IDX combined with cis endows T cells with enhanced migratory capability via a CXCL10-type I IFN axis
Tumor-infiltrating immune cells that can produce cytotoxic mediators are amongst the most critical players toward resolving tumorigenesis; therefore, we next investigated the migratory capacity of PBMCs in response to drug treatment. Substantial increases in the total number of migrating PBMCs towards IDX2+cis1.2-treated cancer cells were observed compared to DMSO control (2.4-fold, p=0.0052; Figure 4A). Similarly, an increased number of migratory CD3+ cells were observed in combination treatments (IDX1+cis1.2 and IDX2+cis1.2) (53.03 % ± 8.859, p=0.0236 and 68.18 % ± 7.299, p=0.0029, respectively; Figure 4B). Comparing T cells subpopulations, exposure of IDX alone to tumor cells could dose-dependently enhance CD4+ and CD8+ cell migration (Supplementary Figure 2B). Particularly, the increased proportion of migratory CD8+ T cell (mean 61.2% ±1.937 and 64.99% ±8.539, respectively) was slightly greater than CD4+ T cell (mean 36.66% ±9.446 and 33.5% ±3.87, respectively), but markedly lower than double-positive (DP) CD4+ CD8+T cell subsets (mean 78.39% ±11.81 and 82.19% ±15.48, respectively) upon combination treatments (IDX1+cis1.2 and IDX2+cis1.2) when compared to DMSO. These results suggest that combination treatment could favor the migration of DP T cells towards tumor site (Figure 4C-E, Supplementary Figure 2C).
We next investigated the effect of IDX, cis alone or in combination on downstream signaling in cancer cells. The action of IDX has been suggested to be restricted to cells with defective DNA repair capacity, such as cancer cells and likely modulated by the stimulator of IFN genes (STING) signalling pathway. In Figure 4F, STING protein was constitutively expressed in NPC cell lines, with levels sustained with IDX1. STING expression was, unexpectedly, reduced upon IDX2 treatment, and to a greater extent, by cis1.2. We speculate that cis may compromise STING expression, impairing the type I IFN production pathway and suppressing proinflammatory chemokine production. As expected, administration of cis downregulated IRF3 phosphorylation at Ser396, but not Ser386, in NPC cell lines as compared to control and IDX treatment. This suppressive effect was also observed for the type I IFN protein IFN-α, while its expression was restored upon combination treatment. Interestingly, the detection patterns of apoptotic caspase-7 cascade were opposite to those of IFN-α, where cleaved caspase-7 was found to be greatly increased in NPC cell lines following cis treatment, as compared to control and IDX alone. This finding is consistent with those previously reported on the apoptotic caspase cascade functions where caspase-7 was shown to potentially cleave and suppress STING-mediated type I IFN production (Ning et al., 2019; White et al., 2014). Notably, we observed that IDX2 alone greatly induced the production of IFN I downstream component CXCL10, known to be involved in T cells chemotactic recruitment into tumor (densitometries demonstrated a 2.93-fold increase in protein expression as compared to DMSO; Supplementary Figure 2D) (Berghuis et al., 2011; Muthuswamy et al., 2012). We further demonstrated that CXCL10 was present in the supernatant of tumor cells, and that its up-regulation was particularly influenced by IDX2 (1.5-fold) and IDX2+cis1.2 combination treatment (1.32-fold; Figure 4G) as compared to DMSO. Moreover, the levels of additional proinflammatory chemokines (CXCL8 [1.93-fold and 2.11-fold, IDX2 and IDX2+cis respectively], CCL8 [1-fold and 1.17-fold], CXCL9 [1.26-fold and 1.18-fold], TNFSF1B [1.48-fold and 1.24-fold], and TNFRSF1A [3.66-fold and 3.45-fold]) involved in T cell activation and chemotaxis were found to be significantly augmented upon treatment with IDX, either alone or in combination with cis1.2.