Genome-wide screening reveals that miR-195 targets the TNF-α/NF-κB pathway by down-regulating IκB kinase alpha and TAB3 in hepatocellular carcinoma

Authors

  • Jie Ding,

    1. State Key Laboratory of Oncogenes and Related Genes, Shanghai Cancer Institute, Renji Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Shenglin Huang,

    1. State Key Laboratory of Oncogenes and Related Genes, Shanghai Cancer Institute, Renji Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Ying Wang,

    1. State Key Laboratory of Oncogenes and Related Genes, Shanghai Cancer Institute, Renji Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China
    Search for more papers by this author
  • Qi Tian,

    1. State Key Laboratory of Oncogenes and Related Genes, Shanghai Cancer Institute, Renji Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China
    Search for more papers by this author
  • Ruopeng Zha,

    1. State Key Laboratory of Oncogenes and Related Genes, Shanghai Cancer Institute, Renji Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China
    Search for more papers by this author
  • Haibing Shi,

    1. State Key Laboratory of Oncogenes and Related Genes, Shanghai Cancer Institute, Renji Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China
    Search for more papers by this author
  • Qifeng Wang,

    1. State Key Laboratory of Oncogenes and Related Genes, Shanghai Cancer Institute, Renji Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China
    Search for more papers by this author
  • Chao Ge,

    1. State Key Laboratory of Oncogenes and Related Genes, Shanghai Cancer Institute, Renji Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China
    Search for more papers by this author
  • Taoyang Chen,

    1. Qi Dong Liver Cancer Institute, Qi Dong, China
    Search for more papers by this author
  • Yingjun Zhao,

    1. State Key Laboratory of Oncogenes and Related Genes, Shanghai Cancer Institute, Renji Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China
    Search for more papers by this author
  • Linhui Liang,

    1. State Key Laboratory of Oncogenes and Related Genes, Shanghai Cancer Institute, Renji Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China
    Search for more papers by this author
  • Jinjun Li,

    1. State Key Laboratory of Oncogenes and Related Genes, Shanghai Cancer Institute, Renji Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China
    Search for more papers by this author
  • Xianghuo He

    Corresponding author
    1. State Key Laboratory of Oncogenes and Related Genes, Shanghai Cancer Institute, Renji Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China
    • Address reprint requests to: Xianghuo He, State Key Laboratory of Oncogenes and Related Genes, Shanghai Cancer Institute, Renji Hospital, Shanghai Jiao Tong University School of Medicine, No.25/Ln.2200, Xie Tu Road, Shanghai 200032, China. E-mail: xhhe@shsci.org; fax: +86-21-64436539.

    Search for more papers by this author

  • Potential conflict of interest: Nothing to report.

  • Supported by grants from the National Natural Science Foundation of China (91029728, 81101481, 81125016), Shanghai Rising-Star and Young Medical Talent Training Program (11QA1406100, XYQ2011048), Shanghai Municipal Education Commission, and Shanghai Municipal Health Bureau (11SG18, 20114Y159).

Abstract

Nuclear factor kappa B (NF-κB) is an important factor linking inflammation and tumorigenesis. In this study we experimentally demonstrated through a high-throughput luciferase reporter screen that NF-κB signaling can be directly targeted by nearly 29 microRNAs (miRNAs). Many of these miRNAs can directly target NF-κB signaling nodes by binding to their 3′ untranslated region (UTR). miR-195, a member of the miR-15 family, is frequently down-regulated in gastrointestinal cancers, especially in hepatocellular carcinoma (HCC). The expression level of miR-195 is inversely correlated with HCC tumor size. We further show that miR-195 suppresses cancer cell proliferation and migration in vitro and reduces tumorigenicity and metastasis in vivo. Additionally, miR-195 may exert its tumor suppressive function by decreasing the expression of multiple NF-κB downstream effectors by way of the direct targeting of IKKα and TAB3. Conclusion: Multiple miRNAs are involved in the NF-κB signaling pathway and miR-195 plays important inhibitory roles in cancer progression and may be a potential therapeutic target. (Hepatology 2013;58:654–666)

Abbreviations
HBV

hepatitis B virus

HCC

hepatocellular carcinoma

HCV

hepatitis C virus

IKK

IκB kinase

miRNA

microRNA

NF-κB

nuclear factor kappa B

TNF-α

tumor necrosis factor alpha

UTR

untranslated region.

Accumulating evidence has shown that chronic infection and inflammation are important causes of tumorigenesis in humans, accounting for ∼20% of human cancers worldwide.[1] For example, chronic infections with the hepatitis B virus (HBV) and the hepatitis C virus (HCV) are well-recognized risk factors for human hepatocellular carcinoma (HCC).[2] During the past decade, numerous investigations have supported the important role of nuclear factor kappa B (NF-κB) as a link between inflammation and tumorigenesis.[3] Since it was first identified as part of the immune system, NF-κB has become widely accepted as a crucial transcription factor that regulates inflammation, innate and adaptive immunity, cell proliferation, cell differentiation, and apoptosis.[4, 5] The incorrect regulation of NF-κB has been linked to many diseases, including cancer. NF-κB is constitutively active in many tumors[6, 7] and plays an important role in cancer development and progression. Thus, NF-κB is regarded as a potential therapeutic target in human cancers.

NF-κB activity is tightly controlled at multiple levels by positive and negative regulatory elements. In their inactive status, NF-κB family members exist as dimers, with a predominance of p65/p50 heterodimers, and are sequestered in the cytoplasm by members of the I-κB family. The stimulation of cells with tumor necrosis factor alpha (TNF-α) or interleukin (IL)-1 initiates a cascade of signaling events leading to the activation of NF-κB. When TNF-α binds to the TNF receptor (TNFR), IκB kinase (IKK) is activated. Activated IKK phosphorylates the NF-κB-bound IκB protein and targets them for ubiquitin-dependent degradation. Released NF-κB homo- or heterodimers translocate to the nucleus, where they are able to bind promoter and enhancer regions containing κB sites, and then mediate the transcription of related target genes.[8] These NF-κB signaling nodes are essential to the NF-κB signaling pathway and can be regulated by posttranslational modifications such as phosphorylation and acetylation.9-13[9-13] Another emerging concept of NF-κB signaling control is regulation by noncoding RNAs, especially microRNAs (miRNAs).

miRNAs are an evolutionarily conserved class of endogenous, small, noncoding RNAs that negatively regulate gene expression at the posttranscriptional level. miRNAs mainly bind to the 3′ untranslated regions (UTRs) of target messenger RNAs (mRNAs), resulting in mRNA degradation or translation repression. Bioinformatics analysis indicates that >60% of protein-coding genes may be directly targeted by miRNAs. Meanwhile, a single mRNA can be modulated by multiple types of miRNAs. Growing evidence has demonstrated that these small RNAs are involved in almost every aspect of biological processes and human diseases. Previous studies showed that a few miRNAs can directly or indirectly modulate NF-κB signaling.[14, 15] For example, miR-146 down-regulates IRAK1 and TRAF6, reducing the activation of NF-κB in human breast and pancreatic cancers.[7, 16] However, the identity of the miRNAs that target NF-κB signaling remains largely unknown.

In this study we performed a high-throughput luciferase reporter assay to screen for miRNAs that target the NF-κB signaling pathway. Interestingly, ∼29 miRNAs were identified as having the potential to reduce the induction of NF-κB activities upon TNF-α addition. Many of these miRNAs can directly target NF-κB signaling nodes by binding to their 3′ UTRs. miR-195, a member of the miR-15 family, was frequently down-regulated in HCC and suppressed cancer cell growth and metastasis in vitro and in vivo.

Materials and Methods

Cell Culture

HEK 293T, HEK 293, and Huh-7 cells were cultured in Dulbecco's Modified Eagle's Medium (DMEM) with 10% fetal bovine serum (FBS) and antibiotics at 37°C with 5% CO2. AGS and LOVO cells were cultured in F-12K Nutrient Medium with 10% FBS. HCT-116 cells were cultured in McCoy's 5A Medium with 10% FBS. SGC-7901 cells were cultured in RPMI-1640 Medium with 10% FBS. SMMC-7721 cells were cultured in DMEM with 10% newborn calf serum. For the NF-κB stimulation, the cells were treated with 5 ng/mL TNF-α (R&D Systems) for 5 hours before assaying luciferase activity levels.

Luciferase Reporter Constructs and Luciferase Assay

The 3′ UTR sequences of NF-κB signaling pathway-related genes were amplified from the genomic DNA of normal liver tissues and then subcloned directly downstream of the stop codon of luciferase. The primer sequences are shown in Supporting Table S1. Both wild-type and mutant 3′ UTR sequences were confirmed by sequencing. For the luciferase screens, HEK 293T cells were seeded in 96-well plates at a density of 5,000 cells per well. After 24 hours, the cells were transiently transfected with 5 ng of pRL-CMV Renilla luciferase reporter, 30 ng of either p-LUC or p-LUC-IKKα UTR, and 5 pmol of miRNA mimics. After 48 hours the luciferase activity was measured using the dual-luciferase reporter assay system (Promega, Madison, WI).

Oligonucleotide Transfection and Quantitative Real-Time Polymerase Chain Reaction (PCR)

All miRNAs mimics were synthesized by Genepharma (Shanghai, China). The miR-15 family member inhibitors (2'-O-methyl modified), small interfering RNAs (siRNAs) specific for TAB3 and IKKα were synthesized by Ribobio (Guangzhou, China). The sequences are shown in Supporting Table S2. The oligonucleotide transfection was performed using the Lipofectamine RNAiMax reagent (Invitrogen, Carlsbad, CA).

Total RNA was extracted using the TRIzol reagent (Invitrogen). Complementary DNA (cDNA) was synthesized with the PrimeScript RT reagent kit (TaKaRa, Tokyo, Japan) using 500 ng of RNA. The real-time PCR analyses were performed using SYBR Premix Ex Taq (TaKaRa). β-Actin was measured as internal controls. The primers used are displayed in Supporting Table S3. To quantify the expression levels of the mature miR-15 family members, we used TaqMan miRNA assays (Applied Biosystems, Foster City, CA). U6 small nuclear (snRNA) was measured as internal control. The sequences of primers are provided in Supporting Table S4.

To analyze the DNA copy variations for miR-195 locus in HCC samples, the primers were designed to amplify miR-195 locus including pre-miR-195 sequences. The β-actin region was used as internal control. The primers used are displayed in Supporting Table S1. The relative genomic level of tumor tissues was compared with that of normal liver tissues.

Lentivirus Production and Infection

The pri-miR-195 sequence was amplified from normal genomic DNA and cloned into the pWPXL lentiviral vector (a gift from Dr. Didier Trono) to generate pWPXL-miR-195. The primer sequences are shown in Supporting Table S1. The virus particles were harvested 48 hours after HEK 293T cells were transfected with pWPXL-miR-195, the packaging plasmid psPAX2, and the VSV-G envelope plasmid pMD2.G (a gift from Dr. Didier Trono) using the Lipofectamine 2000 reagent (Invitrogen). SMMC-7721, Huh-7, HCT-116, SGC-7901, LOVO, and AGS cells were infected with recombinant lentivirus-transducing units plus 6 μg/mL polybrene (Sigma, St. Louis, MO).

Human Tissues

HCC and adjacent nontumorous liver tissues were collected from the surgical specimen archives of the First Affiliated Hospital of Medical School of Zhejiang University, Guangxi Medical University, and the Qidong Liver Cancer Institute, Jiangsu Province, China. A set of human colorectal cancer (CRC) tissue samples, which included 79 pairs of CRC and the corresponding noncancerous colon tissues, was provided by the Fudan University Shanghai Cancer Center. The 76 gastric cancer (GC) tissues and matched precancerous tissues used for the real-time PCR were obtained from the Fudan University Shanghai Cancer Center. All human materials were obtained with informed consent and approved by the Ethical Review Committee of the World Health Organization of the Collaborating Center for Research in Human Production authorized by the Shanghai Municipal Government. The HCC clinical information was collected from patient records and the details are listed in Table 1 with the correlation of miR-195 expression. miR-195 level was stratified into two groups for comparison according to the down-regulation or not in HCC. The cutoff point of the level of miR-195 is 0.5 (two-fold change of down-regulation). Fisher's exact test was used for analysis of categoric data.

Table 1. Correlation of the Clinicopathological Features With Tumor miR-195 Expression in HCC
VariablemiR-195P
No Down-Regulation, No. of CasesDown-Regulation, No. of Cases
Sex   
Female10130.081
Male39116
Age (years)   
≤5537910.707
>551236
Hepatitis   
Yes18770.0307
No3162
Liver cirrhosis   
Yes5290.1301
No44109
Alpha-fetoprotein   
≤20 ng/mL10400.1312
∼20 ng/mL3673
Tumor size   
≤5 cm20230.0426
>5 cm2055
Vascular invasion   
Yes5120.7899
No4281
Lymphatic metastasis   
Positive080.1065
Negative49110
Metastasis   
Positive030.5568
Negative49116

Electrophoretic Mobility Shift Assay (EMSA)

EMSA was performed using the Gelshift Chemiluminescent EMSA Kit (Active Motif, Carlsbad, CA) according to the manufacturer's instructions. Nuclear proteins from HEK 293T cells were extracted using the Nuclear Extract Kit (Active Motif) according to the manufacturer's instructions.

Western Blot

Proteins were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to nitrocellulose membrane (Bio-Rad, Hercules, CA). The membranes were blocked with 5% nonfat milk and incubated with the appropriate antibody. The antigen-antibody complex was detected with enhanced chemiluminescence reagents (Pierce, Rockford, IL). The antibodies used in this study are shown in Supporting Table S5.

Cell Proliferation and Colony Formation Assays

The cell proliferation assay was measured using the Cell Counting Kit-8 (CCK-8) (Dojindo, Japan) on days 1, 3, and 5 after Lenti-miR-195 or mock infection. The cells were seeded into each well of a 96-well plate, and 10 μL of CCK-8 was added to 90 μL of the culture medium at the indicated time. The cells were subsequently incubated at 37°C for 2 hours, and the optical density was measured at 450 nm. For the colony formation assay, 500 or 1,000 cells were plated into each well of a 6-well plate and incubated at 37°C for 2 weeks. The colonies were fixed and stained in a dye solution containing 0.1% crystal violet and 20% methanol, and the number of colonies was counted. These assays were replicated three times.

In Vivo Assays for Tumor Formation

For the in vivo tumor formation assays, 2 × 106 SMMC-7721 or HCT-116 cells infected with the Lenti-miR-195 or mock vector were suspended in 200 μL serum-free DMEM and subcutaneously injected into the flank of each nude mouse (female BALB/c-nu/nu, 8 per group). The tumor sizes were measured twice a week as soon as the tumors were measurable and the tumor volumes were calculated: V (cm3) = width2 (cm2) × length (cm) / 2. On day 49 the mice were sacrificed. The mice were manipulated and housed according to protocols approved by the Shanghai Medical Experimental Animal Care Commission.

In Vivo Migration and Invasion Assays

For the transwell migration assays, 5 × 104 cells were plated in the top chamber of each insert (BD Biosciences, Franklin Lakes, NJ) with a noncoated membrane. For the invasion assays, 1 × 105 cells were placed into the upper chamber with a Matrigel-coated membrane (BD Biosciences). After several hours of incubation at 37°C, cells that migrated or invaded were fixed and stained in dye solution containing 0.1% crystal violet and 20% methanol. The number of cells that had migrated or invaded was counted and imaged using an IX71 inverted microscope (Olympus, Tokyo, Japan).

In Vivo Assays for Metastasis

For the metastasis assays, 2 × 106 SMMC-7721 or HCT-116 cells infected with the Lenti-miR-195 or mock vector were suspended in 200 μL of serum-free DMEM and subcutaneously injected into nude mice through the tail vein (female BALB/c-nu/nu mice, 6 per group). After 8 weeks the mice were sacrificed and their lungs were dissected, fixed with phosphate-buffered neutral formalin, and prepared for standard histological examination. The mice were manipulated and housed according to protocols approved by the Shanghai Medical Experimental Animal Care Commission.

Statistical Analysis

The results are presented as the means ± standard error of the mean (SEM). The data were subjected to Student t test (two-tailed, with P < 0.05 considered significant) unless otherwise specified (paired t test, χ2 test, or Pearson's correlation).

Results

Screening for Candidate miRNAs That Modulate the NF-κB Signaling Pathway

To screen for miRNAs that modulate the NF-κB signaling pathway, we utilized HEK 293 cells stably expressing a luciferase reporter with two inducible NF-κB binding sites upstream of the luciferase gene (denoted 293-NFκB-luc) (Fig. 1A). In total, 870 human miRNAs (miRBase 15.0) were investigated. These miRNAs and control RNAs were individually transfected into the 293-NFκB-luc cells. Luciferase activity was assayed 48 hours after transfection and stimulation with TNF-α for 5 hours. We observed 29 miRNAs that decreased the induction of luciferase following TNF-α addition in the 293-NFκB-luc cells (0.7 as the cutoff, Fig. 1B and Table 2). Notably, these miRNAs included miR-146a, miR-29, and miR-155, which were previously identified to target NF-κB signaling.17-19[17-19]

Figure 1.

Screening for candidate miRNAs that modulate NF-κB signaling. (A) Schematic diagram of the screening strategy. 293-NFκB-luc cells were plated into each well of a 96-well plate and transfected with individual miRNA mimics. Luciferase activity was assayed 48 hours after transfection and stimulation with TNF-α for 5 hours. (B) Results of luciferase screen. The ratio of luciferase activity between individual miRNA and control RNA was plotted. Based on the known miRNAs that were previously identified to target NF-κB pathway (miR-146a, 0.66; miR-29, 0.61; miR-155, 0.68), we chose 0.7 as the cutoff and totally identified 29 miRNAs (show in red) that suppress the NF-κB pathway. (C) A schematic diagram of the TargetScan prediction binding sites of the IKKα and TAB3-regulating miRNAs in the wild-type IKKα and TAB3 3′ UTR (top). Luciferase activity assays of the full-length wild-type IKKα and TAB3 3′ UTRs luciferase reporters (bottom). The luciferase activity was normalized to the Renilla luciferase activity in each sample. A representative experiment in triplicate (means ± SEM) is shown. *P < 0.05 and **P < 0.01 by Student t test. (D) Panoramic view of miRNAs and their potential target genes in classical and nonclassical NF-κB pathway after dual luciferase activity assays. Notably, miR-297 targeting TAB3 does not show, as we found no change in TAB3 expression level upon miR-297 addiction.

Table 2. miRNAs That Modulate NF-κB Signaling and Their Potential Targets
IDChromosomeLuciferase Activity (Ratio)Potential Target Genes
hsa-miR-296-3p20q13.320.61 
hsa-miR-12621p31.30.66 
hsa-miR-499-3p20q11.220.65TRAF5,TAB3
hsa-miR-302c4q250.66 
hsa-miR-520f19q13.420.63TAB3
hsa-miR-28-5p3q280.69RIPK1,IKKβ
hsa-miR-146a5q340.66 
hsa-miR-127-5p14q32.20.68 
hsa-miR-129-3p7q32.1/11p11.20.63TAK1
hsa-miR-13217p13.30.68IKKβ
hsa-miR-1941q41/11q13.10.65 
hsa-miR-1282q21.3/3p22.30.60TAB3
hsa-miR-128920q11.22/5q31.10.66 
hsa-miR-15521q21.30.68TAB2
hsa-miR-15a/15b/16/195/49713q14.2/3q25.33/17p13.10.68IKKα, TAB3
hsa-miR-199a-3p19p13.2/1q24.30.67TAB3
hsa-miR-222Xp11.30.66 
hsa-miR-223Xq120.69IKKα
hsa-miR-2974q250.56IKKα, TAK1,TRAF5
hsa-miR-29abc7q32.3/1q32.3/1q32.20.61TNFR1,TRAF5
hsa-miR-319p21.30.59NIK,RIPK1
hsa-miR-503Xq26.30.68IKKβ, IKKγ, TRAF5
hsa-miR-520d-5p19q13.420.56IKKβ, TAB2, TAB3, TRAF5
hsa-miR-64720q13.330.69IKKγ

To explore whether these 29 miRNAs can directly target NF-κB signaling nodes, we first used a miRNA target-prediction tool, TargetScan, to predict 18 NF-κB signaling nodes and their potential to be targeted by these miRNAs. The analysis revealed that most of the genes could be targeted by at least one of these miRNAs (Supporting Information, Table S6). In total, there are 79 target sites among the 3′ UTRs of these 18 genes. To further confirm this result, we constructed various luciferase reporters with the full-length 3′ UTR of each gene in the list and then screened the reporters using a dual-luciferase assay. Representative results are shown in Fig. 1C. We found that ∼28 out of the 79 predicted target sites (35.4%) of the predicted targets were validated by the experimental reporter assays (Fig. 1D and Table 2).

miR-15 Family Targets NF-κB Signaling-Related Genes

Interestingly, the miR-15 family was found to directly target the 3′ UTRs of the IKKα and TAB3 mRNAs. The miR-15 family contains five members (miR-15a, miR-15b, miR-16, miR-195, and miR-497) with the same seed sequence (Fig. 2A). MiR-15a and miR-16 are the first identified miRNAs that serve as tumor suppressors.[20] To determine whether IKKα and TAB3 are regulated by the miR-15 family by way of the direct binding of their 3′ UTRs, we constructed binding site mutant fragments of the IKKα and TAB3 mRNA 3′ UTRs and inserted them immediately downstream of the luciferase reporter gene (Fig. 2B). For the luciferase activity assays, the miR-15a mimic or control RNAs were cotransfected with different 3′ UTR luciferase constructs. The results showed that miR-15a decreased the relative luciferase activity in the cells with the wild-type 3′ UTRs of IKKα and TAB3 (Fig. 2C). Such regulation was sequence-specific because the relative luciferase activity did not drop as sharply in cells containing the 3′ UTRs with mutant binding sites as in those that contained the wild-type counterparts (Fig. 2C). In agreement with these results, we observed that the overexpression of the miR-15 family members caused a clear decrease in the endogenous mRNA and protein levels of the IKKα and TAB3 genes (Fig. 2D,E). Subsequently, the overexpression of miR-15 family members suppressed the activation of p65, whereas its protein levels were not altered (Fig. 2E). To assess NF-κB transcriptional activation, nuclear extracts prepared from cells overexpressing miR-15 family members and control cells were assessed for NF-κB binding to a consensus NF-κB response element by EMSA. We found that nuclear extracts from cells overexpressing miR-15 family members possessed lower NF-κB DNA-binding activation than the negative control nuclear extracts (Fig. 2F). Taken together, these results suggest that miR-15 family members inhibit the expression of IKKα and TAB3 by directly targeting their 3′ UTRs and thus attenuate the activity of NF-κB signaling.

Figure 2.

miR-15 family targets NF-κB signaling pathway-related genes. (A) Sequence alignment of five miR-15 family members. The list of the miR-15 family members shows the 5′AGCAGCAC conserved sequences highlighted in red. (B) The putative miR-15a binding site in the IKKα and TAB3 3′ UTRs. (C) Luciferase activity assays of luciferase reporters with wild-type or mutant IKKα and TAB3 3′ UTRs were performed after the cotransfection with miR-15a mimics or the negative control (NC) in HEK 293T cells. The luciferase activity of each sample was normalized to the Renilla luciferase activity. A representative experiment in triplicate is shown with the means ± SEM. (D) The mRNA levels of IKKα and TAB3 were determined by quantitative real-time PCR analyses after transfection with miR-15 family mimics or the NC in HEK 293T cells. β-Actin served as an internal control. (E) The protein levels of IKKα, TAB3, p65, and p-p65 were determined by western blot analyses after transfection of HEK 293T cells with the mimics of different miR-15 family members or the NC. β-Actin served as an internal control. (F) EMSA was conducted to measure the DNA-binding activity of NF-κB in HEK 293T cells transfected with miR-15 family members or the NC. *P < 0.05, **P < 0.01, and ***P < 0.001 by Student t test.

miR-195 Is Frequently Down-Regulated in Gastrointestinal Cancers and Is Inversely Correlated with HCC Tumor Size

The down-regulation of miR-15 family members has been observed in various types of cancers.21-26[21-26] Here, we analyzed the expression levels of miR-15 family members in gastrointestinal cancers, including 189 HCCs, 76 GCs, 79 CRCs, and their corresponding adjacent noncancerous tissues by real-time PCR. Unexpectedly, we found that only miR-195 expression was significantly down-regulated in all three types of cancer compared with their noncancerous counterparts (Fig. 3A). When comparing primary cancers with their corresponding nontumorous tissues, the down-regulation of miR-195 (greater than a two-fold change) was observed in 60% of HCCs, 41% of GCs, and 43% of CRCs (Fig. 3B), indicating that the down-regulation of miR-195 is common in gastrointestinal cancers, especially in HCC. The analysis of the genomic DNA region of the miR-195 locus revealed that the DNA copy number of the miR-195 region is reduced in ∼35% of HCC tissues (Fig. 3C). Furthermore, we investigated the correlation between the expression of miR-195 and clinical-pathological features of HCC. Down-regulation of the miR-195 level was found to be associated significantly with larger tumor size (>5 cm) (P = 0.0426) and presence of hepatitis (P = 0.0307) (Table 1). The expression level of miR-195 was significantly down-regulated in patients with HCC tumor size >5 cm (Fig. 3D), suggesting that miR-195 may play a role in controlling cell proliferation. The expression levels of other miR-15 family members (miR-15a, miR-15b, miR-16, and miR-497) were not consistently significantly different in these tumor tissues (Supporting Information, Fig. S1).

Figure 3.

miR-195 is frequently down-regulated in gastrointestinal cancer and associated with HCC tumor size. (A) The relative expression level of mature miR-195 was determined between 189 HCCs, 76 GCs, and 79 CRCs and adjacent noncancerous tissues using the ABI TaqMan miRNA assay. U6 snRNA served as an internal control. (B) The down-regulation of miR-195 in paired HCC/nontumorous, CRC/nontumorous, and GC/nontumorous samples was defined as a log two-fold change < -1 (two-fold). (C) Result of genomic real-time PCR screening for the analysis of miR-195 DNA copy numbers in 100 HCC pairs. The analysis was performed using SYBR Premix Ex Taq assays. β-Actin was used for normalization. Blue and red, respectively, indicate the chromosomal regions that are commonly deleted or amplified in HCC. (D) Relative expression of mature miR-195 in 118 HCC tissue samples, with tumor size (>5 cm or <5 cm). The statistical analysis was performed using Student t test (A) and Mann Whitney U test (D).

miR-195 Suppresses Cancer Cell Proliferation and Tumorigenicity In Vitro and In Vivo

To better understand the role of miR-195 in the development of gastrointestinal cancer, we first constructed a lentiviral vector expressing miR-195 and established stable cell lines, denoted SMMC-7721-195, Huh-7-195, AGS-195, SGC-7901-195, LOVO-195, and HCT-116-195, after lentivirus transduction (Supporting Information, Fig. S2). The colony formation and cell proliferation assays revealed that the ectopic expression of miR-195 in these stable cancer cells resulted in a significant reduction in colony number and cell proliferation (Fig. 4A,B; Supporting Information, Fig. S3). In contrast, the silencing of miR-195 by way of the transfection of SMMC-7721, SGC-7901, and LOVO cells with miR-195 inhibitors increased the cell proliferation rates compared with those of the negative control (Fig. 4C). To further examine the effect of miR-195 on the tumorigenicity in vivo, SMCC-7721 and HCT-116 cells stably expressing miR-195 or the vector control were subcutaneously injected into nude mice. Intriguingly, we found no tumors in the mice injected with the HCT-116-195 cells, and significantly reduced tumor growth in the mice with the SMCC-7721 cells (Fig. 4D). Taken together, these findings demonstrated that miR-195 suppresses gastrointestinal cancer cell proliferation in vitro and in vivo.

Figure 4.

miR-195 inhibits cancer cell proliferation in vitro and in vivo. (A) The colony formation assay for Huh-7, AGS, and HCT-116 cells infected with lentivirus expressing miR-195 or the mock control. A total of 500 or 1,000 cells per well were seeded and cultivated for 2 weeks. The colonies were fixed and stained in a dye solution containing 0.1% crystal violet and 20% methanol. (B) Cell proliferation assays for the Huh-7, AGS, and HCT-116 cells infected with lentivirus expressing miR-195 or the mock control were conducted using the cell counting kit-8 (CCK-8) assay. (C) Cell proliferation assays for SMMC-7721, SGC-7901, and LOVO cells transfected with miR-195 inhibitor or NC inhibitor were conducted using the cell counting kit-8 (CCK-8) assay. (B,C) The mean values are plotted as shown and the bars indicate SEM in triplicate. (D) The tumor growth curve is shown. HCT-116 and SMMC-7721 cells transfected with lentivirus expressing miR-195 or the mock control were subcutaneously injected into nude mice. The tumor sizes were measured twice a week as soon as the tumors were measurable. *P < 0.05, **P < 0.01, ***P < 0.001, and P = ns (not significant) by Student t test.

miR-195 Reduces Cancer Cell Invasion and Metastasis In Vitro and In Vivo

miR-195 was recently reported to impair the invasion of breast cancer cells in vitro,[23] thus prompting us to investigate whether miR-195 could also inhibit cell invasion and metastasis in gastrointestinal cancers. Transwell assays without or with Matrigel demonstrated that the overexpression of miR-195 significantly decreased both migration and invasion, respectively, in gastrointestinal cancer cells (Fig. 5A,B; Supporting Information, Fig. S4). In contrast, the silencing of endogenous miR-195 in SMMC-7721, SGC-7901, and LOVO cells promoted cell migration and invasion (Fig. 5C,D). Notably, the incubation times for the migration and invasion assays were 16 hours and 40 hours, respectively, and at those timepoints the cell growth was not significantly affected by miR-195 (Supporting Information, Fig. S5). Thus, the inhibitory effects on cell migration and invasion were not caused by a reduction in cell number. Furthermore, the SMMC-7721-195, HCT-116-195, and control cells were injected into nude mice through the tail vein. Eight weeks later, the lungs were prepared for standard histological examination. The results showed that the metastatic nodules could be found only in the control group (n = 5 out of 6) but not in the miR-195 group (n = 6) in SMMC-7721 cells, and significantly reduced in the mice with HCT-116-195 cells (Fig. 5D). Taken together, these findings indicate that miR-195 inhibits cancer cell migration and invasion in vitro and in vivo.

Figure 5.

miR-195 inhibits cancer cell invasion and metastasis in vitro and in vivo. (A,B) Transwell migration and invasion assays of SMCC-7721, AGS, and HCT-116 cells were performed after transduction with miR-195-expressing or control lentivirus, respectively. The results are representative of at least three independent experiments. (C,D) Transwell migration and invasion assays of SMMC-7721, SGC-7901, and LOVO cells transfected with the miR-195 inhibitor or the negative control. (A-D) The representative images are shown on the left. The statistical analysis was performed using Student t test. (E) Hematoxylin and eosin (H&E)-stained section of metastatic nodules in the lung formed by SMMC-7721-195 or HCT-116-195 and mock control cells at the eighth/seventh week after tail-vein injection. The numbers of metastatic nodules in the lungs of each mouse were counted and analyzed using Student t test. Data are presented as means ± SEM.

miR-195 Suppresses Multiple NF-κB Downstream Effectors by Targeting IKKα and TAB3

To further explore whether miR-195 exerts its tumor suppressive effect through the inhibition of the NF-κB signaling pathway, we investigated the expression levels of a set of NF-κB downstream effectors that are central coordinators of cell proliferation, metastasis, and apoptosis. We observed that the ectopic expression of miR-195 decreased the levels of NF-κB downstream effectors. Conversely, the silencing of miR-195 expression increased the levels of these effectors (Fig. 6A,B). To examine whether miR-195 affects the NF-κB downstream effectors by targeting IKKα and TAB3, we synthesized siRNAs specific for IKKα and TAB3. The siRNAs directed against IKKα or TAB3 sharply decreased IKKα and TAB3 expression (Supporting Information, Fig. S6) and subsequently inhibited the expression of the NF-κB downstream effectors (Fig. 6A,B). When the miR-195 inhibitor was cotransfected with IKKα or TAB3 siRNA into SMMC-7721 cells, we found that the knockdown of IKKα and TAB3 abrogated the effects of the miR-195 inhibitor on the expression of the NF-κB downstream effectors (Fig. 6C). These findings indicate that IKKα and TAB3 are indeed functional targets of miR-195 in the NF-κB signaling pathway.

Figure 6.

miR-195 suppresses multiple NF-κB downstream effectors in the cell. (A,B) The mRNA and protein levels of the NF-κB downstream effectors were determined by quantitative real-time PCR analyses and western blot analyses, respectively, after transfection of SMMC-7721 cells with miR-195 mimics, miR-195 inhibitor, siIKKα, siTAB3, or the NC. β-Actin served as an internal control. *P < 0.05, **P < 0.01, and ***P < 0.001 by Student t test. (C) The protein levels of the NF-κB downstream effectors in SMMC-7721 cells were determined by western blot analyses after transfection with anti-miR-195, siIKKα, siTAB3, or the NC.

Discussion

 In the present study we identified a set of miRNAs that modulates the NF-κB signaling pathway. We further determined that approximately half of these miRNAs can directly target NF-κB signaling nodes by binding to their 3′ UTRs. Other miRNAs may indirectly affect NF-κB signaling by targeting some NF-κB signaling-associated regulators. These findings provide us with abundant information to further decode the functional implications of miRNAs and NF-κB signaling in development and human diseases. Notably, many of these miRNAs have been reported to play important roles in tumor development and progression, such as miR-15, miR-29, miR-31, miR-124, miR-199a, and miR-223.27-32[27-32] Thus, we have found an alternative mechanism by which these miRNAs exert their tumor suppressive functions, by targeting the NF-κB signaling pathway.

Interestingly, we found that the miR-15 family can directly regulate two NF-κB signaling nodes, IKKα and TAB3, by binding to their 3′ UTRs. This combined regulation may facilitate the ability of the miRNA to effectively control this signaling pathway. IKKα is part of the IκB kinase complex, which regulates NF-κB release and translocation to the nucleus. Consistent with our result, IKKα has also been identified to be targeted by miR-15a, miR-16, and miR-223 during macrophage differentiation.[20] TAB3, transforming growth factor β-activated kinase (TAK1)-binding protein 3, is involved in the activation of NF-κB. So far, three mammalian TABs have been identified, which have been named TAB1, TAB2, and TAB3. TAB3, a scaffold protein, acts as an adaptor that links TAK1 with TRAF2 or TRAF6 and mediates TAK1 activation in a TNF-α-dependent and IL-1-dependent manner, respectively. TAB3 is widely expressed and constitutively overexpressed in certain tumor tissues, such as skin, testis, and small intestinal cancers.[33, 34] Previous findings have indicated that TAB3 is posttranslationally regulated through the ubiquitination pathway or phosphorylated at Ser-506 by MAPKAPK2 and MAPKAPK3.[33, 35] Here, we have shown that the miR-15 family members could bind directly to the 3′ UTR of TAB3 mRNA and decrease the expression level of TAB3, providing new evidence in support of a mechanism for TAB3 regulation at the posttranscriptional level by miRNAs.

miR-15 family members are among a subset of human-expressed miRNAs that have been implicated in many human cancers. To date, five members of the miR-15 family have been identified in humans: miR-15a, miR-15b, miR-16, miR-195, and miR-497. We investigated the expression levels of these miRNAs in three types of gastrointestinal cancers: HCC, CRC, and GC. Unexpectedly, we found that only miR-195 expression was significantly down-regulated in all three cancers. Approximately 60% of HCCs down-regulate miR-195 at least two-fold compared with their noncancerous tissue counterparts. miR-195 is a highly conserved miRNA located on chromosome 17p13.1, which is a common, recurrent deletion region in some types of human cancer, especially HCC.36-39[36-39] Our results showed that the miR-195 genomic DNA regions are lost in ∼35% of HCC tissues (Fig. 3C), indicating that genomic deletion may result in the down-regulation of miR-195 in HCC. Notably, we did not find a good correlation of RNA and DNA level of miR-195 in our examined samples (r = 0.14, P = 0.24) when the correlation of RNA and DNA levels of miR-195 were analyzed. We reasoned that other factors might also result in the down-regulation of miR-195. Supporting with this notion, a recent study found that DNA hypermethylation was present upstream of miR-195 DNA regions in breast cancer,[23] suggesting epigenetic alteration may also be responsible for the reduced miR-195 expression in human cancers.

In a previous study, Xu et al.[22] described the tumor suppressive properties of miR-195 in HCC proliferation and tumorigenesis. We further showed that miR-195 not only can reduce cancer cell proliferation and the tumorigenicity, but can also suppress migration invasion and metastasis in vitro and in vivo. Subsequently, miR-195 may exert a tumor suppressive function by decreasing the expression of multiple NF-κB downstream effectors by way of the direct targeting of IKKα and TAB3. Notably, previous studies have identified some miR-195 target genes, such as Bcl-2, Raf-1, CCND1, CCND3, CDK6, and E2F3, most of which are regulated by NF-κB.[22, 23, 40, 41] In such cases, miR-195 may directly target these genes by binding to their 3′ UTRs or indirectly regulate these genes by disrupting NF-κB signaling.

In conclusion, we have experimentally demonstrated that the NF-κB signaling pathway can be modulated by multiple miRNAs, many of which may regulate NF-κB signaling nodes by directly binding to their 3′ UTRs. Interestingly, miR-195, which is frequently down-regulated in gastrointestinal cancers, can directly target IKKα and TAB3 mRNAs and thus affect NF-κB signaling and multiple NF-κB downstream effectors. Furthermore, miR-195 was found to markedly suppress cancer cell proliferation and metastasis. This newly identified miR-195/NF-κB signaling axis provides us with a better understanding of the connection between inflammation, miRNAs, and cancer and may facilitate the development of therapeutics for gastrointestinal cancers, especially HCC.

Acknowledgment

We thank Dr. T. Didier for kindly providing the pWPXL, psPAX2, and pMD2.G lentivirus plasmids.

Ancillary