CMT64 cells were established from a spontaneous alveolar lung carcinoma of a 17-month-old female C57BL/ICRF a^t mouse. CMT167 and 170, with increased metastatic ability, were selected and cultured from pooled lung metastases from this model (39). Mouse CMT64 cells were provided by Cell Services at Cancer Research UK, and human lung cancer cell line H460 was obtained from the American Type Culture Collection (ATCC). Lewis lung carcinoma cells (LLC, metastatic lung squamous cell carcinoma, C57BL/6) were obtained from CRUK, Clare Hall, Herts, UK. Mouse CMT167, as well as CMT170 lung cancer cell line, human lung cancer cell lines H1299 and A549 and non-cancerous cell lines NIH3T3 (murine) and NHBE (human) were obtained from the Centre for Biomarkers & Biotherapeutics, Barts Cancer Institute, Queen Mary University of London, UK. CV1 (African monkey kidney) cells were obtained from ATCC.

Construction and production of VVΔTKΔN1L, VV CTRL and VVL-m21 were described previously (17, 25). Briefly, homologues recombination was used to delete the thymidine kinase (TK) gene of Vaccinia virus Lister strain, resulting in a virus with improved tumor-specific viral replication (40). Homologous recombination was then employed to remove the N1L region, creating VVΔTKΔN1L, which further improved tumor specificity of the virus and enhanced the generation of anti-viral and anti-tumor immunity after infection (25). In VV CTRL and subsequent versions of the virus (including VVL-m21), the TK gene was replaced with a modified virus B5R protein termed STC that improves EEV production and therefore virus spread after infection (17). In these viruses the original B5R gene has been retained, but the signal peptide, stalk (S), transmembrane (T) and cytoplasmic tail (C) regions (STC) of B5R placed, as second copies, under H5 promoter control within the TK region to obtain VV CTRL. To insert cytokine payloads, murine and human IL-12 were amplified from pUNO1-mIL12 and pUNO1-hIL12 (Invivogen) using the primers mIL12_F:5'-AAGCTTATGTGTCCTCAGAAGCTAACCATCTC-3' where HindIII is underlined and mIL12_R:5'- ATTTAAATCCATACCACATTTGTAGAGG-3' where SwaI is underlined or hIL12_F:5'- AAGCTTATGTGTCACCAGCAGTTGG-3' and hIL12_R:5'- ATTTAAACCATACCACATTTGTAGAGG-3'. These fragments were inserted into the HindIII/SwaI-digested H5-RFP-H5 subclone prior to creation of the entire N1L shuttle vector to create N1L Left Arm-H5-H5-RFP-H5-m/hIL-12-Right Arm (17). Homologous recombination was used to insert this payload into the N1L region of the VV CTRL virus. VVΔTK-ΔN1L-msPD-1 (referred to hereafter as VV-sPD1, containing mouse soluble extracellular PD-1 fragment, msPD-1, amino acids 1-167) was constructed by the insertion of the msPD-1 into the TK region of the Lister vaccine strain of VV (VV Lister) under the control of the H5 promoter using an in vitro intracellular recombination technique previously described (41). The msPD-1 DNA fragment containing SalI and NheI enzyme sites was synthesized and inserted into pUC57 vector (GENEWIZ). msPD-1 DNA fragment was digested by SalI and NheI, and inserted into the TK shuttle vector containing RFP flanked by LoxP sites described previously to construct the TK-msPD-1 shuttle vector (42). The recombination of VV-ΔN1L and TK-msPD-1 shuttle vector was performed in CV1 cells. Viral purification and production were carried out as described previously (41).

The cytotoxicity of the viruses in each cell line was assessed six days after infection with virus using an MTS non-radioactive cell proliferation assay kit (Promega) according to the manufacturer’s instructions. Cell viability was determined by measuring absorbance at 490 nm using a 96-well plate absorbance reader (Dynex) and a dose response curve created by non-linear regression allowing determination of an EC50 value (dose required to kill 50% of cells). A minimum of two biological repeats were carried out.

Appropriate cell lines were seeded in triplicate and infected 16 hours later with virus at a multiplicity of infection (MOI) of 1 PFU/cell. Cells and supernatant (or supernatant only to assess EEV release) were collected at 24-, 48- and 72-hours post-infection and titers were determined by measuring the median tissue culture infective dose (TCID50) on indicator CV1 cells. Cytopathic effect (CPE) was determined by light microscopy seven days after infection. The Reed-Muench mathematical method was used to calculate the TCID50 value of each sample (43). Viral burst titers were converted to PFU per cell based on the number of cells present at viral infection. One-way ANOVA followed by Tukey’s multiple comparison post-test was used to assess significance. A minimum of two biological repeats were carried out.

Appropriate cell lines were seeded in triplicate and infected 16 hours later with virus at a MOI of 1 PFU/cell. Supernatant was collected at 24-, 48- and 72-hours post-infection, IL-12 levels measured by Mouse IL-12 ELISA Kit or Human IL-12 2nd Generation ELISA Kit (eBioscience) and PD-1 levels evaluated by Mouse PD-1 DuoSet ELISA (R&D Systems).

All animal studies carried out were approved by the Animal Welfare and Research Ethics Committee of Zhengzhou University (Zhengzhou, China). The efficacy study of different VVs as monotherapy in CMT64 models was carried out under the terms of the UK Home Office Project License PPL 70/6030 and subject to Queen Mary University of London ethical review, according to the guidelines for the welfare and use of animals in cancer research. Power calculations were carried out to determine required sample sizes using G*Power 3, F-test ANOVA repeated measures setting parameters of α = 0.1, power = 90%, effect size = 0.5.

CMT64 (3×10⁶ cells/mouse) or LLC (1×10⁶ cells/mouse) cells suspended in 100 μl normal saline were implanted subcutaneously into the right flank of female C57BL/6 mice aged six weeks. When tumors were palpable (around 100 mm³), animals were stratified into treatment groups and received therapy according to the protocols. Tumor growth was measured using electronic calipers until tumors reached 1.44 cm² at which point the animals were sacrificed, and tumor volume was estimated twice weekly using the formula:

Where w is width, l is length.

Intratumoural (I.T) injections of 1×10⁸ PFU of virus or vehicle buffer in a total volume of 50 µL were introduced through a single central tumor puncture site and 3-4 needle tracts were made radially from the center; virus was injected as the needle was withdrawn. α-PD1 antibody was kindly provided by Shengdian Wang (CAS Key Laboratory of Infection and Immunity, Institute of Biophysics, Chinese Academy of Sciences, Beijing, China). α-PD1was resuspended in PBS at final concentration of 200 μg/mouse and injected via intraperitoneally (I.P) on day 1, 4 and 7 or 7, 9 and 11, as indicated in the results. Tumor growth curves were terminated on the death of the first mouse in each group, but group survival was monitored until the experimental endpoint, and Kaplan-Meier survival plots generated (one biological repeat was carried out).

In subcutaneous CMT64 models, when the tumor volume reached 150 mm³, mice were divided into three groups to receive I.T. injections of 1×10⁸ PFU VV CTRL, VVL-m12 or PBS only on day 1. On days 3, 8 and 15, subcutaneous tumors, spleens, lungs, livers, kidneys and lymph nodes were harvested from three mice in each group for further investigation, including flow cytometry (FC) analysis, IFN-γ release assay, real-time quantitative PCR, and immunohistochemistry (IHC).

Blood samples were collected from the ocular venous plexus of each mouse for assessing the influence of VVs on hepatic and renal function. Serum alanine aminotransferase (ALT), aspartate aminotransferase (AST) and gamma glutamyl transpeptidase (GGT) were performed using a colorimetric method to evaluate the levels and activity of metabolic enzymes (Elabscience). Serum creatinine (Cr) was determined by Colorimetric Assay Kit following the instructions of the manufacturer (Elabscience). The level of blood urea nitrogen (BUN) was measured using a BUN detection kit (Solarbio Life Sciences) according to the protocol of the manufacturer.

Spleens, lymph nodes and tumors from CMT64 subcutaneous mouse models were extracted and combined with T cell culture medium (RPMI medium, 10% FBS, 1% penicillin-streptomycin, and 1% sodium pyruvate), homogenized, then pushed through a 70-μm cell strainer to create a single cell suspension. Cells were resuspended in red blood cell (RBC) lysis buffer (Sigma-Aldrich), washed in PBS, and the pellet was resuspended in T-cell medium. All fluorophore-conjugated antibodies were used at 1:200 dilution. Single-cell suspensions of spleen, tumor or lymph node were prepared in FC buffer (FB; phosphate-buffered saline (PBS) containing 1% heat-inactivated bovine calf serum (BCS)). Cells were blocked with goat serum (Beyontime) prior to incubation with fluorophore-conjugated antibody for 30 minutes. Cells were fixed in 2% formalin and analyzed using an LSRFortessa multichannel flow cytometer (Beckton Dickinson (BD) Biosciences). Raw data were analyzed using FloJo v10. Live cells were gated and from these, CD8+ T cells (CD3+CD8+), central CD8+ T (CD8+CD44^hiCD62L^hi, TCM) or effector memory CD8+ T (CD8+CD44^hiCD62L^lo, TEM) cells, CD4+ T cells (CD3+CD4+), NK cells (NK 1.1+), Treg cells (CD4+FOXP3+), dendritic cell (DC, CD11c+), and macrophages (F4/80+ CD11b+, M1: CD86+MHCII+CD206-, M2: CD206+) were identified.

CMT64 subcutaneous tumors were collected, sectioned, snap-frozen, and stored at-80°C. Frozen tissues were processed for IHC analysis of CD8+ T (BioLegend; 100702) and CD4+ T (BioLegend; 100402) cell infiltration according to the manufacturer’s instructions. Tumor tissues were fixed immediately in 10% buffered formalin phosphate and embedded in paraffin. IHC staining was performed using a labeled streptavidin-biotin method for IHC detection of CD31(Abcam; 182981).

MVD was assessed by IHC analysis using antibodies against the endothelial marker CD31 and determined according to the method of Weidner and colleagues (44). Briefly, the immunostained sections were screened at low magnifications (40× and 100×) to identify hot spots, which are the areas of highest neovascularization. Any yellow-brown stained endothelial cell or endothelial cell cluster that was clearly separate from adjacent microvessels, tumor cells, and other connective tissue elements was considered a single, countable microvessel. Within the hot spot area, the stained microvessels were counted in a single high-power (200×) field, and the average vessel count in five hot spots was considered the value of MVD. All counts were performed by three investigators in a blind manner. Microvessel counts were compared between the observers and discrepant results were reassessed. The consensus was used as the final score for analysis.

Spleens were collected from animals in the CMT64 subcutaneous mouse model and cells were separated using a 70 μm cell strainer and T cell culture medium. Red blood cells were lysed using RBC lysis buffer (Sigma-Aldrich) and cells re-suspended in complete T cell culture medium. 5×10⁵/well/100μL splenocytes were seeded into each well of a 96-well plate in triplicate. Cells were re-stimulated with 5×10⁵ mitomycin-C (MMC)-treated CMT64 or LLC cells. Separately, in order to assess immune response to the viruses, 100 μl aliquots of splenocyte suspensions were treated with Vaccinia virus B8R peptide (TSYKFESV) (GL Biochem) at a concentration of 5 mg/ml in a 96-well plate. Re-stimulated splenocytes were incubated for 72 h at 37 °C, and the supernatant was collected for quantitative detection of IFN-γ using Mouse IFN gamma uncoated ELISA kit (Invitrogen) according to the manufacturer’s instructions.

Statistical analysis was carried out using Graph Pad Prism 8 and SPSS 19.0 software. To compare differences between groups, one-way or two-way ANOVA with Bonferroni or Tukey’s post-test was used, and results were expressed as mean± SEM. Survival was estimated by the Kaplan-Meier method and differences between groups were analyzed using log-rank test. A value of p < 0.05 (*) was considered as statistically significant.

We have previously demonstrated that VVLΔTKΔN1L has potent efficacy and tumor selectivity (25). To boost the anti-tumor potency of VV, we have previously described the generation of VV CTRL that contains a second, mutated copy of the viral protein B5R. This addition improves the production of the EEV, critical for long-range spread of the virus in the host (17). To maximize the oncolytic potential of VV, IL-m12 was incorporated into the N1L region of VV CTRL under control of the H5 promoter creating VVL-m12 (Figure 1A).

To determine the biological characteristics of VVL-m12 in vitro, cytotoxicity, replication and IL-12 expression of the virus were compared to the unarmed viruses including VVLΔTKΔN1L and VV CTRL in lung cancer cell lines. CMT64 as well as its sublines CMT167 and 170, representing late-stage invasive lung cancer, and LLC, representing metastatic lung squamous cell carcinoma, were all derived from mouse models of lung cancer (25, 39, 45, 46). A549, H460 and H1299 cells, representing the common primary malignant cell lines for human lung cancer in vitro studies, were also examined, but a virus expressing the human IL-12 cytokine (VVL-h12) was used in these instances.

VVL-m/h IL12 was more cytotoxic compared to VVLΔTKΔN1L in all tumor cell lines examined, as evidenced by a decrease in the infectious dose required to kill 50% of cells (EC50) (Figure 1B; Supplementary Figure S1A). VVL-m/h12 also replicated efficiently in all lung cancer cell lines examined, whereas both replication and cytotoxicity was significantly attenuated in normal cells including NIH3T3 and NHBE cell lines (Figures 1B, C; Supplementary Figures S1B, D).

Interestingly, arming VV CTRL with IL-12 led to improved replication in both murine and human lung cancer cell lines, an observation that we are as yet unable to explain. However, as arming the virus with IL-12 did not improve the cytotoxicity of the virus (compared to VV CTRL) it seems reasonable to infer that any biological effects of the IL-12 payload is mediated by its immune promoting properties as opposed to potential effects on viral replication, although this cannot be excluded as a mechanism of action.

As expected, due to the inclusion of the STC payload in VV CTRL and VVL-m/h12, both of these viruses were more competent than the parental virus VVLΔTKΔN1L, that does not contain the STC insertion, at promoting EEV release in murine and human lung cancer cell lines (Figure 1D; Supplementary Figure S1E).

To assess IL-12 expression by VVL-m/h12, lung cancer cells were infected with virus and the supernatant collected for IL-12 detection by ELISA at 24, 48 and 72 h timepoints post-infection. These results demonstrated effective production of IL-12 at all-time points in the above cell lines. The concentration of IL-12 in the supernatant reached its peak at 48 h (H1299) or 72 h post-infection. (Figure 1E; Supplementary Figure S1F).

CMT64 subcutaneous tumors were established in female C57/BL6 mice and the animals received intratumoral (I.T) injections of 1×10⁸ PFU of VVL-m12, VV CTRL, VVL-m21 or PBS on day 1, 3, 5, 7, 9 and 11. The dose of selected VVs was 10 times lower than the most commonly reported experimental dose of 1×10⁹ PFU/injection (47). In the CMT64 subcutaneous model, VVL-m12 was significantly more effective at controlling tumor growth and improving survival compared to VV CTRL, VVL-m21 and PBS and no toxic side effects were noted (Figures 2A, B). Efficacy of VVL-m12 reached its peak at day 22 after therapy after which point therapeutic response waned. There was no difference in the weight of mice between the VVL-m12 and other groups (Figure 2C), suggesting that IL-12 did not add toxicity to the therapeutic regime.

Monoclonal antibodies that block the inhibitory PD-1 or PD-L1 pathways have shown remarkable efficacy in patients suffering with lung cancer. As an alternative strategy for antitumor therapy, a soluble form of PD-1 containing the extracellular domain of the membrane bound molecule has shown promising potential in animal models and clinical antitumor immunotherapy (48, 49). VV-msPD1 was rationally engineered by inserting a soluble PD-1 peptide in the TK region (Supplementary Figure S2A). VV-msPD1 replicated efficiently in and was selectively cytotoxic to all lung cancer murine cell lines examined (Supplementary Figures S2B, D, E). Effective production of PD-1 in the supernatant of lung cancer cells after infection of VV-msPD1 was detected by ELISA at 24, 48 and 72 h timepoints post-infection (Supplementary Figures S2C, F).

We investigated whether the combination of VVL-m12 with α-PD1 antibody or the sequential provision of VVL expressing IL-m12 then VV-msPD1 was able to enhance the antitumor efficacy in murine lung cancer. Both CMT64 and LLC lung cancer models were established and treated according to the protocols described in the methods. The efficacy of α-PD1 antibody treatment alone was limited with no significant difference compared with PBS in CMT64 animals (Figures 3A, B). Using mIL-12 armed viruses in the treatment regime showed a significant advantage in this model, and the addition of PD1 antibody may be beneficial. 9/10 animals treated with VV-m12/VV PD1 antibody cleared tumor completely by day 30, while 8/10 animals treated with VV-mIL12/PBS or VV-m12/VV-msPD1 cleared tumors by day 50. These animals were kept for their natural lifespan (around 1 year further) and no tumor recurrence was noted. Superiority of IL-12 was also noted using an aggressive LLC model compared with both PBS and α-PD1 antibody treatment. Overall survival was significantly improved after treatment with these viruses and targeting PD1 expression using either VV-msPD1 or α-PD1 antibody in combination with virus was significantly more effective than treatment with VV-mIL12 alone (Figures 3C, D). Overall, this data suggests that arming the virus with IL-12 and including a further therapeutic boost with either α-PD1 or a VV expressing soluble PD1 (VV-msPD1) creates a highly effective therapeutic regime for long-term elimination of lung cancer in murine models.

Successful immunotherapeutic regimes aim not only to inhibit the primary tumor, but also to induce durable long-term immunity to prevent tumor recurrence. Thus, mice that had effectively cleared primary tumors in sequential VVL-m12 and α-PD1 antibody as well as VVL-m12 monotherapy groups were re-challenged with CMT64 cells 30 days after the primary tumors had completely cleared. Treatment with both therapeutic regimes resulted in long-term antitumor efficacy to CMT64 lung cancer cells as evidenced by rapid tumor clearance that necessitated no further immunotherapy. Although sequential VVL-m12 and α-PD1 antibody treatments were able to clear the secondary tumor more quickly, the difference between the two groups was not significant (Figure 3E).

The infiltration of immune cells into tumors, referred to as tumor-infiltrating lymphocytes (TILs), consequent to immunotherapy is a strong predictor of treatment success. To investigate TILs in response to our treatment regime, a CMT64 lung cancer subcutaneous model was established and treated once I.T. with PBS or VVs (1×10⁸ PFU). Tumor tissue was collected on days 3, 8 and 15 for analysis using flow cytometry (FC) and immunohistochemistry (IHC). At all time-points, treatment with VVL-m12, but not VV CTRL or PBS resulted in significant increase in CD4+ TIL, and more predominantly CD8+ TIL, in the tumor tissue demonstrating that VVL-m12 acts primarily via adaptive CD4+ TIL and CD8+ TIL to mediate antitumor effects in this model (Figures 4A–C).

TCM subsets including CD8+TCM (CD44^hiCD62L^hi CD8+) and CD4+TCM (CD44^hiCD62L^hiCD4+), and TEM cells consisting of CD8+TEM (CD44^hiCD62L^loCD8+ cells) and CD4+TEM (CD44^hiCD62L^loCD4+ cells) were analyzed using FC on day 3, 8 and 15 after treatment according to gating criteria for these cells (Supplementary Figures S4A, B). VVL-m12 significantly promoted CD8+ and CD4+ TCM formation by day 15 compared to VV CTRL or PBS (Figure 5A). CD8+ TEM and CD4+ TEM populations, which are critical for the early phase antitumor immune responses, increased on days 3 and 8 post-treatment. VVL-m12 treatment was able to clear tumor more quickly via induction of TEM together with TCM and more consistently owing to a greater TCM influx compared with VV CTRL (Figure 5A).

We examined interferon (IFN)-γ production from splenocytes isolated from these models and found that VVL-12 treatment was able to modestly induce IFN-γ production from ex vivo re-stimulated splenocytes at 3 or 8 days after VV therapy, but significantly enhanced ex vivo IFN-γ production by splenocytes stimulated with mitomycin C-treated CMT64 and LLC lung cancer cells harvested at 15 days post-treatment. The same trend of IFN-γ release occurred after restimulation with the viral immunogen B8R (Figure 5B).

Using the same model described above, innate immune cells in the spleen and lymph nodes were examined at 3, 8 and 15 days post-treatment. VVL-m12 treatment did not enhance NK cell counts in the spleen (Figure 6A). We demonstrated that VVL-m12 treatment significantly increased DC populations on day 3, reduced Treg cells from day 3 to 8 (Figure 6B), and was effective at re-polarization of M2 macrophages to an M1 phenotype on day 3 and 8 in vivo and in vitro (Figure 6A; Supplementary Figures S3A, B, S4A), all of which may contribute to a more powerful antitumor potential in the early stage of antitumor immunity.

Both VV and IL-12 have been reported to have powerful anti-angiogenic effects that may provide a further anti-tumor mechanism to the treatment. To assess this, MVD was determined from CMT64 tumor sections treated once with virus or PBS. Seciotns were stained with an antibody reactive to CD31 and the number of the micro-vessels per high-power field (HPF) counted. Three days after therapy, MVD in PBS, VV CTRL, and VVL-m12 groups was 32.67± 3.53, 27.67± 5.24 and 1.75± 0.48, respectively. The difference between the three groups was more significant over time using one-way ANOVA analysis (MVD in PBS, VV CTRL, and VVL-12 at day 8: 49± 5.57, 25.5± 3.66 and 14± 2; day 15:66.67 ± 2.40, 21.5 ± 3.01 and 1 ± 0.41). Quantification revealed VVL-12 treatment resulted in significant decrease in vascularity compared with PBS and VV CTRL (Figure 6C).

As toxicity reactions have been associated with systemic delivery of IL-12, we evaluated the safety associated with I.T (localized) administration of VVL-m12 compared with VV CTRL and PBS. Liver and renal function was examined after VVL-12 (1×10⁸ PFU), VV CTRL (1×10⁸ PFU) or PBS were administered once into subcutaneous CMT64 tumors. Liver toxicity was assessed by measuring ALT, AST and GGT. Renal toxicity was evaluated by measuring Cr and BUN. No elevations were noted in any enzymes indicating that liver and renal toxicity was not a side-effect associated with VVL-m12 treatment (Supplementary Figures S5A, B). Furthermore, VVL-m12 also exhibited strong tumor selectivity in vivo. Tumor tissues, liver, lung, and kidney were examined after I.T. administration of VVL-m12 into C57/BL6 mice bearing subcutaneous CMT64 lung cancer on days 3, 8 and 15. Vaccinia virus late transcription factor 2 (VLTF-2) gene, detected after VVL-m12 treatment, was confined to lung cancer tissue and absent from liver, kidney and normal lung tissues (Supplementary Figure S5C).