SCD2-mediated cooperative activation of IRF3-IRF9 regulatory circuit controls type I interferon transcriptome in CD4+ T cells

Abstract

Type I interferons (type I-IFN) are critical for the host defense to viral infection, and at the same time, the dysregulation of type I-IFN responses leads to autoinflammation or autoimmunity. Recently, we reported that the decrease in monounsaturated fatty acid caused by the genetic deletion of Scd2 is essential for the activation of type I-IFN signaling in CD4⁺ Th1 cells. Although interferon regulatory factor (IRF) is a family of homologous proteins that control the transcription of type I-IFN and interferon stimulated genes (ISGs), the member of the IRF family that is responsible for the type I-IFN responses induced by targeting of SCD2 remains unclear. Here, we report that the deletion of Scd2 triggered IRF3 activation for type I-IFN production, resulting in the nuclear translocation of IRF9 to induce ISG transcriptome in Th1 cells. These data led us to hypothesize that IRF9 plays an essential role in the transcriptional regulation of ISGs in Scd2-deleted (sgScd2) Th1 cells. By employing ChIP-seq analyses, we found a substantial percentage of the IRF9 target genes were shared by sgScd2 and IFNβ-treated Th1 cells. Importantly, our detailed analyses identify a unique feature of IRF9 binding in sgScd2 Th1 cells that were not observed in IFNβ-treated Th1 cells. In addition, our combined analyses of transcriptome and IRF9 ChIP-seq revealed that the autoimmunity related genes, which increase in patient with SLE, were selectively increased in sgScd2 Th1 cells. Thus, our findings provide novel mechanistic insights into the process of fatty acid metabolism that is essential for the type I-IFN response and the activation of the IRF family in CD4⁺ T cells.

Keywords: CD4+ T cells; ChIP-seq; IFR9; IRF3; RNA-seq; SCD2; fatty acid metabolism.

Figure 1

The transcriptome of ISGs in sgScd2 Th1 cell depends on IRF3 and IRF9, but not IRF7. (A) A scatter plot of gene expression by RNA-sequencing (n = 5 per sample) compares in control and sgScd2 Th1 cells. The genes with over 1.5-fold changes are marked with black dots, and ISGs are marked with red dots (Control, n = 5; sgScd2, n = 5 biologically independent sample). (B) GSEA reveals the upregulation of the ISGs in sgScd2 Th1 cells. Genes are ranked into an ordered list on the basis of fold change in control and sgScd2 Th1 cells. Genes below the picture indicate leading edge subset. (C) Western blot analysis of SCD2 from control, sgScd2, sgScd2/sgIrf3, sgScd2/sgIrf7 and sgScd2/sgIrf9 Th1 cells. The summary of relative intensity was shown. Band intensity was determined by image j. (D) The Venn diagram showed overlaps and differences between 1.5-fold increased genes in sgScd2, sgScd2/sgIrf3, sgScd2/sgIrf7 or sgScd2/sgIrf9 Th1 cells compared to control Th1 cells. (E) A heat map depicts the gene relevant to (B). (F–H) A scatter plot of gene expression by RNA-sequencing (n = 5 per genotype) compares sgScd2/sgIrf3 (F), sgScd2/sgIrf9 (G) or sgScd2/sgIrf7 (H) Th1 cells against control Th1 cells. (I) GSEA reveals the upregulation of the ISGs in sgScd2/sgIrf7 Th1 cells. Genes are ranked into an ordered list on the basis of fold change in control and sgScd2/sgIrf7 Th1 cells. Genes below the picture indicate leading edge subset. Three experiments were performed and showed similar results (C). Data represent mean ± SD (one-way ANOVA test followed by Tukey’s post-hoc test for multiple comparisons, P****<0.0001).

Figure 2

Gene deletion of SCD2 resulted in the translocation of IRF3 and IRF9 from cytosolic to nuclear. (A, B) Western blot analysis of phospho-IRF3 (pIRF3), total IRF3 (A) and IRF9 (B) from control and sgScd2 Th1 cells. (C) qRT-PCR analyses of the relative expression of Irf7 from control, sgIrf7, sgScd2 and sgScd2/sgIrf7 Th1 cells. Relative expression (normalized to Hprt) with SD is shown. (D) Intracellular staining and flow cytometry analyzing of IRF7 in control, sgIrf7, sgScd2 and sgScd2/sgIrf7 Th1 cells. Mean fluorescence intensity (MFI) of IRF7 are shown. Summary data of three independent experiments of IRF7 expression are shown here. Each dot represents one experiment. Data are means ± SD. (n = 3 per each group biologically independent sample). (E) Western blot analysis of pIRF3, total IRF3 and IRF9 from control, sgScd2 and sgScd2/sgTmem173 Th1 cells. (F, G) Western blot analysis of pSTING, total STING (F) and pTBK1, TBK1 (G) from control, sgScd2, sgScd2/sgIrf3, sgScd2/sgIrf7 and sgScd2/sgIrf9 Th1 cells. (H) The amount of IFNα in the cell supernatant was measured by ELISA. Data are means ± SD. (I) Western blot analysis of IRF9 from control, sgScd2 and sgScd2/sgTmem173 Th1 cells. (J) Western blot analysis of pIRF3 and total IRF3 from control, sgScd2, sgScd2/sgIrf3 and sgScd2/sgIrf9 Th1 cells. (K) Western blot analysis of IRF9 from sgScd2 Th1 cells treated with 10μg/ml IFNAR neutralizing antibody. Isotype antibody was used as control. Band intensity was determined by image j and summary of three independent experiments was shown. Three technical replicates were performed with quantitative RT-PCR and relative expression (normalized to Hprt) with SD is shown (C). Three experiments were performed and showed similar results. Data represent mean ± SD (unpaired two-tailed student t tests or one-way ANOVA test followed by Tukey’s post-hoc test for multiple comparisons, P*<0.05, P**<0.01, P***<0.001, P****<0.0001). ns means “not significant”.

Figure 3

IRF9 ChIP-seq analysis revealed similar genome wide binding profile of IRF9 in sgScd2 Th1 cells IFNβ-treated Th1 cells. (A) Projections of PC1 and PC2 for normalized IRF9 ChIP-seq signal intensities of derived from non-treat, IFNβ, IFNβ-treated sgIrf9, mock, sgScd2 and sgScd2/sgIrf9 Th1 cells. Duplicates of each group are shown (ChIP-seq: PC1 77.6%, PC2 12.6%). (B) A heat map depicts the normalized IRF9 ChIP-seq signal intensities relevant to (A). (C) Pie chart showed the distribution and genomic location of transcription factor IRF9, showing the percentage for each genomic location category in IFNβ-treated and sgScd2 Th1 cells. (D) Results of the known motif enrichment analysis de novo motif analysis from IFNβ-treated and sgScd2 Th1 cells. (E, F) GREAT analysis of IRF9 ChIP-seq peaks in IFNβ-treated (E) and sgScd2 (F) Th1 cells. The enriched terms for GO Biological Process and mouse phenotype are shown.

Figure 4

IRF9 in sgScd2 Th1 cells bound to a region that was not bound by IFNβ-treated Th1 cells. (A) Venn diagrams showing the peak count for IRF9 ChIP-seq. Peaks were divided into IFNβ-unique, shared and sgScd2-unique sections. (B, C) Heatmaps and average aggregate plots were generated using deeptools plotHeatmap and plotProfile in non-treat, IFNβ and IFNβ-treated sgIrf9 (B), or mock, sgScd2 and sgScd2/sgIrf9 Th1 cells (C). TSS from 2kb upstream of the transcriptional start site to -2 kb downstream was analyzed. (D, E) Example of shared IRF9 binding sites at the genomic region for Ifi27, Irf7, Irf9, Oasl2 and Rtp4 in non-treat, IFNβ and IFNβ-treated sgIrf9 (D) or mock, sgScd2 and sgScd2/sgIrf9 Th1 cells (E). The arrow indicates the direction of transcription. (F, G) Results of the HOMER de novo motif analysis from IFNβ-unique peaks (F) or sgScd2-unique peaks (G) relevant to (A). (H, I) Example of IRF9 binding sites at the genomic region for Trim26, Herc3 and Asb13 in IFNβ-unique peaks (H) and F830016B08Rik, Gbp2b and Oas2 in sgScd2-unique peaks (I).

Figure 5

Integrated analysis of RNA-seq and IRF9 ChIP-seq was performed to investigate ISG transcriptome regulated by IRF9. (A) Scatter plots depicting IRF9 binding in sgScd2 Th1 cells (x axes) and IFNβ-treated Th1 cells (y axes) for all shared IRF9-bound genes. Genes with signal intensities higher than 5.64 (Log2) are marked with red dots. A heat map shows 1.5-fold upregulated genes in sgScd2 Th1 cells. (B, C) A heat map shows normalized IRF9 ChIP-seq signal intensities relevant to Figure 4A and relative gene expression using qRT-PCR in sgScd2-unique (B) or IFNβ-unique peaks (C). (D) A dot plot of gene expression by RNA-sequencing relevant to Figure 1 compares in sgScd2 Th1 cells/control or sgScd2+ sgIrf9 Th1/control Th1 cells. The genes related to sgScd2-unique, IFNβ-unique peaks or shared were marked with blue, orange or red dots, respectively. (E) A heat map shows relative expression of auto immunity related genes in control, sgScd2 and sgScd2/sgIrf9 Th1 cells. Three technical replicates were performed with quantitative RT-PCR (B, C). Three experiments were performed and showed similar results.