ShRNA-Targeted MAP4K4 Inhibits Hepatocellular Carcinoma Growth

Hepatocellular carcinoma (HCC) is the third leading cause of cancer mortality worldwide, with more than 600,000 new cases diagnosed annually (1). The most frequent risk factors for HCC include chronic viral hepatitis, alcohol abuse, and dietary exposure to aflatoxins (2–5). Early HCC is potentially curable by resection, liver transplantation, or percutaneous treatment, and 5-year survival rates of up to 60% to 70% can be achieved in well-selected patients (6). Unfortunately, most patients are diagnosed at the advanced stage in developing countries where regular tumor surveillance has not been implemented and therefore precluded from the curative treatments. For patients with advanced HCC, the multikinase inhibitor sorafenib is available as a systemic therapy, offering a significant survival advantage (7, 8). Despite significant progress, advanced HCC still remains incurable. The identification of molecular markers and pathways that contribute to the aggressive phenotype of HCC is obviously essential for development of more effective, targeted therapies.

Mitogen-activated protein kinases (MAPK) are a family of conserved serine/threonine protein kinases that function to transmit extracellular signals into the cytoplasm. The different MAPKs can be divided into 3 main subgroups, that is, the extracellular signal-regulated protein kinase (ERK), c-Jun N-terminal kinase/stress-activated protein kinases (JNK/SAPK), and p38 (9). MAPKs are commonly activated through phosphorylation cascades, and subsequently phosphorylate a number of effectors that contribute to the regulation of diverse cellular events, including differentiation (10, 11), proliferation (12, 13), survival (14), migration (15), and invasion (16). MAPK kinase kinase kinase isoform 4 (MAP4K4), a germinal centre protein kinase that belongs to the mammalian STE20/MAP4K family, is implicated in the activation of the JNK/SAPK pathway (17, 18). Recent studies (17, 19) have documented that MAP4K4 is overexpressed in many types of human cancer. The prognostic significance of MAP4K4 overexpression has been shown in patients with stage II pancreatic ductal adenocarcinoma (19). Silencing of MAP4K4 using small interfering RNAs (siRNA) causes reduced tumor cell motility in ovarian carcinoma-derived cell line, SKOV-3 (17). These findings suggest an involvement of MAP4K4 in cancer progression. However, little is known about its expression pattern, clinical relevance, and biological function in HCC.

Here, we showed that MAP4K4 is frequently overexpressed in HCC tissues as well as several HCC cell lines including HepG2 and Hep3B, and that this upregulation is significantly associated with worse prognosis of HCC patients. Short hairpin RNA (shRNA)-mediated silencing of the MAP4K4 in HepG2 and Hep3B cells decreased cell proliferation, blocked cell cycle progression at S phase, and increased spontaneous apoptosis. In vivo studies further showed that MAP4K4 silencing dramatically retarded xenograft tumor growth in nude mice. Using a quantitative real-time array, we found that the antitumor effects of MAP4K4 downregulation were likely associated with the inactivation of multiple tumor progression–related signaling pathways, including JNK, NFκB, and toll-like receptors (TLR). Together, our data highlight an important role for MAP4K4 in controlling HCC progression.

Cell culture, construction of shRNA-expressing stable cell lines, and assay for proliferation, colony formation, and apoptosis were performed as described in Supplementary Methods.

Analysis of gene expression by quantitative reverse transcriptase polymerase chain reaction (qRT-PCR), Western blotting, and immunohistochemistry were conducted using standard procedures detailed in Supplementary Methods.

The study protocol was approved by the Ethics Committee of Second Military Medical University (Shanghai, China), and written informed consent was obtained from each patient. Tumor samples from resection specimens were collected from 2 consecutive cohorts of patients with HCC, who underwent surgical resection for the disease at Changhai Hospital and Institute of Liver Diseases (Shanghai, China) between August 2007 and October 2008 for cohort A and between January 2001 and December 2002 for cohort B. The patients of both cohorts were selected on the basis of (a) distinctive pathologic diagnosis of HCC, (b) receiving curative resection, defined as macroscopically complete removal of the neoplasm, and (c) availability of detailed clinicopathologic data. Patients with preoperative anticancer treatment or with evidence of other malignancies were excluded from the study. Cohort A consisted of 20 patients (14 men and 6 women; median age, 51 years; range, 29–76 years), from whom fresh tumor samples coupled with adjacent nontumor liver tissues were obtained for analysis of MAP4K4 gene expression. After removal at surgery, all the fresh tissues were cut into small pieces, snap frozen in liquid nitrogen immediately, and stored at −80°C until RNA or protein extraction. Cohort B comprised 400 patients (286 men and 114 women; median age, 50 years; range, 27–78 years), whose paraffin-embedded tumor specimens were available at the surgical pathology archive of Changhai Hospital, allowing for immunohistochemical analysis.

The clinicopathologic characteristics of both the study cohorts are summarized in Supplementary Table S1. In the following text, we focus on the description of the patients of cohort B, who were selected for further elucidation of the clinical significance of MAP4K4 expression. Of the 400 studied patients, 305 (76%) and 245 (61%) patients were positive for hepatitis B surface antigens (HBsAg) and hepatitis B e antigen (HBeAg), respectively. Liver cirrhosis was found in 275 (69%) cases. Seventy-eight patients (20%) showed alcohol abuse and 12 (3%) had non-alcoholic steatohepatitis. All the subjects were negative for hepatitis C virus RNA. Patients with large (>2 cm) tumors comprised 86%, and intrahepatic metastasis was seen in 62% patients. Tumors were staged according to the 2002 tumor-node-metastasis (TNM) classification system (20). A total of 85 (21%) patients were of stage I or II, and 315 (79%) of stage III or IV. Histologic grading of tumors was made according to the 2000 World Health Organization histologic classification of tumors of the liver and intrahepatic bile ducts (well, n = 95; moderate, n = 236; poor, n = 69). Of the 400 patients, 78 (19%) were lost to follow-up. The median follow-up time was 36 months (range, 6–66 months). By the end of follow-up, 249 patients (62%) died. Tumor recurrence was diagnosed on the basis of serum alpha-fetoprotein (AFP) elevation, imaging techniques including computed tomography, ultrasound, and magnetic resonance imaging, as well as histology.

For tumorigenicity assays, 4-week-old, male BALB/c nude mice were purchased from Experimental Animal Center of Shanghai, Shanghai, China. Mice were randomly divided into 2 groups (n = 4 for each group) to receive s.c. injection of 2×10⁶ HepG2 transfectants expressing MAP4K4-A1 or scrambled control shRNAs. Tumor volumes were measured every 5 days with a caliper and calculated according to the formula: 0.5 × length × width². At 30 days after the cell inoculation, mice were sacrificed, and all tumors were excised and processed for immunohistochemical staining. All experimental manipulations were undertaken in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals, with the approval of the Scientific Investigation Board of the Second Military Medical University, Shanghai, China.

Human TLR Signaling Pathway RT² Profiler PCR Array (SABiosciences) was used to assess the molecular signaling pathways affected by MAP4K4 knockdown. This technique combines together the quantitative performance of real-time PCR and the multiple gene profiling capability of microarrays. All steps were done according to the manufacturer's protocol for the ABI PRISM7900 system (Applied Biosystems). Briefly, HepG2 transfectants expressing MAP4K4-A1 or scrambled shRNAs were seeded in 6-well plates at a density of 10⁴ cells per well. At 72 hours after culture, cells were harvested and total RNA was isolated using TRIzol for real-time RT-PCR analysis. The cycling conditions were as follows: initial denaturation at 95°C for 10 minutes and then 40 cycles of denaturation at 95°C for 5 seconds, annealing at 60°C for 1 minute, and elongation at 72°C for 15 seconds. Data normalization was based on correcting all Ct values for the average Ct values of several constantly expressed housekeeping genes present on the array (21). Three independent experiments were performed with similar results, and the mean fold changes of gene expression are shown in Supplementary Table S2.

All statistical analyses were carried out using the SPSS.11 software. If not otherwise stated, mean of at least 3 independent experiments ± SD are shown. Using classic criteria (significance, 0.05; power, 0.8; 2-sided test), the sample size calculation revealed that 99 patients per group were required to detect an expected 1.5-times higher risk of death for the MAP4K4-H group than the MAP4K4-L group (22). Given the possibility of imbalance between the 2 groups, some more patients were included in the present study. Significance between controls and treatment samples was calculated by Student's t test. The relationship between MAP4K4 protein expression and clinicopathologic features of HCC was analyzed using the Pearson χ² test. Because early tumor recurrence (i.e., recurrence within 2 years after surgery) more likely occurs due to primary tumor dissemination rather than de novo carcinogenesis arising from cirrhotic liver (23), in this study we used early tumor recurrence as an endpoint. Time to recurrence was calculated from the date of surgery to the date of first documented recurrence; patients without recurrence within 2 years were censored. Another primary endpoint was overall survival, which was calculated from the date of diagnosis to the date of death or the date of last follow-up (if death did not occur). Overall survival and time to recurrence were calculated using the Kaplan–Meier method. The differences between patient groups were analyzed using the log-rank test. The prognostic significance of clinicopathologic factors was determined using univariate Cox regression analysis. Multivariate survival analysis was done with a Cox proportional hazards model. Statistical significance was assumed for a 2-tailed P value less than 0.05.

We first determined the MAP4K4 expression in 20 pairs of fresh HCC and adjacent non-tumor samples by qRT-PCR and Western blot analyses. As illustrated in Figure 1A, the expression of MAP4K4 was significantly enhanced in HCC tissues both at mRNA and protein levels relative to non-tumor tissues. To further examine the frequency and the localization of MAP4K4 protein expression, we carried out an immunohistochemical analysis on additional, larger sample sets. Notably, MAP4K4-H was detected in 194 (48.5%) of 400 HCCs. Surrounding non-neoplastic liver tissues displayed less cytoplasmic staining for MAP4K4, whereas HCC tissues showed variable proportions of MAP4K4-positive cells that gradually increased with the grades of differentiation from well differentiated to poorly differentiated (Fig. 1B). Additionally, we also analyzed the MAP4K4 expression in several human HCC cell lines (HepG2, Huh7, Hep3B, and SMMU7721). MAP4K4 was found to be expressed at higher levels in HepG2 and Hep3B than in Huh7 and SMMU7721 cells (Fig. 1C). The results indicate that MAP4K4 is aberrantly upregulated in a subset of HCC tumors.

Next, we investigated whether MAP4K4 overexpression was associated with clinical characteristics or survival of HCC patients. A significantly higher proportion of patients in the MAP4K4-H group was positive for HBsAg (P = 0.033) and HBeAg (P = 0.022) compared with that in the MAP4K4-L group. High MAP4K4 expression was significantly correlated with tumor size (P = 0.006), histologic grade (P < 0.001), intrahepatic metastasis (P = 0.002), and tumor stage (P = 0.012; Table 1). The patients in the MAP4K4-H group had a shorter median overall survival (30 versus 48 months, P < 0.0001) than those in the MAP4K4-L group (Fig. 1D). The early recurrence rate was significantly (P = 0.0001) higher in the MAP4K4-H group than in the MAP4K4-L group (Fig. 1D). A total of 12 clinicopathologic variables were included in the univariate analysis to study their correlations with patients’ survival, including sex, age, serum AFP level, serum HBsAg, serum HBeAg, tumor size, histologic grade, liver cirrhosis, capsule, intrahepatic metastasis, tumor stage, and MAP4K4 expression. As shown in Table 2, serum HBsAg (P = 0.036 and 0.008), serum HBeAg (0.025 and 0.007), intrahepatic metastasis (P = 0.043 and 0.032), tumor stage (P = 0.026 and 0.013), and MAP4K4 expression (P = 0.019 and 0.018) were significantly associated with the overall survival and early recurrence, respectively. The multivariate survival analysis revealed that MAP4K4 expression was a predictor of overall and early recurrence (P = 0.019 and 0.018, respectively) independent of serum HBsAg and HBeAg status, intrahepatic metastasis, and tumor stage (Table 3).

To explore the biological significance of MAP4K4 in HCC, we specifically knocked down its expression in HepG2 cells using RNA interference (RNAi). This HCC cell line was chosen because of its high abundance of MAP4K4. The transfection of HepG2 cells with MAP4K4-A1 shRNA caused an approximately 75% downregulation of MAP4K4 transcript levels compared with control shRNA-transfected cells (Supplementary Fig. S1A). Moreover, the amounts of MAP4K4 protein were considerably reduced in cells expressing MAP4K4-A1, showing efficient knockdown (Supplementary Fig. S1B). Similar results were observed in MAP4K4-A3–producing cells, although the extent of MAP4K4 downregulation was smaller. However, MAP4K4 expression levels were little affected by the presence of MAP4K4-A2 shRNA.

We next studied the impact of MAP4K4 silencing on HepG2 cell proliferation in vitro. The results of the MTT assay showed that downregulation of MAP4K4 by MAP4K4-A1 and -A3 significantly reduced the proliferation rate of HepG2 cells compared with cells expressing MAP4K4-A2 or control shRNA (P < 0.01; Fig. 2A). Colony formation assay further showed that MAP4K4-A1–mediated and MAP4K4-A3–mediated knockdown of MAP4K4 resulted in a substantial decrease in the HepG2 cell growth in comparison with scrambled shRNA-expressing cells (P < 0.05; Fig. 2B). However, MAP4K4-A2 did not significantly affect the number of colonies. Cell cycle analysis revealed that MAP4K4 silencing caused a considerable inhibition of cell cycle progression, leading to a selective accumulation of cells in the S phase compared with control shRNA transfectants (Fig. 2C; Supplementary Table S3). Accompanying with S phase arrest, a characteristic sub-G₁ peak, indicative of DNA fragmentation and a hallmark of apoptosis, was seen in cells with downregulated MAP4K4. Quantification of apoptosis by Annexin V/PI double labeling indicated that a remarkably higher apoptotic index was detected in MAP4K4-A1 and -A3 transfectants relative to control cells (Fig. 2D; Supplementary Table S3). By contrast, the delivery of MAP4K4-A2 had no significant influence on cell apoptosis. The knockdown of MAP4K4 expression was also conducted in Hep3B cells, another HCC cell line with high endogenous levels of MAP4K4. Similar results were achieved, that is, decreasing cell proliferation and increasing cell apoptosis by MAP4K4 downregulation (Supplementary Fig. S2). These findings collectively highlight an important role for MAP4K4 in regulation of HCC cell growth and survival.

To get more insight into the relevance of MAP4K4 silencing in vivo, an equal number of HepG2 transfectants producing MAP4K4-A1 or control shRNA were injected s.c. into the left flank of nude mice. Consistent with the in vitro studies, MAP4K4 downregulation by MAP4K4-A1 significantly slowed the xenograft tumor growth compared with the control shRNA (P < 0.01; Fig. 3A). Expression of MAP4K4 in the xenografts was determined by immunohistochemistry. Strong cytoplasmic MAP4K4 staining of cells in the control xenografts was observed (Fig. 3B). In contrast, the MAP4K4-A1 transfectant-derived tumor xenografts displayed a considerably lower proportion of MAP4K4-positive cells.

Having identified MAP4K4 as a potential oncogene in HCC, we next sought to investigate its downstream signaling pathways using the human TLR signaling pathway PCR array. This array was chosen because it can profile the expression of multiple genes related to the TLR, NFκB, and JNK/p38 signaling pathways, which contribute to HCC progression (24–27). The most differentially expressed genes whose expression changed 2-fold or more in MAP4K4-A1–producing HepG2 cells compared with control shRNA-expressing cells are summarized in Supplementary Table S4. Among the 54 genes, 53 (98.1%) were downregulated and only 1 (1.9%) was upregulated. The downregulated genes included MAP3K1, MAP2K4, MAPK8 (JNK), FOS, ELK1, NFκB1, NFκB2, TLRs, MyD88, and TRAF6, suggesting an important role for MAP4K4 in the regulation of the JNK, NFκB, and TLR signaling pathways. To confirm the qRT-PCR array data, we performed Western blot analysis in HepG2 cells harboring MAP4K4-A1 or control shRNA. Although the total protein amounts of JNK and NFκB were little affected by MAP4K4-A1, phosphorylation of these 2 proteins was significantly disrupted in MAP4K4-A1 transfectants relative to the control cells (Fig. 3C). In contrast, the phosphorylation levels of ERK1/2 or p38 MAPK were comparable between the MAP4K4-A1–expressing and control shRNA-expressing cells, suggesting that MAP4K4 selectively regulates the JNK rather than ERK1/2 or p38 MAPK pathway. Additionally, we found that the expression of TLR5, TLR4, MyD88, and TRAF6 was also downregulated in MAP4K4-A1–transfected cells (Fig. 3C). These data suggest that in HCC, multiple tumor progression–related signaling pathways including JNK, NFκB, and TLRs were coordinated by MAP4K4.

Recent studies have shown that MAP4K4 levels are elevated in many solid neoplasms, including colon cancer, lung cancer, ovarian cancer, kidney cancer, and brain cancer (17, 28). MAP4K4 overexpression has been reported to be associated with worse prognosis and to serve as a prognostic marker for stage II pancreatic ductal adenocarcinomas (19). These findings imply an oncogenic role for MAP4K4; however, little is known about its expression pattern and biological significance in HCC, one of the most common malignancies worldwide. In this study, we showed that MAP4K4 expression determined by qRT-PCR and Western blotting is upregulated in HCCs relative to adjacent nontumor liver tissues. Immunohistochemical analysis performed in a large cohort of archival HCC specimens revealed that high MAP4K4 expression is significantly correlated with tumor size, histologic grade, and intrahepatic metastasis. Most interestingly, MAP4K4 overexpression predicts poor prognosis of HCC patients. To extend our clinical studies and investigate its biological function, we generated stable MAP4K4-downregulated cell lines in HepG2 cells by delivery of its specific shRNA. Knockdown of endogenous MAP4K4 attenuated in vitro proliferation of HepG2 cells and reduced the growth of HepG2-derived xenografts in mice. These data strongly suggest that MAP4K4 is a key player in HCC progression.

The anticancer effects of MAP4K4 downregulation are associated with inhibition of cell cycle progression and induction of apoptosis. MAP4K4 silencing caused a cell cycle arrest at the S phase coupled with increased apoptosis in HepG2 cells. This phenomenon coincides with inactivation of several tumorigenesis-related signaling pathways, including JNK, NFκB, and TLRs. Despite lack of direct evidence, the antitumor effects of MAP4K4 depletion is likely mediated through the JNK pathway. This hypothesis is supported by several indirect observations. First, similar to MAP4K4, JNK expression is elevated in human HCCs (25, 29). Second, sustained JNK activity contributes to hepatocarcinogenesis. Inhibition of JNK by pharmacologic inhibitors or siRNAs reduces HCC cell growth and sensitizes cells to apoptosis in vitro and in vivo (24, 25, 29, 30). Third, our data and previous studies (17, 18) show that MAP4K4 regulates the JNK activation. In contrast to JNK, the other 2 MAPK pathways, ERK and p38, are not affected by MAP4K4 inhibition, implying a unique mediator role for JNK in MAP4K4 signaling transduction.

Besides JNK, another 2 MAP4K4-repressed pathways, TLR and NFκB, are also involved in tumorigenesis (26, 27). NFκB has been shown to link inflammation to cancer (31–33). In different contexts, NFκB displays opposing roles in carcinogenesis. Inhibition of NFκB activity by chemical reagents has yielded anti-inflammatory and anticancer effects (34). However, aberrant inactivation of NFκB through deletion of its essential modulator (NEMO, NF-κB essential modulator) triggers chronic inflammation and spontaneous HCC development (31). Likewise, TLRs, key players in innate immunity to pathogens and ubiquitously expressed on cancer cells, also play a complex role in regulating inflammation and tumorigenesis (35, 36). Given the regulatory role in the activation of the TLR and NFκB pathways, we assume that MAP4K4 might represent a promising therapeutic target for chronic liver inflammation and associated cancer progression. Indeed, the anti-inflammatory activities of MAP4K4 depletion have been documented by a recent study, in which orally delivered siRNA targeting macrophage MAP4K4 suppresses systemic inflammation (37).

In conclusion, we identified a frequent overexpression of MAP4K4 in HCCs. This overexpression is an independent predictor of adverse prognosis, and contributes to tumor growth through coordinating several signaling pathways involved in cancer progression, thereby providing a potential therapeutic target for HCC.

No potential conflicts of interest were disclosed.

This work was supported in part by grants from National Nature Science Foundation of China (No.81072020), the Shanghai Natural Science Foundation of China (No. 08ZR1419600) and Key Research Foundation of Shanghai Municipal Education Commission of China (No. 08ZZ123) to S.H. Zhang, and from the Jiangxi Natural Science Foundation of China (No. 2009GZY0161) to A.W. Liu.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.