Characterization of the oral and gut microbiome in children with obesity aged 3 to 5 years

Abstract

The ever-increasing global prevalence of obesity has trended towards a younger age. The ecological characteristics and changes of the oral and gut microbial community during childhood are poorly understood.In this study, we analyzed the salivary and fecal microbiota of 30 children with obesity and 30 normal weight children aged 3-5 years via third-generation long-range DNA sequencing,with the aim of understanding the structure of childhood microbiota and identifying specific oral and gut microbial lineages and genera in children that may be associated with obesity.The results revealed significant variation in alpha diversity indices among the four groups (Chao1: P < 0.001; observed species: P < 0.001; Shannon < 0.001). Principal coordinate analysis (PCoA) and nonmetric multidimensional scaling (NMDS) revealed significant differences in oral and gut microbial community structure between obesity and controls. The Firmicutes/Bacteroidetes (F/B) abundance ratios of oral and intestinal flora among children with obesity were higher than those of controls. The most abundant phyla and genera found in oral and intestinal flora were Firmicutes, Proteobacteria, Bacteroidetes, Neisseria, Bacteroides, Faecalibacterium, Streptococcus, Prevotella and so on. Linear discriminant analysis effect size (LEfSe) revealed higher proportions of Filifactor (LDA= 3.98; P < 0.05) and Butyrivibrio (LDA = 2.54; P < 0.001) in the oral microbiota of children with obesity, while the fecal microbiota of children with obesity were more enriched with Faecalibacterium (LDA = 5.02; P < 0.001), Tyzzerella (LDA=3.25; P < 0.01), Klebsiella (LDA = 4.31; P < 0.05),which could be considered as dominant bacterial biomarkers for obesity groups.A total of 148 functional bacterial pathways were found to significantly differ in the oral and gut microbiota among controls and obesity using PICRUSt 2. Most predicted functional pathways were clustered in biosynthesis. In conclusion, This work suggests there were significant differences in oral and gut microbiota in controls and obesity groups, microbiota dysbiosis in childhood might have significant effect on the development of obesity.

Keywords: gut microbiome; high-throughput sequencing; obesity; oral microbiome; preschool children.

Figure 1

Study flowchart.

Figure 2

Alpha and beta diversity of microbiota among the four groups. (A) A violin diagram of Chao 1; (B) Observed-species index data; and (C) Shannon index data. The horizontal line within a box represents the median, a dot indicates an observed value, box margins are interquartile ranges (50% of the observations) and whisker lines extend for 1.5 times the interquartile range. Asterisks (*); (**); and (***) represent significant differences at P < 0.05, P < 0.01, and P < 0.001, respectively. (D) Principal coordinate analysis (PCoA) data of bacterial communities from the four groups. (E) Non-metric dimensional scaling (NMDS) analysis of bacterial β-diversity from the four groups. ns, no significant.

Figure 3

(A) Venn diagram. Taxonomic composition and abundance distribution at (B) phylum; (C) family; (D) genus; and (E) species levels.

Figure 4

(A) Heatmap clustering analysis of bacterial communities among all four groups at the genus level; (B) Principle component analysis (PCA); (C) Orthogonal Partial Least Squares Discriminant Analysis (OPLS-DA).

Figure 5

Taxonomic heatmap at the phylum (A) and genus (B) levels. From left to right: 1) phylogenetic tree map, colored amplicon shared variance (ASV) feature sequences and their connected branches according to taxonomic level; 2) abundance heat map; 3) differential heat map that shows each group with its most abundant species (filled in blue), the significance of the difference between this group and other samples tested; significant differences are denoted with pink as determined via the Wilcoxon rank-sum test (considering a false-discovery rate [FDR]-corrected P value < 0.05 as significant); 4) FDR correction is “BH” multiple test correction (Benjamini Y. and Hochberg Y, 1995). (C) Linear discriminate analysis effect size (LEfSe) taxonomic cladogram. The colored nodes from the inner to outer circles represent the hierarchical relationship of all taxa from phylum to genus levels. Taxa enriched in different groups are shown with different colors; taxa with non-significant changes are colored white. The diameter of each small circle represents taxa abundance. (D) Enriched taxa with linear discriminate analysis (LDA) scores >2 are shown in the histogram. The greater the LDA score was, the more significant the phylotype microbiota was in comparison.

Figure 6

PICRUST2 analysis based on the MetaCyc Pathway Database. (A) Relative abundance of metabolic pathways at levels 1 and 2. (B) Differential analysis of metabolic pathways in oral and gut flora of controls. (C) Differential analysis of metabolic pathways in oral and gut flora of children with obesity.

Figure 7

Composition of phylum involved in metabolic pathways. (A) NAD-BIOSYNTHESIS-II, NAD salvage pathway II. (B) PWY-5837, 1,4-dihydroxy-2-naphthoate biosynthesis I. (C) PWY-5863, superpathway of phylloquinol biosynthesis.