Final Report Summary - VIRAL&IMMUNE THERAPY (Breaking tolerance - combination of virotherapy and immunotherapy for cancer treatment) Aims The main aim of this project is to enhance the efficiency of a tumour-killing virus and induce an immune response against the primary tumour and metastatic deposits. Background Tumour-killing viruses called oncolytic adenoviruses are genetically programmed to selectively replicate within cancer cells, thereby killing them and spread to adjacent cancer cells. The genetic modifications make these viruses sense changes or mutations usually present in cancer cells and leave normal cells undamaged. In our research group, human adenovirus 5 (Ad5) that normally causes a common cold was genetically modified to sense changes in the regulation of the cell division that is typical of cancers. To achieve this, the key viral gene involved in virus replication (E1A) was placed under the control of the promoter E2F1 that is active only in highly replicative cells, such as cancer cells. In addition, the key virus protein E1A has a small deletion (Delta24) that makes it inactive in normal cells because it cannot bind and inhibit pRB (a tumour suppressor gene that regulates cell division). To increase the virus infectivity to cancer cells, a binding site of a cell surface protein called integrin (RGD) was inserted in the fibre (a protein found in the shell or capsid of the virus). These mechanisms make the tumour-killing virus ICOVIR15K cancer selective and highly potent. Proteins that are usually present in cancer cells but not in normal tissues are called tumour-associated proteins. The National Cancer Institute (NCI) has classified a number of tumour-associated antigens (proteins and other molecules that can trigger an immune response) by priority in order to accelerate their translational research. In this research project, we are aiming to display small parts of tumour-associated antigens (called epitopes) within the viral capsid in order to trigger a strong anti-tumoural response. Amongst those priority tumour-associated antigens, we have chosen Wilms tumour (WT)1, Survivin and GD2 as they are common in many solid cancer types, such as melanoma and neuroblastoma which is an infantile cancer that has already been subject of treatment with an oncolytic adenovirus created in our group. Survivin and WT1 vaccine therapies based in epitopes (small protein fragments or peptides) are currently being evaluated in the clinic and have showed positive immune (T cell) responses, but with limited efficacy. The anti-GD2 antibody (ch14.18) is being currently used in neuroblastoma trials as a passive immunotherapy showing improved 3-year overall survival (68.5 % with ch14.18 therapy versus 56.6 % with maintenance chemotherapy). However, the response rate of all the mentioned treatments could be augmented if small parts of WT1, Survivin and GD2 were presented in a more immunogenic environment, such as in the capsid of a tumour-killing virus, such as ICOVIR15K. Methods We have genetically modified the tumour-killing virus ICOVIR15K to display in its capsid ten tumour-associated epitopes that are usually present in tumour cells but not in normal cells; here we call this new technology viro-immunotherapy. We employed a two-step homologous recombination in bacteria. WT1 and Survivin epitopes are well characterised; however, as GD2 is a carbohydrate antigen, we are employing a mimotope which is a small protein fragment (10 aminoacid peptide) that mimics the structure of GD2 and is recognised by human ch14.18 and murine14G2a antibodies. These tumour epitopes and mimotope were inserted in different locations of the capsid of the tumour-killing virus. Successful epitope insertion can be measured by viral tumour-killing potency in vitro. Results We have created a number of plasmid libraries with tumour-associated epitopes inserted in hipervariable region 1 (HVR1), HVR5 and HVR7 of hexon viral capsid. This was achieved by two-step homologous recombination performed in bacteria containing the genome of ICOVIR15K. Ten different tumour-associated epitopes were inserted as single epitope or as pair combinations for each HVR. Five of them belonging to WT1 protein, four to Survivin protein and one to the GD2 mimotope. When inserted as two-epitopes (pair), then the total possible combinations of insertions are 82 for each HVR and only 10 possibilities when inserted as single epitopes. These protein sequences contain tumour-associated epitopes that bind the immune system presenting proteins called major histocompatibility complex (MHC) class I and /or II. The tumour-associated epitopes were designed as 'short' to bind only the MHC class I, or 'long' to bind both MHC class I and II. We find out that no viable virus was produced when tumour-associated epitopes were inserted in HVR7. Only two viruses displaying tumour-associated epitopes in the HVR1 were viable as single epitopes. Those viruses with a pair of tumour-associated epitopes in the HVR5 of the hexon (10 combinations in total) showed no loss in tumour-killing activity when the epitopes were shorter; however, those with longer epitopes showed reduced tumour-killing activity (created by directed insertion). It was possible to identify viruses that displayed epitopes from all three WT1, Survivin and GD2. We evaluated the binding of the anti-GD2 antibody (14G2a) to the viruses displaying the GD2 mimotope in the HVR1 and HVR5 of the hexon, but the antibody was not able to bind the purified virus perhaps indicating that the GD2 mimotope structure may have been modified when presented in the context of the HVRs of the viral hexon. We have evaluated the immune response against a long WT1 epitope displayed from the HVR5 in vivo. All animals responded against viral epitopes; however, only 30 % of animals responded against the targeted long WT1 epitope, thereby indicating the difficulty to instigate an immune response against a tumour antigen. We are currently evaluating other viro-immunotherapies to find out better candidates. The most potent tumour-killing viruses contained two short epitopes of WT1 protein or a combination of Survivin and WT1 shot epitopes, followed by longer epitope versions. Multiple epitope insertion in different HVRs did not form any viable virus. The most potent tumour-killing virus was called ICOVIR15K-WT1 (3/1). The WT1 (3/1) epitopes were also inserted in two other locations of the viral capsid such as the HI loop of the fibre protein and the protein IX (PIX) protein; however, viruses were only produced when these epitopes were inserted in the HI loop of the fibre. We are currently evaluating which display site of the viral capsid is more immunogenic in an immunocompetent murine model. Conclusions Genetic modifications in the adenoviral capsid are extremely hard to obtain because these modifications normally interfere with capsid structure and no viable adenoviral particles are formed. We have successfully been able to introduce tumour epitopes of WT1 and Survivin in the hexon. The majority of the insertions have been inserted in HVR5, followed by HVR1; whereas HVR7 could not generate any viral particle that contained tumour epitopes. Successful epitope insertion was also achieved in the HI loop of the fibre. Potential impact and use If these tumour-killing viruses are able to promote a specific immune response against the tumour-associated epitopes displayed in the viral capsid, then it is expected that this immune response could be translated into better anti-tumour efficacy. Given that our research group has already experience employing tumour-killing viruses in neuroblastoma patients, we envisage the clinical use of this viro-immunotherapy. Socio-economic impact of the project If this viro-immunotherapy treatment is able to show anti-tumoural response, we envisage that the technology should be patented, licensed and subject to the clinical testing. Consequently, cancer patients could benefit from the treatment. References: 1. Rojas, J. J., et al., Minimal RB-responsive E1A promoter modification to attain potency, selectivity, and transgene-arming capacity in oncolytic adenoviruses. Mol Ther, 2010. 18(11): p. 1960-71. 2. Alonso, M. M., et al., ICOVIR-5 shows E2F1 addiction and potent antiglioma effect in vivo. Cancer Res, 2007. 67(17): p. 8255-63. 3. Cheever, M. A., et al., The prioritisation of cancer antigens: a national cancer institute pilot project for the acceleration of translational research. Clin Cancer Res, 2009. 15(17): p. 5323-37. 4. Garcia-Castro, J., et al., Treatment of metastatic neuroblastoma with systemic oncolytic virotherapy delivered by autologous mesenchymal stem cells: an exploratory study. Cancer Gene Ther, 2010. 17(7): p. 476-83. 5. Ryan, B. M., N. O'Donovan, and M. J. Duffy, Survivin: a new target for anti-cancer therapy. Cancer Treat Rev, 2009. 35(7): p. 553-62. 6. Oka, Y. and H. Sugiyama, WT1 peptide vaccine, one of the most promising cancer vaccines: its present status and the future prospects. Immunotherapy, 2010. 2(5): p. 591-4. 7. Navid, F., M. Armstrong, and R.C. Barfield, Immune therapies for neuroblastoma. Cancer Biol Ther, 2009. 8(10): p. 874-82. 8. Fest, S., et al., Characterisation of GD2 peptide mimotope DNA vaccines effective against spontaneous neuroblastoma metastases. Cancer Res, 2006. 66(21): p. 10567-75. Contact Details: Miriam Bazan Peregrino, PhD (Oxon) Instituto de Investigación Biomédica de Bellvitge (IDIBELL) Hospital Duran i Reynals 3a planta - Gran Via de l'Hospitalet, 199 08908 L'Hospitalet de Llobregat Barcelona - Spain miriam.bazan-peregrino@balliol.oxon.org
