VAX014 Activates Tumor-Intrinsic STING and RIG-I to Promote the Development of Antitumor Immunity

rBMC production methods

Flow cytometry and in vivo antibody reagents

Western blot reagents

Full Western blot images

Supplemental Figure S1. Bifurcation of patients into high or low expression cohorts based on STING1 (STING) and RIG-I mRNA expression. A) Percentage of patients in all TCGA Pan-Cancer Atlas tumor studies excluding leukemia exhibiting STING or RIG-I mRNA expression levels greater than the 25th percentile or the best cut-off value determined by KMplot.com. B) Kaplan-Meier progression-free survival plots of patients categorized by low or high STING or RIG-I expression based on the best cut-off values determined by KMplot.com. The data sets cover all time points with a display cutoff of 120 months.

Supplemental Figure S2. VAX014 does not facilitate production of IFN-α in vitro and cytotoxicity of VAX014 is not affected by the presence of type I IFN. A) IFN-α (all isoforms) content in supernatants of MB49 cells treated with VAX-I or VAX014 (MOI 300) for 20 to 24 hours as measured via ELISA (n = 3). B) Cell killing by VAX014 in the presence or absence of universal type I IFN (IFN-α) in MB49 and LLC1 cells. Assays were run in triplicate and histograms shown are representative of one experiment. Histograms depict mean ± SEM.

Supplemental Figure S3. VAX014 does not activate endosomal TLR3 in MB49 cells. MB49 or J774A.1 cells were plated (5.0 × 105 cells/well) and treated with the TLR3 agonist Poly(I:C) (0.65 μg/mL), VAX-I or VAX014 rBMCs for ∼20 hours. A) IFN-β in supernatants of J774A.1 cells measured via ELISA (n = 3). B) Surface expression of PD-L1 and MHC-I in MB49 cells co-incubated with the supernatants from treated J774A.1 cells, as determined by flow cytometry (n = 3–6). C) IFN-β in supernatants of MB49 cells (n = 3). D) Surface expression of PD-L1 and MHC-I in MB49 cells as measured by flow cytometry (n = 6). Histograms depict mean ± SEM of all combined experiments and the dashed lines represent baseline expression of the untreated control for each cell type.

Supplemental Figure S4. Activation of murine STING and human IFN-β promoter by VAX014 in the 293 Dual mSTING reporter cell line. A) SEAP enzymatic activity (mSTING reporter) and B) luciferase activity (huIFN-β transcriptional reporter) of cells treated with VAX-I or VAX014 rBMCs (MOI 70) or DMXAA (10 μM) for ∼20 hours. Histograms depict mean ± SEM.

Supplemental Figure S5. VAX014-mediated IFN-β production is recapitulated in MC38 in vitro. A) Expression of cGAS, STING, RIG-I, and TBK1 in MC38 cells analyzed via Western blot. Vinculin was used as a loading control on each independent blot and representative blot is shown. Each blot was conducted at least twice. B) Constitutive surface expression and surface expression of PD-L1 and MHC-I induced by VAX014 in MB49 WT vs. MC38 WT cells (n = 1). C-E) MC38 WT cells were plated (5.0 × 105 cells/well) and treated with DMXAA (30 μg/mL), VAX014 (MOI 300) or VAX-I (MOI 300) ± H-151 (4 μg/mL) or Amx (100 μg/mL) for ∼20 hours, then supernatants transferred to MB49 WT recipient cells + H-151. C) Surface expression of PD-L1 and MHC-I in MB49 WT recipient cells + H-151 treated with MC38 WT supernatants, as determined by flow cytometry (n = 3). D) IFN-β in supernatants of treated MC38 WT cells as measured via ELISA (n = 3). E) Surface expression of PD-L1 and MHC-I in MB49 WT recipient cells ± IFNAR1 blockade co-incubated with VAX014 treated MC38 WT supernatants (n = 3) and surface expression of IFNAR1 in MC38 WT cells as analyzed by flow cytometry (n = 3). Histograms depict mean ± SEM of all combined experiments and dashed lines represent the baseline expression value of the untreated control.

Supplemental Figure S6. In vitro confirmation of all MB49 STING KO clones. A) Expression of STING and RIG-I in MB49WT (WT) and MB49STING KO (STING KO) clones using Vinculin or GAPDH as a loading control, representative of at least two independent experiments. B-C) Cells were plated (1.0 × 106 cells/well) and treated with DMXAA (30 μg/mL) for ∼20 hours. B) IFN-β in supernatants of treated cells as measured via ELISA (n = 3). C) Surface expression of PD-L1 and MHC-I in cells as determined by flow cytometry (n = 3). Histograms depict mean ± SEM of all combined experiments and dashed lines represent the baseline expression value of the corresponding untreated control. #indicates which STING KO clone is represented in Figure 3.

Supplemental Figure S7. Knockout of STING does not dampen VAX014-mediated oncolysis in MB49 cells. MB49WT (WT) and MB49STING KO (STING KO) cells were plated (5,000 cells/well) and treated with a titration (MOI 10,000:1–5:1) of VAX014 rBMCs for 2 hours and assayed for cell killing and release of LDH. A) EC50 values and B) LDHhalf-max values calculated by GraphPad Prism 10. Individual data points are plotted with the mean (n = 3). ns = not significant, compared to MB49 WT. #indicates which STING KO clone is represented in Figure 3.

Supplemental Figure S8. In vitro characterization of all MB49 STING KO clones. A-D) MB49WT (WT) or MB49STING KO (STING KO) cells were plated (5.0 × 105 - 1.0 × 106 cells/well) and treated with DMXAA (30 μg/mL) or VAX014 (MOI 300) ± Amx (100 μg/mL) for ∼20 hours. A) IFN-β in supernatants of treated cells as measured via ELISA (n = 3–6). B) Ratio of pIRF3 to IRF3 in cells 4 hours post-treatment analyzed via Western blot. Histogram includes values from two independent experiments. C) Representative Western blot for data in B). D) Surface expression of PD-L1 and MHC-I in cells as determined by flow cytometry (n = 3). Histograms depict mean ± SEM of all combined experiments and dashed lines represent the baseline expression value of the corresponding untreated control. #indicates which STING KO clone is represented in Figure 3.

Supplemental Figure S9. MB49 STING KO clones still respond to IFN-β by upregulating PD-L1 and MHC-I. A) Surface expression of IFNAR1 in MB49STING KO (STING KO) cells as analyzed by flow cytometry. MFI of unstained control was subtracted out and a secondary antibody control was included (n = 3). B) Surface expression of PD-L1 and MHC-I in MB49 STING KO cells ± IFNAR blockade co-incubated with VAX014-treated MB49 WT supernatants as determined by flow cytometry (n = 3). Histograms depict mean ± SEM of all combined experiments and dashed lines represent the baseline expression value of the corresponding untreated control. ns = not significant. #indicates which STING KO clone is represented in Figure 3.

Supplemental Figure S10. In vitro characterization of MC38 STING KO clones. A) STING expression in MC38WT (WT) and MC38STING KO (STING KO) clones. MB49STING KO cells were used as a negative control and Vinculin as a loading control (n= 1). B) RIG-I expression in MC38 WT and MC38 STING KO clones using GAPDH as a loading control (n= 1). C) MC38 WT and MC38 STING KO cells were plated (5,000 cells/well) and treated with a titration (MOI 2,000:1 to 4:1) of VAX014 rBMCs for 24 hours and assayed for cell killing (n= 1). D) Surface expression of IFNAR1 in MC38 WT and MC38 STING KO cells as measured by flow cytometry (n= 1). E-F) Cells were plated (5.0x105 cells/well) and treated with DMXAA (30 μg/mL) or VAX014 (MOI 300) ± Amx (100 μg/mL) for ∼20 hours and supernatants transferred to MB49 WT recipient cells + H-151. E) Surface expression of PD-L1 and MHC-I in MB49 WT recipient cells was analyzed by flow cytometry and F) IFN-β in supernatants of treated cells was measured via ELISA. Fresh MB49 WT cells ± H-151 were used as a control (n= 2-3). Histograms depict mean ± SEM of all combined experiments and dashed lines represent the baseline expression value of the corresponding untreated control. #indicates which STING KO clone was used in vivo (see Supplementary Figure S21).

Supplemental Figure S11. In vitro confirmation of all MB49 RIG-I KO clones and verification of intact STING activity. A) Expression of STING and RIG-I in MB49WT (WT) and MB49RIG-I KO (RIG-I KO) clones, using Vinculin or GAPDH as a loading control, representative of at least two independent experiments. B-C) Cells were plated (5.0x105 cells/well) and treated with DMXAA (30 μg/mL) ± H-151 (4 μg/mL) for ∼20 hours. B) IFN-β in supernatants of treated cells as measured via ELISA (n= 3). C) Surface expression of PD-L1 and MHC-I in cells as determined by flow cytometry (n= 3). Histograms depict mean ± SEM of all combined experiments and dashed lines represent the baseline expression value of the corresponding untreated control. #indicates which RIG-I KO clone is represented in Figure 4.

Supplemental Figure S12. Knockout of RIG-I does not dampen VAX014-mediated oncolysis in MB49 cells. MB49WT (WT) and MB49RIG-I KO (RIG-I KO) cells were plated (5,000 cells/well) and treated with a titration (MOI 10,000:1 to 5:1) of VAX014 rBMCs for 2 hours and assayed for cell killing and release of LDH. A) EC50 values and B) LDHhalf-max values calculated by GraphPad Prism 10. Individual data points are plotted with the mean (n= 3). ns = not significant, compared to MB49 WT. #indicates which RIG-I KO clone is represented in Figure 4.

Supplemental Figure S13. In vitro characterization of all MB49 RIG-I KO clones. A-D) MB49WT (WT) and MB49RIG-I KO (RIG-I KO) cells were plated (5.0x105 cells/well) and treated with DMXAA (30 μg/mL) or VAX014 (MOI 300) ± H-151 (4 μg/mL) for ∼20 hours. A) IFN-β in supernatants of treated cells as measured via ELISA (n= 3-6). B) Ratio of pIRF3 to IRF3 in cells 4 hours post-treatment analyzed via Western blot. Histogram includes values from two independent experiments. C) Representative Western blot for data in B). D) Surface expression of PD-L1 and MHC-I in cells as determined by flow cytometry (n= 5). Histograms depict mean ± SEM of all combined experiments and dashed lines represent the baseline expression value of the corresponding untreated control. #indicates which RIG-I KO clone is represented in Figure 4.

Supplemental Figure S14. MB49 RIG-I KO clones still respond to IFN-β by upregulating PD-L1 and MHC-I. A) Surface expression of IFNAR1 in MB49RIG-I KO (RIG-I KO) cells as analyzed by flow cytometry. MFI of unstained control was subtracted out and a secondary antibody control was included (n= 3). B) Surface expression of PD-L1 and MHC-I in MB49 RIG-I KO cells ± IFNAR blockade co-incubated with VAX014-treated MB49 WT supernatants as determined by flow cytometry (n= 3). Histograms depict mean ± SEM of all combined experiments and dashed lines represent the baseline expression value of the corresponding untreated control. ns = not significant. #indicates which RIG-I KO clone is represented in Figure 4.

Supplemental Figure S15. In vitro characterization of MC38 RIG-I KO clones. A) RIG-I expression in MC38WT (WT) and MC38RIG-I KO (RIG-I KO) clones as analyzed by Western blot. MB49 RIG-I KO cells were used as a negative control and GAPDH as a loading control (n= 1). B) STING expression in MC38 WT and MC38 RIG-I KO clones, using Vinculin as a loading control, analyzed via Western blot (n= 1). C) Cells were plated (5,000 cells/well) and treated with a titration (MOI 2,000:1 to 4:1) of VAX014 rBMCs for 24 hours and assayed for cell killing (n= 1). D) Surface expression of IFNAR1 in cells as determined by flow cytometry (n= 1). E-F) Cells were plated (5.0x105 cells/well) and treated with DMXAA (30 μg/mL) or VAX014 (MOI 300) ± H-151 (0.4 μg/mL) for ∼20 hours and supernatants transferred to MB49 WT recipient cells + H-151. E) Surface expression of PD-L1 and MHC-I in supernatant-treated MB49 WT recipient cells was analyzed by flow cytometry and F) IFN-β in supernatants of treated cells measured via ELISA. Fresh MB49 WT cells ± H-151 were used as a control (n= 2-3). Histograms depict mean ± SEM of all combined experiments and dashed lines represent the baseline expression value of the corresponding untreated control.

Supplemental Figure S16. In vitro characterization of all MB49 Double KO clones. A) RIG-I expression in MB49WT (WT) vs. MB49STING/RIG-I KO (Double KO) cells analyzed by Western blot. GAPDH was run as a loading control (n= 1). B) STING expression in MB49 WT vs. MB49 Double KO cells analyzed via Western blot, using Vinculin as a loading control (n= 1). C) MB49 WT, MB49 single KO (STING KO or RIG-I KO), and MB49 Double KO cells were plated (5,000 cells/well) and treated with a titration (MOI 2,000:1 to 4:1) of VAX014 rBMCs for 24 hours and assayed for cell killing. EC50 values calculated by GraphPad Prism 10 are shown (n= 2-3). D) Surface expression of PD-L1 in MB49 Double KO clones as determined by flow cytometry (n= 3). Cells were plated (5.0x105 cells/well) and treated with DMXAA (30 μg/mL) or VAX014 rBMCs (MOI 300) for ∼20 hours. E) IFN-β content in supernatants from D) as measured via ELISA (n= 2). F) Surface expression of PD-L1 and MHC-I in MB49 single KO (STING KO or RIG-I KO) and MB49 Double KO cells co-incubated with VAX014-treated MB49 WT supernatants for ∼20 hours, analyzed by flow cytometry (n= 1). Histograms depict mean ± SEM of all combined experiments and dashed lines represent the baseline expression value of the corresponding untreated control. SRKO=Double KO cell line where RIG-I was ablated from MB49 STING KO cells; RSKO=Double KO cell line where STING was ablated from MB49 RIG-I KO cells. ns = not significant, all compared to STING KO or RIG-I KO. #indicates which Double KO clone is represented in Figure 4.

Supplemental Figure S17. In vitro characterization of MC38 Double KO clones. A) STING expression in MC38WT (WT) and MC38STING/RIG-I KO (Double KO) clones analyzed via Western blot using Vinculin as a loading control (n= 1). B) RIG-I expression in MC38 WT and MC38 Double KO clones, analyzed by Western blot using GAPDH as a loading control (n= 1). C) Cells were plated (5,000 cells/well) and treated with a titration (MOI 2,000:1 to 4:1) of VAX014 rBMCs for 24 hours and assayed for cell killing (n= 1). D) Surface expression of IFNAR1 in MC38 WT vs. MC38 Double KO cells as analyzed by flow cytometry (n= 1). E-F) Cells were plated (5.0x105 cells/well) and treated with DMXAA (30 μg/mL) or VAX014 (MOI 300) for ∼20 hours and supernatants were transferred to MB49 WT recipient cells + H-151. E) Surface expression of PD-L1 and MHC-I in supernatant-treated MB49 WT recipient cells was analyzed by flow cytometry and F) IFN-β in supernatants of treated cells was measured via ELISA. Fresh MB49 WT cells ± H-151 were used as a control (n= 2-3). Histograms depict mean ± SEM of all combined experiments and dashed lines represent the baseline expression value of the corresponding untreated control.

Supplemental Figure S18. Evaluation of STING and RIG-I signaling genes in MB49 WT tumors. Heatmap of transcripts involved in STING and RIG-I signaling from i.d. B16F10 (n= 2) and MB49 WT (n= 3) tumors following i.t. treatment with saline for 24 hours. Heatmap was generated using Rosalind.

Supplemental Figure S19. Growth of MB49 STING KO and RIG-I KO i.d. tumors. 9-week-old female C57BL/6 mice were implanted with a single i.d. tumor on Day 0 and growth monitored until termination criteria were reached. A) Mean tumor growth rates and B) Kaplan-Meier survival curves for MB49WT (WT) and MB49STING KO (STING KO) tumors. C) Mean tumor growth rates and D) Kaplan-Meier survival curves for MB49 WT and MB49RIG-I KO (RIG-I KO) tumors. Mean tumor growth rates (mean ± SEM) were plotted until the median survival of the WT control. Student’s t test was used to analyze the significance of growth rates on Day 32. (n= 5-8/group). ns = not significant.

Supplemental Figure S20. Efficacy of i.t. VAX014 in i.d. MC38 WT tumors. 7-10-week-old female C57BL/6 mice were implanted with a single i.d. MC38 WT tumor on Day 0. Animals received weekly i.t. injections of saline or VAX014 starting on Day 5 until complete response (CR) was achieved or termination criteria were reached. A) Mean tumor growth rates and B) Kaplan-Meier survival curves for saline- and VAX014-treated animals (n= 14/group). Mean tumor growth rates (mean ± SEM) were plotted until the median survival of the saline group (Day 25). C) Kaplan-Meier survival curves for mice treated with saline or VAX014 in the presence or absence of depletion antibodies against CD4, CD8, and NK cells (n= 6/group). Unless otherwise indicated, the saline group was used as the comparator for statistical testing. D) Representative data of depletion confirmation in blood from mice in C) analyzed by flow cytometry (n= 3/group). Student’s t test was used to analyze the significance of tumor growth rates on Day 25.

Supplemental Figure S21. VAX014 efficacy is entirely dependent on tumor-intrinsic STING in i.d. MC38 tumors. 7-10-week-old female C57BL/6 mice were implanted with a single i.d. tumor on Day 0. Animals received weekly i.t. injections of saline or VAX014 starting on Day 4-6 until complete response (CR) was achieved or endpoint was reached. Tumor growth data and Kaplan-Meier survival curves for A) MC38WT (WT) and B) MC38STING KO (STING KO) tumors. Mean saline tumor growth rates (mean ± SEM) were plotted until the median survival of the corresponding saline group and individual VAX014-treated tumor growth rates were plotted until animal expiration (n= 12/group).

Supplemental Figure S22. Expression of type I IFN transcripts in tumors treated with VAX014 in vivo. A) Expression of IFN-β and several isoforms of IFN-α transcripts in excised i.d. MB49WT tumors treated with i.t. VAX014 at various treatment timepoints. Dashed line represents baseline normalized expression. B) Heatmap of IFN transcript log2 normalized expression at various treatment timepoints (n= 3). Violin plots show log2 normalized expression values as calculated by Rosalind and IFN transcript heatmap was generated using GraphPad Prism 10.

Supplemental Figure S23. Overall survival of high or low expression cohorts based on STING and/or RIG-I mRNA expression. Kaplan-Meier overall survival plots of patients, categorized by low or high STING and/or RIG-I expression based on the 25th percentile cut-off of all studies except for leukemia, for A) all solid tumors, B) BRCA, C) HNSC, and D) SKCM. The data sets cover all time points with a display cutoff of 120 months.

Supplemental Figure S24. STING expression is prevalent and associates with improved survival across solid tumor types. A) Percentage of patients exhibiting high STING mRNA expression levels (defined as expression value higher than the 25th percentile among all TCGA Pan-Cancer Atlas expression data) from the cBioPortal database, excluding those solid tumor indications with fewer than 100 patients. The acronyms for the individual cancer indications follow the standard TCGA study abbreviations. B) Forest plot of hazard ratios ±95% confidence intervals as determined by Cox regression analysis of progression-free survival (PFS) based on the STING expression status. C) Violin plots of RNA-Seq by Expectation Maximization (RSEM) normalized STING mRNA expression levels. Orange and blue lines represent the median and quartiles, respectively. The dashed line represents the 25th percentile of all solid tumors. D) Kaplan-Meier curves based on PFS of patients as categorized by low or high STING expression. The data sets cover all time points with a display cut-off of 120 months. †Hazard ratio is greater than 10.

Supplemental Figure S25. RIG-I expression is prevalent and associates with improved survival across solid tumor types. A) Percentage of patients exhibiting high RIG-I mRNA expression levels (defined as expression value higher than the 25th percentile among all TCGA Pan-Cancer Atlas expression data) in the cBioPortal database, excluding those solid tumor indications with fewer than 100 patient samples. The acronyms for the individual cancer indications follow the standard TCGA study abbreviations. B) Forest plot of hazard ratios ± 95% confidence intervals as determined by Cox regression analysis of progression-free survival (PFS) based on the RIG-I expression status. C) Violin plots of RNA-Seq by Expectation Maximization (RSEM) normalized RIG-I mRNA expression levels. Orange and blue lines represent the median and quartiles, respectively. The dashed line represents the 25th percentile of all solid tumors. D) Kaplan-Meier curves based on PFS of patients as categorized by low or high RIG-I expression. The data sets cover all time points with a display cut-off of 120 months. $Hazard ratio is greater than 10.

Supplemental Figure S26: TBK1 inhibition with BX795 abrogates VAX014-mediated IFN-β gene expression in additional human solid tumor cell lines positive for STING and RIG-I. Expression of cGAS, STING, and RIG-I was analyzed via Western blot. Each blot was run independently and included loading controls. Cells were treated with VAX-I or VAX014 rBMCs (±BX795) for 4 hours and analyzed by RT-qPCR (n= 2-4).