Pediatric migraine is characterized by traits of ecological and metabolic dysbiosis and inflammation

Abstract

Background: Recently, there has been increasing interest in the possible role of the gut microbiota (GM) in the onset of migraine. Our aim was to verify whether bacterial populations associated with intestinal dysbiosis are found in pediatric patients with migraine. We looked for which metabolic pathways, these bacteria were involved and whether they might be associated with gut inflammation and increased intestinal permeability.

Methods: Patients aged between 6 and 17 years were recruited. The GM profiling was performed by the 16S rRNA metataxonomics of faecal samples from 98 patients with migraine and 98 healthy subjects. Alpha and beta diversity analyses and multivariate and univariate analyses were applied to compare the gut microbiota profiles between the two group. To predict functional metabolic pathways, we used phylogenetic analysis of communities. The level of indican in urine was analyzed to investigate the presence of metabolic dysbiosis. To assess gut inflammation, increased intestinal permeability and the mucosal immune activation, we measured the plasmatic levels of lipopolysaccharide, occludin and IgA, respectively.

Results: The α-diversity analysis revealed a significant increase of bacterial richness in the migraine group. The β-diversity analysis showed significant differences between the two groups indicating gut dysbiosis in patients with migraine. Thirty-seven metabolic pathways were increased in the migraine group, which includes changes in tryptophan and phenylalanine metabolism. The presence of metabolic dysbiosis was confirmed by the increased level of indican in urine. Increased levels of plasmatic occludin and IgA indicated the presence of intestinal permeability and mucosal immune activation. The plasmatic LPS levels showed a low intestinal inflammation in patients with migraine.

Conclusions: Pediatric patients with migraine present GM profiles different from healthy subjects, associated with metabolic pathways important in migraine.

Keywords: Dysbiosis; Gut microbiota; migraine; tryptophan; Gut permeability; IgA; Indican; Inflammation; Lipopolysaccharide; Occludin.

Fig. 1

Gut microbiota biodiversity. A α-diversity analysis based on the Shannon-Weiner, Simpson and Chao1 indexes. The Mann–Whitney test was applied for the comparisons between migraine patients (MIMIC) and control subjects (CTRL). B β-diversity was calculated by the Bray–Curtis algorithm and presented by PCoA plots. PERMANOVA test results are statistically significant, p-value = 0.001

Fig. 2

Compositional analysis of migraine patients (MIMIC) and control subjects (CTRL) gut microbiota at genus level. A Unsupervised multivariate analysis. Principal Component Analysis [PCA] plot showing the top 25 loadings; B Supervised multivariate analysis plot (Partial Least Squares Discriminant Analysis [PLS-DA]); C Loading variables plot (filtered for VIP > 0.85). The Root Mean Square Error (RMSE) = 0.247, R² value = 0.759, p-value ≤ 0.05 and Q² = 0.632, p value ≤ 0.05; D Univariate analysis (Linear Discriminant Analysis [LDA] Effect Size [LEfSe] plot. Bacterial taxa enriched in migraine patients have positive LDA scores (red), while bacterial enriched in CTRL have negative scores (green). E Receiver operating characteristic (ROC) analysis of the PLS-DA model. The value of AUROC = 0.969 indicates a high accuracy of the prediction model

Fig. 3

PICRUSt2 functional prediction using the KEGG pathway Database. Predicted metabolic pathways statistically associated to migraine patients (MIMIC) and control subjects (CTRL). Red bars represent pathways increased in MIMIC; green bars represent pathways increased in CTRL. LDA, linear discriminant analysis

Fig. 4

Bacterial and metabolic pathways correlation network in migraine patients (MIMIC) (A) and in control subjects (CTRL) (B). Each node represents bacteria (orange circles) and metabolic pathways (blue circles). Green and red edges indicate positive and negative correlation values, respectively. Only correlations statistically significant (p-value < 0.05) are reported

Fig. 5

Plasmatic levels of LPS, occludin, IgA and urinary level of indican for patients with migraine (MIMIC). Bar plots of LPS (ng/ml), occludin (ng/ml) and IgA (g/L) plasmatic concentrations and of urinary indican concentration (mg/L)