Improved efficacy of chemotherapy by parvovirus-mediated sensitisation of human tumour cells

Induction of an antitumoral immune response by wild-type adeno-associated virus type 2 in an in vivo model of pancreatic carcinoma

We analyzed the immunologic impact of adeno-associated virus type 2 (AAV-2), a small single-stranded parvovirus with tumorsuppressive properties, on DSL6A pancreatic carcinoma in syngeneic rats. Established tumors of animals treated with AAV-2 or mock infected were resected (Ro), and DSL6A cells were rechallenged on the different site. Eleven (92%) of 12 mock-infected animals but only 3 (25%) of 12 AAV-2-treated animals redeveloped tumors. Adeno-associated virus type 2 infection provoked systemic raises in monocytes and neutrophils numbers and in levels of the proinflammatory monocyte chemoattractant protein 1 and interleukin 10. Adeno-associated virus type 2-treated tumors were infiltrated with monocytes, macrophages, natural killer cells, CD4+ T cells, and especially CD8+ T cells. In cytotoxicity assays, AAV-2-infected DSL6A tumor cells were recognized by lymphocytes from AAV-2-treated animals and from controls. Yet, uninfected DSL6A cells were exclusively killed by lymphocytes from AAV-2-treated animals. Additionally, those lymphocytes displayed high natural killer cell activity but failed to attack unrelated tumor targets. Taken together, these results suggest that the antiviral response toward AAV-2 cross-activates the immune system toward simultaneously present tumor disease. This and the known potential to significantly reduce toxic side effects of chemotherapy make nonpathogenic viruses such as AAV-2 as "1-agent combination therapy" to an interesting treatment option of residual tumor disease.

Suicide gene therapy of sarcoma cell lines using recombinant adeno-associated virus 2 vectors

Soft-tissue sarcomas are mesenchymal tumors that respond poorly to systemic chemotherapy. Suicide gene therapy may be an alternative treatment strategy. Here we show a high susceptibility of human sarcoma cell lines for recombinant adeno-associated virus 2 (rAAV-2) suicide vectors: connective tissue sarcoma (HS-1), fibrosarcoma (HT-1080), Ewing sarcoma (RD-ES), Askin tumor (SK-N-MC), rhabdomyosarcoma (A-204) and soft-tissue sarcoma (WSKL-1). Several vectors containing the thymidine kinase (TK) gene under the control of either the cytomegalovirus promoter or the elongation-factor 1 alpha (EF1alpha) promoter were cloned and tested. Higher expression levels of the transgene were observed in the sarcoma lines when using the EF1alpha-suicide gene-containing vectors. A complete eradication of rAAV-2-EF1alpha-TK/eGFP (TK/enhanced green fluorescent protein fusion gene)-transduced tumor cells was shown following exposure to ganciclovir (2.5 microg/ml) in vitro, while at this dose level > 90% of mock-transduced tumor cells survived. Xenotransplantation tumor models (intraperitoneal, subcutaneous) for the human sarcoma cell line HS-1 were established in nonobese diabetic/severe-combined immunodeficient mice. Mice transplanted with rAAV-2-EF1alpha-TK/eGFP-transduced and ganciclovir-exposed tumor cells survived > 5 months while in the nontransduced group all mice had died approximately 1 month after inoculation. These data hold promise for further development of rAAV-2-based suicide gene therapy of sarcomas.

Adeno-associated virus serotypes 1 to 5 mediated tumor cell directed gene transfer and improvement of transduction efficiency

BACKGROUND

Gene therapy is an attractive new approach for the treatment of cancer. Therefore, the development of efficient vector systems is of crucial importance in this field. Different adeno-associated virus (AAV) serotypes have been characterized so far, which show considerable differences in tissue tropism. Consequently, we aimed to characterize the most efficient serotype for this application.

METHODS

To exclude all influences other than those provided by the capsid, all serotypes contained the same transgene cassette flanked by the AAV2 inverted terminal repeats. We systematically compared these vectors for efficiency in human cancer cell directed gene transfer. In order to identify limiting steps, the influence of second-strand synthesis and proteasomal degradation of AAV in a poorly transducible cell line were examined.

RESULTS

AAV2 was the most efficient serotype in all solid tumor cells and primary melanoma cells with transduction rates up to 98 +/- 0.3%. Transduction above 70% could be reached with serotypes 1 (in cervical and prostate carcinoma) and 3 (in cervical, breast, prostate and colon carcinoma) using 1000 genomic particles per cell. In the colon carcinoma cell line HT-29 proteasomal degradation limited AAV1-AAV4-mediated gene transfer. Moreover, inefficient second-strand synthesis prevents AAV2-mediated transgene expression in this cell line.

CONCLUSIONS

Recent advances in AAV-vector technology suggest that AAV-based vectors can be used for cancer gene therapy. Our comparative analysis revealed that, although AAV2 is the most promising candidate for such an application, serotypes 1 and 3 are valid alternatives. Furthermore, the use of self-complementary AAV vectors and proteasome inhibitors significantly improves cancer cell transduction.

Prevention of chemotherapy-related toxic side effects by infection with adeno-associated virus type 2

Drug resistance and toxic side effects are major limiting factors in the clinical use of antineoplastic chemotherapy. Patients with pancreatic cancer generally do not benefit from chemotherapy. The nonpathogenic adeno-associated virus type 2 (AAV-2) has been shown to sensitize human tumor cells to gamma irradiation and chemotherapeutic drugs. In the present study, we characterized the therapeutic role of AAV-2 infection in combination with 5-fluorouracil (5-FU)-based chemotherapy on pancreatic cancer cells in an animal model. In Lewis rats bearing s.c. implants of syngeneic DSL6A pancreatic cancer cells, intratumoral infection with AAV-2 (MOI 10E8 i.u.) in combination with 5-FU (5 or 50 mg/kg body weight) resulted in significantly reduced tumor growth and prolonged survival time compared with 5-FU single therapy. Most surprisingly, AAV-2-infected rats remained in a much better physical condition compared to their noninfected counterparts. While rats treated with 5-FU single therapy lost weight, were sluggish and died within 4 months after tumor implantation, animals infected with AAV showed much better vigilance, with body weight, leukocyte number and hemoglobin levels similar to healthy rats. In particular, 5-FU-related side effects like thrombocytopenia and leukopenia were significantly reduced in animals treated with the combination regimen. By in vitro analysis, human (Capan-1 and DANG) pancreatic cancer cell lines were shown to be sensitized to 5-FU chemotherapy to an extent similar to DSL6A cells. AAV-2 infection enhanced 5-FU-induced apoptosis by a factor of 8 to 14 in both human and rat pancreatic cancer cell lines. The data suggest that infection with the nonpathogenic AAV-2 significantly improves both chemotherapy efficacy and physical appearance and offers a novel strategy in cancer treatment.

Enhancement of cisplatin-induced apoptosis by infection with adeno-associated virus type 2

The non-pathogenic human adeno-associated virus, AAV, has been shown to sensitize human cancer cells and experimental tumors towards the action of chemotherapeutic agents such as cisplatin. Since chemotherapeutic drugs mainly involve the induction of apoptosis, we investigated whether 1 possible mechanism of AAV-mediated sensitization of human tumor cells may result from an enhancement of cisplatin-induced apoptosis. In HeLa and A549 cells, infection with AAV type 2 (AAV-2) increased cisplatin-induced DNA fragmentation but had no cytotoxic effect by itself. This enhanced apoptosis appeared to be mediated at least in part by a component of the viral capsid since empty or UV-inactivated AAV-2 particles were also able to boost cisplatin-induced DNA fragmentation. Interestingly, these effects were not observed after infection with AAV type 5 (AAV-5) or the autonomous parvovirus, H-1. AAV-2-mediated enhancement of apoptosis was not associated with a modification of the expression of CD95 ligand, CD95 receptor or other death receptors, as shown by RT-PCR and RNase protection assay. In contrast, using the mitochondrial fluorescent dye, JC-1 in flow cytometry, AAV-2 infection was found to further reduce the mitochondrial transmembrane potential after treatment with cisplatin in a caspase-independent manner, suggesting that increase of apoptosis by AAV-2 occurred at the mitochondrial level. In contrast, in cells of the small cell lung cancer line, P693, an enhancement of cisplatin-induced DNA fragmentation was not observed after infection with AAV-2. In these cells, sensitization to cisplatin-toxicity was associated with cell cycle arrest in G2/M. The data indicate that in the absence of viral gene expression, AAV-2-mediated sensitization to cisplatin involves multiple cellular pathways promoting cell death signals in a cell type-dependent manner. The results further support that AAV-2 particles may be appropriate adjuvants for improving cancer chemotherapy and may also have consequences regarding AAV-2-based vectors for gene therapy.

Adenoassociated virus type 2-induced inhibition of the human papillomavirus type 18 promoter in transgenic mice

The epithelium of the cervix uteri has been reported to be frequently coinfected with both human papillomaviruses (HPV) and helper virus-dependent adenoassociated viruses (AAV). Seroepidemiological data suggest that AAV infection could inhibit cervical cancer that is caused by specific ("high-risk") types of papillomaviruses. In vitro, infection with AAV type 2 (AAV-2) or transfection of AAV-2 early (rep) genes has been shown to inhibit transformation by papillomaviruses. To analyze the effects of AAV on HPV in vivo, we studied the influence of AAV-2 infection on the promoter activity of high-risk HPV type 18 (HPV-18) in mice, transgenic for sequences of the upstream regulatory region (URR) of HPV-18 controlling transcription of the reporter gene, lacZ. Transgenic animals (or tongue cells thereof, explanted and grown in culture) were treated with dexamethasone to induce the HPV-18 promoter. Simultaneously they were (i) infected with AAV, (ii) inoculated with AAV virus-like particles (VLPs; empty capsids), or (iii) mock infected. Inoculation with AAV-2 or VLPs inhibited activation of the HPV-18 promoter. In vitro, in baby hamster kidney cells transfected with the HPV-18-lacZ construct, tissue extracts from AAV-infected animals suppressed the HPV-18 URR to a similar extent as AAV infection did. Down-regulation of the HPV-18 promoter was less efficient with extracts from animals inoculated with VLPs and was not observed with extracts from uninfected or dexamethasone-treated animals. This indicates that AAV induces cellular factor(s) in vivo capable of mediating down-regulation of the HPV-18 promoter also in cells in vitro. In contrast, promoters of the low-risk HPV types (HPV-6, HPV-11) were not influenced by AAV infection as opposed to promoters of the high-risk types (HPV-18 and HPV-16).

Superior gene transfer into solid tumour cells than into human mobilised peripheral blood progenitor cells using helpervirus-free adeno-associated viral vector stocks

Autologous peripheral blood progenitor cell (PBPC) grafts can be contaminated with tumour cells that potentially give rise to relapse following myeloablative therapy and PBPC transplantation. Adeno-associated virus (AAV)-based vectors produced by a new adenovirus-free technique are a gene delivery system which may be applicable for tumour cell purging. To test for the host range of these vectors, solid tumours of clinical relevance and normal CD34+ PBPC were selected as target cells for an AAV-vector, encoding the green-fluorescent protein (GFP) as the indicator gene. At a multiplicity of infection (MOI) of 100: 79.94% +/- 14.36% (mean +/- SEM) of the connective tissue sarcoma cell line (HS-1) and 64.84% +/- 6.91% of the cervical carcinoma cell line cells (HeLa-RC) expressed GFP while the other cell lines tested (1 ovarian tumour, 1 germ cell tumour, 1 osteosarcoma, 2 small cell lung cancer) ranged between 2.82% and 11.94%. Optimising the transduction protocol by use of higher MOIs of up to 500 and by pretreatment with the tyrosine kinase inhibitor, genistein, resulted in up to 95.97% and 94.10% green-fluorescent HS-1 and HeLa-RC cells, respectively. In contrast, only 1.39% +/- 0.51% of the normal haematopoietic CD34+ progenitor cells expressed GFP at a MOI of 100. The differential infectivity between HS-1 and CD34+ cells was maintained after tumour cell spiking in leucapheresis products. Our observations suggest that AAV-based vectors may prove useful for purging of autologous PBPC grafts from solid tumour cells.

Enhanced sensitivity of small cell lung cancer cell lines to cisplatin and etoposide after infection with adeno-associated virus type 2

In previous studies we have reported the sensitisation of human tumour cells to gamma irradiation and chemotherapeutic drugs upon infection with the human non-pathogenic adeno-associated virus type 2 (AAV-2) in vitro and in vivo. Treatment of small cell lung cancer (SCLC) is consistently hampered by relapses due to the selection of chemotherapy-resistant cell clones. Hence, we were interested to test whether selection of chemotherapy-resistant SCLC cells might be reduced or even prevented if chemotherapy is applied in combination with AAV-2 infection. In vitro proliferation assays indicated that the number of proliferating cells, after combined treatment with cisplatin and etoposide, can be significantly reduced by concomitant AAV-2 infection, as compared with treated but non-infected controls. H446 SCLC cells, which show resistance to etoposide/cisplatin chemotherapy (compared with a cell line which was never chemotherapeutically treated before, like NCI-H209) were significantly more sensitive after AAV-2 infection, suggesting that the therapeutic efficacy of chemotherapy in SCLC can be enhanced even if the cells are already relatively resistant to chemotherapy. Similarly, in vivo growth of tumours induced by inoculation of SCLC cells into immunocompromised nude mice was reduced more efficiently in AAV-2-infected animals compared with tumours in mice treated with chemotherapeutic drugs alone. These data extend and further support our previous reports on AAV functions which might be useful in improving the efficacy of chemotherapeutic drugs used in human cancer treatment.

Dose-dependent regression of HeLa cell-derived tumours in SCID mice after parvovirus H-1 infection

Parvoviruses of rodents are endowed with oncosuppressive properties. In particular, parvoviral infections protect host animals from spontaneous and chemical- or virus-induced tumour initiation in laboratory animals. The present study was undertaken to substantiate the capacity of parvovirus H-1 to inhibit therapeutically the growth of established tumours originating from human carcinoma cells implanted in recipient mice. To this end, quickly growing s.c. carcinomas were established by injection of human cervical carcinoma cells (HeLa) into immunodeficient (SCID) mice. Tumour-bearing mice subsequently were inoculated with H-1 at various multiplicities of infection. H-1 virus infection led to regression of tumours, the onset and efficiency of which were dose-dependent.

Infection of primary cells by adeno-associated virus type 2 results in a modulation of cell cycle-regulating proteins

It has been demonstrated that infection of primary human cells with adeno-associated viruses (AAV) leads to a decrease in cellular proliferation and to growth arrest. We analyzed the molecular basis of this phenomenon and observed that infection with AAV type 2 (AAV2) had an effect on several factors engaged in the control of the mammalian cell cycle. In particular, all of the pRB family members, pRB, p107, and p130, which are involved in G1 cell cycle checkpoint control, were affected. After infection, a shift from hyper- to hypophosphorylated forms was observed. Cyclins A and B1, which are required for G1/S transition and progression into mitosis, respectively, were downregulated at the transcriptional level as well as at the protein level, whereas the G1 cyclins D1 and E remained unaffected. In addition, the steady-state levels of cyclin-dependent kinases CDK1 and CDK2 and of transcription factor E2F-1 were diminished. Of all the factors known to be involved in phosphorylation of pRB family proteins, only the CDK inhibitor p21WAF1 exhibited a response to AAV2 infection. p21WAF1 mRNA was quickly and progressively upregulated in a p53-independent manner over at least 72 h. Consistent with the increased p21WAF1 protein levels, cyclin E- and cyclin A-dependent kinase activities declined to low levels and E2F-p130-cyclin-CDK2 complexes were disrupted. From these data, we conclude that the major effect of AAV2 infection on primary human fibroblasts appears to be upregulation of p21WAF1 gene expression and thus cell cycle arrest by the suppression of pRB family protein phosphorylation.