The 4T1 mouse mammary carcinoma cells were grown in RPMI 1640 medium supplemented with 10% Fetal Bovine Serum (FBS) (ThermoFisher, MA, United States) and Primocin (ThermoFisher). VC2 was constructed as described previously (Naidu et al., 2020). Briefly, the VC2 recombinant virus was constructed utilizing the two-step Double-Red Recombination protocol using the HSV-1(F) viral genome cloned as a bacterial artificial chromosome (BAC). The virus was cultivated in African green monkey kidney (Vero) cells maintained in complete Dulbecco’s Modified Eagle Medium (DMEM) (ThermoFisher) supplemented with 10% FBS.

Female 8- to 10-week-old BALB/cJ mice () were purchased from Jackson Laboratories (Bar Harbor, ME, United States) and were housed in the Louisiana State University School of Veterinary Medicine (LSU-SVM) Animal Biosafety Level 2 facility. The mice were anesthetized using 4% isoflurane and oxygen was maintained at 0.5L/min. 10⁵/100 µL 4T1 cells were injected in the mammary fat pad to establish the tumor at day 0. After 10 days, when the primary tumor reached a size of 4 mm, animals were injected with 50 µL PBS/UV-VC2/VC2 in RPMI 1640 complete medium every other day for a total of 3 times. Animals were monitored every day and euthanized on day 28 by CO₂ inhalation followed by cervical dislocation. All animal work was approved by LSU IACUC.

One Hematoxylin and Eosin (H&E) stained section including all lung lobes for each treated and untreated animal was evaluated with standard light microscopy by a blinded board-certified anatomic veterinary pathologist. The samples were subjectively classified according to the percentage of pulmonary tissue obliterated by tumor cells originating from the main mammary tumor (metastasis). Metastasis to lymph nodes and cardiac tissue within the slides were not included in this assessment. We refer to this as a semi-quantitative analysis of metastasis.

Following euthanasia, the primary tumors were collected in PBS, were minced and were processed with the Mouse Tumor Dissociation Kit (Miltenyi, Germany) and gentleMACS Octo Dissociator (Miltenyi). The homogenized solution was passed through a 100 μm, 70 μm and 30 μm filter to prepare a single cell solution. This cell suspension was incubated with a pre-titrated antibody mixture for 30 min at 4°C and was then washed and fixed with 2% paraformaldehyde. The next day, samples were analyzed using the BD LSR-Fortessa™ Cell Analyzer and data was processed using the FCS Express 7 software. The following reagents and anti-mouse antibodies used for flow cytometry were purchased from Miltenyi - CD45-APC Vio770, CD11C-PE, CD24 PE-Vio770; Thermofisher - CD3-PerCP-ef710, CD4-FITC, Live/Dead Aqua dye; BD - Gr-1 PE-CF594, CD8a-BV650, gdTCR-BV711, MHCII-BV711, CD64-BV605, CD49b-BV421, CD11b BV786CD38 Percp ef710, Egr2 APC and TCRβ BV605.

Following euthanasia, tissues were immediately fixed using 10% formalin for 3 days and processed in the Histology Core Facility of LSU-SVM. For IHC, 5-μm thick paraffin sections were deparaffinized in xylene and rehydrated through graded alcohols. Antigen retrieval was performed by incubating the slides in boiling Tris-based solution (Vector Laboratories) for 1 h. Endogenous peroxidases were inactivated using 3% H₂O₂ for 10 min. Slides were then blocked with 2.5% Normal Horse Serum (Vector Laboratories, United States) and incubated with 1/100 diluted rabbit anti-mouse CD3 (Abcam). After overnight incubation, slides were incubated with goat anti-rabbit IgG (Vector Laboratories) followed by horse anti-goat horseradish peroxidase (HRP, Vector Laboratories). Slides were developed with Vector VIP substrate (Vector Laboratories) and scanned using the NanoZoomer Slide scanner (Hamamatsu Photonics, Japan).

Tumor specific cytotoxicity was measured using flow cytometric assay. Fluorescein isothiocyanate-labeled 4T1 cells were plated on a 12-well plate on the day before the experiment. On the day of the experiment, spleen was harvested and a single cell solution was obtained from each spleen. One million splenocytes were incubated with 4T1 cells for 3 days in RPMI1640 complete medium. After the incubation period, splenocytes were washed out and 4T1 cells were incubated with 1 µL 7-aminoactinomycin D (7-AAD, ThermoFisher) for 30 min at 4°C. Cells were washed, trypsinized and analyzed using the BD LSR-Fortessa™ Cell Analyzer, and data was processed using the FCS Express 7 software. Cell death was calculated as percent of 4T1 positive for 7-AAD.

Tumor infiltrating T cell proliferation assessment was done using CD3/CD28 Dynabeads. Briefly, tumor infiltrating T cells were isolated using CD4/CD8 beads (Miltenyi). After isolation, cells were labelled using the CellTrace violet dye according to the manufacturer’s instructions (ThermoFisher). Cells were activated using CD3/CD28 beads (ThermoFisher) for 3 days at 37°C in the presence of 30 IU IL-2. After the incubation, the cells were analyzed using the BD LSR-Fortessa™ Cell Analyzer and data was processed using the FCS Express 7 software. Splenic T cells from naïve animals were used as the positive control.

Manual RNAscope duplex was used to quantify gene transcription in 5-μm thick, formalin-fixed, paraffin-embedded lung sections. To stain RNA for specific genes, 5-μm thick tissue sections were prepared according to the ACD guideline. The slides were dried in the oven at 60°C for 1 h followed by 5 min incubation in Xylene and 1 min incubation in 100% alcohol. The slides were air dried before adding H₂O₂ for 10 min at room temperature (RT). All slides were then washed in distilled water and placed in boiled target retrieval agent provided with RNAScope kit. After 15 min incubation slides were submerged in distilled water and transferred in 100% alcohol for 3 min. Slides were dried in the oven and a barrier was created and proteas plus treatment was done before proceeding for RNAScope. WE used RNAScope 2.5 HD Duplex assay kit from ACDBio (Cat. no 322430) (Advanced Cell Diagnostics, United States). For all incubation with Probes and Amps, a hybridization chamber from ACDBio was used at 40°C. After protease plus treatment slides were washed with distilled water and assay probes were added followed by 2-h incubation. Following incubations Amp1-6 were added with washing in between using washing buffer provided with the kit. After Amp6 incubation red color was developed using red reagents. After red color development Amp7-10 were added with washing in between steps. Green color was developed after Amp10 using green reagents. All slides were washed and transferred in H&E solution briefly before adding 0.2% ammonium hydroxide as bluing reagents. All slides were dried before adding mounting medium. As control positive and negative control probes were used from ACDBio. The following probes were used: V-HSV186-UL48 (VP16), Mm-CD274(PD-L1)-C1 and Mm-VEGF-A-C2. All images were scanned using the NanoZoomer Slide scanner (Hamamatsu Photonics) and analyzed using Visiopharm Image Analysis Software.

To detect different cell populations within tissue sections immobilized on slides, first an app was created to distinguish the tissue from background. Empty slides were labelled as background and whole lung as tissue sample. The app was trained at ×0.5 magnification using the Decision Forest classification method with red, green and blue filters. After initial training both false positive and negative readings were optimized using a post-processing step where a minimum threshold (500,000 µm²) was set to define a tissue section. This app was referred to as Tissue Detection App. To detect metastasis a second app was created and trained for metastasis regions from normal tissue. From a normal and metastatic lung several parameters were defined and labelled as empty space, metastasis, normal tissue, muscle, leukocytes and RBC. This input was trained by the Decision Forest classification at ×4 magnification and several filters of Red, Green, Blue, Chromaticity red, Contrast Red Blue and H&E were used. After combined training from several control slides a post processing step was used to optimize the detection by setting a minimum threshold. To calculate metastasis from a slide, first, tissue detection was run followed by application of the metastasis app, which generated the percentage of metastasis within the tissue area. We refer to this as a quantitative analysis of metastasis as it accurately calculates the area.

This was done similarly by generating apps or using premade apps downloaded from the VisioPharm website and running them after the tissue detection app. To calculate the RNAScope and IHC data within a metastatic area, at first metastatic area was defined by a board-certified pathologist and IHC/RNAScope apps were run only within the metastatic area.

Statistical analysis was performed using GraphPad Prism 9 software. For analysis between three groups one-way ANOVA and Kruskal-Wallis test were performed. To compare results between two groups, the Mann-Whitney test was utilized. The statistical significance level was set at p = 0.05.

We have reported previously that VC2 replicated efficiently in Vero and other cells, while it cannot enter neuronal axons and establish latency in mice (Saied et al., 2014; Naidu et al., 2020; Nabi et al., 2021). VC2 replicated efficiently in 4T1 cells as parental HSV-1(F) (Figure 1A) and viral antigens could be detected by flow cytometry (Figure 1B).

To investigate the ability of VC2 to replicate in 4T1-generated tumors in BALB/cJ mice, 0.5 × 10⁶ cells were inoculated in the mammary fat pad of BALB/cJ mice. The schematic of experiment is shown in Figure 2A. Tumors were palpable and could be easily visualized at 7 days post implantation. Implanted tumors grew substantially at the original site of implantation (Figure 2B) and metastasized to the lungs over the course of the experimentation. Examination of the lung at 28 days post tumor implantation revealed substantial metastasis in the form of multiple variably sized nodules (Figure 2C). Individual tumors were injected with approximately 10⁵ PFU of VC2 at days 10, 12 and 14 post tumor implantations. Using the RNAScope assay, HSV-1 VP16 mRNA was localized in multiple tissue sites corresponding to the sites of virus infection (Figure 2D).

The 4T1 mammary carcinoma in mice is a transplantable aggressive tumor that metastasizes spontaneously to draining lymph nodes followed by metastasis in the lung and other tissues similar to human breast cancer (Dexter et al., 1978; Aslakson and Miller, 1992). This characteristic of spontaneous metastasis can be used to evaluate the strength of anti-tumor immune responses following VC2 inoculation of primary 4T1 tumors. Examination of the VC2-treated tumors at day 28 post tumor implantation revealed that there were no significant differences in the primary tumor weights and overall appearances (Figure 3: A). Microscopic examination of the mouse lungs by a board-certified veterinary anatomic pathologist revealed that the VC2-treated animals had significantly less metastasis compared to the mock-treated control group of mice (Figure 3B). Furthermore, animals treated with UV-irradiated VC2 (UV-VC2/inactivated) had significantly higher metastasis compared to VC2-treated mice (Figures 3B, D). Quantitative analysis showed a similar significant reduction in lung metastasis in VC2-treated mice (Figure 3C) in comparison with the mock and VC2-UV inactivated mice, with a high degree of correlation with the semiquantitative data (Figure 3E).

To evaluate the systemic cell-mediated adaptive immune responses following VC2 intratumor therapy, cytotoxicity against 4T1 cells was measured using the membrane impermeant dye 7-amino actinomycin D (7-AAD) that is excluded from viable cells. VC2-treated animals had significantly higher splenocyte-mediated cytotoxicity against 4T1 cell in-vitro compared to the mock- and the UV-irradiated VC2-treated animals (Figure 3F).

Tumor-infiltrating leukocytes were characterized to identify specific immune cell populations associated with the observed reduction in metastasis. The presence of both myeloid and lymphoid population was evaluated within the tumor on day 28. There were no differences in the myeloid compartment for myeloid derived suppressor cells (MDSCs), Natural Killer (NK) Cells, macrophages, M1/M2 macrophages, Dendritic cells and dendritic cell CD11b ± subsets (Figures 4A–F). However, a significant difference was noted in the T cell compartment, with significant T cell infiltration observed in the tumors of VC2-treated animals at 14- to 18-day post infection (Figure 5A).

T cell infiltration within tumors is beneficial in controlling tumor development and metastasis (Desfrancois et al., 2010; Kmiecik et al., 2013; Ali et al., 2014; Waugh et al., 2016; Barnes and Amir, 2017; Kurozumi et al., 2019; Xia et al., 2019; Kuroda et al., 2021) and was associated with improved outcomes (Dexter et al., 1978; Aslakson and Miller, 1992; Desfrancois et al., 2010; Ali et al., 2014; Barnes and Amir, 2017; Bernstein et al., 2019; Nabi et al., 2021; Uche et al., 2021). The IHC highlighted the presence of T cells within 4T1 tumors (Figure 5B). Further phenotyping of T cells revealed that most of the T cells were CD4⁺ and CD4⁺CD8⁺, but there was no difference between the CD4⁺ and CD4⁺CD8⁺ population within each animal group (Figure 5C). However, VC2 treated animals had significantly less T regulatory cells (Tregs) compared to mock-treated animals (Figure 5C) in agreement with previous studies from our laboratory using the mouse syngeneic and immunocompetent melanoma model (Uche et al., 2021). Several studies showed Tregs as a pro-tumor cell type because of their immunosuppressive characteristics (Spranger et al., 2013; Overacre-Delgoffe et al., 2017; Togashi et al., 2019). Thus, reduction of these population suggest an improvement in TME.

It is well known that the tumor microenvironment is immunosuppressive and induces T cell anergy (Webb et al., 2011; Krysko et al., 2012; Quail and Joyce, 2013; Liberti and Locasale, 2016; Palucka and Coussens, 2016). To analyze the activation and exhaustion status of infiltrating T cells, the phenotypic expression of both activation and exhaustion markers in the T cell population was investigated. Infiltrating CD4⁺ T cells in VC2-treated animals exhibited less exhaustion compared to mock-treated animals as evidenced by the lower expression of exhaustion markers PD1 and CTLA4 (Figure 6A). On the other hand, the CD4⁺CD8⁺ T cell population exhibited lower CTLA4 expression, and the population control group appeared more exhausted compared to VC2 treated animals (Figure 6B). Exhausted T cells within the tumor microenvironment generally lack the ability to respond and proliferate following T cell activation signals (Zhang et al., 2020; van der Heide et al., 2022). The ability of VC2 treatment to retain or improve T cell proliferation was tested using the fluorescent tracking dye CellTrace Violet and anti-CD3/CD8 antibodies. Following activation, the purified tumor infiltrating CD3⁺ T cells from VC2 treated animals showed significant improvement in their proliferation ability compared to control (Figure 6C). These findings suggest V2 OVT induces an improved T cell response in the tumor and retains their activation state despite the immunosuppressive tumor microenvironment.

VC2 treated animals had significantly higher T cell infiltration per metastatic area of the lung (Figures 7A, B). Although T cells were identified throughout the lung, it appeared they were very concentrated in the metastatic areas especially in the VC2 treated animals (Figure 7B). The transcription of the pro-tumor markers PD-L1 and VEGF was investigated using a two-color RNAScope assay. The transcription of both PD-L1 and VEGF were significantly lower in metastatic areas of VC2-treated animals compared to mock treatment (Figure 7C). Although the transcripts of PD-L1 and VEGF were easily detectable in both mock- and VC2-treated animals, a significantly lower transcription was observed in the VC2-treated animals (Figures 7C, D).