Targeted cancer immunotherapy with oncolytic adenovirus coding for ...

5 downloads 274 Views 764KB Size Report
Nov 10, 2011 - oncolytic adenovirus producing a full-length human mAb. High mAb .... oncolytic control viruses lacking aCTLA4 and recombinant anti-.
Gene Therapy (2011), 1 - 11 & 2011 Macmillan Publishers Limited All rights reserved 0969-7128/11 www.nature.com/gt

ORIGINAL ARTICLE

Targeted cancer immunotherapy with oncolytic adenovirus coding for a fully human monoclonal antibody specific for CTLA-4 JD Dias1,2,6, O Hemminki1,2,6, I Diaconu1,2, M Hirvinen1,2, A Bonetti3, K Guse1,2, S Escutenaire1,2, A Kanerva1,4, S Pesonen1,2, A Lo¨skog5, V Cerullo1,2 and A Hemminki1,2 Promising clinical results have been achieved with monoclonal antibodies (mAbs) such as ipilimumab and tremelimumab that block cytotoxic T lymphocyte-associated antigen-4 (CTLA-4, CD152). However, systemic administration of these agents also has the potential for severe immune-related adverse events. Thus, local production might allow higher concentrations at the target while reducing systemic side effects. We generated a transductionally and transcriptionally targeted oncolytic adenovirus Ad5/3-D24aCTLA4 expressing complete human mAb specific for CTLA-4 and tested it in vitro, in vivo and in peripheral blood mononuclear cells (PBMCs) of normal donors and patients with advanced solid tumors. mAb expression was confirmed by western blotting and immunohistochemistry. Biological functionality was determined in a T-cell line and in PBMCs from cancer patients. T cells of patients, but not those of healthy donors, were activated by an anti-CTLA4mAb produced by Ad5/3-D24aCTLA4. In addition to immunological effects, a direct anti-CTLA-4-mediated pro-apoptotic effect was observed in vitro and in vivo. Local production resulted in 43-fold higher (Po0.05) tumor versus plasma anti-CTLA4mAb concentration. Plasma levels in mice remained below what has been reported safe in humans. Replication-competent Ad5/3-D24aCTLA4 resulted in 81-fold higher (Po0.05) tumor mAb levels as compared with a replication-deficient control. This is the first report of an oncolytic adenovirus producing a full-length human mAb. High mAb concentrations were seen at tumors with lower systemic levels. Stimulation of T cells of cancer patients by Ad5/3-D24aCTLA4 suggests feasibility of testing the approach in clinical trials. Gene Therapy advance online publication, 10 November 2011; doi:10.1038/gt.2011.176 Keywords: oncolytic adenoviruses; anti-CTLA4 monoclonal antibody; CD152 INTRODUCTION A key revelation in cancer immunotherapy has been the realization that induction of an antitumor immune response is not sufficient for efficacy. Instead, because of the immunesuppressive nature of advanced tumors, downregulation of inhibitory mechanisms is also required.1,2 One of these key regulatory pathways involves cytotoxic T lymphocyte-associated antigen-4 (CTLA-4, CD152), which acts against B7/CD28-mediated stimulation. The antitumor efficacy of antibodies (Abs) antagonistic to CTLA-4 has been shown previously in several tumor models.3,4 Presently, two fully human monoclonal Abs (mAbs) against CTLA-4 are in clinical development: IgG1 ipilimumab (formerly MDX-010) and IgG2 tremelimumab (formerly CP-675206). Several published studies attest to the biological and clinical activity of ipilimumab and tremelimumab in patients with melanoma and other cancers.5 - 7 However, although antitumor activity has been seen in many trials, including a landmark randomized phase-3 melanoma trial,6 also severe and even fatal side effects have been reported.5 - 7 Side effects relate to normal tissue exposure whereas efficacy is determined by reduction of immune suppression at the tumor. Therefore, clinical data suggest that local production of an anti-CTLA4mAb would be appealing for increasing both safety and efficacy.

CTLA-4 is a type-I transmembrane protein of the Ig superfamily, expressed upon activation by T lymphocytes as a covalent homodimer that functions as an inhibitory receptor for the co-stimulatory molecules B7.1 (CD80) and B7.2 (CD86).7 CTLA-4 blockade with mAbs results in increased interleukin-2 (IL-2) and interferon-g (IFN-g) production by lymphocytes, and increased expression of major histocompatibility complex class-I molecules.8,9 Based on extensive pre-clinical studies several antitumor mechanisms have been proposed: (A) Anti-CTLA4mAbs can block immune-suppressive signaling derived from CTLA-4 molecules on the surface of activated T cells.10 (B) An important inhibitory subset of T cells (regulatory T cells, T-Regs) constitutively expresses CTLA-4 and can bind to B7 molecules on dendritic cells, which are key antigen-presenting cells.11 Subsequent upregulation of indoleamine-2,3-dioxygenase and other suppressive circuits results in the tolerization of T cells, instead of cytotoxic action.12 (C) CTLA-4-expressing T cells (including T-Regs) may also bind directly to activated T cells, because B7 co-stimulatory molecules are expressed on the surface of activated human T cells. CTLA-4blocking mAbs interfere with this suppressive signaling and result in the local expansion of tumor antigen-specific T cells.11 (D) CTLA-4 is expressed on the surface of many tumor cells,13 presumably

1 Cancer Gene Therapy Group, Molecular Cancer Biology Program & Transplantation Laboratory & Haartman Institute, & Finnish Institute for Molecular Medicine, University of Helsinki, Helsinki, Finland; 2HUSLAB, Helsinki University Central Hospital, Helsinki, Finland; 3Molecular Neurology Research Program, University of Helsinki, Helsinki, Finland; 4 Department of Obstetrics and Gynecology, Helsinki University Central Hospital, Helsinki, Finland and 5Department of Immunology, Genetics and Pathology, Uppsala University, Helsinki, Sweden. Correspondence: Dr A Hemminki or V Cerullo, Biomedicum, Haartmaninkatu 8, University of Helsinki, Helsinki 00014, Finland. E-mail: vincenzo.cerullo@helsinki.fi or akseli.hemminki@helsinki.fi 6 These authors equally contributed to this work. Received 4 April 2011; revised 15 August 2011; accepted 5 October 2011

Oncolytic adenovirus expressing anti-CTLA4mAb JD Dias et al

2

reflecting direct immune-suppressive action on B7 of cytotoxic T-cells and dendritic cells. Interestingly, CTLA-4-blocking mAbs can induce direct killing of tumor cells by triggering apoptosis,7,13 and in vivo this is enhanced by Ab-dependent cellular cytotoxicity.14 (E) Tumor-expressed CTLA-4 may trigger increased indoleamine2,3-dioxygenase in tumor-infiltrating dendritic cells, and CTLA-4specific mAbs would also block this effect.7 Thus, five different mechanisms make anti-CTLA-4 an appealing and potent antitumor approach, but side effects, including mortality, associated with systemic exposure have resulted in safety concerns.6,15 Adenoviral gene therapy has been proposed as a novel treatment alternative for advanced cancer,16 and oncolytic adenoviruses have been explored for enhanced tumor transduction and amplification of antitumor effect.17 These viruses have a cytolytic nature, that is, the replicative life cycle of the virus results in host cell death. We have shown previously that adenoviruses featuring 5/3 serotype chimerism show significantly enhanced gene delivery and tumor cell-killing effect compared with viruses with wild-type capsids.18 - 21 Also, the oncolytic adenoviruses featured here feature a loss-of-function 24-bp deletion (D24) in the constant region-2 domain of the adenoviral E1A gene,22,23 which renders the virus unable to bind retinoblastoma (Rb), required for replication in normal cells. In tumor therapy this deletion is transcomplemented by cells deficient in the Rb/p16 pathway, including most if not all cancer cells.24 Tumor specificity of such adenoviral mutants has been demonstrated previously.17,22,23,25,26 In this study, we constructed a transductionally and transcriptionally targeted oncolytic adenovirus armed with a fully human mAb specific for CTLA-4. Ad5/3-D24aCTLA4 was evaluated in vitro and in vivo for antitumor efficacy. Importantly, we saw stimulation of T cells of patients with advanced cancer, but not of T cells of healthy individuals. RESULTS Expression of anti-CTLA4mAb from Ad5/3-D24aCTLA4 We generated two different 5/3 chimeric adenoviruses bearing the cDNA sequence for an IgG2-type anti-CTLA4mAb (Figure 1). The coding sequence of anti-CTLA4mAb was introduced into the 6.7K/gp19K deletion of adenoviral E3A (replication-competent Ad5/3-D24aCTLA4) or into the deleted E1 driven by the cytomegalovirus (CMV) promoter (replication-deficient Ad5/3-aCTLA4). Viral structures were confirmed by PCR and sequencing of modified regions. In western blotting, cells infected with Ad5/3-D24aCTLA4 and Ad5/3-aCTLA4 expressed the approximately 150-kDa anti-CTLA4mAb under native conditions, and the approximately 50-kDa heavy chains and the approximately 25-kDa light chains under denaturing conditions in supernatants at 48 h after infection (Figure 2a). Tumor cells infected with oncolytic Ad5/3-D24aCTLA4 expressed three-fold more anti-CTLA4mAb compared with replication-deficient Ad5/3-aCTLA4 in vitro (Supplementary Figure 1). Functionality of anti-CTLA4mAb expressed from Ad5/3-D24aCTLA4 The functionality of anti-CTLA4mAb produced by Ad5/3-D24aCTLA4 and Ad5/3-aCTLA4 was examined by measuring the IL-2 production of stimulated Jurkat cells (clone 6.1), as described previously.27,28 The anti-CTLA4mAb binds to cell-surface CTLA-4, blocking the immunosuppressive interaction with recombinant B7 (rB7). This analysis indicates that anti-CTLA4 inhibitory activity was found in the supernatants of cells infected with Ad5/3-D24aCTLA4 and Ad5/3-aCTLA4 (Supplementary Figure 2 and Figure 2b). Ad5/3-D24 and Ad5/3Luc1 are the isogenic oncolytic and nononcolytic control viruses lacking aCTLA4 and recombinant antiCTLA4mAb was used as an additional positive control. Gene Therapy (2011), 1 -- 11

Figure 1. Construction of a double-targeted and anti-CTLA4-armed oncolytic adenovirus. A schematic illustration of the genome of the oncolytic adenovirus Ad5/3-D24aCTLA4 and replication-deficient Ad5/3-aCTLA4, as well as the unarmed oncolytic Ad5/3-D24 (positive control) and replication-deficient Ad5/3-Luc1 (negative control).

In order to further confirm the ability of virus-expressed anti-CTLA4mAb to block signaling through CTLA-4, recombinant CTLA-4 (rCTLA-4) was added to the previous assay (Supplementary Figures 2e and f, and Figure 2c). rCTLA-4 binds to the antiCTLA4mAb in the growth media, freeing CTLA-4 on the cell surface to interact with rB7 and repress the T-cell activation seen as a decrease in IL-2 production. rCTLA-4/anti-CTLA4mAb interaction was observed with supernatant from cells infected with Ad5/3-D24aCTLA4 and Ad5/3-aCTLA4. The highest loss of function was observed with supernatant from Ad5/3-D24aCTLA4-infected cells, even higher than the positive control. Taken together, our data indicate that infection of cancer cells with Ad5/3-D24aCTLA4 leads to high expression of a functional human mAb against human CTLA-4, which results in induction of T-cell activity as measured by IL-2. However, it is difficult to directly compare the effects of virus-produced aCTLA4 to purified and concentrated recombinant aCTLA4 because other factors are likely to be present in the supernatant from virus-infected cells. For example, the data shown in Figure 2 suggest that supernatant from virus-infected tumor cells may contain immune-suppressive proteins as IL-2 levels were lower than with mock. Tumor cell lines and low-passage tumor explants express high levels of CTLA-4 As it has been reported that almost 90% of the tumor cell lines express CTLA-4, and that anti-CTLA4mAb might have direct antitumor activity,13 we sought to investigate whether that was true also for the cells used in this study. The A549, SKOV3-ip1 and PC3-MM2 tumor cell lines presented 99.5%, 96.6% and 96.6% CTLA-4 expression, respectively, whereas UT-SCC8 low-passage tumor explants presented a 90.3% CTLA-4 expression (Figure 2d). Analysis of the replication and cell-killing ability of a newly constructed oncolytic adenovirus expressing anti-CTLA4mAb The anti-CTLA4 expression cassette was cloned to replace E3 gp19k, leaving the E3 promoter intact. The entire inserted sequence measures 2769 bp and hence the overall adenovirus genome size gain is about 1800 bp when the 965-bp E3 gp19k deletion and 24-bp E1 deletions are taken into account.29 This is & 2011 Macmillan Publishers Limited

Oncolytic adenovirus expressing anti-CTLA4mAb JD Dias et al

3

Figure 2. Evaluation of virus-produced anti-CTLA4mAb expression and function. (a) Serum-free supernatants of virus-infected A549 (lanes 1 and 2 from left to right) or PC3-MM2 (lanes 3 and 4) cells were analyzed by western blotting specific for human IgG (both heavy and light chains) under native (upper row) or denaturing conditions (lower row). Lanes 1 and 5: Cells infected with replication-competent Ad5/3-D24aCTLA4. Lanes 2 and 6: Cells infected with replication-deficient Ad5/3-aCTLA4. Lanes 3 and 7 are oncolytic controls, whereas lanes 4 and 8 are E1-deleted controls. (b) Functional assay to demonstrate activity of virus-produced anti-CTLA4mAb. Jurkat cells were incubated with ionomycin, PMA and recombinant human B7, resulting in CD28- and CTLA-4-positive cells resembling T cells stimulated by antigenpresenting cells, but inhibited by suppressor cells, a situation often found in advanced tumors. Supernatant from Ad5/3-D24aCTLA4-infected cells resulted in significant inhibition of B7/CTLA-4-mediated suppression in comparison with infection by replication-competent Ad5/3-D24. This is seen as increased production of IL-2, a T-cell activation marker. Recombinant anti-CTLA4mAb (aCTLA4) was included as a positive control. Note that recombinant aCTLA4mAb has been subjected to purification and concentration, whereas the unconcentrated, unpurified supernatant from infected cells was used in the other groups. (c) Loss-of-function assay to demonstrate the affinity of rCTLA-4 to virusproduced anti-CTLA4mAb. Jurkat cells were activated as before but now rCTLA-4 also was added to bind to B7, preventing CTLA-4 activation. When supernatant containing aCTLA4 was added, rCTLA-4 was blocked and B7 was able to bind CTLA-4 for reduced IL-2 production. Columns: Mean of three independent experiments. Bars: S.e. *Po0.05; ***Po0.001. (d) CTLA4 expression levels in A549, SKOV3.ip1 and PC3-MM2 tumor cell lines, and in the UT-SCC8 low-passage tumor explant. All tumor cell lines were positive for CTLA4 in fluorescence-assisted cell sorting. The solid line represents isotype control.

just below the proposed 105% threshold of compatibility with adenovirus packaging and functionality.30 To study whether the increase in genome size or expression of anti-CTLA4 affects viral replication, we infected A549 cells with phosphate-buffered saline (PBS), Ad5/3-D24 and Ad5/3-D24aCTLA4, & 2011 Macmillan Publishers Limited

and measured viral genomes as a function of time (Figures 3a and b). No significant differences were seen between the virus groups suggesting intact replicativity. We evaluated the oncolytic efficacy of Ad5/3-D24aCTLA4 on different tumor cell lines and on a head and neck squamous cell Gene Therapy (2011), 1 - 11

Oncolytic adenovirus expressing anti-CTLA4mAb JD Dias et al

4

Figure 3. Analysis of the replication and cell-killing efficiency of anti-CTLA4-armed adenoviruses and control vectors. The anti-CTLA4mAb expression cassette was cloned to replace the E3 gp19k open reading frame, leaving the E3 promoter intact. To study whether the transgene affects viral replication, A549 cells were infected with PBS, Ad5/3-D24 and Ad5/3-D24aCTLA4 (a, b). Quantitative PCR was performed at different time points using the supernatant and the cells. No significant difference was found between virus groups. For evaluation of oncolytic potency, tumor cell lines (c) PC3-MM2, (d) SKOV3.ip1 and (e) A549, and (f ) the low-passage tumor explant UT-SCC8, were infected. Ad5/3-D24aCTLA4 had oncolytic potency similar to the positive control virus Ad5/3-D24 in all cell lines. Also replication-deficient Ad5/3-aCTLA4 had antitumor activity because tumor cells express CTLA4. MTS assay was read as follows: UT-SCC8, 7 days; PC3-MM2, 6 days; A549, 6 days; SKOV3-ip1, 7 days. Columns: Mean of triplicate assays. Bars: S.e. **Po0.01; ***Po0.001.

carcinoma (HNSCC) low-passage tumor cell culture (Figures 3c -- f). Ad5/3-D24aCTLA4 infection resulted in 97.6, 78.2, 69.1 and 57.3% killing of PC3-MM2, SKOV3-ip1, A549 and UT-SCC8 cells, respectively. Ad5/3-D24aCTLA4 and its counterpart with no transgene (Ad5/3-D24) showed no statistical difference in cytotoxicity in any of the analyzed tumor cell lines, confirming that the oncolytic potency of Ad5/3-D24aCTLA4 is retained despite the transgene. With the non-replicating Ad5/3-aCTLA4, no cytotoxicity was observed with lower virus doses (Figure 3). However, cytotoxicity was observed with the highest doses resulting in maximum cell killing rates of 96.2%, 74.3%, 49.1% and 56.2% on PC3-MM2, SKOV3-ip1, A549 and UT-SCC8, respectively (Figure 3). This result is in accord with previous demonstration of direct binding of Gene Therapy (2011), 1 -- 11

anti-CTLA4mAb with CTLA-4 expressed on the surface of cancer cells for induction of cell death.31 Ad5/3-D24aCTLA4 has antitumor activity in human cancer xenograft models Human anti-CTLA4mAb does not bind to mouse CTLA-432 and xenograft experiments require T-cell-deficient nude mice. Thus, these models only assay oncolytic and pro-apoptotic effects. Nevertheless, two subcutaneous xenograft models were tested. In both models Ad5/3-D24aCTLA4 showed a significant antitumor effect (Po0.01 by t-test) compared with mock (Figures 4a and d). No statistical difference was observed between Ad5/3-D24aCTLA4 and Ad5/3-D24. & 2011 Macmillan Publishers Limited

Oncolytic adenovirus expressing anti-CTLA4mAb JD Dias et al

5

Figure 4. Tumor growth suppression and apoptosis in immune-deficient mice using human cancer xenografts treated with an antiCTLA4mAb-expressing oncolytic adenovirus. PC3-MM2 (a -- c) and A549 (d) tumors (n ¼ 8 per group) were established in nude mice, where human anti-CTLA-4 does not have immunological activity. Therefore, the model measures only oncolysis and the pro-apoptotic effect of anti-CTLA-4. After 7 days, tumors (5 -- 8 mm in diameter) were treated intra-tumorally with 1  108 VPs on days 0, 2 and 4 (arrows). Mock mice were injected with growth media only. (a) Ad5/3-D24aCTLA4 caused significant antitumor activity compared with mock in this highly aggressive prostate cancer model. Points: mean. Bars: s.e. **Po0.01 (Student’s t-test on day 7). (b) On day 5, tumor cryosections were stained for expression of anti-CTLA4 (human IgG, upper row) or apoptosis (active caspase-3, lower row) by immunohistochemistry (brown). Images were taken at  40 magnification. Mock indicates staining of uninfected tumors, whereas the isotype control groups lack a primary Ab. (c) On day 7, human IgG levels in tumors and plasma of mice treated with Ad5/3-D24aCTLA4 or Ad5/3-aCTLA4 were measured. We found 81-fold more anti-CTLA4mAb in Ad5/3-D24aCTLA4-treated tumors in comparison with Ad5/3-aCTLA4-treated tumors (Po0.05). In addition, we found 43.3-fold more anti-CTLA4mAb in tumors treated with Ad5/3-D24aCTLA4 than in the plasma of the same animals (Po0.05). The average plasma concentration was 392.6 mg g1 (SE 312.0), which is below concentrations reported tolerated in humans treated with ipilimumab (561 mg ml1)33 and tremelimumab (450 mg ml1)34,35, respectively. In Ad5/3-D24aCTLA4-treated tumors, mAb concentration was 16 977 mg g1 and therefore much higher than in plasma. Columns: Mean of triplicate assays. Bars: S.e. (d) Ad5/3-D24aCTLA4 caused significant antitumor activity compared with mock (Po0.01) in this lung cancer model. Significant difference between Ad5/3-D24aCTLA4 and Ad5/3-D24 was not seen. Groups are shown until the first mouse from the group had to be killed per animal regulations. Points: Mean. Bars: S.e.

Ad5/3-D24aCTLA4-treated tumors express anti-CTLA4mAb resulting in enhanced apoptosis Given that CTLA-4-blocking Abs may induce direct killing of tumor cells by triggering apoptosis,7,13 we assessed whether antitumor efficacy was associated with human anti-CTLA4mAb production & 2011 Macmillan Publishers Limited

and subsequently increased apoptosis. We observed human mAb staining in tumors treated with Ad5/3-D24aCTLA4 and Ad5/3-aCTLA4, but not with Ad5/3-D24 or Ad5/3-Luc1 (Figure 4b). Expression of human mAb seemed to correlate with increased apoptosis (Figure 4b). Gene Therapy (2011), 1 - 11

Oncolytic adenovirus expressing anti-CTLA4mAb JD Dias et al

6

Local versus systemic levels of anti-CTLA4mAb in Ad5/3-D24aCTLA4-treated mice On day 7, we measured human IgG levels in tumors and plasma of mice treated with Ad5/3-D24aCTLA4 or Ad5/3-aCTLA4. In Ad5/3-D24aCTLA4-treated tumors, anti-CTLA4mAb concentration was 16 977 mg g1, which is 81-fold higher than that seen in Ad5/3-aCTLA4-treated tumors (Po0.05; Figure 4c). In addition, we found 43.3-fold more anti-CTLA4mAb in tumors treated with Ad5/3-D24aCTLA4 than in the plasma of the same animals (Po0.05; Figure 4c). The average plasma concentration was 392.6 mg g1 (SE 312.0), which is below concentrations reported tolerated in humans treated with ipilimumab (561 mg ml1)33 and tremelimumab (450 mg ml1),34,35 respectively (1 ml of plasma weighs circa 1 g). Therefore, viral delivery of anti-CTLA4mAb led to increased tumor concentrations without increase in systemic levels. Effect of aCTLA4 on antiviral immunity Immunocompetent mice were treated three times with PBS, Ad5/3-D24, Ad5/3-D24 and mouse aCTLA4 Ab, or with mouse aCTLA4 Ab only. After 2 weeks we collected the spleens and performed IFN-g ELISPOT on splenocytes. Peripheral blood mononuclear cells (PBMCs) were stimulated with UV-inactivated Ad5/3 chimeric virus or with an Ad5 peptide pool. No significant differences were observed suggesting that mouse aCTLA4 Ab does not affect antiviral T-cell responses resulting from injection of human oncolytic adenovirus into mice (Supplementary Figure 5). Although the combination group in this experiment mimics oncolytic virus expressing aCTLA4, it should be kept in mind that human adenovirus does not replicate productively in mice and therefore the experiment may not be fully representative of T-cell responses caused by adenovirus in humans. Immunomodulation of cancer patient T-cells by anti-CTLA4mAbexpressing viruses To extend our preclinical findings into humans, we studied PBMCs of patients with advanced solid tumors refractory to chemotherapy. In all four patients, supernatant from anti-CTLA4mAbexpressing viruses was able to increase T-cell activity, as measured by IL-2 and IFN-g (Figure 5). The rationale for the experiment is shown in Supplementary Figures 2a -- d. Interestingly, whereas with Jurkat cells, which is an immortalized T-cell line, the recombinant mAb was more effective, with patient samples the viral supernatants were often more potent. Similar data were obtained in a loss-of-function assay (Supplementary Figure 3; rationale in Supplementary Figures 2e and f). Some variation of the effect was seen between patients. We assume that patient PBMC samples feature different degrees of immunosuppression because of differences in the tumors the patients have. Some tumors could be more immune-suppressive than others. Also, many immune-suppressive mechanisms (some related to CTLA4 and some not) may be having a role concurrently and thus the effects of aCTLA4 vary between individuals. Effect of anti-CTLA4mAb on PBMCs from healthy individuals In contrast to cancer patient PBMCs, supernatant from Ad5/3-D24aCTLA4-infected cells did not seem to significantly increase IL-2 or IFN-g production in PBMCs from healthy individuals (Figure 6), although further samples should be analyzed to confirm this. Significant changes were not seen in the loss-offunction assay either (Supplementary Figure 4). However, as the positive control recombinant mAb was effective in both cases (it increased IL-2 and IFN-g in Figure 6, and decreased them in Supplementary Figure 4), we assume this is a dose effect. This is supported by the non-significant trend seen for Ad5/3-D24aCTLA4 in the loss-of-function assay (Supplementary Figure 4) and several cases of significant effect by supernatant from cells infected by Gene Therapy (2011), 1 -- 11

the replication-defective control Ad5/3-aCTLA4 (Figure 6 and Supplementary Figure 4), which produces less anti-CTLA4mAb (Figure 4c). Nevertheless, the effect of anti-CTLA4mAb on T cells was more pronounced in cancer patients, perhaps because they have a higher degree of systemic immunosuppression owing to the advanced tumor present.36 DISCUSSION Cancer is a major public health problem in all parts of the world. Despite several improvements in diagnostics and treatment modalities, still over 7 million cancer-related deaths occur every year worldwide.37 With regard to immunotherapy of cancer, a key realization has been that induction of an antitumor immune response is not sufficient to eradicate the disease because of tumor immune evasion mechanisms. In particular, advanced tumors are often highly immune-suppressive.1,2 In this regard, one particularly promising approach is reduction of tumor immunosuppression by anti-CTLA4mAb, which recently resulted in a positive randomized phase-3 study in melanoma.6 Replication-deficient vectors have been used to reduce side effects of therapeutic agents by promoting gene expression localized to tumors.38 Oncolytic viruses are even more useful because their self-amplifying nature can be used for enhanced tumor transduction and amplification of effect.39 Also, we have demonstrated previously that oncolytic adenoviruses expressing immunomodulatory agents are able to induce tumor-specific immunity by triggering the innate and adaptive pathways of the immune system.40,41 Of note, the oncolytic platform may be useful for increasing ‘danger signals’, useful for immunity versus tolerance. One of the main mechanisms that tumors use to escape antitumor immunity is through T-Regs. Among several approaches that have been studied thus far to downregulate T-Regs, the antiCTLA4mAb is the only one whose efficacy has been proven in a large randomized study.6 Although the trial by Hodi et al. represents a breakthrough for tumor immunotherapy, it was preceded by a negative phase-3 study with a similar antiCTLA4mAb, suggesting that the approach would benefit from enhancement of efficacy.42 Also, in all anti-CTLA4mAb trials serious immune-related adverse events have caused severe morbidity, including mortality, resulting in concern over the safety of the approach.6 Therefore, utilization of a gene therapy platform is appealing because it might increase local concentration for increased efficacy while reducing side effects related to systemic exposure. This hypothesis was validated in a preclinical study where local secretion of mouse anti-CTLA4 enhanced therapeutic efficacy with reduced systemic autoimmunity.43 Therefore, we armed an oncolytic adenovirus with fully human mAb specific for CTLA-4. In this approach, tumor cells are killed owing to viral replication, antitumor immune activation and direct pro-apoptotic tumor cell killing. Additional benefit may result from tumor antigen release owing to viral replication, which can improve the efficacy of antiCTLA4mAb therapy by potentially allowing a more specific immune response against tumor targets.6,44,45 Also, the side effects of oncolysis and anti-CTLA4mAb therapy are non-overlapping, which might facilitate increased efficacy without increasing toxicity. In this study, the human IgG2 subtype was chosen to ensure that CTLA-4-blocking Abs would not kill activated T cells by fixing complement or triggering Fc receptor-mediated, Ab-dependent, cell-mediated cytotoxicity. IgG2 induces minimal complement activation and Ab-dependent, cell-mediated cytotoxicity,46 and decreases the possibility of cytokine release syndrome in humans.7 We showed that oncolytic adenoviruses can effectively express functional anti-CTLA4mAb as a transgene in cancer cell lines & 2011 Macmillan Publishers Limited

Oncolytic adenovirus expressing anti-CTLA4mAb JD Dias et al

7

Figure 5. Increased IL-2 and IFN-g production in stimulated PBMCs of cancer patients mediated by Ad5/3-D24aCTLA4. PBMCs from four patients with advanced chemotherapy-refractory cancer were stimulated with ionomycin, PMA and recombinant human B7 to mimic tumorinduced immune suppression; and treated with filtered supernatants from virus-infected PC3-MM2 cells. IL-2 (a) and IFN-g (b) are markers of T-cell activation, and were measured by a fluorescence-activated cell sorting (FACS) array. Recombinant anti-CTLA4mAb (aCTLA4) was included as a positive control. Columns: Mean of three independent experiments. Bars: S.e. *Po0.05; **Po0.01; ***Po0.001.

(Figure 2). Also, we found that it was possible to combine oncolytic adenovirus replication with anti-CTLA4mAb and retain the oncolytic potency of the virus (Figures 3 and 4). This has been of concern, as CTLA-4 blockade with mAbs results in increased production of IFN-g and major histocompatibility complex class-I molecules that potentially can inhibit viral replication.47,48 However, this may not be a problem in tumor cells, which are often deficient in IFN signaling and major histocompatibility complex class-I presentation.24 Importantly, we saw much higher local concentration of mAb in tumors in comparison with blood. Also, the replication-competent platform was superior to a replicationdeficient virus. With regard to safety, it may be promising that & 2011 Macmillan Publishers Limited

plasma mAb was lower than what has been tolerated in humans trials.33,34,49,50 Emphasizing the utility of local versus systemic CTLA4 blockage, it has been reported that combination of tumor-localized antiCTLA4 scFv expression with systemic T-Reg depletion has a synergistic effect.51 Further, the authors did not observe autoimmune reactions in a murine model when anti-CTLA-4 scFv Ab was expressed by the tumors and anti-CD25 Abs were administered intraperitoneally.51 By contrast, when antiCTLA4mAb was administered systemically together with T-Reg depletion autoimmune reactions were observed.52,53 One possible reason for the synergy between these approaches is that Gene Therapy (2011), 1 - 11

Oncolytic adenovirus expressing anti-CTLA4mAb JD Dias et al

8

Figure 6. IL-2 and IFN-g production in PBMCs of healthy donors by Ad5/3-D24aCTLA4. PBMCs from two healthy donors were stimulated with ionomycin, PMA and recombinant human B7; and treated with filtered supernatants from virus-infected PC3-MM2 cells. The IL-2 (a) or IFN-g (b) levels in the growth media were measured by fluorescence-activated cell sorting (FACS) array. Columns: mean of three independent experiments. Bars: S.e. *Po0.05; ***Po0.001.

depletion of T-Regs reduces the number of regulatory cells, whereas anti-CTLA4 reduces the activity of regulatory cells. Moreover, antiCTLA4 can also reduce the suppression of antigen-presenting cells in addition to T-Regs. We have previously reported treatment of cancer patients with oncolytic viruses expressing granulocyte -- macrophage colonystimulating factor (GM-CSF) in combination with low-dose metronomic cyclophosphamide, which was used to reduce T-Regs.41 In addition, cell-mediated delivery of mouse anti-CTLA4 from GM-CSF-secreting tumor cells activated potent antitumor responses and prolonged overall survival, with reduced evidence of systemic autoimmunity.43 Therefore, the efficacy of the approach could perhaps eventually be further improved by combining with GM-CSF and/or reduction of the number of T-Regs. Also, multimodal approaches featuring radiotherapy or chemotherapy might be appealing because of synergism without overlapping side effects.18 To our knowledge, this is the first study showing that a fulllength mAb can be produced from an oncolytic adenovirus. Also, it is the first fully human anti-CTLA4mAb expressed by a tumortargeted, replicative-competent platform. Although there were no side effects seen in the mouse experiment that was performed, this requires further studies as human anti-CTLA4 would not be predicted to affect mouse tissues.32 Studying the safety of the approach in the laboratory is challenging as human mAbs are not active in non-human species and most animals, including mice, are non-permissive for human adenoviral oncolysis. Using mouse anti-CTLA4mAbs in mouse tumor models might yield unreliable results if mouse CTLA4 biology is distinct from human CTLA4 biology. Likewise, the full therapeutic benefit of a human oncolytic adenovirus coding for human anti-CTLA4mAb can only be realized in human cancer patients owing to species incompatibilities with regard to both oncolysis and CTLA4. Our data from cancer patient PBMCs provide a rationale for clinical translation of Ad5/3-D24aCTLA4 for treating patients with advanced cancer. A particularly appealing target might be melanoma, where activity of anti-CTLA4mAbs has now been convincingly established. However, as the p16 -- Rb pathway is defective in many if not all solid tumors,24 Ad5/3-D24aCTLA4 is an attractive candidate agent for treatment of many other types of cancer refractory to available treatments. Future studies are Gene Therapy (2011), 1 -- 11

needed to optimize these results, especially in a multimodal context, and to obtain more safety data.

MATERIALS AND METHODS Cell culture and patients The human HNSCC low-passage tumor cell culture UT-SCC8 (supraglottic larynx) 54 was cultured in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal calf serum (PromoCell GmbH, Heidelberg, Germany), 1% non-essential amino acids (Gibco, Invitrogen, Carlsbad, CA, USA) 2 mmol l1 glutamine, 100 U ml1 penicillin and 100 U ml1 streptomycin (all from Sigma, St Louis, MO, USA). The UT-SCC cells were used in low passage, typically passage 15 - 30. Human transformed embryonic kidney cell line 293, human lung cancer cell line A549, human prostate cancer cell line PC-3MM2 and Jurkat (clone E6 -- 1) human leukemic T-cell lymphoblast cells were obtained from the American Type Culture Collection (ATCC, Manassas, VA, USA). The human ovarian cancer cell line SKOV3.ip1 was obtained from obtained from Janet Price (MD Anderson Cancer Center, Houston, TX, USA). All cell lines were maintained under recommended conditions. PBMCs from healthy individuals and patients with advanced metastatic tumors refractory to conventional therapies were obtained with informed consent.

Construction of adenoviruses To create the oncolytic Ad5/3-D24aCTLA4, we constructed pTHSN-aCTLA4, which contains the heavy and light chains of IgG2-type anti-CTLA4mAb in the E3 region of the adenoviral genome deleted for 6.7K/gp19K. The equimolarity of the heavy and light chains is achieved with an internal ribosome entry site (IRES) between the chains. pAdEasy-1.5/3-D24-aCTLA4 was generated by homologous recombination in Escherichia coli BJ5183 cells (Qbiogene Inc., Irvine, CA, USA) between FspI-linearized pTHSNaCTLA4 and SrfI-linearized pAdEasy-1.5/3-D24,29 a rescue plasmid containing the serotype 3 knob and a 24-bp deletion in E1A. The genome of Ad5/3D24aCTLA4 was released by PacI digestion and subsequent transfection of A549 cells. The virus was propagated on A549 cells and purified on cesium chloride gradients. The viral particle concentration was determined at 260 nm, and standard TCID50 on 293 cells was performed to determine infectious particle titer. & 2011 Macmillan Publishers Limited

Oncolytic adenovirus expressing anti-CTLA4mAb JD Dias et al

9 To construct the E1-deleted control virus Ad5/3-aCTLA4, both chains of the anti-CTLA4mAb cDNA were ligated into pShuttle-CMV. Homologous recombination was performed between pAdEasy-1.5/3 and PmeI-linearized pShuttle-CMV-aCTLA4 to construct pAdEasy-1.5/3- aCTLA4. The genome of Ad5/3-aCTLA4 was released by PacI and transfected into 293 cells. The virus was propagated on 293 cells and purified on cesium chloride gradients. The viral particle concentration was determined at 260 nm, and a standard plaque assay on 293 cells was performed to determine infectious particles. Viruses were propagated as reported and their main features, including those of the control viruses, Ad5Luc1(ref. 29) and Ad5/3-D24,26 are described in Figure 1.

Quantitative PCR analysis A549 cells were plated on six-well plates in serum-free Dulbecco’s modified Eagle’s medium and infected with 1000 virus particles (VPs) per cell. Supernatant and cells were harvested at the indicated time points. After lysing the cells, quantitative PCR was performed as described.40 The limit of detection and limit of quantification for the assay is 500 VPs ml1.

infected for 1 h at 37 1C, washed and incubated in 5% fetal calf serum. Cell viability was determined according to the manufacturer’s protocol (Cell Titer 96 Aqueous One Solution Cell Proliferation Assay; Promega, Nacka, Sweden).

Antitumor activity of Ad5/3-D24aCTLA4 in subcutaneous tumor models All animal experiments were approved by the Experimental Animal Committee of the University of Helsinki and the Provincial Government of Southern Finland. Human xenografts were established by injecting 5  106 PC3-MM2 or A549 cells into the flanks of 5- to 6-week-old female NMRI/ nude mice (Taconic, Ejby, Denmark). After 7 days, tumors (n ¼ 8 per group, 5 - 8 mm in diameter) were injected with a volume of 50 ml for three times every other day with 3  108 VPs (days 0, 2 and 4) and control tumors were injected with Dulbecco’s modified Eagle’s medium only. The formula (length  (width)2  0.5) was used to calculate tumor volume. Mice were killed when a tumor reached an average diameter of 15 mm.

Effect of murine aCTLA4 Ab on antiviral immunity

Western blot analysis A549 or PC3-MM2 tumor cells were infected with Ad5/3-D24aCTLA4 or Ad5/3-aCTLA4 at 10 VPs per cell. After 48 h, the supernatants of virusinfected cells were filtrated using 0.02-mm filters (Whatman Anotop, London, UK); 15 ml were run on a 7.5% sodium dodecyl sulfate-PAGE gel under reducing or native conditions, and transferred onto a nitrocellulose membrane. The membrane was incubated with goat anti-human IgG (heavy and light chains) (AbD Serotec, MorphoSys, Du¨sseldorf, Germany), washed and incubated with a secondary Ab coupled to horseradish peroxidase (Dako, Glostrup, Denmark). Signal detection was performed by enhanced chemiluminescence (GE Healthcare, Amersham, UK).

C57Bl/6 immunocompetent mice (n ¼ 5) were treated three times every third day with PBS, Ad5/3-D24, Ad5/3-D24 and mouse aCTLA4 Ab (Bio X Cell catalog no. BE0164; West Lebanon, NH, USA) or with a mouse aCTLA4 Ab only. The virus (3  10e8 VPs) was administered subcutaneously and the mouse aCTLA4 Ab (0.2 mg) was administered intraperitoneally in 0.1 ml of PBS. If both virus and Ab were administered, the Ab was administered first followed by administration of the virus 15 min later. After 2 weeks we collected the spleens of the mice and performed an ELISPOT assay (mouse INF-gamma ELISpot, Mabtech, Nacka, Sweden) on the splenocytes. PBMCs were stimulated by UV-inactivated Ad5/3-Luc1 or by an anti-Ad5 peptide pool.

Human IgG and apoptosis immunostaining

Anti-CTLA4 biological activity and FACS array 1

PBMCs or Jurkat cells were stimulated with 0.3 mg ml l ionomycin (SigmaAldrich Co., St Louis, MO, USA), 0.03 mg ml1 phorbol myristate acetate (PMA) (Sigma-Aldrich Co.) and 1 mg ml1 recombinant human B7 Fc Chimera (R&D Systems, Abingdon, England, UK), and treated with 0.02-mm filtered (Whatman Anotop) supernatants of virus-infected PC3-MM2 cells. In the loss-of-function assay, 0.1 mg ml1 recombinant human CTLA-4/Fc chimera (R&D Systems) was added to ionomycin, PMA and recombinant B7. Twenty-four hours after PBMC stimulation or 48 h after Jurkat cell stimulation, the IL-2 or IFN-g levels in the growth media were analyzed by using the BD Cytometric Bead Array Human Soluble Protein Flex Set (Becton Dickinson, Helsinki, Finland) according to the manufacturer’s instructions. Ten VPs per cell were used and supernatants were collected 48 h later. Mouse anti-human CTLA-4 ( ¼ CD152) mAb (BD Pharmingen Europe, Helsinki, Finland) was used as positive control at a dilution of 1:10000 as recommended by the vendor. The FCAP Array v.1.0.2 (Soft Flow, Pe´cs, Hungary) software was used for analysis.

Immunofluorescence and flow cytometry Indirect immunofluorescence was performed in the low the low-passage tumor cell culture UT-SCC8 or the tumor cell lines A549, SKOV3-ip1 and PC3-MM2 for analyzing surface CTLA-4. Briefly, cell pellet was incubated for 30 min at 4 1C with mouse anti-human CTLA-4mAb (BD Pharmingen Europe) as the primary Ab followed by incubation for a further 30 min at 4 1C with an Alexa Fluor-488 donkey anti-mouse IgG (Invitrogen) as the secondary Ab. The fluorescence intensity was measured on an LSR flow cytometer (BD Pharmingen Europe). At least 40 000 cells per sample were counted. A Clontech Discovery Labware immunocytometry system (BD Pharmingen Europe) and the FlowJo 7.6.1 software were used for analysis.

Cryosections (4 -- 5 mm) of frozen tumors embedded in Tissue Tek OCT (Sakura, Torrance, CA, USA) were prepared and fixed in acetone for 10 min at 20 1C. As primary Ab we used a goat anti-human IgG (heavy and light chains) (AbD Serotec, MorphoSys) or a rabbit mAb against active caspase-3 at a dilution of 1:200 for 1 h at room temperature (BD Pharmingen, AB559565). Further, sections were incubated according to manufacturer’s instructions using the LSAB2 System-HRP kit (K0673; DakoCytomation, Carpinteria, CA, USA). Bound Abs were visualized using 3,30 -diaminobenzidine (Sigma). Lastly, sections were counterstained with hematoxylin and dehydrated in ethanol, clarified in xylene and sealed with Canada balsam. Representative pictures were captured at  40 magnification using a Leica DM LB microscope equipped with an Olympus DP50 color camera.

mAb concentration in tumor and plasma samples Tissues were minced with a scalpel and 50 mg were incubated with 5 ml of protease inhibitor (P8340; Sigma-Aldrich) and 500 ml of a digestion mixture consisting of RPMI-1640 medium with 10 mmol l1 HEPES buffer and 1.6 mmol l1 phenylmethylsulfonyl fluoride (Sigma-Aldrich), 40 mg ml1 gentamycin (Amresco, Solon, OH, USA), 100 mg ml1 bovine serum albumin (Sigma-Aldrich) and 100 mg ml1 Zwittergent 3 - 12 (Merck4Biosciences, Darmstadt, Germany). After incubation for 90 min at 37 1C under continuous agitation, the digestates were subjected to 30 s of sonication and centrifuged at 2000 g for 10 min at 4 1C. Supernatants were collected and stored. Human IgG concentrations in tissue digestion supernatants and plasma samples were measured by ELISA for human IgG (Immunology Consultants Laboratory, Portland, OR, USA) according to the manufacturer’s instructions. A plasma density of 1 g ml1 was used for conversion between weight and volume.

In vitro cell killing HNSCC low-passage tumor cell culture or tumor cell lines were seeded at 1.5  104 or 1.0  104 cells per well on 96-well plates. On the next day, viruses were diluted in growth media with 2% fetal calf serum, cells were & 2011 Macmillan Publishers Limited

Statistical analysis Statistical analyses were performed by two-tailed Student’s t-test. For all analyses, a P-value of o0.05 was deemed statistically significant. Gene Therapy (2011), 1 - 11

Oncolytic adenovirus expressing anti-CTLA4mAb JD Dias et al

10

CONFLICT OF INTEREST Akseli Hemminki is cofounder and shareholder in Oncos Therapeutics Ltd.

ACKNOWLEDGEMENTS We thank Pa¨ivi Hannuksela, Sirkka-Liisa Holm, Eerika Karli and Aila Karioja-Kallio for technical assistance. Financial support: Financial support was received from Helsinki Biomedical Graduate School, European Research Council, Helsinki University Central Hospital, Sigrid Juselius Foundation, Academy of Finland, Emil Aaltonen Foundation, ASCO Foundation, Biocenter Finland, the Biomedicum Foundation, Finnish Cancer Society, Biocentrum Helsinki, the K Albin Johansson Foundation, University of Helsinki and the Finnish Cultural Foundation. Akseli Hemminki is K Albin Johansson Research Professor of the Foundation for the Finnish Cancer Institute.

REFERENCES 1 Dranoff G. Cytokines in cancer pathogenesis and cancer therapy. Nat Rev Cancer 2004; 4: 11 - 22. 2 de Visser KE, Eichten A, Coussens LM. Paradoxical roles of the immune system during cancer development. Nat Rev Cancer 2006; 6: 24 - 37. 3 Leach DR, Krummel MF, Allison JP. Enhancement of antitumor immunity by CTLA4 blockade. Science 1996; 271: 1734 - 1736. 4 Kwon ED, Foster BA, Hurwitz AA, Madias C, Allison JP, Greenberg NM et al. Elimination of residual metastatic prostate cancer after surgery and adjunctive cytotoxic T lymphocyte-associated antigen 4 (CTLA-4) blockade immunotherapy. Proc Natl Acad Sci USA 1999; 96: 15074 - 15079. 5 Kirkwood JM, Lorigan P, Hersey P, Hauschild A, Robert C, McDermott D et al. Phase II trial of tremelimumab (CP-675206) in patients with advanced refractory or relapsed melanoma. Clin Cancer Res 2010; 16: 1042 - 1048. 6 Hodi FS, O’Day SJ, McDermott DF, Weber RW, Sosman JA, Haanen JB et al. Improved survival with ipilimumab in patients with metastatic melanoma. N Engl J Med 2010; 363: 711 - 723. 7 Ribas A, Hanson DC, Noe DA, Millham R, Guyot DJ, Bernstein SH et al. Tremelimumab (CP-675206), a cytotoxic T lymphocyte associated antigen 4 blocking monoclonal antibody in clinical development for patients with cancer. Oncologist 2007; 12: 873 - 883. 8 Lee KM, Chuang E, Griffin M, Khattri R, Hong DK, Zhang W et al. Molecular basis of T cell inactivation by CTLA-4. Science 1998; 282: 2263 - 2266. 9 Paradis TJ, Floyd E, Burkwit J, Cole SH, Brunson B, Elliott E et al. The antitumor activity of anti-CTLA-4 is mediated through its induction of IFN gamma. Cancer Immunol Immunother 2001; 50: 125 - 133. 10 Chambers CA, Kuhns MS, Egen JG, Allison JP. CTLA-4-mediated inhibition in regulation of T cell responses: mechanisms and manipulation in tumor immunotherapy. Annu Rev Immunol 2001; 19: 565 - 594. 11 Paust S, Lu L, McCarty N, Cantor H. Engagement of B7 on effector T cells by regulatory T cells prevents autoimmune disease. Proc Natl Acad Sci USA 2004; 101: 10398 - 10403. 12 Munn DH, Sharma MD, Hou D, Baban B, Lee JR, Antonia SJ et al. Expression of indoleamine 2,3-dioxygenase by plasmacytoid dendritic cells in tumor-draining lymph nodes. J Clin Invest 2004; 114: 280 - 290. 13 Contardi E, Palmisano GL, Tazzari PL, Martelli AM, Fala F, Fabbi M et al. CTLA-4 is constitutively expressed on tumor cells and can trigger apoptosis upon ligand interaction. Int J Cancer 2005; 117: 538 - 550. 14 Jinushi M, Hodi FS, Dranoff G. Therapy-induced antibodies to MHC class I chainrelated protein A antagonize immune suppression and stimulate antitumor cytotoxicity. Proc Natl Acad Sci USA 2006; 103: 9190 - 9195. 15 Sanderson K, Scotland R, Lee P, Liu D, Groshen S, Snively J et al. Autoimmunity in a phase I trial of a fully human anti-cytotoxic T-lymphocyte antigen-4 monoclonal antibody with multiple melanoma peptides and Montanide ISA 51 for patients with resected stages III and IV melanoma. J Clin Oncol 2005; 23: 741 - 750. 16 Russell WC. Update on adenovirus and its vectors. J Gen Virol 2000; 81: 2573 - 2604. 17 Kanerva A, Hemminki A. Modified adenoviruses for cancer gene therapy. Int J Cancer 2004; 110: 475 - 480. 18 Dias JD, Guse K, Nokisalmi P, Eriksson M, Chen DT, Diaconu I et al. Multimodal approach using oncolytic adenovirus, cetuximab, chemotherapy and radiotherapy in HNSCC low passage tumour cell cultures. Eur J Cancer 2009; 46: 625 - 635. 19 Rajecki M, Kanerva A, Stenman UH, Tenhunen M, Kangasniemi L, Sarkioja M et al. Treatment of prostate cancer with Ad5/3Delta24hCG allows non-invasive detection of the magnitude and persistence of virus replication in vivo. Mol Cancer Ther 2007; 6: 742 - 751.

Gene Therapy (2011), 1 -- 11

20 Kanerva A, Wang M, Bauerschmitz GJ, Lam JT, Desmond RA, Bhoola SM et al. Gene transfer to ovarian cancer versus normal tissues with fiber-modified adenoviruses. Mol Ther 2002; 5: 695 - 704. 21 Sa¨rkioja M, Kanerva A, Salo J, Kangasniemi L, Eriksson M, Raki M et al. Noninvasive imaging for evaluation of the systemic delivery of capsid-modified adenoviruses in an orthotopic model of advanced lung cancer. Cancer 2006; 107: 1578 - 1588. 22 Fueyo J, Gomez-Manzano C, Alemany R, Lee PS, McDonnell TJ, Mitlianga P et al. A mutant oncolytic adenovirus targeting the Rb pathway produces anti-glioma effect in vivo. Oncogene 2000; 19: 2 - 12. 23 Heise C, Hermiston T, Johnson L, Brooks G, Sampson-Johannes A, Williams A et al. An adenovirus E1A mutant that demonstrates potent and selective systemic antitumoral efficacy. Nat Med 2000; 6: 1134 - 1139. 24 Sherr CJ. Cancer cell cycles. Science 1996; 274: 1672 - 1677. 25 Bauerschmitz GJ, Guse K, Kanerva A, Menzel A, Herrmann I, Desmond RA et al. Triple-targeted oncolytic adenoviruses featuring the cox2 promoter, E1A transcomplementation, and serotype chimerism for enhanced selectivity for ovarian cancer cells. Mol Ther 2006; 14: 164 - 174. 26 Kanerva A, Zinn KR, Chaudhuri TR, Lam JT, Suzuki K, Uil TG et al. Enhanced therapeutic efficacy for ovarian cancer with a serotype 3 receptor-targeted oncolytic adenovirus. Mol Ther 2003; 8: 449 - 458. 27 Greenwald RJ, Boussiotis VA, Lorsbach RB, Abbas AK, Sharpe AH. CTLA-4 regulates induction of anergy in vivo. Immunity 2001; 14: 145 - 155. 28 Krummel MF, Allison JP. CTLA-4 engagement inhibits IL-2 accumulation and cell cycle progression upon activation of resting T cells. J Exp Med 1996; 183: 2533 - 2540. 29 Kanerva A, Zinn KR, Peng KW, Ranki T, Kangasniemi L, Chaudhuri TR et al. Noninvasive dual modality in vivo monitoring of the persistence and potency of a tumor targeted conditionally replicating adenovirus. Gene therapy 2005; 12: 87 - 94. 30 Kennedy MA, Parks RJ. Adenovirus virion stability and the viral genome: size matters. Mol Ther 2009; 17: 1664 - 1666. 31 Ribas A. Anti-CTLA4 antibody clinical trials in melanoma. Update Cancer Ther 2007; 2: 133 - 139. 32 Hanson DC, Canniff PC, Primiano MJ et al. Preclinical in vitro characterization of anti-CTLA4 therapeutic antibody CP-675206. Proc Am Assoc Cancer Res 2004; 45: abstract #3802. 33 Weber JS, O’Day S, Urba W, Powderly J, Nichol G, Yellin M et al. Phase I/II study of ipilimumab for patients with metastatic melanoma. J Clin Oncol 2008; 26: 5950 - 5956. 34 Ribas A, Camacho LH, Lopez-Berestein G, Pavlov D, Bulanhagui CA, Millham R et al. Antitumor activity in melanoma and anti-self responses in a phase I trial with the anti-cytotoxic T lymphocyte-associated antigen 4 monoclonal antibody CP-675206. J Clin Oncol 2005; 23: 8968 - 8977. 35 Tarhini AA, Iqbal F. CTLA-4 blockade: therapeutic potential in cancer treatments. Onco Targets Ther 2010; 3: 15 - 25. 36 Stagg J, Johnstone RW, Smyth MJ. From cancer immunosurveillance to cancer immunotherapy. Immunol Rev 2007; 220: 82 - 101. 37 Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, 2002. CA Cancer J Clin 2005; 55: 74 - 108. 38 Harvey BG, Maroni J, O’Donoghue KA, Chu KW, Muscat JC, Pippo AL et al. Safety of local delivery of low- and intermediate-dose adenovirus gene transfer vectors to individuals with a spectrum of morbid conditions. Hum Gene Ther 2002; 13: 15 - 63. 39 Dias JD, Liikanen I, Guse K, Foloppe J, Sloniecka M, Diaconu I et al. Targeted chemotherapy for head and neck cancer with a chimeric oncolytic adenovirus coding for bifunctional suicide protein FCU1. Clin Cancer Res 2010; 16: 2540 - 2549. 40 Cerullo V, Pesonen S, Diaconu I, Escutenaire S, Arstila PT, Ugolini M et al. Oncolytic adenovirus coding for granulocyte macrophage colony-stimulating factor induces antitumoral immunity in cancer patients. Cancer Res 2010; 70: 4297 - 4309. 41 Koski A, Kangasniemi L, Escutenaire S, Pesonen S, Cerullo V, Diaconu I et al. Treatment of cancer patients with a serotype 5/3 chimeric oncolytic adenovirus expressing GMCSF. Mol Ther 2010; 18: 1874 - 1884. 42 Ribas A, Hauschild A, Kefford R et al. Phase III, open-label, randomized, comparative study of tremelimumab (CP-675206) and chemotherapy (temozolomide [TMZ] or dacarbazine [DTIC]) in patients with advanced melanoma. J Clin Oncol 2008; 26 (Suppl): abstract LBA9011. 43 Simmons AD, Moskalenko M, Creson J, Fang J, Yi S, VanRoey MJ et al. Local secretion of anti-CTLA-4 enhances the therapeutic efficacy of a cancer immunotherapy with reduced evidence of systemic autoimmunity. Cancer Immunol Immunother 2008; 57: 1263 - 1270. 44 Mokyr MB, Kalinichenko T, Gorelik L, Bluestone JA. Realization of the therapeutic potential of CTLA-4 blockade in low-dose chemotherapy-treated tumor-bearing mice. Cancer Res 1998; 58: 5301 - 5304. 45 Wolchok JD, Saenger Y. The mechanism of anti-CTLA-4 activity and the negative regulation of T-cell activation. Oncologist 2008; 13 (Suppl 4): 2 - 9.

& 2011 Macmillan Publishers Limited

Oncolytic adenovirus expressing anti-CTLA4mAb JD Dias et al

11 46 Bruggemann M, Williams GT, Bindon CI, Clark MR, Walker MR, Jefferis R et al. Comparison of the effector functions of human immunoglobulins using a matched set of chimeric antibodies. J Exp Med 1987; 166: 1351 - 1361. 47 McCart JA, Puhlmann M, Lee J, Hu Y, Libutti SK, Alexander HR et al. Complex interactions between the replicating oncolytic effect and the enzyme/prodrug effect of vaccinia-mediated tumor regression. Gene Ther 2000; 7: 1217 - 1223. 48 Nakamura H, Mullen JT, Chandrasekhar S, Pawlik TM, Yoon SS, Tanabe KK. Multimodality therapy with a replication-conditional herpes simplex virus 1 mutant that expresses yeast cytosine deaminase for intratumoral conversion of 5-fluorocytosine to 5-fluorouracil. Cancer Res 2001; 61: 5447 - 5452. 49 Bashey A, Medina B, Corringham S, Pasek M, Carrier E, Vrooman L et al. CTLA4 blockade with ipilimumab to treat relapse of malignancy after allogeneic hematopoietic cell transplantation. Blood 2009; 113: 1581 - 1588. 50 Vonderheide RH, LoRusso PM, Khalil M, Gartner EM, Khaira D, Soulieres D et al. Tremelimumab in combination with exemestane in patients with advanced breast cancer and treatment-associated modulation of inducible costimulator expression on patient T cells. Clin Cancer Res 2010; 16: 3485 - 3494.

51 Tuve S, Chen BM, Liu Y, Cheng TL, Toure P, Sow PS et al. Combination of tumor site-located CTL-associated antigen-4 blockade and systemic regulatory T-cell depletion induces tumor-destructive immune responses. Cancer Res 2007; 67: 5929 - 5939. 52 Sutmuller RP, van Duivenvoorde LM, van Elsas A, Schumacher TN, Wildenberg ME, Allison JP et al. Synergism of cytotoxic T lymphocyte-associated antigen 4 blockade and depletion of CD25(+) regulatory T cells in antitumor therapy reveals alternative pathways for suppression of autoreactive cytotoxic T lymphocyte responses. J Exp Med 2001; 194: 823 - 832. 53 Takahashi T, Tagami T, Yamazaki S, Uede T, Shimizu J, Sakaguchi N et al. Immunologic self-tolerance maintained by CD25(+)CD4(+) regulatory T cells constitutively expressing cytotoxic T lymphocyte-associated antigen 4. J Exp Med 2000; 192: 303 - 310. 54 Erjala K, Sundvall M, Junttila TT, Zhang N, Savisalo M, Mali P et al. Signaling via ErbB2 and ErbB3 associates with resistance and epidermal growth factor receptor (EGFR) amplification with sensitivity to EGFR inhibitor gefitinib in head and neck squamous cell carcinoma cells. Clin Cancer Res 2006; 12: 4103 - 4111.

Supplementary Information accompanies the paper on Gene Therapy website (http://www.nature.com/gt)

& 2011 Macmillan Publishers Limited

Gene Therapy (2011), 1 - 11