Introduction

Antiangiogenic therapy shows promise as a strategy for cancer treatment because most tumor growth is dependent on the formation of new blood vessels ¹. Investigations into the molecular basis of angiogenesis have demonstrated that tumors express a number of autocrine and paracrine factors that activate or facilitate this process ². Among them, vascular endothelial growth factor (VEGF) plays an important role. Many studies have demonstrated a positive correlation between VEGF expression and malignant potential in various cancers ³, which indicates VEGF as an attractive target for therapeutic intervention.

Soluble Flt-1 (sFlt-1) is an endogenous selective inhibitor of VEGF, which can inhibit the mitogenic response to VEGF in culture by directly sequestering VEGF or heterodimerizing with the extracellular ligand-binding region of the membrane-spanning Flt-1 and KDR ^4,5,6. Previous studies applied the sflt-1 gene (approved gene symbol FLT1) to an antiangiogenic therapy approach and achieved good results, such as ex vivo transfection of cancer cells with a plasmid encoding sFlt-1 ⁴, regional or systemic administration of an adenovirus-mediated sflt-1 cDNA ^7,8, and intraperitoneal transduction of a soluble flt-1 cDNA using HVJ cationic liposomes ⁹. In addition, the mutant consisting of the first three ectodomains of Flt-1 was reported to be able to bind VEGF with affinity similar to that of sFlt-1 ¹⁰, indicating the sFlt-1(1–3) could be sufficient as a potent inhibitor of VEGF-mediated angiogenesis ¹¹.

However, efficient delivery of antiangiogenic genes to tumors remains a major obstacle. Nonviral and replication-deficient viral vectors have thus far been used with limited success in cancer gene therapy, due mainly to low transduction efficiency and poor distribution throughout the solid tumor mass. Because the presence of sustained and high-level antiangiogenic proteins is essential to maintain tumor neovessel inhibition and hence tumor growth suppression ¹², better vectors need to be explored.

In recent years, conditionally replicative adenoviruses have been developed and the primary advantage is that viral replication will amplify the initial input dose in a tumor-dependent fashion, allowing extensive tumor infection and expansive oncolysis through the cytopathic effect ^13,14,15. Among them, an E1B-55-kDa-deleted adenovirus, ONYX-015, has been reported as a good example that preferentially targets and kills p53-dysfunctional tumor cells rather than normal tissues ¹⁶. However, it is of note that ONYX-015 is safe but not potent enough as a monotherapy to cause complete tumor regression or generate sustained clinical responses ^17,18. Therefore, it is appealing to build a gene delivery system within the context of an oncolytic virus so as to complement and synergize with the lytic function of virus ¹⁹. The feasibility of this approach was demonstrated by arming the oncolytic adenovirus with a suicide gene such as CD or HSV-tk ^20,21. In previous studies, we also constructed an armed oncolytic adenovirus system, ZD55-gene, which not only is deleted of the E1B-55-kDa gene, similar to ONYX-015, but also possesses the capability of integrating foreign therapeutic genes ²². It exhibited a significant antitumor efficacy because of the additional functions of such therapeutic genes as CD, SMAC, and TRAIL ^22,23,24.

As a natural extension, we constructed ZD55-sflt-1, an E1B-55-kDa-deleted oncolytic adenovirus vector engineered to express sFlt-1(1–3). It showed improved antitumor and antiangiogenic effects both in vitro and in vivo compared with ONYX-015 and replication-deficient Ad-sflt-1. Moreover, the potency of ZD55-sflt-1 could be further increased by the addition of a chemotherapeutic agent, 5-FU.

Results

Characterization of ZD55-sflt-1

ZD55 is an E1B-55-kDa-deleted oncolytic adenovirus, which is similar to ONYX-015 ²². Here, we inserted the sflt-1(1–3) expression cassette driven by the human CMV promoter into ZD55 to construct ZD55-sflt-1. To characterize ZD55-sflt-1, we infected SW620 colon cancer cells with various viruses and examined the expression of E1A and E1B-55-kDa. Like ONYX-015, ZD55-sflt-1 expressed E1A protein but failed to express E1B-55-kDa protein. In contrast, wild-type adenovirus 5 expressed both proteins, whereas the replication-deficient Ad-sflt-1 did not express either of them (Fig. 1A).

Figure 1.

Characterization of ZD55-sflt-1. (A) Western blot assay of E1A and E1B-55-kDa expression. SW620 cells were mock infected (lane 2) or infected with ONYX-015 (lane 3), Ad-sflt-1 (lane 4), ZD55-sflt-1 (lane 5), or Ad-WT (lane 6) at a m.o.i. of 10. Uninfected HEK293 cells were used as a positive control (lane 1). Cell lysates were subjected to Western blot assay with anti-Ad2 E1A Ab and anti-Ad5 E1B-55-kDa Ab. (B) Tumor selective cytotoxicity of ZD55-sflt-1. Cell monolayers were infected with the indicated viruses at different m.o.i. Uninfected cells were included as a control. Seven days later, cells were stained with crystal violet. Similar results were obtained in three independent experiments.

Full figure and legend (142K)

To examine further the viral properties of ZD55-sflt-1, we performed an in vitro cytotoxicity assay. ZD55-sflt-1 showed marked cytopathic effect (CPE) against all tumor cell lines tested (SW620, BEL7404, Bcap37, HeLa), even at very low multiplicities of infection (m.o.i.) (0.1–1), which was comparable to that of ONYX-015 and about 100-fold greater than that of Ad-sflt-1 (CPE could be detected only at a m.o.i. of 100). When tested in normal cell lines (NHLF, WI38), all three viruses caused no significant CPE, even at a m.o.i. of 100. It is of note that both ZD55-sflt-1 and ONYX-015 caused about 100-fold greater attenuation in killing normal cells compared to tumor cells (Fig. 1B). These results indicate that ZD55-sflt-1 has the correct structure and can induce tumor-specific cytopathic effect like ONYX-015.

Increased Expression of sflt-1 from ZD55-sflt-1

Oncolytic adenovirus can act as a potential vehicle to increase the potency of its armed therapeutic gene ²⁵, so we first quantified the expression level of sFlt-1 from ZD55-sflt-1. As shown in Fig. 2A, the secretion of sFlt-1 generated by SW620 cells after ZD55-sflt-1 infection was 102 21.4 ng/ml, about 30-fold greater than that after Ad-sflt-1 injection (3.5 0.9 ng/ml). However, in a normal cell line (NHLF), very low sFlt-1 expression was detected in both groups. In HEK293 cells, which can render replication-deficient adenovirus replication competent, similar amounts of sFlt-1 were produced after ZD55-sflt-1 (229 15.6 ng/ml) or Ad-sflt-1 (223 18.2 ng/ml) infection (P > 0.05). These data indicate that ZD55-sflt-1 can specifically increase the expression level of sFlt-1 in tumor cells because of its tumor-specific replication ability.

Figure 2.

Increased expression of sFlt-1 from ZD55-sflt-1. (A) sFlt-1 expression in vitro. Cells were infected with ZD55-sflt-1 or Ad-sflt-1 at a m.o.i. of 0.5. The conditioned medium was harvested at 7 days postinfection and measured using a commercial sVEGFR1 ELISA. The means SD (n = 3) are shown (^**P < 0.01). (B) sFlt-1 expression in tumors. After intratumoral injection of ZD55-sflt-1 or Ad-sflt-1 (1 10⁹ pfu), tumors were collected at indicated days and sFlt-1 expression was quantified as above. The means SD from three mice are shown (^**P < 0.01).

Full figure and legend (72K)

We obtained similar results in vivo. A considerable amount of sFlt-1 was detected from the ZD55-sflt-1-treated SW620 tumors on day 7 after intratumoral injection (12 1.6 ng/mg) and it remained detectable on day 20 (5 0.8 ng/mg). In contrast, the expression of sFlt-1 from Ad-sflt-1-treated tumor was low, peaking on day 7 (3 0.4 ng/mg) and then dropping rapidly (Fig. 2B).

Increased Bioactivity of sFlt-1 in ZD55-sflt-1

We subsequently evaluated whether the biological function of sFlt-1 as an antagonistic agent to VEGF would be augmented along with the increased expression level from ZD55-sflt-1. In the virus-inactivated medium prepared from both Ad-sflt-1- and ZD55-sflt-1-infected SW620 cells, VEGF-induced human umbilical vein endothelial cell (HUVEC) proliferation was strongly inhibited, and the effect of ZD55-sflt-1 was more obvious than that of Ad-sflt-1, with cell viability of 35 11.3 and 53 7.8%, respectively (P < 0.01). In contrast, in the medium from ONYX-015-infected SW620 cells, no significant inhibitory effect was observed (Fig. 3A).

Figure 3.

In vitro biological activity of sFlt-1. (A) sFlt-1 inhibited the proliferation of HUVECs. SW620 cells were uninfected (control) or infected with viruses at a m.o.i. of 10. Supernatants were obtained after 48 h and virus was inactivated. HUVECs were incubated in the above conditioned media (CM) and stimulated with hVEGF (10 ng/ml). Cell proliferation was measured 72 h later using the MTT assay. Data are expressed as percentage of control and shown as means SD of three independent experiments (^**P < 0.01). (B) sFlt-1 inhibited the tube formation of HUVECs. HUVECs were incubated with the CM and stimulated with hVEGF. 48 h later, the tube formations were quantified, and the means SD (n = 3) are shown (^**P < 0.01). (C) sFlt-1 did not affect tumor cell growth in vitro. Subconfluent SW620 cells were incubated with the CM and cell viability was determined at the indicated days using the MTT assay. The means SD (n = 3) are shown (P > 0.05).

Full figure and legend (103K)

We further tested the in vitro antiangiogenic activity of sFlt-1 in a differentiation inhibition assay. While the ability of HUVECs to form tube-like structures on an ECMatrix surface was nearly unaffected by incubation with the virus-inactivated medium from ONYX-015-infected SW620 cells, the incubation with medium from Ad-sflt-1- or ZD55-sflt-1-infected cells strongly reduced this activity (with tube formation ratios of 66 9.2 and 29 6.3%, respectively, P < 0.01; Fig. 3B).

In addition, we also evaluated the bioactivity of sFlt-1 on tumor cells by observing the proliferation of SW620 cells when incubated in the virus-inactivated medium as above. No significant difference was observed among any of the groups (P > 0.05; Fig. 3C). All these data indicate that sFlt-1 exerts a therapeutic effect toward endothelial cells but not tumor cells. This bioactivity is correlated with its expression level.

Antitumoral Efficacy of ZD55-sflt-1 in Nude Mice

Given the potential of ZD55-sflt-1 in vitro, we then investigated its antitumor efficacy in vivo. Because antiangiogenic therapy is likely to be most effective in the low tumor burden state ²⁶, we administered various treatments to mice with small tumor volumes (about 30 mm³). As shown in Fig. 4A, we observed much more rapid tumor growth in the PBS-treated group than in the other groups. On day 33, the last day when all animals were alive, the mean tumor volumes of PBS-, ONYX-015-, Ad-sflt-1-, and ZD55-sflt-1-treated mice were 641.4, 208.0, 246.1, and 115.8 mm³, respectively. Although no significant difference existed between the ONYX-015- and the Ad-sflt-1-treated groups, significant antitumor efficacy was observed in the ZD55-sflt-1-treated group relative to them (P < 0.05). Tumors in the ZD55-sflt-1-treated mice experienced a slower growth in the first 6 weeks, followed by further regression. Established tumors were completely eradicated in two of eight mice. Moreover, we observed a long-term survival rate in the ZD55-sflt-1-treated group (50%), whereas the rate in the ONYX-015-treated mice was 12.5% and no mouse survived in the PBS- and Ad-sflt-1-treated groups at the end of this experiment (P < 0.05, Fig. 4B).

Figure 4.

Antitumor efficacy of ZD55-sflt-1 in nude mice bearing SW620 xenograft tumors. (A) When tumor volumes reached about 30 mm³, mice were intratumorally injected with 5 10⁸ pfu of ZD55-sflt-1, Ad-sflt-1, ONYX-015 (n = 8), or PBS (n = 6) every other day for four injections. The tumor volume was measured at 3-day intervals and is presented as means SD. (B) Kaplan–Meier survival analysis of mice with established SW620 tumor after treatment with ZD55-sflt-1, Ad-sflt-1, ONYX-015, or PBS. (C) Histopathological responses in liver after a single i.t. injection of 10⁹ particles of ONYX-015 (b), Ad-sflt-1 (c), or ZD55-sflt-1 (d). Liver sample from PBS-treated mice served as the control (a). (H&E staining; original magnification 400).

Full figure and legend (251K)

In addition, we examined liver toxicity associated with high expression of sFlt-1 by ZD55-sflt-1. ZD55-sflt-1 did not significantly alter serum liver enzyme levels after i.t. administration (Table 1). Hematoxylin and eosin (H&E)-stained sections from ZD55-sflt-1-treated mice also showed no pathological signs of hepatotoxicity (Fig. 4C).

Table 1 - Serum liver chemistries.

Full table

Mechanisms Underlying the Antitumor Effect of ZD55-sflt-1

To obtain an insight into the mechanisms underlying the significant antitumor effect of ZD55-sflt-1, we first subjected tumors to in situ apoptosis detection. Apoptotic cells were detected in all groups, while the apoptotic index of the ZD55-sflt-1-treated group (8 0.7%) was eightfold higher than that of PBS-treated group (P < 0.01) and two- to threefold higher than that of the ONYX-015- and Ad-sflt-1-treated groups (P < 0.01; Figs. 5A and B). Then, we examined the microvessel density (MVD) by immunostaining with anti-CD31 antibody. The MVD in tumors derived from the ZD55-sflt-1-treated group (5 2.4) was significantly reduced compared with tumors from the other groups (23 3.7 in the PBS-treated group and 18 2.6 in the ONYX-015-treated group, P < 0.01; 8 2.2 in the Ad-sflt-1-treated group, P < 0.05; Figs. 5C and D).

Figure 5.

Increased apoptotic rates and decreased vessel densities in tumors from ZD55-sflt-1-treated mice. (A) 30 days after injection, tumor sections derived from each group were subjected to an in situ cell apoptosis detection kit to detect apoptotic cells and (C) immunostained with CD31 antibody to assess tumor angiogenesis ((a) PBS; (b) ONYX-015; (c) Ad-sflt-1; (d) ZD55-sflt-1. Original magnification, 400). (B) The apoptotic index and (D) MVD were the mean values of sections from three mice in each group and are presented as means SD (n = 3, ^*P < 0.05, ^**P < 0.01).

Full figure and legend (366K)

Based on the above observations, we presumed that the antitumor effect of ZD55-sflt-1 was the consequence mainly of both direct oncolytic effects of ZD55 and indirect antiangiogenic effects of sFlt-1, which was confirmed by histological examination (Fig. 6). Four days after ZD55-sflt-1, Ad-sflt-1, or ONYX-015 injection, all tumor sections stained positively for Adv hexon protein, while only samples from ZD55-sflt-1- or ONYX-015-treated mice remained positive after 20 days (data not shown). This suggests that the long persistence of the viral replication may cause the destruction of tumor cells. Moreover, consistent with previous ELISA (Fig. 2B), we detected much higher sFlt-1 expression in ZD55-sflt-1-treated tumors than in Ad-sflt-1-treated tumors, whereas no positive staining was observed in the ONYX-015- or the PBS-treated group.

Figure 6.

Adv hexon and sFlt-1 expression. Adv hexon and sFlt-1 expression in each treated group (PBS, ONYX-015, Ad-sflt-1, and ZD55-sflt-1) was detected by immunostaining with anti-hexon or anti-sFlt-1 antibody, respectively (original magnification, 400).

Full figure and legend (334K)

Synergistic Antitumor Efficacy of ZD55-sflt-1 in Combination with 5-FU

For the future possibility of applying ZD55-sflt-1 to clinical use, we investigated whether a synergistic effect could be achieved by combining ZD55-sflt-1 with a chemotherapeutic agent, such as 5-FU. As shown in Fig. 7A, although neither an m.o.i. of 0.1 of ZD55-sflt-1 nor 1 ng/ml 5-FU caused significant cytotoxicity by day 7, the cell viability was strongly decreased to 25.9% when we applied both ZD55-sflt-1 and 5-FU under the same conditions (P < 0.05).

Figure 7.

Synergistic effect of ZD55-sflt-1 with 5-FU. (A) SW620 cells were treated with ZD55-sflt-1 at a m.o.i. of 0.1 or 5-FU at 1 ng/ml alone or in combination. Cell viability was determined by comparing the OD₅₅₀/OD₆₅₀ of treated cells with that of mock-infected wells. Results are means SD of triplicates (^**P < 0.01). (B) When tumor volumes reached about 100 mm³, mice were injected with 5-FU (i.p., 30 mg/kg for 7 days) or ZD55-sflt-1 (i.t., 5 10⁸ pfu, every other day for four injections) alone or in combination. The tumor volume was measured at 3-day intervals and is presented as means SD (n = 8).

Full figure and legend (95K)

We also evaluated the synergistic efficacy in the SW620 tumor model. When tumors reached about 100 mm³, we administered 5-FU, ZD55-sflt-1, or ZD55-sflt-1 followed by 5-FU. 5-FU did not suppress tumor growth significantly, whereas administration of ZD55-sflt-1 resulted in a significant antitumor effect. One of eight ZD55-sflt-1-treated mice experienced a complete eradication of established tumor. Most importantly, treatment with ZD55-sflt-1 followed by 5-FU further augmented this activity. Analysis of fractional tumor volume indicated a synergistic effect between ZD55-sflt-1 and 5-FU. For example, on day 33, the combination group showed a 1.2-fold higher inhibition of tumor growth over an additive effect. With time, there was a progressive improvement in antitumor activity, which induced a complete tumor regression in four of eight treated animals (Fig. 7B).

Discussion

In this study, with the aim of achieving synergy between antiangiogenic gene therapy and viral therapy, we constructed a novel E1B-attenuated oncolytic adenovirus, ZD55-sflt-1. The significant antitumor effect and long-term survival of mice suggested the potential of using ZD55-sflt-1 as a vector for cancer therapy.

Several classes of gene therapy-based therapeutics have been traditionally associated with nonreplicating viral-based gene delivery vehicles ²⁵. However, since transduction of cytotoxic genes will cause death of the transduced cells and subsequently lead to attenuation of viral replication, choosing the appropriate gene(s) with which to arm an oncolytic virus is still a question. Genes with indirect antitumor effects would be a better choice ²⁷. Here, we demonstrated that the incorporation of an antiangiogenic gene does not interfere with viral characters. Like ONYX-015, ZD55-sflt-1 could specifically induce cytopathic effects in a wide range of human tumor cells but not normal cells (Fig. 1B). Immunohistochemical staining of tumor sections with antibodies against Adv hexon protein also confirmed that ZD55-sflt-1 could effectively spread throughout the tumor as ONYX-015 did (Fig. 7).

In addition, virally expressed angiogenic inhibitor is a complement to viral therapy. The antiangiogenic approach targets the tumor endothelium instead of the tumor, which can amplify the tumor killing effect ²⁸. Although sFlt-1 did not affect tumor cell growth in vitro (Fig. 3C), it could be efficacious against tumor growth in vivo (Fig. 4A). The antitumor effect of sFlt-1 was correlated with a decreased tumor vascularization and an increase in apoptotic tumor cells (Fig. 5).

On the other hand, the potency of antiangiogenic inhibitor could be increased when using oncolytic adenovirus as vehicle. To impede the growth of tumor-associated vascularization, continuous infusion or large doses of angiogenic inhibitor(s) are often necessary ¹². Several studies on sFlt-1 gene therapy using plasmids, adenovirus, or AAV as vector exhibited inhibition of tumor growth, but their efficiency was always limited or transient because of the insufficient production of sFlt-1 ^4,5,6,7,8. Oncolytic adenovirus is a much better transgene carrier. We observed more effective antiangiogenic effects of sFlt-1 mediated by ZD55-sflt-1 than by the replication-deficient Ad-sflt-1 in SW620 xenografts of nude mice (Fig. 4). Consistent with this result, a higher expression of sFlt-1 in tumor cells and in tumors treated with ZD55-sflt-1 vs Ad-sflt-1 was observed (Fig. 2).

Another problem with antiangiogenic therapy is that angiogenic inhibitors just induce tumor dormancy rather than kill tumor cells, so there is a need for combination of other agents that kill tumor cells directly. The oncolytic adenoviral vector can efficiently play this role as it can specifically induce tumor cell death. However, it does not exclude the use of chemotherapy or radiation. The synergistic effects of ZD55-sflt-1 and 5-FU were observed with decreased tumor cell viability and inhibited tumor growth: four of eight mice experienced complete eradication of their established tumors (Fig. 7). This probably is attributable to both the chemosensitivity effect of adenovirus E1A protein ¹⁷ and the synergistic effects of chemotherapy and antiangiogenic drugs ²⁹. In addition, tumor-associated angiogenesis can be promoted by several other cytokines or growth factors (FGF, IL-8, TNF-, and so on) that are not involved with the inhibition effect of sFlt-1 ⁸. Hence, combining multiple antiangiogenic factors would achieve a potent inhibition in a wide variety of tumors. Indeed, we are currently investigating such an approach by constructing an oncolytic adenovirus to express both sFlt-1(1–3) and the kringle 5 of plasminogen. A combination of genes that target totally different aspects of tumor biology (e.g., prodrug-converting enzyme, immunostimulatory) could also be used.

While it is important to consider the therapeutic factors and how they may synergize with the oncolytic virus to maximize therapeutic benefit ²⁵, it is equally important to take into account some undesired side effects that may exist. Mahasreshti et al. recently reported that intravenous administration of the sflt-1 gene via replication-deficient adenoviral vectors caused unacceptable hepatotoxicity in an ovarian tumor model ³⁰. Bearing this point in mind, we specifically targeted the virus to tumor cells by deleting the E1B-55-kDa gene. Although many investigations have suggested that the replication of this mutant adenovirus may not be entirely dependent upon p53 status ^31,32, clinical trails demonstrated the safety of ONYX-015 even when it was systemically administered ¹⁸. In this study, prolonged survival time in ZD55-sflt-1-treated mice was observed (Fig. 4B). Histopathological examination and serum biochemistry assay also confirmed the absence of hepatotoxicity (Fig. 4C, Table. 1). However, it is of note that only intratumoral injection of ZD55-sflt-1 was used here to treat local tumor. The safety of ZD55-sflt-1 needs to be further elucidated, especially if it were to be systemically administered to treat tumor metastasis. Future studies aiming at either transductional targeting (e.g., through genetic modification of the viral capsid ³³) or transcriptional targeting (e.g., using tumor-selective promoters to control foreign gene expression) may offer additional safety features in regulating the long-term expression of an antiangiogenic gene, such as sflt-1, using oncolytic adenoviruses in vivo.

In summary, we have successfully extended our technological platform by creating ZD55-sflt-1. Antiangiogenic gene therapy could enhance the therapeutic effects of viral therapy in a tumor cell-specific manner, which could be further enhanced by other modalities. Our results are extended by a recent report using a human telomerase-targeted replicative adenovirus as an antiangiogenic gene transfer vector to treat human tumor ³⁴. A virtue of our strategy is that it could be targeted to different types of tumor cells. Tumor cells retaining functional p53 are more responsive to antiangiogenic therapy, whereas those with mutated p53 would be more susceptible to ZD55-induced cytolysis, even if less responsive to antiangiogenic therapy ³⁵. Although more investigations are required for future use, our preliminary data offer the promise of using antiangiogenic genes in the context of oncolytic viruses as a novel approach for cancer therapy.

Materials and Methods

Cell lines and culture

The human cell lines SW620 (colorectal cancer), HeLa (cervical cancer), BEL7404 (hepatocarcinoma), NHLF (normal lung fibroblasts), and WI38 (normal embryonic lung cell) were purchased from ATCC (Rockville, MD, USA); Bcap37 (breast cancer) was purchased from the Shanghai Cell Collection, CAS; HEK293 was obtained from Microbix Biosystems, Inc. (ON, Canada). Cells were cultured in DMEM or RPMI 1640 (GIBCO BRL, Grand Island, NY, USA) supplemented with 10% heat-inactivated fetal bovine serum (FBS; GIBCO BRL), 4 mM glutamine, 50 U/ml penicillin, and 50 g/ml streptomycin. HUVECs were originally isolated from umbilical cord veins ³⁶ and cultured in M199 supplemented with 20% FBS.

Virus construction

Plasmid pZD55 was constructed in our lab ²² by deleting the E1B-55-kDa region (amino acids 215–354) from pXC1 (Microbix). The expression cassette for the sflt-1(1–3) gene was digested from pCA13-sflt-1 using BglII and cloned into pZD55 to generate pZD55-sflt-1. Adenovirus ZD55-sflt-1 was constructed by homologous recombination techniques using pZD55-sflt-1 and the adenovirus packaging plasmid pBHGE3 (Microbix Biosystems) in HEK293 cells with Effectene Transfection Reagent (Qiagen, Germany). A standard replication-deficient adenovirus, Ad-sflt-1, was constructed through cotransfection of the adenovirus shuttle vector containing sflt-1(1–3) cDNA with an E1A/B-deleted adenoviral backbone vector. ONYX-015 was kept in our lab.

Viruses were plaque purified, propagated on HEK293 cells, and purified by CsCl gradient according to standard techniques. Functional particle titers of all adenoviruses were determined by plaque assay on HEK293 cells.

Western blot

Cells were infected with viruses for 48 h and lysed in sample buffer (62.5 mM Tris–HCl, pH 6.8, 2% SDS, 10 mM glycerol, 1.55% DTT). Cell lysates were separated by 12% SDS–PAGE and transferred to nitrocellulose membranes (Amersham). The expression of E1A and E1B-55-kDa protein was determined by hybridization with the rabbit polyclonal antibody anti-Ad2E1A (Santa Cruz Biotechnology, Santa Cruz, CA, USA) and the rat polyclonal antibody anti-Ad5E1B-55kD (Oncogene), respectively. Reactivity was visualized by an enhanced chemiluminescence system (Amersham Life Sciences, Inc., Arlington Heights, IL, USA).

Cytotoxicity assay

Subconfluent cell lines were infected with viruses at various m.o.i. Seven days later, cells were exposed to 2% crystal violet in 20% methanol for 15 min, washed, and documented by photographs.

Measurement of sFlt-1 in culture media and tumors

The expression of sFlt-1 in culture media and tumors were measured using a commercial sVEGFR1 ELISA following the instructions of the manufacturer (R&D Systems, Germany). Cells were infected with ZD55-sflt-1 or Ad-sflt-1 at a m.o.i. of 0.5. The conditioned media (CM) were harvested on day 7 and stored at -70°C until the assay. For sFlt-1 expression in tumors, mice were injected i.t. with 10⁹ pfu of ZD55-sflt-1 or Ad-sflt-1. Tumors were extracted on the indicated days and homogenized in cold lysis buffer (50 mM Tris–HCl, pH 8.0, 150 mM NaCl, 0.02% sodium azide, 100 g/ml PMSF, 1 g/ml aprotinin, 1% NP-40). The homogenates were centrifuged at 12,000g and the supernatants were collected.

Proliferation assay

SW620 cells were infected with viruses at a m.o.i. of 10 for 72 h. The supernatants were collected and heated at 56°C to inactivate the viruses. HUVECs (10⁴/well) were seeded in serum-free DMEM for 24 h and then exchanged with 50% diluted CM as above. Fifteen minutes later, the supernatants were supplement with 10 ng/ml rhVEGF (PeproTech EC Ltd.). The relative VEGF-induced proliferation of HUVECs was assayed after 72 h using a method based on the metabolization of MTT (Sigma). The effect of sFlt-1 on a tumor cell line (SW620) was examined in a similar way.

Tube formation assay

A 96-well plate was coated with 50 l ECMatrix (Chemicon, Temecula, CA, USA) and HUVECs incubated in 50 l of CM were seeded onto its surface. After 30 min, 10 ng/ml rhVEGF was added to the medium. The tube formations were quantified 2 days later by counting the number of connecting branches between two discrete endothelial cells (20 magnification).

Animal experiments

Female Balb/c nude mice at 4–6 weeks were obtained from the Animal Research Committee of the Institute of Biochemistry and Cell Biology (Shanghai, China). The xenograft tumor model was established by subcutaneously injecting 10⁶ SW620 cells into the right flank of mice. Once the tumors reached about 30 mm³, the animals were randomly divided into ZD55-sflt-1, Ad-sflt-1, ONYX-015 (n = 8), and PBS (n = 6) groups. A dose of 5 10⁸ pfu in 100 l of PBS was administered intratumorally every other day four times. Tumor growth was measured at 3-day intervals. The tumor volumes were estimated as tumor volume (mm³) = length width²/2. The survival time was also calculated.

For toxicology studies, mice were injected i.t. with 10⁹ pfu of ZD55-sflt-1, Ad-sflt-1, ONYX-015, or PBS. Four days after treatment, liver samples were subjected to histology examination and the serum levels of AST, ALT, ALP, and GGT were determined by automated colorimetric assays.

Immunohistochemistry assay

Paraffin tumor sections (8–10 m) were treated with xylene, rehydrated in graded ethanol, and transferred to PBS. To detect the expression of sFlt-1 or Ad hexon, sections were incubated with anti-sFlt-1 antibody (20 g/ml; Santa Cruz Biotechnology) or anti-hexon antibody (1:20; Chemicon), followed by peroxidase-conjugated anti-rabbit or anti-goat IgG (1:200; Santa Cruz Biotechnology). The staining signal was amplified by peroxidase-conjugated avidin–biotin complex (Vector Laboratories, Burlingame, CA, USA) and developed with DAB solution. Hematoxylin was used as counterstain. The stained sections were examined in a Zeiss photomicroscope (Carl Zeiss, Inc., Thornwood, NY, USA) equipped with a three-chip charge-coupled device color camera (Model DXC-960 MD; Sony Corp., Tokyo, Japan).

Detection and quantification of apoptosis

Apoptotic cells in tumor specimens were detected using an in situ cell apoptosis detection kit (Sino-American Biotech Co., China) according to the manufacturer's instruction. Apoptotic cells were counted under a light microscope (400 magnification) in five randomly chosen fields, and the apoptosis index was calculated as a percentage of all cancer cells in these fields.

Microvessel density assay

Tumor sections were incubated with polyclonal anti-CD31 (DAKO, Glostrup, Denmark) for immunohistochemistry as described above. The microvessel density was determined by the method of Weidner et al. ³⁷. Microvessel counting was performed at 200 magnification in three fields (0.74 mm²/field).

Statistical analysis

Differences among the treatment groups were assessed by the unpaired Student t test and the one-way analysis of variance. Differences among the results of in vivo survival experiments were assessed by the Kaplan–Meier assay. P values <0.05 were regarded as statistically significant.

References

1. Regulier, E. et al. (2001). Adenovirus-mediated delivery of antiangiogenic genes as an antitumor approach. Cancer Gene Ther. 8: 45–54. | Article | PubMed | ISI | ChemPort |
2. Huss, W. J. et al. (2001). Angiogenesis and prostate cancer: identification of a molecular progression switch. Cancer Res. 61: 2736–2743. | PubMed | ISI | ChemPort |
3. Sako, A. et al. (2004). Transduction of soluble Flt-1 gene to peritoneal mesothelial cells can effectively suppress peritoneal metastasis of gastric cancer. Cancer Res. 64: 3624–3628. | Article | PubMed | ChemPort |
4. Goldman, C. K. et al. (1998). Paracrine expression of a native soluble vascular endothelial growth factor receptor inhibits tumor growth, metastasis, and mortality rate. Proc. Natl. Acad. Sci. USA 95: 8795–8800. | Article | PubMed | ChemPort |
5. Kendall, R. L. and Thomas, K. A. (1993). Inhibition of vascular endothelial cell growth factor activity by an endogenously encoded soluble receptor. Proc. Natl. Acad. Sci. USA 90: 10705–10709. | Article | PubMed | ChemPort |
6. Kendall, R. L. et al. (1996). Identification of a natural soluble form of the vascular endothelial growth factor receptor, Flt-1, and its heterodimerization with KDR. Biochem. Biophys. Res. Commun. 226: 324–328. | Article | PubMed | ISI | ChemPort |
7. Kong, H. L. et al. (1998). Regional suppression of tumor growth by in vivo transfer of a cDNA encoding a secreted form of the extracellular domain of the flt-1 vascular endothelial growth factor receptor. Hum. Gene Ther. 9: 823–833. | PubMed | ISI | ChemPort |
8. Takayama, K. et al. (2000). Suppression of tumor angiogenesis and growth by gene transfer of a soluble form of vascular endothelial growth factor receptor into a remote organ. Cancer Res. 60: 2169–2177. | PubMed | ISI | ChemPort |
9. Mori, A. et al. (2000). Soluble Flt-1 gene therapy for peritoneal metastases using HVJ-cationic liposomes. Gene Ther. 7: 1027–1033. | Article | PubMed | ISI | ChemPort |
10. Barleon, B. et al. (1997). Mapping of the sites for ligand binding and receptor dimerization at the extracellular domain of the vascular endothelial growth factor receptor Flt-1. J. Biol. Chem. 272: 10382–10388. | Article | PubMed | ISI | ChemPort |
11. Ye, C. et al. (2004). sFlt-1 gene therapy of follicular thyroid carcinoma. Endocrinology 145: 817–822. | PubMed | ChemPort |
12. Lalani, A. S. et al. (2004). Anti-tumor efficacy of human angiostatin using liver-mediated adeno-associated virus gene therapy. Mol. Ther. 9: 56–66. | Article | PubMed | ISI | ChemPort |
13. Kirn, D. H. and McCormick, F. (1996). Replicating viruses as selective cancer therapeutics. Mol. Med. Today 2: 519–527. | Article | PubMed | ISI | ChemPort |
14. Russell, S. J. (1994). Replicating vectors for cancer therapy: a question of strategy. Semin. Cancer Biol. 5: 437–443. | PubMed | ChemPort |
15. Burroughs, K. D. et al. (2004). Potentiation of oncolytic adenoviral vector efficacy with gutless vectors encoding GMCSF or TRAIL. Cancer Gene Ther. 11: 92–102. | Article | PubMed | ISI | ChemPort |
16. Bischoff, J. R. et al. (1996). An adenovirus mutant that replicates selectively in p53-deficient human tumor cells. Science 274: 373–376. | Article | PubMed | ISI | ChemPort |
17. Khuri, F. R. et al. (2000). A controlled trial of intratumoral ONYX-015, a selectively-replicating adenovirus, in combination with cisplatin and 5-fluorouracil in patients with recurrent head and neck cancer. Nat. Med. 6: 879–885. | Article | PubMed | ISI | ChemPort |
18. Vasey, P. A. et al. (2002). Phase I trial of intraperitoneal injection of the E1B-55-kd-gene-deleted adenovirus ONYX-015 (dl1520) given on days 1 through 5 every 3 weeks in patients with recurrent/refractory epithelial ovarian cancer. J. Clin. Oncol. 20: 1529–1562.
19. Bauzon, M. et al. (2003). Multigene expression from a replicating adenovirus using native viral promoters. Mol. Ther. 7: 526–534. | Article | PubMed | ChemPort |
20. Kathrin, M. et al. (2002). Enzyme-activated prodrug therapy enhances tumor-specific replication of adenovirus vectors. Cancer Res. 62: 6089–6098. | PubMed | ChemPort |
21. Wildner, O. et al. (1999). Adenoviral vectors capable of replication improve the efficacy of HSVtk/GCV suicide gene therapy of cancer. Gene Ther. 6: 57–62. | Article | PubMed | ISI | ChemPort |
22. Zhang, Z. L. et al. (2003). An armed oncolytic adenovirus system, ZD55-gene, demonstrating potent antitumoral efficacy. Cell Res. 13: 481–489. | Article | PubMed | ChemPort |
23. Qiu, S. B. et al. (2004). Combination of targeting gene–virotherapy with 5-FU enhances antitumor efficacy in malignant colorectal carcinoma. J. Interferon Cytokine Res. 24: 219–230. | Article | PubMed | ChemPort |
24. Pei, Z. F. et al. (2004). Augmenting antitumor activity of TRAIL in hepatocellular carcinoma by Smac using tumor selective replicating adenovirus. Hepatology 39: 1371–1381. | Article | PubMed | ISI | ChemPort |
25. Hermiston, T. W. and Kuhn, I. (2002). Armed therapeutic viruses: strategies and challenges to arming oncolytic viruses with therapeutic genes. Cancer Gene Ther. 9: 1022–1035. | Article | PubMed | ISI | ChemPort |
26. Bergers, G. et al. (1999). Effects of angiogenesis inhibitors on multistage carcinogenesis in mice. Science 284: 808–812. | Article | PubMed | ISI | ChemPort |
27. Motoi, F. et al. (2000). Effective gene therapy for pancreatic cancer by cytokines mediated by restricted replication-component adenovirus. Hum. Gene Ther. 11: 223–235. | Article | PubMed | ISI | ChemPort |
28. Denekamp, J. (1993). Angiogenesis, neovascular proliferation and vascular pathophysiology as targets for cancer therapy. Br. J. Radiol. 66: 181–196. | PubMed | ISI | ChemPort |
29. Bello, L. et al. (2001). Low-dose chemotherapy combined with an antiangiogenic drug reduces human glioma growth in vivo. Cancer Res. 61: 7501–7506. | PubMed | ISI | ChemPort |
30. Mahasreshti, P. J. et al. (2003). Intravenous delivery of adenovirus-mediated soluble Flt-1 results in liver toxicity. Clin. Cancer Res. 9: 2701–2710. | PubMed | ChemPort |
31. Harada, J. N. and Berk, A. J. (1999). p53-independent and -dependent requirements for E1B-55 K in adenovirus type 5 replication. J. Virol. 73: 5333–5344. | PubMed | ISI | ChemPort |
32. Rothmann, T. et al. (1998). Replication of ONYX-015, a potential anticancer adenovirus, is independent of p53 status in tumor cells. J. Virol. 72: 9470–9478. | PubMed | ISI | ChemPort |
33. Roelvink, P. W. et al. (1999). Identification of a conserved receptor-binding site on the fiber proteins of CAR-recognizing Adenoviridae. Science 286: 1568–1571. | Article | PubMed | ISI | ChemPort |
34. Zhang, Q. et al. (2004). Effective gene–viral therapy for telomerase-positive cancers by selective replicative-competent adenovirus combining with endostatin gene. Cancer Res. 64: 5390–5397. | Article | PubMed | ISI | ChemPort |
35. Yu, J. L. et al. (2002). Effect of p53 status on tumor response to antiangiogenic therapy. Science 295: 1526–1528. | Article | PubMed | ISI | ChemPort |
36. Jaffe, E. A. et al. (1973). Culture of human endothelial cells derived from umbilical cord veins: identification by morphologic and immunologic criteria. J. Clin. Invest. 52: 2745–2756. | PubMed | ISI | ChemPort |
37. Weidner, N., Semple, J. P., Welch, W. R. and Folkman, J. (1991). Tumor angiogenesis and metastasis—correlation in invasive breast carcinoma. N. Engl. J. Med. 324: 1–8. | PubMed | ISI | ChemPort |

Acknowledgements

We thank Li Ma for the provision of sflt-1(1–3)-bearing plasmids and also thank Lanying Sun, Songbo Qiu, and Buqing Shu for professional technical assistance and useful suggestions. This work was supported by the Key Project of the Chinese Academy of Sciences (No. KSCX2-3-06), the National Natural Science Foundation of China (No. 30120160823), a Chinese National "863" High Tech Project Foundation grant (No. 2002AA216021), the 973 Project (No. 2004CB518804), the UTE project of CIMA, and a grant from the Instituto Carlos III C03/02 (Spain).