OV delivery of GP33 to solid tumor cells redirect the activity and cytotoxicity of P14 T cells in vitro

The vesicular stomatitis virus (VSV) is a potential oncolytic viral vector.30 In order to reduce neurotoxicity while retaining the lytic potency and wide-ranging tumor tropism of VSV,31,32,33 the G protein of VSV was replaced with the G protein of Lymphocytic Choriomeningitis Virus (LCMV), and the modified recombinant virus was named rVSV-LCMVG (Fig. 1a). Electron microscopy showed that rVSV-LCMVG maintained the original bullet-shaped particles, and the expression of viral proteins VSV-N, VSV-P, and VSV-M was detectable in anti-VSV-Rat serum, and the presence of LCMVG protein could be detected using anti-LCMVG monoclonal antibodies (mAb) (Fig. 1b). rVSV-LCMVG displayed remarkable cytotoxic effects on diverse tumor cell lines, even at multiplicity of infection (MOI) levels below 0.01. Notably, it exhibited particularly strong killing effects on liver and lung cancer cell lines, highlighting its potential as an oncolytic virus (Fig. 1c and Supplementary Fig. 1a). IC50 of rVSV-LCMVG for various cancer cell lines was in Supplementary Fig. 1b. To evaluate its efficiency in infecting and producing LCMVG in tumor cells, we infected B16-OVA cells with rVSV-LCMVG at different multiplicities of infection (MOI), 16 h later the expression of VSV-P and LCMVG could be detected (Supplementary Fig. 1c, d). We also conducted an assessment of the proportion of tumor cells that exhibited positivity towards LCMVG and VSV-P antigens following exposure to different rVSV-LCMVG MOIs for 12, 24, and 48 h. Flow cytometry analysis demonstrated a significant increase in the proportion of tumor cells expressing LCMVG and VSV-P, which was dependent on the MOI (Supplementary Fig. 1e). Compared to the wild-type VSV, the rVSV-LCMVG also exhibited significantly enhanced safety. The intracranial injection of 1 × 102 plaque-forming units (PFU) of the wild-type VSV led to the mortality of all mice, whereas all mice that received 1 × 106 PFU of rVSV-LCMVG survived (Fig. 1d and Supplementary Fig. 1f). Compared to the wild-type VSV, the modified rVSV-LCMVG demonstrated a low propensity to induce the production of neutralizing antibodies after multiple intravenous doses (Fig. 1e). C57BL/6 J mice were intravenously injected with rVSV-LCMVG at 1 × 107 PFU and no significant weight loss between day 1 to day 30 postinjection when compared with PBS-treated controls (Supplementary Fig. 2a, b). To provide a more detailed analysis of potential toxicity, the same doses of rVSV-LCMVG were injected intravenously and serum ALT (Alanine aminotransferase) well as AST (Aspartate aminotransferase) were determined. Both levels were not elevated in any of the group throughout the observation period, indicating a lack of potential toxicity (Supplementary Fig. 2c, d). Quantitative RT-PCR showed that the rVSV-LCMVG virus genome decreased gradually with the extension of infection time in the blood, heart, liver, spleen, lung, kidney and brain of the treated animals (Supplementary Fig. 2e). These results suggested that rVSV-LCMVG exhibited safety as an oncolytic virus in the treatment of tumors, and it could be employed in a multi-injection, multi-course administration strategy to mitigate the influence of neutralizing antibodies.

Fig. 1

Characterization of rVSV-LCMVG, which could effectively deliver GP33 to tumor cells to direct the activation and cytotoxicity of P14-TCR-T cells in vitro. a Schematic of oncolytic virus rVSV-LCMVG showing the G gene of VSV genome replaced by the G gene of LCMV. b Electron micrographs of VSV and rVSV-LCMVG, and identification of N, P, M, LCMVG protein expression via western blot analysis. c Murine and human cancer cells were infected with rVSV-LCMVG at the indicated MOIs. Cell viability was analyzed at 48 h after virus infection, using CCK8 cell viability assay kits. d Inoculation with wild-type virus VSV or rVSV-LCMVG via intracranial injection, to monitor the survival of mice. e Inoculation 1 × 107 PFU of rVSV-LCMVG or VSV by intravenous injection, one dose every three days and every three injections is a course of treatment, the blood is collected to detect the content of neutralizing antibodies in serum. f Results of B16-OVA tumor cell killing assay, as visualized by phase-contrast microscopy. Representative images are shown. Scale bar, 50 μm. g Expression of cell surface CD69, ICOS, and CD107a on P14 cells after 16-hour coculture with B16-OVA tumor cells in the presence or absence of the indicated MOI of rVSV-LCMVG. h IFN-γ production in supernatants measured by enzyme-linked immunosorbent assay (ELISA) collected from cocultures with P14 at indicated MOIs for 16 h

To assess the susceptibility of rVSV-LCMVG-infected cells to specific T-cell-mediated killing, B16-OVA cells were infected with rVSV-LCMVG for 16 h and co-cultured with P14 cells, which can recognize LCMV-GP33, at effector: target (E:T) ratios of 1:1. The group that underwent combined rVSV-LCMVG infection and P14 cells coculture exhibited significantly higher levels of killing compared to B16-OVA cells infected with rVSV-LCMVG alone or co-cultured with P14 cells alone at each time point (Fig. 1f). CD69 and ICOS were employed as T-cell activation surface markers, while CD107a levels on the cell surface and the concentration of interferon-γ (IFN-γ) in the supernatant were used to assess P14 cells function. The activity of P14 T cells, which were co-cultured for 16 h with rVSV-LCMVG infected B16-OVA cells, exhibited robustness that was dependent on the rVSV-LCMVG MOI (Fig. 1g, h). These findings suggest that OVs have the capability to deliver antigens, in this case LCMVG to tumors and enhance antigen-specific T-cell-mediated antitumor responses.

Tumor-reactive and OV-reactive P14 T cells conferred stronger antitumor immunity

Given the limitations of OVs and adoptively transferred T-cell monotherapy for the treatment of solid tumors, we conducted a study to investigate the potential of combination therapy. In this study, we utilized the B16-GP33 melanoma model, which expresses the exogenous antigen GP33, to assess the effectiveness of combination therapy involving rVSV-LCMVG and P14 cells. Once the tumor size reached approximately 100 mm3 following the subcutaneous injection of B16-GP33 cells, we transferred P14 cells (2 × 106 cells per mouse) on day 0, relative to treatment. The next day, on day 1, the tumors were intratumorally (i.t.) injected with rVSV-LCMVG 1 × 107 PFU per dose for every 3 days for 12 consecutive days (Fig. 2a). To investigate the impact of each component of combinatorial treatment on B16-GP33 tumor growth, groups of mice with established tumors were assigned to four treatment groups: PBS (control), rVSV-LCMVG alone, P14 alone, or combination therapy with rVSV-LCMVG and P14. Tumor growth was assessed every three days. As expected, mice treated with either P14 cells alone or rVSV-LCMVG alone exhibited slower tumor growth compared to the control group treated with PBS. Combination therapy resulted in significant tumor regression and a substantial increase in survival time. 10 days after the injection of rVSV-LCMVG, mice treated with either P14 T cells or rVSV-LCMVG alone showed a moderate reduction in tumor volume, whereas the P14 combined with rVSV-LCMVG group completely eliminated the tumor after 19 days. Furthermore those receiving dual treatment survived for more than 35 days until the conclusion of the experiment (Fig. 2b and Supplementary Fig. 3a). Therefore, in an attempt to address the limited therapeutic impact of systemically administering OVs, we sought to enhance the therapeutic efficacy by combining rVSV-LCMVG with P14 through intravenous injection at an equivalent dosage to the previous intratumoral injection. We transferred P14 into B16-GP33-bearing mice one day before the administration of rVSV-LCMVG (Fig. 2c). Tumors that progressed in the group receiving intravenous administration of rVSV-LCMVG maintained similar levels compared with those in the group receiving PBS treatment. However, when P14 was combined with intravenous administration of rVSV-LCMVG, there was a significant improvement in tumor treatment efficacy and survival rates. (Fig. 2d and Supplementary Fig. 3b).

Fig. 2

Antitumor efficacy of rVSV-LCMVG combined with P14 cells in B16-GP33 tumor models. a Schematic of B16-GP33 tumor-bearing mice treated with rVSV-LCMVG and P14 T cells. b Tumor volumes are shown as mean values with SEM (n = 5 per group). Survival curves of C57BL/6 J mice from the experiment described in a are shown. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001, based on two-way ANOVA with post hoc Holm–Sidak test; survival analysis was conducted using log rank test. c Schematic of the treatment was the same as that in (a), except oncolytic virus was administered intravenously. d Tumor volumes are shown as mean values with SEM (n = 5 per group). Survival curves of C57BL/6 J mice from the experiment described in c are shown. e Representative flow cytometric analysis showing abundance of inhibitory receptors (PD-1 and LAG3) and activation molecules ICOS on tumor-infiltrating P14 cells isolated from the tumor. f Quantification (geometric mean of fluorescence intensity) of the expression levels of PD-1, LAG3 and ICOS on tumor-infiltrating P14 cells. Each dot represents one mouse. g Flow cytometry plot showing the fraction of P14 (CD90.1+) cells in the total CD8+ T-cell gate from the tumor, draining lymph node, and spleen of a representative mouse. h Quantification of P14 in g. Each dot represents one mouse. i Representative intracellular staining for the cytokines IFN-γ and GZMB. j Summary of cytokine production by P14 cells. Each dot represents one mouse. Horizontal bars show the minimum and maximum values (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns means not significant based on the Mann Whitney test)

To investigate the mechanisms by which rVSV-LCMVG enhances the effects of adoptive T-cell therapy, we examined the number, phenotype, and function of P14 cells in tumors and the peripheral region using flow cytometry. This analysis was performed after intratumoral administration of two doses of the oncolytic virus on the fifth day. We observed changes in the functions of P14 cells in the P14 transferred group. These cells upregulated inhibitory receptors, such as PD-1 and LAG3, downregulated the expression of the co-stimulatory molecule ICOS. Furthermore, P14 cells in the P14 combined with rVSV-LCMVG group showed reduced expression levels of LAG3 and PD-1, as well as increased expression levels of ICOS (Fig. 2e, f). Additionally, P14 cells obtained from the tumors, including draining lymph nodes and spleen, showed higher abundance in the group treated with a combination of P14 cells and rVSV-LCMVG, compared to the group treated with P14 cells alone. (Fig. 2g, h). P14 cells in the P14 transferred group also exhibited lower levels of IFN-γ and GZMB upon ex vivo restimulation. P14 cells in the P14 combined with rVSV-LCMVG group produced more IFN-γ and GZMB upon restimulation ex vivo, compared with P14 transferred group (Fig. 2i, j), indicating improved P14 cells function in the TME. Consistent with the rapid development of P14 cells dysfunction, the aggressive growth of the B16-GP33 melanoma tumor could only be controlled by the adoptive transfer of P14 cells during the early stages of the disease. We also analyzed the expression of 36 soluble cytokines and chemokines in the B16-GP33 tumor using the Luminex beads method, in addition to detecting specific T cells cytokine production. In tumors treated with the combination therapy of rVSV-LCMVG and P14 cells, the intratumor levels of IFN-γ, TNF-α, IL-2, IL-12, IL-15, and GM-CSF were significantly higher compared to the single-strategy group. These elevated levels of cytokines could induce tumor regression and stimulate systemic immunity. Furthermore, treatment with the combination therapy also led to significantly higher levels of CCL5 and CXCL10, which may attract inflammatory cells to the injection site (Supplementary Fig. 3c). Therefore, the continuous injections of rVSV-LCMVG after infiltration of P14 cells into the tumor altered the cytokine profile in the TME, as the infiltrating cells responded to rVSV-LCMVG treatment. Correspondingly, the combination of P14 and rVSV-LCMVG treatment significantly enhanced the survival rates of mice bearing B16-GP33 tumors. These findings suggest that the improved tumor control observed after the combination therapy was mediated by the oncolytic virus, which promotes greater infiltration of T cells and enhances their antitumor capacity within the reconstituted tumor immune microenvironment. Consequently, the combination therapy with oncolytic virus has a profound impact on the responses to adoptively transferred T-cell therapy.

OVs enhanced the antitumor function of tumor-specific T cells and OV-specific T cells

To further investigate the changes of tumor-specific and virus-specific T cells when combined oncolytic virus therapy, the antitumor activity of this combination approach was tested in a syngeneic tumor model using C57BL/6 J mice bearing subcutaneous B16-OVA tumors, a melanoma cell line engineered to express the exogenous antigen chicken ovalbumin (OVA). The melanoma cell line B16-OVA was subcutaneously injected firstly, and when the tumor size reached approximately 100 mm3, an appropriate amount of P14 and OT-I (2 × 106 cells per mouse) were transferred on day 0 (relative to treatment). On the next day, followed by four doses of oncolytic virus therapy, one dose every three days, rVSV-LCMVG 1 × 107 PFU/dose (Fig. 3a). A modest decrease in tumor burden and an enhancement in overall survival were observed in mice with intratumoral injection of four doses of 1 × 107 PFU rVSV-LCMVG. Furthermore, significant tumor suppression was observed in the group receiving combined treatment, leading to a more effective extension of the survival rates of mice. (Fig. 3b and Supplementary Fig. 4a). In addition to intratumoral administration, we also assessed the therapeutic efficacy of intravenous administration of the oncolytic virus in conjunction with T cells in the B16-OVA tumor model (Fig. 3c). Compared to the single treatment group, the co-administration of T cells along with intravenous administration of oncolytic virus demonstrated a notable therapeutic effect in inhibiting tumor growth and extending the lifespan of mice (Fig. 3d and Supplementary Fig. 4b). Tumor-specific OT-I T cells isolated from the tumors exhibited high levels of PD-1 and LAG3, whereas bystander P14 cells isolated from the same tumors displayed much lower levels of these markers. Furthermore, the expression of PD-1 and LAG3 decreased in OT-I cells when combined with rVSV-LCMVG treatment, while the expression of ICOS increased (Fig. 3e, f). Furthermore, 5 days after transfer, both OT-I and P14 cells infiltrated the tumors in the OT-I&P14 treatment group, with OT-I cells showing higher levels of infiltration compared to non-specific P14 cells in B16-OVA tumors, while enhanced recruitment of virus-specific P14 T cells was observed in the presence of rVSV-LCMVG. Additionally, tumors, draining lymph nodes, and spleen exhibited a similar trend (Fig. 3g, h and Supplementary Fig. 4c). In addition, OT-I tumor-infiltrating lymphocytes (TILs) demonstrated decreased production of IFN-γ compared to OT-I cells in the spleen. However, when mice were treated with a combination of OT-I and P14 cells along with rVSV-LCMVG, the levels of cytokines secreted by both cells significantly increased in tumors, draining lymph nodes, and spleens (Fig. 3i, j and Supplementary Fig. 4d). The findings demonstrate that the tumor-specific T cells infiltrating the tumor site show signs of exhaustion. Nevertheless, when administered in combination with the oncolytic virus rVSV-LCMVG therapy, the exhaustion phenotype of the tumor antigen-specific T cells (OT-I) can be reversed. Additionally, the detection of cytokines revealed an augmented secretion by the combined OT-I cells and oncolytic virus, thereby intensifying the antitumor effect.

Fig. 3

Antitumor efficacy of combination therapy of rVSV-LCMVG and P14, OT-I cells in B16-OVA tumor models. a Schematic of B16-OVA tumor-bearing mice treated with rVSV-LCMVG and OT-I and P14 T cells. b Tumor volumes are shown as mean values with SEM (n = 5 per group). Survival curves of C57BL/6 J mice in a are shown. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001, based on two-way ANOVA with post hoc Holm–Sidak test; survival analysis was conducted by log rank test. c Schematic of the treatment was the same with a, except oncolytic virus was administered intravenously. d Tumor volumes are shown as mean values with SEM (n = 5 per group). Survival curves of C57BL/6 J mice from the experiment are described in c. e Representative flow cytometric analysis showing abundance of inhibitory receptors (PD-1 and LAG3) and activation molecules ICOS on tumor-infiltrating OT-I and P14 cells isolated from the tumor. f Quantification (geometric mean of fluorescence intensity) of the expression levels of PD-1, LAG3, and ICOS in tumor-infiltrating OT-I and P14 cells. Each dot represents one mouse. g Flow cytometry plot showing the fraction of OT-I (CD45.1+) cells in the total CD8+ T-cell gate, in the tumor, draining lymph node, or spleen of a representative mouse. h Quantification of the OT-I and P14 in the total CD8+ T-cell gate, in the tumor, draining lymph node, or spleen. Each dot represents one mouse. i Representative intracellular staining for the cytokines IFN-γ and GZMB. j Summary of cytokine production by OT-I and P14 cells upon restimulation with cognate peptides. Each dot represents one mouse. Horizontal bars show the minimum and maximum values (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns means not significant based on the Mann Whitney test)

Next, multiplex immunofluorescence imaging was performed to better characterize structures within the tumor and draining lymph nodes at the cell-cell interaction level. When examining the draining lymph nodes, it was observed that the ratio of OT-I and P14 T cells in the combined treatment group was significantly higher compared to the single treatment group (Supplementary Fig. 5a, b). In the tumor region, there was a notable increase in the overall infiltration of CD8 T cells, OT-I, and P14 T cells in tumors treated with T cells in conjunction with rVSV-LCMVG, as opposed to the adoptive transfer of OT-I and P14 alone. Moreover, we found a close colocalization of PD-1 and CD8 expression in tumors, with a relatively low expression level of PD-1 in the combined treatment group (Supplementary Fig. 5c, d). This finding was consistent with previous flow cytometry results.

Transcriptional signature of OT-I and P14 TILs

The aforementioned studies demonstrated distinct proliferative and differentiation responses of tumor-specific T-cell OT-I and virus-specific T-cell P14 to various treatments. Thus, it became crucial to explore the disparities in the transcriptional signatures of these T cells expanded after adoptive transfer of T-cell monotherapy or combined oncolytic virus therapy. To accomplish this, RNA sequencing (RNA-seq) analysis was conducted on the sorted tumor-specific and virus-specific T cells obtained from mice in both treatment groups with B16-OVA tumors. The RNA-seq results indicated significant alterations in the gene expression profiles of both tumor-specific and virus-specific T cells in mice treated with combination therapy as opposed to those treated with monotherapies (Fig. 4a, b and Supplementary Fig. 5e).

Fig. 4

OT-I and P14 T cells have distinct transcriptional profile when combined with rVSV-LCMVG in B16-OVA. Transcriptome kinetics of OT-I and P14 T cells following a transfer of OT-I and P14 T cells into B16-OVA tumor-bearing mice on day 0, then followed by two doses of rVSV-LCMVG intratumorally administered on day 1 and day 4 (or not). For bulk RNA-seq, OT-I and P14 T cells were harvested and sorted on day 5. a Principal components analysis of mRNA matrix from all cells in combination treatment group or the monotherapy. b Venn-diagram showing differential RNA-seq peaks for OT-I and P14 T cells in combination treatment group compared to the monotherapy. c Differences in pathway activity scores of OT-I T cells between the combination treatment and monotherapy groups. d Differences in pathway activity scores of P14 T cells between the combination treatment and monotherapy groups. e Volcano plot of differentially expressed genes fold changes in OT-I T cells between the combination treatment and monotherapy groups. f Heatmap depicting representative protein export genes of OT-I T cells from the combination treatment and monotherapy groups. g Volcano plot of differentially expressed genes fold changes in P14 T cells between the combination treatment and monotherapy groups. h Heatmap depicting representative protein export genes of P14 T cells from the combination treatment and monotherapy groups

Pathway enrichment by gene set variation analysis was performed at the same time as the previous flow analysis on day 5 after adoptive transferred. In the oncolytic virus combined with adoptive T-cell therapy group, both OT-I and P14 cells showed enrichment for cytokine activity, the granzyme-mediated cell death pathway, and positive regulation of T-cell proliferation. Notably, combination treatment resulted in pathway enrichment in granzyme-mediated cell death in P14 CD8+ T cells. (Fig. 4c, d). The expression of various inhibitory receptors and transcription factors, such as Tox, Slamf6, Egr2, and Eomes, known to be associated with T-cell exhaustion, was found to be downregulated in OT-I cells from mice that received combination therapy of rVSV-LCMVG and T cells, compared to the cells isolated from the monotherapy group. In contrast, there was an upregulation in the expression of genes encoding effector molecules and inflammatory cytokine receptors, including Gzmb, Gzmk, Gzma, Ccr5, Ifng, and Stat1, in mice receiving the combination therapy. (Fig. 4e, f). Furthermore, the combined therapy not only reversed the exhausted phenotype of tumor antigen-specific T cells OT-I but also amplified the antitumor effects by enhancing the production of cytokines by virus-specific T cells (Fig. 4g, h). The study findings indicated that both OT-I and P14 T cells treated with the combination therapy exhibited a reduction in exhaustion signature, while demonstrating an increase in effector signatures.

Transcriptional profiling of antitumor T cells in TME by scRNA-seq

To gain a deeper understanding of how the differentiation process of tumor-specific and virus-specific T cells was affected by rVSV-LCMVG, we conducted single-cell RNA sequencing (scRNA-seq) analysis on these T cells following various in vivo treatments. We focused specifically on tumor-specific OT-I T cells obtained from B16-OVA engrafted C57BL/6 J mice. These T cells were then categorized into ten major clusters based on their characteristics: early activated T cells, Xcl1+ T cells, Il7r+ Tem cells, Nme1+ T cells, ISG+ Teff cells, Tcf7+ Tex cells, Gzmb+ Teff cells, S phase Tex cells, Temra cells, and G2m phase Tex cells. (Fig. 5a). OT-I T cells without rVSV-LCMVG stimulation were primarily found in exhausted T-cell clusters (G2m phase Tex, S phase Tex, and Tcf7+ Tex). However, when the TME was remodeled by rVSV-LCMVG, OT-I T cells predominantly belonged to effector T-cell clusters, including early activated T cells, ISG+ Teff cells, Temra cells, and Il7r+ Tem cells (Fig. 5b). In addition, OT-I cells exhibited elevated expression of Runx3 following treatment with combined OVs. This indicates that these OT-I cells may persist in tumor tissues for an extended duration, thereby exerting antitumor effects (Fig. 5c). Consistent with our previous findings, the administration of rVSV-LCMVG ameliorated the exhaustion phenotype of tumor-specific T cells by promoting the differentiation of Tex into effector T cells.

Fig. 5

Transcriptional profiling of OT-I tumor-specific CD8+ T cells using scRNA-seq. a Uniform manifold approximation and projection (UMAP) visualization of the scRNA-seq clusters of OT-I tumor-specific CD8+ T cells from 6 samples in different groups. b Bar plot demonstrating percentages of cells in clusters as a fraction of total cells for each sample, related to the UMAP plot in a. c Dot plot representing the relative average expression of a subset of marker genes of OT-I tumor-specific CD8+ T cells in different groups. d Dot plot representing the relative average expression of a subset of marker genes across all clusters. e Single-cell transcription levels of representative genes illustrated in the UMAP plot from a. Transcription levels are color coded: gray, not expressed; blue, expressed

Next, we further validated the differentially expressed gene patterns of clusters that were significantly perturbed by rVSV-LCMVG treatment. G2M phase Tex expressed canonical exhaustion-related genes (Pdcd1, Ung, Mcm2, Ccnb2, and Top2a). Tcf7+ Tex was identified as the proliferative progenitor of terminally exhausted T cells. Nme1+ T cells expressing Nme1, Ccr7 and Npm1, were highly connected to Tcf7+ Tex cells. ISG+ effector T cells were further categorized based on Stat1, Isg15, Ifit3, and Gzmb expression. Il7r+ Tem highly expresses the signature of memory T cells (Il7r, Zfp36l2, Gpr183, Cxcr4, and Sell). Taken together, rVSV-LCMVG administration promotes tumor-specific exhausted (Tex) differentiation into effector (Teff) and memory (Tmem) cells with a significant decline in Tex proportion (Fig. 5d, e).

When analyzing virus-specific T cells, we observed that all samples could be classified into 13 distinct clusters. These clusters include early activated T cells, G2m phase Tex cells, Gzmb+ Teff cells, Il7r+ Tem cells, ISG+ Teff cells, ISG+ Bystander cells, Naïve-like T cells, Nme1+ T cells, S phase Tex cells, Regulator-like CD8 cells, Tcf7+ Tex cells, Xcl1+ T cells, and Terminally Tem cells. (Supplementary Fig. 6a, b). Combined with OVs, P14 virus-specific T cells differentiate from naive T cells into Teff and Tmem cells. In contrast, tumor-specific T cells undergo differentiation from Tex to Teff and Tmem cells (Supplementary Fig. 6c). This indicated that the adoptive transfer of tumor-specific T cells alone resulted in their differentiation into exhausted and disabled T cells upon tumor infiltration. However, when tumor-specific T cells were used in combination with the oncolytic virus rVSV-LCMVG, they effectively reversed exhaustion and improved their antitumor ability.

mRNA tumor vaccine combined with oncolytic virus improved the therapeutic effect of B16 tumor

Given the high cost and challenges associated with personalized CAR T or TCR-T treatment, the induction of specific T cells through mRNA vaccines holds the potential to establish a more transformative therapeutic strategy. In this study, we explored the possibility of indirectly inducing tumor-specific T cells to replace the direct reinfusion of T cells. Instead of transferring P14 cells, we employed LCMV-Armstrong virus to induce specific T cells that recognize the gp33 epitope. Subsequently, we detected a certain proportion of these specific T cells in the peripheral blood, spleen, and lymph nodes of the abdominal groove. (Supplementary Fig. 7a, b). In the B16-GP33 model, we utilized LCMV-Armstrong immune-induced specific T cells along with the rVSV-LCMVG oncolytic virus, this combined approach demonstrated a notable efficacy in inhibiting tumor growth. It is important to highlight that the treatment effect was significantly superior to that of using the oncolytic virus alone. Furthermore, the combination therapy also led to a noticeable extension in the survival rate of mice as compared to the monotherapy treatment involving immune LCMV-Armstrong. (Supplementary Fig. 7c–e). In the B16-GP33 tumor model, immune LCMV-Armstrong effectively generated GP33-specific T cells, which successfully suppressed tumor growth. Then, we applied the same treatment strategies in the B16-OVA model to validate the results. The findings demonstrated that GP33-specific T cells induced by LCMV-Armstrong, which solely targeted the antigens carried by rVSV-LCMVG OVs and did not recognize tumor-associated antigens, when combined with rVSV-LCMVG could effectively restrain the growth of B16-OVA tumors and significantly prolonged the survival of mice. (Supplementary Fig. 7f–h). These results suggested that in addition to tumor-specific T cells, in the combination therapy using virus-specific T cells could also achieve better therapeutic effects.

Next, we prepared the mRNA tumor vaccine which could express gp33 epitope and we verified the expression of gp33 at the cellular level by immunofluorescence using an earlier G2B1 antibody that specifically recognizes the gp33 epitope (Supplementary Fig. 8a). Mice were immunized intramuscularly with a dose of 10 μg per mouse. The specific T cells capable of recognizing the gp33 epitope were identified seven days after immunization. After an interval of 14 days since the initial dose, the same vaccine dose was administered to enhance the immune response. Subsequently, after five days, an increase in the number of specific T cells in the spleen was observed (Fig. 6a, b). IFN-γ enzyme-linked immunosorbent spot (ELISpot) test results showed T cells from the immunized mice spleen had a strong response when stimulation with gp33-41 antigenic peptides ex vivo (Fig. 6c, d).

Fig. 6

mRNA tumor vaccine combined with oncolytic virus improved the therapeutic effect. a Representative flow cytometry plot showing the fraction of gp33-specific T cells in the total CD8+ T cells gate from the spleen. b Quantification of the gp33-specific T cells. The proportion of cells in CD8 (left) and the total cell number (right). Each dot represents one mouse. c Representative well images of the IFN-γ ELISpot response of the gp33-specific T cells isolated from spleen in different groups. d Numbers of IFN-γ SFCs (spot-forming cells) of the gp33-specific T cells isolated from spleen were quantified after stimulation with GP33-41 peptide. e Schematic of B16-GP33 or B16-OVA tumor-bearing mice treated with gp33-mRNA and rVSV-LCMVG. f B16-GP33 tumor volumes are shown as mean values with SEM. Tumor response data derived from mice (n = 5) are shown. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001, based on two-way ANOVA with post hoc Holm–Sidak test. g Survival curves of C57BL/6 J mice from the experiment described in f are shown; survival analysis was conducted by log rank test. h B16-OVA tumor volumes are shown as mean values with SEM, (n = 5). *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001, based on two-way ANOVA with post hoc Holm–Sidak test. i Survival curves of C57BL/6 J mice from the experiment described in h are shown; survival analysis was conducted by log rank test

To determine the efficacy of mRNA vaccines combined with OVs in eliminating established tumors in vivo, we administered subcutaneous injections of 2 × 106 B16-GP33 cells per mouse. Once tumor formation was evident at the injection site, muscular immunization was conducted, with each mouse receiving a dose of 10 µg mRNA. Following a 7-day interval, oncolytic virus therapy was administered (Fig. 6e). The intratumoral injection of rVSV-LCMVG or mRNA vaccine monotherapy resulted in a moderate inhibition of tumor growth compared to the PBS group. Combination therapy with intratumoral or intravenous injection of rVSV-LCMVG in combination with mRNA vaccines largely improved the responsiveness of B16-GP33 tumors and prolonged the survival of these mice (Fig. 6f, g and Supplementary Fig. 8b). In combination therapy using mRNA vaccines, the therapeutic efficacy of the rVSV-LCMVG oncolytic virus was found to be superior in the intravenous injection group compared to the intratumoral administration. This may be attributed to the fact that, after immunization with mRNA vaccines, the intravenous administration of OVs stimulated a stronger systemic immune response than the intratumoral administration. As a result, there was an increase in the production of specific T cells and improved therapeutic outcomes. Even in the B16-OVA model, mRNA was only able to induce the generation of virus-specific T cells, also emphasizing that the combination therapy approach yielded better therapeutic results (Fig. 6h, i and Supplementary Fig. 8c). Our results highlight that while using mRNA to induce oncolytic virus-specific T cells or tumor-specific T cells, combined therapy with oncolytic virus would lead to a better therapeutic effect, especially when the mRNA-induced specific T cells could recognize both tumor and OVs, even if the oncolytic virus was administered intravenously, mice would gain a better therapeutic effect compared to monotherapy.