Tumor survival and PD-L1 expression are promoted by increased levels of IL-10 in DLBCL. (A) Baseline expression of IL10 across normal GC B cells and human DLBCL data sets classified according to COO gene expression or LymphGen genetic subtypes; merged expression arrays (left) from GSE56315 and GSE12195; RNA-seq (right) from Schmitz/Wright et al^2,4 (t test). (B) Expression of Il10 by RNA-seq and RT-qPCR between normal GC B cells (B220⁺CD38^lowFAS⁺YFP⁺) from control YC mice and lymphoma cells (B220⁺GFP⁺) from pBIC tumors (t test and Mann-Whitney, respectively). Values from our previous GSE116290 (triangles) or current GSE255140 (circles) RNA-seq experiments were normalized and analyzed combined here. (C) Model depicting the IL-10/JAK/STAT3 autocrine pathway that promotes survival and upregulation of PD-L1 in DLBCL cells, showing different strategies to inhibit this survival loop. (D) IL-10 expression and pSTAT3 quantification measured as MFI by intracellular flow cytometry of lymphoma (CD19⁺GFP⁺) and adjacent normal B cells (CD19⁺GFP^–), respectively (Mann-Whitney). (E) Normalized cell viability assays of lymphoma cells (B220⁺GFP⁺) cells compared with adjacent normal B cells (B220⁺GFP^–) studied by flow cytometry after 24-hour ex vivo culture with increasing doses of anti–IL-10R, ruxolitinib, or pyrimethamine inhibitors (n ≥ 3) and 50% inhibitory concentration (IC50) calculations from baseline-corrected data. (F) Relative changes in PD-L1 surface expression between adjacent normal B cells (B220⁺GFP^–) and lymphoma cells (B220⁺GFP⁺) after 24 hours untreated vs treated with concentrations above the IC50 of ruxolitinib (100 μM) and pyrimethamine (200 μM; n ≥ 3; analysis of variance [ANOVA]). ∗P < .05; ∗∗P < .01; ∗∗∗P < .001. MFI, median fluorescence intensity.