age (Hazard ratio, 0.976; 95% confidence interval (CI), 0.953–0.999; p = 0.044), tumour size (Hazard ratio, 2.336; 95% CI, 1.226–4.453; p = 0.01), histological differentiation (Hazard ratio, 1.563; 95% CI, 1.008–2.425; p = 0.046), TNM stage (Hazard ratio, 1.795; 95% CI, 1.178–2.735; p = 0.006), miR-378a-3p expression (Hazard ratio, 0.205; 95% CI, 0.087–0.484; p < 0.001) , and miR-378a-5p expression (Hazard ratio, 0.154; 95% CI, 0.057–0.413; p < 0.001) were significantly associated with the prognosis of CRC patients (Table 2). Multivariable Cox regression analysis showed that tumour size, TNM stage, miR-378a-3p expression (Hazard ratio,
0.154; 95% CI, 0.059–0.404; p < 0.001) (Table 3) and miR-378a-5p expression (Hazard ratio, 0.147; 95% CI, 0.053–0.413; p < 0.001) (Table 4) were independent prognostic factors for CRC patients. 3.5. Ectopic miR-378a-3p or miR-378a-5p expression inhibited cellular proliferation and colony formation in CRC cells To investigate the effect of ectopic miR-378a-3p or miR-378a-5p expression on cellular proliferation of CRC, over-expression of miR-378a-3p or miR-378a-5p was established in LOVO and SW620 cell lines by miR-378a-3p mimic or miR-378a-5p mimic transfection. MiR-378a-3p or miR-378a-5p control mimic was used as the control group (Fig. 3A). Our result showed that transfection of miR-378a-3p mimic and miR-378a-5p mimic dramatically inhibited LOVO (p = 0.001 and p = 0.011) and SW620 (p = 0.001 and p = 0.001) cellular proliferation compared with the control group by MTT assay, respectively (Fig. 3B). Moreover, colony formation assay demonstrated that miR-378a-3p mimic and miR-378a-5p mimic significantly inhibited colony formation of LOVO (p = 0.001 and p = 0.042) and SW620 (p = 0.002 and p = 0.003) cell line by less in colony number and smaller in colony size compared with the control group, respectively (Fig. 3C). We also
Table 2 Univariate Cox regression analysis of the relationship between miR-378a-3p and miR-378a-5p expression level, clinicopathological features and survival of colorectal cancer patients. Variable
Subset
Hazard ratio (95% confidence interval (CI))
P value
Sex Age (years) Location Tumour size Histological differentiation TNM stage T classification N classification M classification MiR-378a-3p expression MiR-378a-5p expression
Male versus female 20–82 (median, 63) Colon versus rectum 65 cm versus >5 cm Well, moderately, poorly I, II, III, IV T1, T2, T3, T4 N0, N1, N2 M0 versus M1 Low versus high Low versus high
1.086 0.976 0.964 2.336 1.563 1.795 1.716 1.222 2.246 0.205 0.154
0.797 0.044 0.846 0.01 0.046 0.006 0.148 0.546 0.128
Available at www.sciencedirect.com
ScienceDirect journal homepage: www.ejcancer.com
Clinical and biological significance of miR-378a-3p and miR-378a-5p in colorectal cancer Hui Li 1, Sujuan Dai 1, Tiantian Zhen 1, Huijuan Shi, Fenfen Zhang, Yang Yang, Lili Kang, Yingjie Liang, Anjia Han ⇑ Department of Pathology, The First Affiliated Hospital, Sun Yat-Sen University, Guangzhou, China
KEYWORDS MiR-378a-3p MiR-378a-5p Insulin-like growth factor 1 receptor Colorectal cancer
Abstract To investigate miR-378a-3p and miR-378a-5p expression and their relationships with the clinicopathological features of colorectal cancer (CRC). Our results showed that miR-378a-3p and miR-378a-5p expression were dramatically lower in CRC cell lines and tissues than that in adjacent normal colorectal mucosal tissues, respectively. MiR-378a-3p and miR-378a-5p expression were significantly associated with histological differentiation and TNM stage, respectively. CRC patients with low miR-378a-3p and miR-378a-5p expression had a significantly shorter survival time than those patients with high miR-378a-3p and miR-378a-5p expression (p < 0.001, p < 0.001), respectively. Univariate and multivariable Cox regression analysis showed that tumour size, TNM stage, miR-378a-3p expression and miR-378a-5p expression were independent prognostic factors for CRC patients. Ectopic miR-378a-3p or miR-378a-5p expression inhibited cellular proliferation and colony formation, induced apoptosis and G1-phase cell cycle arrest in CRC cells, but had no effect on migration and invasion of CRC cells. Furthermore, miR-378a-3p over-expression or downregulation could inhibit or enhance insulin-like growth factor 1 receptor (IGF1R) expression in CRC cells. There was a significantly negative correlation between IGF1R protein expression and miR-378a-3p expression in CRC tissues. MiR-378a-3p over-expression or down-regulation suppressed or enhanced phosphorylated-ERK1/2 protein level, but had no effect on phosphorylated-Akt protein level. In conclusion, miR-378a-3p and miR-378a-5p expression might play an important role as tumour suppressor gene in the initial stage of carcinogenesis of CRC. Ó 2013 Elsevier Ltd. All rights reserved.
⇑ Corresponding author: Address: Department of Pathology, The First Affiliated Hospital, Sun Yat-Sen University, 58, Zhongshan Road II, Guangzhou 510080, China. Tel./fax: +86 20 87332235. E-mail address: [email protected] (A. Han). 1 Hui Li, Sujuan Dai and Tiantian Zhen contributed equally to this work.
0959-8049/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ejca.2013.12.010
Please cite this article in press as: Li H. et al., Clinical and biological significance of miR-378a-3p and miR-378a-5p in colorectal cancer, Eur J Cancer (2014), http://dx.doi.org/10.1016/j.ejca.2013.12.010
2
H. Li et al. / European Journal of Cancer xxx (2014) xxx–xxx
1. Introduction Colorectal cancer (CRC) is one of the most common malignancies worldwide. The prognosis of patients with CRC remains poor despite the improvement of current treatment modalities including surgical resection, radiotherapy and chemotherapy. The 5-year overall survival rate for CRC patients is just 50–60% mainly because of tumour recurrence and/or metastasis [1]. Thus, it is imperative to find novel biomarkers and develop novel treatment strategies for CRC. MicroRNAs (miRNAs) which are composed of 21–25nucleotides are small, endogenous non-coding RNAs and regulate the expression of a variety of genes based on sequence complementarities with their target mRNA molecules. It has been proposed that miRNAs imperfectly complementary to the binding sites in the 30 untranslated regions (UTR) of their target mRNA repressprotein expression through translational inhibition, whereas those miRNAs perfectly complementary to mRNA targets induce mRNA degradation through the RNA-mediated interference pathway [2]. miRNAs are involved in diverse cellular biological functions, including proliferation, development, differentiation, apoptosis, migration, invasion and angiogenesis. An increasing number of evidence shows that miRNAs contribute to development and progression of a variety of human cancers [3,4]. The transcript of miR-378 via post-transcriptional processing generates two mature miRNAs, namely, miR-378a-3p and miR-378a-5p. An increasing number of evidence shows that miR-378 induces malignant transformation of human breast cancer cell [5], increases cell survival, colony formation, contributes to multiple drug resistance in human malignant glioma cells [6], and promotes migration, invasion and tumour angiogenesis of human lung cancer cell [7]. MiR-378 up-regulation might have an adverse impact on prognosis in acute myeloid leukaemia [8]. Circulation of miR-378 in serum is a potential biomarker for renal cell carcinoma and early detection of gastric cancer [9,10]. However, microarray-based miRNA profiling study reveals that miR-378 is down-regulated in oral squamous carcinoma, basal cell carcinoma of the skin, gastric cancer and CRC [11–14]. Whether miR-378 is an oncogenic or tumour suppressor gene in CRC remains unknown. Moreover, whether miR-378a-3p or miR-378a-5p plays an important role in development and progression of CRC needs clarify. Herein, we investigated the clinical and biological significance of miR-378a3p and miR-378a-5p in human CRC.
mucosal tissues were obtained from our Hospital between January 2004 and December 2005. Twelve pairs of fresh CRC and adjacent normal colorectal mucosal tissues were collected from May 2012 through October 2012. The fresh CRC specimens and paired adjacent normal colorectal tissues from surgery were immediately stored at 4 °C for 24 h in RNA later solution (Sigma, St. Louis, MO), then at –80 °C until RNA extraction. No patients had received chemotherapy and/or radiotherapy before surgery. The histopathology of the disease was determined by two pathologists according to the criteria of the World Health Organisation (2010). Clinical staging was done according to the Union for International Cancer Control classification. For the research purposes of these clinical materials, prior patient’s consents and approval from the Institutional Research Ethics Committee were obtained. Follow-up information was available for all patients. Detailed clinical information about these patients including sex, age, tumour location, histological differentiation, tumour size, TNM stage, T classification, N classification and distant metastasis status, is summarised in Table 1. 2.2. Cell lines and transfection Human CRC cell lines LOVO and SW1116 were maintained in RPMI-1640 medium (Invitrogen, Carlsbad, CA). The CRC cell line HT29 was cultured in Dulbecco’s modified Eagle’s medium (Invitrogen). The CRC cell lines SW480 and SW620 were grown in Leibovitz’s L-15 Medium (Invitrogen, Carlsbad, CA). The CRC cell line HCT116 was maintained in McCoy’s 5A Medium (Invitrogen). All medium was supplemented with 10% (v/v) foetal bovine serum (GIBCO, Carlsbad, CA), 1 antibiotic/antimycotic (100 units/mL streptomycin, 100 units/mL penicillin and 0.25 mg/mL amphotericin B). All cell lines were cultured in a humidified incubator at 37 °C with 5% CO2. 2 105 cells were plated in 6-well plates and transfected with 100 nmol/L miRNAs using LipofectamineÒ RNAiMAX Transfection Reagent (Invitrogen) according to the manufacturer’s protocol. The miR-378a-3p mimic, miR-378a-5p mimic, non-specific control mimic, miR-378a-3p inhibitor, miR-378a-5p inhibitor and non-specific control inhibitor were purchased from Life Technologies (Carlsbad, CA). 2.3. Reliability analysis of miRNA in CRC FFPE tissues by quantitative real time polymerase chain reaction (PCR)
2. Materials and methods 2.1. Clinical samples and patient information Ninety-six pairs of formalin-fixed paraffin-embedded (FFPE) archived CRC and adjacent normal colorectal
Total RNA extracted from 12 pairs of CRC FFPE tissues and frozen tissues using Trizol reagent (Invitrogen) according to the manufacturer’s instructions was used to determine whether miR-378a-3p could be detected reliably in archival specimens by real-time
Please cite this article in press as: Li H. et al., Clinical and biological significance of miR-378a-3p and miR-378a-5p in colorectal cancer, Eur J Cancer (2014), http://dx.doi.org/10.1016/j.ejca.2013.12.010
H. Li et al. / European Journal of Cancer xxx (2014) xxx–xxx
3
Table 1 MiR-378a-3p and miR-378a-5p expression levels and their relationship with clinicopathological features of colorectal cancer. Variable
No. (%)
Expression level of miR-378a-3p
P value
Expression level of miR-378a-5p
P value
Sex Male Female
53(55.21) 43(44.79)
–1.44 ± 1.52 –1.45 ± 1.27
0.965
–1.22 ± 1.49 –1.45 ± 1.49
0.447
Age (years) 663 >63
53(55.21) 43(44.79)
–1.54 ± 1.28 –1.34 ± 1.55
0.496
–1.52 ± 1.44 –1.08 ± 1.52
0.154
Location Colon Rectum
50(52.08) 46(47.92)
–1.49 ± 1.41 –1.40 ± 1.41
0.748
–1.23 ± 1.51 –1.43 ± 1.47
0.520
Histological differentiation Well Moderately & poorly
20(20.83) 76(79.17)
–0.76 ± 1.36 –1.63 ± 1.36
0.013
–0.62 ± 1.40 –1.51 ± 1.46
0.016
TNM stage I–II III–IV
47(48.96) 49(51.04)
–1.14 ± 1.58 –1.74 ± 1.15
0.033
–0.98 ± 1.58 –1.66 ± 1.33
0.025
T classification T1–T2 T3–T4
18(18.75) 78(81.25)
–0.84 ± 1.24 –1.59 ± 1.41
0.032
–0.82 ± 1.35 –1.44 ± 1.50
0.094
N classification N0 N1–N2
45(46.88) 51(53.12)
–1.21 ± 1.61 –1.66 ± 1.17
0.121
–1.06 ± 1.53 –1.56 ± 1.42
0.103
M classification M0 M1
91(94.79) 5(5.21)
–1.48 ± 1.39 –0.85 ± 1.64
0.329
–1.28 ± 1.39 –2.19 ± 2.87
0.182
PCR. Relative to U6 snRNA, the miR-378a-3p expression level was determined using 2–DCT, where DCT = CT (miR-378a-3p in FFPE tissues)–CTU6 or DCT = CT (miR-378a-3p in Frozen tissues)–CTU6. 2–DCT value was not fitted to normality distribution and was subsequently transformed at log 2. Then linear regression was used to investigate the correlation between miRNA expression in CRC FFPE tissues and frozen tissues.
and adjacent normal colon tissues. The quantity of miR-378a-3p and miR-378a-5p in each CRC tissue relative to its paired adjacent normal colon tissue was calculated using the equation [RQ = 2–DDCT, DDCT = (CTmiRNA-CTU6)T – (CTmiRNA-CTU6)N]. The expression level of miR-378a-3p or miR-378a-5p was classified into low expression group and high expression group compared with the mean expression level of miR-378a-3p or miR-378a-5p, respectively.
2.4. Expression of miR-378a-3p and miR-378a-5p by quantitative real time PCR
2.5. Western blot analysis
Reverse transcription was performed using One Step PrimeScript miRNA cDNA Synthesis Kit (TaKaRa Biotechnology Co., Ltd., Dalian, China), and quantitative real time PCR was performed using SYBR Premix Ex Taq II (TaKaRa Biotechnology). RNU6B snRNA was used for sample loading normalisation. The specific forward primer of miR-378a-3p was 50 -CTGGACTTGGAGTCAGAAGGAA-30 (TaKaRa Biotechnology). The specific forward primer of RNU6B was 50 -ACGCAAATTCGTGAAGCGTT-30 . The specific forward primer of miR-378a-5p was 50 -CTGACTCCAGG TCCTGTGTA-30 . Reverse primer for miR-378a-3p, miR-378a-5p and U6B snRNA was Uni-miR qPCR primer (TaKaRa code: D350A). Each sample was analysed in triplicate. Relative expression levels of miR-378a-3p and miR-378a-5p were calculated using 2–DCT in CRC
As described previously [15], 30 lg of total proteins were loaded onto 10% sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS–PAGE) and transferred onto polyvinylidene difluoride (PVDF) membranes (Millipore, Billerico, MA) that were subsequently blocked in 5% non-fat milk in TBST (20 mmol/L Tris, pH 7.6, 137 mmol/L NaCl, 0.1% Tween-20). The membranes were incubated with primary antibodies including rabbit anti-insulin-like growth factor 1 receptor (IGF1R), rabbit anti-phospho-AKT (Ser473), rabbit anti-p44/42MAPK (ERK1/2), rabbit anti-phospho-p44/ 42 MAPK (p-ERK1/2) (Thr202/Tyr204) and rabbit anti-glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (Cell Signalling Technology, Danvers, MA) at 4 °C overnight. After washing, the membranes were incubated with secondary antibody HRP-conjugated goat
Please cite this article in press as: Li H. et al., Clinical and biological significance of miR-378a-3p and miR-378a-5p in colorectal cancer, Eur J Cancer (2014), http://dx.doi.org/10.1016/j.ejca.2013.12.010
4
H. Li et al. / European Journal of Cancer xxx (2014) xxx–xxx
anti-rabbit (Cell Signalling Technology) and visualised by enhanced chemiluminescence detection kit (Millipore). 2.6. Immunohistochemistry staining and evaluation As described previously [16], immunohistochemistry staining for IGF1R was performed on CRC FFPE tissue sections. The sections were deparaffinised, rehydrated in serially graded ethanol and heated in citric buffer (pH 6.0) once for 3 min in a microwave oven for antigen retrieval. They were then washed with distilled water, blocked with 3% hydrogen peroxide and incubated with rabbit anti-IGF1R (Cell Signalling Technology) monoclonal antibody, at dilution of 1:100 at 4 °C for 12 h. The tissues were then stained using EnVision-HRP secondary antibody (Dako, Carpinteria, CA) for 30 min at room temperature, washed with a 0.01 mol/L concentration of phosphate buffered saline (PBS), stained with 0.5% diaminobenzidine and counterstained with Mayer’s haematoxylin, then air dried, and mounted with glycerol gelatin. The degree of IGF1R immunostaining was based on both the proportion of positively stained tumour cells and intensity of staining. The proportion of tumour cells was scored as follows: 0 (25–50% positive tumour cells), 3 (>50–75% positive tumour cells) and 4 (>75% positive tumour cells). Meanwhile, staining intensity was scored as follows: 0 (no staining); 1 (weak staining = light yellow), 2 (moderate staining = yellow brown) and 3 (strong staining = brown). Staining index was calculated as the proportion score staining intensity score. In each case, 10 high-power fields of representative areas were counted. The staining was independently assessed by two researchers who were unaware of the specific proteins being assessed or their putative role in CRC. 2.7. Cell proliferation assay Cells were plated in 96-well plates at 2 103 per well and transfected with miR-378a-3p or miR-378a-5p mimics or inhibitors. Non-specific control miRNA mimic or inhibitor was used as the control group. After 24, 48, 72, 96, 120 and 144 h transfection, cell proliferation was determined by 3-(4,5-dimethyl thiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT) assay (CellTiter 96 NonRadioactive Cell Proliferation Assay Kit, Promega Corporation, Madison, WI). The experiment was performed in triplicate. 2.8. Adhesion-dependent colony formation assay After 24 h transfection with miR-378a-3p or miR378a-5p mimics, the cells were plated in 6-well plates 2 102 per well and grown for 2 weeks. Then, the cells
were washed twice with PBS, fixed with 4% paraformaldehyde and stained with 0.5% crystal violet for 15 min. The number of colonies in 10 random view fields was counted under a microscope and the average number of colonies was achieved. The experiment was triplicated independently. 2.9. Analysis of cell cycle and apoptosis 6 105 cells were seeded in 6-well plates and incubated overnight till 50–60% confluence. The cells were transfected with 100 nM miR-378a-3p or miR-378a-5p mimic or miR-378a-3p or miR-378a-5p inhibitor and harvested at 48 h, washed in cold PBS, fixed with 80% ethanol for 8 h at 4 °C, then stained with propidium iodide buffer (50 mg/ml propidium iodide, 0.1% sodium citrate and 0.1% Triton X-100) for 3 h at 4 °C. Non-specific control miRNA mimic or inhibitor was used as control group. 2 104 cells were analysed for cell cycle and apoptosis using a Becton Dickinson FACScan (Becton Dickinson Immunocytometry Systems, San Jose, CA). The percentage of cells in each phase of the cell cycle and apoptotic cells was quantified using Cell Quest software, respectively. This experiment was performed in triplicate. 2.10. Scratch wound migration assay 2 105 LOVO cells were seeded in 6-well plates and grown to 60% confluency in complete medium. The cells were transfected with 100 nM miR-378a-3p mimic or miR-378a-5p mimic, non-specific control miRNA mimic was used as the control group. After 24 h transfection, vertical scratches were then made using a 200 ll plastic filter tip to create a ‘wound’ of approximately 200 lm in diameter. To eliminate dislodged cells, culture medium was removed and wells were washed with PBS. The average distance of migrating cells was determined under an inverted microscope at 0, 12, 24 and 48 h. The experiment was performed in triplicate. 2.11. Transwell invasion assay Transwell invasion assay was carried out using 24-well MILLIcell Hanging Cell Culture Inserts 8 lm membrane coated with matrigel in the upper chamber (MILLIPORE). After 24 h transfection, cells were seeded into the upper chamber at a density of 3 104/ chamber well and incubated at 37 °C for 24 h, followed by removal of cells inside the upper chamber with cotton swabs. Invasive cells on the lower membrane surface were fixed with 95% ethanol for 30 min and stained with 0.2% Crystal Violet solution for 20 min. Five random fields from each invasion assay were counted under an inverted microscope. Three independent experiments were performed and the data were presented as mean ± standard deviation (SD).
Please cite this article in press as: Li H. et al., Clinical and biological significance of miR-378a-3p and miR-378a-5p in colorectal cancer, Eur J Cancer (2014), http://dx.doi.org/10.1016/j.ejca.2013.12.010
H. Li et al. / European Journal of Cancer xxx (2014) xxx–xxx
2.12. Statistical analysis All statistical analyses were performed using SPSS 16.0 statistics software. Student’s t test was used to compare the levels of cellular proliferation, apoptosis, migration, invasion and colony formation between different groups. Chi-square test was used to compare the levels of miR-378a-3p or miR-378a-5p expression and various clinicopathological parameters of CRC patients. The correlation of miR-378a-3p or miR-378a-5p expression level in CRC FFPE tissues and frozen tissues was analysed by Spearman’s correlation coefficients and linear regression analysis. Overall survival time of CRC patients was analysed by Kaplan–Meier analysis. Independent prognostic factor for CRC patients was analysed using Cox regression analysis. The correlation between miR-378a-3p expression and IGF1R expression in CRC was analysed by Spearman’s correlation coefficients. p < 0.05 was set to be statistically significant. 3. Results 3.1. Reliability of miRNA expression detection in CRC FFPE tissues by quantitative real time PCR To determine the reliability of miRNA expression detection in CRC FFPE tissues by quantitative real time PCR, we detected miR-378a-3p expression in 12 pairs of CRC FFPE tissues and frozen tissues by quantitative real time PCR. As shown in Fig. 1A, there was a significantly positive correlation of miR-378a-3p expression between CRC FFPE tissues and frozen tissues (R2 = 0.526, p = 0.008). The mean value of –DCTmiR-378a-3p in CRC FFPE tissues was –6.52. The mean value of –DCTmiR-378a-3p in their respective CRC frozen tissues was –6.17. The result indicated that it was reliable to detect miR-378a-3p expression level in CRC FFPE tissues by quantitative real time PCR. 3.2. MiR-378a-3p and miR-378a-5p expression in CRC cell lines and tissues As shown in Fig. 1B, the expression level of miR-378a3p and miR-378a-5p was lower in six CRC cell lines including HT29, HCT116, SW480, SW1116, SW620 and LOVO than that in normal colorectal mucosal tissue, respectively. Consistently, the expression level of miR378a-3p and miR-378a-5p was lower in 12 samples of fresh CRC tissues than that in their respective adjacent normal colorectal mucosal tissues, respectively (Fig. 1C). Furthermore, our result showed that the expression level of miR-378a-3p (–5.64 ± 1.34) and miR-378a-5p (–9.35 ± 1.58) was significantly lower in 96 samples of CRC FFPE tissues than that (–4.15 ± 0.99; –8.02 ± 1.10) in their respective adjacent normal colorectal mucosal tissues, respectively
5
(p < 0.001 and p < 0.001) (Fig. 1D). Moreover, there was a significantly positive correlation between miR378a-3p and miR-378a-5p expression in 96 samples of adjacent normal colorectal mucosal tissues (R2 = 0.572, p < 0.001) and 96 samples of CRC tissues (R2 = 0.770, p < 0.001), respectively. After Log2 transformation of miR-378a-3p and miR-378a-5p expression level, a significantly positive correlation was also found in 96 samples of CRC tissues (R2 = 0.689, P < 0.001) (Fig. 1E). 3.3. Relationship between miR-378a-3p and miR-378a-5p expression and clinicopathological features of CRC The mean expression level of miR-378a-3p in 96 samples of CRC was –1.45 ± 1.40 (median, –1.56; range, –4.68–3.11). The mean expression level of miR-378a5p in 96 samples of CRC was –1.32 ± 1.49 (median, –1.45; range, –6.88–4.22). Then, CRC patients were divided into two groups according to the mean expression level of miR-378a-3p and miR-378a-5p, respectively. Forty-five cases (46.9%) were low miR-378a-3p expression. Fifty-one cases (53.1%) were high miR378a-3p expression. Forty-one (42.7%) cases were low miR-378a-5p expression and 55 (57.3%) cases were high miR-378a-5p expression. The relationships between miR-378a-3p and miR-378a-5p expression and clinicopathological features of CRC were summarised in Table 1. The expression of miR-378a-3p was significantly associated with histological differentiation (p = 0.013), T classification (p = 0.032) and TNM stage (p = 0.033) of CRC. The expression of miR-378a-5p was significantly associated with histological differentiation (p = 0.016) and TNM stage (p = 0.025) of CRC. Moreover, miR-378a-3p and miR-378a-5p expression was lower in CRC with moderately and poorly-differentiation, high TNM stage (III–IV) and large tumour size (T3–T4) than that in CRC with well-differentiation, low TNM stage (I–II) and small tumour size (T1–T2), respectively (Fig. 2A–C). We also found that the greatest diameter of CRC was significantly larger (6.09 ± 3.17 cm) in low miR-378a-3p expression group than that (4.80 ± 1.88 cm) in high miR-378a-3p expression group (p = 0.019). Likewise, the greatest diameter of CRC was significantly larger (6.29 ± 2.80 cm) in low miR-378a-5p expression group than that (4.88 ± 1.86 cm) in high miR-378a-5p expression group (p = 0.007) (Fig. 2D). However, the expression level of miR-378a-3p and miR-378a-5p was not significantly associated with sex, age, tumour location, N classification and distant metastasis status, respectively (Table 1). 3.4. The relationship between miR-378a-3p and miR378a-5p expression and prognosis of CRC patients Kaplan–Meier survival analyses showed that CRC patients with low miR-378a-3p and miR-378a-5p
Please cite this article in press as: Li H. et al., Clinical and biological significance of miR-378a-3p and miR-378a-5p in colorectal cancer, Eur J Cancer (2014), http://dx.doi.org/10.1016/j.ejca.2013.12.010
6
H. Li et al. / European Journal of Cancer xxx (2014) xxx–xxx
Fig. 1. MiR-378a-3p and miR-378a-5p expression in colorectal cancer (CRC) cell lines and tissues. (A) Comparison of miR-378a-3p expression pattern from matched samples of CRC formalin-fixed paraffin-embedded (FFPE) and frozen tissues. (B) MiR-378a-3p (left) and miR-378a-5p (right) expression in CRC cell lines compared with adjacent normal colon mucosa epithelial cells (Control) by quantitative real time polymerase chain reaction (PCR) analysis. (C) MiR-378a-3p (left) and miR-378a-5p (right) expression in twelve pairs of fresh CRC tissues (T) and matched adjacent normal colorectal mucosa tissues (N). (D) MiR-378a-3p (left) and miR-378a-5p (right) in 96 pairs of CRC FFPE tissues (T) and matched adjacent normal colorectal mucosal tissues (N). The term DCT (DCT = CTmiR-378a-CTU6) was used to describe the expression level of miR-378a3p and miR-378a-5p. Small nuclear RNA U6 served as internal normalised references for miR-378a-3p and miR-378a-5p. X ± SD are represented in the images. (E) The correlation between miR-378a-3p expression and miR-378a-5p expression in 96 samples of adjacent normal colorectal mucosal tissues (left) and CRC tissues (middle and right).
Please cite this article in press as: Li H. et al., Clinical and biological significance of miR-378a-3p and miR-378a-5p in colorectal cancer, Eur J Cancer (2014), http://dx.doi.org/10.1016/j.ejca.2013.12.010
H. Li et al. / European Journal of Cancer xxx (2014) xxx–xxx
7
Fig. 2. The relationship between miR-378a-3p and miR-378a-5p expression and clinicopathological characteristics of colorectal cancer (CRC). (A) The expression level of miR-378a-3p (left) and miR-378a-5p (right) in moderately to poorly differentiated CRC group was lower than that of well differentiated CRC group. (B) The expression level of miR-378a-3p in T3–T4 group was lower than that in T1–T2 group (left), while the expression level of miR-378a-5p was not significantly associated with T classification (right). (C) The expression level of miR-378a-3p (left) and miR-378a-5p (right) in III–IV stage was lower than that of I–II stage. (D) Low expression of miR-378a-3p (left) and miR-378a-5p (right) was significantly associated with large tumour size, respectively. (E) Kaplan–Meier survival curve and log-rank test for CRC patients with different expression levels of miR-378a-3p (left) and miR-378a-5p (right). Please cite this article in press as: Li H. et al., Clinical and biological significance of miR-378a-3p and miR-378a-5p in colorectal cancer, Eur J Cancer (2014), http://dx.doi.org/10.1016/j.ejca.2013.12.010
8
H. Li et al. / European Journal of Cancer xxx (2014) xxx–xxx
expression had a significantly shorter survival time than those patients with high miR-378a-3p and miR-378a-5p expression (p < 0.001, p < 0.001), respectively (Fig. 2E). The 3- and 5-year median survival rate for CRC patients was 56.9% (range, 43.4–70.4%) and 50.8% (range, 37.1–64.5%) in low miR-378a-3p expression group compared with 93.3% (range, 91.9–94.4%) and 86.7% (range, 76.7–96.7%) in high miR-378a-3p expression group, respectively. The 3 and 5-year median survival rate for CRC patients was 58.2% (range, 45.1–71.3%) and 52.5% (range, 39.2–65.8%) in low miR-378a-5p expression group compared with 95.1% (range, 94.1–96.0%) and 87.8% (range, 77.8–97.8%) in high miR-378a-5p expression group, respectively. Univariate Cox regression analysis showed that age (Hazard ratio, 0.976; 95% confidence interval (CI), 0.953–0.999; p = 0.044), tumour size (Hazard ratio, 2.336; 95% CI, 1.226–4.453; p = 0.01), histological differentiation (Hazard ratio, 1.563; 95% CI, 1.008–2.425; p = 0.046), TNM stage (Hazard ratio, 1.795; 95% CI, 1.178–2.735; p = 0.006), miR-378a-3p expression (Hazard ratio, 0.205; 95% CI, 0.087–0.484; p < 0.001) , and miR-378a-5p expression (Hazard ratio, 0.154; 95% CI, 0.057–0.413; p < 0.001) were significantly associated with the prognosis of CRC patients (Table 2). Multivariable Cox regression analysis showed that tumour size, TNM stage, miR-378a-3p expression (Hazard ratio,
0.154; 95% CI, 0.059–0.404; p < 0.001) (Table 3) and miR-378a-5p expression (Hazard ratio, 0.147; 95% CI, 0.053–0.413; p < 0.001) (Table 4) were independent prognostic factors for CRC patients. 3.5. Ectopic miR-378a-3p or miR-378a-5p expression inhibited cellular proliferation and colony formation in CRC cells To investigate the effect of ectopic miR-378a-3p or miR-378a-5p expression on cellular proliferation of CRC, over-expression of miR-378a-3p or miR-378a-5p was established in LOVO and SW620 cell lines by miR-378a-3p mimic or miR-378a-5p mimic transfection. MiR-378a-3p or miR-378a-5p control mimic was used as the control group (Fig. 3A). Our result showed that transfection of miR-378a-3p mimic and miR-378a-5p mimic dramatically inhibited LOVO (p = 0.001 and p = 0.011) and SW620 (p = 0.001 and p = 0.001) cellular proliferation compared with the control group by MTT assay, respectively (Fig. 3B). Moreover, colony formation assay demonstrated that miR-378a-3p mimic and miR-378a-5p mimic significantly inhibited colony formation of LOVO (p = 0.001 and p = 0.042) and SW620 (p = 0.002 and p = 0.003) cell line by less in colony number and smaller in colony size compared with the control group, respectively (Fig. 3C). We also
Table 2 Univariate Cox regression analysis of the relationship between miR-378a-3p and miR-378a-5p expression level, clinicopathological features and survival of colorectal cancer patients. Variable
Subset
Hazard ratio (95% confidence interval (CI))
P value
Sex Age (years) Location Tumour size Histological differentiation TNM stage T classification N classification M classification MiR-378a-3p expression MiR-378a-5p expression
Male versus female 20–82 (median, 63) Colon versus rectum 65 cm versus >5 cm Well, moderately, poorly I, II, III, IV T1, T2, T3, T4 N0, N1, N2 M0 versus M1 Low versus high Low versus high
1.086 0.976 0.964 2.336 1.563 1.795 1.716 1.222 2.246 0.205 0.154
0.797 0.044 0.846 0.01 0.046 0.006 0.148 0.546 0.128
