Cancer Diagnosis and Therapy
Gene therapy for intraperitoneally disseminated pancreatic cancers by Escherichia coli uracil phosphoribosiltransferase (UPRT) gene mediated by restricted replication-competent adenoviral vectors
Article first published online: 5 SEP 2002
Copyright © 2002 Wiley-Liss, Inc.
International Journal of Cancer
Volume 102, Issue 1, pages 51–59, 1 November 2002
How to Cite
Oonuma, M., Sunamura, M., Motoi, F., Fukuyama, S., Shimamura, H., Yamauchi, J.-i., Shibuya, K., Egawa, S.-i., Hamada, H., Takeda, K. and Matsuno, S. (2002), Gene therapy for intraperitoneally disseminated pancreatic cancers by Escherichia coli uracil phosphoribosiltransferase (UPRT) gene mediated by restricted replication-competent adenoviral vectors. Int. J. Cancer, 102: 51–59. doi: 10.1002/ijc.10650
- Issue published online: 25 SEP 2002
- Article first published online: 5 SEP 2002
- Manuscript Accepted: 3 JUL 2002
- Manuscript Revised: 2 MAY 2002
- Manuscript Received: 26 NOV 2001
- Japanese Ministry of Education, Science, Sports and Culture. Grant Numbers: 10470251, 12877187
- gene therapy;
- replicating adenovirus;
- pancreatic cancer;
- peritoneal dissemination
Although patients with unresectable pancreatic tumors have been treated with 5-fluorouracil (5FU)-based combination chemotherapy, the drug resistance of cancer cells presents a crucial therapeutic problem. It was reported that UPRT overcomes 5FU resistance. UPRT catalyzes the synthesis of 5-fluorouridine monophosphate (FUMP) from Uracil and phosphoribosylpyrophosphate (PRPP). The antitumor effect of 5FU is enhanced by augmenting 5-fluorodeoxyuridine monophosphate (FdUMP) converted from FUMP, which inhibits thymidylate synthetase (TS). We first demonstrated that injecting an E1-deficient adenoviral vector (Adv) expressing UPRT (AxCAUPRT) followed by 5-FU treatment resulted in a volume reduction of xenotransplanted human tumors. In examining the therapeutic effect of AxCAUPRT/5-FU against peritoneal dissemination, we found that non-selective gene transduction of AxCAUPRT caused severe adverse effects arising from the increase of F-dUMP in normal intestine. Because the therapeutic gene delivered by a restricted replication-competent Adv lacking 55 kDa E1B protein (AxE1AdB) is speculated to be expressed selectively in tumors, mice with established tumors were injected with AxE1AdB and E1-deleted Adv expressing the lacZ reporter gene (AxCAlacZ). The expression of the reporter gene (lacZ) was selectively enhanced in disseminated tumors. The therapeutic advantage of restricted replication competent Adv that expresses UPRT (AxE1AdB-UPRT) was evaluated in an intraperitoneal disseminated tumor model. To study the anti-tumor effect of AxE1AdB-UPRT/5FU, mice with disseminated AsPC-1 tumors were administered the Adv, followed by the 5FU treatment. It was shown that the treatment with AxE1AdB-UPRT/5FU caused a dramatic reduction of the disseminated tumor burden without toxicity in normal tissues. Our results showed that the AxE1AdB-UPRT/5FU system is a promising tool for intraperitoneal disseminated pancreatic cancer. © 2002 Wiley-Liss, Inc.
Pancreatic cancer is the fifth leading cause of cancer death in the United States.1 The great majority of the patients present with the advanced disease when they are diagnosed. Consequently, many cases are unresectable at the time of diagnosis because of invasion to the retroperitoneal space together with the involvement of major vessels, lymph nodes, peritoneal dissemination and distant metastases. Although patients with unresectable pancreatic tumors are treated with 5-fluorouracil (5-FU) based-chemotherapy, the response rate of the agent against the disease is from 18–28% and the median survival is only 6 months. Accordingly, conventional chemotherapy provides little survival benefit in patients with pancreatic cancer.2
The drug resistance of cancer cells presents a crucial therapeutic problem. 5FU requires enzymatic conversion to the nucleotide to exert its cytotoxic activity: 1) incorporation of 5-fluorouridine triphosphate into RNA, 2) incorporation of 5-fluorodeoxyuridine triphosphate into DNA and 3) inhibiting thymidylate synthetase (TS) by forming a ternary complex that is synthesized by TS, 5,10-methylenetetrahydrofolate (5,10-CH2THF) and fluorodeoxyuridine monophosphate (F-dUMP).3 Therefore, inhibiting TS is an important strategy for augmenting the anti-tumor effect of 5FU. To inhibit TS other anticancer drugs such as cisplatin (CDDP), leucovorin (LV) and methotrexate (MTX) are introduced as modulators of the essential cofactors for ternary complex formation.4 Although several reports have described a relatively high response rate in other cancers, such as gastric cancer and colorectal cancer,5, 6 the median survival time remains extremely short in pancreatic cancer when treated with such combined chemotherapy.7, 8
Recently, the virus-mediated enzyme/prodrug system is thought to be a promising therapeutic approach in cancer gene therapy. The basis of the system is to exert a cytotoxic effect by the transduction of a therapeutic gene that codes an enzyme converting a non-toxic prodrug into an active antimetabolite intracellularly. There is a bystander effect in these systems that include herpes simplex virus thymidine kinase (HSV-tk) with gancyclovir (GCV)11, 12 and E. coli cytosine deaminase (CD) with 5-fluorocytosine (5FC).13 Uracil phosphoribosyl transferase (UPRT) with 5FU9 is a new approach to overcome 5FU resistance. Transduction of the UPRT gene improves the sensitivity of cancer cells resistant to 5FU. UPRT catalyzes the synthesis of fluorouridine monophosphate (FUMP) from Uracil and phosphoribosylpyrophosphate (PRPP). The anti-tumor effect of 5FU is enhanced by augmenting FdUMP converted from FUMP, which inhibits TS.9, 10 Although the bystander effect of these systems causes the killing of untransduced cells, the transduction efficiency of gene therapy using replication incompetent vectors is limited by their inability to spread and transfect the untransduced cells, especially in solid tumors.
To overcome this problem, gene transduction mediated by an Adv lacking the 55 kDa E1B protein (E1B55K) has been studied.14 This Adv can selectively replicate in and destroy tumor cells with an abnormal P53 gene.15, 16 Recent studies have reported that pancreatic cancer is associated with P53, K-RAS and other gene abnormalities. P53 genes are frequently mutated in about 50% of pancreatic ductal cancers.17 These genetic abnormalities are candidates for developing therapeutic approaches including gene therapy. As we previously reported, the Adv lacking E1B55K induces selective replication and cytopathic effects on pancreatic cancer cells with an abnormal P53 gene.18 Moreover, in contrast to replication incompetent Adv (E1-deficient Adv), this vector takes advantage of concentrating the prodrug-converting enzyme activity to the area of the infected cancer cells with the abnormal P53 gene.19
We investigated whether UPRT-expressing Adv lacking E1B55K (AxE1AdB-UPRT) selectively replicates and enhances the expression of UPRT gene in pancreatic cancer cells with an abnormal P53 gene. The anti-tumor effect of AxE1AdB-UPRT/5-FU treatment in vivo was examined in intraperitoneal disseminated AsPC-1 tumors.
MATERIAL AND METHOD
Five human pancreatic cancer cell lines were used in the experiments. AsPc-1, BxPc-3, Miapaca-2 and Panc-1 were purchased from American Types Culture Collection (ATCC, Manassas, VA) and maintained in RPMI 1640 medium containing 10% FBS, 2 mmol/L L-glutamine, 50 U/ml penicillin and 50 μg/mL streptomycin at 37°C in a humidified, 5% carbon dioxide atmosphere. All of these pancreatic cancer cells tested have the P53 mutation 20). The PK-1 human pancreatic cancer cell line was established in our department.21
The human embryonic kidney 293 cells were purchased from ATCC. They were maintained in DMEM containing 10% FBS, 2 mmol/L L-glutamine, 50 U/ml penicillin and 50 μg/ml streptomycin. The WI-38 human fibroblast cells were purchased from RIKEN Gene Bank (Tsukuba, Japan). They were maintained in DMEM containing 10% FBS 2mmol/L L-glutamine, 50 U/ml penicillin and 50 μg/ml streptomycin.
The recombinant adenovirus expressing UPRT gene driven by the CAG promoter (AxCA-UPRT),9 the recombinant adenovirus containing lacZ gene (AxCA-lacZ),22 the recombinant adenovirus lacking the E1B55K (AxE1AdB)18 and Ad5dlx23 that is an Ad5 mutant with E3 deletion, lacking 1879 bp from nt 2859 to nt 3047, were prepared as described previously.
Recombinant adenovirus AxE1AdB-UPRT was constructed essentially by the COS-TPC method.23 The 2832 bp Sal I/Hind III fragment from pCAupp that contains cDNA of UPRTase (649 bp) under the control of the CAG promoter, was inserted at the Sal I/Hind III site of the E1AdB cassette in pSKd(Xh)E1AdB, resulting in pCAupp-E1AdB. The blunted fragment that contained both the CAupp and E1AdB expression cassettes was ligated to the Swa I site of pAdex1cw, resulting in pAxE1AdB-UPRT. AxE1AdB-UPRT was obtained by cotransfecting the 293 cells with pAxE1AdB-UPRT cosmid DNA together with Ad5dlx DNA-TPC. The TGA stop codon at the 3rd codon of E1B55K in AxE1AdB-UPRT was confirmed by DNA sequencing. Adenovirus vectors were propagated in 293 cells and the titration was determined by the standard plaque assay.
Transduction efficiency of recombinant adenovirus in pancreatic cancer cells
Cells were plated into 12-well plates at a density of 1 × 105 cell/well. After 24 hr incubation, AxCA-lacZ was infected at a multiplicity of infection (MOI) of 10, 50 and 100. The cells were incubated at 37°C in a humidified atmosphere of 5% CO2 for 24 hr. To detect the expression of lacZ, cells were fixed with 0.5% glutaraldehyde and stained by X-gal. Blue stained cells were counted in 6 independent microscopic fields (×400).
Estimation of intracellular F-dUMP
To estimate the intracellular level of F-dUMP in cells transduced with the UPRT gene, Competitive Ligand Binding Assay for F-dUMP was carried out.24 Cells were infected with AxCA-UPRT and AxCA-lacZ at a MOI of 100 for 1 hr. After culturing for 24 hr, they were exposed to 0.001 mg/ml 5-FU (Kyowa Hakko Corp., Tokyo, Japan) for 48 hr. Cells (1 × 107) were harvested and homogenate buffer (2-mercaptoethanol, NaF, KH2PO4 containing cytidine monophosphate, KOH buffer), 3H-FdUMP and acetic acid were added. The supernatant was removed after centrifuging. Tris-HCl (PH 8.0) 5 mL was added and the supernatant was passed through a column of DEAE-cellulose. Fifty microliters of 3H-FdUMP, 100 μl thymidylate synthase and buffer (tetra hydrofolic acid, ascorbic acid sodium salt, HCHO) were added to this sample and the sample tubes were incubated at 25°C for 4 hr. The charcoal suspension was then added. The tubes were mixed and centrifuged and the supernatants were removed for assay by scintillation counting to measure the covalent ternary complex with TS-3H-FdUMP-5,10-methylenetetrahydrofolate in the supernatants. A linear standard curve was obtained by measuring F-dUMP.
Sensitivity of AxCA-UPRT infected pancreatic cancer cells to 5-FU
A cytotoxic assay was carried out to determine whether the pancreatic cancer cells transduced with the UPRT gene were increased in sensitivity to 5FU. Pancreatic cancer cells were plated onto 96-well plates at a density of 2 × 103/well. Cells were infected either AxCA-UPRT or AxCA-lacZ at a MOI of 100. The medium was replaced with those containing various concentrations of 5-FU at 24 hr after infection. After culturing for 72 hr, the relative cell number was measured by MTT assay.9 The concentration of 5-FU yielding IC50 was calculated using curve-fitting parameters.
Effect of UPRT-expressing E1 deficient adenoviral vector and 5-FU in vivo
The therapeutic advantage of AxCAUPRT/5-FU treatment was evaluated in an intraperitoneal disseminated AsPC-1 tumor model. For the development of an animal model that closely mimics human advanced pancreatic cancer, 6-week-old male SCID mice (CLEA, Tokyo, Japan) received an intraperitoneal injection of AsPC-1 (2 × 106 cells/mouse).25 The mice with intraperitoneal disseminated tumors were treated with an intraperitoneal injection of Adv 10 days after the initial tumor inoculation. From the following day, animals were administered 5-FU (10 mg/kg wt) for 7 days. Animals were sacrificed and all intraperitoneal tumors were weighed 4 weeks after the initial tumor inoculation. Within 10 days after the tumor inoculation, disseminated tumors in the peritoneal cavity could be detected macroscopically. The intraperitoneal tumors showed moderately to poorly differentiated adenocarcinoma confirmed by histological examination (data not shown).
In vitro viral replication assay
AsPC-1 and WI38 cells were plated into 6-well plates at a density of 2 × 104/well. The cells were infected with AxE1AdB, AxE1AdB-UPRT and Ad5dlx at a MOI of 5. After 24 hr infection the cells were cultured at 37°C for 5 or 10 days. The cells were harvested and sonicated for virus release. The titers of adenovirus were determined by standard plaque assay of the 293 cells 5 or 10 days after infection.26 Cultured cells infected with the adenoviral vector were fixed with 1% methanol and permeabilized with 0.1% Triton X-72 hr after infection.
Immunostaining was carried out by the streptavidin-biotin peroxidase method using LSABkit (DAKO, Kyoto, Japan). Primary antibodies against adenoviral hexon protein (MAB805, Chemicon International, Inc., Temecula, CA) were applied at 1:200 dilution. The cells were incubated for 1 hr at room temperature. A biotinylated goat anti-mouse secondary antibody was applied, followed by streptavidin peroxidase conjugate. 3-Amino-9-ethylcarbazole (AEC) was used as the chromogen.18
Selective replication of AxE1AdB and E1-deficient adenovirus expressing lacZ gene in vivo
Selective replication of AxE1AdB in vivo was evaluated by staining β-galactosidase by X-gal. Four weeks after the initial tumor inoculation, mice received an injection of 1 × 108 pfu of Adv (AxCA-lacZ, AxE1AdB, AxE1AdB+AxCA-lacZ). Three days after injection of the Adv, mice were sacrificed and the peritoneal cavity was washed by PBS and fixed by submersion in cold PBS containing 2% formaldehyde and 0.2% glutaraldehyde for 3 hr. To detect the expression of β-galactosidase, the peritoneal cavity was stained by X-gal for 4 hr.27 To evaluate the selective replication of AxE1AdB in intraperitoneal tumors, mice were sacrificed 3 days after the injection with AxE1AdB and AxCA-lacZ. Intraperitoneal tumors and intestines were removed and stained by HE or immunostaining with the antibody against adenoviral hexon protein (MAB805). Tumors and intestines were embedded in O.C.T. compound, cut into 5-μm sections, fixed in 1% methanol, stained using LSABkit (DAKO, Kyoto, Japan) as described above and counterstained with eosin.
Estimation of amplified gene delivered by replication-competent adenoviral vector
To determine the amplification of the UPRT gene delivered by replication-competent Adv in cells with abnormal p53, Southern blot analysis was carried out. AsPC-1 cells and 293 cells were infected with AxCA-UPRT and AxE1AdB-UPRT at a MOI of 10. The infected cells were harvested 72 hr after infection. The cells were suspended in a equal volume of 50 mM Tris-HCl, 100 mM EDTA and 0.1%SDS containing proteinase K at 200 μg/ml and incubated at 50°C for 2hr. The mixture was extracted twice with phenol/chloroform and DNA was precipitated with ethanol. Samples were digested with XbA1 and separated with 0.8% agarose gel electrophoresis and the restriction fragments were transferred to a nylon membrane.
To make the DNA probe, DNA that was extracted from AxCA-UPRT infected 293 cells was subjected to PCR analysis by using two primers (sense: GCGAATTCCACCATGAAGATCGTGGAAGTCAAACAC, antisense: GGCGGATCCTTATTTCGTACGAAAGATTTTGTCACCGG) specific to the UPRT sequence that should yield a 649-bp fragment. The PCR product was labeled with digoxigenin (DIG) by PCR using the DIG DNA Labeling Kit.28 The transferred nylon membrane was hybridized with DIG labeled probe using DIG Hybridization Kit and the UPRT gene was detected using the DIG Luminescent Detection Kit.
Therapeutic effect of UPRT-expressing restricted replication-competent adenoviral vectors in vivo
The therapeutic effect of AxE1AdB-UPRT/5-FU treatment was evaluated in the intraperitoneal disseminated ASPC-1 tumor model. The mice were treated with intraperitoneal injection Adv 10 days after the initial tumor inoculation. From 3 days later, the animals were administered 5-FU (10 mg/kg) for 7 days. The animals were sacrificed and all intraperitoneal tumors were weighed 4 weeks after the initial tumor inoculation. All disseminated tumors were collected and weighed.
Data presentation and statistical analysis
Results are presented at mean ± SD. Statistical analysis was carried out by Mann-Whitney's U-test using StatView software and p < 0.05 was considered to be statistically significant.
Transduction efficiency of recombinant adenovirus in pancreatic cancer cell line
The transduction efficiency of adenoviral vectors in pancreatic cancer cells was evaluated by lacZ gene expression. The number of cells expressing lacZ gene increased in a MOI-dependent manner. The lacZ gene was expressed in 20–30% of the cells at a MOI of 10, in 50–60% of the cells at a MOI of 50 and in 80–95% of the cells at a MOI of 100. There was no significant difference in the gene transfer efficiency among the 5 cell lines (data not shown).
UPRT gene transduction caused the increase of F-dUMP
To estimate the intracellular level of F-dUMP in cells transduced with the UPRT gene, Competitive Ligand Binding Assay for FdUMP was carried out. AsPC-1 cells infected with AxCA-UPRT or AxCA-lacZ were exposed to 0.001 mg/ml 5-FU for 48 hr and the intracellular levels of F-dUMP were compared. The levels of F-dUMP were 46.8 ± 13.9 pmol/1 × 107 cell in cells infected with AxCA-UPRT and 10.9 ± 1.67 pmol/1 × 107 cell in cells infected with AxCA-lacZ, respectively. The level of F-dUMP in cells treated with AxCA-UPRT/5FU was significantly increased compared to the control group (p < 0.0143). Transduction of the UPRT gene caused an increase of intracellular F-dUMP (Fig. 1).
Sensitivity to 5-FU of UPRT-transduced pancreatic cancer cells
Cytotoxic assay was carried out to determine whether the pancreatic cancer cells transduced with the UPRT gene were increased in 5FU sensitivity. The IC50 of either parental cells or cells infected with control vectors (AxCA-lacZ) were similar, ranging from 0.1–18 μg/ml. AsPC-1 cells were the most resistant to 5FU among the tested cells. The transduction of the UPRTase gene increased the sensitivity of the AsPC-1 cells. The IC50 of 5FU was decreased from 18 μg/ml to 0.25 μg/ml by the gene transduction, indicating a 72-fold increase in sensitization. The sensitivity to 5FU of all pancreatic cancer cells tested was increased by the UPRT gene transduction (Table I).
|(μ g/ml)||(μ g/ml)|
Effect of UPRT-expressing E1 deficient adenovirus vector and 5-FU in vivo
The antitumor effect of AxCA-UPRT/5-FU treatment was evaluated in the animal model of intraperitoneally disseminated AsPC-1 tumors. As shown in Figure 2, animals treated with AxCA-UPRT/5FU showed a significant regression of tumors at the mesentery. The mice with AxCA-lacZ/5FU showed diffuse metastatic tumors spreading all over the mesentery. The mean tumor weights were 0.220 ± 0.029 g, 0.542 ± 0.084 g and 0.530 ± 0.073 g in the AxCA-UPRT/5FU, AxCA-UPRT and AxCA-lacZ/5FU groups, respectively (n = 6). The tumor weight in animals treated with AxCA-UPRT/5FU was significantly reduced compared to that of the control group (p < 0.009) (Fig. 3). A dose escalation of 5FU to 20 mg/kg/day could not augment the antitumor effect compared to 5FU at 10 mg/kg/day (data not shown).
When using high dose AxCA-UPRT (3 × 108 pfu) to augment the antitumor effect further, all such treated mice had watery, unformed stool 5 days after beginning the 5FU treatment. Mice that had diarrhea were sacrificed and the jejunum was removed and observed under a light microscope. The jejunum of mice treated with AxCA-UPRT/5FU showed a loss of cilia. In contrast, either AxCA-UPRT alone or 5FU alone caused no apparent damage in the small intestines (data not shown). We also estimated the in vivo transduction efficiency and spread by the intra-abdominal administration of various doses of replication-incompetent vector (AxCAlacZ). As demonstrated in Figure 6, lacZ gene expression was detected in disseminated tumors, normal intestines and peritoneum. The level of gene expression was intensified both in tumor and normal tissue in a dose-dependent manner.
These results suggested that high dose administration of Adv expressing the therapeutic gene could intensify the expression of the gene, but result in adverse effects induced by the gene transduction to the normal tissue. To augment the gene expression in disseminated tumors selectively, we developed AxE1AdB-UPRT that selectively replicates in tumor cells and expresses the UPRTase gene.
In vitro viral replication assay
Viral replication assay was carried out to evaluate the replication of AxE1AdB, AxE1AdB-UPRT and Ad5dlx in AsPC-1 and WI38 in vitro. In WI38, they grew 250∼1,900 times less efficiently 10 days after Adv infection. In AsPC-1 cells, both AxE1AdB and AxE1AdB-UPRT replicated as efficiently as Ad5dlx (Fig. 4).
We also carried out immunostaining with monoclonal antibody (MAb) against Adv-hexon protein to confirm the selective replication in cancer cells. AxE1AdB-UPRT infected AsPC-1 showed marked expression of adenoviral hexon protein. In contrast, AxE1AdB-UPRT infected WI38 showed little expression. Both AxE1AdB and AxE1AdB-UPRT replicated in AsPC-1 cells with an abnormality in the p53 gene, but in normal cells the replication of these viruses was highly restricted (Fig. 5).
Selective replication of AxE1AdB co-administered with E1-deficient adv expressing lacZ Gene in vivo
To determine the selective replication of AxE1AdB in vivo, mice with intraperitoneal disseminated tumors were treated with intraperitoneal injection of AxE1AdB and AxCA-lacZ. Staining for the expression of β-galactosidase by X-gal, the intensity and selectivity of expression were evaluated. As shown in Figure 6a–c, the peritoneal cavity of a mouse that received AxE1AdB (1 × 108 pfu) with AxCA-lacZ (1 × 108 pfu) showed selectively enhanced expression of β-galactosidase in disseminated tumors. In contrast, that of a mouse receiving AxCA-lacZ (1 × 108 pfu) showed weak expression of β-galactosidase in both tumors and normal organs such as intestine and peritoneum (Fig. 6a). As described above, the dose escalation of AxCA-lacZ (1 × 109 pfu) caused an unselective augmentation of the gene expression in both tumors and normal tissue (Fig. 6b). Immunohistochemical analysis also showed the expression of adenoviral hexon protein in tumors but not in normal intestines (Fig. 7).
Transgene amplification mediated by replication competent adv
To quantify the amplification of the UPRT gene, AsPc-1 cells were infected with AxCA-UPRT or AxE1AdB-UPRT. The viral DNA from the infected cells was digested with the restriction enzyme XbaI. The XbaI digestion of AxE1AdB-UPRT gene yielded 2.6 kb fragments containing the UPRT gene sequence. That of AxCA-UPRT yielded 1.5 kb fragments containing the UPRT gene sequence. Southern blot analysis was carried out using DIG labeled probes of the UPRT gene sequence that hybridize to both the 2.6 kb fragment from AxE1AdB-UPRT and the 1.5 kb fragment from AxCA-UPRT. We demonstrated previously that the viral DNA in the AxE1AdB-infected cells is increased to 3,000 copies/infected cell 30 hr after infection.18 As shown in Figure 8, AsPC-1 cells infected with AxE1AdB-UPRT contained large amounts of the UPRT gene at levels similar to those of 293 cells infected with the vector for viral propagation. Compared to the cells infected with AxE1AdB-UPRT, cells infected with the replication-incompetent vector had little UPRT gene expression after longer exposure (data not shown). These data indicated that restricted replication competent Adv amplified the UPRT transgene in P53 dysfunctional cells.
Therapeutic effect of UPRT-expressing restricted replication-competent Adv and 5-FU in vivo
The antitumor effect of AxE1AdB-UPRT/5-FU treatment was tested in the animal model of intraperitoneal disseminated AsPC-1 tumors. As shown in Figure 9, the tumor weight in animals treated with AxE1AdB-UPRT/5FU was significantly reduced compared to that of the AxCA-UPRT/5FU group (p = 0.0281). The mean tumor weights were 0.137 ± 0.061 g, 0.242 ± 0.064 g and 0.478 ± 0.043 g in the AxE1AdB-UPRT/5FU, AxCA-UPRT/5FU and AxCA-UPRT groups, respectively. The treatment with AxE1AdB-UPRT/5FU showed a dramatic tumor reduction without adverse effects.
We studied the effectiveness of gene therapy with UPRT mediated by Adv for overcoming the 5FU resistance of pancreatic cancers. The transduction of UPRT gene mediated by Adv resulted in an increase of the FdUMP (Fig. 1) and consequent sensitivity of various pancreatic cancer cells to 5FU (Table I). Although in vivo gene transduction of UPRT followed by the administration of 5FU brought about regression of intraperitoneal pancreatic tumors (Figs. 2,3), the Adv caused adverse effects such as severe diarrhea with dehydration at a dose high enough to obtain a complete reduction of the tumors. As indicated in Figure 6, mice receiving a high dose of AxCA-lacZ showed marked expression of β-galactosidase both in tumors and other organs including normal intestine. F-dUMP converted from 5-FU in normal mucosal cells inhibits the cell growth, thereby causing the gastrointestinal toxicity of 5FU.29 Because Adv-mediated gene delivery is a powerful tool for gene expression in vivo, strategies to selectively amplify the expression of therapeutic genes in tumors are essential for the treatment of disseminated tumors.
One of the possible strategies includes Adv expressing the therapeutic gene driven by a tumor specific promoter. It was reported that CEA-producing gastric cancer was selectively destroyed by Adv expressing the suicide gene driven by CEA. Not all cancers produce CEA, however, and the activity of tumor-specific promoters is usually weaker than that of the constitutive promoters.13, 27
According to significant advances in the understanding of genetic abnormalities of pancreatic cancers, it is demonstrated that pancreatic cancer often has mutations in P53, K-RAS, SMAD4/DPC4 and P16/MTS1.17 Targeting these genetic mutations offers some hope for developing a novel approach to pancreatic cancer, however a definitive strategy has not been established for the clinical situation. We studied the tumor-selectivity of gene transduction mediated by adenovirus lacking the 55 kDa E1B. As described in previous reports, an adenovirus lacking the 55 kDa E1B selectively replicates in cancer cells with a P53 abnormal gene.15 As we described previously, the coinfection of AxE1AdB and E1 deficient adenovirus results in the replication of both AxE1AdB and E1 deficient adenovirus.18, 30 In our present study, selective replication in the intraperitoneal tumors was demonstrated by coinfection of AxE1AdB and AxCA-lacZ. The marked expressions of β-galactosidase only in tumors indicated the selective replication of AxE1AdB and augmented expression of the coinfected transgene in disseminated tumors. As compared to gene delivery by the conventional (replication-incompetent) Adv, the restricted replication-competent adenovirus has the advantage of being able to increase the transgene.
We developed restricted replication-competent Adv expressing UPRT (AxE1AdB-UPRT). As we expected, selective replication and amplification of the UPRT gene in cells with an abnormal P53 gene were shown to have occurred. In contrast, human fibroblast cells with a normal P53 gene were shown to be resistant to the replication of AxE1AdB-UPRT. The restricted replication-competent Adv augmented the anti-tumor effect without adverse effects compared to the replication incompetent Adv (Fig. 9). Previous researchers reported that treatment with intravenous injection of ONYX-015 (replication competent Adv lacking E1B55K) to liver metastases was effective. ONYX-015 localized and replicated within the tumor, thereby demonstrating a significant anti-tumor effect without adverse effects.31 It is speculated that the treatment with AxE1AdB-UPRT followed by 5FU would be also effective against metastatic liver tumors. We are now investigating the efficacy of AxE1AdB-UPRT plus 5FU for metastatic liver tumors.
The Phase I clinical trial for head and neck cancer has already been carried out with ONYX-015 because of the safety of gene therapy with restricted replication competent Adv. In this clinical trial, intra-tumoral injection with a maximum dose of 1011 pfu did not cause any serious adverse effects and ONYX-015 was demonstrated to replicate in P53 dysfunctional tumors but not in normal tissue.32 Although recent studies showed that the viral replication does not entirely depend on the P53 status,26, 33 the phase II clinical trial of intra-tumoral ONYX-015 in combination with CDDP and 5FU in patients with recurrent head and neck cancer showed a high response rate.34 Because various combination strategies with ONYX-015 have been investigated, further studies about the relationship between the viral replication and the status of the P53 pathway are needed to determine the clinical indications for restricted replication competent Adv.
It was reported that a replication competent adenovirus containing 2 independent prostate-specific promoters improves the selectivity of the viral replication in prostate specific antigen (PSA)-producing cancer cells. In PSA non-producing cancer cells, the Adv replicates 10,000∼100,000 times less efficiently than in PSA-producing cancer cells.35 Another approach is to engineer another cytolytic virus capable of replicating in tumor cells. Kasuya et al.36 described a gene therapy with a mutant herpes simplex virus (hrR3) that replicates only in the rapidly dividing cells of disseminated pancreatic cancers. Intraperitoneal delivery of hrR3 and ganciclovir improved the survival in a murine model of peritoneal dissemination of pancreatic cancer. The clinical safety of replicating vectors largely depends on whether the replication is restricted to tumors. Further studies are necessary to develop tumor-specific replicating vectors for the clinical situation.
In summary, AxE1AdB-UPRT can replicate and amplify UPRT gene expression in pancreatic cancer cells with an abnormality in P53. The AxE1AdB-UPRT/5FU system is a promising tool for targeting intraperitoneal disseminated pancreatic cancer.
We thank the Kyowa Hakko Corporation for assaying intracellular F-dUMP. We also appreciate the expert technical assistance of H. Fujimura, E. Shibuya and K. Inabe. Our study was supported in part by two grants-in-aid for scientific research (10470251, 12877187) from the Japanese Ministry of Education, Science, Sports and Culture.
- 3Chemotherapy of neoplastic disease. In: HardmanJG, LimbirdLH, MolinoffPB, RuddonRW, GilmanAG. The pharmacological basis of therapeutics, 9th ed. New York: Pergamon Press, 1995. 1225–87., .
- 30Restricted replication-competent adenovirus: a novel strategy for cancer gene therapy. Int J Clin Oncol 2000;5: 147–52., , .