Recombinant adenovirus carrying the hepatocyte nuclear factor-1alpha gene inhibits hepatocellular carcinoma xenograft growth in mice


  • Potential conflict of interest: Nothing to report.

  • This work was supported by the National Natural Science Foundation of China (nos. 81070347, 30971346, and 30901775), National Science Fund for Distinguished Young Scholars (no. 30825020), and Creative Research Groups (no. 30921006).


Hepatocyte nuclear factor-1alpha (HNF1α) is one of the key transcription factors of the HNF family, which plays a critical role in hepatocyte differentiation. Substantial evidence has suggested that down-regulation of HNF1α may contribute to the development of hepatocellular carcinoma (HCC). Herein, human cancer cells and tumor-associated fibroblasts (TAFs) were isolated from human HCC tissues, respectively. A recombinant adenovirus carrying the HNF1α gene (AdHNF1α) was constructed to determine its effect on HCC in vitro and in vivo. Our results demonstrated that HCC cells and HCC tissues revealed reduced expression of HNF1α. Forced reexpression of HNF1α significantly suppressed the proliferation of HCC cells and TAFs and inhibited the clonogenic growth of hepatoma cells in vitro. In parallel, HNF1α overexpression reestablished the expression of certain liver-specific genes and microRNA 192 and 194 levels, with a resultant increase in p21 levels and induction of G2/M arrest. Additionally, AdHNF1α inhibited the expression of cluster of differentiation 133 and epithelial cell adhesion molecule and the signal pathways of the mammalian target of rapamycin and transforming growth factor beta/Smads. Furthermore, HNF1α abolished the tumorigenicity of hepatoma cells in vivo. Most interestingly, intratumoral injection of AdHNF1α significantly inhibited the growth of subcutaneous HCC xenografts in nude mice. Systemic delivery of AdHNF1α could eradicate the orthotopic liver HCC nodules in nonobese diabetic/severe combined immunodeficiency mice. Conclusion: These results suggest that the potent inhibitive effect of HNF1α on HCC is attained by inducing the differentiation of hepatoma cells into mature hepatocytes and G2/M arrest. HNF1α might represent a novel, promising therapeutic agent for human HCC treatment. Our findings also encourage the evaluation of differentiation therapy for tumors of organs other than liver using their corresponding differentiation-determining transcription factor. (HEPATOLOGY 2011)

Hepatocellular carcinoma (HCC) is one of the most common cancers, with a very dismal outcome.1, 2 The family of hepatocyte nuclear factors (HNFs) consists of a series of liver-enriched transcription factors, including HNF1, HNF3, HNF4, HNF6, and CCAAT/enhancer binding proteins (CEBPs).3, 4 These factors form a systemic transcriptional network that finely regulates the expression of a large number of hepato-specific and -enriched genes.5, 6 They also play crucial roles in the development and regeneration of the liver. In addition, the expression levels of HNFs are associated with the differentiation of HCC.7, 8 Specifically, the expression of HNF4α and 1α is reduced simultaneously in cells derived from human HCC, hepatoblastomas, and immortalized hepatocytes.7 We have previously shown that forced reexpression of HNF4α induced the differentiation of both hepatoma cells and their cancer stem cells into more mature phenotypes and abolished their tumorigenesis. Moreover, we showed that HNF4α could protect mice against the metastasis of HCC and also inhibit the growth of HCC xenograft.9 More recently, we also documented the striking suppressive effect of HNF4α on hepatic fibrosis and hepatocarcinogenesis.10, 11

Hepatocyet nuclear factor-1alpha (HNF1α), another key transcription factor of the HNF family, belongs to the POU-homeodomain family and interacts with DNA in the form of homo- or heterodimers with its isoform, HNF1β. HNF1α can bind to the cis-acting elements of at least 200 genes in human hepatocytes and contribute to many of the important biological functions of hepatocytes, such as carbohydrate synthesis and storage, lipid metabolism, detoxification, and synthesis of serum proteins.3, 12HNF1α knockout mice exhibited abnormal liver function, with less than 15% of the mice surviving to day 42 and less than 1% living longer than 3 months. More interestingly, these HNF1α-deficient mice also represented two tumor-associated phenotypes (i.e., dramatic liver enlargement caused by an increased proliferation of hepatocytes and degeneration of individual hepatocytes).13, 14 On the other hand, it has been proven that the expression of HNF1α was decreased in chemically induced liver tumors in mice15 and in human HCC tissues.16, 17 The expression of HNF1α was also reduced in the course of progression from slowly growing differentiated HCC to the fast-growing dedifferentiated HCC variant.8 In addition, substantial evidence has demonstrated that inactivation or mutation of HNF1α is an important genetic event in the development of liver adenoma and might contribute to hepatocarcinogenesis.18, 19 Taken together, these findings suggest that suppression of HNF1α may play a critical role in the development of HCC. Therefore, we speculate that up-regulation of HNF1α might restrain the proliferation of HCC and exert a substantial therapeutic effect on HCC.

In the current study, we demonstrated that forced reexpression of HNF1α could induce the differentiation and G2/M arrest of hepatoma cells, leading to the inhibition of their reproductive activity and abrogation of cell tumorigenesis. More intriguingly, we documented that HNF1α overexpression could significantly inhibit HCC xenograft growth in vivo. These data strongly indicate that HNF1α, as one of the differentiation-determining transcription factors for hepatocytes, might represent a promising, novel therapeutic agent for human HCC.


ATP, adenosine triphosphate; CD, cluster of differentiation; CEBPs, CCAAT/enhancer binding proteins; CYP, cytochrome P450; DEN, diethylinitrosamine; EpCAM, epithelial cell adhesion molecule; GFP, green fluorescent protein; HCC, hepatocellular carcinoma; HNFs, hepatocyte nuclear factors; HNF1α, hepatocyte nuclear factor-1alpha; IT, intratumoral; mRNA, messenger RNA; miRNA, microRNA; mTOR, mammalian target of rapamycin; NOD/SCID, nonobese diabetic/severe combined immunodeficiency; PCK, phosphoenolpyruvate carboxykinase; PCNA, proliferating cell nuclear antigen; pfu, plaque-forming units; RT-PCR, reverse-transcriptase polymerase chain reaction; SC, subcutaneous; SD, standard deviation; siRNA, small interfering RNA; TAFs, tumor-associated fibroblasts; TGFβ, transforming growth factor beta.

Materials and Methods

See Supporting Methods for some detailed experimental methods.

Cells, Animals, and Tissue Specimens

HCC cell lines and the human embryonic kidney cell line, 293, were obtained from the American Type Culture Collection (Manassas, VA). Male BALB/c nude mice and nonobese diabetic/severe combined immunodeficiency (NOD/SCID) mice were purchased from the Shanghai Experimental Animal Center of the Chinese Academy of Sciences (Shanghai, China). The rat HCC model was induced by intraperitoneal injection of diethylinitrosamine (DEN), as previously described.11 The human mature hepatocytes were separated from liver fragments of donors for liver transplantation, and the human HCC and its adjacent tissues were obtained from 20 patients with HCC who underwent hepatectomy or liver transplantation in Shanghai Eastern Hepatobiliary Surgery Hospital (Shanghai, China).

Adenoviral Vectors

AdHNF1α, a replication-deficient E1 and E3 adenovirus vector expressing the HNF1α gene under the transcriptional control of the cytomegalovirus promoter, was constructed as described previously.9 Additionally, AdGFP, an adenovirus only expressing green fluorescent protein (GFP), was yielded as a control.

Quantitative Real-Time Reverse-Transcriptase Polymerase Chain Reaction, Western Blotting Analysis, Immunocytochemistry, and Immunohistological Analysis

Quantitative real-time reverse-transcriptase polymerase chain reaction (RT-PCR), western blotting analysis, immunocytochemistry, and immunohistological analysis were carried out according to the manufacturer's recommended methods. The primers for real-time RT-PCR are listed in Supporting Table 1. The primary antibodies used for western blotting analysis, immunocytochemistry, and immunohistological staining are shown in Supporting Table 2.

Cell Proliferation and Clonogenic Assays

For cell proliferation assays, the number of metabolically active mitochondria and viable cells was determined daily after virus infection or microRNA (miRNA) mimics or inhibitor transfection. The effects of adenovirus or miRNA mimics on the colony-formation ability of hepatoma were also evaluated.

In Vitro Adenosine Triphosphate/Tumor Chemosensitivity Assay of Human Cancer Cells and Tumor-Associated Fibroblasts

Human cancer cells were isolated from 7 human HCC tissues, and tumor-associated fibroblasts (TAFs) were separated from 2 human HCC tissues, according to the method previously established. Adenosine triphosphate (ATP)/tumor chemosensitivity assay was used to confirm the effect of HNF1α on them.

Flow Cytometry Analysis

Cell-cycle distribution, cell apoptosis, and percentage of CD133+ cells were identified by flow cytometry analysis.

Transfection of Small Interfering RNA, miRNA Mimics, and miRNA Inhibitors

Small interfering RNA (siRNA), miRNA mimics, miRNA inhibitors, and their negative controls were synthesized by GenePharma (Shanghai GenePharma Co., Ltd., Shanghai, China) and transfected into Huh7 cells using Lipofectamine2000 (Invitrogen, Carlsbad, CA) in triplicate, according to the manufacturer's instructions. All sequences are listed in Supporting Table 3.

Tumor Formation in Animals

Hep3B and Huh7 cells infected with AdHNF1α or control virus AdGFP were injected subcutaneously (SC) into the left and right armpits of each BALB/c nude mouse, respectively. The kinetics of tumor formation was evaluated to detect the effect of HNF1α on the tumorigenicity of hepatoma cells.

Antitumor Effect of HNF1α in Mice

To investigate the antitumor effect of HNF1α in vivo, an SC HCC model was generated in BALB/c nude mice by the injection of 5 × 106 Hep3B to the armpits. Animals were treated with an intratumoral (IT) injection of 2 × 109 pfu (plaque-forming units) of AdHNF1α or AdGFP three times a week for up to 2 weeks when the volumes of tumors reached approximately 350 mm3 (Supporting Fig. 1A). In addition, a liver xenograft model was established in NOD/SCID mice by the direct injection of 5 × 106 Hep3B into the liver in situ. Tumor nodules in the liver were visible macroscopically 10 days after Hep3B cell injection. Then, 2 × 109 pfu of adenoviruses AdHNF1α or AdGFP were administrated via the tail vein twice a week for up to 3 weeks (Supporting Fig. 1B). Mice were sacrificed 1 week after final virus delivery.

Figure 1.

Characterization of HNF1α expression. The gene expression fold of HNF1α in hepatoma cell lines versus human mature hepatocytes was detected by real-time RT-PCR (A). Real-time RT-PCR (B) and immunohistochemistry analysis (C) were used to examine HNF1α expression in livers from DEN-treated mice. The mRNA expression level was normalized against β-actin. Real-time RT-PCR analysis for HNF1α mRNA expression was performed in 20 paired clinical HCC samples and their surrounding liver tissues (D). mRNA expression levels were normalized against β-actin. The expression of HNF1α in cancerous tissues was reduced in 52.94% (I), up-regulated in 23.53% (II), and not apparently altered in 23.53% (III) of HCC patients, according to immunohistochemical analysis (E). Representative images of HNF1α in an HCC cancerous tissue or its surrounding tissue by immunohistochemistry from 17 patients (F). Scare bars, 100 μm. *P < 0.05; **P < 0.01; ***P < 0.001.

Statistical Analysis

Analysis of variance and the Student's t test were used for comparison among the groups and between paired data, respectively. HCC-free survival was calculated with the Kaplan-Meier method, and significance was determined by the log-rank test. A P value < 0.05 was considered significant, and P < 0.01 was considered very significant. Correlation analysis was performed using Pearson's method, and a P value < 0.01 was considered statistically significant.


HCC Is Associated With Decreased Expression of HNF1α

Previous study has shown that the expression of HNF1α may be associated with the development and differentiation of HCC.8, 19 Herein, we found that HNF1α messenger RNA (mRNA) transcription levels in hepatoma cell lines were markedly decreased, compared with human mature hepatocytes (Fig. 1A) (P < 0.001). We also demonstrated that HNF1α expression in the liver was gradually decreased during the development of HCC induced by DEN (Fig. 1B,C). Additionally, we found that HNF1α levels in HCC tissues were significantly lower than those in the adjacent tissues in the most paired tissues from 20 patients. Specifically, HNF1α mRNA transcription levels were reduced in 70% of the HCC tissue specimens, compared with their adjacent tissues, whereas HNF1α mRNA expression remained unchanged in 20% and elevated in 10% of the HCC tissue specimens (Fig. 1D). Immunohistochemical examination further confirmed these results at the protein level (Fig. 1E,F).

HNF1α Inhibits the Proliferation and Colonogenicity of Hepatoma Cells In Vitro

AdHNF1α infection resulted in a noticeable increase in both mRNA and protein levels of HNF1α 3 days after infection (Supporting Fig. 2). The expression of HNF1α in AdHNF1α-infected Hep3B and Huh7 cells was more than 10-fold, in comparison to healthy human liver tissues, whereas the physiological levels of HNF1α in AdHNF1α-infected MHCC-H and MHCC-L were not reached by the adenoviral delivery, which may have been the result of the extremely low level of HNF1α expression in these two cells lines. Moreover, we found that, at day 7 postinfection, HNF1α overexpression significantly suppressed the proliferation of all the four detected hepatoma cell lines (Fig. 2A,B; Supporting Fig. 3) (P < 0.01). In addition, clonogenic assays showed that hepatoma cells overexpressing HNF1α yielded a markedly reduced number of colonies, compared with their respective controls (Fig. 2C,D; Supporting Fig. 4) (P < 0.01).

Figure 2.

HNF1α suppresses the proliferation of HCC in vitro. The number of viable cells was counted daily after the hepatoma cell line, Hep3B, infected with adenovirus carrying HNF1α or GFP (A). The inhibition effect was significant at day 7 after AdHNF1α infection, compared with that of AdGFP or control group (B). Colony formation was also detected after virus infection (C and D). HNF1α delivery significantly reduced colony-formation ability (**P < 0.01). The proliferation of cancer cells (E) and TAF (F) isolated from human HCC tissues was significantly suppressed in AdHNF1α group (P < 0.05).

Figure 3.

HNF1α up-regulates the expression of liver-specific genes and inhibits the expression of CD133 and EpCAM. The expression of liver-specific genes and cancer stemness markers were detected by real-time RT-PCR after the HNF1α gene transferred to Hep3B and Huh7. The gene expression fold of liver-specific genes in the HNF1α versus the GFP group was analyzed after infection with adenovirus for 3 (A and B) or 7 days (C and D). The expression of CD90, CD133, and EpCAM was detected 3 days after HNF1α force overexpression (E). Each value represents the mean ± standard deviation (SD) for triplicate samples. *P < 0.05; **P < 0.01.

Figure 4.

HNF1α induces G2/M arrest in hepatoma cells. Flow cytometry analysis was used to detect the effect of HNF1α on cell cycle in Hep3B cells (A) and in Huh7 cells (B) after virus infection for 3 days. The percentage of cells in the G2/M phase was remarkably increased after HNF1α delivery. Real-time RT-PCR (C and D) and western blotting (E and F) were carried out to examine the expressions of cyclin A2, cyclin B1, cdc2, and p21 in Hep3B and Huh7. Expression of β-actin mRNA and protein was used as an internal control. Each value represents the mean ± SD for triplicate samples. HNF1α delivery significantly enhanced the expressions of cyclin A2, cyclin B1, and p21, whereas it suppressed the level of cdc2 expression. *P < 0.05.

To further confirm the effect of HNF1α on HCC, cancer cells and TAFs were isolated from 7 human HCC tissues and 2 human HCC tissues, respectively. In vitro ATP/tumor chemosensitivity assay showed that HNF1α inhibited the proliferation of all the 7 isolated human cancer cells and 2 TAFs at 7 days after adenovirus induction in a dose-dependent manner (P < 0.05) (Fig. 2E,F).

HNF1α Overexpression Promotes the Expression of Liver-Specific Genes and Inhibits the Expression of CD133 and Epithelial Cell Adhesion Molecule in Hepatoma Cells

Considering the critical role of HNF1α in the maintenance of hepatocyte function, we hypothesized that HNF1α might induce the differentiation of hepatoma cells by provoking the reexpression of hepatocyte marker genes. Indeed, HNF1α delivery for 3 days increased the expressions of a cluster of liver-specific genes in both Hep3B and Huh7 hepatoma cells, including glucose-6-phosphatase, alcohol dehydrogenase 1, C-reactive protein, apolipoprotein C|||, transthyretin, cytochrome P450 (CYP) 7A1, Na+/taurocholate transport protein, phosphoenolpyruvate carboxykinase (PCK) 1, PCK2, CYP1A2, CYP1A3, 4-hydroxyphenylpyruvate dioxygenase, ferripyochelin binding protein, aldolase B, and ornithine aminotransferase. The enhanced expression of biliverdin reductase and lipase A was only observed in Hep3B (Fig. 3A,B). More hepatic-specific gene expression was up-regulated 7 days after AdHNF1α infection, including ATP citrate lyase, fatty acid synthase, stearoyl-coenzyme A desaturase, fatty acid desaturase 1, and ELOVL fatty acid elongase 2 (Fig. 3C,D). Nevertheless, the expression of some other liver-specific genes, such as albumin, serum glutathione S epoxide transferase, CYP2E, and insulin receptor, was not remarkably enhanced after HNF1α delivery. Also, HNF1α overexpression did not significantly induce the transcription of HNF4α and 1β in all four detected hepatoma cell lines (data not shown). Quantitative real-time RT-PCR showed that the transcription of CD133 and epithelial cell adhesion molecule (EpCAM) was reduced 3 days after AdHNF1α delivery in Hep3B and Huh7 (Fig. 3E). Flow cytometry confirmed the decreased percentage of CD133+ cells (Supporting Fig. 5).

Figure 5.

HNF1α activates p21 partly by up-regulating the expression of miR-192 and miR-194. Real-time RT-PCR analysis for HNF1α, miR-192, and miR-194 expression was performed in 20 paired clinical HCC samples and their surrounding tissues (A). The transcription of miR-192, miR-194, and p21 was detected by real-time RT-PCR after AdHNF1α (B) or siRNA-HNF1α (C) introduction to Huh7 for 3 days. The transcription of p21 was also measured 3 days after the delivery of miR-192/miR-194 mimics (D) or miR-192/miR-194 inhibitors (E) to the hepatoma cell line, Huh7. Cell proliferation (F) was measured after miR-192 or miR-194 mimics transfected into Huh7. *P < 0.05.

Effect of HNF1α on Tumorigenic Cell-Signaling Pathways

It has been shown that inhibition of HNF1α could activate protumorigenic cell-signaling pathways, such as the mammalian target of rapamycin (mTOR) pathway.20 We then detected the effect of AdHNF1α on the mTOR pathway by real-time RT-PCR. As expected, HNF1α overexpression inhibited the expression of most genes in the mTOR pathway (Supporting Fig. 6A,B). Additionally, both real-time RT-PCR and western blotting showed that HNF1α delivery suppressed the expression of transforming growth factor β1 (TGFβ1) and its receptor, TβR1, TβR2, p-Smad2, and p-Smad1/5 in Hep3B (Supporting Fig. 6C,D).

Figure 6.

HNF1α decreases the tumorigenicity of hepatoma cells in vivo. Hep3B cells or Huh7 cells expressing HNF1α or GFP were injected SC into the left and right armpits of each nude mouse, respectively. The incidence of tumorigenicity (A and C) and volume of tumors (B and D) were significantly different between the HNF1α overexpression group and the GFP control group (P < 0.05). Data of volume are presented as means ± SD.

HNF1α Induces G2/M Arrest in Hepatoma Cells

Flow cytometry analysis revealed that HNF1α treatment for 72 hours significantly increased the percentage of cells in the G2/M phase in all the four kinds of hepatoma cells (Fig. 4A,B; Supporting Fig. 7; Table 1). Nevertheless, no apoptosis was induced by AdHNF1α in hepatoma cells, except MHCC-L (Supporting Fig. 8). Moreover, both real-time RT-PCR and western blotting showed that the expressions of cyclin A2, cyclin B1, and p21 was increased after HNF1α introduction, whereas the level of cdc2 expression was reduced (Fig. 4C-F).

Figure 7.

HNF1α exhibits potent antitumor effect in vivo. Adenoviruses carrying the HNF1α gene or GFP gene were IT injected into SC inoculated tumors. The tumor size after treatment was estimated by serial calculation (A), and the tumor nodules were collected at the end of treatment (B). The final tumor volume (C) and weight (D) in AdHNF1α-treated mice were significantly smaller than that in control mice. The orthotopic liver cancer model was established in NOD/SCID mice by direct injection of 5 × 106 Hep3B cells into the liver in situ. Liver samples were collected after 3 weeks of treatment (E). Tumor nodules are indicated with yellow arrows. *P < 0.05.

Table 1. Effect of HNF1α on Cell Cycles of Hepatoma Cells
GroupG0/G1 Phase(%)G2/M Phase(%)S Phase(%)
  • Abbreviations: HNF1α, hepatocyte nuclear factor-1alpha; GFP, green fluorescent protein.

  • Compared with GFP group:

  • *

    P < 0.05;

  • **

    P < 0.01.

 Control77.01 ± 1.1814.34 ± 1.958.65 ± 0.77
 GFP72.72 ± 1.2915.44 ± 0.0511.85 ± 1.33
 HNF1α50.59 ± 1.3635.12 ± 0.22**14.30 ± 1.58*
 Control81.51 ± 0.917.19 ± 0.2011.30 ± 0.42
 GFP80.93 ± 0.417.71 ± 0.5711.36 ± 0.35
 HNF1α71.73 ± 1.2117.24 ± 1.19**11.04 ± 0.04
 Control68.82 ± 1.318.58 ± 0.2922.60 ± 1.42
 GFP77.66 ± 0.829.25 ± 0.6713.09 ± 0.45
 HNF1α69.80 ± 0.2113.13 ± 0.59**17.07 ± 0.96
72.13 ± 0.758.58 ± 0.6119.29 ± 0.32 
 GFP69.66 ± 2.4115.81 ± 0.9714.53 ± 1.30
 HNF1α63.78 ± 1.0322.98 ± 0.69**13.24 ± 0.54

miRNA 192 and miRNA 194 Contribute to the Activation of p21 by HNF1α

It has been proven that miR-192 and miR-194 could inhibit clonogenicity and cell adhesion by inducing p21 accumulation and cell-cycle arrest,21 and that HNF1α could bind to the core element of the pri-miR-194-2 promoter.22 We then explored the possible role of miRNA in the up-regulation of p21 by HNF1α. Real-time RT-PCR analysis revealed that both miR-192 and miR-194 transcription levels were reduced in more than half of the HCC tissue specimens (Fig. 5A), which were positively correlated with HNF1α levels (r = 0.738, P < 0.001; r = 0.790, P < 0.001) by Pearson's correlation analysis. Moreover, HNF1α overexpression significantly increased the levels of miR-192 and miR-194 transcripts (Fig. 5B). Transfection of Huh7 cells with siRNA against HNF1α, on the other hand, resulted in a significant reduction in the transcription levels of miR-192 and miR-194 (Fig. 5C). Additionally, both miR-192 and miR-194 mimics enhanced the expression of p21 (Fig. 5D), which was suppressed by the inhibitors of both miR-192 and miR-194 (Fig. 5E). In addition, the inhibition effect of HNF1α on cell proliferation was partly alleviated by miR-192 and miR-194 inhibitors (Supporting Fig. 9). Cell proliferation and clonogenic assays further demonstrated that miR-192 and miR-194 suppressed the growth and clonogenicity of hepatoma cells (Fig. 5F; Supporting Fig. 10).

HNF1α Abolishes the Tumorigenicity of Hepatoma Cells In Vivo

To evaluate whether HNF1α would affect the tumorigenicity of hepatoma cells in vivo, we investigated the effect of HNF1α overexpression on xenograft growth in nude mice bearing Hep3B or Huh7 xenograft. We could detect the presence of tumors in nude mice bearing Hep3B xenograft expressing GFP as early as day 12 postinoculation, and all these mice developed tumor nodules by day 28. Interestingly, no animals had identifiable tumors in nude mice bearing Hep3B xenograft overexpressing HNF1α up to 6 weeks postinoculation (Fig. 6A,B). Similarly, at day 14 postinoculation, tumors developed in half of the nude mice bearing Huh7 xenograft expressing GFP, whereas no tumor was observed in nude mice bearing Huh7 xenograft overexpressing HNF1α. All mice bearing Huh7 xenograft in the GFP group developed tumor nodules, with 1 mouse dying of excess tumor burden, whereas only 1 tumor was observed in the HNF1α group at week 6 (Fig. 6C,D).

HNF1α Significantly Inhibits HCC Xenograft Growth in Mice

Based on the strong inhibitive effect of HNF1α on tumorigenesis, we further pursued the efficacy of AdHNF1α on liver cancer using two different models. First, an IT injection of AdHNF1α or AdGFP was performed at day 21 postinoculation. We found that AdHNF1α attenuated the growth of Hep3B xenograft in nude mice and, at day 21 post-treatment, resulted in a markedly reduced tumor volume and tumor weight, compared with mice receiving IT injection of AdGFP (Fig. 7A-D) (P < 0.05). Furthermore, real-time RT-PCR showed that exogenous HNF1α expression could provoke a cluster of liver-specific genes expression and decreased the expression of CD133 and EpCAM in SC xenograft models, which was consistent with results in vitro. (Supporting Fig. 11).

In addition, mice bearing orthotopic liver cancer were administered AdHNF1α or AdGFP via the tail vein. GFP expression was detected in more than 80% of hepatic cells at day 3 of adenoviruses injection in both AdGFP and AdHNF1α groups, whereas the increased expression of HNF1α in the liver was only observed in the AdHNF1α group (Supporting Fig. 12). Interestingly, we found that only 33.33% of the mice (2 of 6) treated with AdHNF1α developed tumor nodules in the liver, whereas no macroscopic or microscopic tumor in the liver was observed 3 weeks after treatment in the other mice. In contrast, 100% of the mice (6 of 6) receiving AdGFP developed gross tumor nodules in the liver, and these nodules were larger than those of mice receiving AdHNF1α (Fig. 7E). Additionally, none of the mice treated with AdHNF1α developed ascites, wheras 66.67% of the mice (4 of 6) treated with AdGFP developed severe ascites, including 3 mice with hemorrhagic ascites. Furthermore, immunohistochemical analysis revealed that AdHNF1α delivery significantly reduced the expression of proliferating cell nuclear antigen (PCNA), Ki67, CD31, and CD34 (Supporting Fig. 12).


HNF1α, considered as a marker of the epithelial phenotype of hepatocytes,23 is indispensable for hepatocytic cell lineage and liver development.3, 6 The suppression and genetic alterations of HNF1α in hepatic adenomas and HCC have been well characterized. Nevertheless, there has been no previous report on the efficacy of HNF1α in treating HCC. With this report, we demonstrated that HNF1α overexpression could restrain the proliferation of hepatoma cells and abolish their tumorigenicity both in vitro and in vivo. Most excitingly, our results clearly revealed the substantial efficacy of HNF1α in forestalling liver tumor progression in two distinct animal models.

Several studies have indicated that reduced expression of liver-enriched transcription factors is associated with the dedifferentiation of hepatocytes, which is a key early event in the pathogenesis of HCC and partly determines its progression.8, 15, 19 In the present study, we demonstrated the significantly decreased expression of HNF1α in hepatoma cell lines and experimental cancerous tissues in liver. Consistent with the report by Hellerbrand et al.,19 we also documented the reduced expression of HNF1α in human HCC tissues. Xu et al., however, found increased expression of HNF1α in 4 of 6 human HCC tissues.24 The discrepancy could be explained by the relatively small number of the samples they analyzed.

It is known that cell differentiation is defined by the activity of tissue-specific functional genes.6, 8 The expression of hepatic functional proteins in hepatocytes reflects the degree of hepatocyte differentiation, and measuring their expression can thus estimate the degree of differentiation of hepatoma cells.7, 8 In this study, we found that HNF1α overexpression up-regulated the expression of a cluster of liver function-specific genes both in vitro and in vivo. As indicators of hepatocyte maturation, these genes play important roles in the cardinal functions of the liver, including glucose and lipid metabolism, xenobiotics and drug metabolism, ammonia detoxification, and nutrient transportation. Our data suggested that forced reexpression of HNF1α in hepatoma cells could induce the differentiation of HCC cells into a more mature hepatocyte phenotype. This reversion of dedifferentiation status may underlie the anticancer effect of HNF1α. Nevertheless, the up-regulated genes in HNF1α-treated hepatoma cells were distinct from those with overexpression of HNF4α.9 This discrepancy may be attributed to the delicated regulation of HNFs on hepatocyte differentiation. The expression of HNF4α starts at the phase of embryonic development and is crucial for hepatocyte differentiation and epithelial morphogenesis,3, 4 whereas the embryonal expression of HNF1α starts 10.5 days after gestation and is essential for maintenance of liver-differentiated state and epithelial phenotype after birth.3, 7

mTOR and TGFβ/smads are both important tumorigenic signaling pathways. It has been described, by Pelletier et al., that impaired expression of HNF1α could activate the mTOR pathway and was associated with the development of hepatocellular adenomas.20 Our study further demonstrated that HNF1α overexpression could inhibit the mTOR pathway. In addition, we found that HNF1α could inhibit the activation of the TGFβ/smads pathway in hepatoma cells.

Checkpoints of cell cycle control the proper timing of cell-cycle events by enforcing the dependency of late events on the completion of early events. Consequently, checkpoint block can result in cell-cycle arrest and significantly change the activity of cell proliferation.25 Previous study has suggested that down-regulation of HNF1α with siRNA in HepG2 could activate the mTOR-signaling pathway and lead to constant accumulation of cyclin D1,20 the key cell-cycle regulatory protein related to G0/G1 arrest. Thus, we expected that up-regulation of HNF1α may reduce the expression of cyclin D1 and induce G0/G1 arrest. Nevertheless, our in vitro study revealed that HNF1α delivery induced significant G2/M arrest in all the four types of hepatoma cells, along with the suppression of cdc2 and cyclin D1, whereas no G0/G1 arrest was detected. It has been well demonstrated that G2/M transition is regulated by a complex composed of cdc2 and cyclin B1 or cyclin A, and the decrease in cdc2 levels would ultimately lead to G2/M arrest.26, 27 Therefore, the induction of G2/M arrest of hepatoma cells by HNF1α may be attributed to the reduction of cdc2. The cyclin-dependent kinase inhibitor gene, p21, a pleiotropic molecule, could inhibit cdc2 activity via directly binding to cdc2/cyclin complexes, interfering with the activating phosphorylation of cdc2, and inactivating PCNA.28, 29 Our current study found that up-regulation of HNF1α increased p21 at both transcriptional and protein levels in hepatoma cells. These findings implicate that cell-cycle arrest by HNF1α is likely the result of the induction of G2/M arrest through the accumulation of p21.

It has been reported that both miR-194 and miR-192 are highly expressed in the healthy liver.30 HNF1α could enhance the transcription of miR-194 through direct interaction with its promoter during intestinal epithelial cell differentiation.22 miR-192, the clustering partner of miR-194, may share the same regulatory sequences with miR-194. Additionally, it has been reported that these two miRNAs could induce G2/M arrest by enhancing p21 levels, and that miR-194 could suppress the metastasis of liver cancer cells in mice.21, 31 Herein, we demonstrated that the expression of miR-194 and miR-192 was closely correlated with HNF1α in human liver cancer tissues, and that the inducible expression of miR-194 and miR-192 is regulated by HNF1α. More intriguingly, miR-194 and miR-192 could activate the expression of p21, and the effect of HNF1α on cell proliferation was partly eliminated by the inhibitors of miR-194 or miR-192. Therefore, we presume that the HNF1α-induced expression of p21 was partly dependent on up-regulating the expression of miR-194 and miR-192.

Differentiation therapy is defined as inducing the differentiation of malignant cells to benign or normal cells.32 Treatment of acute premyelocytic leukemia with all-trans retinoic acid has yielded a high-quality remission and survival, represented as a classic model of differentiation therapy.33 Nevertheless, the application of differentiation therapy for other leukemia and solid malignant tumors has not obtained favorable results in clinical practice, suggesting that more effective differentiation-inducing agents are urgently needed. Our previous study has demonstrated the striking effect of HNF4α on liver cancer and hepatic fibrosis.9-11 With this study, we also revealed the similar differentiation effect of HNF1α on hepatoma cells. Therefore, we strongly believe that differentiation therapy with HNFs might present as an ideal strategy for the treatment of human HCC. Moreover, this strategy may be extended to other cancer types through the induction of differentiation using corresponding differentiation-determining transcription factors.34 On the other hand, it has been well proven that TAFs could markedly encourage the growth of HCC,35 and that tumor-associated, high-density neovascularization was responsible for the development of tumor growth.36 Therefore, the strong inhibition of the proliferation of TAFs and tumor vascularization by HNF1α might also contribute to the striking in vivo antitumor effects of HNF1α. Additionally, the suppression of CD133 and EpCAM may be in favor of the antitumor effect of HNF1α.

In conclusion, our current findings not only deepen our understanding of the biological significance of HNF1α, but also strongly implicate that HNF1α may serve as a promising agent for the differentiation therapy of HCC in clinical practice. Moreover, this study also encourages the evaluation of differentiation therapy for tumors of organs other than the liver using their corresponding differentiation-determining transcription factor.