Blockade of innate inflammatory cytokines TNF α, IL-1 β, or IL-6 overcomes virotherapy-induced cancer equilibrium to promote tumor regression

Abstract

Cancer therapeutics can lead to immune equilibrium in which the immune response controls tumor cell expansion without fully eliminating the cancer. The factors involved in this equilibrium remain incompletely understood, especially those that would antagonize the anti-tumor immune response and lead to tumor outgrowth. We previously demonstrated that continuous treatment with a non-replicating herpes simplex virus 1 expressing interleukin (IL)-12 induces a state of cancer immune equilibrium highly dependent on interferon-γ. We profiled the IL-12 virotherapy-induced immune equilibrium in murine melanoma, identifying blockade of innate inflammatory cytokines, tumor necrosis factor alpha (TNFα), IL-1β, or IL-6 as possible synergistic interventions. Antibody depletions of each of these cytokines enhanced survival in mice treated with IL-12 virotherapy and helped to overcome equilibrium in some tumors. Single-cell RNA-sequencing demonstrated that blockade of inflammatory cytokines resulted in downregulation of overlapping inflammatory pathways in macrophages, shifting immune equilibrium towards tumor clearance, and raising the possibility that TNFα blockade could synergize with existing cancer immunotherapies.

Keywords: IL1B; IL6; TNF; cytokine blockade; equilibrium; macrophage.

Graphical Abstract

Figure 1.

Injection of d106S-IL12 virus generates immune equilibrium and results in pro-inflammatory cytokine secretion. (A) Mice bearing B16^Nectin1 tumors were injected intratumorally every 3 days starting on day 7 with the listed treatments (N = 5 mice per group). (B) Mice were treated as in (A), but therapy was stopped at day 19 after five total injections (N = 5/10 mice for PBS/d106S-IL12). (C/D) Following seven days of B16^Nectin1 tumor growth, mice were injected intratumorally every three days with the listed treatments. For better visualization, cytokines are grouped based on high, medium, and low concentrations present in the tumor. (C) At day 10 post-tumor challenge, following one injection or (D) at day 17 post-tumor challenge, following four injections, mice were sacrificed. Tumor lysates were subjected to cytokine/chemokine analysis by multiplex array. Protein lysates were normalized using a BCA assay. Each column represents one mouse (N = 5 mice per group, except N = 4 for d106S-IL12 day 17). (E) Data from (D) day 17 cytokine levels. Groups were compared with a one-way ANOVA and Dunnett’s multiple comparisons test; *P < 0.05, ***P < 0.001.

Figure 2.

Pro-inflammatory cytokine blockade slows tumor growth and overcomes d106S-IL12-induced immune equilibrium. B16^Nectin1 tumors were treated intratumorally starting at day 7 every 3 days for total of five injections (N = 5 for PBS, n = 10 for d106S-IL12 groups). At day 10, blocking antibodies against (A/B) TNFα, (C/D) IL-1β, or (E/F) IL-6R (100 µg per antibody) were injected intraperitoneally every 3 days for a total of six injections. Survival curves (B/D/F) are across two independent experiments (N = 15 mice per group). (G) Tumor-free mice from survival experiments of single cytokine blockade. (H) Pooled survival across two independent experiments (N = 15/45 for no blockade/cytokine blockade groups). Log-rank test; *P < 0.01. Arrows indicate intratumoral treatments (PBS/d106S-IL12), green circles indicate cytokine blockade.

Figure 3.

Cytokine blockade does not profoundly change levels of immune cells infiltrating tumors. B16^Nectin1 tumors were treated intratumorally starting at day 7 every 3 days for total of four injections of either PBS, d106S, or d106S-IL12. At day 10, blocking antibodies against either TNFα, IL-1β, or IL-6R (100 µg per injection) were injected every 3 days for a total of three injections (N = 5 per treatment group). Mice were sacrificed on day 16 (PBS groups) or day 17 (d106S, d106S-IL12 groups) post-tumor challenge. Tumors from each treatment group were pooled (N = 5) and sorted for CD45+ cells and subjected to single-cell RNA-sequencing. (A) t-SNE plot of all analyzed cells colored by unsupervised clustering. (B) Cell cluster frequency within each treatment sample. (C) Distribution of cell clusters across treatment groups from (A). (D/E) Flow cytometry on individual tumor samples from the scRNA-seq. (D) Flow cytometry analysis was compared to scRNA-seq clustering showing similar patterns. Gating strategy in Supplementary Fig. 2. Wilcoxon signed rank test *P < 0.05 for scRNA-seq; unpaired t-test for pooled flow cytometry samples *P < 0.05, ***P < 0.001; one-way ANOVA and Tukey’s multiple comparisons test for d106S-IL12 no blockade vs. cytokine blockades ^##P < 0.01, ^###P < 0.001. This experiment was analyzed in part previously [6].

Figure 4.

Pro-inflammatory cytokine blockades transcriptionally reprogram macrophages and reduce similar inflammatory pathway genes, regardless of targeted cytokine. (A) Differentially expressed gene (DEG) analysis was performed comparing d106S-IL12 no blockade to d106S-IL12 plus pro-inflammatory cytokine blockade groups (+anti-TNFα, IL-1β, or IL-6R) from the scRNA-seq data. Shown are Venn diagrams of shared DEGs among cytokine blockade groups. (B) Volcano plot of shared DEGs upregulated or downregulated by pro-inflammatory cytokine blockade. (C) Gene Set Enrichment Analysis for hallmark TNF or IL6 signaling and response genes for the three cytokine blockade groups (+d106S-IL12) versus d106S-IL12 no blockade. NES = normalized enrichment score. (D) Normalized expression of cytokines across top expressing clusters. (E) t-SNE plot where each cell is shaded according to the number of shared DEGs that it expresses. High expression of genes largely falls within the macrophage/monocyte clusters (see Fig. 3A). (F) Top significantly downregulated inflammatory genes in macrophage/monocyte clusters with greater than 0.6 average log-fold change versus no blockade d106S-IL12. Genes are grouped by function. GC = glucocorticoids, TF = transcription factor.