Identification of MicroRNA-214 as a negative regulator of colorectal cancer liver metastasis by way of regulation of fibroblast growth factor receptor 1 expression

Authors

  • Dong-liang Chen,

    1. State Key Laboratory of Oncology in South China, Sun Yat-sen University Cancer Center, Collaborative Innovation Center for Cancer Medicine, Guangzhou, China
    2. Department of Medical Oncology, Sun Yat-sen University Cancer Center, Guangzhou, China
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Zhi-qiang Wang,

    1. State Key Laboratory of Oncology in South China, Sun Yat-sen University Cancer Center, Collaborative Innovation Center for Cancer Medicine, Guangzhou, China
    2. Department of Medical Oncology, Sun Yat-sen University Cancer Center, Guangzhou, China
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Zhao-lei Zeng,

    1. State Key Laboratory of Oncology in South China, Sun Yat-sen University Cancer Center, Collaborative Innovation Center for Cancer Medicine, Guangzhou, China
    2. Department of Experimental Research, Sun Yat-sen University Cancer Center, Guangzhou, China
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Wen-jing Wu,

    1. State Key Laboratory of Oncology in South China, Sun Yat-sen University Cancer Center, Collaborative Innovation Center for Cancer Medicine, Guangzhou, China
    2. Department of Medical Oncology, Sun Yat-sen University Cancer Center, Guangzhou, China
    Search for more papers by this author
  • Dong-sheng Zhang,

    1. State Key Laboratory of Oncology in South China, Sun Yat-sen University Cancer Center, Collaborative Innovation Center for Cancer Medicine, Guangzhou, China
    2. Department of Medical Oncology, Sun Yat-sen University Cancer Center, Guangzhou, China
    Search for more papers by this author
  • Hui-yan Luo,

    1. State Key Laboratory of Oncology in South China, Sun Yat-sen University Cancer Center, Collaborative Innovation Center for Cancer Medicine, Guangzhou, China
    2. Department of Medical Oncology, Sun Yat-sen University Cancer Center, Guangzhou, China
    Search for more papers by this author
  • Feng Wang,

    1. State Key Laboratory of Oncology in South China, Sun Yat-sen University Cancer Center, Collaborative Innovation Center for Cancer Medicine, Guangzhou, China
    2. Department of Medical Oncology, Sun Yat-sen University Cancer Center, Guangzhou, China
    Search for more papers by this author
  • Miao-zhen Qiu,

    1. State Key Laboratory of Oncology in South China, Sun Yat-sen University Cancer Center, Collaborative Innovation Center for Cancer Medicine, Guangzhou, China
    2. Department of Medical Oncology, Sun Yat-sen University Cancer Center, Guangzhou, China
    Search for more papers by this author
  • De-shen Wang,

    1. State Key Laboratory of Oncology in South China, Sun Yat-sen University Cancer Center, Collaborative Innovation Center for Cancer Medicine, Guangzhou, China
    2. Department of Medical Oncology, Sun Yat-sen University Cancer Center, Guangzhou, China
    Search for more papers by this author
  • Chao Ren,

    1. State Key Laboratory of Oncology in South China, Sun Yat-sen University Cancer Center, Collaborative Innovation Center for Cancer Medicine, Guangzhou, China
    2. Department of Medical Oncology, Sun Yat-sen University Cancer Center, Guangzhou, China
    Search for more papers by this author
  • Feng-hua Wang,

    1. State Key Laboratory of Oncology in South China, Sun Yat-sen University Cancer Center, Collaborative Innovation Center for Cancer Medicine, Guangzhou, China
    2. Department of Medical Oncology, Sun Yat-sen University Cancer Center, Guangzhou, China
    Search for more papers by this author
  • Lucia J. Chiao,

    1. Department of Translational Molecular Pathology, University of Texas M.D. Anderson Cancer Center, Houston, TX, USA
    Search for more papers by this author
  • Helene Pelicano,

    1. Department of Translational Molecular Pathology, University of Texas M.D. Anderson Cancer Center, Houston, TX, USA
    Search for more papers by this author
  • Peng Huang,

    1. State Key Laboratory of Oncology in South China, Sun Yat-sen University Cancer Center, Collaborative Innovation Center for Cancer Medicine, Guangzhou, China
    2. Department of Translational Molecular Pathology, University of Texas M.D. Anderson Cancer Center, Houston, TX, USA
    Search for more papers by this author
  • Yu-hong Li,

    1. State Key Laboratory of Oncology in South China, Sun Yat-sen University Cancer Center, Collaborative Innovation Center for Cancer Medicine, Guangzhou, China
    2. Department of Medical Oncology, Sun Yat-sen University Cancer Center, Guangzhou, China
    Search for more papers by this author
  • Rui-hua Xu

    Corresponding author
    1. State Key Laboratory of Oncology in South China, Sun Yat-sen University Cancer Center, Collaborative Innovation Center for Cancer Medicine, Guangzhou, China
    2. Department of Medical Oncology, Sun Yat-sen University Cancer Center, Guangzhou, China
    • Address reprint requests to: Prof. Rui-hua Xu, M.D. Ph.D., State Key Laboratory of Oncology in South China, Department of Medical Oncology, Sun Yat-sen University Cancer Center, No. 651 Dong Feng East Road, Guangzhou 510060, China. E-mail: xurh@sysucc.org.cn; fax: +86-20-87343468.

    Search for more papers by this author

  • Supported by the National High Technology Research and Development Program of China (863 Program), China (No.2012AA02A506); National Natural Science Foundation of China (No.81372570).

  • Potential conflict of interest: Nothing to report.

Abstract

The purpose of this study was to identify microRNAs (miRNAs) involved in the pathology of colorectal cancer (CRC) liver metastasis and investigate their underlying mechanisms. A total of 39 miRNAs were identified to be differentially expressed between 16 primary CRC tissues with liver metastases and 16 CRC tissues without liver metastases from 32 patients by Affymetric miRNA microarrays. A panel of eight miRNAs were confirmed to be significantly and differentially expressed between CRC tissues with and without liver metastases through quantitative reverse-transcription polymerase chain reaction (RT-PCR) analysis in the 32 patients. In a validated cohort of 99 CRC patients (44 with and 55 without liver metastases), only miR-214 was validated to be significantly down-regulated in CRC with liver metastases, which was associated with an unfavorable prognosis. Ectopic expression of miR-214 suppressed proliferation, migration, and invasion in vitro, tumor growth and liver metastasis in an in vivo xenograft mouse model, whereas miR-214 knockdown promoted proliferation, migration, and invasion in CRC cell lines. Further studies indicated that fibroblast growth factor receptor 1 (FGFR1) was a potential target of miR-214. Restoring miR-214 expression in CRC cells decreased endogenous FGFR1 messenger RNA (mRNA) and protein levels. FGFR1 knockdown mimicked the tumor suppressive effect of miR-214 on CRC cells, while reintroduction of FGFR1 abolished the tumor suppressive effect of miR-214 on CRC cells. Moreover, miR-214 expression levels were inversely correlated with FGFR1 in CRC patients. Conclusion: Down-regulation of miR-214 expression was correlated with increased FGFR1 expression levels, which may contribute to increased CRC liver metastasis. miR-214 may serve as a potential marker to predict survival, and the miR-214-FGFR1 axis may be a therapeutic target in CRC patients. (Hepatology 2014;60:598–609)

Abbreviations
CRC

colorectal cancer

FGFR1

fibroblast growth factor receptor 1

IHC

immunohistochemistry

miRNA

microRNA

qPCR

quantitative polymerase chain reaction.

Metastasis is a complex process that includes a series of steps in which tumor cells grow, detach from the primary tumor site, and migrate to a distant organ. In colorectal cancer (CRC), the liver is the most common site for distant metastasis, and liver metastasis is a common cause of cancer-related deaths in CRC patients.[1, 2] Clinically, ∼25% of CRC patients present with synchronous liver metastases at diagnosis, while other advanced CRC patients develop metachronous liver metastases within 3 years of treatment.[3] Most CRC patients with liver metastases are not suited for surgery; as a result, the 5-year survival rate after liver metastasis diagnosis is no higher than 10%.[4] In contrast, the 5-year survival rate for patients with early-stage CRC is more than 90%.[5, 6] Therefore, it is urgently necessary to unravel the underlying molecular mechanisms and genetic alterations that lead to CRC metastases.

Studies have found that genetic alterations in tumor cells lead to cellular heterogeneity, which might promote cancer cell invasiveness and colonization in specific organs during the metastatic process.[7] Recent reports have shown that several genetic signaling pathways including MACC1, Notch, and transforming growth factor beta (TGF-β)/Smad signaling[8-11] may be involved in CRC liver metastasis. In addition to protein-coding genes, increasing evidence has revealed critical roles for noncoding RNAs in human cancer metastasis, particularly microRNAs (miRNAs).[12]

miRNAs are a class of short, endogenous, noncoding RNAs (∼20-24 nucleotides) that regulate the expression of a wide variety of genes.[13] Through base pairing with the 3′-untranslated region (3′UTR) of target genes, miRNAs enhance mRNA degradation or inhibit posttranscriptional translation.[14] Increasing evidence has shown that miRNA alteration and dysfunction play critical roles during tumorigenesis and metastasis by way of the regulation of cancer cell proliferation, differentiation, apoptosis, and invasion.[12, 15, 16] For instance, miR-140-5p inhibits tumor growth and metastasis in hepatocellular carcinoma by targeting TGFBR1 and FGF917; Down-regulation of miR-224 and the passenger strand of miR-221 increase MBD2, suppressing maspin and promoting colorectal tumor growth and metastasis in mice18; miR-137 suppresses CRC invasion and metastasis by way of regulating FMNL2.19 Although miRNAs have been extensively investigated in recent years, the molecular regulatory mechanisms of miRNAs and their significance in CRC liver metastasis remain largely unknown and need exploring.

Materials and Methods

Human Tissue Specimens

A total of 16 CRC tissue samples were collected from patients with liver metastasis at diagnosis, and 16 CRC tissue samples were obtained from stage III patients who did not have metastasis at diagnosis or in the follow-up period of more than 2 years after resection. A validated cohort including 99 patients (44 with liver metastases and 55 without liver metastases) was also included in this study. Their primary CRC tissues (n = 99), adjacent normal tissues (n = 99), and corresponding matched liver metastatic tissues (n = 44) were obtained immediately after surgery and stored at −80°C or paraffin-embedded. All patients underwent radical resection at Sun Yat-sen University Cancer Center (Guangzhou, China) from January 2002 to July 2009. Detailed information is provided in the Supporting Materials and Methods.

RNA Extraction and Microarray Analysis

Total RNA was extracted using the mirVana miRNA Isolation Kit (Ambion, CA) according to the manufacturer's instructions. The RNA quality was assessed by formaldehyde agarose gel electrophoresis and quantified spectrophotometrically. The GeneChip miRNA 1.0 array (Affymetrix, Santa Clara, CA) with coverage of miRBase v.11 was used in this study according to a previously described method.[20] Details are provided in the Supporting Materials and Methods. The microarray data have been put into the NCBI Gene Expression Omnibus database and are available through GEO with accession number GSE53159.

In Vivo Proliferation and Metastasis Assays

Female BABL/c athymic nude mice purchased from the Animal Center of Guangdong Province (Guangzhou, China) aged 5-6 weeks were used. All animal experiments were performed under the experimental animal use guidelines of the National Institutes of Health. To evaluate the in vivo tumorigenic effects, HCT116-Lv-miR-214, HCT116-Lv-miR-NC, and HCT116 cells (1 × 106 cells per mouse) were subcutaneously injected into the flanks of the nude mice. Tumor size was measured every 4 days and tumor volume was calculated. Five weeks later the mice were sacrificed and the tumors were collected. The capacity to metastasize to the liver was determined following a previously described method.[21] Briefly, the mice were anesthetized by inhalation of isoflurane (0.5-1.0%) and oxygen. Through a 1-cm incision in the upper left lateral abdomen the spleen was exposed and 106 cells suspended in 20 μL phosphate-buffered saline (PBS) were injected into the distal tip of the spleen using a Hamilton syringe. After injection the spleen was replaced in the abdomen and the incision was closed with staples. The animals were sacrificed after 5 to 6 weeks and the spleen and liver were dissected out and embedded in paraffin.

Immunohistochemistry (IHC) Assays

The paraffin-embedded tissue blocks were cut into 4-μm slides. A fibroblast growth factor receptor 1 (FGFR1) rabbit antibody (#ab71928) was used. IHC analysis and qualification of FGFR1 expression was performed according to a previously described method.[22]

The details for cell culture and transfections, quantitative real-time PCR, cell proliferation assays, colony formation assays, in vitro cell wound healing and invasion assays, lentivirus production and transduction, western blot analyses, vector constructs, transfections, and luciferase activity assays are described in the Supporting Materials and Methods.

Statistical Analyses

Statistical analyses were performed using the SPSS software package (v. 13.0) or GraphPad Prism 5.0. Statistical significance was determined using Student t test, Fisher's exact test, or one-way analysis of variance (ANOVA) as appropriate. The Spearman correlation test was used for evaluating the correlations between miR-214 expression levels and FGFR1 protein levels in CRC patients. Survival curves were generated using the Kaplan-Meier method and assessed using the log-rank test. The Cox proportional hazard regression model was performed to identify independent prognostic factors. P < 0.05 was considered statistically significant.

Results

miR-214 Is Significantly Down-Regulated in CRC With Liver Metastasis

To study the miRNA profile that regulates human CRC liver metastasis, we performed an miRNA microarray analysis of the CRC tissues from patients with (n = 16) and without (n = 16) liver metastases. The clinicopathological information of the 32 CRC patients is listed in Supporting Table 1. We identified 39 microRNAs whose expression levels differed between the two groups with a fold-change of >2 and false discovery rate (FDR) <0.05 (Supporting Fig. 1). Among these miRNAs, 16 were up-regulated and 23 were down-regulated in CRC tissues with liver metastases compared with those without liver metastasis. We then validated these miRNAs in the 32 CRC tissues using real-time PCR analysis. The results showed only eight miRNAs were significantly differentially expressed between the twp groups. miR-29a, miR-196a, and miR-25 were up-regulated, whereas miR-21, miR-99b, miR-143, miR-345, and miR-214 were down-regulated in CRC tissues with liver metastases (Supporting Table 2). To further select the miRNAs that are significantly involved in CRC liver metastasis regulation, the expression levels of the eight miRNAs were measured in another validated cohort of 99 patients (44 with and 55 without liver metastases). To ensure that the reference gene snRNA U6 did not change between different samples, we calculated the mean CT values of U6. The U6 levels did not differ between CRC tissues without metastases (n = 55) and CRC tissues with liver metastases (n = 44; P = 0.072), tumor tissues (n = 99), and adjacent normal tissues (n = 99; P = 0.474) or primary CRC tissues (n = 44) and corresponding matched liver metastases (n = 44; P = 0.666) (Supporting Fig. 2). The results indicated that miR-214 was significantly down-regulated in CRC with liver metastases (Fig. 1A; P < 0.001); miR-345 and miR-143 were slightly reduced in CRC with liver metastasis (P = 0.078 and P = 0.081, respectively, data not shown), whereas other miRNAs, including miR-29a, miR-196a, miR-21, miR-99b, and miR-25 showed no significant difference between CRC tissues with and with liver metastasis (data not shown). Moreover, miR-214 was significantly reduced in tumor tissues compared with the corresponding normal tissues (Fig. 1B; P < 0.001). However, no significant difference in miR-214 expression was observed between the primary CRC tissues and the paired liver metastases (Fig. 1C; P = 0.068). In situ hybridization (ISH) analysis confirmed the expression pattern of miR-214 in tissues (Supporting Fig. 3). Based on these results, we focused on miR-214 in this study. Then we analyzed the clinicopathological implication of miR-214 in CRC patients. The patients were divided into two groups based on the median level of miR-214 expression levels; high miR-214 levels were negatively associated with tumor size (P = 0.030), lymph node invasion (P = 0.018), liver metastasis (P = 0.003), and TNM stage (P = 0.023; Table 1). Kaplan-Meier survival curves showed that patients with high miR-214 levels demonstrated a more favorable clinical outcome (P < 0.001; Fig. 1D). In addition, lymph node invasion, liver metastasis, and TNM stage were also associated with overall survival as demonstrated by univariate analysis (Table 2). However, age, gender, histological grade, tumor size, and tumor depth did not have prognostic value in this study. Notably, multivariate analysis indicated that miR-214 expression and liver metastasis were independent prognostic factors for CRC patients (P = 0.012 and P = 0.001, respectively; Table 2).

Table 1. Correlation Between Clinicopathological Parameters and miR-214 Expression Levels in 99 CRC Patients
VariablesnLow miR-214 Expression (%)High miR-214 Expression (%)P Value
  1. a

    P < 0.05, chi-square test.

  2. b

    m, tumor invasion of mucosa; sm, submocosa; mp, muscularis propria; ss, subserosa; se, serosa penetration; si, invasion to adjacent structures.

Age   0.093
<607433(67.3)41(82.0) 
≥602516(32.7)9(18.0) 
Gender   0.113
Male6830(61.2)38(76.0) 
Female3119(39.8)12(24.0) 
Tumor size   0.030a
<5cm289(18.3)19(38.0) 
≥5cm7140(81.7)31(62.0) 
Tumor depthb   0.123
m/sm/mp4217(34.7)25(50.0) 
ss/se/si5732(65.3)25(50.0) 
Histological grade   0.154
Well198(16.3)11(22.0) 
Moderate2911(22.4)18(36.0) 
Poor and others5130(61.3)21(42.0) 
Lymph node invasion   0.018a
Absent299(19.5)20(40.0) 
Present7040(80.5)30(60.0) 
Liver metastasis   0.003a
Absent5520(40.8)35(70.0) 
Present4429(59.2)15(30.0) 
TNM stage   0.023a
I-II3311(22.4)22(44.0) 
III-IV6638(77.6)28(56.0) 
Figure 1.

miR-214 is down-regulated in CRC tissues with liver metastases. (A) miR-214 expression levels in CRC with liver metastases (n = 44) are significantly lower than in CRC tissues without liver metastases (n = 55; P < 0.001). (B) miR-214 expression levels in tumor tissues (n = 99) are significantly lower than in adjacent normal tissues (n = 99; P < 0.001). (C) The miR-214 expression levels in primary CRC tissues (n = 44) are comparable to paired liver metastases (n = 44; P = 0.068). (D) Kaplan-Meier analysis of overall survival based on miR-214 levels in 99 CRC patients shows that patients with high miR-214 levels possess a better clinical outcome than patients with low miR-214 levels (P < 0.001).

Table 2. Univariate and Multivariate Analyses of Various Potential Prognostic Factors in 99 CRC Patients
FactorsUnivariate AnalysisMultivariate Analysis
HR (95% CI)PHR (95% CI)P
  1. a

    P < 0.05. HR, hazard ratio; CI, confidence interval.

Age1.02(0.83-1.21)0.168
Gender1.23(0.87-1.62)0.436
Histological grade1.23(0.91-1.76)0.073
Tumor size0.98(0.69-1.82)0.543
Tumor depth1.21(1.01-1.71)0.075
Lymph node invasion1.72(1.21-2.09)0.041a0.97(0.68-1.35)0.775
Liver metastasis1.97(1.36-2.95)<0.001a1.86(1.36-3.08)0.001a
TNM stage1.37(1.12-2.04)<0.001a1.31(0.89-1.73)0.253
miR-214 expression2.02(1.52-2.75)<0.001a1.39(1.11-1.98)0.012a
Figure 2.

Exogenetic overexpression of miR-214 suppresses CRC cell proliferation, migration, and invasion in vitro. (A) Relative expression levels of miR-214 in HCT116 and LoVo cells after transfection with miR-214 mimics (*P < 0.05). (B) Ectopic miR-214 expression significantly inhibits cell viability in HCT116 and LoVo cells (*P < 0.05). (C) Ectopic miR-214 expression significantly inhibits colony formation in HCT116 and LoVo cells (*P < 0.05). (D) miR-214 overexpression inhibits wound healing in HCT116 cells. (E,F) Ectopic miR-214 expression inhibits cell invasion in HCT116 and LoVo cells (*P < 0.05).

Figure 3.

Ectopic miR-214 expression inhibits CRC cell growth and liver metastasis in vivo. (A) The infection efficiency of lentivirus in HCT116 cells. Because this vector contains a GFP fragment, the cells emit green fluorescence when infected by the virus. The CRC cells were observed under light and green fluorescence microscopy. (B,D) Ectopic miR-214 expression significantly inhibits CRC cell growth in nude mice (P = 0.024). (C) Compared with mice injected with HCT116-Lv-miR-NC and HCT116 cells, fewer mice injected with HCT116-Lv-miR-214 formed liver metastases (P = 0.003). (E,F) The micro-metastatic nodules in the liver were significantly fewer in the mice injected with HCT116-Lv-miR-214 compared with those injected with HCT116-Lv-miR-NC and HCT116 cells (P = 0.004).

Exogenetic Overexpression of miR-214 Suppresses CRC Cell Proliferation, Migration, and Invasion In Vitro

As miR-214 is associated with tumor metastasis in CRC patients, we investigated its role in CRC cells. The expression level of miR-214 was measured in CRC cell lines and a normal colon cell line, CCD-112CoN. Interestingly, compared with that of CCD-112CoN, miR-214 expression levels were significantly reduced in the CRC cell lines (all P < 0.05; Supporting Fig. 4). Among the six CRC cell lines, HCT116 presented with the lowest level of miR-214 while HT-29 displayed highest level of miR-214. The effects of miR-214 on cell proliferation and invasion were then evaluated in CRC cells through gain-of-function and loss-of-function assays. HCT116 and LoVo cells, which possess relatively low miR-214 levels, were transfected with miR-214 mimics to overexpress miR-214, whereas HT-29 cells, which display relatively high miR-214 levels, were transfected with an miR-214 inhibitor to knockdown the endogenous miR-214. The transfection efficiency was confirmed through real-time PCR (both P < 0.05; Fig. 2A; Supporting Fig. 5A). As shown in Fig. 2B, miR-214-transfected HCT116 and LoVo cells demonstrated a significantly slower growth rate than NC-transfected and blank cells (P = 0.027 and P = 0.038, respectively). In contrast, the miR-214 inhibitor-transfected HT-29 cells showed increased growth compared with scramble-transfected and blank cells (P = 0.041; Supporting Fig. 5B). A colony formation assay also showed that miR-214 overexpression resulted in significant tumor growth inhibition in HCT116 and LoVo cells (P = 0.013 and P = 0.029, respectively; Fig. 2C). In addition, the effects of miR-214 on the motility and invasive capacities of CRC cells were evaluated by wound healing and transwell assays. miR-214 overexpression markedly reduced the migration and invasion ability of HCT116 and LoVo cells (all P < 0.05; Fig. 2D-F). In contrast, miR-214 knockdown in HT-29 cells led to increased cell invasion (P = 0.018; Supporting Fig. 5C).

Figure 4.

FGFR1 is a direct target of miR-214 in CRC cells. (A) The binding sites of miR-214 in the FGFR1 3′untranslated region. (B) The FGFR1 mRNA levels in CRC cell lines transfected with miR-214 mimics or negative control (NC) vector (*P = 0.015). (C) The FGFR1 protein levels in CRC cell lines treated with miR-214 mimic or NC vector. (D) The pcDNAmiR-214 expression plasmid or the negative control, a pGL3 luciferase vector containing the wild or mutant type of FGFR1 3′UTR, were cotransfected into HEK293a and HCT116 cells, and the relative firefly luciferase activity was measured. The data are presented as the mean ± SD of three independent experiments (*P < 0.05).

Figure 5.

miR-214 represses FGFR1 to suppress CRC growth and metastasis. (A) Western blot assays were performed to check the efficiency of miR-214 overexpression, FGFR1 knockdown, and FGFR1 reintroduction in HCT116 cells. (B-D) Cell invasion, wound healing, and colony formation in HCT116 cells transfected with different plasmids (*P < 0.05).

Ectopic miR-214 Expression Inhibits CRC Cell Growth and Liver Metastasis In Vivo

To investigate the in vivo effects of miR-214 on CRC cells, we constructed two stable cell lines by using the lentivirus vector to mediate miR-214 expression in HCT116 cells, designated HCT116-Lv-miR-214 and HCT116-Lv-miR-NC, respectively. The ectopic expression efficiency was confirmed through real-time PCR and observation of green fluorescence (Fig. 3A). To investigate the effect of miR-214 on in vivo tumor growth, the cells were injected into the flanks of nude mice. Tumor size was measured every 4 days; after 5 weeks, the mice were sacrificed and the tumors were collected. The results showed that the volume of the tumors from the mice injected with HCT116-Lv-miR-214 cells was significantly less than those of HCT116-Lv-miR-NC and HCT116 cells (n = 9 mice per group; P = 0.024; Fig. 3B,D). To evaluate the inhibition of in vivo liver metastasis, the cells were injected into the distal tip of the spleen using a Hamilton syringe. Six weeks later the mice were sacrificed and the spleen and liver were removed and embedded in paraffin. All mice formed tumors in the spleen. Most of the mice injected with HCT116-Lv-miR-NC (8 of 10) and HCT116 (9 of 10) formed liver metastases; in contrast, few of the mice injected with HCT116-Lv-miR-214 (2 of 10) formed liver metastases (P = 0.003; Fig. 3C). Moreover, the numbers of metastatic nodules in the livers were significantly reduced in mice injected with HCT116-Lv-miR-214 compared with those injected with HCT116-Lv-miR-NC and HCT116 cells (P = 0.004; Fig. 3E,F). In addition, to confirm whether the endogenous miR-214 level is associated with cell growth and invasion ability, the CCD-112CoN, HT-29, and HCT116 cells were injected into the flank and the distal tip of the spleen of nude mice as described above. Interestingly, none of 6 mice injected with CCD-112CoN, 3 of 6 mice injected with HT-29, and 6 of 6 mice injected with HCT116 formed tumor at the end of 5 weeks (data not shown). Likewise, none of 7 mice injected with CCD-112CoN, 2 of 7 injected with HT-29, and 7 of 7 mice injected with HCT116 formed liver metastasis by the end of 6 weeks (data not shown).

FGFR1 Is a Direct Target of miR-214 in CRC Cells

It is well known that miRNAs function by regulating target genes. In the present study, the putative miR-214 target genes were predicted using bioinformatic algorithms such as miRanda, PicTar, and TargetScan. The results showed that FGFR1 possesses potential complementary sites for the miR-214 seed region in the 3′UTR (Fig. 4A). To explore whether FGFR1 expression is negatively regulated by miR-214 in the CRC cellular environment, HCT116 and LoVo cells were transfected with miR-214 mimics, while HT-29 cells were transfected with an miR-214 inhibitor and the FGFR1 mRNA and protein levels were measured. The results showed that ectopic miR-214 expression markedly reduced FGFR1 mRNA and protein levels in both HCT116 and LoVo cells (*P = 0.015; Fig. 4B,C), whereas miR-214 knockdown resulted in increased FGFR1 mRNA and protein expression (*P = 0.034; Supporting Fig. 6A,B). To further examine the direct binding and repression effect between miR-214 and FGFR1, a luciferase assay was performed. The FGFR1-wt-3′UTR and its mutant type (FGFR1-mt-3′UTR) were amplified and cloned downstream of a luciferase reporter gene in the pGL3-basic vector. Cotransfection of pcDNAmiR-214 and FGFR1-wt-3′UTR in HEK293a and HCT116 cells markedly reduced luciferase activity compared with the negative control and blank cells (both P < 0.05; Fig. 4D). However, no difference in luciferase activity was observed in cells cotransfected with FGFR1-mt-3′UTR and pcDNAmiR-214 (Fig. 4D).

Figure 6.

FGFR1 overexpression in CRC and its relationship with miR-214 expression. (A) FGFR1 is overexpressed in tumor tissues compared with adjacent normal tissues as well as in liver metastases compared with primary CRC tissues. (B) The FGFR1 protein levels are inversely correlated with miR-214 levels in CRC patients (r = −0.928; P < 0.001).

miR-214 Represses FGFR1 to Suppress CRC Growth and Metastasis

The above results prompted us to investigate whether miR-214 suppresses CRC growth and metastasis through FGFR1 suppression. To this end, we first evaluated whether FGFR1 knockdown could mimic the effect of miR-214 overexpression. HCT116 cells were transfected with si-RNA for FGFR1; western blotting analysis confirmed the reduction of FGFR1 expression (Fig. 5A). As expected, compared with the control group, HCT116 cells transfected with si-FGFR1 displayed significantly suppressed cell invasion capacity, reduced wound healing, and fewer cell colonies (all P < 0.05; Fig. 5B-D), and the effects were similar to that of miR-214 overexpression. To investigate whether ectopic FGFR1 expression could counteract the effects of miR-214 in CRC cells, HCT116 cells stably transduced with miR-214 (HCT116-Lv-miR-214) were transfected with a pcDNA3.1-FGFR1 plasmid that contains the coding sequence but lacks the FGFR1 3′UTR. Interestingly, ectopic FGFR1 expression could at least partially rescue the proliferation, migration, and invasion capacities inhibited by miR-214 (Fig. 5B-D). To further confirm the target relationship of miR-214 and FGFR1 in CRC patients, we examined FGFR1 expression in 99 paired tumor/adjacent nontumor tissues using IHC. The FGFR1 protein levels were significantly higher in tumor tissues compared with adjacent nontumor tissues (Fig. 6A). Likewise, compared with primary CRC tissues, higher FGFR1 expression was observed in liver metastases (Fig. 6A). Moreover, a significant inverse correlation was found between miR-214 levels and FGFR1 protein levels in the CRC cohort (r = −0.928; P < 0.001; Fig. 6B). To explore if FGFR1 can be used as a therapeutic target in CRC, HCT116 cells were treated with AZD4547, a new selective small-molecule FGFR inhibitor. As a result, cell proliferation was significantly inhibited, as demonstrated by 3-(4,5-dimethylthiazol-2-yl)−2,5-diphenyltetrazolium bromide (MTT) and colony formation assays (Supporting Fig. 7A,B), and cell invasion ability was also suppressed (Supporting Fig. 7C).

Discussion

Recent studies have highlighted the role of miRNAs in cancer initiation and progression.[23, 24] It has been revealed that more than 50% of miRNAs are located at tumor-related genomic regions or in fragile sites.[25] Previously, Okamoto et al.[26] identified miR-493 as an inhibitory miRNA for colon cancer liver metastasis in a mouse model. In this study we used a microarray approach to identify miRNAs associated with CRC cancer liver metastasis in patients. Among the 39 miRNAs identified to be associated with CRC liver metastasis using our comprehensive approach, there were several miRNAs such as miR-146a, miR-214, miR-345, miR-143, and the miR-200 family that have been reported to be deregulated in human tumors and associated with cancer metastasis.[27-31] To our knowledge, this is the first study to systematically investigate the expression profile of miRNAs in CRC liver metastasis. In the validated cohort of CRC patients, only miR-214 was confirmed as a negative regulator of CRC liver metastasis and associated with favorable survival in CRC patients. We found that miR-214 was significantly down-regulated in CRC tissues with liver metastasis compared with CRC tissues without liver metastasis. Moreover, miR-214 expression was apparently lower in CRC tissues than in adjacent normal tissues. However, miR-214 levels showed no significant differences between primary tissues and the corresponding matched liver metastatic tissues. Previously, miR-214 has been found to be deregulated in several tumors such as pancreatic cancer, melanoma, and hepatocellular carcinoma.[28, 32, 33] However, the biological role of miR-214 in CRC remains unclear. Importantly, miR-214 was associated with a favorable patient prognosis and a less aggressive tumor phenotype, indicating that miR-214 may serve as a marker to predict recurrence, liver metastasis, and outcome for CRC patients.

Given that miR-214 is down-regulated in CRC patients with liver metastasis, we speculated that the up-regulation of miR-214 may be able to suppress the aggressive tumor phenotype in CRC. The results obtained from in vitro cell proliferation, colony formation, cell migration, and invasion as well as in vivo tumor growth and liver metastasis assays confirmed that miR-214 inhibited CRC tumor growth and metastasis. Interestingly, previous studies have reported that miR-214 could be tumor promoting or suppressive in different tumor types. For example, miR-214 is up-regulated and stimulates tumor progression by targeting TFAP2C in melanoma28; however, miR-214 is down-regulated and inhibits tumor progression and metastasis in other tumors such as cervical cancer, breast cancer, and human hematoma.[34-36] These results indicate that the expression pattern and biological role of miR-214 varies depending on the genetic background. These results highlight the role of miR-214 in CRC liver metastasis, and targeting miR-214 could be a promising therapeutic strategy in CRC patients.

The critical role of miRNAs is to regulate the expression of their target genes through mRNA cleavage and/or by inhibition of translation, depending on the degree of complementarity with the 3′UTR of target genes. Computational algorithms have become a useful tool in predicting miRNA targets, which are mainly based on base pairing between target gene 3′UTRs and miRNAs.[37] To explore the underlying mechanism of miR-214 function in CRC, we searched for direct target genes regulated by miR-214 using bioinformatics analysis. As miR-214 inhibited tumor progression and metastasis, we focused on the genes that could stimulate CRC growth and metastasis. As a result, we identified several putative targets of miR-214 including HDGF, FGFR1, CDK6, RUNX1, FOXO4, and DEK. However, further investigation showed that only FGFR1 was a direct functional target of miR-214 in CRC cells. First, miR-214 overexpression significantly decreased FGFR1 mRNA and protein levels in CRC cells. Second, ectopic miR-214 expression markedly reduced the activity of a luciferase reporter containing the 3′UTR sequence of FGFR1; third, FGFR1 knockdown mimicked the tumor suppressive effect of miR-214 on CRC cells, while reintroduction of FGFR1 partly abolished the tumor suppressive effect of miR-214 on CRC cells. Interestingly, a recent report revealed that miR-214 inhibits tumor metastasis by way of targeting FGFR1 in hepatocellular carcinoma,[33] which is in accord with our results in CRC.

FGFR1 is a member of the FGFR tyrosine kinase family that contains four kinases: FGFR1, FGFR2, FGFR3, and FGFR4.38 FGFR1 has been found to be up-regulated in various tumors and is implicated in tumor metastasis in prostate cancer, nonsmall-cell lung cancer, and breast cancer.[39-42] More important, targeting FGFR1 with small-molecule inhibitors has exhibited antitumor activity for malignant disease in preclinical trials.[42, 43] In CRC, FGFR1 overexpression is associated with CRC liver metastasis,[44] whereas FGFR1 inhibition prevents liver metastasis of colon cancer xenografts by modulating the premetastatic niche.[45] In the present study we found that FGFR1 knockdown mimicked the tumor suppressive effect of miR-214, and the reintroduction of FGFR1 in a cell line with stable miR-214 overexpression abolished the suppressive effect of miR-214. Moreover, FGFR1 overexpression is frequently observed in CRC tissues compared with adjacent normal tissues as well as in liver metastases compared with primary CRC tissues, and an inverse correlation was found between the expression levels of miR-214 and FGFR1 in CRC patients. In addition, cell proliferation and invasion capacity was markedly suppressed upon treatment of AZD4547 in HCT116 cells. These data suggest an important role for FGFR1 in the development of CRC liver metastasis, and the miR-214-FGFR1 axis may shed more light on finding new treatments for CRC liver metastasis.

In conclusion, we identified miR-214 as a novel regulator of CRC liver metastasis; down-regulation of miR-214 leads to up-regulation of its target FGFR1, which in turn leads to CRC proliferation and metastasis. Identification of tumor-specific miRNAs and their targets is essential for understanding the molecular mechanisms of CRC progression and metastasis and is important for designing novel therapeutic targets. miR-214 may be a potential prognostic marker, and the miR-214-FGFR1 axis could be used as a therapeutic target for CRC patients with liver metastasis.

Acknowledgment

We thank Prof. Zhi-zhong Pan and Yun-fei Yuan for providing assistance in obtaining tissue samples and Prof. Tie-bang Kang for technical advice.

Ancillary

Advertisement