SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Transcription factor 1 (Tcf1; hepatocyte nuclear factor 1α [HNF1α]) is critical for hepatocyte development and function. Whether Tcf1 also regulates hepatic microRNAs (miRNAs) has not been investigated yet. Here we analyzed Tcf1-dependent miRNA expression in adult mice in which this transcription factor had been genetically deleted (Tcf1−/−) using miRNA microarray analysis. The miR-192/-194 cluster was markedly down-regulated in liver of Tcf1−/− mice. MiR-192/-194 levels were also decreased in two other tissues that express Tcf1, kidney and small intestine, although to a lesser extent than in liver. In order to identify targets of miR-192/-194 in vivo we combined Affymetrix gene analysis of liver in which miR-192/-194 had been silenced or overexpressed, respectively, and tested regulated messenger RNAs (mRNAs) with multiple binding sites for these miRNAs. This approach revealed frizzled-6 (Fzd6) as a robust endogenous target of miR-194. MiR-194 also targets human FZD6 and expression of miR-194 and Fzd6 are inversely correlated in a mouse model of hepatocellular carcinoma (Dgcr8flox/flox p53flox/flox × Alb-Cre). Conclusion: Our results support a role of miR-194 in liver tumorigenesis through its endogenous target Fzd6. These results may have important implications for Tcf1-mediated liver proliferation. (HEPATOLOGY 2012;55:98–107)

Transcription factor 1 (Tcf1; hepatocyte nuclear factor 1α [HNF1α]) is a homeodomain protein that is expressed in several epithelial organs such as liver, intestine, kidney, and pancreas. In liver, Tcf1 plays an important role as demonstrated by occupancy of >200 promoters in human hepatocytes1 and by the dysregulation of >400 genes in livers of Tcf1-deficient mice.2 Tcf1 regulates many biochemical processes such as bile acid and cholesterol metabolism2 and is critical for hepatocyte proliferation.3

MicroRNAs (miRNAs) are a family of noncoding RNAs, 19-22 nucleotides (nt) in length, that regulate gene expression at the posttranscriptional level.4 Research over the last years has implicated miRNAs in almost every biological process, particularly in growth control and cancer. Recently, several miRNAs have been implicated in hepatocellular carcinoma.5-11 However, evidence for endogenous targets of these miRNAs in vivo is still lacking. Identification of miRNA targets remains a major challenge, as the determinants of miRNA-target interaction are largely unknown beyond the requirement for 6 to 8 nucleotide complementarity at the miRNA 5′ end (the seed motif). Accordingly, it has been estimated that commonly used programs for miRNA target prediction have a >60% false-positive rate.12

There is increasing evidence that transcription factors and miRNAs coordinately regulate gene expression as demonstrated, for example, in tumor suppressive signaling.13 It is reasonable to assume that miRNAs are also involved in regulating gene expression downstream of transcription factors in hepatocytes. Interestingly, Tcf1 has been shown to activate the promoter region of the miR-192/-194-2 gene during intestinal epithelium maturation,14 but the regulation of miRNAs by Tcf1 in hepatocytes in vivo has not been reported yet. In this study we therefore screened for Tcf1-mediated miRNA expression in adult liver. Loss of Tcf1 leads to specific down-regulation of the miR-192/miR-194 cluster. To identify endogenous targets of these miRNAs we used a strategy that combines alteration of miRNA levels in vivo and subsequent analysis of regulated mRNAs with multiple binding sites for the respective miRNAs. We identify frizzled-6 (Fzd6) as a target of miR-194 in vivo and provide evidence for its implication in liver tumorigenesis.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Animals.

All animal models were maintained on a 12-hour light/dark cycle in a pathogen-free animal facility. For antagomir studies, 6 to 12-week-old male C57/Bl6J mice were injected in the tail vein with either phosphate-buffered saline (PBS), antagomirs, Ad-122 (Ctrl),15 or Ad-192/194. Antagomirs (sequences in Supporting Methods) were administered at the indicated doses at 0.2 mL per injection on 3 consecutive days. Mice were injected with adenoviruses at 5 × 109 plaque-forming units (PFU) in 0.2 mL PBS. Mice with Dgcr8/p53 null livers were generated by combining floxed Dgcr8,16 p5317 and one transgenic Alb-Cre allele by appropriate breeding schemes.18 Mice with liver-specific inactivating mutations of DGCR8 and p53 develop hepatocellular carcinoma (HCC) (˜30% at 3 months of age). In 70% of mice a single HCC tumor develops; the remainder (˜30%) develop spontaneous, multifocal HCC that histologically ranges from differentiated to undifferentiated. All tumors are α-fetoprotein (AFP)-positive and this marker can be measured in the serum of 100% of mice that have HCC. Genotyping of transgenic mice was performed on DNA isolated from 3-weeks-old mice by polymerase chain reaction (PCR).19 All animals received humane care according to the criteria outlined in the “Guide for the Care and Use of Laboratory Animals” prepared by the National Academy of Sciences and published by the National Institutes of Health and the studies were approved by the local Ethical Committee of the kantonale Veterinäramt.

Generation of Recombinant Adenovirus.

The recombinant adenovirus used to express miR-192/194 was generated by inserting a PCR-amplified 374 nt murine genomic sequence from chromosome 19 including 42 nt upstream of the miR-194 stem loop sequence and 42 nt downstream of the miR-192 stem loop sequence into a green fluorescent protein (GFP) expressing shuttle vector Ad5CMV K-NpA (ViraQuest).

RNA Isolation, Affymetrix Microarray, Northern Blotting Analysis, and miRNA Microarray.

Total RNA was isolated using the Trizol reagent (Invitrogen). For Affymetrix analysis, RNA was pooled from three animals in each group. The generation and analysis of Affymetrix microarray data were performed as described.15 Northern blotting was essentially performed as described15 with a detailed description included in the Supporting Methods. For miRNA microarray analysis, total RNA was isolated using TriZol reagent and pooled from the livers of three mice per group and analyzed by the miRXplore microarray platform according to the manufacturer's protocol (Miltenyi Biotec).

Real-Time PCR.

Two μg of total RNA was used for complementary DNA (cDNA) preparation with random hexamer primers using Super Script III Reverse Transcriptase (Invitrogen). Steady-state mRNA expression was measured by quantitative real-time PCR using the LightCycler 480 SYBR Green Master I Mix (Roche) with a Mx3005P Real Time PCR System (Stratagene). Transcript levels were normalized to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) or 36B4. Primer sequences for real-time PCRs are available on request. MiRNA levels were measured using the TaqMan miRNA Assays for miR-122, mir-22, mir-192, mir-194, let7b, or U6 (Applied Biosystems) and PCR results were normalized to U6 levels.

Assay of Luciferase Activity.

Mouse or human 3′ untranslated region (UTR) sequences were PCR-amplified with specific primers and cloned downstream of the stop codon in pRL-TK (Promega) as described.15 HEK-293 cells were cultured in 24-well plates and each well was transfected with 50 ng of the respective pRL-TK 3′ UTR constructs, 50 ng of pGL3 control vector (Promega), and 200 ng of double-stranded small interfering RNA (siRNA) (Dharmacon). Cells were harvested and assayed 24 to 30 hours after transfection. Results were normalized to the control vector and are expressed relative to the average value of the control miRNA.

Cell Culture, Infection, and Transfection.

HEK293 cells were maintained in growth medium, Dulbecco's modified Eagle's medium (Invitrogen) containing 4.5 g/L glucose supplemented with 10% fetal bovine serum (FBS) and penicillin/streptomycin. HEK293 cells were transfected with siRNAs at the indicated concentration using Lipofectamine 2000 (Invitrogen).

Bioinformatic Analysis.

A detailed description of the bioinformatic analysis is included in the Supporting Methods.

Statistical Analysis.

All bars show mean ± standard error of the mean (SEM). Significance was calculated using Student's t test (*P < 0.05; **P < 0.01; ***P < 0.001).

Results

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Tcf1 Regulates the MiR-192/-194 Cluster in Adult Liver.

Tcf1 controls the expression of many genes that are important for hepatocyte function, but the Tcf1-dependent miRNA expression in the liver has not been studied so far. Therefore, we analyzed miRNA expression in liver from mice in which Tcf1 had been deleted using miRNA microarrays. In comparison to 954 synthetic miRNA sequences, 54 miRNAs were robustly detected in our liver samples (Supporting Table 1). Only three miRNA sequences (miR-192, miR-193, miR-194) fulfilled the criteria to be <0.5-fold down-regulated and to obtain a P-value < 0.0001 in at least three out of the four possible gene array comparisons. On the other hand, no miRNA was up-regulated >2-fold with a P-value < 0.0001 in at least three out of the four possible gene array comparisons.

We selected 10 miRNAs that are enriched in endodermal tissues,20, 21 including the three miRNA sequences that were significantly regulated in our miRNA analysis, and compared their expression between mice in which Tcf1 had been genetically deleted (Tcf1−/−) and their wildtype littermates using northern blotting. We confirmed two miRNAs, miR-192 and miR-194, that were down-regulated in Tcf1−/− livers, whereas the expression of all other miRNAs tested, among them the liver-specific miRNA miR-122, was unchanged (Supporting Fig. 1). Mir-192 and miR-194 are transcribed from the same precursor, the miR-192/-194 pri-miRNA, indicating that regulation occurred at the transcriptional level. Indeed, both miRNAs were markedly down-regulated in the liver, and expression levels also decreased in Tcf1 null mice in two other tissues which express Tcf1, including kidney and small intestine, although to a lesser extent (Fig. 1). Together, our results indicate that in adult mice Tcf1 controls the expression of only a specific subset of miRNAs, the miR-192/-194 cluster. This regulation occurs predominantly in the liver.

thumbnail image

Figure 1. Regulation of the miR-192/-194 cluster in Tcf1−/− mice. Northern blot analysis of total RNA (30 μg) isolated from liver, kidney, or small intestine of Tcf1−/− mice or their wildtype littermates, respectively. Probing for snU6 served as loading control. **P < 0.01; ***P < 0.001 according to Student's t test.

Download figure to PowerPoint

Identification of MiRNA Targets In Vivo.

Available cloning data from adult liver suggested that the miR-192/-194 cluster is expressed at low levels representing less than 1% of all sequenced miRNA reads.20 Because the impact of this miRNA in liver tissue may be difficult to infer simply based on target predictions, we used the following approach to identify in vivo targets. First, we silenced hepatic miR-192 and miR-194 expression in vivo using simultaneous intravenous injection of two separate antagomirs, antagomir-192 and antagomir-194. For the reverse approach, overexpression of these miRNAs, we used adenoviral gene transfer (Ad-192/-194). Both interventions were efficient as confirmed using northern blotting (Supporting Fig. 2). Next, we obtained Affymetrix gene profiles from livers in which miR-192 and miR-194 had been silenced using antagomirs. In all, 812 transcripts were regulated at least 1.4-fold in antagomir-treated mice compared with controls. Silencing of an miRNA relieves repression of its endogenous targets and leads to enrichment of the binding motif for the miRNA in up-regulated genes. Indeed, the Affymetrix analysis revealed a significant enrichment for the miR-194 motif in the group of up-regulated transcripts (Table 1). In contrast, the miR-192 seed-complementary motif was only marginally enriched, suggesting that this miRNA might be less relevant for adult liver. We also considered that miR-192 and miR-194 have family members at other genomic sites (on mouse chromosome 19 miR-194-2/mir-192, on mouse chromosome 1 miR-194-1/miR-215). Although miR-194-1 and miR-194-2 have identical sequences, miR-192 and miR-215 differ by three nucleotides, but share the same binding motif (seed region). However, northern blotting could detect miR-215 only in small intestine and quantification of relative expression levels of miR-215 and miR-192 in liver and small intestine using Taqman miRNA assays demonstrated that miR-215 is preferentially expressed in small intestine, whereas miR-192 is predominantly expressed in liver (Supporting Fig. 3). Furthermore, miR-215 was not detectable in the liver samples in our miRNA microarray analysis (data not shown). We conclude that miR-215 expression is unlikely to contribute to the regulation of miR-192 targets in adult liver tissue.

Table 1. Analysis of Binding Motif Enrichment After Silencing or Overexpression of miR-192 and miR-194 In Vivo
miR-192nt2-7nt2-8nt1-7nt1-8
Antagomir-192/-194-up--
  P = 0.032  
Ad-miR-192/-194downdown--
 P = 0.OI7P = 0.007  
miR-194nt2-7nt2-8nt1-7nt1-8
  1. Liver RNA was isolated after silencing of miR-192/-194 using antagomirs (Antagomir-192/-194) or overexpression of the microRNAs (Ad-miR-192/-194), respectively. Subsequently, Affymetrix microarrays were performed and compared to controls. Data sets were analyzed for enrichment of different binding motifs of miR-192 or miR-194, respectively (nucleotides 2-7 [nt 2-7], nucleotides 2-8 [nt2-8], nucleotides 1-7 [nt1-7], nucleotides 1-8 [nt1-8]) as described in Methods. “Up” indicates enrichment in the upregulated transcripts and “down” indicates enrichment in the downregulated transcripts.

Antagomir-192/-194--upup
   P = 0.005P = 0.049
Ad-miR-192/-194downdowndowndown
 P = 0.014P = 0.002P < 0.001P < 0.001

As expected, miRNA overexpression (Ad-192/-194) induced a strong enrichment of the seed-complementary motif from both miRNAs in the group of down-regulated transcripts. Similar results were obtained by comparing differential gene regulation of mRNAs carrying seed matches to miR-192/-194 with mRNAs carrying no seed matches to miR-192/-194 (Supporting Fig. 4A). Because the binding of the six nucleotide seed motif to the target mRNA is increased by an adenine opposite of miRNA position 1 regardless of complementarity,22 we also analyzed differential seed regulation requiring an adenine at position 1 in the binding site. For miR-194, this analysis did not change the results because the complementary nucleotide is already an adenine. For miR-192, we indeed observed a stronger gene regulation with an adenine at position 1, but this regulation was still less strong than observed for miR-194 (Supporting Fig. 4B). Together, these data indicated that we successfully silenced and overexpressed miR-194 in adult mouse liver in vivo.

To zoom in on the relevant targets of these miRNAs, we combined the information of mRNA expression changes in silencing and overexpression experiments (Fig. 2A). The group of transcripts showing up-regulation after silencing and down-regulation after overexpression of the miRNAs (in the lower right quadrant of Fig. 2A) are of particular interest because they behave as direct targets of the miRNAs. This group contained 488 mRNAs (Fig. 2B), which harbor an miR-192 or miR-194 seed-complementary motif. Although the distribution of transcripts containing miR-192 or miR-194 seed-complementary motifs across the quadrants is not distinguishable from the distribution of transcripts irrespective of whether or not carrying these motifs (Fig. 2C), the total count of miR-194- and miR-192-complementary sites was enriched in the lower right quadrant (Fig. 2D). This suggests that in vivo targets of these low-expressed miRNAs harbor more than one seed-complementary motif. There was no evidence that transcripts carrying one seed match to miR-192 and one seed match to miR-194 are more strongly regulated than transcripts carrying two seed matches to the same miRNA (data not shown), suggesting that the two miRNAs do not act synergistically in adult liver. Based on this preliminary analysis, we analyzed all transcripts that were significantly up-regulated >1.3-fold and down-regulated <1.3-fold after silencing or overexpression of miR-192/-194, respectively, and that contained more than two seed matches for either miR-192 or miR-194. Three genes fulfilled these criteria, all of which carry miR-194- rather than miR-192-complementary motifs. Of the three candidate targets, qPCR revealed that Fzd6 expression was most strongly and significantly up-regulated in antagomir-treated mice (Fig. 3). Together, these results suggested that our approach successfully identified Fzd6 as an endogenous target of miR-194.

thumbnail image

Figure 2. Combination of Affymetrix gene array profiles after silencing or overexpression of the miR-192/-194 cluster in vivo. Total RNA was isolated from liver after treatment of mice with intravenous injections of antagomirs or adenovirus as described in the legend of Supporting Fig. 2. (A) Affymetrix gene array analysis was performed and results from overexpression of the miRNAs plotted against results from silencing of the miRNAs. (B) Number of total amount of mRNAs in the four corners as depicted in (A). (C) Number of mRNAs with a miR-192 or miR-194 binding motif (seed). (D) Total number of all miR-192 and miR-194 seeds in the four corners. P < 0.005 according to Fisher's exact test.

Download figure to PowerPoint

thumbnail image

Figure 3. qPCR screening for endogenous miR-194 targets after antagomir treatment. Candidates for direct miR-194 targets were selected as described in the Results and qPCR analysis was performed on liver RNA that was isolated from mice treated with antagomir-192/-194 (black bars) or scrambled antagomir (control) (gray bars). GAPDH was used as housekeeping gene and results expressed as relative units (relU). **P < 0.01 according to Student's t test.

Download figure to PowerPoint

Fzd-6 Is an Endogenous and Conserved Target of MiR-194 in Adult Liver.

To confirm that Fzd6 is an endogenous target of miR-194 in vivo, we performed additional control experiments using overexpression or silencing of miR-194 in adult liver. Fzd6 expression always correlated inversely with the expression level of miR-194 (Fig. 4A,B). Importantly, Fzd6 expression was also up-regulated in liver from mice treated only with antagomir-194 (Fig. 4C), confirming that Fzd6 is a target of miR-194 from the miR-192/-194 cluster. Fzd6 also showed a trend for up-regulation in Tcf1−/− livers (Fig. 4D), but to a lesser extent than compared with livers in which miR-194 was completely silenced using antagomirs. The 3′ UTR of the mouse Fzd6 transcript contains three predicted miR-194 binding sites, the first two binding sites being conserved to human FZD6.23 Indeed, FZD6 was also regulated by miR-194 in human cells (Fig. 4E). Furthermore, we cloned mouse and human 3′ UTR of FZD6 into a luciferase reporter system. When cotransfected with miR-194 both reporters exhibited significant repression (Fig. 4F,G). Together, these data confirm that Fzd6 is an endogenous target of miR-194 in vivo and is conserved between mouse and human.

thumbnail image

Figure 4. Fzd6 is a target of miR-194 in vivo. (A-E) qPCR analysis of Fzd6 expression in liver using RNA that was isolated from mice after overexpression (Ad-192/-194) or silencing of miR-192 and miR-194, respectively. Results were analyzed as described in the legend of Fig. 3. In (D,E), RPLP0 (36B4) was used as a housekeeping gene. (E) Regulation of FZD6 in HEK293 cells after transfection with a synthetic miR-194 (s-miR-194). (F,G) Mir-194-directed repression of renilla luciferase reporter genes bearing 3′ UTRs from mouse Fzd6 or human FZD6, respectively, after cotransfection into HEK293 cells with either miR-194 (s-miR-194) or control (s-miR-122). *P < 0.05; **P < 0.01; ***P < 0.001 according to Student's t test.

Download figure to PowerPoint

Fzd6 and MiR-194 Are Inversely Correlated in a Genetic Mouse Model of HCC.

Frizzled membrane receptors can mediate Wnt signaling by activating canonical β-catenin, or the noncanonical c-Jun N-terminal kinase (JNK) and protein kinase C (PKC) pathways, which are known to promote malignancy in virus-related HCCs.24 Fzd6 is commonly up-regulated in HCC25 and also down-regulation of miR-194 has been reported in different mouse models of liver tumors26, 27 and in human liver mesenchymal-like cancer cell lines.6 We used a genetic approach to test whether miR-194 and Fzd6 are correlated in HCC. For this we employed a mouse HCC model in which Dgcr8, an RNA-binding protein that is essential for miRNA biogenesis, and p53 alleles were selectively inactivated in the liver by crossing conditional alleles with a transgenic mice expressing the Cre-recombinase under the control of the albumin promoter (Alb-Cre).16-18 As reported for mice lacking the RNase Dicer in liver (Dicerflox/flox × Alb-Cre),28 we observed spontaneous multifocal HCC formation in Dgcr8flox/flox × p53flox/flox × Alb-Cre mice. In contrast to the Dicer deficient mice, Dgcr8/p53 null livers develop tumors between 3 and 6 months of age. As expected, tumor tissue was depleted of >90% of miR-194 (Fig. 5A). Importantly, this decrease of miR-194 was associated with a 2-fold increase in Fzd6 expression (Fig. 5B). We also asked if a selective rescue of miR-194 expression in miRNA-depleted hepatocytes would lead to the repression of Fzd6 transcript levels. To test this we established cell lines from liver tumors of Dgcr8flox/flox × p53 × Alb Cre mice. The loss of miR-194 in these cells was confirmed and a transfection protocol was established that allowed the delivery of synthetic double-strand miRNAs in these cells (Fig. 5C). Reexpression of miR-194 resulted in a ˜70% inhibition of Fzd6 expression, demonstrating that this miRNA is a major posttranscriptional regulator of Fzd6 in hepatocytes (Fig. 5D). These results confirm that miR-194 and its endogenous target Fzd6 are inversely correlated in one murine model of HCC and provide a mechanistic basis for this observation.

thumbnail image

Figure 5. Inverse correlation of Fzd6 and miR-194 transcripts in spontaneous liver tumors. (A-D) qPCR analysis of miRNAs and Fzd6 in livers and hepatoma cells. Expression levels of indicated miRNAs (A) and Fzd6 (B) by qPCR in Dgcr8/p53 null liver tumor tissues and liver tissue of Dgcr8flox/flox littermate controls, n = 3, *P ≤ 0.015, **P ≤ 0.002. (C) Expression levels of miR-194 in a cell line derived from a spontaneous liver tumor of a liver-specific Dgcr8/p53 mouse, ***P ≤ 0.001. (D) Relative expression levels of Fzd6 in livers of Dgcr8- and p53-deleted cells that were transfected with a synthetic miR-194 (s-miR-194) or scrambled miRNA (s-Ctrl) *P < 0.05.

Download figure to PowerPoint

To address the question whether miR-194/Fzd6 regulation is potentially important in the pathogenesis of HCC, we analyzed miR-194-dependent regulation of genes that are downstream of Wnt/β-catenin signaling. Activation of Wnt/β-catenin signaling in mouse liver leads to HCC29 and several groups have identified endogenous targets that are activated by the wnt/β-catenin pathway in hepatocytes.30, 31 These genes are relevant to ammonia metabolism (Glul, Slc1a2, Oat, Rhbg), regulation of Wnt signaling (axin2), chemotaxis (Lect2), and other functions (RNAse4). Importantly, silencing or overexpression of miR-194, respectively, led to a coordinated response in the expression of these Wnt/β-catenin target genes in liver tissue from our different mouse models (Fig. 6A-C). Silencing of miR-194 in vivo using antagomirs or by genetic deletion of Dgcr8 increased the expression of β-catenin target genes (Fig. 6A,B). Conversely, reexpression of miR-194 resulted in down-regulation of these genes (Fig. 6C). Together, these data identify miR-194 as a suppressor of wnt/β-catenin signaling in mouse liver in vivo and provide evidence that miR-194/Fzd6 could have a role in liver carcinogenesis.

thumbnail image

Figure 6. Regulation of hepatic Wnt/β-catenin targets in liver and hepatoma cells with either gain- or loss-of-function of miR-194. (A-C) qPCR analysis of hepatic Wnt/β-catenin targets in livers and cell lines derived from Dgcr8/p53 null liver tumors. (A) Hepatic expression levels in antagomir-treated mice as described in Fig. 4B. (B) Expression levels of indicated genes in Dgcr8/p53 null liver tumor tissues compared with wildtype control livers (Dgcr8fl/fl). (C) Expression levels in cell lines derived from Dgcr8/p53 null liver tumors that were transfected with a synthetic miR-194 (s-miR-194) or scrambled miRNA (s-Ctrl) as described in Fig. 5D. RPLP0 (36B4) was used as a housekeeping gene. *P < 0.05.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

A major finding of our study is the regulation of the miR-192/-194 cluster by Tcf1 in adult liver in vivo. The decreased expression of miR-192/-194 in liver from Tcf1−/− mice was remarkably specific and more highly expressed as well as liver-enriched miRNAs (miR-22, let7b, and miR-122, respectively) were not affected by the loss of Tcf1. The regulation of miR-194 by Tcf1 could be relevant for liver tumorigenesis for several reasons. Previous studies have reported miR-194 as being down-regulated in HCC.26, 27 Furthermore, we demonstrate that miR-194 regulates hepatic Fzd6 and that the levels of miR-194 and Fzd6 are inversely correlated in one mouse model of HCC. Dysregulation of WNT/FZD receptor elements is a common feature in human hepatocarcinogenesis25 and activating mutations of β-catenin, the downstream mediator of Wnt signaling, can be detected in 20%-30% of HCC.32 Indeed, we could confirm that miR-194 acts as a suppressor of Wnt/β-catenin signaling in adult liver. The role of miR-194 for tumorigenesis in liver is further supported by a recent study that demonstrated suppression of metastasis of the liver cancer cell line SK-Hep-1 in mice by miR-194.6 However, in this study no endogenous miRNA targets were reported using loss of miR-194 function studies in vivo.

Overexpression of miRNAs may regulate many nonendogenous targets, especially when miRNAs with low expression levels are studied. Indeed, in our analysis overexpression of miR-194 in liver caused many more transcripts with an miR-194 seed to be regulated compared with the silencing of miR-194. However, mRNAs with several seed matches still changed consistently between the silencing and overexpression experiments. These data suggest that the most robust in vivo targets of miRNAs that are little expressed harbor more than one seed-complementary motif. Our study demonstrates a successful implementation of a strategy to identify these endogenous miRNA targets in vivo. The approach involves the analysis of global gene expression in gain- and loss-of miRNA function models using microarrays followed by testing of all regulated transcripts containing multiple seed sequences using qPCR. Although Fzd6 is predicted as an miR-194 target (i.e., by ElMMo),33 its rank is rather low and therefore would not be a priority for experimental validation. There could be many reasons, most prominently the embedding of miRNA targets into larger regulatory networks, that could be responsible for the lack of response of predicted targets in our particular set of experiments. The approach presented here may be especially relevant for the identification of targets of miRNAs with low expression levels in a particular tissue of interest.

Decreased expression of the miR-192/-194 cluster in Tcf1 mice is in line with a previous report in which expression of this cluster was also controlled by Tcf1 in intestinal cell lines.14 This study could also establish a direct interaction of Tcf1 and the miR-194-2 promoter in Caco-2 cells. In our study, however, loss of Tcf1 had a greater effect on miR-192/miR-194 expression in liver tissue than small intestine. Differential regulation of Tcf1 targets between distinct tissues is not an unusual observation.3

Our finding that Tcf1 is upstream of only a very limited subset of miRNAs in liver in vivo is of relevance for the interpretation of miRNA profiling data obtained from human samples. For example, Ladeiro et al.34 analyzed miRNA expression using quantitative reverse-transcriptase PCR (RT-PCR) in different human hepatocellular tumors, including hepatocellular adenomas (HCA) with mutations in the TCF1 gene. Fifteen miRNAs were found to be dysregulated in HCA with hepatocyte nuclear factor 1a (HNF1α) mutations. Mir-192 was down-regulated ˜2.7-fold with the lowest P-value of all miRNAs tested, whereas miR-194 was not included in this subanalysis. These results provide compelling evidence that the miR-192/miR-194 cluster is also downstream of TCF1 in human liver. It is also important to note that whereas we analyzed miRNA expression in mice with a germline loss-of-function of Tcf1 >80% of the adenomas analyzed by Ladeiro et al. contained somatic mutations,35 indicating that the regulation of miR-192/miR-194 is not limited to germline mutations. The fact that Ladeiro et al. identified a significantly higher number of regulated miRNAs than our study might be attributed to environmental influences inherent to clinical studies, e.g., differences in use of oral contraceptives, diet, body mass index, tobacco use, or different degrees of liver steatosis.

Tcf1 represents an important driver of hepatocellular adenoma development and could represent an early step in malignant transformation of hepatocytes.36 In humans, 30%-50% of hepatocellular adenomas harbor a biallelic Tcf1 mutation.37 MiR-194 and its endogenous target Fzd6 may contribute to this effect. However, we only observed a trend for up-regulation of Fzd6 in livers of Tcf1−/− mice. The lesser degree of regulation is consistent with the lesser down-regulation of miR-194 in the liver of these mice compared with antagomir-194 treatment, which leads to a complete inactivation of miR-194. However, stronger regulation of Fzd6 in Tcf1−/− mice might be observed at the protein level, which could not be tested in our study because appropriate antibodies are lacking. Whether FZD6 is downstream of TCF1 in human liver also deserves further testing. Global gene expression analysis based on microarray studies between human hepatocellular adenomas with biallelic inactivating mutations of TCF1 did not reveal significant alterations of the probe sets of FZD6,38 but more direct analysis of mRNA or protein levels have yet to be performed. In addition, we cannot exclude that miR-194 might regulate other targets that are involved in liver proliferation. It would therefore be interesting to investigate if hepatocellular adenomas/carcinomas harbor mutations in the gene encoding miR-192/-194. In any case, our results provide evidence that miR-192/-194 is not relevant for the downstream metabolic effects of Tcf1. Neither silencing nor overexpression of the miR-192/-194 cluster altered several metabolic parameters such as bodyweight, plasma glucose, or cholesterol levels (data not shown).

Together, our results support a role of miR-194 in liver tumorigenesis. Further functional studies of miR-194 could enhance our understanding of how Tcf1 governs hepatocyte development and proliferation.

References

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. References
  7. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
HEP_24658_sm_SuppFig1.tif8519KSupporting Fig. 1. Expression of endodermal-tissue-enriched microRNAs in liver from Tcf1-/- mice. Northern Blots of total RNA (30μg) isolated from livers of Tcf1-/- mice or their wildtype littermates. Probing for snU6 (U6 small nuclear RNA) served as loading control.
HEP_24658_sm_SuppFig2.tif8518KSupporting Fig. 2. Hepatic expression of miR-192 and miR-194 after antagomir or adenovirus-treatment in vivo. Northern blots of total RNA isolated after 3 consecutive injections of antagomir-192 and antagomir-194 or control antagomir, respectively (80mg/kg/bw per day). For overexpression, liver was harvested 6 days after the intravenous injection of the respective adenovirus constructs as described in Methods. Ethidium bromide staining of tRNA is shown as a loading control.
HEP_24658_sm_SuppFig3.tif8519KSupporting Fig. 3. Preferential expression of miR-192 and miR-215 in liver and small intestine, respectively. (A) Multiple tissue Northern blots were probed for miR-194, miR-192 and miR-215, respectively. Probing for snU6 served as loading control. (B) For RT-PCR analysis, 10ng of total RNA from murine liver and small intestine (n=4) were transcribed and amplified using the manufacturer's protocol (Applied Biosystems). Results were normalized for the data obtained by the RNU6B primer set.
HEP_24658_sm_SuppFig4.tif904KSupporting Fig. 4. Differential regulation of putative miR-192/-194 targets in the antagomir and overexpression experiments. Total RNA was isolated from liver after treatment of mice with intravenous injections of antagomirs or adenovirus as described in the legend of Supporting Fig. 2. Subsequently, Affymetrix gene array analysis was performed. (A) Distribution of log2-fold changes for mRNAs carrying a seed match to miR-192 exclusively (red line), miR-194 exclusively (green line), or no seed match to miR-192/-194 (grey line) in their 3'UTR in the antagomir experiment. The green, red, and grey dots represent the average log2 fold change of mRNAs with a seed match to miR-192, miR-194 or no seed match. Fold-changes of mRNAs with miR-192 or miR-194 seed matches were compared to mRNAs carrying no seed match using Wilcoxon's two-sided rank sum test. The corresponding p-values are indicated in red and green for miR-192 and miR-194, respectively. Ad-miR-192/-194 indicates the identical analysis for the overexpression of miR-192/-194 using adenovirus. (B) Same analysis as in A for the seed matches of miR-192 or miR-194, respectively, but requiring an “A” at the nucleotide opposite of position 1 of the microRNA.
HEP_24658_sm_SuppTable.doc1329KSupporting Table 1. Microarray analysis of microRNA expression in livers of Tcf1 null mice. Tcf1_1 and Tcf1_2 refer to two independent pools of RNA obtained from liver tissue of Tcf null mice. Con_1 and Con_2 are two independent RNA pools from littermate controls, respectively. Reratio indicates the fold-regulation obtained from the microRNA microarray analysis. $ indicates p-value <0.0001. The Gaussian error function and a preset p-value were used for the decision whether an averaged ratio was marked as significantly regulated.
HEP_24658_sm_SuppInfo.doc52KSupporting Information

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.