Original Article | Published:

Inhibition of angiogenesis and HGF-cMET-elicited malignant processes in human hepatocellular carcinoma cells using adenoviral vector-mediated NK4 gene therapy


NK4 is an hepatocyte growth factor (HGF)-antagonist and a broad angiogenesis inhibitor. NK4 gene therapy has confirmed antitumor efficacy on cancers with intact HGF-cMET signaling pathway. However, the feasibility to treat tumors in which the effect of the HGF-cMET signaling pathway is less unambiguous or may even be inhibitory on carcinogenesis, such as hepatocellular carcinoma (HCC) with NK4 needs further assessment. Therefore, we evaluated the effects of adenoviral vector-mediated expression of NK4 on the biological behavior of a series of HCC cell lines in vitro and on HepG2 xenografts in vivo. In vitro, transduction of HCC cell lines with the replication-deficient recombinant adenoviral vector AdCMV.NK4 resulted in significant inhibition of proliferation over and above the antimitogenic effects of HGF. In addition, HGF-induced scattering and invasion through matrigel were inhibited effectively. Moreover, transduced HCC cells produced sufficient amounts of NK4 protein to achieve bystander effects involving reduced migration of nontransduced tumor cells and reduced proliferation of endothelial cells. Finally, treatment of established HepG2 xenografts with AdCMV.NK4 resulted in significant tumor growth delay and significant reduction of intratumoral microvessel density. In conclusion, NK4 gene therapy is a promising strategy to treat HCC based on the pleiotropic functions of NK4 interfering with tumor growth, invasion, metastasis and angiogenesis.


Human hepatocellular carcinoma (HCC) is a highly malignant disease with a poor prognosis. To date, surgical resection is the only potentially curative therapy for HCC. However, the high recurrence rate with intrahepatic metastatic spread is a major obstacle for improving survival of patients.1 Thus, development of novel therapeutic strategies targeting the malignant behavior of HCC cells, especially their invasiveness, is important to improve the prognosis of patients.2 With respect to this, the hepatocyte growth factor (HGF) antagonist and angiogenesis inhibitor NK4 has been demonstrated to effectively inhibit multiple cellular processes important for the malignant behavior of tumors. These processes include primary tumor growth, metastasis formation and angiogenesis.3, 4, 5, 6 NK4, which consists of the N-terminal hairpin domain and the four subsequent kringle domains of HGF, competitively inhibits the multiple biological activities of HGF through binding to cMET and is suggested to have a broad antiangiogenic activity through its kringle domain structure.3, 4, 7 Previous in vivo studies confirmed therapeutic efficacy of NK4 gene therapy for glioblastoma, prostate, breast, gastric, pancreatic, gall bladder and colon cancers8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20 and recommended further preclinical evaluation of NK4 gene therapy as anticancer agent. However, NK4 gene therapy for treatment of HCC could be argued against because the antitumor effects of NK4 in part rely on its antagonistic activity to HGF3 and the role of the HGF-cMET system in hepatocarcinogenesis has not been clearly established. On one hand, cMET and HGF have been reported to promote HCC development and progression. cMET was shown to be overexpressed in liver tumors compared to nontumorous tissue.21, 22, 23, 24 This cMET overexpression correlated with increased occurence of metastases and shorter 5-year survival time for patients.25 In addition, HGF was found to promote development of HCC in transgenic mice by HGF-cMET autocrine signaling in tumor cells and by stimulating angiogenesis.26, 27 On the other hand, HGF has been identified as an inhibitor of hepatocarcinogenesis. HGF exerted antiproliferative activity on HCC cell lines28, 29, 30, 31 and inhibited growth of HCC tumors in animal models.32, 33, 34

In this study, we explored the therapeutic efficacy of NK4 gene therapy for HCC in vitro and in vivo. To this end, we employed adenoviral vector-mediated NK4 gene transfer on three HCC cell lines that respond to HGF with enhanced migration, but reduced proliferation.31 We report antimotogenic and antimitogenic activity of NK4 on all three HCC cell lines. In addition, NK4 gene transfer inhibited angiogenesis and delayed growth of established HepG2 xenografts in nude mice.

Materials and methods

Cell culture

HepG2, Hep3B and HuH7 hepatocellular carcinoma cell lines, SaOS-2 osteosarcoma cell line and LNCaP prostatic carcinoma cell line were obtained from the American Type Culture Collection (Manassas, VA). All cell lines were maintained in DMEM-F12, supplemented with 10% fetal calf serum (FCS), L-glutamin, penicillin and streptomycin at 37°C in a 5% CO2-humidified atmosphere. The human umbilical vein endothelial cell (HUVEC) line EC-RF2435 was cultured in 1%-gelatin-coated culture flasks and in culture medium consisting of Medium199 supplemented with 10% FCS, 10% human serum, penicillin, streptomycin, L-glutamine and endothelial cell growth factor (ECGF). All culture media and supplements were purchased from Life Technologies, Inc., B.V., Breda, The Netherlands. The status of cMET and HGF expression of the cells was confirmed by RT-PCR as previously described36, 37 using U118MG cells, known to express cMET and HGF,38 as positive control.

Adenoviral vector

AdCMV.NK4. is an E1/E3-deleted replication-deficient adenovirus type 5 (Ad5) vector, which expresses myc-tagged NK4 protein and GFP, both under the control of a CMV promoter.12 AdCMV.GFP is the control Ad5 vector, which only contains GFP under the control of a CMV promoter. Both viruses were propagated on the permissive 293 cell line (obtained from ATCC) and purified by CsCl gradient banding using standard techniques. Particle titers were determined by OD260 measurements and functional plaque forming units (PFU) titers were determined by limiting dilution plaque titration on 293 cells according to standard techniques. Infectious units (IU) titers were determined by limiting dilution titration on 911 cells and hexon staining using the Adeno-X™ Rapid Titer Kit according to the instructions of the manufacturer (BD Biosciences Clontech, Alphen aan de Rijn, The Netherlands). All studies were performed with a single batch of AdCMV.NK4 and AdCMV.GFP. All virus aliquots were stored at −80°C until use.

Preparation of conditioned medium and measurement of HGF, NK4 and VEGF concentrations

HepG2 cells were seeded in 24-well plates at 2 × 105 cells/well. After 24 hours, cells were transduced with AdCMV.NK4 or AdCMV.GFP at an MOI of 5, 20, 50, 200 and 500 IU/cell. After 1 hour virus was replaced by 300 μl standard culture medium. Mock-infected cells were taken along as controls. After 72 hours the conditioned media were collected and centrifuged. Conditioned medium was used for scatter assays, and HUVEC proliferation assay. Concentrations of NK4, HGF and vascular endothelial growth factor (VEGF) were determined by ELISA (Quantikine, R&D systems, Minneapolis, MN) according to the instructions of the manufacturer. Supernatants of three independent experiments performed in triplicate were measured.

Scatter assay

Cells were plated at 1 × 103 cells/well in 96-well plates and subsequently transduced with AdCMV.NK4. or AdCMV.GFP at an MOI of 150 IU/cell. At 24 hours postinfection, cells were incubated with culture medium supplemented with or without 10 ng/ml HGF (R&D systems, Minneapolis, MN), or conditioned media of HepG2 cells, collected 72 hours after transduction with AdCMV.NK4 or AdCMV.GFP at an MOI of 200 IU/cell and supplemented with or without 10 ng/ml HGF. After 24 hours, cells were fixed with 3.7% formaldehyde in PBS and stained with 1% crystal violet in 70% EtOH. Scattering was assessed using a light microscope and photographed.

Multicellular spheroid outgrowth model

Multicellular tumor spheroids were formed using a spin method. Single-cell suspensions were rotated at 140 rpm for 24 hours at 37°C, 5% CO2 at 1 × 104 cells/well in a 96-well plate coated with 2% agarose, and further cultured for 3–5 days to establish spheroids. The spheroids were transduced with 2x106 IU AdCMV.NK4 or AdCMV.GFP for 24 hours in 50 μl culture medium. Subsequently, the spheroids were transferred to 24-well plates containing 500 μl culture medium supplemented with or without 25 ng/ml HGF. Mock-infected spheroids were taken along as controls. Spheroids were allowed to attach to the bottom and outgrowth of cells was followed. After 9 days, cells were fixed using 3.7% formaldehyde in PBS and stained with 1% crystal violet in 70% EtOH. Outgrowth was assessed using a light microscope and photographed.

Tumor cell invasion assay

Cells were transduced with AdCMV.NK4. or AdCMV.GFP at MOI 100 IU/cell or mock-infected for 24 hours. The next day, 2x104 cells were seeded on top of the upper chamber of transwell chambers (Corning Costar, Corning, NY), equipped with a 6.5-mm-diameter polycarbonate filter (pore size 8 μm), and precoated with matrigel (4 mg/ml) (BD Biosciences, Bedford, MA). In all, 900 μl of culture medium supplemented with or without 20 ng/ml HGF was added into the lower chamber. After 96 hours incubation at 37°C 5% CO2, the noninvasive cells (upper chamber) were removed with cotton swabs. Tumor cell invasion was determined by visualizing cells that migrated through the filter by staining (Diff Quick staining set; Dade-Behring AG, Düdingen, Switzerland) and counting under an inverted microscope (five fields/filter at × 32 magnification).

Cell proliferation assay

Cells were seeded in 96-well plates at 1 × 103 cells/well (proliferating population) or 3 × 104 cells/well (confluent population). After 24 hours, cells were transduced with AdCMV.NK4 or AdCMV.GFP at an MOI of 100 IU/cell in 50 μl. After virus incubation at 37°C for 1 hour, medium was replaced by culture medium containing HGF concentrations ranging from 0 to 20 ng/ml. Mock-infected cells were taken along as controls. After 120 hours, relative cell growth was determined by WST-1 conversion assay (Roche Diagnostics, Mannheim, Germany) according to the manufacturer's instructions.

HUVEC proliferation assay

HUVECs were seeded at 1 × 103cell/well in 100 μl medium in 96-well tissue culture plates. After 24 h, 100 μl of 10-times diluted conditioned medium (CM)-samples of nontransduced HepG2 cells or of HepG2 cells transduced with AdCMV.NK4 or AdCMV.GFP at an MOI of 500 IU/cell, were added to three replicate wells. VEGF (1 and 10 ng/ml), HGF (10 and 100 ng/ml) and ECGF (25 μg/ml) were taken along as controls. After 72 hours, relative cell growth was assessed by WST-1 conversion assay.

In vivo effects of AdCMV.NK4. on HepG2 xenografts in nude mice

Female, specified pathogen-free, athymic nu/nu mice, weighing 20–25 g(Harlan, Horst, The Netherlands) were housed in sterile filtertop cages and fed with irradiated nutrients and filtered acidified water ad libitum. The care and handling of the animals were undertaken in accordance with the directive on Animal Experimentation of the European Community. The protocol was approved by the institutional Animal Welfare Committee. HepG2 cells (5 × 106) were s.c. injected on the hind leg flanks. When tumor nodules reached 150–300 mm3, mice were randomly divided over three groups and injected intratumorally on day 0 and 4 with 109 PFU AdCMV.NK4 (n=8), 109 PFU AdCMV.GFP (n=7), or PBS (n=8). Tumor size was monitored two times a week, using digital calipers. Volume was calculated from the average of tumor length and width according to the formula 4/3πr3. Median tumor growth rates were determined from time of individual tumors to reach five times initial tumor volume. Mice were euthanized when tumors reached a size of 2000 mm3. Tumor specimens were explanted, snap-frozen and stored in liquid nitrogen. Treatment was not associated with toxic effects as evidenced by lack of body weight loss and unchanged grooming habits.

Immunohistochemical analysis of HepG2 xenografts

Expression of cMET, HGF, NK4 and CD34 in HepG2 xenografts was analyzed by immunohistochemistry as described before.12 Vessel density was determined by counting “hot spot” areas as previously described.12


Statistical analysis of data was performed using ANOVA test and Kruskal–Wallis test. For each single analysis, the statistical test used is stated in the text. Differences were considered significant when P<.05.


Inhibition of HCC cell proliferation and migration by NK4 gene transfer

We used the HCC cell lines HepG2, Hep3B and HuH7 to explore the biological effects of adenoviral vector-mediated NK4 gene delivery on HCC. All three HCC cell lines express cMET, but no HGF mRNA (Fig 1). We next examined the effects of AdCMV.NK4 transduction on proliferation of HCC cell lines in the presence and absence of HGF. cMET-positive SaOS-2 and -negative LNCaP cells were included as controls. HGF had no effect on LNCaP cells, while SaOS-2 cells demonstrated marked stimulation of proliferation upon exposure to HGF in a dose-dependent manner up to HGF concentrations of 20 ng/ml (Fig 2). This type of proliferation profile represents the mitogenic response seen for most cMET-positive tumors. To the contrary, while HGF significantly stimulated proliferation of the three HCC cell lines up to 1.5-fold at concentrations up to 2.5 ng/ml, it markedly inhibited proliferation at concentrations ≥10 ng/ml (Fig 2). This kind of response profile distinguishes HCC from most other cMET-positive tumors. Transduction of the HCC cell lines by AdCMV.NK4 significantly inhibited proliferation mediated by HGF at low concentrations, similar to its activity on the SaOS-2 cell line, which has a common HGF mitogenic response profile (Fig 2). Importantly, even in the presence of HGF at concentrations ≥2.5 ng/ml, the proliferation of AdCMV.NK4-transduced HCC cells was effectively reduced, over and above the antimitogenic effects of HGF (Fig 2). The viability of quiescent, confluent, nonproliferating HCC cells was not affected by AdCMV.NK4 (data not shown), excluding that the observed effects could be explained by cytotoxic effects of the adenoviral vector or the NK4 protein. These data thus indicate effective antiproliferative activity of adenoviral vector-mediated expression of NK4 on HCC cell lines.

Figure 1

Expression of cMET and HGF by HCC, LNCaP and SaOS-2 cells. Shown are the results of RT-PCR on the cell lines HepG2, Hep3B, HuH7, LNCaP, SaOS-2 and positive control U118MG, known to express HGF and cMET.38 cMET PCR product is 190 bp and HGF PCR product is 749 bp. M, Marker; B, blank (no template RNA).

Figure 2

Effects of HGF and AdCMV.NK4 on proliferation of HCC, LNCaP and SaOS-2 cells in vitro. HCC cells were transduced with AdCMV.NK4 (gray bars) or AdCMV.GFP (black bars) at an MOI of 100 IU/cell and cultured at indicated HGF concentrations. Mock-infected cells (white bars) were taken along as controls. After 120 hours, viability was measured by WST-1 conversion assay, and relative growth compared to mock-infected cells cultured without HGF was determined. Three independent experiments performed in triplicate were carried out. Data shown are means±SD of a representative experiment. ANOVA: **P<0.001; *P<0.01.

We next investigated the effect of AdCMV.NK4 transduction on the migratory and invasive capacity of the HCC cell lines by scatter assay, spheroid outgrowth assay and matrigel transwell system. In accordance with the motogenic response profile observed in cMET-positive cells, HCC cells were induced to migrate after exposure to HGF, and AdCMV.NK4 had a dose-dependent inhibitory effect on the HGF-induced migration, reaching complete abrogation of responsiveness to HGF after transduction of cells with AdCMV.NK4 at MOI ≥100 (Fig 3), while control virus exerted no effect. Together, these findings indicate effective inhibition of HGF-induced migratory and invasive capacities of HCC cell lines by adenoviral vector-mediated expression of NK4.

Figure 3

Inhibition of HGF-induced migration of HCC cells by AdCMV.NK4. (a) Effect of AdCMV.NK4 on HGF-induced migration of HCC cells in monolayer colonies. HCC cells were transduced with AdCMV.NK4 or AdCMV.GFP at an MOI of 150 IU/cell, and allowed to form colonies. Then, colonies were treated with or without 10 ng/ml HGF and scattering was evaluated after 24 hours. Three independent experiments performed in triplicate were carried out. Shown are microscopic pictures of a representative experiment. (b) Inhibitory effect of AdCMV.NK4 on HCC spheroid outgrowth. Shown are microscopic pictures of outgrowth of spheroids transduced with AdCMV.NK4 or AdCMV.GFP cultured in medium with or without 25 ng/ml HGF. Mock-infected spheroids were taken along as controls. Three independent experiments performed in triplicate were carried out. Shown are results of a representative experiment. (c) Inhibitory effect of AdCMV.NK4 on HGF-induced invasion of HCC cells into matrigel. Cells were transduced with AdCMV.NK4 or AdCMV.GFP, loaded onto matrigel in the upper compartment of a transwell chamber and exposed to medium supplemented with or without 20 ng/ml HGF in the lower compartment. Mock-infected cells were taken along as controls. After 96 hours, filters were fixed and invaded cells were stained. Three independent experiments performed in triplicate were carried out. Data shown are means±SD of a representative experiment. M, mock-infected cells; GFP, AdCMV.GFP-infected cells; NK4, AdCMV.NK4-infected cells; ANOVA: *P<0.001; ns, AdCMV.GFP and AdCMV.NK4 have no significant effect on basal cell migration (P>0.05).

Bystander-effect of AdCMV.NK4. gene transfer on nontransduced HCC and endothelial cells

Transduction of HepG2 with AdCMV.NK4 resulted in secretion of NK4 protein into the culture medium in an MOI-dependent manner as determined by ELISA (data not shown). Transfer of conditioned medium from AdCMV.NK4-transduced HepG2 cells onto colonies of nontransduced HepG2 cells demonstrated complete abrogation of HGF-induced scattering (data not shown). Thus, AdCMV.NK4-transduced HepG2 cells secrete sufficient amounts of NK4 into the culture medium to counteract the motogenic effects of HGF on nontransduced cells. In addition, conditioned medium from AdCMV.NK4-transduced HepG2 cells significantly inhibited proliferation of HUVEC (Fig 4, ANOVA: P<0.001), which express cMET, but do not produce HGF (as determined by RT-PCR and ELISA, data not shown). HepG2 cells produce, among other angiogenic factors, biologically effective amounts of VEGF (5 ng/ml per 72 hours per 106cells) that stimulated HUVEC to proliferate. VEGF levels in conditioned medium of transduced and control HepG2 cells were similar (ELISA, data not shown). The inhibition of HUVEC proliferation by conditioned medium of AdCMVNK4-transduced HepG2 cells thus suggests a broad antiangiogenic activity of NK4.

Figure 4

Bystander effect of NK4 produced by AdCMV.NK4-transduced HepG2 cells on vascular endothelial cell proliferation. HUVEC were cultured with ECGF (25 μg/ml), VEGF (1 or 10 ng/ml) or conditioned medium of control-, AdCMV.NK4- or AdCMV.GFP-transduced HepG2 cells. After 72 hours, viability of HUVEC was measured by WST-1 conversion assay, and relative growth was determined. Three independent experiments were performed in triplicate. Data are given as mean±SD of a representative experiment performed in triplicate. ANOVA: *, P<0.001; CM-HepG2, conditioned medium of HepG2 cells treated as indicated.

Inhibition of HCC tumor growth and angiogenesis by AdCMV.NK4. gene transfer in vivo

The effects of adenovirus-mediated expression of NK4 on tumor growth in vivo were determined using s.c. HepG2 xenografts in athymic nude mice. These xenografts consist of cMET expressing HepG2 tumor cells and HGF-producing mouse stromal cells as demonstrated by immunohistochemistry (Fig 5a). In a pilot experiment, where we treated minimal size HCC xenografts (around 60 mm3) with two intratumoral injections of 109 PFU AdCMV.NK4, we observed that one of the three treated mice were long-term tumor free (complete remission, data not shown), while none of the control mice were cured. Next, we explored NK4 gene transfer in a more clinically relevant model of HCC using established HepG2 tumors (150–300 mm3). Treatment with 109PFU AdCMV.NK4 resulted in reduced tumor growth rates, whereas growth of control tumors was not affected (Fig 5b). Tumors from mice treated with PBS or AdCMV.GFP grew rapidly and increased to five times their initial size in 8.7±3.3 and 10.2±3.2 days, respectively. AdCMV.NK4 caused a significant tumor growth delay of 9 days compared to controls (19.4±4.5 days, Kruskal–Wallis: P<.001). However, although AdCMV.NK4 delayed tumor growth, it did not cure animals from their tumor. When animals were killed, their tumors were explanted and analyzed for NK4 expression and vessel density. NK4 expression was detectable in all AdCMV.NK4-treated tumors at the time the tumors were explanted 4–5 weeks after the first adenovirus injection (Fig 5c). Intratumoral vessel densities were quantified after staining histological sections for CD34, which specifically labels endothelial cells (Fig 5d). In NK4-treated mice the mean intratumoral vessel density was reduced by 52% relative to tumors in the control groups (Fig 5e, ANOVA: P<.001).

Figure 5

Inhibition of human HCC xenograft growth in nude mice after intratumoral injection with AdCMV.NK4. (a) Expression of cMET and HGF in s.c. HepG2 tumors. Shown are representative pictures of HepG2 xenograft cryosections after immunohistochemistry using antibodies C-28 (anti-cMET) and H-145 (anti-HGF). Note that the HepG2 tumor cells express cMET (immunoreactivity was localized on the cell membrane and in the cytoplasm) and the mouse stromal cells produce HGF (immunoreactivity in the cytoplasm). (b) Subcutaneous HepG2 tumors at 150–300 mm3 were injected intratumorally on days 0 and 4 with PBS (line - - - - , n=8), AdCMV.GFP (line —•—, 109 PFU, n=7) or AdCMV.NK4 (line —▪—, 109 PFU, n=8). The volume of the tumors was measured twice weekly, and relative tumor volume was determined. The mean relative tumor volumes (±SEM) are shown. Arrows indicate the time points of adenoviral vector injection. (c) Expression of NK4 in s.c. AdCMV.NK4-treated HepG2 tumors. Shown is a representative picture of an AdCMV.NK4-treated HepG2 xenograft 28 days after the first adenovirus injection. Sections were stained by immunohistochemistry using antibody Ab9106 (anti-myc tag). Note the fields of NK4-expressing cells throughout the xenograft section (left). The right-hand pictures show fields of NK4-expressing cells at a higher magnification. (d) Immunostaining for CD34 revealed numerous small vessels in PBS-treated tumors (data not shown) and AdCMV.GFP-treated tumors (left), but only sparse vascularization in tumors of AdCMV.NK4-treated mice (right). (e) Vessel density was determined by counting the number of anti-CD34 stained vessels per field (0.28 mm2). The mean vessel density per field of vision (±SD) per treatment group (PBS: n=8, AdCMV.GFP: n=7 and AdCMV.NK4: n=8) is shown. ANOVA: *P<0.001.


The quest to develop more effective therapeutic modalities for human tumors has led to an upsurge of interest in approaches that simultaneously target multiple cellular processes important for the malignant behavior of cancers, including tumor cell proliferation, metastasis and angiogenesis. In this regard, gene therapy using NK4 is a rational new therapeutic approach. NK4 is known to effectively inhibit a variety of malignant processes by antagonizing HGF as well as several angiogenic factors, like VEGF and bFGF.3, 4, 5, 6, 7, 12 As a result of these pleiotropic activities NK4 is preferred among other potential inhibitors of HGF-cMET signaling, like cMET selective tyrosine kinase inhibitors, neutralizing anti-HGF antibodies or cMET-blocking antibodies.39 So far, NK4 gene therapy has demonstrated efficacy in tumor models in which the involvement of the HGF-cMET signaling pathway in pathogenesis has been clearly established.8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20 Based on the multifunctional properties of NK4 on various growth factors,3, 4, 12, 40 we propose that NK4 gene therapy would have therapeutic value for a wider range of tumors than only cMET overexpressing carcinomas. However, application of NK4 gene therapy to specific tumor types such as HCC for which the effect of the HGF-cMET signaling is less explicit or may even be inhibitory on carcinogenesis32, 33, 34 needed further assessment. Therefore, we evaluated NK4 gene therapy on a series of HCC cell lines that respond to HGF by enhanced migration but reduced proliferation31 and found that even on such cells NK4 gene therapy is effective.

In vitro, adenoviral vector-mediated expression of NK4 resulted in significant antagonism to HGF-induced invasiveness, and importantly, exhibited strong antiproliferative activity. These results may be explained by the structure of NK4 being identical to the HGFalpha-chain and lacking the intracellular signaling pathway-stimulating beta-chain.3 NK4 is able to block the receptor for binding with HGF, and thereby is hindering the promitogenic as well as the proliferative intracellular signaling pathways activated upon exposure to HGF. Importantly, NK4 over-ruled the effect of high concentrations of HGF (≥10 ng/ml), at which HGF was antimitogenic by itself. Therefore, these data may remove concerns about application of an HGF-antagonist to HCC, and strongly suggest therapeutic potential of NK4 as anticancer agent for treatment of HCC. Interestingly, NK4 additionally inhibited growth factors other than HGF, which together resulted in significant inhibition of proliferation of endothelial and tumor cells (Figs 2 and 4). This might be explained by NK4 interfering with receptor systems other than HGF-cMET via its kringle domains.4, 40, 41 Our in vitro data thus indicate that NK4 has also therapeutic value for tumors on which HGF may exert inhibitory activity on carcinogenic processes, like HCC.

In vivo, evaluation of adenoviral vector-mediated NK4 gene therapy using established subcutaneous HepG2 tumors in nude mice revealed significant tumor growth delay. We also found a significant reduction in vessel density in tumors from AdCMV.NK4-treated mice compared to controls, suggesting that in addition to direct inhibitory effects of NK4 on tumor cells inhibition of angiogenesis contributed to the antitumor effect. Complete cures were not observed in the stringent model of established tumors. Clearly, our animal data are encouraging, although they also indicate that NK4 gene therapy requires optimization before clinical application. Potential improvements include readministration of adenoviral vectors without challenging neutralizing immunity, application of selectively replicating adenoviral vectors expressing NK4 to overcome low tumor transduction and to gain effect by high level production of NK4 plus the inherent oncolytic property of these viruses,42, 43, 44, 45 and other treatment strategies such as using the liver as systemic supplier of NK4.8, 17 Obviously, further development of NK4 gene therapy for HCC should also include safety evaluation in animal models of cirrhotic liver and chemical hepatocarcinogenesis representing the pathogenesis of human HCC, and examination of toxicity of NK4 on angiogenesis during normal processes like wound healing. In this regard, it is reassuring that trials with other antiangiogenic factors such as angiostatin and endostatin, have been shown to inhibit tumor angiogenesis with little or no effect on wound healing in mice.46

In conclusion, our data suggest the potential of NK4 gene therapy for treatment of HCC based on pleiotropic functions of NK4 on tumor and endothelial cells. Our results support NK4 gene therapy to have a broad therapeutic value, not only for tumors in which involvement of HGF-cMET signaling is clearly established, but also for tumors in which HGF is more indirectly implicated in cancer progression via promoting angiogenesis. This study thus recommends further preclinical evaluation of NK4 gene therapy.


  1. 1

    Curley SA, Jones DV . Management of hepatocellular carcinoma. In: Meyers MA ed. Neoplasms of the digestive tract: imaging, staging and management. 1st edn. Philadelphia: Lippincott-Raven; 1998: 93–109.

  2. 2

    Prieto J, Herraiz M, Sangro B, et al. The promise of gene therapy in gastrointestinal and liver diseases. Gut. 2003;52 (Suppl 2):49–54.

  3. 3

    Date K, Matsumoto K, Shimura H, et al. HGF/NK4 is a specific antagonist for pleiotrophic actions of hepatocyte growth factor. FEBS Lett. 1997;420:1–6.

  4. 4

    Kuba K, Matsumoto K, Date K, et al. HGF/NK4, a four-kringle antagonist of hepatocyte growth factor, is an angiogenesis inhibitor that suppresses tumor growth and metastasis in mice. Cancer Res. 2000;60:6737–6743.

  5. 5

    Date K, Matsumoto K, Kuba K, et al. Inhibition of tumor growth and invasion by a four-kringle antagonist (HGF/NK4) for hepatocyte growth factor. Oncogene. 1998;17:3045–3054.

  6. 6

    Jiang WG, Hiscox SE, Parr C, et al. Antagonistic effect of NK4, a novel hepatocyte growth factor variant, on in vitro angiogenesis of human vascular endothelial cells. Clin Cancer Res. 1999;5:3695–3703.

  7. 7

    Matsumoto K, Nakamura T . NK4 (HGF-antagonist/ angiogenesis inhibitor) in cancer biology and therapeutics. Cancer Sci. 2003;94:321–327.

  8. 8

    Maemondo M, Narumi K, Saijo Y, et al. Targeting angiogenesis and HGF function using an adenoviral vector expressing the HGF antagonist NK4 for cancer therapy. Mol Ther. 2002;5:177–185.

  9. 9

    Hirao S, Yamada Y, Koyama F, et al. Tumor suppression effect using NK4, a molecule acting as an antagonist of HGF, on human gastric carcinomas. Cancer Gene Ther. 2002;9:700–707.

  10. 10

    Maehara N, Nagai E, Mizumoto K, et al. Gene transduction of NK4, HGF antagonist, inhibits in vitro invasion and in vivo growth of human pancreatic cancer. Clin Exp Metastasis. 2002;19:417–426.

  11. 11

    Saimura M, Nagai E, Mizumoto K, et al. Intraperitoneal injection of adenovirus-mediated NK4 gene suppresses peritoneal dissemination of pancreatic cancer cell line AsPC-1 in nude mice. Cancer Gene Ther. 2002;9:799–806.

  12. 12

    Heideman DAM, van Beusechem VW, Bloemena E, et al. Suppression of tumor growth, invasion and angiogenesis of human gastric cancer by adenovirus-mediated expression of NK4. J Gene Med. 2004;6:317–327.

  13. 13

    Fujiwara H, Kubota T, Amaike H, et al. Suppression of peritoneal implantation of gastric cancer cells by adenovirus vector-mediated NK4 expression. Cancer Gene Ther. 2005;12:206–216.

  14. 14

    Ueda K, Iwahashi M, Matsuura I, et al. Adenoviral-mediated gene transduction of the hepatocyte growth factor (HGF) antagonist, NK4, suppresses peritoneal metastases of gastric cancer in nude mice. Eur J Cancer. 2004;40:2135–2142.

  15. 15

    Kushibiki T, Matsumoto K, Nakamura T, Tabata Y . Suppression of the progress of disseminated pancreatic cancer cells by NK4 plasmid DNA released from cationized gelatin microspheres. Pharm Res. 2004;21:1109–1118.

  16. 16

    Tanaka T, Shimura H, Sasaki T, et al. Gallbladder cancer treatment using adenovirus expressing the HGF/NK4 gene in a peritoneal implantation model. Cancer Gene Ther. 2004;11:431–440.

  17. 17

    Wen J, Matsumoto K, Taniura N, Tomioka D, Nakamura T . Hepatic gene expression of NK4, an HGF-antagonist/angiogenesis inhibitor, suppresses liver metastasis and invasive growth of colon cancer in mice. Cancer Gene Ther. 2004;11:419–430.

  18. 18

    Brockmann MA, Papadimitriou A, Brandt M, et al. Inhibition of intracerebral glioblastoma growth by local treatment with the scatter factor/hepatocyte growth factor-antagonist NK4. Clin Cancer Res. 2003;9:4578–4585.

  19. 19

    Davies G, Mason MD, Martin TA, et al. The HGF/SF antagonist NK4 reverses fibroblast- and HGF-induced prostate tumor growth and angiogenesis in vivo. Int J Cancer. 2003;106:348–354.

  20. 20

    Martin TA, Parr C, Davies G, et al. Growth and angiogenesis of human breast cancer in a nude mouse tumour model is reduced by NK4, a HGF/SF antagonist. Carcinogenesis. 2003;24:1317–1323.

  21. 21

    Luo YQ, Wu MC, Cong WM . Gene expression of hepatocyte growth factor and its receptor in HCC and nontumorous liver tissues. World J Gastroenterol. 1999;5:119–121.

  22. 22

    Tavian D, De Petro G, Benetti A, et al. u-PA and c-MET mRNA expression is co-ordinately enhanced while hepatocyte growth factor mRNA is down-regulated in human hepatocellular carcinoma. Int J Cancer. 2000;87:644–649.

  23. 23

    Okano J, Shiota G, Kawasaki H . Expression of hepatocyte growth factor (HGF) and HGF receptor (c-met) proteins in liver diseases: an immunohistochemical study. Liver. 1999;19:151–159.

  24. 24

    Ljubimova JY, Petrovic LM, Wilson SE, et al. Expression of HGF, its receptor c-met, c-myc and albumin in cirrhotic and neoplastic human liver tissue. J Histochem Cytochem. 1997;45:79–87.

  25. 25

    Ueki T, Fujimoto J, Suzuki T, et al. Expression of hepatocyte growth factor and its receptor c-met proto-oncogene in hepatocellular carcinoma. Hepatology. 1997;25:619–623.

  26. 26

    Sakata H, Takayama H, Sharp R, et al. Hepatocyte growth factor/scatter factor overexpression induces growth, abnormal development, and tumor formation in transgenic mouse livers. Cell Growth Differ.. 1996;7:1513–1523.

  27. 27

    Horiguchi N, Takayama H, Toyoda M, et al. Hepatocyte growth factor promotes hepatocarcinogenesis through c-Met autocrine activation and enhanced angiogenesis in transgenic mice treated with diethylnitrosamine. Oncogene. 2002;21:1791–1799.

  28. 28

    Tajima H, Matsumoto K, Nakamura T . Hepatocyte growth factor has potent anti-proliferative activity in various tumor cell lines. FEBS Lett. 1991;291:229–232.

  29. 29

    Shiota G, Rhoads DB, Wang TC, et al. Hepatocyte growth factor inhibits growth of hepatocellular carcinoma cells. Proc Natl Acad Sci USA. 1992;89:373–377.

  30. 30

    Miyazaki M, Gohda E, Tsuboi S, et al. Human hepatocyte growth factor stimulates the growth of HUH-6 clone 5 human hepatoblastoma cells. Cell Biol Int Rep. 1992;16:145–154.

  31. 31

    Neaud V, Faouzi S, Guirouilh J, et al. Human hepatic myofibroblasts increase invasiveness of hepatocellular carcinoma cells: evidence for a role of hepatocyte growth factor. Hepatology. 1997;26:1458–1466.

  32. 32

    Liu ML, Mars WM, Michalopoulos GK . Hepatocyte growth factor inhibits cell proliferation in vivo of rat hepatocellular carcinomas induced by diethylnitrosamine. Carcinogenesis. 1995;16:841–843.

  33. 33

    Shiota G, Kawasaki H, Nakamura T, et al. Inhibitory effect of hepatocyte growth factor on metastasis of hepatocellular carcinoma in transgenic mice. Res Commun Mol Pathol Pharmacol. 1996;91:33–39.

  34. 34

    Santoni-Rugiu E, Preisegger KH, Kiss A, et al. Inhibition of neoplastic development in the liver by hepatocyte growth factor in a transgenic mouse model. Proc Natl Acad Sci USA. 1996;93:9577–9582.

  35. 35

    Fontijn R, Hop C, Brinkman HJ, et al. Maintenance of vascular endothelial cell-specific properties after immortalization with an amphotrophic replication-deficient retrovirus containing human papilloma virus 16 E6/E7 DNA. Exp Cell Res. 1995;216:199–207.

  36. 36

    Heideman DAM, Snijders PJF, Bloemena E, et al. Absence of tpr-met and expression of c-met in human gastric mucosa and carcinoma. J Pathol. 2001;194:428–435.

  37. 37

    Borset M, Lien E, Espevik T, et al. Concomitant expression of hepatocyte growth factor/scatter factor and the receptor c-MET in human myeloma cell lines. J Biol Chem. 1996;271:24655–24661.

  38. 38

    Koochekpour S, Jeffers M, Rulong S, et al. Met and hepatocyte growth factor/scatter factor expression in human gliomas. Cancer Res. 1997;57:5391–5398.

  39. 39

    Maulik G, Shrikhande A, Kijima T, et al. Role of the hepatocyte growth factor receptor, c-Met, in oncogenesis and potential for therapeutic inhibition. Cytokine Growth Factor Rev. 2002;13:41–59.

  40. 40

    Nakabayashi M, Morishita R, Nakagami H, et al. HGF/NK4 inhibited VEGF-induced angiogenesis in in vitro cultured endothelial cells and in vivo rabbit model. Diabetologia. 2003;46:115–123.

  41. 41

    Kuba K, Matsumoto K, Ohnishi K, et al. Kringle 1–4 of hepatocyte growth factor inhibits proliferation and migration of human microvascular endothelial cells. Biochem Biophys Res Commun. 2000;279:846–852.

  42. 42

    Alemany R, Balague C, Curiel DT . Replicative adenoviruses for cancer therapy. Nat Biotechnol. 2000;18:723–727.

  43. 43

    Heise C, Kirn DH . Replication-selective adenoviruses as oncolytic agents. J Clin Invest.. 2000;105:847–851.

  44. 44

    Motoi F, Sunamura M, Ding L, et al. Effective gene therapy for pancreatic cancer by cytokines mediated by restricted replication-competent adenovirus. Hum Gene Ther. 2000;11:223–235.

  45. 45

    Maemondo M, Saijo Y, Narumi K, et al. Gene therapy with secretory leukoprotease inhibitor promoter-controlled replication-competent adenovirus for non-small cell lung cancer. Cancer Res. 2004;64:4611–4620.

  46. 46

    Folkman J . Antiangiogenic gene therapy. Proc Natl Acad Sci USA. 1998;95:9065–9066.

Download references


We thank Dr K. Matsumoto and Dr T. Nakamura (Osaka University Medical School, Suita, Osaka, Japan) for providing the NK4 cDNA, and Jeroen Mastenbroek and Diederik van den Berg for technical assistance. We thank Elisabeth Bloemena for helping with immunohistochemical examinations and Paul van Diest for help with determination of microvessel density. This work was supported by a research grant from the Pasman Foundation. Renée Overmeer and Daniëlle Heideman are supported by grants from the Dutch Digestive Diseases Foundation (WS02-31 and WS95-12, respectively). Victor van Beusechem is supported by a research fellowship of the Royal Netherlands Academy of Arts and Sciences (KNAW).

Author information

Correspondence to Daniëlle A M Heideman.

Rights and permissions

Reprints and Permissions

About this article


  • tumor growth
  • invasion
  • angiogenesis
  • hepatocyte growth factor
  • c-MET

Further reading