Fig 2: Overexpression of Irg-1 activates the Nrf2 pathway.A, B The translocation of Nrf2 into the nucleus and cytoplasm of HO-1 and NQO1 expression were detected by western blot and quantified by Image J software. C, D The Nrf2 was detected by immunofluorescence combined with DAPI staining for nuclei (scale bar: 50 µm). E, F The silencing efficiency of Si-Nrf2 was detected by western blot. G, H After treatment with Si-Nrf2, Nrf2 in the nucleus and HO-1 and NQO1 in the cytoplasm were detected with western blot and quantified by Image J software. I, J The protein expression of iNOS, COX2, IL-6 were detected by western blot. The data are presented as the mean ± SD. n = 5. NS not significant, *P < 0.05, **P < 0.01.

Fig 3: CircDYM directly bound to TAF1 to regulate microglial activation. (a) Microarray heat map assessing the variation in mRNA expression in primary mouse microglia with overexpressed circDYM and vector treated with or without LPS. (n = 3 for each group) (b) Prediction of transcriptor-mRNA interaction using the GTRD algorithm. (c) qPCR analysis of the expression of potential TAF1 target genes in LPS-induced primary mouse microglia after circDYM overexpression. All data were presented as mean ± SEM of three independent experiments. *P < 0.05, **P < 0.01, ***P < 0.001 versus the Control + Vector group; # P < 0.05 versus the LPS + Vector group using two-way repeated measures ANOVA followed by the Holm-Sidak post hoc multiple comparison test. (d) Prediction of circDYM-TAF1 interaction using the catPAPID algorithm and schematic of TAF1 with functional protein domains. (e) qPCR analysis of relative enrichment of endogenous circDYM in TAF1 RIP. All data were presented as mean ± SEM of three independent experiments. ***P < 0.001 versus the IgG group using Student's t-test. (f) Western blot analysis of relative enrichment of endogenous TAF1 in circDYM RIP. All data were presented as mean ± SEM of three independent experiments. **P < 0.01 versus the Vector group using Student's t-test. (g) Co-localization of circDYM and TAF1 in the cytoplasm of primary mouse microglia by FISH analysis. Green: circDYM; Red: TAF1; Blue: DAPI. Scale bar: 20 µm. (h–j) Western blot analysis of TAF1 protein levels (h) and cytoplasmic (i) or nuclear (j) localization of TAF1 in primary mouse microglia after vector or circDYM lentivirus transduction with or without treated with LPS (100 ng ml–1) for 24 h. The purity of cell subcellular fractions was assessed by western blotting against specific markers. All data were presented as mean ± SEM of three independent experiments. *P < 0.05, **P < 0.01 versus the Control + Vector group; # P < 0.05, ### P < 0.001 versus the LPS + Vector group using two-way ANOVA followed by the Holm-Sidak post hoc multiple comparison test. (k) Distribution of TAF1 in primary microglia. Primary mouse microglia were subjected to immunocytochemistry analysis of TAF1 and Iba-1 proteins and DAPI staining of genomic DNA with or without LPS treatment. Scale bar: 20 µm. (l) Overexpression of circDYM significantly inhibited iNOS expression induced by TAF1 in primary mouse microglia. All data were presented as mean ± SEM of three independent experiments. **P < 0.01 versus the Control + Vector group; ## P < 0.01 versus the TAF1 +Vector group using two-way repeated measures ANOVA followed by the Holm-Sidak post hoc multiple comparison test. (m) qPCR analysis of the expression of potential TAF1 target genes in primary mouse microglia after overexpression of circDYM or TAF1. All data were presented as mean ± SEM of three independent experiments. **P < 0.01 versus the Control + Vector group; # P < 0.05, ## P < 0.01 versus the TAF1 +Vector group using two-way repeated measures ANOVA followed by the Holm-Sidak post hoc multiple comparison test. Images of unedited full blots in Figure S19

Fig 4: DCL suppresses LPS/IFN?-induced inflammatory response in murine macrophages. (A) Cell viability of LPS/IFN?-stimulated RAW264.7 cells treated with different concentrations of DCL or BAY was measured by CCK-8 assay. (B) The effects of DCL on NO production in LPS (0.5 µg/ml)/IFN? (10 ng/ml)-stimulated murine RAW264.7 macrophage cell line and murine primary PMs. (C) The effect of DCL on LPS/IFN?-induced PGE2 release in RAW264.7 cells. (D) Western blot analysis of iNOS and COX-2 from RAW264.7 cells subjected to LPS/IFN? stimulation and treated with different concentrations of DCL or BAY. The densitometry analysis of iNOS (E) and COX-2 (F), normalized against ß-Actin. Data are represented as mean ± SEM, n = 3. # p < 0.05, ## p < 0.01, ### p < 0.001, #### p < 0.0001 compared to the control check (CK) group. * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001 compared to LPS/IFN? group.

Fig 5: M1-phenotype and M2-phenotype of Microglia in all groups. a Representative photomicrographs of iNOS and Iba1, iNOS and NeuN double labeling cells in APP/PS1 mice. b Representative photomicrographs of iNOS-labeled cells morphology. c The number of microglia expressing iNOS and d neuron expressing iNOS in PtA and hippocampus of all groups. The data are presented as mean ± SEM (n = 4/group) in the different experimental groups. e The mRNA levels of IL-1ß in cortex and hippocampus. The data are presented as mean ± SEM (n = 4/group) in the different experimental groups. f Representative photomicrographs of CD206 positive cells and g the number of positive cells in PtA and hippocampus of all groups. h IL-10, i Fizz1, and j YM1 mRNA levels in PtA and hippocampus. The data are presented as mean ± SEM (n = 4/group) in the different experimental groups. #p < 0.05, ##p < 0.01 vs WT mice; *p < 0.05, **p < 0.01, ***p < 0.001 vs APP/PS1 mice