Differences in the intestinal microbiome of healthy children and patients with newly diagnosed Crohn's disease

Abstract

The aetiology of inflammatory bowel diseases (IBD) seems to be strongly connected to changes in the enteral microbiome. The dysbiosis pattern seen in Crohn's disease (CD) differs among published studies depending on patients' age, disease phenotype and microbiome research methods. The aims was to investigate microbiome in treatment-naive paediatric patients to get an insight into its structure at the early stage of the disease in comparison to healthy. Stool samples were obtained from controls and newly diagnosed patients prior to any intervention. Microbiota was analysed by 16SrRNAnext-generation-sequencing (NGS). Differences in the within-sample phylotype richness and evenness (alpha diversity) were detected between controls and patients. Statistically significant dissimilarities between samples were present for all used metrics. We also found a significant increase in the abundance of OTUs of the Enterococcus genus and reduction in, among others, Bifidobacterium (B. adolescentis), Roseburia (R.faecis), Faecalibacterium (F. prausnitzii), Gemmiger (G. formicilis), Ruminococcus (R. bromii) and Veillonellaceae (Dialister). Moreover, differences in alpha and beta diversities in respect to calprotectin and PCDAI were observed: patients with calprotectin <100 µg/g and with PCDAI below 10 points vs those with calprotectin >100 µg/g and mild (10-27.7 points), moderate (27.5-40 points) or severe (>40 points) CD disease activity had higher richness and diversity of gut microbiota. The results of our study highlight reduced diversity and dysbiosis at the earliest stage of the disease. Microbial imbalance and low abundance of butyrate-producing bacteria, including Bifidobacterium adolescentis, may suggest benefits of microbial modification therapy.

Figure 1

Alpha diversity analysis of control and CD patients. Within-sample diversity measured by Shannon index (A), observed OTUs (B), Faith’s phylogenetic diversity (C) and Pielou’s measure of species evenness (D). Kruskal-Wallis test was performed to analyse statistical significance.

Figure 2

PCoA 2D plots of beta diversity analysis of control and CD patients. Between-sample dissimilarities were measured by Jaccard distances (A), Bray-Curtis distance (B), unweighted UniFrac distances (C) and weighted UniFrac distances (D). Permutational multivariate analysis of variance (PERMANOVA) was performed to analyse statistical significance.

Figure 3

Composition of the bacterial community at the genus (L6) level for control and CD samples. Relative abundance of top 20 genera are shown.

Figure 4

Volcano plots of differentially expressed OTUs between control and CD patients at genus (A) and species (B) levels. ANCOM analysis was performed to analyse a statistical significance. Statistically significant OTUs are represented as red rounds.

Figure 5

Alpha diversity analysis of control and CD patients with different level of calprotectin. Within-sample diversity measured by Shannon index (A), observed OTUs (B), Faith’s phylogenetic diversity (C) and Pielou’s measure of species evenness (D). Kruskal-Wallis with Post-hoc was performed to analyse statistical significance. Statistically significant values between control and other groups were represented as “#”. Statistically significant values between PCDAI < 10 group and with other groups: mild (10–27.5 points; 21 patients), moderate (>27.5–39 points, 17 patients) and severe (>40 points; 21 patients) CD disease activity (Supplementary Fig. S1 online) were represented as “*”.

Figure 6

PCoA 2D plots of beta diversity analysis of control and CD patients with different level of calprotectin. Between-sample dissimilarities were measured by Jaccard distances (A), Bray-Curtis distance (B), unweighted UniFrac distances (C) and weighted UniFrac distances (D).