Eggerthella lenta down regulated flavone and flavonol biosynthesis promoted Kawasaki disease

Abstract

Kawasaki Disease (KD) is a multisystemic vasculitis of unknown aetiology in children. The incidence of KD varies by geographic area and correlates with differences in gut microbiota patterns, with the highest incidence in Asian. This study aimed to investigate alterations in faecal microbiota and assess their relationship with systemic inflammation in KD patients. A total of 59 patients and 55 matched controls were included. Fecal samples were collected at the onset of KD. The V3/V4 regions of 16S rDNA were sequenced using the MiSeq platform. PICRUSt 2 was used to analyse the potential functional pathways involved in gut dysbiosis. Alpha (p < 0.042) and beta (p < 0.001) diversity in KD were significantly decreased when compared to the control group. After multivariate regression, among the seven critical microbes, increased Eggerthella lenta (p = 0.016) and decreased Bacteroides ovatus (p = 0.014) could also predict KD risk using receiver operating characteristic curve (ROC) analysis (Eggerthella lenta: area under the ROC curve, AUC = 0.841, odds ratio = 23.956; Bacteroides ovatus: AUC = 0.816, odds ratio = 31.365). Notably, Bacteroides ovatus was positively correlated with blood segment cells (p = 0.006), but negatively correlated with blood lymphocytes (p = 0.013). After multivariate regression, flavone and flavonol biosynthesis decreased in children with KD (p < 0.001). Our results indicated that both Bacteroides ovatus and Eggerthella lenta may deregulate flavone and flavonol biosynthesis, consequently modulating immune cells and potentially triggering KD. This study suggests that alterations in the gut microbiota are closely associated with immune responses and provides a new perspective on the aetiology, pathogenesis, and treatment of KD.

Keywords: Bacteroides ovatus; Eggerthella lenta; Kawasaki disease; flavone; flavonol.

Graphical abstract

Figure 1.

Gut microbiota composition change in Kawasaki disease (KD) patients. (a) Shannon alpha diversity (including richness and evenness) was significantly decreased in Kawasaki disease patients (p = 0.042). (b) Jaccard beta diversity bacteria composition was significantly changed in Kawasaki disease (p < 0.001). A p value less than 0.05 was considered statistically significant.

Figure 2.

Core microbiota analysis between healthy control and Kawasaki disease. (a) the cladogram plots of core microbiota analysis. Red colour is presented healthy control. Green colour is presented Kawasaki disease patients. (b) LEfSe bar plots shown significantly changed gut microbiota in healthy control and Kawasaki disease. The cut-off value of LDA score is 4.0.

Figure 3.

Eggerthella lenta and Bacteroides ovatus are significantly changed in Kawasaki disease. (a) receiver operating characteristic (ROC) curve were used to analyse the Eggerthella lenta and Bacteroides ovatus and get the cut-off value (>0.041 and < 0.026, respectively) and area under curve (0.841 and 0.816, respectively). (b) the correlation plot between Eggerthella lenta and Bacteroides ovatus. Eggerthella lenta shown negative correlation with age and Bacteroides ovatus (r = −0.284, p = 0.002, and r = −0.326, p < 0.001, respectively). Bacteroides ovatus shown positive correlation with blood segments cells (r = 0.352, p = 0.006) and negative correlation with blood lymphocytes (r = −0.323, p = 0.013). A p value less than 0.05 was considered statistically significant.

Figure 4.

The critical signalling pathway in Kawasaki disease is flavone and flavonol biosynthesis. a. functional pathways analysis by PICRUSt and false positive discovery rate (FDR) analysis was utilized by Benjamini-Hochberg method and adj p-value (q value) less than 0.05 were consider as statistically significant. b. flavone and flavonol biosynthesis was downregulated in Kawasaki disease. c. flavone and flavonol biosynthesis shown negative correlation with Eggerthella lenta and blood lymphocytes (r = −0.267, p = 0.004 and r = −0.354, p = 0.006, respectively); and shown positive correlation with bacteroides ovatus and blood segments cells (r = 0.781, p < 0.001, and r = 0.344, p = 0.009, respectively). d. flavone and flavonol biosynthesis shown positive correlation with C-reactive protein (CRP, r = 0.253, p = 0.050). A p value less than 0.05 was considered statistically significant.

Figure 5.

Decision tree analysis for Kawasaki disease prediction based on the Eggerthella lenta/Bacteroides ovatus ratio. the root node shows 114 samples (55 controls and 59 KD patients), with a split at a threshold ratio of 0.011 (adjusted p < 0.001, chi-square = 55.284, df = 1). For samples with a ratio of 0.011 or lower, 82.1% are classified as controls (node 1), while for samples with a ratio greater than 0.011, 84.5% are classified as KD patients (node 2). A p value less than 0.05 was considered statistically significant.