The Proinflammatory Role of Guanylate-Binding Protein 5 in Inflammatory Bowel Diseases

Abstract

NLRP3 inflammasome is implicated in the pathogenesis of inflammatory bowel diseases (IBD). Since guanylate-binding protein 5 (GBP5) induces the NLRP3 inflammasome activity, we aim to investigate the potential role of GBP5 in IBD pathogenesis. The expression of GBP5, NLRP3 inflammasome, and related cytokines and chemokines was examined in two cohorts of IBD patients and healthy controls, by microarray transcriptome analysis and quantitative real-time PCR. Cellular localization of GBP5 in colonic biopsies was examined by immunohistochemistry and immunofluorescence with confocal microscopy. For functional studies, GBP5 was induced by interferon γ or silenced by siRNA or CRISPR/CAS9 technique, and inflammatory activities were evaluated at mRNA and protein levels. We found that the expression of GBP5 was elevated in colonic mucosa in two geographically and culturally distinct IBD cohorts. In colonic tissues of IBD patients, GBP5 expression was mainly confined to immune cells and the levels of GBP5 expression were correlated with those of the inflammatory cytokines and chemokines. In cultured T and macrophage cells, the expression of proinflammatory cytokines and chemokines was increased when GBP5 was induced, while GBP5 deficiency leads to decreased expression of proinflammatory mediators including gasdermin D, caspase 1, cytokines, and chemokines. We conclude that GBP5 is required in the expression of many proinflammatory cytokines and chemokines in intestinal immune cells. In addition, GBP5 may upregulate inflammatory reactions through an inflammasome-mediated mechanism. Since GBP5 plays a proinflammatory role at the early steps of the inflammatory cascades of IBD pathogenesis, and is implicated in IBD patients of distinct genetic and environmental backgrounds, targeting GBP5 could be an effective strategy for the management of IBD.

Keywords: Crohn’s disease; guanylate binding protein; inflammasome; pyroptosis; ulcerative colitis.

FIGURE 1

GBP5 is highly expressed in the inflamed intestinal tissue of IBD patients. (A) Messenger RNA expression of GBP5 in the colonic mucosa of healthy controls (n = 6), patients with Crohn’s disease (CD, n = 19), and ulcerative colitis (UC, n = 24). Data are from a microarray dataset generated from a European cohort. **P < 0.01, Dunn’s multiple comparison test. (B) Quantitative RT-PCR analysis of GBP5 mRNA in colonic mucosa from a Chinese cohort including healthy controls (n = 35), and patients with CD (n = 38) and UC (n = 17). ***P < 0.001, Dunn’s multiple comparison test. (C) Immunohistochemical staining of GBP5 in colonic mucosa from four representative patients with CD. Images of inflamed and non-inflamed sites from the same patient are compared side by side. Bar = 50 μm. (D) Quantitation of the GBP5 staining in the inflamed and non-inflamed sites in (C). **P < 0.01, paired Student’s t-test. (E,F) Spearman’s correlation analysis of the mucosal GBP5 expression levels and endoscopic severities. GBP5 expression levels are from quantitative RT-PCR results. SES-CD, simple endoscopic score for Crohn’s disease; UC-EIS, ulcerative colitis endoscopic index of severity.

FIGURE 2

Immune cell-specific GBP5 expression in patients with CD. (A) Immunohistochemical staining of a colonic biopsy section from a representative CD patient with an anti-GBP5 antibody: an inflamed area. The top image shows a panoramic view of the section. Bar = 2 mm. Details are shown for boxed areas representing different anatomic structures. I, mucosa; II, muscularis mucosa and submucosa; III, circular muscle; IV, longitudinal muscle; V, serosa; VI, mesentery, black arrow indicates a blood vessel; VII, Peyer’s patch in mucosa; VIII, lymph node in submucosa. Bar = 50μm. (B) Immunohistochemical staining of a colonic biopsy section from a representative CD patient with an anti-GBP5 antibody: a non-inflamed area. The top image shows a panoramic view of the section. Bar = 2 mm. Details are shown for boxed areas representing different anatomic structures. I, mucosa; II, lamina propria, muscularis mucosa and submucosa; III, a blood vessel in submucosae layer; IV, circular muscle and longitudinal muscle; V, serosa; VI, mesentery. Bar = 50μm. (C) Immunofluorescence staining of inflamed colon tissue from a representative CD patient with antibodies against GBP5, CD3, CD40, and CD68. White arrow indicates the overlap of GBP5 and CD68. Bar = 50 μm.

FIGURE 3

GBP5 is implicated in inflammatory processes in IBD according to transcriptome analysis. (A) Heatmap for mRNA expression levels of GBP5 (arrow), available cytokines, and chemokines based on transcriptome data generated from colonic mucosal biopsies of healthy controls (n = 6), patients with Crohn’s disease (CD, n = 19), and ulcerative colitis (UC, n = 24). Unsupervised hierarchical clustering of GBP5, cytokine, and chemokine genes was performed. Genes in the red box are more closely clustered with GBP5. (B) Venn plot of the number of genes associated with GBP1, GBP2, GBP4, and GBP5, based on their mRNA expression levels. Associations with Pearson’s correlation coefficient no less than 0.6 are counted. (C) Gene Ontology (GO) analysis of GBP5 associated genes. The cutoff value for the input gene list is Pearson’s correlation coefficient no less than 0.6. BP, biological process; CC, cellular compartment; MF, molecular function. All samples from GSE16879 are included (n = 90).

FIGURE 4

GBP5 knockout downregulated the expression of proinflammatory mediator genes in THP-1 cells. The transcriptomes of wild-type and GBP5^–/– THP-1 cells, with or without stimulation [IFNγ and lipopolysaccharide (LPS)], respectively, were analyzed by RNAseq technique. N = 3 for each treatment group. (A) Generation of GBP5 knockout (GBP5^–/–). THP-1 cell line by CRISPR/Cas9-mediated genome editing. The two CRISPR target sites are highlighted in red. Loss of GBP5 was confirmed by Western blot. WT, wild-type control. (B) Messenger RNA expression of GBP5 and related genes in wild-type and GBP5^–/– THP-1 cells (clone B2), with or without stimulation (IFNγ and LPS), respectively. FPKM, fragments per kilobase million. ***P < 0.001; ****P < 0.0001; unpaired Student’s t-test. (C) Differential gene expression between wild-type and GBP5^–/– THP-1 cells: volcano plot. Vertical and horizontal dashed lines indicate the cutoff values for differentially expressed genes:| Log₂ (Fold Change)| > 1 and p.adjust < 0.05. (D) GO (Gene Ontology) analysis of downregulated genes in GBP5 knockout cells. The cutoff values for selecting differentially expressed genes for the input gene list: Log₂(Fold Change) < -2, and p. adjust < 0.05. BP, biological process; CC, cellular compartment; MF, molecular function. (E) Gene set enrichment analysis (GSEA) of GBP5^–/– THP-1 cells (clone B2) transcriptome compared to wild-type THP-1 cells. Upper panel: the enrichment score curves of the top KEGG (Kyoto Encyclopedia of Genes and Genomes) pathways exhibiting decreased expression in GBP5^–/– THP-1 cells: immune related pathways. Middle panel: distribution of the genes related to the pathways indicated in the upper panel. The genes were ranked according to their differential expression between wild-type and GBP5^–/– THP-1 cells. Genes of higher rank (left) exhibit relatively higher expression in wild-type THP-1 cells. Lower panel: Graphical representation of the correlations of the gene expression levels with the phenotypes: wild-type or GBP5 knockout. Genes on the right are more negatively correlated with GBP5 deficiency.

FIGURE 5

Decreased chemokine and cytokine secretion in GBP5 knockout THP-1 cells. The protein levels of chemokines and cytokines in the cell culture supernatant of wild-type (WT) and GBP5 knockout (KO) THP-1 cells are plotted. Cells were primed with IFNγ and LPS before sample collection. The concentrations of cytokines and chemokines were determined by Luminex liquid suspension chip. Data were normalized as (x-mean)/SD. Gray blocks indicate no detection. P-values were from Student’s t-tests.

FIGURE 6

Diminished expression of inflammasome related proteins in GBP5 knockout THP-1 cells was subjected to wild-type THP-1 cells, and GBP5 knockout THP-1 cell clones (A1, B1, and B2) were left untreated (NT), treated with IFNγ (25 ng/ml) only, or treated with IFNγ (25 ng/ml) plus LPS (500 ng/ml) for 16 h, before Western blot analysis with antibodies against GBP5, gasdermin D (GSDMD), caspase 1 (CASP1), pro-IL1β, and GAPDH, respectively.