Melittin prevents liver cancer cell metastasis through inhibition of the Rac1-dependent pathway


  • Potential conflict of interest: Nothing to report.


Melittin, a water-soluble toxic peptide derived from bee venom of Apis mellifera was reported to have inhibitory effects on hepatocellular carcinoma (HCC). However, its role in antimetastasis and the underlying mechanism remains elusive. By utilizing both HCC cell lines and an animal model based assay system, we found that Rac1, which has been shown to be involved in cancer cell metastasis, is highly expressed in aggressive HCC cell lines and its activity correlated with cell motility and cytoskeleton polymerization. In addition, Rac1-dependent activity and metastatic potential of aggressive HCC cells are remarkably high in both cellular and nude mouse models. We provide evidence here that melittin inhibits the viability and motility of HCC cells in vitro, which correlates with its suppression of Rac1-dependent activity, cell motility, and microfilament depolymerization. Furthermore, melittin suppresses both HCC metastasis and Rac1-dependent activity in nude mouse models. The specificity of the effect of melittin on Rac1 was confirmed in HCC cells both in vitro and in vivo. Conclusion: Melittin inhibits tumor cell metastasis by reducing cell motility and migration via the suppression of Rac1-dependent pathway, suggesting that melittin is a potential therapeutic agent for HCC. (HEPATOLOGY 2008;47:1964–1973.)

Hepatocellular carcinoma (HCC) is one of the most aggressive malignant tumors highly prevalent in Asia and Africa.1 Recent studies show that the incidence of HCC in the United States and the United Kingdom has increased substantially in the past 2 decades.2, 3 Despite intense efforts to improve its prognosis, the overall survival rate of patients with HCC is very low.4–6 The major obstacles to survival are metastasis and recurrence after HCC resection.7 However, the molecular mechanism for HCC metastasis remains unclear, and the means to inhibit HCC metastasis are limited.

It has been shown that HCC cell invasiveness and metastasis are regulated by multiple cues, including extracellular molecules such as cell adhesion molecules, proteases, angiogenesis factors, cytokines, and growth factors, and the underlying signaling transduction components such as Rac1. Rac1, a member of the Ras superfamily of small GTP (guanosine triphosphate)–binding protein is known to play important roles in the regulation of distinct microfilament-based structures, which is required for cell adhesion, migration, and invasion. It has been demonstrated that Rac1 can promote tumor cell migration and invasion for multiple types of cancer such as renal, breast, and liver carcinomas. Rac1 is involved in the activation of c-Jun N-terminal kinase (JNK) and JNK-dependent cell motility.8, 9 A dominant-negative form of Rac1 (Rac12 DN or Rac1 N17) blocks the changes in cell shape and the formation of adhesion complexes induced by growth factors. Conversely, microinjection of a constitutively active form of Rac1 (Rac1 DA or Rac1 V12) into fibroblasts induces changes in cytoskeleton and cell morphology in the absence of growth factors.10 Thus, the Rac1 pathway is likely to play an important role in the control of HCC cell motility and can be used as a target pathway for screening biologically active substances in the treatment of HCC.

Melittin is a water-soluble toxic peptide derived from the venom of the bee, Apis mellifera. It is a small peptide composed of 26 amino acids with a hydrophobic N-terminal region (1-20 amino acids) and a hydrophilic C-terminal region (21-26 amino acids) that has a stretch of positively charged amino acids.11 Melittin is able to disrupt membranes and exert toxic effects on H-ras transformed cells.12 It also possesses anti-HCC effects.13, 14 However, the underlying mechanism is unknown.

In this study, we screened for HCC cell lines with high expression levels of Rac1 to study the relationship between the inhibitory effect of melittin on HCC metastasis and the Rac1-mediated signaling pathway both in vitro and in vivo. We found that Rac1 plays a crucial role in the control of HCC cell motility and metastasis. Melittin prevents HCC metastasis via inhibition of Rac1.


HCC, hepatocellular carcinoma; HGF, hepatocyte growth factor; JNK, c-Jun N-terminal kinase; MKK4, mitogen activated protein kinase kinase 4.

Materials and Methods


We purchased metastatic HCC cell lines, MHCC97L and MHCC97H, from the Liver Cancer Institute, Zhongshan Hospital (Shanghai, China). All others were preserved by the Second Military Medical University (Table 1). Nude mice, purchased from Chinese Academy of Sciences (Shanghai) 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 (NIH publication 86-23 revised 1985). We obtained melittin from Sigma; control polypeptide (IETQVPEKKKGIFRR) from the core facility of the MD Anderson Cancer Center; monoclonal antibody for Rac1 from BD Biosciences; and antibodies for phospho-mitogen-activated protein kinase kinase 4 (MKK4), phospho-JNK, and phospho-extracellular signal-regulated kinase (ERK) from Cell Signaling. Antibody for phospho-c-Jun and rabbit antibody for hepatocyte growth factor (HGF) were from Santa Cruz. pRK5.RAC1 V12 (Rac1-DA), pRK5.RAC1N17 (Rac1-DN), pCMS·EGFP.RAC1V12 (Rac1-DA), and pCMS·EGFP.RAC1 N17 (Rac1-DN) have been described.15

Table 1. The Origin and Characteristics of 10 Liver Cell Lines
Cell LineOriginCharacteristics
Bel-7402Human HCC 
Hep3BHuman HCC 
HepG2Human HCC 
HuH7Human HCC 
PLC/RPF/5Human hepatoblastoma carcinoma 
SMMC7721Human HCCLow metastatic potential
MHCC97LHuman HCCLow metastatic potential
MHCC97HHuman HCCHigh metastatic potential
Chang liverNormal human hepatic cell 
LO2Normal human hepatic cell 

Cell Culture and Transfection.

We cultured cell lines in Dulbecco's modified Eagle medium (Mediatech) supplemented with 10% fetal calf serum and transfected them with Lipofectamine 2000 (Invitrogen).

Quantitative Real-Time Polymerase Chain Reaction.

We prepared total RNA with an RNeasy Mini Kit (Qiagen, Valencia, CA). We reverse-transcribed total RNA to complementary DNA using Superscript III (Invitrogen). We quantified Rac1 expression by real-time quantitative polymerase chain reaction (PCR) using the Sybergreen PCR Master Mix kit (Applied Biosystems, Foster City, CA) and the ABI-Prism 7300 System. We used the following primers: GGAGACGGAGCTGTAGGTAA and TCTCTTCCTCTTCTTCACAA. We used glyceraldehyde 3-phosphate dehydrogenase as internal control. For relative quantification, we calculated the copy ratios of Rac1 messenger RNA (mRNA) normalized to normal liver cells and used them as an indication of the relative expression levels.

XTT Cytotoxicity Assay.

We detected the cytotoxic effect with a proliferation kit (XTT II, Boehringer Mannheim, Germany). Briefly, we plated the cells in 96-well culture plates (5×104 cells/well). We added melittin at various final concentrations in quadruplicate. Later, we added 100 μL of XTT reaction solution. We read the optical density in an enzyme-linked immunosorbent assay plate reader 4 hours later.

Cell Migration and Invasion Assays.

We assayed the invasion and migration activity of cells using a transwell cell-culture chamber as described.16 For transfection experiments, we seeded cells 24 hours after transfection. For the invasion assays, we added Matrigel (BD Biosciences) to the transwell chambers and incubated it for 5 hours before we seeded the cells. We performed the migration and invasion assays in quadruplicate for each cell line tested.

Measurement of Rac1 Activity, Western Blot, and Immunohistochemistry.

We assessed Rac1 activity by pull-down assay with the Rac1-binding domain as described.17 We performed western blot and immunohistochemical staining as described.15

Establishment of Orthotopic Transplanted Nude Mice Model of Human HCC Metastasis.

We injected MHCC97H or Rac1-DA-transfected-MHCC97H cells (1 × 107 cells/animal) subcutaneously into nude mice to produce implanted tumors. We measured tumor volumes with a slide caliper; the volume = (the larger diameter) × (the smaller diameter)2/2.18 When their larger diameter grew to around 0.5 cm to 0.7 cm (20 days), we dissected the implanted tumors and cut them into pieces of around 2 × 2 × 1 mm3. We then transplanted them into the liver parcel of other nude mice.19

Antimetastasis Effect of Melittin on Nude Mouse Model with Aggressive HCC.

We randomly divided 40 nude mice into four groups as shown in Table 2. We injected groups 1 and 2 with saline via the tail vein and injected groups 3 and 4 with melittin (80 μg/kg/day).20 We treated nude mice for 10 days after orthotopic transplantation. After 35 days, we sacrificed the mice for liver tumor and lung tissue preparations.21 We measured the volume of the implanted tumors and the body weights of mice to evaluate for cachexia. We counted the number of metastases in the lungs.

Table 2. Comparison of the Changes in the Volumes of Implanted Tumors and the Weights of Nude Mice with Implanted Tumors Among Different Groups Treated with Melittin (x̄ ± S)
GroupnVolume (cm3)Body Weight (g)Melittin-Treated (80 μg·kg−1 day−1)Nude Mice ModelDay
  • *

    P < 0.05 versus group 2.

  • P < 0.01 versus group 1.

  • P > 0.05 versus group 2.

1103.73 ± 0.0223.7 ± 0.47b10
2102.47 ± 0.0426.5 ± 0.14a10
3101.90 ± 0.01*30.3 ± 1.48*+a10
4102.33 ± 0.0228.1 ± 2.07+b10

Statistical Analysis.

We determined the significance of the differences among different measured indicators with the Student t test (for normally distributed data), or the Mann-Whitney U test (for nonnormally distributed data). We performed correlation analysis by Z test.


Rac1 Expression Correlates with the Metastatic Capacity of HCC Cell Lines.

Rac1 is known to play an important role in tumor cell metastasis. It is plausible that Rac1 also regulates the metastasis of HCC cells. To test this hypothesis, we first determined Rac1 expression in the different HCC cell lines listed in Table 1. Our data shows that Rac1 mRNA levels in all hepatoma cell lines are higher than that in normal liver cell lines (Fig. 1A). The Rac1 expression in HCC cell lines with higher metastatic capacity was higher than that in cells without metastatic capacity. MHCC97H cells, which have the highest metastatic capacity,22 exhibited the highest levels of Rac1 mRNA expression, with a ratio of 297 compared with normal liver cell lines (Fig. 1A). In contrast, SMMC (Second Military Medical College) 7721, HuH7, PLC/RPF/5, MHCC97L, Hep3B, Bel-7402, and HepG2 with less metastatic capacity had lower levels of expression with ratios of 86, 67, 50, 17, 12, 10, and 9, respectively.

Figure 1.

Expression of Rac1 in different HCC cell lines. (A) Real-time PCR assay for Rac1. Values obtained on Chang liver and LO2 cell lines served as controls and the mean value of the two served as a calibrator for normalization of Rac1 mRNA levels in other HCC cells. The Rac1 mRNA steady-state levels in HCC cell lines are shown as the relative abundance of Rac1 mRNA. (B) Cell lysates from HCC cell lines were subjected to western blot analysis with anti-Rac1 antibody and the blot was reprobed for glyceraldehyde 3-phosphate dehydrogenase to verify equal loading.

To confirm this result, we further determined Rac1 protein expression in different cell lines. The Rac1 protein expression level in the MHCC97H cell line, like its transcriptional level, was markedly higher than that in other cell lines, suggesting that Rac1 expression in HCC tumor cells positively correlates with the cell's metastatic potential (Fig. 1B). We therefore chose the MHCC97H cell line as the system for our study.

Rac1 Activity Correlates with HCC Cell Migration.

Rac1 regulates cytoskeleton reorganization and cell movement and has been indicated as an important factor in the progression and metastasis of cancer cells.9, 23 We examined the relationship between cell motility and Rac1 activity/expression in MHCC97L cells and MHCC97H cells, which have different Rac1 expression levels (Fig. 1). The levels of the activated or the guanosine triphosphate–bound form of Rac1, as well as Rac1 expression, were significantly higher in the MHCC97H cells (Fig. 2A). Transfection of a dominant-negative form of Rac1 (Rac1 DN) into MHCC97H cells reduced their migratory ability (Fig. 2A, C).

Figure 2.

High levels of Rac1 activity/expression in MHCC97H cells correlates with cytoskeleton polymerization and cell motility. (A) Rac1 activity was analyzed by pull-down assay; Rac1 and PCNA protein were detected separately. (B) Confocal imaging of cytoskeleton polymerization in cells labeled with FITC-conjugated phalloidin (500 ng/mL). (C) Cell motility was assessed using a transwell chamber assay. The values (% of cells) are expressed as mean ± standard deviation (SD) from four separate measurements. t test, **P < 0.01 as compared with values obtained with MHCC97L cells. (D) Correlation between Rac1 activity and % of total migrated cells (both migrated-adherent and migrated nonadherent). Z test of correlation P = 0.0089.

Microfilaments play a key role in cytoskeleton reorganization induced cell morphological change and cell motility.8 The organization of actins and cell motility have been shown to correlate with tumor metastatic potential.24 Therefore, we examined the microfilaments in both MHCC97H and MHCC97L cells. Compared with MHCC97L cells, there was a significant amount of microfilament depolymerization in MHCC97H cells. Overexpression of Rac1-DN decreased microfilament depolymerization (Fig. 2B). Notably, cell motility was evidently different among different cells (Fig. 2C) but their proliferating cell nuclear antigen (PCNA) levels were similar (Fig. 2A, C). Furthermore, Rac1 activity correlated with the percentage of total migrated cells (both migrated-adherent and migrated-nonadherent) in the tested cell lines (Fig. 2A, D).

Melittin Inhibits HCC Cell Viability and Migration.

We have shown that melittin has an anti-HCC effect.13, 14 It is possible that melittin affects HCC via the Rac1-mediated signaling pathway. To test this hypothesis, we examined the effect of melittin on the cell viability of MHCC97L, MHCC97H, and these cells transfected with Rac1-DA or Rac1-DN by XTT. The time- and dose-dependence of the cytotoxic effect on the four cell lines are shown in Fig. 3A -C (MHCC97L cells, median inhibitory concentration (IC50) = 9.24 μg/mL; MHCC97H cells, IC50 = 4.06 μg/mL; Rac1-DA-transfected MHCC97H cells, IC50 = 3.83 μg/mL; Rac1-DN-transfected MHCC97H cells, IC50 = 25.69 μg/mL). The viability of melittin-treated MHCC97H cells was significantly lower than that of melittin-treated MHCC97L cells, while the viability of melittin-treated Rac1-DA-transfected MHCC97H cells was lowest (Fig. 3A-C). These results suggest that melittin is more effective on aggressive cell lines, especially those with high levels of Rac1 activity, than cells with lower metastasis potential.

Figure 3.

Inhibitory effects of melittin on cytotoxic and cell mobility of MHCC97L, MHCC97H, Rac1-DA, or Rac1-DN transfected MHCC97H cells. The cytotoxicity of melittin was evaluated using XTT cell proliferation assay 12 hours (A) 24 hours (B) and 36 hours (C) in the presence of various concentrations of melittin as indicated. **P < 0.01: MHCC97H versus MHCC97L, Rac1-DN transfected MHCC97H cells; *P < 0.05: MHCC97H versus Rac1-DA transfected MHCC97H cells; ΔP > 0.05: Rac1-DN transfected MHCC97H cells versus MHCC97L cells. (D) Inhibitory effect of melittin on cell motility in MHCC97L and MHCC97H cells (4 μg/mL melittin, 24 hours). The values were calculated by as a percentage of the control and expressed as the mean ± SD (*P < 0.05, **P < 0.01).

We further investigated the effect of melittin on MHCC97H and MHCC97L cell motility by transwell chamber assay. As shown in Fig. 3D, the number of migrating cells was significantly higher in MHCC97H cells than in MHCC97L cells (39 ± 1 versus 7 ± 2, P < 0.01). When treated with melittin, the migration of MHCC97H cells decreased significantly, but the migration of MHCC97L cells did not. This result suggests that melittin is more potent in suppressing the motility of HCC with high metastatic potential. In addition, our data demonstrates that melittin's anti-HCC effect depends on Rac1.

Melittin Inhibits Rac1 Activity and Microfilament Depolymerization in Cells.

To determine the relationship between the anti-HCC effect of melittin and the Rac1 signaling pathway, we evaluated the effect of melittin on Rac1 activity and expression in MHCC97H cells. As shown in Fig. 4A and E, melittin, but not a control polypeptide, suppressed Rac1 activity and expression in a dose-dependent manner without affecting PCNA expression in MHCC97H cells. In addition, melittin also inhibited cell motility (Fig. 4B, F) and reduced the amount of disorganized microfilaments in MHCC97H cells (Fig. 4C).

Figure 4.

Melittin suppresses Rac1 activity/expression as well as Rac1-dependent HCC cell motility. (A) Melittin significantly inhibited Rac1 activity/expression in a dose-dependent manner. (B) Melittin inhibited cell migration. The inhibitory effect was reversed by Rac1-DA into MHCC97H cells. The values were presented as the mean ± SD of five separate measurements. t test, **P < 0.01 as compared with values obtained with MHCC97L cells. (C) Apparently increased organization of microfilament was observed in melittin-treated MHCC97H cells but was not in MHCC97H cells expressing Rac1-DA. (D) Direct correlation between Rac1 activity and % of total migrated cells in MHCC97H cells, MHCC97H cells expressing Rac1-DA and those cells treated or untreated with 4 or 8 μg/mL of melittin as indicated. Z test of correlation. P = 0.0061. (E,F) Melittin, but not control polypeptide, prevents Rac1 expression and cell motility of MHCC97H cells, **P < 0.01: melittin treated MHCC97H versus control polypeptide treated MHCC97H cells.

To further verify the specific effect of melittin on Rac1, we introduced the constitutively active form of Rac1 (Rac1-DA) into MHCC97H, and found that it suppressed the inhibitory effect of melittin on microfilament disorganization and cell motility, as well as on Rac1 expression/activity (Fig. 4 A-C). In addition, Rac1 activity correlated directly with cell mobility in MHCC97H cells, MHCC97H cells expressing Rac1-DA and cells treated with melittin (Fig. 4D).

Melittin Prevents Rac1-JNK Activation.

JNK is one of the downstream effectors of Rac1 and it is crucial for cell migration. Thus, melittin may inhibit HCC metastasis through the Rac1-JNK pathway. We therefore determined whether JNK is involved in HCC metastasis. The levels of activated forms of JNK (P-JNK), c-Jun (P-c-Jun, a substrate of JNK), and MKK4 (P-MKK4, upstream kinase of JNK) in MHCC97H were significantly higher than those in MHCC97L cells (Fig. 5A). Furthermore, melittin inhibited the activity of MKK4, JNK, and c-Jun in MHCC97H cells in a dose-dependent manner (Fig. 5B, upper panels). In contrast, this inhibitory effect was much less obvious in cells transfected with Rac1-DA (Fig. 5B, lower panels). This suggests that the JNK pathway mediates the high metastatic potential of MHCC97H cells and further supports the notion that melittin is a Rac1 inhibitor.

Figure 5.

Rac1-dependent JNK activity is high in MHCC97H cells and it can be suppressed by melittin. (A) Cell lysates from MHCC97L, MHCC97H, and Rac1-DN-transfected MHCC97H cells were subjected to western immunoblotting with antiphosphorylated forms of MKK4, JNK, and c-Jun antisera and the blot was reprobed for glyceraldehyde 3-phosphate dehydrogenase to verify equal loading. (B) Melittin treatment leads to a dose-dependent decrease in MKK4, JNK, and c-Jun activity in MHCC97H cells, but not in Rac1-DA-transfected cells.

HGF has been shown to play a role in HCC proliferation, migration and invasion.25 It is likely that HGF can interact with its receptor, c-met, to activate Rac1 and downstream kinases including ERK and JNK.26, 27 Therefore, we investigated the effect of melittin on HGF treated HCC cells. To reinforce our study, we analyzed Rac1 expression and cell motility further in additional HCC cell lines that have relatively high levels of Rac1 expression, including SMMC7721, Huh7, and PLC/RPF/5. Melittin suppressed cell motility and Rac1 expression in aggressive HCC cells treated with HGF more obviously than in less aggressive ones (Fig. 6A -H). Interestingly, melittin could inhibit HGF expression as well as the activity of ERK (P-ERK) in MHCC97H cells (Fig. 6A). Meanwhile, HGF could also partially suppress the effect of melittin on cell migration, especially in MHCC97H cells (Fig. 6A-H).

Figure 6.

Melittin suppresses the expression and/or activity of metastasis associated proteins and cell motility in different HCC cell lines. (A) Effects of melittin on Rac1 and HGF expression and ERK activity in MHCC97H cells. The inhibitory effect of melittin on Rac1, HGF and P-ERK was partially suppressed by HGF (20 ng/mL). (B-D) Melittin inhibited Rac1 expression in SMMC7721, Huh7, and PLC/RPF/5. (E-H) Melittin inhibits cell motility of MHCC97H, SMMC7721, Huh7 and PLC/RPF/5 cells induced by HGF. The effect can be suppressed by HGF, especially in MHCC97H cells. The values are calculated by a percentage of the control and presented as the mean ± SD (*P < 0.05, **P < 0.01).

Melittin Inhibits HCC Growth and Metastasis In Vivo.

Overgrowth and metastasis are two major characteristics of malignant tumors. Pulmonary metastasis happens in 90% of HCC patients who have metastasis.28 Here, we used an orthotopic transplantation nude mouse model of human HCC metastasis LCI-D20 (100% tumorigenesis and pulmonary metastasis) to study the role and the mechanism of melittin-mediated cytotoxicity toward HCC in vivo.19

Melittin can suppress the activity of Rac1, MKK4, JNK, c-Jun, and the migration of HCC cells in vitro. We therefore investigated whether melittin could suppress HCC tumor growth and metastasis in vivo in an orthotopic transplanted nude mouse tumor model of human HCC metastasis established using MHCC97H cells (Model A) or using MHCC97H cells transfected with the constitutively activated form of Rac1 (Model B). We determined the inhibitory effect of melittin in vivo on HCC tumor growth by examining orthotopic transplanted tumor volume and nude mice body weights, as described in Table 2 and Fig. 7A,B. We found that there was a significant decrease in tumor volume accompanied by an increase in body weight in Model A nude mice treated with melittin, suggesting an inhibitory effect of melittin on HCC tumor growth and cachexia. We also introduced melittin into Model B. The constitutively activated form of Rac1 clearly reversed the inhibitory effect of melittin on HCC tumor growth and cachexia (Table 2 and Fig. 7A,B).

Figure 7.

Melittin inhibited HCC growth and metastasis in vivo. The in vivo models used were orthotopic transplanted nude mouse tumor models of human HCC metastasis established with MHCC97H cells (Model A) or with MHCC97H cells transfected with constitutively active Rac1 (Model B). (A) Effect of melittin on HCC tumor growth in Models A and B. (B) Effect of melittin on cachexia in Models A and B. (C) Melittin inhibited the lung metastasis in the nude mouse models of HCC metastasis. (D) Left panels: Rac1 expression in orthotopic liver tumor tissues of nude mouse models (10 × HE). D1to D4: Rac1 expression in orthotopic liver tumor tissues from saline-treated/Model A; melittin-treated/Model A; saline-treated/Model B and melittin-treated/Model B respectively. Right panel: Rac1 protein positive rate in orthotopic liver tumor tissues from both models with or without treatment with melittin. (E) Melittin suppressed the activity of Rac1 and the JNK pathway in orthotopic liver tumor tissues in vivo. Cell lysates were subjected to Western immunoblotting with anti-phosphorylated forms of MKK4, JNK, c-Jun and glyceraldehyde 3-phosphate dehydrogenase antisera.

To confirm the effect of melittin on HCC metastasis, we examined the metastasis in the lungs of nude mice. As expected, the lung metastasis rate in both Model A and B was almost 100%. The lung metastasis rate in nude mouse Model B without treatment with melittin was much higher than in nude mouse Model A treated with melittin (60% ± 3% versus 40% ± 4%, P < 0.05) while the lung metastasis rate was similar in the two different nude mouse models treated with isotonic saline (Fig. 7C).

To determine the relationship between Rac1 expression, HCC tumor growth, and the effects of melittin in vivo, we examined the expression of Rac1 in the orthotopic liver tumors in the nude mouse models. We found that the positive rate for Rac1 was higher in nude mouse Model B than that in nude mouse Model A (90% ± 7% versus 74% ± 3%, P <0.05) (Fig. 7D). The Rac1 positive rate was 90% ± 7% and 40% ± 3%, in nude mouse Model B treated by saline and melittin, respectively, (P < 0.01) (Fig. 7D). In nude mouse Model A, the Rac1 positive rate was 74% ± 3% and 15% ± 4% after treatment with saline and melittin, respectively (P < 0.05) (Fig. 7D). These data suggest that Rac1 is involved in HCC metastasis and Rac1 expression can be suppressed by melittin in vivo.

We further determined the effect of melittin on Rac1 activity in the orthotopic liver tumor tissues of the nude mouse models. As shown in Fig. 7E, Rac1 activity, similar to Rac1 protein expression, was suppressed by melittin in nude mouse Model A accompanied by a significantly decrease in the phosphorylation of MKK4, JNK, and c-Jun, as well as HCC metastasis to the lung. To support the hypothesis that Rac1 is the target of melittin, we repeated the above experiments in nude Mouse model B. The presence of the constitutive activated form of Rac1 in Model B led to less sensitivity of HCC cells to melittin. In addition, the Rac1 expression/activity and MKK4, JNK, and c-Jun activities, as well as cell metastasis to the lung were also less affected (Fig. 7A-E).


Tumor metastasis is a complex process involving extensive interactions between tumor cell and host tissues. Tumor cells with high metastatic potential are more motile than nonmetastatic cells.29 Rac1 plays an important role in cell motility as well as tumor metastasis. However, its role in hepatoma has not been well studied, especially in animal models.

In this study, we have demonstrated that Rac1 plays a crucial role in hepatoma metastasis based on the following evidence. Both transcriptional and protein levels of Rac1 are higher in more aggressive HCC cell lines such as MHCC97H cells. In addition, Rac1 activity regulates cell motility and cytoskeleton polymerization in different HCC cells. Furthermore, Rac-dependent JNK and ERK activity and metastasis for aggressive HCC cells are strikingly high in both cellular and nude mouse models. Finally, Rac1-DA-transfected MHCC97H and Rac1-DN-transfected MHCC97H cell lines showed increased and decreased cell motility and metastasis respectively in both cellular and nude mouse models. All this evidence indicates that Rac1 is an important regulator for the metastasis of hepatoma cells.

Melittin suppressed cell motility in aggressive HCC cells more obviously than that in less aggressive cells even in the presence of HGF. Melittin inhibited cell motility accompanied by a decrease in Rac1, ERK, and JNK activity, suggesting that melittin acts through the suppression of Rac1-dependent pathways. HGF could partially suppress the inhibitory effect of melittin on HCC cell motility, as well as on the activity of Rac1, ERK, and JNK. This indicates the crucial role of HGF in the metastasis of HCC cells. Melittin induced a very apparent decrease in migration of hepatoma cells and, to a lesser degree, cell viability, which could independently affect cell invasion/migration. Since PCNA levels were not affected, this then suggests that melittin does not affect cell proliferation significantly. Therefore, it is likely that the major effect of melittin on HCC is inhibition of cell migration. Interestingly, melittin failed to inhibit cell motility and JNK activity in Rac1-DA-transfected MHCC97H cells, indicating that melittin targets Rac1 and inhibits Rac1-induced cell motility.

Melittin could inhibit HCC orthotopic transplantation tumors and cachexia in vivo. Introduction of a constitutively activated form of Rac1 could reverse the inhibitory effect of melittin on cachexia and orthotopic transplantation tumors in the process of HCC invasion. In addition, the lung metastasis rate was significantly decreased in the melittin-treated nude mouse model LCI-D20. Furthermore, in this nude mouse model, the activities of both Rac1 and JNK, were suppressed noticeably by melittin in orthotopic tumor tissues. However, similar to our in vitro studies, melittin had only limited effects on Rac1 expression, Rac1 and JNK activity in orthotopic tissues, or on the orthotopic transplantation tumor volume, cachexia, and lung metastasis rate in a nude mouse model established with MHCC97H cells expressing a constitutively active Rac1. This further supports our hypothesis that melittin specifically inhibits Rac1-dependent JNK pathway activation.

Although we have shown here that the inhibitory effect of melittin on human HCC growth in nude mice is very promising, administration of high doses of melittin in vivo has its side effects, particularly liver injury and hemolysis. Considering that HCC usually develops in a background of chronic liver injury and impaired liver function, caution will be required in the clinical application of melittin. Zhao et al.,30 have shown that mutation of Val 5 to Arg, Ala15 to Arg, and deletion of Leu15 in melittin significantly reduces its adverse side effect of hemolysis but retains its antibacterial effect. We are trying to modify melittin further in order to maintain or even improve its antitumor function while reducing its side effects.

We have shown previously that POSH (plenty of SH3 domain) cooperates with JIPs (JNK interacting proteins) to serve as scaffold proteins that interact with the JNK pathway components including Rac1, MLKs, MKK4/7, and JNKs to form a complex (POSH and JIP apoptotic complex, PJAC) to mediate JNK pathway activation.31 In addition, we demonstrated PJAC is regulated through a self-amplifying feed-forward loop mechanism involving protein phosphorylation and stabilization.32 Therefore, it will be intriguing to investigate, in the future, how Rac1 is inhibited by melittin and whether melittin can affect PJAC formation as well as its related biological functions.

In summary, our present study demonstrates that Rac1 is a critical regulator for the metastasis of HCC cells both in vitro and in vivo. Rac1 expression appears to correlate with the metastatic potential of tumor cells. We also show that melittin can inhibit cell motility drastically and prevent HCC metastasis in vivo through the suppression of the Rac1-dependent pathway. Furthermore, our study concludes that melittin is a potential novel drug for HCC.


We thank Drs. H. Li, S. Perrett, and C. Yang for their comments and suggestions during the preparation of this manuscript.