We then set out to understand what signaling cascade governs cell-in-cell structures. Initially, we sought to test if the observed results were cell fusion. Thus, we incubated tumor cells whose cytosols were labeled with either Wasabi or tdTomato with tumor-reactive T cells. Confocal analysis indicated that each cell type in the cell-in-cell formation maintained its cytoplasm, and no mix between colors was detected (Figure 6—figure supplement 1). Furthermore, long-term follow-up of cell dissemination from this structure indicated that each cell maintains its initial single labelling color (Figure 6—video 1 and Figure 6—figure supplement 1). To corroborate this, we also incubated tumor cells whose membrane, nucleus, and F-actin were labeled with different fluorophores. Similarly, each cell in this formation was separated and maintained the integrity of its original cell components (Figure 6A). Given the similarity of this formation to entosis, we used ROCK inhibitor, which is the key regulator of this process. Indeed, blocking ROCK almost completely prevented T cell-mediated cell-in-cell formation (Figure 6B, C). We then tested whether the molecular machinery reported to mediate tumor spontaneous entosis applies to govern the current cell-in-cell structure. However, we observed no increase or changes in the cellular localization of phosphorylated β catenin, E-cadherin, and phosphorylated integrin β1 (26) (Figure 6—figure supplement 1) suggesting other mediators promote this entosis. Furthermore, there was no reduction in cell-in-cell formation upon blocking of E- and N- cadherins, or inhibition of Wnt signaling (Figure 6D–E). In sharp contrast, disruption of actin filaments, blocking of mRNA synthesis or protein production completely abrogated tumor cells’ capacity to form a cell-in-cell formation, suggesting that the structure requires de novo synthesis of genes (Figure 6D–E).
Figure 6 with 2 supplements with 2 supplements see all Download asset Open asset STAT3 and EGR1 signaling govern T cell-mediated cell-in-cell tumor formation. (A) Representative images of B16F10, co-expressing Lifeact-GFP and H2b-tdTomato or MyrPalm-tdTomato and H2b-GFP, following incubation with gp100-reactive T cells. (B) Mean percentage of cell-in-cell tumor formations in B16F10 following overnight incubation with gp100-reactive CD8+ T cells with or without ROCK inhibitor (n=3). (C) Representative images of cell-in-cell tumor formations in B16F10 following overnight incubation with gp100-reactive CD8+ T cells with or without ROCK inhibitor. (D) Mean percentage of cell-in-cell tumor formations in B16F10 cells following overnight incubation with specific inhibitors and reactive CD8+ T cells (n=4). (E) Representative images of B16F10 cells treated with inhibitors and incubated overnight with gp100-reactive CD8+ T cells. (F) Significantly increased genes in B16F10 cells incubated with T cell-derived granules (Lyso) or isolated directly from relapsed tumors (Tumor), compared to B16F10 control cells (WT) (Bottom) and relative expression of the top 25 genes (Top) (n=3). (G) STAT3 and EGR1 expression levels in B16F10 cells isolated directly from relapsed tumors (Tumor) and after incubation with T-cell-derived granules (Lyso) compared to B16F10 control cells (WT) (n=3) (H) Mean percentage and representative images of cell-in-cell tumor formations in B16F10 48 hours after transfection with STAT3-T2A-iRFP670, EGR1-T2A-GFP or both (n=3). (I) Mean percentage of apoptotic B16F10, transfected with STAT3-T2A-iRFP670, EGR1-T2A-GFP or both, following incubation with tumor reactive T cells (n=3). (J) Mean percentage of cell-in-cell tumor formations in B16F10, transfected with siRNA, following incubation with tumor reactive T cells or T cells secreted granules. (K–L) B16F10 tumor size in mice treated with gp100-reactive T cells (ACT) (K) or Dc adjuvant and anti-TRP1 antibodies (L) with or without Stattic (n=4). Orange arrowheads indicate Stattic treatments and black arrowheads indicate injection of immunotherapies. All experiments were repeated independently at least three times. Statistical significance was calculated using ANOVA with Tukey’s correction for multiple comparisons (**denotes p<0.01, *** denotes p<0.001, **** denotes p<0.0001). Error bars represent standard error. Scale bars = 20 μm. Figure 6—source data 1 Significantly elevated genes in both B16F10 cells incubated with T cell secreted granules and tumor cells sorted from treated animals, related to Figure 6. https://cdn.elifesciences.org/articles/80315/elife-80315-fig6-data1-v1.xlsx Download elife-80315-fig6-data1-v1.xlsx Figure 6—source data 2 STAT3 and EGR1 pathways regulate cell-in-cell tumor formation, related to Figure 6. https://cdn.elifesciences.org/articles/80315/elife-80315-fig6-data2-v1.xlsx Download elife-80315-fig6-data2-v1.xlsx Figure 6—source data 3 Log2 expression of genes related to cancer pathways in tumor cells sorted from treated animals compared to B16F10 WT, reated to Figure 6. https://cdn.elifesciences.org/articles/80315/elife-80315-fig6-data3-v1.xlsx Download elife-80315-fig6-data3-v1.xlsx
To assess what genes govern this formation, we compared the gene signature of untreated B16F10 cells, to cell-in cell formation induced following incubation with T-cell-derived secreted granules and to that of tumor cells sorted from in vivo five days after immunotherapy. Over 400 genes were increased in cell-in-cell formation induced by T cell secreted granules compared to untreated B16F10 cells, 215 of which were also upregulated by the tumor cells sorted from treated animals (Figure 6F, ). KEGG analysis further indicated that multiple signaling cascades, including the JAK/STAT3 axis and FGF-receptors downstream pathways, are enriched in cell-in-cell formation generated in vivo, with approximately 80 significantly elevated genes relating to these pathways (Figure 6—figure supplement 1, ). Indeed, EGR1 and STAT3 expression was significantly elevated in cell-in-cell tumors (Figure 6G). Both mice and human tumors incubated with reactive T cells had elevated their p-STAT3 levels, in comparison to untreated tumor cells (Figure 6—figure supplement 1), suggesting this mechanism is conserved across species. Since these results may also reflect adaptation to the tumor microenvironments, we next corroborate the necessity of these factors to cell-in-cell formation ex vivo. Overexpression of either STAT3 or EGR1 was sufficient to induce cell-in-cell formation without additional stimulation (Figure 6H). Furthermore, these cells were significantly more resistant to killing by CD8+ T cell, compared to sham transfected cells (Figure 6I). We also tested if inhibiting these genes would reduce cell-in-cell tumor formation. Inhibition of STAT3 and EGR1, but not MAPK3 and EGR2, significantly abrogated the capacity of tumor cells to form cell-in-cell structures upon incubation with IFNγ-stimulated T cells and with T cell secreted granules (Figure 6J and Figure 6—figure supplement 1). In order to integrate cell-in-cell inhibition to an in vivo therapy, we first tested the effect of small molecule inhibitors on tumor cell formation. We found that inhibition of STAT3, or EGR pathway, completely prevented cell-in-cell formation upon incubation with reactive T cells or T cell-secreted granules (Figure 6—figure supplement 1). Since blocking EGR1 also reduced T cell viability, we then set out to establish a treatment protocol that combines STAT3 inhibition (which also inhibits T cell activity) and immunotherapy. To this end, mice bearing palpable tumors were injected with Stattic or with PBS for two consecutive days. On the second day, we sub-lethally irradiated the mice and injected them with 5 × 106 gp100-reactive T cells and IL-2. In another model, we treated mice for two days with Stattic, followed by treatments with anti-CD40, TNFα, and anti-TRP1. Recurrent tumors were treated with the same regimen. In both models, injection of Stattic alone had no effect on tumor growth. We found, however, that injection of Stattic prior to administration of immunotherapy partially restored the responsiveness of tumors that re-occur following immunotherapies (Figure 6K–L). While tumor cell sensitization may be a result from multiple mechanisms, these results stress the benefit of combining immunotherapy with STAT3 inhibition.