Shifts in the Human Gut Microbiota Structure Caused by Quadruple Helicobacter pylori Eradication Therapy

Abstract

The human gut microbiome plays an important role both in health and disease. Use of antibiotics can alter gut microbiota composition, which can lead to various deleterious events. Here we report a whole genome sequencing metagenomic/genomic study of the intestinal microbiota changes caused by Helicobacter pylori (HP) eradication therapy. Using approaches for metagenomic data analysis we revealed a statistically significant decrease in alpha-diversity and relative abundance of Bifidobacterium adolescentis due to HP eradication therapy, while the relative abundance of Enterococcus faecium increased. We have detected changes in general metagenome resistome profiles as well: after HP eradication therapy, the ermB, CFX group, and tetQ genes were overrepresented, while tetO and tetW genes were underrepresented. We have confirmed these results with genome-resolved metagenomic approaches. MAG (metagenome-assembled genomes) abundance profiles have changed dramatically after HP eradication therapy. Focusing on ermB gene conferring resistance to macrolides, which were included in the HP eradication therapy scheme, we have shown a connection between antibiotic resistance genes (ARGs) and some overrepresented MAGs. Moreover, some E. faecium strains isolated from stool samples obtained after HP eradication have manifested greater antibiotic resistance in vitro in comparison to other isolates, as well as the higher number of ARGs conferring resistance to macrolides and tetracyclines.

Keywords: Helicobacter pylory eradication; antibiotic resistance; enterococci; gut microbiota; horizontal gene transfer; metagenome-assembled genome.

Figure 1

Schematic visualization of experimental design.

Figure 2

The most abundant genera in the gut microbiota of patients. Colors denote relative abundance of genera obtained by MetaPhlAn2 after clr (centered log-ratio) transformation and substitution of zeros (the higher value corresponding to the higher relative abundance). The figure shows the taxa present in at least 25% of the samples. The columns correspond to the samples/patients; the time points (1st—before HP eradication therapy, 2nd—after) are denoted with a top color bar. Hierarchical clustering was performed using the Euclidean distance and complete linkage. OTUs with no genus information are collapsed into the no-name genus.

Figure 3

Changes of gut microbiota composition of patients. (A) Multidimensional scaling biplot of taxonomic profiles (genera level) of patients metagenomic samples before and after antibiotics therapy using Aitchison distance. Samples from the same patient are connected by lines. (B) CoDa dendrogram which characterizes the association of bacterial and bacteriophage families with balances presented as edges. Decomposition of total variance by balances between groups of families is shown using vertical bars. Mean values of balances are shown using anchoring points of vertical bars (according to Pawlowsky-Glahn and Egozcue, 2011). Red color bar denotes the 1st-time point (before HP eradication), whereas the 2nd-time point is shown using a blue color bar. The red color area indicates Balance 1, and Balance 2 is shown using the blue color area.

Figure 4

Change of the MAGs detection value depending on the time point. Each arc corresponds to one patient. Colors represent time points [red—before HP eradication therapy (1st time point), blue and green—after HP eradication therapy (2nd time points—at least 2 days and 3rd—after 2 weeks)]. Detection value (proportion of nucleotides in a contig that are covered at least 1x (according to http://merenlab.org/2017/05/08/anvio-views) was used as an abundance metric, which is shown as color brightness. Taxonomy at phylum level is shown as outer arc.

Figure 5

Discovery of links between MAGs and antibiotic resistance genes. (A) Resistome profiling of four patients gut metagenomes. Color shows individual patients (red—HP003, blue—HP009, green—HP010, yellow—HP028). Time points are denoted by a right color bar. (B) CFX-group gene within transposon-like structure is associated with three MAGs: Bin 38—unknown Bacteroides, Bin 22—Bacteroides dorei, and Bin 4—Bacteroides uniformis. Color shows areas of the MAGs close to the graph sequences according to the BLAST results (patient HP003, time points 2 and 3). (C) Multiple sets of ARGs such as ermB, ermT, Ant6-Ia, Aph-III, Aph7, Sat4A, catA, tetL, tetM, and tetQ (shown by the blue circle and different colors) are located close to Bin 12—Enterococcus faecalis and Bin 22—B. dorei MAGs (patient HP003, time point 2). (D) Relative abundance of E. faecalis and B. dorei in HP003 patient gut metagenomes before and after the HP eradication therapy according to MetaPhlan2 analysis.

Figure 6

Comparative genomic analysis of Enterococcus spp. isolates. Genomes of E. faecium DO and E. faecalis V583 from NCBI were used as references. (A) Detection value of Enterococcus spp. isolates across related metagenomic samples. Isolates from different time points are shown by different colors (red—first point, blue—second point, green—third point). Every arc corresponds to one patient (captions under each plot). (B) Comparison dendrogram (average nucleotide identity, nucleotide identity and KEGG orthology groups distributions). Isolates from the 1st time point are colored red, isolates of 2nd and 3rd-time points are blue and green. The enterococci MAGs of HP003 and HP010 patients correspond to HP003.bin and HP010.bin signatures. (C) Resistome profiling results by GROOT (absence/presence matrix). Cell colors denoted patients' IDs (red—HP003, blue—HP009, green—HP010, yellow—HP028). Time points shown as top color bars. (D) The in vitro antibiotic susceptibility testing results demonstrate an increase in resistance, which is indicated by the transition from blue color to red.