Oncolytic Vaccinia Virus

Abstract

The present invention relates to an oncolytic vaccinia virus and virus vectors for use in cancer therapy where in the virus comprises at least three vaccinia virus promoters which are positioned in the same orientation.

Claims

1-26. (canceled)

27. A TK-deficient vaccinia virus comprising an inactivated N1L gene, wherein the N1L gene is inactivated by the insertion of a single expression cassette, said expression cassette comprising a nucleic acid sequence, said nucleic acid sequence encoding a single heterologous polypeptide, and wherein the nucleic acid sequence comprises at least three vaccinia virus promoters positioned in the same orientation, wherein the orientation of the vaccinia virus promoters is opposite to that of the adjacent intact reading frame L024.

28. The TK-deficient vaccinia virus according to claim 27, wherein the polypeptide is a cytokine.

29. The TK-deficient vaccinia virus according to claim 27, wherein the polypeptide is selected from the group consisting of GM-CSF, IL-10, IL-12, and IL-21.

30. The TK-deficient vaccinia virus according to claim 27, wherein the vaccinia virus promoters are selected from the group consisting of modified H5, H5, P7.5, and PE/L.

31. A method of treating cancer in a subject, the method comprising administering to said subject the vaccinia virus according to claim 27.

32. A composition comprising a TK-deficient vaccinia virus, wherein the virus comprises an inactivated N1L gene, wherein the N1L gene is inactivated by the insertion of a single expression cassette, said expression cassette comprising a nucleic acid sequence, said nucleic acid sequence encoding a single heterologous polypeptide, and wherein the nucleic acid sequence comprises at least three vaccinia virus promoters positioned in the same orientation, wherein the orientation of the vaccinia virus promoters is opposite to that of the adjacent intact reading frame L024, optionally in the presence of a pharmaceutically acceptable carrier or excipient.

33. The composition according to claim 32, wherein the polypeptide is a cytokine selected from the group consisting of GM-C SF, IL-10, IL-12, and IL-21.

34. A method of treating cancer, the method comprising administering to a subject a TK-deficient vaccinia virus, wherein the virus comprises an inactivated N1L gene, wherein the N1L gene is inactivated by the insertion of a single expression cassette, said expression cassette comprising a nucleic acid sequence, said nucleic acid sequence encoding a single heterologous polypeptide, and wherein the nucleic acid sequence comprises at least three vaccinia virus promoters positioned in the same orientation, wherein the orientation of the vaccinia virus promoters is opposite to that of the adjacent intact reading frame L024.

35. The method according to claim 34, wherein the polypeptide is a cytokine.

36. The method according to claim 35, wherein the cytokine is selected from the group consisting of GM-CSF, IL-10, IL-12, and IL-21.

37. A method of treating cancer and/or a tumour in a subject, the method comprising administering to said subject the TK-deficient vaccinia virus of claim 27, wherein the subject is also receiving a cancer therapy.

38. The method according to claim 37, wherein the heterologous polypeptide is a cytokine.

39. The method according to claim 37, wherein the heterologous polypeptide is selected from the group consisting of GM-CSF, IL-10, IL-12, and IL-21.

40. The method according to claim 37, wherein the cancer therapy is chemotherapy, biological therapy, radiotherapy, immunotherapy, hormone therapy, anti-vascular therapy, cryotherapy, toxin therapy, molecular cancer therapy, gene therapy, or any combination thereof.

41. The method according to claim 37, wherein the gene therapy is tumour suppressor gene therapy, suicide gene therapy, viral vector immunisation strategy, anti-angiogenic therapy, pro-apoptosis gene therapy, or gene replacement therapy.

42. The method according to claim 37, wherein the cancer and/or tumour is a non-resectable cancer and/or tumour.

Description

[0066] The invention will now be further described by way of reference to the following Examples which are present for the purposes of reference only and are not to be construed as being limiting on the invention.

[0067] Reference is made to a number of drawings in which:

[0068] FIG. 1 shows the sequence of the modified vaccinia promoter mH5 (SEQ ID NO:1).

[0069] FIG. 2 shows the sequence of the expression cassette comprising three mH5 promoters (SEQ ID NO:2).

[0070] FIG. 3 shows the formula of the vector comprising a nucleic acid sequence according to the invention.

[0071] FIG. 4 shows a schematic representation of VVL15N1L vectors and the N1L pShuttle plasmid used to create new VVL15N1L vectors. The vectors are named as indicted. The long horizontal bar depicts the double stranded DNA genome of VV. L024, N1L (L025), L026 and TK refer to transcription units.

[0072] FIG. 5 shows the confirmation of N1L deletion in VVL15N1L vectors.

[0073] FIG. 6 shows the biological distribution of VVL15 and VVL15N1L vectors in tumour tissue (FIG. 6a) and off-site locations (FIGS. 6b and 6c).

[0074] FIG. 7 shows the replication of VVL15N1L in different cell lines. The graphs on the left represent replication curves of VVL15 (solid) in comparison to VVL15-N1L (dashed); whereas those on the right correspond to VVL15-RFP (solid) versus VVL15-N1L (dashed).

[0075] FIG. 8 shows the cytotoxic potency of VVL15N1L (hatched) compared with VVL15-RFP (solid) different cell lines.

[0076] FIG. 9 shows the cytotoxic potency of VVL15N1L and VVL15N1L armed with mIL-12 or mGM-CSF.

[0077] FIG. 10 shows IFN-γ production in splenocytes co-cultured with growth-arrested SCC7 cells (FIG. 10a) and heat-inactivated VVL15 (FIG. 10b) and treated with VVL15, VVL15N1L or PBS.

[0078] FIG. 11 shows IFN-γ production in splenocytes co-cultured with DT6606-ova cells (FIG. 11a), ovalbumin antigen (FIGS. 11b) and B8R peptide (FIG. 11c) and treated with VVL15, VVL15N1L or PBS.

[0079] FIG. 12 shows IFN-γ production in splenocytes co-cultured with growth-arrested LLC cells (FIGS. 12a) and B8R peptide (FIG. 12b) following treatment with VVL15, VVL15N1 L or PBS.

[0080] FIG. 13 shows the efficacy of VVL15N1L in pancreatic cancer mouse model—tumour growth rate in DT6606 (FIG. 13a) and CMT-93 (FIG. 13c) following treatment with VVL15, VVL15N1L or PBS and survival rate in DT6606 (FIG. 13b) and CMT-93 (FIG. 13d) flank tumour models following treatment with VVL15, VVL15N1L or PBS.

[0081] FIG. 14 shows the efficacy of VVL15N1L in orthotopic lung cancer mouse model. FIG. 14a demonstrates the individual weight profiles of PBS-treated mice, FIG. 14b demonstrates the mean weight of mice in each group as a function of time, FIGS. 14c and 14d, the corresponding Kaplan-Meier survival curves and median survival plots respectively.

[0082] FIG. 15 shows tumour volumes in LLC tumour model following IT administration of VVL15RFP, VVL15N1 L or PBS (FIG. 15a) and metastases in each treatment group at sacrifice (FIG. 15b).

[0083] FIG. 16 shows tumour growth in DT6606 flank models following treatment with VVL15N1L, VVL15N1L-mIL-12, VVL15N1L-mGM-CSF or PBS (FIG. 16a) and the corresponding Kaplan Meir survival curves (FIG. 16b).

[0084] FIG. 17 shows the assessment of IL-21 armed VV in vitro.

[0085] FIG. 18 shows the anti-tumour efficacy of VV-ΔTkΔN1L-mIL-21, VV-ΔTkΔN1L-hIL-21 and control virus VV-ΔTkΔN1L.

[0086] FIG. 19 shows sequences of human IL-12A (SEQ ID NO:5), IL-12B (SEQ ID NO:6), IL-21 isoform 1 (SEQ ID NO:3), IL-21 isoform 2 (SEQ ID NO: 4) and GM-CSF (SEQ ID NO:7).

EXAMPLES

Example 1: Construction of pUC19-N1L Shuttle Vector

[0087] The inventors constructed a pUC19-N1L shuttle vector illustrated in FIG. 3 comprising a specific expression cassette flanked by a fragment containing L024 as well as 31 bp of L025 (left arm), and a fragment containing 22 bp of L025 as well as L026 (right arm). The expression cassette has the following features: (1) there are three vaccinia virus mH5 promoters, and under each promoter there is a cloning restriction enzyme site for easy insertion of any gene of interest; (2) a reporter gene RFP is driven by one mH5 promoter for positive selection of recombination virus; (3) mH5 promoter only drives the expression of the inserted gene from left to right. The homologous recombination strategy used in this invention was designed to replace almost the entirety of the coding sequence of the L025 (N1L) locus. Sequence analysis at the junctions of L024/25 and L025/26 confirmed that the ORFs upstream (22 bp) and downstream (31 bp) of L025 remained intact.

[0088] The cDNA of m-GM-CSF, h-GM-CSF, m-IL12 and h-IL12 were cloned using standard techniques into the vector for expression under control of the mH5 promoter using the appropriate restriction enzyme to synthesise pUC19 super shuttle vectors.

Example 2: Construction of VVL15N1L

[0089] FIG. 4 depicts vaccinia virus Lister strain and various vaccinia virus constructs created by the inventors. Each pUC19 super-shuttle vector was transfected (using an Effectene-based protocol—Qiagen) into CV1 cells that had been pre-infected (2 hr earlier) with VVL15 (0.1 PFU per cell). 48 hr later, the presence of red fluorescence under fluorescence microscope confirmed expression of the relevant cassette, either from the cytoplasmic plasmid or from the relatively few viruses in which homologous recombination had been successful. These latter were selected out as follows. Cells and supernatant were harvested by scraping the cells from the dish and freeze-thawing twice. 1 μl of this lysate was used to infect all 6 wells of a six-well plate containing CV1 cells grown to 80-90% confluence. This low viral load would ensure the emergence of well separated plaques.

[0090] After a further 48 hr, each well was carefully scrutinized under fluorescence microscope searching for those virus-induced plaques that fluoresced red. Upon identification of positive colonies, their location was marked on the under surface of the plate with a fine-tipped permanent marker. The colony was carefully picked with a 20 μl tip filled with 5 ul 5% FCS CM after aspirating the medium from the well. The tip was then submerged into a cryotube containing 250 μl of 5% FCS CM. Following further freeze-thaw cycles, 5-20 μl of this virus solution was added to each well of a new 6-well plate containing CV1 cells as before. This process was repeated until every plaque fluoresced red, i.e., all viral colonies were due to recombinant virus. In general, it took between 4-8 rounds of plaque purification to obtain a pure batch of recombinant virus. At this point, the viral lysate was scrape-harvested and viral DNA was extracted via a column-based system (i.e., the Blood Mini Kit from Qiagen). A sample of supernatant was also tested for the presence of the relevant cytokine by ELISA. The purity of virus was confirmed by PCR amplification of the N1L gene from extracted viral DNA. Its presence would indicate contamination with the parental virus, VVL15.

[0091] Once preliminary investigations had confirmed the likely creation of a pure recombinant virus that expressed the relevant transgene, 50 μl of viral lysate was added to a T175 flask containing CV1 cells, again grown to 80-90% confluence in approximately 30 ml of 5% FCS CM. Cells and media were scrape-harvested 48 hr later and kept as a “primary viral expansion”.

Example 3: Confirmation of Deletion of N1L in Recombinant VVLs

[0092] The N1L gene was deleted in all novel VVL recombinants (FIG. 5a). Sense and anti-sense N1L gene primers were used to amplify via PCR viral DNA that had been extracted from infected CV1 cells. Only the VVL15 viruses contained this gene. The N1L gene containing segment spanning the primer pair was expected to measure approximately 750 bp. The A52R gene was present in all VVL recombinants. Sense and anti-sense A52R gene primers were used to PCR amplify this locus from DNA extracted from infected CV1 cells. The A52R gene segment spanning the primer pair was expected to measure approximately 880 bp (FIG. 5b).

Example 4: Validation of Transgene Expression from N1L Gene-Deleted Recombinant VVLs

[0093] A sample of supernatant from the final plaque purification round of each transgene-armed recombinant virus was analysed using the relevant cytokine-specific ELISA kit, according to the manufacturer's protocol (ebioscience, Biolegend). To assess whether each cytokine transgene was expressed by the relevant recombinant virus upon tumour cell infection, the same experimental set up as described in the viral replication assay above was conducted. At 24, 48 and 72 hr after viral infection, supernatant was collected from each duplicate set of wells and the concentration of cytokine was determined by ELISA according to the manufacturer's instructions. The control samples were supernatants collected from VVL15-N1L infected wells.

Example 5: Assessment of Viral Biological Distribution of VVL15N1L Vectors

[0094] 2×10.sup.8 CT26 cells in 100 μl of serum free DMEM was injected subcutaneously into the shaved right flanks of 7-week old female BALB/c mice. When tumours were approximately 100 mm.sup.3 they were randomized into two groups. An IV dose of 1×10.sup.8 PFU of either VVL15 or VVL15-N1L was injected via the tail vein. At days 1, 3, 7 and 10 post virus injections, 3 mice from each group were sacrificed via CO.sub.2 inhalation. Blood was collected via cardiac puncture into pre-heparinized 1.5 ml Eppendorf tubes. These along with harvested tumour, brain, lungs, liver, spleen, kidneys and ovaries were immediately snap-frozen on a Petri dish floating on precooled (to −80° C.) isopentane. At a later date they were thawed, weighed and homogenised (or vortexed in the case of blood) in serum-free DMEM. Samples were diluted with a volume of 5 μl per mg (or 5 μl per μl of blood). Following a further freeze-thaw cycle, tissue homogenates were subsequently titrated for live viral PFUs using the TCID50 assay.

[0095] A biological distribution experiment was performed to establish whether IV-delivered virus disseminated to tumour and to determine any off-target replication (i.e. to determine the extent of tumour selectivity and thus safety). The CT26 subcutaneous flank model was utilized for this section as described in the Methods section. Following tail vein injection, both viruses could be recovered from tumour tissue until at least 10 days after injection. The peak titre is between 3 and 7 days. Unexpectedly, the viral recovery of VVL15N1L was not reduced in tumour tissues compared to VVL15 (FIG. 6a).

[0096] Regarding off-target replication, with the exception of lung tissue, virus was not recovered from any other organs or blood within 24 hr post-injection. After 24 hr, VVL15-N1L viral recovery was significantly less than VVL15 from liver and spleen tissue and completely absent from kidney tissue. Neither virus was recovered at detectable levels from brain, heart, ovaries or the circulation at any time point in this experiment (FIG. 6b). In contrast, virus persisted in the lungs until at least 3 days. Even in this organ, recovery of VVL15-N1L was significantly less in comparison to VVL15. The novel backbone VVL15N1L therefore appeared to have an even greater selectivity for tumour tissue than VVL15.

Example 6: Replication of VVL15N1L in Different Cell Lines

[0097] There are multiple mechanisms by which a tumour cell may be killed by vaccinia virus. These include innate host defences triggering apoptosis, death from virus-mediated cellular burst and host immunological defence mechanisms. If a virus is excessively cytotoxic to a cell, it may not generate enough progeny to self-propagate throughout a tumour. Furthermore its ability to replicate might be expected to correlate positively with expression of its therapeutic transgene since there will be more copies of the virus present.

[0098] All cell lines were permissive to infection with VVL15 and VVL15N1L. The most permissive tumour cell line was SCC7, resulting in viral titres of over 1,000 PFU/per cell just three days after initially being infected with 1 PFU/per cell. This result contrasts to that from the cell cytotoxicity assay (MTS), in which SCC7 was the most resistant cell line. The cell line least permissive to replication was CMT93. Viral titres plateaued by day 3 in CMT93 and DT6606 cell lines. This is likely to reflect the out-performance of viral replication in comparison to the replication of uninfected cells (with virions effectively running out of cells to infect); coupled with possible viral degradation from proteases released from lysed cells. Statistically significant attenuation of replication of VVL15N1L was seen only in CMT-93 (FIG. 7).

Example 7: Cytotoxic Potency of VVL15N1L

[0099] VVL15-N1L was compared with VVL15-RFP for its cytotoxicity in a range of murine cancer cell lines in vitro. Cells were seeded at 1×10.sup.3 or 1×10.sup.4 cells per well, depending on growth rates, in 96-well plates, and infected with viruses 16-18 hours later. Cell survival on day 6 after viral infection was determined by MTS assay and EC50 value (viral dose killing 50% of tumour cells) was calculated as described in Wang et al (J Clin Invest, 2009, 119:1604-1615). All assays were performed at least three times. Based on the EC50 values, (i.e. the PFU required to kill 50% of cells), there was no significant difference in cytotoxicity between the two viruses in CT26 and DT6606 cells. In contrast, VVL15-N1L was significantly more potent compared to VVL15 at killing CMT93, LLC and SCC7 cells (FIG. 8).

[0100] The cytotoxic potency of armed VVL15N1L was also compared. EC50 values were significantly higher than VVL15-N1L (i.e., they were less potent than VVL15-N1L at killing the relevant cell line) in all cell lines except LLC and CMT93. The VVL15-N1L-mIL12 recombinant appeared to be more potent than VVL15-N1L-mGMCSF in all cell lines, a feature that reached statistical significance in SCC7 and DT6606 cells (FIG. 9).

Example 8: VVL15N1L Induces a Higher Level of Host Immune Response against Tumour Antigen

[0101] Three syngeneic in vivo tumour models were used: SCC7 cells in C3H/HeN strain mice; DT6606-ova cells in C57BL/6 strain mice and LLC cells in C57BL/6 strain mice.

[0102] Subcutaneous flank tumour models were established and treated with a single dose of virus or PBS as outlined in Table 2. To ensure timely generation of viral- and tumour-specific T cells, spleens were harvested at 14 days post infection. IFN-γ release assays were performed on the subsequently generated splenocyte suspensions.

(1) Preparation of a Single Cell Suspension of Splenocytes from Harvested Spleens

[0103] Syngeneic subcutaneous flank tumours were established for the relevant tumour cell line as described below (see also Table 2). When tumours were approximately 100 mm.sup.3 in volume, mice were randomized into three groups. 1×10.sup.8 PFU of either VVL15 or VVL15-N1 L virus in 50 μl of PBS was injected intratumorally (IT) using a 1 ml insulin syringe attached to a 29-gauge needle. The needle was passed a number of times in different directions throughout the tumour during virus deployment in order for broad dissemination. The third group was injected with the equivalent volume of vehicle buffer, i.e. 50 μl of PBS. 14 days after infection, animals were euthanized via CO2 inhalation. Spleens were harvested under sterile conditions, mashed through 70 μm cell strainers (Becton Dickinson Falcon) using the flat end of the plunger of a 2 ml syringe and flushed through with T cell culture media (TCM) (RPMI-1640, 10% FCS, 1% streptomycin/penicillin, 1% sodium pyruvate) into 50 ml conical flasks. Pelleted splenocytes were re-suspended in 5 ml of RBC lysis buffer (Sigma-Aldrich) following centrifugation at 1,200 rpm and left on ice for 5 minutes. After a wash-centrifugation cycle, they were re-suspended with TCM to a final concentration of 5×10.sup.6 cells/ml.

(2) Preparation of Growth-Arrested Whole Tumour Stimulator Cells

[0104] A single cell suspension of 5×10.sup.6/ml of stimulator cells (i.e., the relevant target or control tumour cell line-SCC7, LLC or DT6606-ova) in cell culture medium (CM) was prepared in a 50 ml conical flask. A 1 mg/ml solution of Mitomycin C (MMC) (Roche) was added to this suspension to achieve a final concentration of 100 μg/ml and incubated in a humidified incubator at 37° C. in air with 5% CO.sub.2 for 1 hr. The cells were subsequently washed twice with 40 ml of PBS, re-suspended in 40 ml CM and incubated until ready to seed (within 30-60 minutes). The now growth-arrested stimulator cells were re-suspended in TCM to achieve a final concentration of 5×10.sup.5 cells/ml.

(3) The IFN-γ Release Assay as a Surrogate Marker for Tumour/Antigen-Specific T Cell Activation

[0105] This assay is based on the release of IFNγ when memory T cells are activated by their cognate epitope-MHC complex. The splenocyte pool should contain all the cellular types (e.g. APCs, Th cells) necessary for the stimulation of CD8+ T cells. 100 μl of each of the splenocyte suspensions from (1) were co-cultured with 100 μl of the target-tumour stimulator cell suspension from above in triplicate wells of a round-bottomed 96-well plate (i.e. 5×10.sup.5splenocytes with 5×10.sup.4 growth arrested tumour cells). Splenocyte-only control wells contained 5×10.sup.5 splenocytes in 200 μl TCM. Where appropriate, splenocytes were also co-cultured with 100 μl of ova-peptide (H-2Kb/SIINFEKL, Proimmune) in TCM (to achieve a final concentration of 5 μg/ml) or 100 μl of TCM containing 5×10.sup.4 MHC-compatible, growth-arrested control tumour cells (B16-F10 when a C56BL/6 mouse-derived tumour model was used).

[0106] Furthermore, in order to prove that virus had been administered and importantly that the animal was able to mount an immune response per se, splenocytes were additionally co-cultured as above with either heat-inactivated VVL15 (100 PFU per cell, heated to 56° C. for 2 hr) or a VV B8R peptide (H-2Kb/TSYKFESV, ProImmune), a strongly antigenic Vaccinia viral epitope (to achieve a final concentration of 5 μg/ml). This experiment would also serve as a positive control for the assay itself. Plates were incubated at 37° C. in air and 5% CO.sub.2 for three days, after which they were centrifuged at 1,200 rpm for 5 minutes. The concentration of IFN-γ in supernatants taken from each of the wells was established using a murine-specific IFN-γ ELISA kit (Biolegend). The final concentration of IFN-γ, averaged across duplicate wells was determined after deduction of corresponding values obtained from wells containing splenocytes alone.

[0107] The SCC model is an aggressive murine head and neck squamous cancer model that like its counterpart in human head and neck cancers is poorly immunogenic. As demonstrated in FIG. 10a-b, splenocytes from the VVL15-N1 L-treated group produced significantly higher levels of IFN-γ than the VVL15 group in response to co-culture with growth-arrested SCC7 cells. There was a low but significant level of IFN-γ production from PBS-treated splenocytes. This likely reflected the host's natural immune response against tumour. As expected, there was no IFN-γ production by splenocytes in response to co-culture with heat-inactivated VVL15 in the PBS group. Surprisingly there was no significant difference between the other two groups in IFN-γ levels upon co-culture with inactivated virus (FIG. 10b). It should be noted that absolute levels of IFN-γ following co-culture with inactivated VVL15 were lower in magnitude in comparison to that following co-culture with growth-arrested tumour cells. This likely resulted from variations in presented immunogenic tumour or viral epitopes. In subsequent experiments, to ensure epitope standardization, an immunogenic VV B8R epitope was used in place of inactivated VVL15.

[0108] Host immunity induced by VV in a pancreatic cancer model was also investigated. As the tumour associated antigen (TAA) profile of the DT6606 cell line had not been defined, the cell line DT6606-ova which stably expressed the foreign antigen ovalbumin was created to demonstrate the putative generation of an antigen-specific immune response (in this case an anti-ovalbumin response). This cell line was used to create a syngeneic subcutaneous flank model as described in Table 2. At 14 days post-IT injection of virus, the VVL-N1 L-treated group demonstrated a significantly higher IFN-γ response from harvested splenocytes compared to the VVL15 or PBS treatment groups upon co-culture with growth-arrested DT6606-ova cells (FIG. 11a). There was a relatively high level of IFN-γ production in the PBS group, indeed no different from the VVL15 treatment group. Ovalbumin is a foreign antigen, with the propensity to stimulate a robust anti-ova response, so this result is unsurprising. The N1L treatment group also produced the highest level of IFN-γ when splenocytes were co-cultured with the ovalbumin antigen (although this did not reach statistical significance) (FIG. 11b).

[0109] Again, the splenocyte IFN-γ response between viral groups was not statistically different following co-culture with the B8R epitope, although the magnitude of the response was nearly 10-fold higher in comparison with tumour/tumour antigen co-culture assays.

[0110] Foreign selection markers such as RFP are likely to be immunogenic and could arguably have caused the in vivo results thus far obtained. To control for this possibility, VVL15-RFP was used as the control virus (instead of VVL15) in a subcutaneous syngeneic LLC flank model (see Table 2). The experimental set up was again otherwise identical to those in SSCVII and DT6606 experiments.

[0111] The previous results were replicated in this model, with the highest IFN-γ production demonstrated by splenocytes from the VVL15-N1 L treatment arm co-cultured with growth-arrested LLC cells (FIG. 12a). This was statistically significant compared to VVL15-RFP and PBS arms. There was no significant difference between viral treatment arms upon splenocyte co-culture with B8R peptide (FIG. 12b). For the VVL15-N1L group, co-culture of splenocytes with growth-arrested B16-F10 cells (as an MHC haplotype-specific control stimulator cell population) led to a lower but statistically non-significant IFNγ level in comparison to co-culture with growth-arrested LLC cells (p=0.0594). It is likely that a number of tumour epitopes are shared between these and other solid tumour cell lines. CTLs generated against these could have been responsible for the IFN-γ levels obtained in the B16-F10 group.

Example 9: Efficacy of VVL15N1L in Pancreatic Cancer Model

[0112] Either 5×10.sup.6 CMT93 cells or 3×10.sup.6 DT6606 cells were subcutaneously implanted into the shaved right flanks of C57BL/6 male mice as described above. Once tumour volumes had reached approximately 100 mm.sup.3, they were randomised into three groups and a dose of 1×10.sup.8 PFU of virus in 50 μl PBS or 50 μl PBS vehicle buffer control was injected as per the treatment schedules outlined in Table 3 (schedule 1 and 2). Tumour volumes were monitored via twice-weekly calliper measurement and mice were weighed weekly. Tumour growth was tracked twice weekly and animals were euthanized as governed by Home Office guidelines when tumour volumes approached 1000 mm.sup.3. There was a statistically significant reduction in tumour growth rate and prolonged survival favouring the VVL15-N1L agent upon treatment of the DT6606 flank tumour model (FIG. 13a-b), while in the CMT93 model there was no difference between tumour growth rates following IT administration of either viral agent, although both were significantly slower-growing than the PBS group (FIG. 13 c-d).

Example 10: Survival in Orthotopic Lung Cancer Following VVL15N1L Administration

[0113] In order to assess whether the viruses were efficacious when administered intravenously, an orthotopic lung cancer model was utilised. 5×10.sup.6 LLC cells in 100 μl PBS were injected into the tail veins of 7-week old female C57BL/6 mice.

[0114] Non contrast-enhanced CT scans of the lungs were used to assess the lung volumes of individual mice over a period of three weeks and any reduction used to extrapolate tumour burden. At a time, determined by the initial presence of tumour on CT, three doses of IV virus/PBS were administered as outlined in Table 3 (schedule 5). Mice were weighed twice weekly and were sacrificed if they showed signs of distress or if weight loss exceeded 20% of their maximal weight.

[0115] All mice developed tumours, with deaths occurring between 14 to 21 days, at which time thoracotomy confirmed extensive lung tumours. Tumours were initially apparent on CT between 4 and 7 days post-injection of LLC cells, thus day 5 post-injection was chosen as the start time for therapy.

[0116] 21 mice were administered with tail vein injections of 0.5×10.sup.6 LLC cells in 100 μl serum-free DMEM. They were randomised into three groups and treatment (see Table 3, schedule 5) commenced from day 5. All mice in the PBS treatment group were symptomatic after 10 days post-injection of LLC cells as evidenced by weight loss and all had died by 21 days (FIG. 14). The median survival was extended by 5 days with VVL15 and 6.5 days with VVL15-N1 L viral treatments respectively, in comparison to the PBS group, although there was no statistically significant difference in survival between the viral groups.

Example 11: Metastatic Dissemination in Lung Cancer Model Following VVL15N1L Administration

[0117] LLC is a very aggressive tumour model with a propensity to metastasise to the lung following subcutaneous flank injections. Indeed it has been reported that surgical excision of subcutaneously grown LLC tumour enhanced the rate of lung metastases, perhaps by the removal an angiogenesis inhibitor secreted by the primary.

[0118] To investigate whether IT injection of VV recombinants can reduce this metastatic rate, 1×10.sup.6 LLC cells were injected subcutaneously into the flanks of 7-week old female C57BL/6 mice. When tumour volumes were approximately 100 mm.sup.3, they were randomised into three groups. Injections of virus/PBS were administered IT as per the treatment schedule in Table 3 (schedule 3). Tumours were monitored via calliper measurement until a group reached the end point of requiring sacrifice (approximately 17-20 days post implantation). All animals were euthanized at the same time, their lungs were harvested and any gross tumour deposits noted. Lung lobes were separated, fixed in 4% formalin, embedded in paraffin, stained with haematoxylin and eosin and sectioned through the largest cross-sectional dimension. For each lobe, slices were also performed above and below the largest cross section. All three sections were scrutinized for tumour deposits by a pathologist who was blinded to the treatment schedule.

[0119] There were no significant differences between groups with regards to tumour volumes at sacrifice (FIG. 15a). However the percentage of mice with lung metastases at the endpoint of the experiment was 14, 43 and 57% for N1L, VVL15 and PBS groups respectively (FIG. 15b). These figures were statistically different from each other. It is however difficult to draw inferences with such small numbers of mice per group (n=7) and the experiment would need to be repeated with larger sample sizes. However this result does suggest the possibility that even if VVL15N1L has no impact on the growth of an aggressive primary tumour, viral therapy may minimize dissemination. It could thus prove to be a good adjuvant therapy.

Example 12: Efficacy of IL-12—and GM-CSF-Armed VV15N1L

[0120] To enhance the antitumour efficacy of VVL15N1L, GM-CSF and IL-12 were inserted into the N1L region of the VVL15N1L vector. The potency of each of these recombinants was tested in vivo against a syngeneic DT6606 subcutaneous flank model (see Table 3, schedule 1). When tumour volumes reached an average of 100 mm.sup.3, daily doses (5 in total) of 1×10.sup.8 PFU of virus (VVL15-N1L, VVL15-N1L-mGMCSF or VVL15-N1L-mIL12) or the equivalent volume of vehicle buffer control (50 ul of PBS) were injected IT (n=7 per group). Tumour growth was followed up via twice weekly calliper measurement (FIG. 16a). The corresponding Kaplan Meir survival curves (FIG. 16b) were based on the necessity for humane animal sacrifice when tumour volumes exceeded 1,000 mm3. FIGS. 16 a-b demonstrate that the GMCSF transgene-armed virus alone was not significantly better than VVL15-N1L, however the IL12-armed virus demonstrated significant potency, leading to cures in 6/7 mice and 100% survival at the end of the experiment. These armed viruses will be tested in other models to establish the universality of this result.

Examples 13 and 14: IL-21 Constructs

Materials and Methods

[0121] Murine pancreatic cancer model: DT6606 subcutaneous syngeneic tumours were established in male C57BL/6 mice. When tumours reached 5-6 mm in diameter, PBS, VV-ΔTkΔN1L-mIL-21, VV-ΔTkΔN1L-hIL-21 or control virus VV-ΔTkΔN1L was administered intra-tumourally (5×10.sup.7pfu/injection) on day 1, 3, 7, 9 and 11. Tumour growth was measured twice weekly and animal survival was monitored. Survival data were compared using Prism® (GraphPad Software, CA, USA) and a log rank (Mantel Cox) test was used to determine significance of survival differences. Significance was determined using an unpaired students T test (*p<0.05; **p>0.01; ***p<0.001).

[0122] Syrian hamster cancer models: Syrian hamsters bearing HPD-1NR tumours—1×10.sup.6 HPD-1NR cells were seeded by subcutaneous injection into the right flank of Syrian hamsters bearing HPD-1NR tumours. When tumours reached 313 mm.sup.3, PBS, VV-ΔTkΔN1L-mIL-21, VV-ΔTkΔN1L-hIL-21 or control virus VV-ΔTkΔN1L was administered intra-tumourally (5×10.sup.7 pfu/injection) on day 1, 3, 7, 9 and 11. Tumours were measured twice weekly and animal survival was monitored. Survival data were compared using Prism® (GraphPad Software, CA, USA) and a log rank (Mantel Cox) test was used to determine significance of survival differences. Significance was determined using an unpaired students T test (*p<0.05; **p>0.01; ***p<0.001). Syrian hamster peritoneal cavity disseminated pancreatic cancer model—1×10.sup.7 SHPC6 cells were seeded into the lower right peritoneal cavity of Syrian hamsters. Four days later, 10 hamsters per group were each injected intra-peritoneally (IP) with 500 μl PBS, 2×10.sup.7 PFU of the different VVs on day 4, 6 and 8. The survival rates of hamsters were monitored. Survival data were compared using Prism® (GraphPad Software, CA, USA) and a log rank (Mantel Cox) test was used to determine significance of survival differences (* p<0.05, **p<0.01, ***p<0.001).

Example 13: Generation of IL-21 Armed VV and Assessment of Anti-Tumour Potency In Vitro

[0123] Human and mouse IL-21 cDNA sequences were inserted into the puc19N1L shuttle vector (as shown in FIG. 3). The standard homologous recombinations were carried out in the TK-deleted backbone of Lister strain vaccinia virus (VVL15) by co-transfecting the resultant plasmid pShuttleN1L-mIL-21, or pShuttleN1L-hIL-21into CV-1 (African green monkey kidney) cells that were pre-infected with VVL15 at 0.05 PFU/cell. The transfected CV-1 cell lysates were subjected to plaque assay. Five plaque purification rounds were performed to propagate a single clonal virus and the modification was re-confirmed by PCR for the deletion of N1Lgene. The resultant viruses are named as W-ΔTkΔN1L-mIL-21 and VV-ΔTkΔN1L-hIL-21. To test the potency of these viruses, a cell death assay (MTS assay) was used to detect the cytotoxicity of the three vaccinia viruses in a murine pancreatic cancer cell line (DT6606), which is derived from a mutant Ras-p53 transgenic pancreatic cancer model.

[0124] FIG. 17 shows the assessment of IL-21 armed VV in vitro. FIG. 17 A & B show cytotoxicity of different oncolytic vaccinia viruses in a murine pancreatic cancer cell line. Cultures of the murine pancreatic cancer cell line DT6606 derived from Ras-p53 transgenic pancreatic cancer mice were infected with different viruses, and cell killing detected by MTS assay six days after viral infection. The curve of cell death induced by virus is shown in FIG. 17A; the EC50 values (viral dose to kill 50% of cancer cells) were calculated (FIG. 17B). A higher EC50 value means that the virus has less potency. FIG. 17 C & D show detection of IL-21 expression and viral replication of different mutants of new generation of vaccinia virus in pancreatic cancer cells in vitro. Cultures of the murine pancreatic cancer cell line DT6606 derived from Ras-p53 transgenic pancreatic cancer mice were infected with different viruses, and IL-21 expression was detected by ELISA assay 24 hours after viral infection (FIG. 17C). The viral replication was detected by TCID50 assay is shown in FIG. 17D. The experiments was triplicated.

[0125] As shown in FIG. 17A & B, the new generation of oncolytic vaccinia virus (VV-ΔTkΔN1L) is still very effective in killing cancer cells; arming the virus with the cytokine IL-21 did not attenuate the virus, instead it increased the cytotoxicity against cancer cells by an (as yet) unknown mechanism (p<0.01).

[0126] In order to check whether the therapeutic gene IL-21 can be expressed in the virus-infected cancer cells and at what level, ELISA was used to detect the expression of the IL-21 protein from the VV-ΔTkΔN1L-mIL-21, VV-ΔTkΔN1L-hIL-21 and control virus VV-ΔTkΔN1L infected-pancreatic cancer cells (DT6606). As shown in FIG. 17C, IL-21 was expressed at a very high level in the cells after infection with VV-ΔTkΔN1L-mIL-21, VV-ΔTkΔN1L-hIL-21, but the control virus VV-ΔTkΔN1L infection did not produce any IL-21 protein. The replication of IL-21 armed virus was not attenuated compare to backbone virus (FIG. 17D).

Example 14: Antitumour Efficacy of IL-21 Armed VV15N1L

[0127] To test whether IL21 can enhance the anti-tumour efficacy of VV-ΔTkΔN1L, the potency of each of these viruses (mouse IL21 or human IL21) was tested in vivo against a syngeneic DT6606 subcutaneous flank model.

[0128] FIG. 18 show the anti-tumour efficacy of VV-ΔTkΔN1L-mIL-21, VV-ΔTkΔN1L-hIL-21 and control virus VV-ΔTkΔN1L. Murine pancreatic cancer model—Tumour growth curve (FIG. 18A) and the survival (FIG. 18B) of the mice treated with different agents are presented (n=7/group). Syrian hamsters bearing HPD-1 NR tumours—Tumour growth curve (FIG. 18C) and the survival (FIG. 18D) of the mice treated with different agents are presented (n=7/group). In each case mean tumour size ±SEM are displayed until the death of the first hamster in each group and compared by one-way ANOVA with post-hoc Bonferroni testing. (FIG. 18E) shows the anti-tumour efficacy of the viral strains in the immunocompetent Syrian hamster peritoneal cavity disseminated pancreatic cancer model.

[0129] The IL21-armed virus demonstrated significant potency, regressed tumour growth (FIG. 18A) and marked prolonged the survival (FIG. 18B) of the mice bearing pancreatic cancer. Given that human cytokines may function better in Syrian hamster than mouse, a subcutaneous Syrian hamster pancreatic cancer model was employed to evaluate the antitumour efficacy of IL21-armed virus. Strikingly the IL21-armed virus demonstrated significant potency, leading to cures in 6 of 7 animals and 86% survival at the end of the experiment (FIG. 18C and D). Based on the antitumour efficacy in subcutaneous model, IL12 or IL21-armed VV demonstrate promising efficacy compared to control virus.

[0130] The major barrier for improving the survival of patients with pancreatic cancer is lacking effective therapeutic agent for advanced pancreatic cancer. To this end, a well-characterised Syrian hamster peritoneally disseminated pancreatic cancer model was used for assessment the feasibility, efficacy and safety of IL12 and IL21-armed VV. IL-12-armed VVLΔTKΔN1L showed induced severe toxicity after systemic delivery (FIG. 18E). Strikingly, VVLΔTKΔN1L-IL21 has the highest therapeutic index for treatment of peritoneally disseminated pancreatic cancer, 70% of Syrian hamster were cured.