Supplementary Figure 5: Fumarate drives Nrf2-dependent transcription of IL-6

(a) Comparison of genes that were differentially expressed between non-treated and fumarate-treated monocytes (y-axis) and those differentially expressed between primary MZ-like B cells from HCs (n = 4) and primary B cells from PPBL patients (n = 3) (x-axis). (b) Representative immunoblot probing for histone H3 methylation among primary B cells of a HC vs. PPBL patient. The bar graph summarizes data from HCs (n = 3) and PPBL patients (n = 3). (c) Representative imaging flow cytometry data showing nuclear abundance of Nrf2 in primary PPBL B cells in presence/absence of the Nrf2 inhibitor, K67. Black scale bar, 7 μm. Histograms depict similarity dilate scores of NRF2 nuclear localization in presence or absence of K67. Experiment was performed once. (d) mRNA abundance of canonical NRF2 regulated genes in PPBL B cell lines (n = 3) and cell lines derived from marginal zone-like B cells from HCs (n=3) (left). Abundance of GSH and GSSG from primary B cells (right) (e) Quantification of IL-6 production by MZ-like B-LCLs from HCs (n = 3) of cells incubated with monomethyl fumarate for 48 hours in presence vs. absence of the Nrf2 inhibitor, ML-385. (f) mRNA abundance of NQO1 and IL6 from PPBL (n=3) and healthy MZ (n=3) B-LCLs activated with IL-21 plus CD40L in the presence or absence of the NRF2 activator, CDDO (1 μM). (g) Effect of Nrf2 inhibition using K67 on IL-6 production, and expression of activation markers, in primary naïve and memory B cells from healthy donors activated with CpG (2.5 μg/ml) for 24 hours (n = 5 healthy donors). (h) Schematic summary of the clinical trial performed in Patient 8. (i) Graphical summary of the model proposed from this study. Pooled data are shown as mean ± SEM. Correlation in (a) was assessed using Pearson’s r, statistical significance was assessed by two-sided unpaired t-test (b,e,d [right]), two-way ANOVA (d [left]).