The 5-year survival rate for patients with pancreatic cancer is <5%, and it is always resistant to the current chemoradiotherapy. Therefore, new, effective agents for the treatment of pancreatic cancer are urgently needed. The promising strategy of cancer-targeting gene virotherapy (CTGVT) has demonstrated great anticancer potential. The objective of the current study was to determine whether 1 CTGVT approach, oncolytic virus (OV)-harboring lipocalin-2, is capable of treating pancreatic cancer.
Tissue microarrays were constructed to detect the expression of lipocalin-2 in 60 specimens of pancreatic adenocarcinoma. The clinical significance of lipocalin-2 was investigated in an analysis of correlations between lipocalin-2 expression and matched clinical characteristics. A lipocalin-2–expressing OV, ZD55-lipocalin-2, was constructed by deleting the adenoviral protein E1B55kD. The antitumor efficacy and mechanisms of the OV were investigated in pancreatic cancer cells with v-Ki-ras2 Kirsten rat sarcoma viral oncogene homolog (KRAS) mutations in vitro and in vivo.
Lipocalin-2 expression was correlated with a good prognosis in patients with pancreatic adenocarcinoma. ZD55-lipocalin-2 dramatically inhibited the growth of pancreatic cancer in vitro and in vivo by inducing cytolysis and caspase-dependent apoptosis.
Pancreatic cancer is a malignant tumor with a very poor prognosis. The overall 5-year survival rate of patients with pancreatic cancer is <5%.1 Only 10% of patients are capable of to undergoing radical pancreatectomy2; however, the median survival for these patients is only 13.9 months.3 Meanwhile, pancreatic cancer remains highly chemoradioresistant.2 Gemcitabine or gemcitabine-based combined therapies are standard treatment for patients with pancreatic cancer, but the survival benefits from these treatments are extremely limited.4 Therefore, effective agents are urgently needed to treat this devastating disease.
Oncolytic viral therapy is 1 of the new therapies that have been developed in recent years. It was designed specifically to replicate in tumor cells, which makes it an ideal therapeutic agent. Many oncolytic viruses without foreign anticancer genes have been introduced into clinical trials,5-9 and some have exhibited great potential and have progressed toward phase 3 trials. On the basis of oncolytic viral therapies, we initiated the cancer-targeting gene virotherapy (CTGVT). In CTGVT, an antitumor gene is inserted into a modified oncolytic virus with a cloning site, such as ZD55. The oncolytic virus is able to replicate several hundred-fold within cancer cells. The inserted antitumor gene also is replicated several hundred-fold through accompanying oncolytic virus replication, which makes the antitumor effect of CTGVT much greater than that of either gene therapy or oncolytic viral therapy alone.10 When the infected cancer cells are lysed, numerous replication-competent progeny viruses are released, infecting the neighboring cancer cells. Because virus replication is targeted to the tumor cells,10 it will leave the normal cells intact. In our laboratory, this therapeutic strategy revealed potential antitumor activity in tumors like hepatoma,11, 12 leukemia,13 and colorectal cancer,14, 15 and sometimes even achieved the complete eradication of xenograft tumors in nude mice.
Of all human cancers, pancreatic adenocarcinoma has the highest incidence of v-Ki-ras2 Kirsten rat sarcoma viral oncogene homolog (KRAS) mutations (>80% or approximately 90%).16, 17 KRAS mutations in tumors always predict resistance to chemotherapy and radiotherapy.18-21 It was reported previously that, in RAS-transformed cells, lipocalin-2 reversed the epithelial-to-mesenchymal transition phenotype and inhibited tumor invasiveness, metastasis, and angiogenesis.22, 23 Lipocalin-2–induced apoptosis also was reported in hematopoietic and endometrial carcinoma cells.24, 25 However, it remains unclear whether lipocalin-2 can inhibit or kill pancreatic cancer cells. Our clinical research indicated that lipocalin-2 is associated with a good prognosis in patients with pancreatic cancer. In addition, other interesting recent results indicated that26, 27 chemopreventive treatments inhibited tumor promotion and progression in a preclinical tumor model that activated lipocalin-2 gene and protein expression. All of these results may provide a clue about the use of lipocalin-2 as an antitumor agent against pancreatic cancer.
The therapeutic value of oncolytic viruses, per se, for pancreatic cancer has been reported previously,28, 29 but only a few studies have involved the use of CTGVT for pancreatic cancer. In the current study, the significance of lipocalin-2 expression in pancreatic cancer was investigated, and a lipocalin-2–expressing oncolytic virus, ZD55-lipocalin-2, was constructed to investigate whether it could strongly exhibit the ability to kill pancreatic cancer cells.
MATERIALS AND METHODS
Patients' Data and Construction of Tissue Microarrays
Clinicopathologic data from 60 patients with pancreatic adenocarcinoma were collected from January 2001 to June 2006. The cohort included 37 men and 23 women, and the median age (±standard deviation) was of 61.70 ± 12.41 years. Forty-six patients underwent conventional pancreatoduodenectomy, 3 patients underwent pancreatoduodenectomy plus portal vein resection and reconstruction, 2 patients underwent total pancreatectomy, 8 patients underwent distal pancreatectomy with splenectomy, and 1 patient underwent pylorus-preserving pancreaticoduodenectomy. According to the International Union Against Cancer (UICC) tumor-lymph node-metastasis (TNM) classification for malignant pancreatic tumors (sixth edition; UICC, 2002), there were 34 patients with stage I to IIA disease and 26 patients with stage IIA to IV disease.
The tissue microarrays were constructed by Shanghai Outdo Biotech Company Ltd. (Shanghai, China). Representative regions were selected from hematoxylin and eosin (H&E)-stained tissue sections and marked on individual paraffin blocks. Two tissue cores were obtained from each specimen. Each core was precisely arrayed into a new paraffin block. These microarrays were serially sectioned (4 mm) and stained with H&E to verify tissue sampling and completeness. Then, lipocalin-2 expression was analyzed by EnVision immunohistochemistry (Dako Cytomation, Carpinteria, Calif).
Lipocalin-2 immunoreactivity was evaluated independently by 2 pathologists who were blinded to patients' outcomes. Lipocalin-2 expression was graded as follows: negative was defined as positive cytoplasmic staining in <5% of tumor cells, and positive was defined as positive cytoplasmic staining in ≥5% of tumor cells.
Cell Lines and Culture Conditions
HEK293 cells were purchased from Microbix Biosystems Inc. (Toronto, Ontario, Canada), Pancreatic cancer cell lines (BxPC-3 and PANC-1) and human lung fibroblast cell lines (NHLF-1 and MRC5) were obtained from the American Type Culture Collection (Manassas, Va). Cells were cultured in Dulbecco modified eagle medium supplemented with 10% fetal bovine serum, 4 mmol/L glutamine, 50 U/mL penicillin, and 50 μg/mL streptomycin.
Plasmids and Recombinant Adenovirus Construction
The pZD55 plasmid was constructed in our laboratory30 with the E1B55-kDa encoding gene deleted to restrict the viral replication only to tumor cells. The lipocalin-2 gene was amplified from a combinational DNA library (Clontech Corporation, Mountain View, Calif) by polymerase chain reaction (PCR). Then, the lipocalin-2 gene was cloned into pCA13 plasmid to construct pCA13-lipocalin-2. Next, pCA13-lipocalin-2 was digested with the Bgl II restriction enzyme to obtain an expression cassette that included the cytomegalovirus promoter and lipocalin-2 coding sequences (CDS). This cassette was subcloned into pZD55 plasmid to produce a pZD55-lipocalin-2 shuttle plasmid. All plasmids were confirmed by restrictive enzyme digestion, PCR, and DNA sequencing. Homologous recombination using shuttle plasmids and pBHGE3 (Microbix Biosystems Inc.) was carried out in 293 cells to generate ZD55-lipocalin-2 and Ad-lipocalin-2 (the typical nonreplicative adenovirus). Recombinant viruses were purified by cesium chloride gradient ultracentrifugation. Virus titers were measured by using a standard plaque assay on HEK293 cells. ZD55-enhanced green fluorescent protein (EGFP) and ZD55 were retained in our laboratory.
Reporter Gene Assay
PANC-1 cells were plated onto 24-well plates at approximately 80% confluence and then infected with ZD55-EGFP. After 48 hours and 72 hours of exposure, the same visual fields were examined to investigate whether the ZD55-mediated gene would express the corresponding protein in pancreatic cancer cells.
Viral Production Assay
PANC-1 and NHLF-1 cells were plated onto 6-well plates at 80% confluence and then infected with ZD55-lipocalin-2. After 48 hours, medium and cells were collected, and the virus was released by freeze-thawing for 3 cycles and centrifuged to collect the supernatant. Virus titers were determined by using a standard plaque assay in 293 cells.
Cell-Viability Assay and Cytopathic Effects
Cells were grown in 96-well plates and then treated with various viruses for different times. Cell viability was quantified by using a standard 3-(4,5-dimethylthizol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. To detect the tumor-specific cytopathic effects of ZD55-lipocalin-2 on cells, 2 pancreatic cancer cell lines (BxPC-3 [wild-type KRAS] and PANC-1 [KRAS mutation]) and 1 wild-type p53 normal fibroblast cell line (MRC5) were infected with ZD55-lipocalin-2 and ZD55 at various multiplicities of infection. Five days later, the cells were stained with crystal violet, and significant cytopathic effects were observed.
PANC-1 cells (5 × 103 cells) that had been treated with the different adenoviruses were replated onto 6-well plates. Before incubation, the cells were washed 3 times in phosphate-buffered saline (PBS). After a 9-day incubation, the cells were fixed in methanol and stained with 0.5% crystal violet. Visible colonies were photographed. All experiments were performed in triplicate.
4′,6-Diamidino-2-Phenylindole and Propidium Iodide Staining and Flow Cytometry
To determine the apoptotic effect, PANC-1 cells plated in 6-well plates were treated with the adenoviruses. Cells were stained with 4′,6-diamidino-2-phenylindole (DAPI) and propidium iodide (PI), and the fraction of cells in sub-G1 phase was determined using flow cytometry. Annexin V/PI double staining and flow cytometry also were performed with a FACScan flow cytometer (Becton Dickinson, East Rutherford, NJ).
Western Blot Analysis
Cells infected with virus or mock-infected cells were harvested from the plates. Western blot analyses were done as previously described.12 Antibodies against β-actin, E1A, E1B55-kDa, caspase-3, and poly-ADP-ribose polymerase (PARP) were obtained from Santa Cruz Biotechnology (Santa Cruz, Calif).
Antitumor Efficacy in Xenograft Tumors
PANC-1 cells (1.4 × 107 cells) were implanted by subcutaneous injection into the lower flank of BALB/c nude mice at ages 4 to 5 weeks. When the tumor size reached 100 to 150 mm3, the animals were randomized into 3 groups with 10 mice per group. Adenovirus at daily dose of 5 × 108 plaque-forming units in 100 μL PBS or PBS alone was injected intratumorally for 4 days. Tumor volumes were measured and calculated according to the following formula: (length × width2) × 0.5. All animals were maintained in accordance with institutional, United Kingdom Coordinating Committee on Cancer Research, and National Institutes of Health guidelines.
In Situ Cell Apoptosis Detection by Terminal Deoxynucleotidyl Transferase-Mediated Deoxyuridine Triphosphate Nick-End Labeling
Tumor specimens were fixed, processed, and embedded. Deparaffinized tumor sections were used in this test, and hematoxylin was used as a counterstain. Cell apoptosis in tumors was detected in situ in resected tissues by enzymatic labeling of DNA strand breaks with a terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate nick-end labeling (TUNEL) assay kit (Boehringer Mannheim GmbH, Mannheim, Germany).
Statistical analyses were conducted using the SPSS for Windows statistical software package (SPSS, Inc., Chicago, Ill). All continuous data are presented as means ± standard deviations. Categorical variables were compared using the chi-square test or the Fisher exact test. Independent-sample t tests or analyses of variance were used to compare mean values between different groups. The Kaplan-Meier method was used to estimate survival, and differences in survival were compared using the log-rank test; P values < .05 were considered statistically significant.
Relations Between Lipocalin-2 Expression and Patient Prognosis
Immunohistochemical staining revealed that lipocalin-2 was highly expressed in the cytoplasm of primary pancreatic cancer cells. The positive expression rate of lipocalin-2 was 81.6% in 60 primary pancreatic cancer samples. Weak or negative expression of lipocalin-2 occurred in normal pancreatic tissues and lymph node metastasis. The 5-year survival rate was 0% for patients who had lipocalin-2–negative tumors and 23%, for patients who had lipocalin-2–positive tumors. Univariate analysis indicated that positive expression of lipocalin-2 was associated with a high survival rate in patients with pancreatic cancer (Fig. 1)
Univariate analysis also was performed to investigate the relation between lipocalin-2 expression and clinicopathologic features. Clinicopathologic characteristics that were analyzed included histologic grade of tumor differentiation, tumor size, tumor stage, etc. The analysis indicated that positive lipocalin-2 expression was associated with negative lymph node metastasis (P = .03) or earlier TNM stage (P = .03) (Table.1).
Table 1. Univariate Analysis of Associations Between Lipocalin-2 Expression and Clinicopathologic Features
Abbreviation: SD, standard deviation.
Staging was assigned according to the International Union Against Cancer TNM classification (sixth edition, 2002).
Construction and Characteristics of ZD55-Lipocalin-2
ZD55 was constructed as described in our previous report.30 The lipocalin-2 gene was inserted into ZD55 to form ZD55-lipocalin-2. Then, 1300 base-pair (bp) DNA fragments were detected in products of Bgl II-digested pZD55-lipocalin-2. Approximately 700 bp DNA fragments were amplified from pZD55-lipocalin-2 by PCR. Homologous recombination was carried out between pZD55-lipocalin-2 and pBHGE3 in HEK293 cells to obtain ZD55-lipocalin-2. Approximately 1450 bp DNA products were detected from ZD55-lipocalin-2 DNA by PCR using sense and antisense ZD55 primers.30 Approximately 700 bp bands occurred in the PCR-amplified products from ZD55-lipocalin-2 DNA using primers for lipocalin-2. Wester blot results also showed ZD55-lipocalin-2 expressed E1A and lipocalin-2 proteins, but failed to express E1B-55kDa protein. These results suggested that the construction of ZD55-lipocalin-2 was correct.
ZD55-Mediated Adenovirus Replicates in PANC-1 Cells
PANC-1 cells were infected with the same titer of Ad-EGFP and ZD55-EGFP, and the fluorescence intensity of EGFP in PANC-1 cells infected with ZD55-EGFP increased over time. In contrast, fluorescence intensity was much weaker in PANC-1 cells that were infected with Ad-EGFP. A viral production assay indicated that the reproductive activity of ZD55-lipocalin-2 in PANC-1 cells was dramatically higher than that in normal NHFL-1 cells.
Reduction of Colony Formation Induced by ZD55-Lipocalin-2
PANC-1 cells (5 × 103 cells) infected with ZD55-lipocalin-2, Ad-lipocalin-2, and ZD55-EGFP were plated, and colony-formation assays were performed 9 days after plating. The colony numbers in ZD55-lipocalin-2–infected cells were the lowest of the 3 groups.
Cytopathic Effects and Growth Inhibition of ZD55-Lipocalin-2 In Vitro
Cytopathic effects were observed by crystal violet staining. The cell-killing effect of ZD55-lipocalin-2 was greater than that of ZD55 in vitro, and the cell-killing effect of ZD55-lipocalin-2 was stronger against PANC-1 cells than against BxPC-3 cells. However, both ZD55-lipocalin-2 and ZD55 caused limited cell death in MRC5 cells. MTT experiments demonstrated that the viability of PANC-1 cells infected with ZD55-lipocalin-2 decreased in a time-dependent and concentration-dependent manner. The inhibition of tumor cells from ZD55-lipocalin-2 was much stronger than that from the control group. Ad-lipocalin-2 also had antitumor effects associated with its concentration and its infection time.
Antitumor Efficacy of ZD55-Lipocalin-2 in Vivo
Pancreatic cancer xenografts were established and randomly divided into 3 groups. There was no significant difference in the baseline tumor volume between the different groups in a 1-way analysis of variance (P = .893). Tumors that were treated with ZD55-lipocalin-2 exhibited significant growth inhibition compared with tumors that were treated with ZD55. Both ZD55-lipocalin-2 and ZD55 had significant antitumor efficacy (Fig. 2A). In particular, 2 tumor xenografts in the ZD55-lipocalin-2 group were completely eradicated 20 days after treatment. In survival analysis, failure was defined as the death of an animal or a tumor volume >550 mm.3 The results indicated that the prognosis for the ZD55-lipocalin-2 group was best among the 3 groups (P < .05), and the prognosis for the ZD55 group was better than that for the PBS group (P < .05) (Fig. 2B).
Apoptosis Induced by ZD55-Lipocalin-2
Apoptosis was detected using flow cytometry and DAPI staining to determine whether growth inhibition was associated with apoptosis induced by ZD55-lipocalin-2. By using annexin V/PI double staining, flow cytometry data can detect early (labeled with annexin V alone) and late (labeled with annexin V/PI) apoptotic cells. Our data demonstrated that the early apoptotic rate of cells treated with ZD55-lipocalin-2 increased significantly compared with the rate of control cells. The fraction of cells in sub-G1 phase in the ZD55-lipocalin-2 group was much greater than that in the other groups. DAPI staining revealed shrunken cell shapes and small, pyknotic nuclei in PANC-1 cells that were infected with ZD55-lipocalin-2, but not in normal cells or in PANC-1 cells that were infected with ZD55 (Fig. 3). A TUNEL assay also revealed that ZD55-lipocalin-2 dramatically induced more apoptotic cell death and necrosis in tumors (Fig. 4).
Western Blot Analysis
The level of procaspase-3 in PANC-1 cells infected with ZD55-lipocalin-2 decreased in a time-dependent manner. The level of cleaved PARP in PANC-1 cells infected with ZD55-lipocalin-2 for 48 hours increased in a dose-dependent manner. Cleaved PARP occurred in cells that were infected with ZD55-lipocalin-2 but was negative in cells that were infected with either Ad-lipocalin-2 or ZD55 (Fig. 5). These data indicate that ZD55-lipocalin-2 induced PANC-1 cells apoptosis through a caspase-dependent pathway, and apoptosis induced by ZD55- lipocalin-2 was caused largely by elevated expression of lipocalin-2.
There have been no major breakthroughs in early detection or effective treatments for pancreatic cancer during the last 20 years. Most pancreatic cancers remain untreatable or develop secondary resistance to the existing treatments. A high rate of KRAS mutations (approximately 90%)17 in pancreatic cancer has dramatically contributed to the incidence of refractory disease. Therefore, it is crucial to develop new therapies for patients who have pancreatic cancer with KRAS mutations.
Oncolytic viruses, which have been evaluated in several clinical trials,31, 32 are emerging as attractive anticancer agents. Approximately 50% of patients with pancreatic cancer achieved favorable outcomes in a phase 2 clinical trial using the E1B-deleted adenovirus ONYX-015.33 CRAd-Cans, another E1B55-deleted oncolytic adenovirus, greatly inhibited the growth of BxPC-3 cells.34 However, the CTGVT strategy combines the advantages of gene therapy with oncolytic virus therapy and, thus, produces much better antitumor effects than either therapy alone. Lipocalin-2 is a multifaceted protein; however, it can inhibit the proneoplastic factor heat inducible factor-1a, focal adhesion kinase phosphorylation, and vascular endothelial growth factor synthesis in RAS-transformed cells.23, 35-38 The antitumor effects of lipocalin-2 in pancreatic cancer reportedly may be because of its ability to block invasion and angiogenesis.39, 40 Our clinical data indicated that the positive rate of lipocalin-2 expression in early stage disease was greater than that in advanced disease. The prognosis for our lipocalin-2–negative group was poorer than that in our lipocalin-2–positive group. These data suggested that greater expression of lipocalin-2 was associated with earlier tumor stages and a better prognosis. These findings supported the potential of combined therapy with lipocalin-2 and an oncolytic virus in the CTGVT strategy as a good way to treat pancreatic cancer.
Our data indicated that the antitumor efficacy of ZD55-lipocalin-2 was dose-dependent and time-dependent. Both ZD55 and Ad-lipocalin-2 had antitumor efficacy in pancreatic cancer, but growth inhibition with ZD55-lipocalin-2 was greater than that with either ZD55 or Ad-lipocalin-2 alone. The antitumor activity of ZD55-lipocalin-2 was stronger in PANC-1 cells (KRAS mutated cells41) than in BxPC-3 cells (wild-type KRAS cells41, 42). Because the clinical results indicated that lipocalin-2–negative pancreatic cancers had a poorer prognosis and that pancreatic cancers with KRAS mutations always were refractory to current chemotherapies and radiotherapies, our in vivo studies targeted lipocalin-2–negative pancreatic cancer cells with KRAS mutations. In a PANC-1 tumor xenograft nude mouse model, ZD55-lipocalin-2 had the greatest ability to inhibit tumor growth among the 3 groups. These data suggested that ZD55-lipocalin-2 may be very useful for pancreatic cancer therapy, especially in tumors with KRAS mutations.
Why did ZD55-lipocalin-2 have stronger therapeutic effects than Ad-lipocalin-2 or ZD55? We believe that the inserted lipocalin-2 gene can over express several hundred-fold through accompanying ZD55-lipocalin-2 replication in pancreatic cancer cells, making the antitumor effect of ZD55-lipocalin-2 much greater than that of Ad-lipocalin-2 or ZD55 alone. Our data clearly demonstrated that sub-G1 phase or pyknotic nuclei occurred in PANC-1 cells infected with ZD55-lipocalin-2, but none or few occurred in cells infected with Ad-lipocalin-2 or ZD55. Moreover, TUNEL staining revealed that the percentage of apoptotic cells and the amount of necrosis in the ZD55-lipocalin-2 group were much greater than those in the ZD55 group or in the PBS control group. It has been reported that the interleukin-1 β-converting enzyme/caspase family plays a crucial role in apoptosis.43 In particular, caspase-3 is a key component of the apoptotic mechanism.44, 45 ZD55-lipocalin-2 had stronger power to activate PARP and casapse-3 than Ad-lipocalin-2 or ZD55. Thus, the combined antitumor effects of ZD55-lipocalin-2, originating from apoptosis because of lipocalin-2 overexpression, and cancer cells lysis, caused by oncolytic virus replication, kill pancreatic cancer cells effectively.
Liver metastases are a common finding in pancreatic cancer. Usually, if there is evidence of liver metastases, surgery will not be an option. Moreover, postoperative liver metastases are frequent events and are the leading cause of death in patients with pancreatic cancer. Our current results indicate that ZD55-lipocalin-2 may be a promising clinical agent in the future. Especially for patients who undergo the removal of a primary tumor but have disease accompanied by a solitary hepatic metastasis and immunohistochemistry results indicating a KRAS mutation in the primary tumor, the administration of ZD55-lipocalin-2 by using a technique like transcatheter arterial chemoembolization may be very useful for the prevention and treatment of liver metastasis.
In summary, although lipocalin-2 is a controversial modulator in cancer progression,46-48 the current results clearly indicate that lipocalin-2 has antitumor effects in pancreatic cancer. ZD55-lipocalin-2 provides excellent antitumor effects in pancreatic cancer. It mediates greater caspase-dependent apoptosis in pancreatic cancer cells compared with Ad-lipocalin-2 (the typical gene therapy) or ZD55 (the typical oncolytic virus therapy). It may serve as a potent anticancer drug for pancreatic cancer therapy, especially for patients who have pancreatic adenocarcinoma with KRAS mutations.
This work was supported by the National Natural Science Foundation of China (81001007), the Program for Young Excellent Talents in Tongji University (2008KJ060), and the Youth Fund of the Shanghai Tenth People's Hospital (10RQ105).