Potential conflict of interest: Nothing to report.
Supported in part by NIH grants DK088076 and CA086978.
c-Myc is a well-known oncogene frequently up-regulated in different malignancies, whereas liver-specific microRNA (miR)-122, a bona fide tumor suppressor, is down-regulated in hepatocellular cancer (HCC). Here we explored the underlying mechanism of reciprocal regulation of these two genes. Real-time reverse-transcription polymerase chain reaction (RT-PCR) and northern blot analysis demonstrated reduced expression of the primary, precursor, and mature miR-122 in c-MYC-induced HCCs compared to the benign livers, indicating transcriptional suppression of miR-122 upon MYC overexpression. Indeed, chromatin immunoprecipitation (ChIP) assay showed significantly reduced association of RNA polymerase II and histone H3K9Ac, markers of active chromatin, with the miR-122 promoter in tumors relative to the c-MYC-uninduced livers, indicating transcriptional repression of miR-122 in c-MYC-overexpressing tumors. The ChIP assay also demonstrated a significant increase in c-Myc association with the miR-122 promoter region that harbors a conserved noncanonical c-Myc binding site in tumors compared to the livers. Ectopic expression and knockdown studies showed that c-Myc indeed suppresses expression of primary and mature miR-122 in hepatic cells. Additionally, Hnf-3β, a liver enriched transcription factor that activates miR-122 gene, was suppressed in c-MYC-induced tumors. Notably, miR-122 also repressed c-Myc transcription by targeting transcriptional activator E2f1 and coactivator Tfdp2, as evident from ectopic expression and knockdown studies and luciferase reporter assays in mouse and human hepatic cells. Conclusion: c-Myc represses miR-122 gene expression by associating with its promoter and by down-regulating Hnf-3β expression, whereas miR-122 indirectly inhibits c-Myc transcription by targeting Tfdp2 and E2f1. In essence, these results suggest a double-negative feedback loop between a tumor suppressor (miR-122) and an oncogene (c-Myc). (Hepatology 2014;59:555–566)
MicroRNAs (miRNAs) are a class of small (∼22 nt), noncoding RNAs that posttranscriptionally repress target gene expression by pairing with mRNAs of protein coding genes, mainly in the 3′ untranslated regions (UTR).[1, 2] Recent studies have shown that miRNAs may also regulate gene expression through interaction with coding region or 5′-UTR of target genes, as demonstrated by transcriptome-wide identification of miRNA target sites. Over the past few years, many studies have proved that miRNAs play an important physiological role in almost every aspect of biological processes, including development and differentiation, immune response, metabolism, cell proliferation, and apoptosis. Thus, dysregulation of some miRNAs is involved in the pathogenesis of a variety of diseases, such as vascular diseases, immunological diseases, neurological disorders, and cancers. Aberrations in miRNA expression have been attributed to several mechanisms, including amplification, deletion, or mutation of miRNA genes, altered transcriptional regulation of miRNA genes, or epigenetic regulation, such as DNA methylation.
miR-122 is the most abundant liver-specific microRNA that plays fundamental roles in liver. It has been shown to regulate cholesterol metabolism and hepatitis C virus replication. We have previously demonstrated that it is suppressed in diet-induced liver cancer in rat and mouse. Several investigators including us also showed that miR-122 is significantly repressed in hepatocellular carcinoma (HCC) patients and in HCC cell lines.[12, 13] Moreover, some studies have illustrated a correlation between reduced expression of miR-122 and metastasis and poor prognosis, higher tumor burden, and gene expression signature of aggressive tumors in HCC patients.[14, 15] Recently, we generated liver-specific knockout and germline miR-122 knockout mice, which develop steatohepatitis, fibrosis, and HCC with age, further reinforcing the important physiological role and intrinsic tumor suppressor function of miR-122 in liver. Although miR-122 expression is reduced in HCCs, the mechanism of this down-regulation is still unknown. Several studies have reported regulation of miR-122 expression by liver-enriched transcription factors (LETFs) during liver development and hepatocyte differentiation.[17-19] Whether these LETFs are involved in miR-122 suppression in liver cancer remains to be established.
The Myc proto-oncogene encodes c-Myc transcription factor that is frequently up-regulated in a variety of human cancers, including liver cancer. As a transcription factor, c-Myc dimerizes with Max, binds to E boxes in the promoter region of target genes, and activates transcription of target genes involved in cell growth and proliferation. Activation of c-Myc can initiate tumorigenesis as documented in several c-Myc transgenic mouse models.[21, 22] For example, tet-o-MYC; LAP-tTA bi-transgenic mice harboring a tetracycline (tet)-repressible MYC transgene (tet-o-MYC) and a transgene that produces the tet-transactivator protein (tTA) driven by the liver activator promoter (LAP) develop HCC within a few weeks after c-Myc induction.[21, 23] In addition to transactivation of target gene expression, c-Myc is also known to repress some gene expression by mechanisms that may involve interaction with other transcription factors, such as Myc-interacting zinc finger protein 1 (Miz-1). Interestingly, activation of c-Myc results in widespread miRNA repression by directly binding to the miRNA promoter region, which facilitates tumorigenesis. Although c-Myc has been demonstrated to repress the expression of several miRNAs in liver cancer such as miR-100, let-7a, miR-26a, and miR-125b, there is no study investigating whether c-Myc can regulate miR-122 expression, the most abundant and frequently suppressed miRNA in liver cancer.
Previously, we observed dramatic repression of miR-122 in liver tumors formed by induction of c-Myc in tet-o-MYC; LAP-tTA bi-transgenic mice following doxycycline withdrawal, indicating that c-Myc may inhibit miR-122 expression in liver cancer. In contrast, hepatic c-Myc is significantly up-regulated in miR-122 knockout mice, suggesting a double-negative feedback loop between miR-122 and c-Myc. In this study we investigated the possible mechanisms underlying the inverse regulation between miR-122 and c-Myc.
Materials and Methods
miR-122 knockout mice and tet-o-MYC; LAP-tTA bi-transgenic mice were described previously. LAP-tTA mice were obtained from Jackson Laboratory. All animals were housed in a temperature-controlled room under a 12/12-hour light/dark cycle and under pathogen-free conditions. All animal studies were reviewed and approved by the Ohio State University Institutional Laboratory Animal Care and Use Committee.
Hepatocytes were isolated as described previously. Briefly, miR-122 knockout or wild-type mice (20-30 g) were anesthetized with ketamine and xylazine injected intraperitoneally. Livers were perfused with 25 mL perfusion buffer (5 mL/min) and then with 50 mL of warm (37°C) liver digestion buffer by way of the portal vein. The livers were aseptically removed to a sterile Petri dish containing Dulbecco's modified Eagle's medium (DMEM) at 4°C to stop digestion. The hepatocytes were released by peeling off the hepatic capsule and dispersed by shaking the digested liver in DMEM medium at 4°C, followed by passing through a 70-μm strainer and collected by centrifugation at 50g at 4°C. The cells were resuspended in William E medium supplemented with 10% serum. The hepatocytes were counted and viability was determined by Trypan blue dye exclusion. The cells were plated on 6-well plates coated with rat tail type I collagen (BD Bioscience) at a density of 1 × 106 cells per well. The next day cells were transfected as described below.
Cell Lines and Transfection
Hepa (mouse hepatoma) and human HCC (Huh-7, Hep3B, PLC/PRF5) cell lines were obtained from the American Tissue Culture Collection (ATCC) and cultured as recommended by the supplier. For miR-122 overexpression and knockdown experiments, these cells and mouse hepatocytes were transfected with miR-122 or control miR (NC) mimic (50 nM) (Thermo Scientific), and anti-miR-122 or anti-miR control (anti-NC) (100 nM) (Thermo Scientific), respectively, using Lipofectamine 2000 (Invitrogen) following the manufacturer's protocol. For gene knockdown experiments, Hepa cells and mouse hepatocytes were transfected with control (siNC: D-001206-13-05, Thermo Scientific) or gene-specific small interfering RNA (siRNA) (60 nM) (si-Myc: M-040813-02, si-Tfdp2: M-057863-01, si-E2f1: M-044993-03, Thermo Scientific).
RNA Isolation and Northern Blot Analysis
The total liver and tumor RNA was isolated using TRIzol and subjected to northern blot analysis using 32P-labeled anti-miR-122 or anti-5S-rRNA oligo as described.
The TaqMan miRNA Assay (Applied Biosystems) was used to quantify mature and primary miR-122 expression in total liver and tumor RNA according to the manufacturer's instructions. Normalization was performed with RNU6B and 18S rRNA. For gene expression assay, DNase I treated total RNA was reverse-transcribed into complementary DNA (cDNA) using a high-capacity cDNA reverse transcription kit (Applied Biosystems) and real-time PCR was performed using SYBR Green chemistry. The expression was normalized to that of glyseraldehyde-3-phosphate dehydrogenase (Gapdh). All real-time reactions, including controls without cDNA, were run in triplicate in a thermocycler. Relative expression was calculated using the comparative CT method. Primer sequences are provided in the Supporting Information.
Western Blot Analysis
Whole cell or tissue extracts were prepared in sodium dodecyl sulfate (SDS) lysis buffer followed by immunoblotting with specific antibodies. The signal was developed with ECL reagent (ThermoFisher) after incubation with appropriate secondary antibodies. Western blot signals were quantified by ImageJ software (NIH) following the online manual. The antibodies used were: c-Myc: sc-40X, Tfdp2: sc-1209, E2f1: sc-193, Iqgap1: sc-10792, Mapre1: sc-15347, Hnf-1: sc-6548), Hnf-1β: sc-22840, Hnf-3α: sc-22841, Hnf-3β: sc-6554, Hnf-4α: sc-8987, Hnf6: sc-13050, Pkm2: cs-3198, histone H3: ab1791 and c-MYC (human specific): cs-5605.
Plasmids Construction and Luciferase Assay
Wild-type (WT) or mutant 3′-UTR (deletion of miR-122 targeting sites) of Tfdp2 and E2f1 were cloned into psiCHECK2 (Promega) luciferase reporter vector downstream of the Renilla luciferase coding region. For luciferase assay, HCC cells were cotransfected with 200 ng of each luciferase reporter construct and miR-122 or negative control (NC) mimic (50 nM) (ThermoScientific). After 48 hours, luciferase activity was measured using the Dual-Luciferase Reporter Assay kit (Promega) and Renilla luciferase activity was normalized to that of firefly luciferase.
Chromatin Immunoprecipitation (ChIP) Assay
ChIP assay in c-Myc-induced tumors and control livers was performed as described. Briefly, liver and tumor tissues were homogenized in 10 volumes of 2 mM DSG (0.5 M stock in DMSO) in phosphate-buffered saline (PBS), incubated at room temperature for 10 minutes, and filtered through a ∼70 μm cell strainer. The cells were resuspended in 1% formaldehyde in PBS and cross-linked at room temperature for 10 minutes, which was stopped by adding glycine (0.125 M). The cells were washed twice with cold PBS and resuspended in ice-cold cell lysis buffer (150 mM NaCl, 50 mM Tris.HCl pH 7.5, 5 mM EDTA, 0.5% NP-40, 1% Triton X-100, 1× complete EDTA free protease inhibitor (Roche, 25× stock)). The nuclei were washed with cell lysis buffer and resuspended in nuclear lysis buffer (1% SDS, 5 mM EDTA, 50 mM Tris.HCl pH 8.1, protease inhibitor). The chromatin was sonicated to 300-500 bp, followed by standard ChIP analysis with the following antibodies, c-Myc: sc-40x, Pol II (sc-899X), AcH3K9 (cs-9649), and H3Me3K9 (ab8898). Primer sequences are provided in the Supporting Information.
qRT-PCR and transfection analysis was performed in triplicate. The data are presented as means ± standard deviation (SD). Most of the experiments were repeated twice. Statistical significance was calculated with a Student t test with P < 0.05 considered significant. In the figures, P values ≤0.05 and ≤0.01 are represented as * and **, respectively.
miR-122 Expression Was Suppressed, Whereas Some of Its Targets Were Up-Regulated in c-MYC-Induced Liver Tumors
Bi-transgenic Tet-o-MYC; LAP-tTA mice develop tumors after withdrawal of doxycycline from the diet that initiates liver tumorigenesis due to induction of human MYC transgene, and all liver lobes were progressively transformed with numerous tumors within a few weeks (Supporting Fig. 1). To determine whether miR-122 is down-regulated upon c-MYC induction, tumors and benign livers were removed at early stages of tumor development after withdrawal of doxycycline after 3 weeks and its level was analyzed by northern blotting, which exhibited a dramatic decrease in the miR-122 level in tumors but not in benign liver tissues compared to c-MYC-uninduced (Myc off) and the parental mouse (FVBN) livers (Fig. 1A). Since the miR-122 level in the benign livers was not altered, we measured ectopic MYC level in these tissues by qRT-PCR and immunoblotting with an antibody that specifically detects human c-MYC protein. The results showed that c-MYC RNA level was up-regulated ∼8-fold, whereas the protein level was increased more than 12-fold in tumors compared to the benign livers (Fig. 1B), suggesting stabilization of the protein in tumors. These results indicate that miR-122 expression is specifically suppressed in tumors expressing high levels of c-Myc (Fig. 1A). Precursor miR-122 (pre-miR-122), which could be detected after longer exposure of the northern blot because of its very short half-life, was detectable in benign livers but was barely detectable in tumors (Fig. 1A), suggesting that the Dicer-mediated processing of pre-miR-122 does not play any major role in reducing mature miR-122 level in tumors. qRT-PCR analysis showed a profound decrease (∼70%-90%) in both primary and mature miR-122 in tumors compared to livers (Fig. 1C), indicating that miR-122 was down-regulated primarily due to transcriptional repression upon c-MYC induction. Although a decrease in mature miR was more than that of pri-miR-122, it was not significant.
To investigate the consequence of miR-122 suppression in c-Myc tumors, we searched the microarray data available from the GEO database (GSE28198) for miR-122 targets, which showed up-regulation of several known targets such as Adam10, Agpat1, Ccng1, Ndrg3, Aldoa, Iqgap1, and Mapre1. Among these, Iqgap1 and Mapre1 were significantly up-regulated at the RNA and protein levels in the tumors compared to benign liver tissues (Fig. 1D,E). Notably, Pkm2, a predicted target of miR-122 up-regulated in many cancers including HCC, was also highly elevated in these tumors both at the RNA and protein level (Fig. 1D,E). To substantiate that the mouse Pkm2 is a direct miR-122 target, we performed a luciferase reporter assay. This study demonstrated that ectopic miR-122 inhibited mouse Pkm2 3′-UTR driven luciferase activity by 60%, which could be reversed by deletion of miR-122 seed match from the 3′-UTR (Fig. 1F). Collectively, these data showed that miR-122 is transcriptionally suppressed in c-Myc-induced tumors correlating with up-regulation of its selected targets.
c-Myc Binds Directly to miR-122 Immediate Upstream Promoter Region and Also Suppresses Hnf-3β Level
Next we sought to elucidate the mechanism by which miR-122 was repressed in c-Myc-induced tumors. To this end, we first examined if c-Myc could inhibit miR-122 expression in vitro. Indeed, overexpression of c-Myc in mouse Hepa cells resulted in a ∼60% decrease in miR-122 level compared to vector transfected cells (Fig. 2A). In contrast, siRNA-mediated depletion of c-Myc significantly increased both mature and primary miR-122 expression in these cells (Fig. 2B). These results suggested that c-Myc could negatively regulate miR-122 expression at the transcriptional level.
To gain further insight into the mechanism of suppression, we investigated if c-Myc inhibits miR-122 expression by directly interacting with its promoter. Indeed, searching the miR-122 promoter region for evolutionarily conserved transcription factor binding sites using the rVista program (http://rvista.dcode.org) identified one conserved noncanonical c-Myc binding site in its immediate upstream promoter and two downstream of +1 site (Fig. 2C). To test if c-Myc could bind to this region, ChIP analysis was performed in c-Myc-induced tumor and control liver tissues using an antibody that precipitates both human and mouse c-Myc (see Materials and Methods for details). The results showed that the association of c-Myc at miR-122 promoter region was ∼2.7-fold higher (P = 0.02) in tumors compared to control livers (Fig. 2D). Moreover, the association of RNA polymerase II (Pol II) and histone H3K9Ac, a marker of activate chromatin, was reduced significantly in tumors compared to that of control livers, correlating with reduced transcription of primary miR-122 in c-Myc tumors (Fig. 1C).
Previous studies have shown that the liver enriched transcription factors (LETFs) Hnf-1α, Hnf-3α, Hnf-3β,[15, 18] Hnf-4α, Hnf6 as well as C/ebpα are involved in transcriptional activation of miR-122 during liver development and hepatocyte differentiation. It is possible that c-Myc may suppress miR-122 expression indirectly by inhibiting the expression of these LETFs. To test this possibility, we measured the expression of these transcription factors in c-Myc-induced tumors. Western blot analysis showed that among these transcription factors expression of only Hnf-3β was consistently reduced by at least 50% in both the tumors relative to those in the respective adjacent benign livers (Fig. 2E), suggesting that Hnf-3β down-regulation may contribute to suppression of miR-122 in tumors.
Down-Regulation of c-Myc Expression by miR-122
A key observation we made while characterizing miR-122 knockout mice is the up-regulation of c-Myc in their livers (Fig. 3A). Since c-Myc is an oncogene with multiple functions, it was of interest to elucidate how miR-122 modulates its expression. For this purpose, we transiently overexpressed miR-122 in Hepa cells that express miR-122 but at a significantly reduced level relative to the WT (miR-122+/+) livers (data not shown), and miR-122−/− (KO) hepatocytes by transfecting miR-122 mimic, and measured c-Myc expression. qRT-PCR analysis showed that the c-Myc mRNA level was reduced by ∼20% and ∼40% in miR-122 mimic transfected Hepa and KO hepatocytes, respectively (Fig. 3B). In contrast, the c-Myc mRNA level was significantly elevated in Hepa cells and WT hepatocytes depleted of miR-122 by transfecting anti-miR-122 oligonucleotide (Fig. 3B). Similarly, overexpression of miR-122 in KO hepatocytes reduced the c-Myc protein level (∼45%), while its depletion in WT hepatocytes increased the c-Myc protein level (∼2-fold) (Fig. 3C), suggesting that miR-122 negatively regulates c-Myc expression in vitro. To examine if c-Myc is a direct target of miR-122, we searched for potential miR-122 binding sites in c-Myc 3′-UTR. However, we could not find any miR-122 cognate site in its 3′-UTR searching different databases, indicating that miR-122 may repress c-Myc expression by an alternative mechanism.
miR-122 Suppressed c-Myc Expression Through Its Targets Tfdp2 and E2f1
To gain insight into the molecular mechanism of c-Myc repression by miR-122, we first determined whether c-Myc was up-regulated in miR-122KO liver at the transcriptional or posttranscriptional level. qRT-PCR analysis showed that hepatic c-Myc mRNA and hnRNA (primary transcript) levels were augmented at a comparable level in 5-week-old KO mice compared to those in the control (miR-122fl/fl) mice (Fig. 4A), suggesting its transcriptional regulation by miR-122. Previous studies have shown that the E2f family of transcription factors together with transcription factor dimerization partner (TFDP) form a complex and transactivate c-Myc promoter through E2f binding sites in the promoter region of c-Myc.[31, 32] Tfdp2, a member of the TFDP family, is predicted as a conserved miR-122 target by several databases including TargetScan (Fig. 4B). Interestingly, RNA22 program also predicted one miR-122 targeting site in E2f1 3′-UTR (Fig. 4B). Based on these observations, we hypothesized that miR-122 may negatively regulate c-Myc gene expression through targeting Tfdp2 and E2f1. To this end, we first examined Tfdp2 and E2f1 expression in miR-122KO livers. qRT-PCR revealed a ∼1.65- and ∼2.35-fold rise in Tfdp2 and E2f1 mRNA level in KO livers compared to controls (Fig. 4C). Similarly, their protein levels were increased in KO livers compared to controls (Fig. 4D). Furthermore, in miR-122 mimic transfected Hepa cells and KO hepatocytes, Tfdp2 and E2f1 the mRNA level was reduced by ∼30% and ∼40%, respectively (Fig. 4E). In contrast, up-regulation of their expression in miR-122 depleted Hepa cells and WT hepatocytes was more robust and significant (Fig. 4E). Western blot analysis confirmed that overexpression of miR-122 in KO hepatocytes decreased Tfdp2 and E2f1 protein levels by ∼50%, whereas knockdown miR-122 in WT hepatocytes increased the respective protein levels by ∼1.3- and ∼2.2-fold (Fig. 4F). Taken together, these data imply that miR-122 negatively regulates Tfdp2 and E2f1 expression.
To determine further if Tfdp2 and E2f1 are bona fide targets of miR-122, we cloned their WT and mutant (deleted of miR-122 targeting sites) 3′-UTR into psiCHECK2 dual luciferase vector downstream of Renilla cDNA and performed a reporter assay. The relative luciferase activities (Renilla to firefly) were significantly repressed by miR-122 mimic compared to negative control RNA in both WT 3′-UTR constructs (E2f1 and Tfdp2), while these activities were not affected in the mutant 3′-UTR transfected cells (Fig. 4G), suggesting that miR-122 directly represses Tfdp2 and E2f1 expression by interacting with their 3′-UTRs.
Finally, we explored the possibility that Tfdp2 and E2f1 are involved in the up-regulation of hepatic c-Myc expression in KO mice by siRNA-mediated knockdown of Tfdp2 and E2f1 in KO hepatocytes followed by measurements of c-Myc mRNA and protein levels. The results showed that indeed c-Myc was down-regulated upon depletion of Tfdp2 or E2f1 (Fig. 5A-C), suggesting that they are involved in transcription of c-Myc in KO mouse livers. As expected, both Tfdp2 and E2f1 were up-regulated in c-Myc-induced tumors compared to benign livers (Fig. 5D). Extension of this study to human hepatic (Huh-7) cells depleted of endogenous miR-122 by transfecting anti-miR-122 oligo (Fig. 5E) showed elevated expression of c-MYC and TFDP2 both at the protein (Fig. 5F) and RNA (Fig. 5G) levels. In contrast, ectopic expression of miR-122 in nonexpressing Hep3B cells resulted in down-regulation of these transcription factors (Fig. 5E-G). Interestingly, PLC/PRF5 cells ectopic expression of miR-122 resulted in down-regulation of all three factors at the protein level (Fig. 5E,F).
Based on all these data we propose a model delineating reciprocal regulation of miR-122 and c-Myc in the liver and tumor (Fig. 6). Intriguingly, the regulation of both appears to occur at the transcriptional level; c-Myc overexpression suppresses miR-122 by directly binding to its promoter and, indirectly, by down-regulating liver-enriched transcription factor, Hnf-3β. In contrast, miR-122 indirectly suppresses c-Myc expression by targeting the activator (E2f1) and coactivator (Tfdp2) of c-Myc gene.
It is now well established that miR-122 functions as a tumor suppressor in the liver and its down-regulation is a characteristic of HCCs with poor prognosis. However, the mechanism of its suppression in hepatocellular cancer is not well understood. In the present study, we discovered a novel reciprocal regulation between miR-122 and the pleiotropic oncogene c-Myc in HCC. This study originated from two key observations made while characterizing miR-122KO mice generated in our laboratory. First, hepatic c-Myc was consistently elevated in young adult miR-122KO mice, which was further up-regulated in spontaneous HCCs developed in these mice. Second, while searching for a mouse model of liver cancer to test the therapeutic efficacy of miR-122, we found that c-MYC-induced liver tumors exhibited a dramatic decrease in miR-122 expression. The present study showed that c-MYC induction causes suppression of miR-122 gene transcription in liver tumors that correlated with association of c-MYC with miR-122 immediate upstream promoter region that harbors a conserved noncanonical c-Myc cognate site predicted by rVista, and also by down-regulating the Hnf3β protein level that activates miR-122 expression.[15, 18] Although repression occurs predominantly at the pre-miRNA transcript level, regulation can also occur at the level of mature miR-122 stability in some tumors since it has been reported that addition of a single A residue at the 3′-end of miR-122 by a noncanonical poly(A) polymerase stabilizes miR-122. It would be of interest to explore if there is any regulation at the level of precursor processing or mature miR-122 stability. Notably, several miR-122 targets such as Iqgap1, Mapre1, and Pkm2, with a demonstrated role in oncogenesis, were up-regulated in c-MYC-induced liver tumor. Thus, transcriptional suppression of miR-122 is likely to be one of the mechanisms by which induction of c-MYC promotes HCC development.
c-Myc has been shown to both positively and negatively regulate microRNA expression. For example, c-Myc directly activates transcription of the polycistronic miR-17-92 cluster by interacting with canonical E box located in the promoter region. In contrast, Myc represses expression of several miR genes miR-15a/16-1, miR-26a, miR-29 family members, and miR-34a. Interestingly, these down-regulated miRs exhibit tumor suppressor function. Although the mechanisms underlying Myc-mediated transcriptional activation of target genes is well studied, that of transcriptional repression mediated by c-Myc is poorly understood. c-Myc interacts with its dimerization partner, Max, to promote gene activation upon occupancy of E box. In contrast, interaction of c-Myc with Miz-1 represses expression of certain genes due to displacement of the coactivator p300. It would be of interest to investigate how c-Myc inhibits expression of miR genes including miR-122. Notably, c-Myc represses expression of certain miRs such as let-7 variants by interfering with their processing but not transrepression. However, our northern blot and qRT-PCR data suggest that primarily transcription of miR-122 is affected in c-Myc-induced liver tumors.
c-Myc is a well-known oncogene overexpressed in various neoplasms, including hepatocellular cancer.[36, 37] Amplification of the c-Myc genomic locus and overexpression of c-Myc have been observed in ∼70% of HCV and alcohol-induced HCCs, as well as in animal models of HCC. Additionally, c-Myc expression can be directly regulated at the posttranscriptional level by microRNAs, such as miR-34, let-7, miR-145. Our study revealed that the suppression of miR-122 in c-MYC-induced liver tumors correlated with direct association with the promoter and down-regulation of LETF (HNF3β) (Fig. 6), whereas miR-122 can indirectly repress c-Myc expression by targeting E2f1 (in mouse) and its dimerization partner Tfdp2 (both in mouse and human). Thus, down-regulation of miR-122 could be a critical mechanism of up-regulation of c-Myc in HCC. Interaction of Tfdp2 with E2f1 has been shown to enhance both the DNA binding activity and the transactivation function of the heterodimer. Indeed, knockdown of E2f1 and Tfdp2 exhibited additive effects on c-Myc induction in Hepa cells.
Both miR-122[16, 40] and c-Myc[29, 41] are major players in hepatic metabolism. Recently, it has been shown that energy metabolism is profoundly altered in c-Myc-induced liver tumors compared to benign liver tissues. Since miR-122 gene delivery inhibited c-Myc-induced liver tumor growth in mice, it would be of interest to determine whether miR-122 can reverse the metabolic profile of these tumors and the metabolic pathways affected in this process.
We thank Drs. Michael Bishop, Gustavo Leone, and Addgene for providing TET-o-Myc mice and c-Myc expression vector, respectively, and Julia Shreve and Corie Klepper for technical assistance.