Hyper-Interferon Sensitive Influenza Induces Adaptive Immune Responses and Overcomes Resistance to Anti–PD-1 in Murine Non–Small Cell Lung Cancer

Lung cancer continues to be a leading cause of cancer-related death globally, with non–small cell lung cancer (NSCLC) representing the most common histologic subtype (1, 2). Although immune checkpoint inhibitors (ICI) targeting the programmed death 1/programmed cell death ligand 1 (PD-1/PD-L1) axis have demonstrated improved survival in a subset of patients with NSCLC, most patients do not respond and many develop resistance (3, 4). Despite numerous trials exploring various ICI combinations, treatment options for patients refractory to anti–PD-1/PD-L1 therapy remain limited.

The exclusion of T cells from the tumor or the presence of a dysfunctional T-cell compartment within the tumor microenvironment (TME) constitutes two central hallmarks of resistance to ICI (5). Favorable responses in patients with NSCLC to PD-1/PD-L1 blockade are often associated with high tumor mutational burden (TMB), increased baseline CD8⁺ T-cell infiltration into the tumor, and high PD-L1 expression in the TME (6–8). In contrast, impaired tumor antigen presentation and deficiencies in IFNγ signaling have been linked to immunotherapy resistance (9, 10). Moreover, in patients with KRAS-mutant NSCLC, serine/threonine kinase 11 (STK11/LKB1) inactivating mutations drive resistance to ICI through both transcriptional inhibition of the suppression of stimulator of interferon genes (STING) in tumor cells and CXCR2-mediated accumulation of suppressive neutrophils in the TME (11–13). Therapies designed to combat tumor-mediated immunosuppression and boost antitumor T-cell responses hold promise for patients with NSCLC refractory to immunotherapy.

In situ vaccination (ISV) has emerged as an antigen-agnostic treatment strategy capable of both conditioning the TME and enhancing antitumor T-cell responses (14). One promising approach to ISV involves the intratumoral (IT) delivery of genetically engineered viruses that exert anticancer activity through both direct tumor oncolysis and the induction of immune responses (15). Because engineered viruses are often pathogenic, it is necessary to attenuate their pathogenicity to limit systemic toxicity. However, reduced virulence may compromise tumor lysis and diminish host immune activation (15). Conversely, prolonged inflammation within the TME because of unconstrained viral replication can activate immunosuppressive feedback loops that impede antitumor immune responses (16, 17). Thus, an ideal therapeutic viral vaccine must retain the ability to induce well-controlled host antitumor immune responses while exhibiting restrained capacity for viral replication.

To develop a safe and effective virotherapy for NSCLC with controlled antitumoral immune activation, we recently engineered an attenuated influenza viral vaccine, termed hyper–IFN-sensitive (HIS) virus (18). Viruses typically employ various mechanisms to suppress host type 1 IFN (IFN-I) responses. To identify IFN-sensitive mutations in the entire viral genome at a single-nucleotide resolution, we developed a quantitative high-throughput genetic mutagenesis system coupled to next-generation sequencing (19, 20). Utilizing this approach, we identified multiple IFN-modulating functions across the influenza genome (18). We engineered HIS through rational incorporation of eight IFN-sensitive mutations in the influenza genome (18). Although HIS replicated efficiently in IFN-deficient Vero cells in vitro, intranasal HIS vaccination in immunocompetent mice induced an enhanced host IFN-I response, resulting in reduced viral replication (18). These unique characteristics of the HIS, namely, its hyper-induction of host IFN responses and attenuated replicative capacity, prompted us to assess its antitumor efficacy in NSCLC.

In this study, we have demonstrated the effectiveness of HIS ISV as a potent viral immunotherapy. Despite its reduced in vivo replication capacity compared with wild-type (WT) influenza, HIS ISV elicited a more robust and specific IFN-I response and displayed superior efficacy in multiple murine models of NSCLC. HIS ISV disrupted the immunosuppressive TME and induced an influx of polyfunctional effector Th1 CD4⁺ T and cytotoxic CD8⁺ T cells into the tumor. The addition of anti–PD-1 enhanced tumor inhibition by HIS ISV, leading to tumor growth in only half of the mice. This successful combination approach induced enduring systematic tumor-specific immunity.

The antitumor efficacy of HIS ISV as an approach to overcome resistance to immunotherapy in NSCLC was evaluated using syngeneic murine models of NSCLC with known driver mutations and varying mutational burdens. Treatments included HIS and anti–PD-1, as monotherapies or in combination, as well as corresponding vehicle or isotype controls. Detailed information on statistical analysis and the numbers of mice per group is provided in the statistical methods and figure legends. To ensure consistent tumor sizes, outlier tumors identified by Grubbs’ test were excluded before randomizing mice into treatment groups. In vivo experiments represent at least two replicates unless otherwise specified in figure legends. In vitro experiments were conducted with biological triplicates, and data are representative of at least two technical replicates.

The murine cell line 1940A was established from lung adenocarcinomas of conditional Kras^G12DTp53^+/−Lkb1^−/−Luc (KPL) FVB mice that express firefly luciferase as we described previously (21). WES analysis revealed that KPL cells lost the other allele of Tp53 upon in vitro culture and, therefore, bear a Kras^G12DTp53^−/−Lkb1^−/−Luc genotype (21). The Kras^G12D LKR-13 line, established from a lung adenocarcinoma tumor from a Kras^LA1 mouse, was generously provided in 2009 by Dr. Jonathan Kurie at the University of Texas MD Anderson Cancer Center, Houston, Texas (22). The KPL-3M cell line with increased mutational burden was generated by in vitro exposure a KPL cell line to N-methyl-N-nitrosourea as we previously described (21). The B16F10 melanoma cell line, along with Jak1^ko and Jak2^ko B16F10 cell lines, were generously provided in 2024 by Dr. Antoni Ribas at the University of California, Los Angeles (23). ΔS and ΔL B16F10 cell lines harboring homozygous deletion of 4C4 locus that is syntenic to human chromosome 9p21 locus, were graciously provided in 2024 by Dr. Scott Lowe at Memorial Sloan Kettering Cancer Center (24). MyC-Cap, Lewis lung carcinoma (LLC), HEK-293T (293T), and Madin-Darby Canine Kidney (MDCK) cell lines were purchased from the ATCC. 293T and MDCK cells were cultured in DMEM with 10% FBS (Corning). All other cell lines were maintained in complete culture media [RPMI-1640 with L-glutamine (Corning) supplemented with 10% FBS (Corning) and 1% penicillin/streptomycin (Fisher Scientific)] at 37°C in a humidified atmosphere of 5% CO₂ and used before five passages. Cell lines were tested annually for authentication (DNA IQ and Powerplex 1.2 systems, Promega) and routinely confirmed to be free of mycoplasma contamination (Mycoalert, Lonza).

The pLentiCRISPRv2 system was employed to generate Irf3^KO mouse cell lines using CRISPR/Cas9 technology. A specific single-guide RNA (sgRNA) sequence targeting IRF3 (5′-CCA​GTG​GTG​CCT​ACA​CCC​CG-3′) was chosen, alongside a nontargeting control sgRNA sequence (5′-GCT​GAG​ACG​GGC​CGG​AGC​AGA-3′). Lentiviral vectors carrying IRF3 sgRNA and control sgRNA were packaged in 293T cells along with PAX2 and MD2.G to produce lentiviruses; 1 mL of the virus was used to infect 1 million 1940A-KPL cells for 48 hours. The cells were washed twice with PBS and subjected to selection with 2-µg/mL puromycin for 1 week. Verification of Irf3^KO was performed via Western blotting using IRF-3 rabbit anti-mouse antibody (D83B9, Cell Signaling Technology).

Influenza A/WSN/33 (WSN) WT and HIS virus were generated using an eight-plasmid reverse genetics system, as we previously described (18). Briefly, 1.5 million 293T cells were transfected with 4 μg of plasmid DNA using Lipofectamine 2000 (Cat. #11668019, Invitrogen). Viruses were collected 72 hours posttransfection and subsequently amplified in A549 cells. Supernatants were collected, clarified of debris and stored at −80°C in aliquots. Viral titers were determined by the 50% tissue culture infectious dose (TCID50) titration assay utilizing MDCK cells, as previously described (18). The HIS virus harbors eight amino acid mutations, including three in PB2 (N9D, Q75H, and T76A), three in M1 (N36Y, R72Q, and S225T), and two in NS1 (R38A and K41A; ref. 18). Heat-inactivated (HI) HIS virus was generated by incubating HIS at 65°C for 1 hour.

Ten thousand MDCK cells were seeded per well in a 96-well plate. Virus stocks, diluted 100-fold, were added to the first column of the plate. Serial dilutions were then performed across the plate in half-log increments from the second to the last column. The TCID50 value was determined as the dilution of the virus required to induce cytopathic effects in 50% of the wells after a 72-hour incubation period (18).

KPL cells achieving 40% confluency in a 96-well plate were inoculated with either WT or HIS virus in 100 μL at the specified MOI. Mock infection, employing cell culture media, served as the control. At 2 hours postinfection, the cells were washed, and the media were replaced. At 36 hours postinfection, 10 μL of CCK8 solution (Cat. #96992, Sigma Aldrich) was added to each well and incubated for 4 hours at 37°C. The OD450 values for each well were determined by an iMark microplate reader.

Viral RNA was extracted using the QIAamp Viral RNA Mini Kit (Cat. #52904, Qiagen). Total cellular RNA was extracted from infected cells with the Purelink RNA Mini Kit (Cat. #12183025, Thermo Fisher) or Trizol (Cat. #15596026, Thermo Fisher) and reverse transcribed by Superscript III Reverse Transcriptase (Thermo Fisher) or Qscript (Quanta Biosciences) according to the product manual. Quantitative real-time PCR (LightCycler 480 II, Roche) was performed using Taq polymerase and SYBR Green Master Mix (Cat. #100029283, Thermo Fisher) using 1 μL of cDNA, including at least two technical replicates. For viral copy number, a standard curve ranging from 1 × 10³ to 1 × 10⁸ viral copies was used for quantification. For cellular gene expression, results were calculated using the 2^−∆∆CT method normalized by GAPDH or ACTIN. The primer sequences used to quantify RNAs are included in the Supplementary Table S1.

Female BALB/c (Shanghai SLAC Laboratory Animal Co.) and C57BL/6 mice (Jackson Laboratory) were utilized for safety studies. Both male and female FVB mice, 129-E (Charles River), and C57BL/6 mice (Jackson Laboratory) were used for efficacy studies. Female SCID mice (Jackson Laboratory) were utilized to assess the dependency of the HIS efficacy on T-cell responses. IFNAR^−/− mice on a C57BL/6 background were bred at University of California, Los Angeles (UCLA). All mice were housed in pathogen-free facilities and all animal experiments were approved by the Animal Research Committee at UCLA or Zhejiang University.

For syngeneic efficacy studies, tumor cells were implanted in 7 to 9-week-old mice subcutaneously at optimal doses as indicated in figure legends. Mice bearing ∼100 to 200 mm³ tumors were randomized and treated IT with mock control (DPBS) or 1 × 10⁵ TCID50 of HIS or WT. An equivalent amount of HI HIS virus served as an additional control. For combination immunotherapy studies, mice were also treated intraperitoneally with either 200 μg of anti–PD-1 (Clone RMP1-14, BioXcell) or IgG2a isotype control (Clone 2A3, BioXcell) antibodies. For tumor IFN inhibition studies, mice were also treated IT with 50 μg of anti-IFNAR1 (Clone MAR1-5A3, BioXcell) or IgG1 isotype control (Clone MOPC-21, BioXcell) antibodies. The timing and frequency of treatments are specified in the corresponding figure legends. Tumor length and width were measured by caliper and the volume calculated by the equation: 0.4 × length × width². Tumors were weighed at the time of euthanasia. For survival studies, mice were sacrificed when tumor volumes reached 1,500 mm³. For bioluminescence studies, images were obtained with an IVIS Spectrum imager 10 minutes after intraperitoneal injection of 150 mg/kg of D-luciferin (Cat. #L9504, Sigma Aldrich).

For safety studies, BALB/c mice were anesthetized using isoflurane (Cat. #1182098, Henry Schein) and subsequently inoculated intranasally with 30 μL of HIS or WT virus at doses indicated in figure legends. Body weights were monitored daily. Complete blood count (CBC), alanine transaminase (ALT), and creatinine (Cr) analyses were performed by Servicebio. Blood samples were collected via retro-orbital bleeding. For CBC analysis, blood was collected in EDTA tubes to prevent clotting, whereas samples for Cr and ALT analyses were collected in standard tubes to allow clotting and enable serum collection. CBC was measured using the POCT-DP-H10 automated analyzer, and Cr and ALT levels were quantified using an ELISA-based assay developed by Servicebio.

To quantify viral titer, various tissues, including lung, brain, liver, kidney, and spleen were collected after euthanasia and homogenized for the TCID50 assay utilizing MDCK cells, as previously described (18). To quantify viral genome copy number from the lung, RNA was isolated utilizing TRIzol (Cat. #15596026, Thermo Fisher), as previously described (18). To quantify viral growth and gene expression in established tumors, subcutaneous tumors were harvested and homogenized in TRIzol followed by RNA isolation. Viral copy number and gene expression were quantified by real-time PCR as described above.

Single-cell suspension of murine tumors and spleens were prepared as previously described (21, 25). Murine tumors were dissected into small fragments and treated with 1 mg/mL of Collagenase A (Cat. #10103578001, Sigma Aldrich), incubated for 40 minutes at 37°C in a water bath with intermittent shaking every 10 minutes. After incubation, 10 mL of fresh complete culture media was added, and the specimens were strained through a 70 μm filter and centrifuged at 500 g for 5 minutes at 4°C. The resultant cell pellets were resuspended in 1 mL of eBioscience red blood cell lysis buffer (Cat. #00-433-57, Thermo Fisher) and kept on ice for 5 minutes. Thereafter, 10 mL of fresh complete culture media was added. Cells were then filtered through a 100 μm filter, centrifuged, rinsed with FACS buffer (PBS supplemented with 2% FBS), and enumerated. The spleen and lymph node samples were pressed through a 100-μm filter using the blunt end of a 3-mL syringe and centrifuged at 500 g for 5 minutes at 4°C. Red blood cells were lysed as described above, and the cells were washed with PBS, and counted.

For surface staining, single-cell suspension of tumors, spleens and lymph nodes were incubated with antibody cocktail for 30 minutes on ice, followed by two washes with FACS buffer. For intracellular staining of FOXP3 and Ki67, cells were fixed and permeabilized with eBioscience Foxp3/Transcription Factor staining buffer set (Cat. #00-5523-00, Thermo Fisher) and stained as previously described (25). Cytokine secretions were examined post in vitro stimulation with eBioscience Cell Stimulation Cocktail (Cat. #00-4970-03, Thermo Fisher) containing phorbol-12-myristate 13-acetate (PMA), ionomycin, brefeldin A and monensin for 4 hours, followed by intracellular staining of IFNγ and TNFα. Data acquisition was performed on Attune NxT cytometer (Thermo Fisher), and data were analyzed using FlowJo software (TreeStar). The list of the flow antibodies is provided in Supplementary Table S2. Zombie-NIR LIVE/DEAD staining was used to eliminate dead cells.

Cell sorting was performed to isolate CD45⁺, CD45⁻ or CD3⁺ cells for real-time PCR and single‐cell RNA sequencing (scRNA-seq) analyses (Supplementary Table S1). Single-cell suspensions from tumor were stained with Zombie-NIR LIVE/DEAD stain and antibody cocktail for 30 minutes on ice (Supplementary Table S2). Cell sorting was performed using FACSAria I (II) High-Speed Cell Sorter with 100-μm nozzle.

RNA-sequencing (RNA-seq) libraries were prepared using NEBNext Ultra II RNA Library Prep Kit for Illumina. Multiplex sequencing was performed by 50-bp single-end read with NovaSeq machine at UCLA Technology Center for Genomics and Bioinformatics (TCGB) core. Raw sequencing reads were aligned to the mouse genome assembly (mm10) using Bowtie2 (26). Results were quantified by reads per million total reads (RPM). Differential expression analysis was performed with edgeR (27). We implemented quasi-likelihood F-tests on a generalized linear model for gene expression. Gene ontology enrichment analysis was performed through metascape (28). Genes related to certain cellular pathways were extracted from MsigDB (29).

The scRNA-seq analysis was performed using 10X Genomics 3′ (v3.1) protocol. Live CD45⁺ cells were sorted as described above. Single cell encapsulation, library construction, and sequencing were performed at the UCLA Technology Center for Genomics and Bioinformatics core facility. Libraries were prepared according to the Chromium Single Cell 3′ Reagent Kit v3 (10X Genomics) user guide and sequencing was performed on the NovaSeq 6000 (Illumina). 10X Cell Ranger (Version 3.1.0) software was used to align reads to the mouse genome mm10 and to generate gene count matrices using the Ensemble transcript reference (Version 93). Poor quality and doublet cells were removed based on either detected genes (<200 or >9,000 genes), total UMIs (<1,000 or >100,000 counts), or percentage of reads aligned to mitochondrial genome (>20%). A total of 18,967 cells (from three treatment arms) passed quality control for downstream analysis. Scanpy (Version 1.6.0) was used to normalize, scale, select variable features and perform differential gene expression analysis. Top 4,000 variable genes were selected for dimension reduction. Principal component (PC) analysis was performed on the scaled data, and the first 50 PCs were selected for Uniform Manifold Approximation and Projection for Dimension Reduction visualization. Cell clusters were identified using the Louvain algorithm; t test approach was utilized to quantify differential gene expression associated with treatments (sp.stats.ttest_ind function).

Experiments were conducted at least twice unless otherwise indicated, and results from one representative experiment are presented. Statistical analyses were carried out using Prism (version 9) software (GraphPad) unless otherwise noted, with P < 0.05 considered statistically significant. Results are expressed as mean ± SEM unless otherwise stated. In in vivo safety and efficacy studies, flow phenotyping, gene expression, and in vitro gene expression and viral studies, differences between groups were assessed using either two-tailed unpaired t-tests or one-way ANOVA with Bonferroni multiple-comparisons. For mouse survival studies, differences between groups were evaluated by log-rank (Mantel–Cox) test.

All data associated with this study are present in the main text or the supplementary materials. Gene expression data (accession number PRJNA1016387) are available at https://www.ncbi.nlm.nih.gov/bioproject/PRJNA1016387. Cell lines utilized in this study have not been deposited to any repositories. All materials are available upon request.

We previously demonstrated that HIS virus induced a hyperactivated IFN response as compared with WT virus in human cells (18). To validate in murine cells, four murine lung cancer cell lines, namely, LLC, Kras-mutant lung adenocarcinoma (LKR13), and two cell lines established from lung tumors of conditional Kras^G12DTp53^+/−Lkb1^−/−Luc (1940A-KPL) and Kras^G12DTp53^−/−Luc (1969B-KP) mice, were treated with HIS and WT virus. Across all cell lines, HIS induced a stronger IFN response, as determined by Ifnb1 gene expression, but demonstrated reduced replication compared with that of WT virus (Fig. 1A and B). The robust upregulation of Ifnb1 expression in KPL cells induced by HIS suggests that HIS can overcome the LKB1 deficiency-mediated STING inhibition and IFN suppression (Fig. 1B; ref. 13). Temporal quantification of viral replication in KPL cells indicated a ∼2-log attenuation of HIS compared with the WT virus, consistent with our previous results in A549 human lung carcinoma (Fig. 1C; ref. 18). The cytotoxicity of HIS in KPL cells was impaired compared with the WT virus (Fig. 1D; Supplementary Fig. S1A), which we hypothesize is because of attenuated viral replication.

To assess HIS-induced global gene expression changes in cancer cells, we conducted bulk RNA sequencing of KPL cells 24 hours post infection. HIS induced the upregulation of 233 genes (fold change > 2 and P < 0.001), with the majority of these genes (65.2%) being IFN-I inducible genes (Fig. 1E and F). Gene set enrichment analysis (GSEA) revealed an enrichment of IFNα responses, but not inflammatory responses, suggesting that HIS specifically activates the IFN pathway (Fig. 1G). Heat inactivation of HIS abolished its capacity to induce an IFN response, as evidenced by decreased expression of Ifnb1 and Mx1, indicating that successful viral entry and replication is required for IFN induction (Fig. 1H). These results are consistent with our prior report demonstrating that HIS induces the IFNB1 gene in A549 cells approximately 5- to 10-fold more than WT influenza or mock infection (18).

The safety profile of HIS virotherapy was assessed in C57BL/6 mice. Intranasal inoculation of mice with 1 × 10⁴ TCID50 of WT virus resulted in 50% lethality yet no mortality was observed with HIS virotherapy (Fig. 2A). Both WT and HIS viral replications were limited to the lung following intranasal inoculation, with no evidence of dissemination to distant organs, including brain, liver, kidney and spleen (Fig. 2B; Supplementary Fig. S1B). Within the lung, a ∼3-log reduction in viral titers was observed on day 3 (D3) in mice treated with HIS compared with WT virus (Fig. 2B). On D8 post viral inoculation, WT influenza was still detectable in the lung, whereas HIS was below the level of detection, indicative of improved HIS clearance (Fig. 2B). To determine whether attenuated HIS replication stems from increased host IFN responses, viral replication was evaluated in IFNAR^−/− mice. HIS and WT viruses exhibited similar replication capacity in IFNAR^−/− mice, suggesting that hyper-activated host IFN signaling in immunocompetent hosts likely restrains HIS replication (Fig. 2C).

Systemic toxicities were evaluated with repeated subcutaneous inoculations of HIS at higher viral titers in C57BL/6 mice (Supplementary Fig. S1C). HIS replication was limited to subcutaneous tissue and was detected at 2 and 24 hours postinoculation at the highest dose of 1 × 10⁵ TCID50 at the infection site (Supplementary Fig. S1D). Increasing doses of HIS (1 × 10⁴ and 1 × 10⁵ TCID50) did not result in weight loss or systemic dissemination (Supplementary Fig. S1E and S1F). Neutrophil and lymphocyte counts, as well as liver enzymes and kidney function, remained comparable with the control (Supplementary Fig. S1G and S1H).

HIS ISV was evaluated in multiple syngeneic murine models of NSCLC resistant to ICI. IT treatment of mice bearing 1940A-KPL tumors with HIS, but not WT virus (1 × 10⁵ TCID50, every-other-day for four doses), resulted in the inhibition of tumor growth (Fig. 2D; Supplementary Fig. S2A). Similarly, HIS ISV, but not WT influenza, demonstrated antitumor efficacy in two additional anti–PD-1 resistant murine NSCLC models, namely 1969B-KP and LLC (Fig. 2E and F; Supplementary Fig. S2B and S2C). Intravenous administration of HIS did not inhibit tumor growth, indicating that the antitumor efficacy of HIS ISV likely relies on the delivery of the virus into the tumor (Supplementary Fig. S2D). Furthermore, heat inactivation of HIS abolished its antitumor efficacy, suggesting that viral entry is necessary for the induction of antitumor responses (Fig. 2D and E; Supplementary Fig. S2A and S2B). HIS ISV resulted in a modest but statistically significant enhancement in the overall survival of mice harboring 1940A-KPL or 1969B-KP tumors and a substantial improvement in overall survival of mice bearing LLC tumors (Fig. 2G).

We hypothesized that the superior efficacy of HIS compared with WT virus stems from its enhanced capacity to induce host IFN-I responses. Viral titers and the expression of selected immune genes were determined 36 hours post IT injection (1 × 10⁵ TCID50) of HIS or WT virus in KPL tumors. WT viral copy number in the whole tumor digest was significantly higher compared with HIS, as determined by qPCR (Fig. 3A). Although the WT virus exhibited robust replication in both tumor (CD45⁻) and immune (CD45⁺) cells, HIS gene expression was predominantly restrained to tumor cells (Fig. 3B). HIS promoted a higher Ifnb1 induction in the TME compared with the WT virus (Fig. 3C). In contrast, WT virus induced a stronger activation of inflammatory genes compared with HIS, including Il1b (WT 4-fold vs. HIS 2-fold) and Ccl5 (WT 20-fold vs. HIS 8-fold; Fig. 3C). HIS-induced Ifnb1 expression was restricted to the CD45⁻ tumor cells, consistent with the predominant viral replication in this compartment (Fig. 3D). Although host IFN-I responses are critical for the initiation of antitumor immunity, prolonged IFN-I signaling can lead to activation of immunosuppressive feedback loops that drive immune resistance (16, 17). Evaluation of gene expression within the tumor cells on D2 and D5 revealed that, in contrast to sustained induction of the inflammatory gene Ccl5 by WT virus, the induction of Ifnb1 following HIS was transient (Fig. 3E). To assess whether the efficacy of HIS is dependent on the induction of IFN-I responses within the TME, mice bearing KPL tumors were treated with IFNAR1 blocking antibody in combination with HIS via IT injection. IFNAR1 antibody completely abolished the efficacy of HIS (Fig. 3F), strongly suggesting that HIS-induced host IFN-I responses are essential for its antitumor efficacy.

To evaluate the dependency of HIS-induced antitumor responses on tumor-specific IFN induction, we utilized CRISPR/cas9 technology to knockout (KO) the IFN regulatory factor 3 (Irf3) gene in 1940A-KPL cell lines. IRF3 is a key transcription factor regulating the expression of type I IFN genes, including IFNα and IFNβ (30). Immunoblot analysis confirmed a significant reduction of IRF3 protein in Irf3^KO 1940A-KPL cell lines, though complete ablation was not achieved (Supplementary Fig. S3A). In vitro proliferation assays revealed similar growth kinetics between scramble and Irf3^KO 1940A-KPL cells (Supplementary Fig. S3B). No discernible differences in in vitro replication of HIS were observed between scramble and Irf3^KO 1940A-KPL cells (Supplementary Fig. S3C). However, HIS infection led to diminished Ifnb1 production in Irf3^KO cells, along with heightened cell cytotoxicity (Supplementary Fig. S3D and S3E). Despite the increased in vitro cytotoxicity of HIS in Irf3^KO cells, HIS ISV demonstrated ineffectiveness in reducing tumor growth in vivo in immunocompetent mice bearing Irf3^KO 1940A-KPL tumors (Fig. 3G), suggesting that IFN is needed for antitumoral responses of HIS ISV.

To further evaluate the impact of HIS on tumors with impaired IFN signaling pathways, we utilized syngeneic B16F10 murine melanoma cell lines engineered to mimic specific genetic alterations associated with immune resistance in human cancers. These mutations included the following: (i) knockout of Jak1 or Jak2, resulting in deficiency of all three types of IFN responses or only IFN-II signaling, respectively and (ii) small (ΔS) and large (ΔL) homozygous deletion of 4C4 locus, corresponding to the human chromosome 9p21 locus (24). Homozygous deletions of chromosome 9p21.3 occur in approximately 15% of all human tumors and encompass mutations targeting CDKN2A/B alone (9p small or ΔS) or larger deletions including the entire IFN-I cluster (9p large or ΔL; refs. 24, 31). HIS induced Ifnb1 expression in B16F10-ΔS and B16F10-Jak2^KO cell lines but not B16F10-ΔL cells (lacking Ifnb1 gene) or B16F10-Jak1^KO cells (lacking all IFN signaling; Supplementary Fig. S3F). HIS resulted in the upregulation of IFN downstream genes Mx1 and Cxcl10 in B16F10-ΔL cell lines, suggesting activation through an IFN-I-independent pathway (Supplementary Fig. S3G and S3H).

IFN-I signaling can facilitate cancer immunosurveillance by promoting DC cross-presentation and the activation of host adaptive immune responses (32, 33). We utilized flow cytometry and scRNA-seq to profile the immune composition of the KPL tumors following HIS ISV (Supplementary Fig. S4A–S4C). Studies were performed on D16, 10 days after the initiation of virotherapy, to allow sufficient time for the induction of host adaptive immune responses. scRNA-seq analysis of the TME revealed three immune clusters corresponding to neutrophil, monocyte/macrophage/DC, and NK/T cells (Supplementary Fig. S4D–S4F). HIS virotherapy resulted in the reduction of neutrophils, the most abundant immune population within the 1940A-KPL tumors, consistent with prior reports showing predominance of neutrophils in the TME of LKB1-deficient NSCLC (Fig. 4A; and Supplementary Fig. S4F; refs. 11, 12).

Recent studies have shown heterogeneity within tumor-infiltrating neutrophils, which is conserved in both murine and human lung cancers (34). Four neutrophil sub-clusters (C1–C4) were identified within 1940A-KPL tumors by scRNA-seq and annotated based on reference murine gene sets (C1—mN4, C2—mN1/3, C3—mN2, and C4—mN5; Fig. 4B and C; ref. 34). Both HIS and WT led to a relative reduction of the C1 cluster, corresponding to the tumor-promoting mN4 phenotype, and a concurrent enrichment of C3 (corresponding to mN2) with a transcriptional signature of IFN-I response genes, including Gpb2, Ift3, Ift1, Stat1, and Isg15, as well as MHC-related genes, such as B2m (Fig. 4C and D). HIS ISV induced a greater upregulation of IFN-stimulated pathways in tumor-infiltrating neutrophils compared with both control and WT (Fig. 4E and F). Although no difference was observed in the DC compartment of the TME by flow cytometry, HIS vaccination led to both the enrichment of monocytes and macrophages within the TME (Supplementary Fig. S4G). HIS ISV also resulted in increased expression of IFN-stimulated genes (ISG) within macrophages, as determined by scRNA-seq (Supplementary Fig. S4H). Increased PD-L1 expression was also observed on neutrophils, macrophages, and cDC2s, but not cDC1s following WT and HIS ISV. HIS ISV resulted in the highest magnitude of PD-L1 expression compared with control and WT virus (Supplementary Fig. S4I), consistent with the superior capacity of HIS to induce IFN signaling within the TME.

HIS ISV promoted an enrichment of CD8⁺ and FOXP3⁻CD4⁺ T effector (CD62L⁻) lymphocytes, as well as NK cells in the TME, compared with both WT and vehicle control (Fig. 4G). The composition of the regulatory T cells (Treg) remained unchanged (Supplementary Fig. S4J). The increased NK cells and CD8⁺ and FOXP3⁻CD4⁺ effector T cells in the tumor partly resulted from their enhanced proliferation, as indicated by increased Ki-67 expression in these populations (Fig. 4H). The highest CD8⁺/Treg ratio was observed in tumors following treatment with HIS (HIS ∼ 2.6 vs. WT ∼ 1.5 vs. Control ∼ 0.6; Fig. 4I). HIS ISV resulted in a higher induction of IFNγ secretion from FOXP3⁻CD4⁺ and CD8⁺ effector T cells compared with both WT and control, suggestive of Th1 polarization and improved CD8⁺ cytotoxic function (Fig. 4J). To evaluate the functional importance of host T-cell responses following HIS virotherapy, we assessed its efficacy in immunodeficient SCID mice lacking mature B and T cells (Fig. 4K). The efficacy of HIS ISV was abolished in KPL tumor-bearing SCID mice, highlighting the dependency of the antitumor efficacy of HIS on host adaptive immune responses. Collectively, our data indicate that HIS-mediated induction of IFN-I signaling within the TME serves as the primary mediator to stimulate host antitumor adaptive immune responses.

The induction of IFN-I responses within the TME and T-cell infiltration into the tumor by HIS ISV prompted us to assess its potential in overcoming resistance to anti–PD-1 immunotherapy. Immunocompetent mice bearing 1940A-KPL tumors were treated with IT HIS as monotherapy or in combination with anti–PD-1. Although anti–PD-1 monotherapy provided no efficacy, the combination therapy significantly enhanced the effectiveness of HIS ISV (Fig. 5A). Combination therapy improved overall survival compared with control or monotherapy, but all tumor-bearing mice ultimately succumbed to the disease (Fig. 5B). We also evaluated WT ISV in combination with anti–PD-1 in 1940A-KPL tumor-bearing mice. Although WT ISV and anti–PD-1 monotherapy lacked antitumor efficacy, their combination led to a modest antitumor response (Supplementary Fig. S5A), which was less potent than the combination of HIS ISV and anti–PD-1. These results align with HIS ISV’s enhanced capacity to induce IFN-I signaling and promote the infiltration of Th1 and cytotoxic effector CD8⁺ T cells into the TME compared with WT virus. Given that oncogene-driven murine models of NSCLC, including 1940A-KPL, have a limited number of protein-altering mutations, the paucity of tumor antigens might limit immunotherapy efficacy. We recently established a murine NSCLC model, KPL-3M, with LKB1 deficiency and an increased TMB (TMB = 7.2 mutations/Mb), aiming to better simulate human NSCLC (21). HIS ISV and systemic anti–PD-1 monotherapy exhibited modest antitumor efficacy individually; however, their combination led to robust tumor regression (Fig. 5C), resulting in the eradication of approximately 50% to 60% of KPL-3M tumors and improved long-term survival (Fig. 5D).

To evaluate the cellular mediators of antitumor responses induced by IT HIS combined with anti–PD-1, we examined the composition and function of immune cells in the KPL-3M TME, tumor-draining lymph node (TdLN) and spleen by flow cytometry. Both HIS monotherapy and combination therapy with anti–PD-1 resulted in a near two-fold reduction of neutrophils within the TME compared with anti–PD-1 or vehicle control (Fig. 6A). A statistically significant enrichment of macrophages (>50%) and inflammatory monocyte-derived cells (MDC; ∼2-fold) was observed within the TME following HIS and anti–PD-1 combination therapy (Fig. 6B). Only combination therapy led to a statistically significant but modest (∼50%) increase in cDCs within the KPL-3M tumors (Fig. 6C; Supplementary Fig. S5B), but both HIS monotherapy and combination therapy resulted in an approximately two-fold enrichment of both cDC1 and cDC2 in TdLNs (Fig. 6D). An expansion of antigen-experienced CD44⁺ CD8⁺ and CD44⁺ CD4⁺ T cells was also observed following HIS monotherapy or combination therapy in TdLNs (Fig. 6E), suggesting potential activation of endogenous T cells by antigen-experienced cDCs.

HIS monotherapy and combination therapy induced increased accumulation of CD4⁺ T cells and Th1 polarization in KPL-3M tumors, with the highest magnitude observed following combination therapy (Fig. 6F and G). Within the CD4⁺ compartment, there was enhanced proliferation of conventional FOXP3⁻CD4⁺ T cells but not Tregs following HIS monotherapy and combination therapy, leading to a two-fold increase in the ratio of conventional CD4⁺ T cells to Tregs (Fig. 6H; Supplementary Fig. S5C). Treatment with HIS as monotherapy or in combination with anti–PD-1 resulted in an approximately 2- to 3-fold increase in polyfunctional Th1 and CD8⁺ T cells producing both IFNγ and TNFα within the TME, with the highest magnitude observed following combination therapy (Fig. 6I and J). In accordance with these observations, the highest magnitude of PD-L1 expression, which is downstream to IFNγ signaling, was observed on tumor cells, macrophages and MDCs following combination therapy with HIS and anti–PD-1 (Fig. 6K; Supplementary Fig. S5D). Notably, the expression of this inhibitory protein was not altered on tumor-infiltrating cDCs following therapy (Supplementary Fig. S5E).

Analysis of the spleens from KPL-3M tumor-bearing mice by flow cytometry demonstrated that HIS monotherapy and combination therapy lead to systemic enrichment of CD44⁺ CD62L⁻ effector memory (EM) CD4⁺ and CD44⁺ CD62L⁺ central memory (CM) CD8⁺ T cells (Fig. 7A). Only the combination therapy with HIS and anti–PD-1 resulted in increased IFNγ expression by splenic CD8⁺ T cells, suggesting potential augmentation of systemic CD8⁺ T-cell responses (Fig. 7B). To evaluate whether combination therapy generates systemic tumor-specific immune responses, mice cured of KPL-3M tumors following combination therapy and age-matched naïve mice were inoculated with KPL-3M or the control MyC-CaP tumor cells from the same mouse genetic background. After an initial tumor growth, cured mice rejected all KPL-3M tumors but succumbed to MyC-CaP tumors (Fig. 7C–E). In contrast, naive FVB mice succumbed to both KPL-3M and MyC-CaP tumors (Fig. 7C–E). These data demonstrate the generation of systemic tumor-specific immune memory following the successful eradication of KPL-3M tumors by HIS and anti–PD-1 combination therapy.

The failure of responses to immunotherapy in patients with NSCLC is partly driven by the immunosuppressive milieu of the TME, which impedes both T-cell activation and effector responses. In this study, we demonstrate that HIS ISV induced robust but transient local IFN-I responses within the TME, overcoming tumor-mediated immune suppression and resulting in activation of antitumor adaptive immune responses. HIS ISV potentiates the efficacy of anti–PD-1 therapy in immune-resistant murine LKB1-deficient NSCLC, inducing systemic tumor-specific immune memory.

Oncolytic viruses are engineered to preferentially infect, replicate, and lyse tumor cells (15). Oncolysis can also induce adaptive immune responses against in situ tumor antigens released into the TME. For most oncolytic viruses, both antitumor efficacy and induction of host immune responses are strongly linked to viral replication and tumor cell lysis, often resulting in a narrow therapeutic index (15). In this study, we demonstrate that the inoculation of HIS, rationally engineered with eight IFN-sensitive mutations in the influenza A genome (18), enhances local IFN-I signaling in hosts. This leads to severely attenuated HIS replication and the effective decoupling of immune activation from viral replication, yielding a highly immunogenic cancer vaccine with an excellent safety profile. In multiple murine models of lung cancer, we show that HIS ISV, at a relatively low dose (1 × 10⁵ TCID), induces approximately 2- to 5-fold higher Ifnb1 expression and exhibits superior antitumor efficacy compared with WT influenza, despite reduced viral replication in the tumor. This suggests that the immune modulatory function of HIS virus, rather than its oncolytic function, inhibits tumor growth. Utilizing a similar functional genomic screening and engineering approach, future studies can further enhance the efficacy of HIS by eliminating additional immune suppressive genes or encoding the virus with immunostimulatory cytokines.

The cGAS-STING pathway in mammalian cells detects cytosolic dsDNA as danger-associated molecules and activates host IFN signaling. A recent analysis of multiple cohorts of patients with NSCLC, including 499 patients from the TCGA database, revealed that reduced expression of STING is associated with decreased survival (35). In LKB1-deficient KRAS-mutant NSCLCs, epigenetic silencing of TMEM173, the gene encoding STING, through promoter methylation has been shown to result in defective IFN-I signaling and immune evasion (13). Our in vitro and in vivo results illustrate that HIS therapy can overcome the suppression of IFN-I signaling in LKB1-deficient Kras-mutant NSCLCs to boost host antitumor immune responses and sensitize resistant tumors to ICI. The capacity of HIS, an RNA virus, to overcome IFN-I deficiency in LKB1-deficient cancers with suppressed STING signaling is likely mediated through the orthogonal RNA-sensing RIG-I/MDA pathway in the cytoplasm, as well as the TLR7/8/MyD88 and TLR3/TRIF pathways in the endosome (18, 36, 37). These findings likely have broad translational applications for overcoming immunotherapy resistance in other cancers that commonly exhibit a deficiency in the cGAS-STING and IFN signaling pathways, including hepatocellular, gastric, and colorectal cancers, as well as melanoma (38–41).

Our results suggest that the antitumor efficacy of HIS depends on the induction of IFN-I signaling in the TME. IFN-Is are essential mediators of cancer immunosurveillance and responses to immunotherapy, acting both directly on malignant cells and through the activation of adaptive immune responses (24, 42–46). However, the function of IFN-I signaling within the TME is complex and dependent on the signal's strength and duration. Although acute IFN-I responses are cytotoxic, chronic stimulation of cancer cells with low doses of IFN-I promotes tumor survival through upregulation of a subset of ISGs that belong to the IFN-related DNA damage resistance signature (refs. 42, 47, 48). Moreover, prolonged IFN-I signaling within the TME may lead to the expression of immunosuppressive cytokines, accumulation of Tregs and suppressive myeloid cells, as well as CD8⁺ T-cell exhaustion, ultimately resulting in resistance to ICI (49–52). Here, we show that HIS ISV promoted a robust induction of IFN-I responses locally, with minimal stimulation of suppressive inflammatory responses. Although IFN-I induction was more pronounced following HIS ISV compared with WT influenza, its duration was shorter—approximately 2 versus 5 days for WT ISV—presumably because of the attenuated and transient replication of HIS. Future studies will assess which specific cells within the TME are required for HIS-mediated production of IFN-I responses.

In both 1940A-KPL and KPL-3M murine models of NSCLC, HIS ISV ameliorated the immunosuppressive TME and resulted in an enrichment of effector Th1 CD4⁺ and cytotoxic CD8⁺ T lymphocytes within the tumors. It is well-established that IFN-I promotes antigen uptake and presentation by DCs, serving as the third signal to facilitate clonal expansion, differentiation, and memory formation of CD8⁺ T cells (33, 43, 53). We observed enhanced cDC migration to TdLNs, accompanied by a concurrent increase in antigen-experienced T cells following HIS ISV, suggesting T-cell priming and activation at TdLNs (Fig. 6D–E). Studies have demonstrated that TdLNs harbor a reservoir of stem-like tumor-specific T cells crucial for mediating responses to ICIs (54). The addition of anti–PD-1 to HIS ISV resulted in an enrichment of polyfunctional effector T cells within the TME and an increase in the production of IFNγ by CD8⁺ T cells. These results align with the enhanced antitumor efficacy of the combination therapy, leading to the eradication of KPL-3M tumors in 50% of mice. Although the absence of dominant antigens, such as OVA, in our murine models of NSCLC prevents a direct assessment of antigen-specific T-cell responses, the emergence of tumor-specific immune memory in mice cured of cancer following combination therapy with HIS ISV and anti–PD-1 indicates the establishment of functional systemic tumor-specific T-cell responses.

Although the diversity of murine models of NSCLC in this report encompasses various driver mutations and distinct TME profiles, as well as varying TMB and intertumoral clonal heterogeneity, one limitation of our studies is the use of syngeneic models. These models utilizing subcutaneous tumor implantation do not recapitulate human tumorigenesis, including the dynamic function of host immunosurveillance (21). Orthotopic models of NSCLC, established through transthoracic injection of tumors, better simulate the TME of human disease but share the same limitations, regarding the lack of a latency period.

In summary, our study introduces an attenuated HIS influenza virus with an outstanding safety and efficacy profile. HIS ISV overcame tumor-mediated immune suppression through the induction of robust local IFN-I responses. Moreover, HIS ISV potentiated the efficacy of anti–PD-1 in eliminating resistant murine NSCLC, resulting in improved overall survival and enduring systemic tumor-specific immunity. These findings lay the foundation for potential translation of this “off-the-shelf” vaccine for patients with immune-resistant NSCLC.

R. Sun reports a patent for US20200129612A1 issued. No disclosures were reported by the other authors.

Y. Du: Conceptualization, data curation, formal analysis, investigation, writing-original draft, writing–review and editing. R. Salehi-Rad: Conceptualization, data curation, formal analysis, validation, investigation, writing–original draft, writing–review and editing. T.-h. Zhang: Data curation, formal analysis, investigation. W.P. Crosson: Data curation, formal analysis, investigation. J. Abascal: Data curation, formal analysis, investigation. D. Chen: Data curation, formal analysis. Y. Shi: Data curation, formal analysis, investigation. H. Jiang: Data curation, formal analysis, investigation. Y.-W. Tseng: Data curation, formal analysis, investigation. X. Ma: Data curation, formal analysis. M. Hong: Investigation. S. Wang: Data curation, investigation. X. Wang: Data curation, investigation. K. Tang: Data curation, investigation. S. Hu: Investigation. Y. Li: Investigation. S. Ni: Data curation, Investigation. Y. Cai: Data curation, investigation, writing–review and editing. S. Tappuni: Data curation, writing–review and editing. Y. Shen: Data curation, investigation. B. Liu: Conceptualization, data curation, formal analysis, supervision, funding acquisition, writing–original draft, Project administration, writing–review and editing. R. Sun: Conceptualization, resources, formal analysis, supervision, funding acquisition, writing–original draft, project administration, writing–review and editing.

The authors express gratitude to Dr. Steven Dubinett for insightful discussions and valuable suggestions. The authors thank Lauren Winter and Elvira Liclican for administrative support. We would like to acknowledge the UCLA Technology Center for Genomics and Bioinformatics for their assistance with scRNA-seq, and the UCLA Jonsson Comprehensive Cancer Center (JCCC) Flow Cytometry Core for their support in cell sorting. Additionally, we extend our thanks to the Core Facilities at Zhejiang University School of Medicine for their technical and facility support. This study was supported in part by funding from Leading Innovative and Entrepreneur Team Introduction Program of Zhejiang (Y. Du, 2022R01002), National Natural Science Foundation of China (Y. Du, 82272300, 82102893), Career Development Award-2 from the Department of Veterans Affairs, Biomedical Laboratory Research and Development Service (R. Salehi-Rad, 1IK2BX006194-01), National Heart Lung and Blood Institute (R. Salehi-Rad, T32-HL072752), and Lung Cancer Foundation of America and International Lung Cancer Foundation Young Investigator Award (R. Salehi-Rad). The UCLA Jonsson Comprehensive Cancer Center Research Flow Cytometry Core is supported by the National Institutes of Health P30 CA-16042 and 5P30 AI-28697 and National Institutes of Health (Y. Shi, R. Sun, R01AI143287).