Eradication of Helicobacter pylori in Children Restores the Structure of the Gastric Bacterial Community to That of Non-infected Children

Introduction

H. pylori alters the gastric microbial composition and causes reduced gastric inflammatory disease in children in Chile^1–3. These findings suggest that H. pylori-associated gastric microbiota may influence the pathogenesis of H. pylori infection beginning in childhood, at least in Chile. Here we characterized the gastric microbiota in H. pylori-infected children in an entirely different geographic region of South America, namely Venezuela, where H. pylori also is endemic⁴, and determined the impact of H. pylori eradication on bacterial community structure.

Methods

Sixteen children (≤13 years) with nausea and abdominal discomfort without antibiotic or antacid therapy during the preceding month underwent endoscopic antral biopsy for H. pylori urease testing and histology (both required positive for infection) and microbiota analysis. Eleven subjects were infected and treated with amoxicillin, clarithromycin and omeprazole for 14 days. Two months later, the children were re-biopsied for H. pylori testing and microbiota analysis. Biopsy DNA was isolated, the V4 region of the 16S rRNA bacterial gene was amplified by PCR, and 250 base single end reads were sequenced and analyzed, as previously described³.

Results

The overall gastric bacterial community structure (β-diversity) in the H. pylori-infected children was different from that of non-infected children using unweighted UniFrac (P=.002) (Figure 1A), weighted UniFrac (P=.005) and Bray-Curtis analyses (P=.006). This finding suggested that H. pylori impacted the gastric commensal bacterial composition in infected children and prompted us to determine whether difference in the abundance of taxa with a prevalence >1% correlated with difference in the bacterial communities between infected and non-infected children. The relative abundance of taxa in the H. pylori-infected children was significantly reduced at the class (3 taxa), order (4 taxa), family (5 taxa) and genera (5 taxa) levels compared with taxa harbored by non-infected children (P=.01-.04), as determined by the Kruskal-Wallis test (Figure 1B). Since reduced gut microbial diversity is associated with certain disease processes, we determined the relative phylotype abundances (α-diversity) for the gastric microbiota. H. pylori-infected children harbored significantly less diverse gastric bacterial communities compared with non-infected children by both Shannon (P=.008) and Simpson (P=.02) indices (Figure 1C). These findings suggested that H. pylori impacted the gastric microbial community structure of children residing in this northern region of South America.

Figure 1.

Gastric microbiota in Venezuelan children is modified by H. pylori infection and reverts to that of non-infected children after H. pylori eradication. (A) β-diversity of the gastric microbiota in H. pylori-infected (n=11) and non-infected children (n=5) determined by permutational multivariate analysis of variance (PERMANOVA) and illustrated by unweighted UniFrac distances with each dot representing one subject in the principal coordinate analysis (PCoA) presented in 2-dimensional plots. P value determined by PERMANOVA test in QIIME. (B) Taxa abundances (when >1% of total bacterial DNA) at the class, order, family and genera levels among H. pylori-infected and non-infected children. P values determined by Kruskal-Wallis test with multiple comparison correction and false discovery rate analysis. When a genus could not be assigned, family is listed. (C) α-diversity of the gastric microbiota in H. pylori-infected (n=11) and non-infected children (n=5) using Shannon and Simpson indices. P values determined by Student’s t-test. (D) Taxa abundances of the 10 most abundant gastric bacteria (when >1% of total bacterial DNA) in children without H. pylori (5) and children in whom H. pylori was eradicated (n=7). P>.05, Kruskal-Wallis test. (E) α-diversity of the microbiota in H. pylori-infected children before (n=11) and after successful H. pylori eradication (n=7) using Shannon and Simpson indices. P values determined by Student’s t-test.

To substantiate these findings, we next analyzed the gastric microbial composition in H. pylori-infected children whose H. pylori was successfully eradicated by antibiotic treatment (7 of the 11 treated children). The bacterial community structure after H. pylori clearance was not significantly different from that of non-infected children (β-diversity) based on unweighted UniFrac analysis (P=.107), as well as weighted UniFrac analysis (P=.439) and Bray-Curtis analysis (P=.260). Overall, the relative proportion of the 10 most abundant bacterial taxa after H. pylori eradication did not differ significantly from that of non-infected children (Kruskal-Wallis test; P>0.5) (Figure 1D). Kocuria, an environmental Gram-positive bacteria that may inhabit skin and mucus membranes, is typically amoxicillin-resistant, and may cause systemic infection in immunocompromised hosts, dominated the microbiota in a single child. Finally, gastric bacterial diversity increased significantly after clearance of H. pylori, based on comparisons of Shannon and Simpson indices (both P=.04) (Figure 1E), reflecting the more diverse bacterial composition after H. pylori eradication. Thus, eradication of H. pylori in infected Venezuelan children was associated with restoration of the gastric microbiota to the community structure of non-infected children.

Discussion

This first characterization of the gastric microbiota in Venezuelan children is important because the microbial composition associated with H. pylori infection may impact the risk for gastric disease sequelae, as shown for adults in mountainous Pacific South America⁵. Further, in countries with a high prevalence of H. pylori, gastric cancer in adults is strongly related to acquiring H. pylori in childhood. However, children have reduced levels of H. pylori-associated gastric inflammation, the key risk factor for gastric cancer, compared with infected adults^1,2. Indeed, gastric cancer in children is exceptionally rare, although H. pylori is typically acquired in early childhood. That H. pylori infection in children is associated with an altered gastric microbiome, as we previously reported for Chilean children³ and confirmed here, raises the possibility that the gastric microbiota in infected children may contribute to mucosal changes such as increased regulatory T-cell responses in infected children^1–3. In a mouse model, H. pylori accelerates the pathogenesis of gastric cancer ⁶ and has more impact on the abundance of specific bacterial taxa in younger than older mice⁷. In addition, H. pylori infection in neonatal, but not adult, mice promotes tolerance to H. pylori and protects against preneoplastic lesions⁸.

The complex consortia of generally non-cultivatable bacteria colonizing the stomach requires a molecular approach to identify these microbial communities. Using 16S rDNA sequencing, we show that H. pylori clearance restores the gastric microbial composition to the community structure of non-infected children, suggesting H. pylori eradication impacts the associated microbiota. Thus, should the gastric microbial composition altered by H. pylori contribute to H. pylori-associated disease, eradication of H. pylori with restoration of the microbial community structure to that of non-infected children may modify infection sequelae. Notably, the class, order and family of bacterial taxa impacted by the presence of H. pylori in the Venezuelan children is different from the taxa reported to be impacted by H. pylori in Chilean children³, except for reduced abundance of the acid tolerant order Lactobacillales in both populations. The more prevalent bacterial genera (>1% abundance) associated with H. pylori also did not overlap in the Venezuelan and Chilean children. The lack of concordance in the gastric microbial community structure among children in these two populations may reflect the impact of different diets, nutritional status and geographic location. However, both studies demonstrate that the presence of H. pylori substantially alters the gastric microbiome in children. While acknowledging the small size of our study population, we provide a starting point for future studies on the clinical impact of restoring the microbiota to that of non-infected children. Thus, characterization of the gastric microbiota in pediatric H. pylori infection in endemic regions of the world may provide insight into the role of the microbiome in disease pathogenesis.

Abbreviations

rDNA

ribosomal DNA

PCoA

principal coordinate analysis

QIIME

quantitative insight into microbial ecology

OTU

operational taxonomic unit

PERMANOVA

permutational multivariate analysis of variance