Suppression of metastasis of human pancreatic cancer to the liver by transportal injection of recombinant adenoviral NK4 in nude mice



NK4, a 4-kringle fragment of hepatocyte growth factor (HGF), is an HGF antagonist that also acts as an angiogenesis inhibitor. NK4 strongly inhibits the infiltration, metastasis, and tumor growth of pancreatic cancer. The aim of our study was to evaluate the antitumor effect of adenovirus-mediated NK4 gene transfer to the liver on hepatic metastasis of pancreatic cancer in vivo. We constructed recombinant adenoviral NK4 (Ad-NK4), which encodes a secreted form of human NK4. Intrasplenic injection of Ad-NK4 induced high and relatively maintained expression of NK4 protein in the liver and suppressed the number and growth of metastatic foci in the liver in a nude mouse model. Microscopically, central necrosis was found even in small metastatic foci in Ad-NK4 treated mice. Immunohistochemical analysis of metastatic tumors showed a remarkable decrease in microvessel density and an increase in the number of apoptotic tumor cells after treatment with Ad-NK4. These results indicate that intraportal injection of Ad-NK4 may be a useful therapeutic modality for the clinical control of hepatic metastasis in pancreatic cancer. © 2005 Wiley-Liss, Inc.

Pancreatic cancer remains one of the most malignant neoplasms. The disease is diagnosed frequently at an advanced stage, and only 3% of patients survive 5 years.1 Pancreatic cancer has a high rate of local and systemic recurrence, including liver metastasis, peritoneal dissemination and retroperitoneal recurrence.2 There is no effective treatment for this disease. Radical resection has only a limited effect for the disease,3 but even after curative resection of the primary tumor, liver metastasis occurs frequently and constitutes a major course of this disease.2, 4 Generally, micrometastasis to the liver seems to have already occurred in most patients when pancreatic cancer is diagnosed.5 Pancreatic cancer is highly resistant to the chemotherapy and radiation protocols available currently. Even gemcitabine, which has become the standard drug used for metastatic disease, does not improve median survival. One factor underlying the aggressive local and early systemic tumor growth may be rapid tumor neoangiogenesis, which results in an abundant blood supply. Tumor-induced neoangiogenesis is a common phenomenon during growth, particularly in tumors larger than 1–2 mm in diameter.6, 7

We showed previously that pancreatic cancer cells frequently overexpress c-Met/hepatocyte growth factor (HGF) receptor and that HGF plays important roles in the mitogenic, motogenic and morphogenic activities of these cells. We also identified and prepared NK4 as an antagonist of HGF.8, 9

NK4 is composed of the N-terminal hairpin and 4 kringle domains of HGF. NK4 binds c-Met/HGF receptor but does not induce tyrosine phosphorylation of c-Met. NK4 is a potent antiangiogenic agent and antagonizes not only HGF-induced angiogenesis but also that of other angiogenic factors such as vascular endothelial growth factor and basic fibroblast growth factor.10 NK4 is also a competitive antagonist of HGF that completely inhibits HGF activity at concentrations of at least 1,000-fold the concentration of HGF.11, 12, 13 Creation of an NK4-encoding adenovirus (Ad-NK4) has made it possible to produce sufficient concentrations of NK4 to antagonize the mitogenic, motogenic and morphogenic activities of HGF. In vivo studies, we showed that Ad-NK4 gene therapy inhibits the growth of pancreatic cancer in a subcutaneously (s.c.) implanted model and in an intraperitoneally (i.p.) implanted model.14, 15

The adenovirus vector has higher specificity for the liver than for other organs16, 17, 18 and can induce gene transfer in almost any cell type.19 Intraportal injection of Ad-NK4 would be expected to result in high expression of NK4 in the liver, even at low infection concentrations.

We investigated the antitumor effect of intraportal injection of Ad-NK4 in a mouse model of pancreatic cancer metastasis to the liver.

Material and methods

Construction of recombinant adenovirus

A recombinant adenoviral vector expressing human NK4 (Ad-NK4) was constructed as described previously.20 Briefly, Ad-NK4 was generated by homologous recombination of the pJM17 plasmid21 and the shuttle plasmid vector pSV2+22 containing an expression cassette and the cytomegalovirus early promoter/enhancer followed by human NK4 cDNA11 and a polyadenylation signal. A control vector with expressing the bacterial β-galactosidase enzyme (LacZ) was constructed by the same procedure with pJM17 and pCA17 containing the LacZ gene. Recombinant Ad-NK4 and adenovirus LacZ (Ad-LacZ) were propagated in HEK-293 cells. The adenovirus titer in plaque-forming units (pfu) was determined by plaque formation assays with HEK-293 cells.

Cells and culture conditions

The SUIT-2 human pancreatic cancer cell line was generously provided by Dr. H. Iguchi (Kyushu Cancer Center, Fukuoka, Japan). SUIT-2 cells were cultured in RPMI supplemented with streptomycin, penicillin, and 10% FBS at 37°C in a humidified atmosphere containing 5% CO2.

Animals and intrasplenic implantation of pancreatic cancer cells

Six-week-old female athymic nude mice (BALBc nu/nu) were purchased from Japan SLC (Hamamatsu, Japan). The mice were housed in laminar-flow cabinets under specific pathogen-free conditions in facilities approved by Kyushu University. Suspensions of 2 × 106 SUIT-2 cells/0.1 ml were implanted by open injection into the spleen of nude mice under anesthesia.

Treatment of intrasplenic implantation with Ad-NK4

Mice that received intrasplenic implantation of SUIT-2 cells were assigned to 1 of 3 treatment groups. Mice were re-laparotomised and then, Ad-LacZ or Ad-NK4 at 1 × 108 pfu/0.1 ml was injected into the spleen on Day 3 after intrasplenic implantation of SUIT-2 cells. In a control group, 0.1 ml PBS was injected in the same manner. Three mice from each group were killed on Day 3 and 3 more on Day 7 after intrasplenic implantation of SUIT-2 cells. Metastatic tumors were evaluated by histological examination. Five additional mice per group were killed on Day 21 after intrasplenic implantation and were evaluated with respect to the ratio of metastatic liver tumor volume to liver volume. Samples of metastatic liver tumors were also subjected to immunohistochemical study.

Determination of tumor–liver volume ratio

Livers were sectioned completely at 2-mm intervals. Images of all sections were digitized, and liver volume was determined with the use of NIH Image software. Tumor volume was determined in a similar manner.

NK4 expression in liver, lung, and blood serum after intrasplenic injection of Ad-NK4

To evaluate NK4 expression in abdominal tissues and blood serum, mice that had undergone intrasplenic injection of PBS, Ad-LacZ or Ad-NK4 were killed on Days 3, 7, 14, 21 and 28 after viral injection. Tissue samples of liver and lung were homogenized in ice-cold lysis buffer composed of 1% Triton X-100, 10 mM Tris-HCl, 5 mM EDTA, 50 mM NaCl, 30 mM sodium pyrophosphate, 50 mM sodium fluoride, 0.1 mM sodium orthovanadate, and 0.1% bovine albumin. The supernatants were collected after centrifugation at 12,000 rpm for 10 min. Expression of NK4 was detected with an ELISA kit (Immunis HGF EIA; Institute of Immunology, Tokyo Japan). The lower limit of detection of NK4 was 0.3 ng/1 g protein.

Immunohistochemical staining and cell counting

Livers from animals in each group were subjected to immunohistochemical staining on Day 21. Tissue sections (5-μm thick) of formalin-fixed, paraffin-embedded specimens were deparaffinized in xylene and rehydrated in graded alcohols. For microvessel staining, the peroxidase-conjugated avidin-biotin complex method was carried out with a Vectastain Elite ABC Kit (Vector, Burlingame, CA). Mouse monoclonal CD31 antibody JC/70A (NeoMarkers, Fremont, CA) was used at a dilution of 1:50, followed by incubation with biotinylated anti-mouse IgG (1:100; Vector). Microvessel density (MVD) was assessed in tumor areas showing high staining density. The number of vessels was counted in 10 fields per section at ×200 (0.739 mm2/field), and the mean counts were recorded. Apoptotic cells within tumors were detected by terminal deoxynucleotidyl transferase-mediated dUTP-biotin nick end labeling TUNEL assay (In situ Apoptosis Detection Kit, Takara, Shiga, Japan). Positive cells were counted in each 10 randomly selected fields per section at ×200. Proliferating cells were detected with an antibody generated against proliferating cell nuclear antigen (PCNA) dilution PC10 (DAKO Glostrup, Denmark), and immunoreactivity was visualized in a manner similar to that used for CD31 staining. To quantify PCNA staining, we counted the positive cells in 10 random fields per section at ×200.


In tumor-free mice, Ad-NK4 at the same concentration as in previous experiments was injected into the spleen. On Days 3, 7, 14 and 21 after injection, mice were killed and blood samples were subjected to biochemistry tests such as GOT, GPT, glucose, BUN, creatinine and amylase activity. As a control, blood sera from mice with no treatment were also analyzed.


Mice were treated intrasplenic implantation of SUIT 2 cells and intrasplenic injection of PBS or Ad-NK4 in a similar manner as above. The condition of these mice was checked out twice daily. Dead mice were autopsied as soon as possible at clean ventilator in facilities approved by Kyushu University.

Statistical analysis

Statistical significance was evaluated with a non-parametric Mann-Whitney U-test. All tests were 2-tailed, and a p-value of <0.05 was considered statistically significant. Survival was analyzed by log-rank analysis of Kaplan-Meier curves.


Effect of Ad-NK4 on metastatic liver tumors

On Day 3 after intrasplenic implantation of SUIT-2 pancreatic cancer cells, there were few macroscopic metastatic nodules at the liver surface. Histologically, there were several micrometastatic nodules measuring approximately 30 μm in diameter (Fig. 1a,b). These were thought to be indicative of early liver metastases, which are not detectable by imaging. Therefore, we injected Ad-NK4 at this time. On Day 7 after implantation of SUIT-2 cells, no significant difference in macroscopic liver appearance between the PBS- and Ad-NK4-treated groups was identified (Fig. 1c,f). Histologically, metastatic tumors of the PBS-treated group showed severe invasion into the hepatic parenchyma, and neovascularization was already initiated (Fig. 1d,e). Metastatic tumors of the Ad-NK4-treated group showed expanding growth with pseudo capsules containing fibroblasts and inflammatory cells, but there was no neovascularization in the tumor and central necrosis was present (Fig. 1g,h).

Figure 1.

Macroscopic and microscopic appearances of the liver on Day 3 after implantation of SUIT-2 cells (a,b), and of the liver in the Ad-NK4-treated group (c,d,f) and the PBS-treated group (g,h,i) on Day 7 after implantation of SUIT-2 cells. Scale bars = 10 μm. Long arrow, central necrosis; arrowheads, microvessels.

Suppression of liver metastasis by intrasplenic injection of Ad-NK4

SUIT-2 cells exhibited an aggressive and malignant phenotype in vivo, and intrasplenically implanted cells never failed to produce metastatic nodules in the liver in the absence of Ad-NK4 injection. The macroscopic appearance of the liver showed almost complete coverage by metastatic tumors in both the PBS- and Ad-LacZ-treated groups. Intrasplenic injection of Ad-NK4 inhibited the growth of metastatic liver tumors remarkably compared to metastatic growth in the PBS- or Ad-LacZ-treated groups on Day 21 after implantation (Fig. 2a). The respective tumor volume/liver volume ratio were 52.75 ± 8.23%, 48.39 ± 15.99% and 16.62 ± 6.60% for the PBS-, Ad-LacZ- and Ad-NK4-treated groups. There were no significant differences in liver weight between the groups (Table I, Fig. 2b).

Figure 2.

Suppression of metastatic liver tumors by intrasplenic injection of Ad-NK4. PBS, Ad-LacZ, or Ad-NK4 was injected intrasplenically on Day 3 after implantation of SUIT-2 cells. (a) Macroscopic appearance of metastatic liver tumors after treatment with PBS (control; a), Ad-LacZ (b), and Ad-NK4 (c) on Day 21 after splenic implantation of SUIT-2 cells (n = 5/group). (b) Ratio of tumor volume vs. liver volume. Data are expressed as mean ± SD. *p < 0.005.

Table I. Suppression of Liver Metastasis by Transportal Injection of Ad-NK4 on Day 21 After Implantation of Suit-2 Human Pancreatic Cancer Cell
 Liver weight (g)Liver–body weight ratio (%)Tumor–liver volume ratio (%)
  • *

    Tumor–liver volume ratio was significantly lower in the Ad-NK4 treated group than other groups (p < 0.005).

PBS2.49 ± 0.6211.05 ± 2.8252.75 ± 8.23
Ad-LacZ2.23 ± 0.5110.24 ± 2.4648.39 ± 15.99
Ad-NK41.73 ± 0.287.46 ± 1.0016.62 ± 6.60*

NK4 expression in liver, lung, and blood serum after intrasplenic injection of Ad-NK4

Intrasplenic injection of Ad-NK4 induced extremely high expression of NK4 in the liver (Fig. 3). NK4 expression was first identified on Day 3 (59.85 ± 61.87 ng/g protein), and it peaked on Day 14 after injection (2084.58 ± 1383.41 ng/g protein). The level remained high even on Day 28 after injection (366.62 ± 373.041 ng/g protein). In contrast, NK4 expression in the lung and blood serum was extremely low and was not detectable on Day 7 after injection. NK4 expression was not detected in other organs (data not shown). In the PBS- and the Ad-LacZ- group, NK4 expression was not detectable in all organs through every time points (data not shown).

Figure 3.

Time-course of NK4 expression in livers after intrasplenic injection of Ad-NK4. Mice were killed 3, 7, 14, 21, and 28 days after intrasplenic injection of Ad-NK4. NK4 expression was detected by ELISA. Data are expressed as mean ± SD.

Immunohistochemical examination of pancreatic tumors

Histologically, the pancreatic tumors showed moderate to poorly differentiated adenocarcinomas that were similar in all 3 leading index groups. With respect to PCNA staining, the difference between the 3 groups was not significant; the proliferation index was not altered by NK4 expression. The apoptotic index as determined by TUNEL staining was greater in the Ad-NK4-treated group (17.0 ± 1.7%; p = 0.004) than in the Ad-LacZ-treated group (8.7 ± 1.2%) or the PBS-treated group (9.7 ± 0.6%). MVD was significantly lower in the Ad-NK4-treated group (5.9 ± 1.4 vessels/field; p < 0.001) than in the Ad-LacZ-treated group (10.8 ± 2.5 vessels/field) or in the PBS-treated group (10.1 ± 1.9 vessels/field) (Fig. 4a,b).

Figure 4.

Immunohistochemical staining of metastatic liver tumors on Day 21 after intrasplenic implantation of SUIT-2 cells, with intrasplenic injection of PBS, Ad-LacZ, or Ad-NK4. (a) Apoptosis (a–c), angiogenesis (d–f) and proliferation (g–i) were determined with the use of the TUNEL method, anti-CD31 antibody, and anti-PCNA antibody, respectively. PBS-treated group (a,d,g), Ad-LacZ-treated group (b,e,h), Ad-NK4-treated group (c,f,i). Arrows indicate microvessels. (b) Changes in the numbers of TUNEL-positive cells and blood vessels. Data are expressed as mean ± SD; *p = 0.004, **p < 0.001.

Toxicity of Ad-NK4

Blood biochemistry assays showed mild elevation of GOT on Day 3 after injection of Ad-NK4 that decreased rapidly to base line levels. Other enzymes showed no significant abnormalities (Fig. 5).

Figure 5.

Blood serum enzyme levels after intrasplenic injection of Ad-NK4. Mice were killed 3, 7, 14, and 21 days after intrasplenic injection of Ad-NK4. Data are expressed as mean ± SD.


Almost all mice in our study died of cachexia or severe peritoneal dissemination. Survival times differed between the groups. Mice of the PBS-treated group become moribund starting at Day 19 after implantation, and the last mouse died on Day 36.The Ad-NK4-treated mice gradually died beginning at Day 19, but the last mouse died on Day 48. The median survival times of the PBS- and the Ad-NK4-treated group were 25.6 ± 7.3 and 39.0 ± 6.9 days, respectively (Fig. 6). The survival rate of the Ad-NK4-treated group was significantly greater than that of the PBS-treated group (p = 0.01).

Figure 6.

Survival curves of PBS- (n = 7) and Ad-NK4-treated (n = 6) mice. PBS or Ad-NK4 was injected intrasplenically on Day 3 after implantation of SUIT-2 cells. There was a significant difference in survival times between the 2 groups (p = 0.01 by log-rank analysis of Kaplan-Meier curves).


Our study showed that intrasplenic injection of Ad-NK4 significantly suppressed tumor progression of pancreatic cancer in the liver of nude mice and prolonged survival. A clinical correlate of our study would involve postoperative liver metastasis of pancreatic cancer without other clinical or radiologic evidences of the disease. NK4 expression in the liver in response to Ad-NK4 injection was sustained and high, and pancreatic tumors showed decreased numbers of tumor vessels and an increased tumor cell apoptotic index.

Many pancreatic cancer patients who undergo resection experience disease recurrence, most frequently in the liver2, 23 and usually within 1 year.23 At the time of surgery, K-ras mutation in the liver has occurred in approximately 30% of patients, even when they showed no liver metastasis before surgery.5 When pancreatic cancer is diagnosed, most patients may already have micrometastases in the liver. Therefore, it is vital to halt or limit liver metastasis, to inhibit enlargement of metastatic tumors and to inhibit spread to other parts of the liver or to other organs.

We reported previously that HGF may be involved in the aggressive invasion, dissemination or metastasis of postoperative pancreatic cancer and that NK4 can inhibit HGF-induced invasive and metastatic behaviors.8, 15, 24 In our current study, we showed an antitumor effect of NK4 by intrasplenic injection of Ad-NK4 in a model of metastasis to the liver, suggesting the possibility that Ad-NK4 can inhibit metastatic liver tumors in postoperative pancreatic cancer patients.

The SUIT-2 human pancreatic cancer cell line overexpresses the c-Met receptor, and NK4 dose-dependently inhibits HGF-induced invasion and migration of these cells8 and decreases the number of blood vessels surrounding the tumor.14 Small tumors (1–2 mm in diameter) receive nutrients by diffusion, but when they grow to over 2 mm in diameter, they need an additional blood supply and therefore induce angiogenesis.6, 7 NK4 not only antagonizes HGF-induced angiogenesis but also arrests the angiogenesis induced by other angiogenic factors, such as vascular endothelial growth factor and basic fibroblast growth factor.13 The bifunctional nature of NK4 may lead to its development as a key drug in tumor therapy.

NK4 is a competitive antagonist of HGF,9, 25 but NK4 binding to the c-Met receptor is 10-fold lower in affinity than HGF binding, and for complete inhibition of HGF binding to the c-Met receptor, 1,000-fold higher concentrations of NK4 over than that of HGF are required.11, 12, 13 It is important to express as high a concentration as possible in the target organ. Use of the adenovirus vector is best for this purpose because it has higher specificity for the liver than for other organs16, 17, 18 and can induce gene transfer in almost any cell type.19 NK4 is a secretory protein, so it is not necessary for the vector to limit transduction to cancer cells. The adenovirus vector can show toxicity in the liver, muscle and lung when it is injected at high concentrations,26 and it can also cause host immune responses when administered intravenously.27 Use of minimal vector concentration to express maximum transfected gene product is necessary.

We administered Ad-NK4 to the spleen of mice as a model of intraportal injection. Expression of NK4 was very high in the liver and was prolonged over 28 days after administration; other organs expressed little or no NK4. Ad-NK4 induced minor and transient elevation of transaminase (GOT) levels. We attempted i.p. injection of Ad-NK4 at the same dose, but NK4 expression in the liver was very low in these experiments, and the suppression of tumor growth was disappointing (data not shown).

With respect to survival rate, significant increases were seen in Ad-NK4-treated mice and this was because that the tumor growth and dissemination from metastatic liver tumor were inhibited, so the onset of liver disfunction and intraabdominal bleeding were delayed, but all mice eventually died. One reason is that even though the adenovirus vector can induce high levels of transgene expression, this expression is temporary. We administered Ad-NK4 only once. In actual therapy, however, multiple injections of Ad-NK4 must be considered. Another reason is that NK4 shows cytostatic effect; therefore the combination therapy with a cytocidal drug such as gemcitabine should enhance the anti-tumor effect.

Many studies have presented recently that various types of cancer such as gastric cancer, gallbladder cancer, glioblastoma and colon cancer overexpress c-Met/HGF receptor and NK4 strongly suppressed the tumor growth and invasion of these type carcinoma in vivo.28, 29, 30, 31, 32 These data indicate that NK4 is certainly effective for various types of carcinoma.

In summary, we showed that intraportal injection of Ad-NK4 induced the expression of an effective concentration of NK4 in the liver and significantly suppressed liver metastasis in a nude mouse model. The survival of Ad-NK4-treated mice was significantly longer than that of PBS-treated mice. The therapeutic effect seemed to be due to the bifunctional activities of NK4 as an HGF antagonist and an angiogenesis inhibitor. Intraportal injection of Ad-NK4 may provide a benefit to patients with pancreatic cancer by inhibiting liver metastases after surgery.