Enhanced IL-12 transgene expression improves oncolytic viroimmunotherapy

Introduction Malignant peripheral nerve sheath tumors (MPNSTs) are aggressive sarcomas with unacceptably low cure rates occurring often in patients with neurofibromatosis 1 defects. To investigate oncolytic Herpes Simplex Virus (oHSV) as an immunotherapeutic approach, we compared viral replication, functional activity, and immune response between unarmed and interleukin 12 (IL-12)-armed oncolytic viruses in virus-permissive (B109) and -resistant (67C-4) murine MPNSTs. Methods This study compared two attenuated IL-12-oHSVs with γ134.5 gene deletions (Δγ134.5) and the same transgene expression cassette. The primary difference in the IL-12-oHSVs was in their ability to counter the translational arrest response in infected cells. Unlike M002 (Δγ134.5, mIL-12), C002 (Δγ134.5, mIL-12, IRS1) expresses an HCMV IRS1 gene and evades dsRNA activated translational arrest in infected cells. Results and discussion Our results show that oHSV replication and gene expression results in vitro were not predictive of oHSV direct oncolytic activity in vivo. Tumors that supported viral replication in cell culture studies resisted viral replication by both oHSVs and restricted M002 transgene expression in vivo. Furthermore, two IL-12-oHSVs with equivalent transcriptional activity differed in IL-12 protein production in vivo, and the differences in IL-12 protein levels were reflected in immune infiltrate activity changes as well as tumor growth suppression differences between the IL-12-oHSVs. C002-treated tumors exhibited sustained IL-12 production with improved dendritic cells, monocyte-macrophage activity (MHCII, CD80/CD86 upregulation) and a polyfunctional Th1-cell response in the tumor infiltrates. Conclusion These results suggest that transgene protein production differences between oHSVs in vivo, in addition to replication differences, can impact OV-therapeutic activity.


Introduction
Malignant Peripheral Nerve Sheath Tumors (MPNSTs) are rare, aggressive soft tissue sarcomas, which most often occur as a neoplasm in patients with neurofibromatosis 1 (NF1).MPNSTs exhibit resistance to non-surgical treatment due to their high metastatic potential and rates of relapse, low response rates to chemotherapy, and inclination for fast disease progression (1,2).Prognosis for MPNSTs is poor, with a pooled recurrence rate of 42%, distant metastasis rate of 27%, and mortality rate of 26% (3).Moreover, overall mortality of MPNSTs have been found to remain largely unaffected by degree of tumor resection or use of radiation therapy as an adjuvant therapy (3).As such, there remains a critical need to continue efforts towards developing effective therapies for MPNSTs.Oncolytic virotherapy is a novel therapy alternative to chemo-or radiation therapy used to treat MPNSTs (4).Oncolytic viruses are modified viruses that selectively replicate in malignant tumors to induce direct cytotoxic or immune-related anti-tumor activity and are often equipped to express transgenes that can heighten their therapeutic effect (5).Oncolytic Herpes Simplex Virus (oHSV) is one such oncolytic virus that has been proposed as a promising cancer therapeutic, with ongoing progress in clinical trials and one agent already approved by the FDA for the treatment of melanoma (6)(7)(8)(9).
In the present study, we tested the anti-tumor effects of two different murine (m)IL-12-expressing oHSVs in virus-permissive (B109) and -resistant (67C-4) murine MPNSTs.Both oHSVs contain deletions in the principal neurovirulence gene, g 1 34.5, that render it safe to use in patients but maintains selective replication in tumor cells (10).One of the Dg 1 34.5 IL-12 oHSVs, M002, has previously been demonstrated to enhance survival of mice bearing intracranial brain tumors and lack significant systemic or neurologic toxicity following treatment in a nonhuman primate model (11,12).M032, an oHSV with the exact configuration as M002, but expresses human (h)IL-12 in lieu of mouse IL-12, has completed Phase I clinical trial (Clinical Trial NCT02062827).The other IL-12 oHSV tested, C002, is derived from C134, a Dg 1 34.5 oHSV that expresses the HCMV gene IRS1.Addition of the HCMV gene enables the attenuated oHSV to evade the infected cell's dsRNA-activated translational arrest response enhancing oHSV replication in tumor cells without restoring wild type neurovirulence and thus improving the therapeutic response (13,14).
IL-12 is a heterodimeric cytokine, primarily produced by antigen presenting cells, that induces IFN-g production, promotes NK-cell mediated cytotoxicity, and enhances T cell responses (15).It also facilitates the differentiation of naïve CD4(+) T cells to Thelper 1 cells and promotes the differentiation of naïve B cells into IgM-secreting plasma cells (16,17).The capacity of IL-12 to regulate both the innate and adaptive arms of the immune response, coupled with its intrinsic anti-tumor effects, has rendered it an attractive immunotherapy over the last several decades (18,19).
We hypothesized that IL-12 transgene expression by an oHSV with improved viral gene expression would enhance the oHSV efficacy.We compared the functional effects of the viral gene expression and replication of a Dg 1 34.5 IL-12 expressing oHSV (M002) and Dg 1 34.5, IRS1 IL-12-expressing oHSV (C002) in the treatment of murine MPNSTs, B109 and 67C-4.This study highlights nuanced differences between oHSVs in biologically active IL-12 production despite similar replication within murine sarcomas.The variability in IL-12 production associates with distinct cytokine-related effects within the tumor immune microenvironment.oHSVs generating elevated IL-12 demonstrated superior therapeutic efficacy across both tumor models.These findings underscore the critical role of viral IL-12 expression in initiating antitumoral immune response and enhancing the efficacy of oncolytic virotherapy.

Materials and methods
Cell lines and viruses 67C-4 was kindly provided by Dr. Nancy Ratner (University of Cincinnati, Cincinnati, OH), and B109 by Dr. Steven Carroll (University of South Carolina, Charleston, SC).Both MPNST cell lines were maintained in DMEM, supplemented with 10% FBS.Vero cells were maintained in DMEM with 5% BGS and were obtained from ATCC (Manassas, VA).Viruses HSV-1(F), R3616, C101, and C134 have been described previously (10,13,20).In brief, HSV-1(F) is an F strain wild type HSV, and all the recombinants used in this study were F strain derived.R3616, the Dg 1 34.5 recombinant, was graciously provided by Dr. Bernard Roizman (University of Chicago, Chicago, IL).C101 is a Dg 1 34.5 virus derived from R3616 that expresses EGFP (13), and C134 is a Dg 1 34.5 virus derived from C101 that contains the HCMV IRS1 gene under control of the HCMV IE promoter in the UL3/UL4 intergenic region (13).M002 is a Dg 1 34.5 virus that expresses mIL-12 and was kindly provided by Dr. James Markert (University of Alabama at Birmingham, Birmingham, AL).C002 is a C134derived virus that contains the same mIL-12 bicistronic cassette as M002 in the g 1 34.5 locus and expresses mIL-12.

Viral recovery plaque assay and qPCR
67C-4 and B109 cells were plated into clear, 24-well flat-bottom polystyrene tissue culture plates and allowed to adhere overnight at 37°C.The following day, cells were infected with C101, C134, M002, or C002 at a multiplicity of infection of one (MOI=1) for 2 hours with virus diluted in 225µL infection medium (DMEM, 1% BGS).The medium was replaced with growth medium (DMEM, 10% FBS) after the 2 hours of incubation.At 24-and 48-hours post infection, cells were scraped off wells, lysed, centrifuged to pellet cell debris, supernatants were serially diluted in infection medium (DMEM, 1% BGS), and virus recovery was measured by limiting dilution plaque assay using Vero monolayers.Viral plaques were quantified 2 days post-infection from May-Grünwald/methanol-stained plaque dishes.

Viral load qRT-PCR
Flank tumors were harvested and homogenized by mechanical disruption using scissors and the TissueRuptor II (Qiagen).DNA was extracted from aliquots of tissue homogenates using the QIAamp DNA Mini Kit (Qiagen).200ng gDNA, forward and reverse primers at a final concentration of 300nM, probe at a final concentration of 100nM, and Taqman Gene Expression Mastermix (Applied Biosystems) were mixed and aliquoted into a qPCR plate.Following primers and probe used were used: HSV Pol F: 5'-ACCGCCGAACTGAGCAGAC-3', HSV Pol R: 5'-TGAGCTTGTAATACACCGTCAGGT-3', HSV Pol Probe: 5'-CGCGTACACCAACAAGCGCCTG-3' (14).Amplification protocol included 2 min Hold at 50°C, 10 min Hold at 95°C, 35x Cycle of 15 sec at 95°C and 1 min at 60°C.Samples were quantified using a standard curve generated using a log diluted (+) control standard cosmid DNA and normalized to the amount of starting DNA.

IL-12 qRT-PCR
Flank tumors were harvested and homogenized by mechanical disruption using scissors and the TissueRuptor II (Qiagen).Aliquots of tissue homogenate were stored in RNAlater at -80°C until the day of assay.RNA was extracted from thawed samples using Quick-RNA Miniprep Plus Kit (Zymo Research) and cDNA produced from equivalent mass of RNA from each sample using the RevertAid RT Kit (Thermo Scientific) followed by qPCR using SYBR Green 2X Mastermix (Applied Biosystems).Amplification protocol included 2 min UDG Activation at 50°C, 10 min Dual-Lock DNA Polymerization at 95°C, and a 40x Cycle of 15 sec at 95°C and 30 sec at 62°C.Specificity of reaction was verified with a melt curve analysis, and samples quantified using a standard curve generated using a log diluted mIL12 (+) control standard plasmid.Primers used (500nM final concentration) are as follows: mIL-12 P35 Sybr F: 5'-GAAACATTATTCCTGCACTGCTGA-3', mIL-12 p35 875 Sybr R: 5'-GCAACTCTCGTTCTTGTGTAGTTC-3'.IL-12 transcript quantity was normalized to the amount of starting RNA.

Animal tumor studies
Animal studies were approved by the Nationwide Children's Hospital Institutional Animal Care and Use Committee (IACUC, protocol number AR19-00177) and performed in accordance with guidelines established by the Department of Defense and the National Institutes of Health (NIH) Guide for the Care and Use of Laboratory Animals.Two syngeneic C57BL/6 tumor models, a flank 67C-4 MPNST model and flank B109 MPNST model, were used in these studies.For 67C-4 flank tumor studies, 3-to 4-weekold C57BL/6 mice were obtained from Envigo (Frederick, MD) and implanted subcutaneously with 4×10 6 cells in 50µL of phosphate buffered saline (PBS)/flank.B109 flank tumor studies were conducted with 3-to 4-week-old C57BL/6 mice from Charles River (Wilmington, MA) implanted subcutaneously with 6 x10 6 to 8×10 6 cells in 50µL of phosphate buffered saline (PBS)/flank in independent studies.Tumor sizes in both studies were measured biweekly by caliper after implantation, and tumor volume was calculated.Due to differences in tumor growth rate and tumor establishment rates, tumor size at the time of treatment differed between the 67C-4 and B109 tumors.Animals were randomized into treatment groups when tumors reached 60-200mm 3 (67C-4) or 64-600mm 3 (B109).Tumors were then treated with saline or oHSV C134, M002, or C002 (3×10 7 PFU in 50µL PBS)/flank intratumorally (ITu).Tumor measurement was terminated upon reaching the IACUC endpoint criteria of 3000mm3 volume per mouse, at which point the mice were euthanized.Flank tumor growth studies were repeated to ensure biological reproducibility.Samples sizes for the studies were as follows: 67C-4 tumor growth study: PBS (n=6), C134 (n=6), M002 (n=7), C002 (n=7).B109 tumor growth study: PBS (n=5), C134 (n=4), M002 (n=8), C002 (n=5).Immune phenotype study: n=8/group.
Enzyme-linked immunosorbent assay and multi-parameter analyte immunoassay (ProcartaPlex) For in vitro studies, cells were infected with M002 and C002 at MOI=1.Culture supernatants were collected at 48 hours to quantify secreted IL-12 cytokine using the mouse IL-12 (p70) ELISA MAX Deluxe Set (BioLegend, CA) following the manufacturer's protocol.For in vivo studies, flank tumors were harvested and homogenized on ice by mechanical disruption using scissors and a TissueRuptor II (Qiagen).Aliquots of tissue homogenate were then stored at -80°C until the day of assay.Samples were prepared by thawing and incubating with an equal volume of SDS-free RIPA lysis buffer (diH 2 O, 150mM Sodium Chloride, 50mM Tris-HCL, 1% Nonidet P-40, 0.5% Sodium deoxycholate) for 30 minutes at room temperature.The lysed samples were centrifuged, supernatants were collected and analyzed with the ELISA or cytokine bead array (ProcartaPlex, ThermoFisher, Waltham, MA) according to the manufacturer's protocol.For the multiplex assay, analytes measured include IL-12 (p40), IFN-g, and TNF-a.Samples were acquired using Bio-Plex 200 (Bio-Rad, Hercules, CA).Cytokines values were normalized for volumes of media suspended during mechanical disruption and depicted as picograms (pg) per ml (pg/ ml) or pg/1000mm 3 volume of tumor.

Statistical analysis
For experiments with two groups, Mann-Whitney 2-sample tests were conducted.In experiments involving four groups, the Kruskal-Wallis test was initially applied, followed by pairwise comparisons using Dunn's method.For viral load RT-qPCR experiments, technical replicates were first averaged, transformed by log2, and analyzed with ANOVA.Longitudinal tumor growth measures were log2 transformed to meet normality assumptions.To account for observational dependencies across days for each tumor, mixed effect modeling was performed on the transformed data, followed by group comparisons at specific time points.TGR (tumor growth rates) were estimated and compared using mixed effect model in SAS (proc mixed).To control type I error, Hochberg method was employed for adjusting multiplicities related to primary hypotheses.Adjusted p-value<0.05was considered significant.Dunn's tests were performed using Minitab 21 (Minitab LLC; State College, PA), and other statistical analyses were carried out using SAS 9.4 (SAS Institute; Kary, NC) or Prism GraphPad v10.1.0(GraphPad Software Inc; San Diego, CA).
In contrast, the 67C-4 MPNST cell line restricted efficient oHSV replication (21).Viral recovery studies showed over 500x less virus recovered when compared to B109 MPNST cells (Figures 1A, C).While M002 and C002 demonstrated similar replication kinetics and viral recovery in 67C-4 cells at 24h (3.07x10 4 PFU/ml vs. 2.92x10 4 PFU/ml, p>0.05) and 48h (3.14x10 4 PFU/ml vs. 2.75x10 4 PFU/ml, p>0.05) post-infection (Figure 1C), they both exhibited statistically higher recovery compared to C134 control virus at both time points.At 48 hours, both M002 and C002 also surpassed Dg 1 34.5 oHSV in viral recovery.These results indicate superior replication of both IL-12 viruses compared to the non-IL-12 control viruses.However, it is essential to note that the observed statistical differences in viral recovery between the IL-12 and non-IL-12 oHSVs may not necessarily translate into significant biological implications, as they fall within the titering variability of ½ log.Notably, IL-12 production in 67C-4 cells differed between M002 and C002 (Figure 1D), with C002 demonstrating a 2.2-fold increase in IL-12 levels compared to M002 (158.73 pg/ml vs 71.9 pg/ml, p=0.0003).In summary, while the IL-12 viruses exhibited similar replication dynamics, the C002 demonstrated enhanced IL-12 protein production in the more restrictive 67C-4 MPNST line.

C002 produced sustained IL-12 and suppressed B109 MPNST tumor growth in vivo
To evaluate the therapeutic potential of oHSVs in tumors that support virus replication, we utilized B109 murine MPNSTs, known for their support of oHSV replication in vitro, and performed tumor growth studies in B109 flank tumors.We hypothesized that M002 and C002 exert similar therapeutic effects based on our in-vitro results; however, our results showed significant differences in anti-tumor activity between the two IL-12 oHSVs.We established B109 tumors in immune competent C57BL/6 mice, randomized, and treated tumors with a single dose of saline or oHSV (3x10 7 PFU) intratumorally (Figure 2A).Contrary to our expectations, neither M002 nor C134 provided therapeutic benefit when compared to saline control treatment in this murine MPNST tumor model (Figure 2B).In contrast, compared to saline, C134, and M002 treatments, a single dose of C002 suppressed tumor growth (p<0.0001)and improved survival in treated mice (Supplementary Figure 1A).
Next, we examined HSV DNA levels in treated tumors at Day 1 (D1) and Day 4 (D4) post-treatment by TaqMan qPCR assay.All oHSVs generated nearly identical levels of viral DNA even at D4 following treatment, although at reduced quantities compared to D1 (Figure 2C).
Next, we investigated differences in IL-12 protein production between virus-treated tumors by measuring IL-12 (p70) levels using ELISA (BioLegend) at D1 and D4 post-treatment from clarified tumor tissue homogenates.C002 treatment consistently generated higher IL-12 protein levels compared to both M002 and C134 at both timepoints (Figure 2D).Notably, no significant differences in the IL-12 levels were found between D1 and D4 in C002-treated tumors.In summary, only C002 treatment produced measurable IL-12 levels above background, persisting up to 4 days posttreatment.Conversely, M002 did not produce detectable IL-12 protein above background levels despite equivalent viral recovery.These results underscore the distinct in vivo therapeutic efficacy of C002 in the B109 murine MPNST model.
Next, we investigated whether C002 also maintains a therapeutic advantage over M002 in 67C-4 MPNSTs, which restrict oHSV replication in cell culture.Tumors were established in C57BL/6 mice and treated with 2 doses of PBS or oHSV (1 week apart), as performed previously (8) and summarized in the provided schematic (Figure 3A).Consistent with our previous results (8), the non-cytokine expressing control oHSV (C134) was ineffective in this OV-resistant sarcoma.The tumor growth results suggested that both IL-12 viruses (M002 and C002) exhibited early therapeutic activity in 67C-4 tumors, with M002 improving survival statistically compared to saline (p=0.049),but the M002-treated tumors grew over time and were not significantly different from the salinetreated cohort at endpoint (p>0.05).In contrast, C002 reduced 67C-4 tumor growth when compared to saline (p<0.0001),C134 (p<0.0001), or M002 (p<0.0001) and improved survival in mice treated with saline (p=0.0072) or C134 (p=0.011)(Figure 3B, Supplementary Figure 1B).We also measured HSV DNA levels using TaqMan qPCR assay.Similar to the B109 study, there were no significant differences in the viral DNA counts among the oHSV-treated tumor cohorts, both at day 1 and day 4 post-treatment (Figure 3C).These findings indicate that the observed improvement in the therapeutic efficacy of C002 cannot be attributed to enhanced viral replication.
In this study, M002 generated similar IL-12 protein amounts as the non-IL-12 oHSV-or saline-treated samples did at both time points measured.Additional assessment of IL-12 (p70) protein levels by traditional ELISA (similar to the B109 studies) confirmed that functional p70 IL-12 was consistently present in the C002 treated tumor samples (Supplementary Figure 2A).To evaluate if transcriptional differences between the viruses could account for these IL-12 protein changes after C002 and M002 treatment, we performed qRT-PCR, which showed abundant IL-12 transcript in the M002-treated tumors (Supplementary Figure 2B).In conclusion, these results demonstrate that there was reduced IL-12 cytokine produced in M002-treated 67C-4 tumors, despite increased viral recovery and mIL-12 transcription, suggesting that M002-expressed transgene was not translated into IL-12 protein.

C002 treatment increased immune cell infiltrates within 67C-4 tumors
Next, we examined how IL-12-oHSV treatment affected tumor infiltrating immune cell populations.Mice were sacrificed at Day 11 post-initial treatment (D4 post-2 nd treatment), and tumor infiltrating leukocytes were analyzed by spectral flow cytometry.Phenotypic characterization was initially performed using an unsupervised approach using self-organized cluster maps (FlowSOM) from tumor infiltrating leukocyte populations isolated from the tumors using a concatenated training sample followed by individual sample analyses (Data not shown).Cluster differences were assessed and verified by traditional twodimensional (2-D) flow cytometry analysis between the treatment cohorts to confirm these population changes.A sequential gating strategy was performed for a comprehensive phenotypic analysis of both lymphoid and myeloid subsets (Supplementary Figure 3).
Tumors treated with C002 exhibited significantly greater CD45(+) immune infiltrates compared to those treated with PBS or M002 (Figure 4A).Among the treatment groups, only C002-treated tumors had increased proportions of CD45(+) cells, indicating an enhanced immune activity induced by C002-treatment.Within the CD45(+) infiltrates, both C002 and M002-treated tumors predominantly showed a CD3(+) T cell enriched microenvironment with proportional decreases in the CD11b(+) myeloid populations compared to PBS-treated tumors (Figure 4B).Despite their lower frequencies in the oHSV-treated tumors, we proceeded to investigate the specific myeloid cell composition within tumor infiltrates by further phenotyping CD11b(+) cells based on lineage-specific markers for monocytes, macrophages, granulocytes, and dendritic cells (Supplementary Figures 4A-G).While there was a significant decrease in myeloid cell frequencies in the oHSV-treated tumors, there was an increase in the absolute numbers, specifically in C002-treated tumors compared to M002 and PBS treatments (Supplementary Figures 4A-G).While the increase in myeloid numbers in the C002treated group may have functional implications, it is important to note that this myeloid cell increase was linked to the greater overall CD45 (+) cells abundance after C002 treatment.

C002 treatment increased proinflammatory monocyte and antigen presenting cell function in 67C-4 MPNSTs
To investigate whether there were functional differences within the myeloid phenotypes associated with C002, we investigated the activation status of these myeloid cells, focusing on the upregulation of the co-stimulatory molecules, CD80 and CD86, along with MHC-II expression.These molecular signatures serve as indicators of cellular activation and effective antigen presentation.We used mean fluorescence intensity (MFI) to quantify CD80, CD86, and MHC-II expression level on a single-cell basis and found that CD11b(+) dendritic cells isolated from C002-treated tumors exhibited increased CD80 and MHC-II surface expression compared to those from M002-treated tumors (Figure 5A).We observed no significant changes in the expression pattern of CD11b (-) dendritic cells for MHCII or co-stimulatory markers between the two treatments (Supplementary Figure 4H).Next, we examined Ly6C HI and Ly6C LO monocytes and found CD86 and MHC-II expression was significantly upregulated in both subsets from C002treated tumors (Figures 5B, C) compared to M002-treated samples.The Ly6C LO patrolling monocytes also showed increased CD80 expression (Figure 5C) following C002 treatment.Finally, there was a significant upregulation of all three molecules (MHCII, CD80, &CD86) in F4/80(+) macrophages from C002-treated tumors compared to M002-treated tumors (Figure 5D).These macrophages also produced greater TNF-a after C002-treatment compared to M002-treated tumors (Supplementary Figure 4I).These findings show that there was not only an increase in the number of monocytes and antigen presenting cells (APCs) but that there were some functional differences with increased costimulatory (CD80, CD86) and MHC-II expression following C002 treatment.

C002-treatment increased polyfunctional CD4(+) T cell effectors in 67C-4 tumors
Based upon the increased expression of co-stimulatory molecules (CD80 and CD86) and MHC-II upregulation differences in APCs isolated from C002-treated tumors, we hypothesized that the upregulation of these surface repertoires in APCs lead to an amplified T cell, particularly CD4(+) cell response.To test this, we examined lymphoid cell dynamics (numbers, activation status, and memory phenotypes), from both the PBSand oHSV-treated tumors.Consistent with our and others' past studies (8,(22)(23)(24), virotherapy had its greatest impact on the T cell population and produced no significant change in CD19(+) B cell frequencies when compared to mock-treated cohorts (Supplementary Figure 5A).Both M002 and C002 treatments significantly increased CD3(+) T cell frequencies and numbers when compared to PBS-treated tumors (Figure 4B).Notably, there were increases in the ab-T phenotype population, as evidenced by the observed decrease in gd-T cell proportions (Supplementary Figure 5B).In addition, CD3(+) NK1.1(+) Natural Killer T (NKT) cells and CD3(-) NK1.1(+) NK cell frequencies remained unchanged (Supplementary Figures 5C,  D), highlighting the stability of this distinct T cell subset within the total T cell population.Further phenotyping of T cells revealed that C002 tumor treatment increased both CD4(+) T cell frequency and absolute number (Figure 6A), with a concomitant decrease in the CD8(+) frequency (Supplementary Figure 5E), resulting in an increased CD4: CD8 ratio (Supplementary Figure 5F).
Because we detected an increase in effector/effector memory CD4 (+) T cell frequency within C002-treated tumors, we next examined the functional potential of these intratumoral CD4 T cells.Consistent with recent studies that have highlighted the pro-inflammatory nature of Ly6C(+)CD8(+) T cells (25, 26), we identified a significant increase of Ly6C(+), CD8(+) T cells in C002-treated tumors relative to PBS-treated tumors (Supplementary Figures 5K-M).This finding led us to investigate Ly6C expression on the CD4(+) T cells in the tumors where we anticipated a similar proinflammatory phenotype change akin to the Ly6C(+)CD8(+) T cells.The results showed both increased frequency and absolute counts of Ly6C(+)CD4(+) T cells in C002-treated tumors (Figures 6E, F).The role of these cells in tumor immunity and if they are unique to virotherapy treatment will require further investigation.Next, we assessed the anti-inflammatory T cells in the tumor microenvironment by quantifying FoxP3(+) regulatory T cells (Tregs).As anticipated, oHSV treatment decreased Treg infiltrates in the tumor (Supplementary Figure 5N).Collectively, these results indicate a shift towards a more pro-inflammatory with reduced immunosuppressive microenvironment in C002treated tumors.Given the increase in pro-inflammatory CD4(+) T cells after oHSV treatment, we next analyzed IFN-g, TNF-a, IL-4, IL-17A, and IL-21 intracellular expression within the CD4(+) tumor infiltrates.CD4(+) T cells expressing IL-17 and IL-21 constituted less than 0.89 ± 0.11% of all cytokine-producing CD4(+) T cells, indicating minimal involvement of Th17 and T-follicular helper (Tfh) in the tumor microenvironment (Supplementary Figures 5O, P).Additionally, we detected a significant decrease in IL-4(+) Th2 cells in M002-treated tumors compared to PBS-treated tumors (Supplementary Figure 5Q).The majority of CD4(+) T cells expressed either TNF-a and/or IFN-g, indicating a T-helper 1 (Th1) subset predominance in the tumor infiltrates.Notably, oHSV treatment shifted the cytokine dynamics from TNF-a to IFN-g, resulting in increased frequencies of single (IFN-g) or double (IFN-g and TNF-a) cytokine-producing cells (Figure 6G).Among the oHSV-treated cohorts, C002-treated tumors exhibited the highest number of IFN-g and/or TNF-a producing CD4(+) T cells (Figure 6H), suggesting a skew in cellular response towards an IFN-g expressing polyfunctional Th1 phenotype.Consistent with the flow data, C002-treated tumors also exhibited higher IFN-g protein levels than did PBS-treated tumors (Figure 6I) that correlated with the level of IL-12 protein in the tumors (Figure 6J).
We detected comparable CD45+ tumor infiltrating lymphocyte populations between the two IL-12-oHSV-treated cohorts in B109 tumors (Supplementary Figure 6A).C002-treated tumors showed fewer B cells in frequency and a smaller CD11b+ myeloid cell population than did M002-treated tumors (Supplementary Figures 6B, C).Notably, this reduction in the myeloid and B-cell populations was accompanied by a compensatory increase in T-cell populations (Supplementary Figure 6D).Similar proportions of CD4(+) and CD8(+) T cells were present in M002 and C002 (oHSV-IL-12)-treated B109 tumors (Supplementary Figure 6E).However, T cells differed in memory and functional aspects between tumors treated by each IL-12 oHSV.C002-treated tumors contained a greater proportion of CD8(+) T cells that were effector memory cells (Supplementary Figure 6F) and expressed more granzyme-B (Supplementary Figure 6G) than M002-treated samples did, similar to that observed in the 67C-4 tumors (Supplementary Figure 4I).Furthermore, the majority of CD4(+) T cells in C002-treated tumors were effector memory cells (Supplementary Figure 6H) and showed increased IFN-g expression compared to M002-treated tumors (Supplementary Figure 6I).
In summary, our studies show that despite similar transcriptional activity, a translationally optimized IL-12 expressing oHSV (C002) produced greater IL-12 cytokines, demonstrated greater IL-12 related immune functional activity, and improved the MPNST therapeutic response compared to another IL-12 expressing oHSV (M002).

Discussion
MPNSTs are challenging tumors for oncolytic herpes simplex virus (oHSV) treatment (7,8).The tumors restrict viral replication and oHSV-associated immune therapeutic activity (7,8).To address these issues, we sought ways to improve immune mediated activity using virus-based cytokine expression and examined two immune competent murine MPNST models (B109 and 67C-4) that differ in their oHSV susceptibility (7).Our prior studies (6, 7) highlighted differences in MPNST susceptibility to infection and gene expression in vitro and focused on human MPNSTs.However, a limitation of human tumor studies lies in its dependence on using immunocompromised animal models for evaluating virotherapy efficacy.This approach eliminates an important component: the immune-mediated therapeutic effects.We therefore used immune syngeneic immune-competent syngeneic murine models to examine the oHSV mediated immune landscaping within the tumor microenvironments.
This study highlights differences between oncolytic HSVs, including those expressing the same immune active murine Interleukin-12 (mIL-12) cytokine, in these murine sarcomas.Both M002 (Dg 1 34.5, mIL-12) and C002 (Dg 1 34.5, IRS1, mIL-12) were constructed using the same targeting plasmid and therefore share the same promoter and IL-12 bicistronic cassette.Our results confirmed that both C002 and M002 produced similar IL-12 transcript levels in the treated tumors but only C002 increased IL-12 protein levels.We previously showed that C134 evades dsRNA-activated translational arrest, improving late viral gene expression, in contrast to early generation (Dg 1 34.5) oHSVs (13, 14).To our knowledge, this is the first time that a virus with improved late gene expression has been shown to improve virus encoded protein production in vivo.These IL-12 results suggest that C134-based oHSVs improve therapeutic transgene expression in the in vivo environment when compared to an early generation IL-12 oHSV (M002).
C002 treatment also increased the tumor infiltration of CD44 (+)CD4(+) T cells, suggesting a shift from naïve to effector/effector memory T cell populations.Further, the absence of CD103 expression as tissue resident marker in these CD4(+) T cells implies they may have migrated from nearby draining lymph nodes after being induced by viral presence in the tumors.At this stage of the study, it is still unclear whether these cells, potentially antigen-specific and influenced by IL-12-related MHC-II presentation changes, are licensed at the lymph node or elsewhere outside of the tumor and whether these cells have other anti-tumor functional activity.
Characterization of CD4(+) T cells demonstrated a predominant Th1 phenotype within C002-treated tumors, evidenced by increased IFN-g expression and a proportional decrease in TNF-a production.Intriguingly, our analysis did not detect Th2, Th17, Tfh or Tregs, indicating a skewed differentiation favoring pro-inflammatory Th1 responses within the tumor microenvironment.Of particular interest was the induction of Ly6C expression on CD4(+) T cells by C002.Ly6C, a protein belonging to the Ly-6 superfamily and often used as a monocyte marker, has traditionally been linked to IFN-g expressing proinflammatory characteristics, especially in CD8(+) cells during bacterial lung infection and acute lung injury in mice (25, 26).Additionally, Ly6C is recognized for homing CD8(+) T cells to secondary lymphoid tissues (27).Similarly, previous studies in LCMV infection have shown that Ly6C HI CD4(+) T cells produced higher levels of IFN-g and granzyme B compared to Ly6C LO cells (28).In our investigation, we identified that C002 induced the expression of Ly6C on the surface of CD4(+) T cells, signifying a pro-inflammatory state.The functional relevance of Ly6C(+)CD4(+) T cells in the context of anti-tumor activity in our study remains unknown at this stage.However, this finding prompts further exploration in future studies to decipher the specific role of this population and its potential implications for the overall anti-tumor immune response induced by C002 treatment.
With the increased population of effector CD4(+) T cells in C002-treated tumors, we anticipated an increased activation state within these cells.While our analysis showed heightened expression of the early activation marker CD44, we did not observe significant changes in other activation markers such as CD69 and CD25.The unexpected absence of CD69 and CD25 in C002-induced CD4(+) T cells raises questions about the optimal time for conducting immunophenotyping of tumor infiltrates to capture the nuanced changes in T cell activation.Performing immunophenotyping at multiple timepoints post-treatment in future studies could provide valuable insights into the evolving activation stages induced by these oHSVs.
These studies also highlight the importance of in vivo oHSV activity assessments.Virotherapy involves a complex anti-tumor response encompassing both direct viral activity within the infected cancer cell and a complex immune-mediated anti-tumor bystander effect.It is not surprising that in vivo studies are necessary to assess this complex tumor-associated immune therapeutic activity.These studies, however, also highlight how direct viral functions (replication and gene expression) can also differ between in vitro and in vivo studies.In-vitro, B109 cells supported both M002 and C002 gene expression, IL-12 cytokine protein production, and viral replication (10 6 -10 7 PFU).However, the B109 MPNST tumors behaves similarly to the more resistant 67C4 tumors.Viral recovery was similar across both tumor models for the oHSV tested in vivo.Furthermore, the B109 tumors were more restrictive to the early generation IL-12 oHSV (M002) in terms of IL-12 production.Although B109 tumors were less aggressive than 67C-4 tumors, they were resistant to C134 or M002 therapy.Neither of the murine sarcomas (B109 nor 67C-4) supported robust oHSV replication as occurs in human tumor models (14).The early generation and next generation oHSVs replicated similarly, producing equivalent IL-12 transcript, but only the C134-based virus (C002) increased IL-12 cytokine protein levels in vivo.The enhanced IL-12 production in C002-treated tumors, however, did not suppress viral replication.Interestingly, differences in IL-12 protein production, absolute increases in immune infiltrates, and functional changes related to IL-12 production (T cell and APC activity), did not associate with viral recovery differences between the two IL-12 viruses.While speculative, it is possible that the early IFN response and the restrictive murine tumor cell environment play greater roles in restricting viral replication than does the later adaptive immunemediated antiviral response in these immune-competent murine models.This highlights a limitation to this study that HSV is a human virus, and viral gene expression and replication is less efficient in murine cells.This virus vs host difference may have exaggerated the observed decrease in cytokine production in the early generation oHSV-treated tumors in this study.
Both cytokine viruses require further in vivo evaluation for cytokine expression levels and replication in human tumors.Human-based cell studies and in vivo xenograft studies may reveal improved cytokine expression, replication, and spread from early generation oHSVs that are not evident from these syngeneic studies.Unfortunately, xenograft studies are imperfect and do not permit a fully integrated immune competent model system with the necessary type I IFN response to fully assess oHSV-related IL-12 activity.
It is also pertinent to address potential concerns regarding the use of IL-12 as an immunotherapy considering its severe adverse effects when administered systemically in early clinical trials in the 1990s (29,30).In the decades since, however, IL-12 has remained an enticing target to scientists and subsequently been demonstrated to be safe to use and tolerable when administered locally rather than systemically, with some preclinical results advancing to phase I clinical trials (19,31).One of these therapies includes the aforementioned M032, an (h)IL-12-expressing HSV-1 that is the human equivalent of M002, which has concluded a dose-escalating phase I clinical trial (NCT02062827) designed to test the safety and tolerability of intratumoral administration in patients with recurrent or progressive malignant glioma.Intratumoral M032 had an acceptable adverse-event profile, and no dose-limiting toxicity occurred at the greatest dose (32).
It is possible that cytokine production of the early generation oHSV may be sufficient in human tumors, as preliminary results from the M032 phase I clinical trial reveal positive responses in some patients (32).Alternatively, it is also possible that the improved cytokine production of the next generation IL-12 oHSV (C002) may be too robust in human tumors, leading to cytokine release syndrome-related morbidity that is poorly predicted by the murine models.We are investigating several approaches in ongoing studies to advance these cytokine viruses using FDA approved medications that we anticipate could be readily incorporated into future oHSV-IL-12 clinical trials and may allow more refined control of the cytokine production in vivo.
One advantage of these studies was the extensive TME cytokine and immune cell phenotypic characterization.By using a full spectrum flow cytometry technique and quantifying tumorassociated cytokine production, we were able to analyze both adaptive and innate TIL changes from the limited numbers of tumor infiltrating leukocytes and identify microenvironment changes post-treatment in the MPNSTs.By incorporating intracellular cytokine staining, we were also able to further characterize T cell functional response changes in the oHSVtreated tumors.We focused on timepoints from past studies where we observed adaptive immune cell changes (4-7d post treatment), but this may have biased the results toward adaptive immune changes.While these studies also detected myeloid cell response changes between oHSV-treated samples like MHCII CD80/86 upregulation in the monocyte, macrophage, and myeloid DC populations, it is possible that other myeloid-and innate cell-associated changes are present at earlier timepoints post-oHSV treatment that these studies did not detect.Additionally, functional studies will be necessary in the future, despite our current findings, which are indicative of APC antigen presentation, increases in pro-inflammatory and patrolling monocytes, and increases in polyfunctional cytokine producing CD4 infiltrates.Functional studies will also allow us to assess whether C002 improves the response of tumor antigen-specific CD4(+) T cells.Lastly, a limitation with oHSV treatment is that anti-tumor responses to treatment may not be uniform.In the B109 tumor model, some tumors were cleared while others grew irrespective of treatment.In the aggressive 67C-4 MPNST model, oHSVs did not produce any cures and only had transient tumor suppressive activity but required repeated dosing.In human studies we anticipate that repeated dosing may be required.

1
FIGURE 1Viral recovery and IL-12 production levels of M002 and C002 in B109 and 67C-4 MPNST cells in-vitro.B109 or 67C-4 MPNST cells were plated into polystyrene tissue culture plates and allowed to adhere overnight at 37°C.Cells were infected the following day with C101, C134, M002, and C002 at a multiplicity of infection (MOI) of 1.At 24-and/or 48-hours post infection, cells were harvested, and virus recovery was measured by limiting dilution plaque assay.Culture supernatants were quantified for IL-12 cytokine protein using mouse IL-12 (p70) ELISA MAX Deluxe Set following the manufacturer's protocol.(A) Viral recovery of Dg 1 34.5 oHSV, C134, M002, and C002 from B109 cells in vitro were similar at 24 hours post infection, but viral recovery of C002 was greater than Dg 1 34.5 oHSV at 48 hours post infection.(B) M002 and C002 produced similar levels of IL-12 protein in B109 cells in vitro.(C) Viral recovery of M002 and C002 from 67C-4 cells in vitro was greater than that of C134 at 24 hours post infection, while M002 and C002 viral recovery was greater than that of both Dg 1 34.5 oHSV and C134 at 48 hours post infection.(D) C002 produced greater levels of IL-12 protein than did M002 in 67C-4 cells in vitro.*p<0.05,**p<0.01,***p<0.001,****p<0.0001.ns, not significant.