Low abundance of healthy bacteria in the gastric microbiota can be a potential biomarker for gastrointestinal diseases: A pilot study

Sir,

The gastrointestinal (GI) microbiota plays a crucial role in host defence, influencing factors such as GI motility, vitamin production, bile acid metabolism and more1. Dysbiosis disrupts these benefits and external factors such as diet and medication further complicate the situation2. Metagenomics research has linked the GI microbiota to immune and metabolic disorders3. Faecalibacterium, an indicator of intestinal health, was found to be depleted in GI disorders, raising interest in its potential as a probiotic4. The study aims to identify bacterial indicators, such as Faecalibacterium, for GI disorders and explores functional dynamics within the microbiota, providing insights into host–microbe interactions and disease prediction.

This study sequenced genomic DNA of 44 antral gastric biopsies taken during endoscopy by a gastroenterologist at Max Super Specialty Hospital, Vaishali, Ghaziabad, India, between August 2018 to December 2019 to understand the dysbiosis of healthy bacteria in GI disorders such as duodenal ulcer (DU), gastritis and gastro-oesophageal reflux disease (GERD). Ethical clearance and patient’s informed consent were taken. Patient clinical information was recorded. The participants were selected based on the inclusion criteria (age 18-90 yr, informed consent for upper GI endoscopy and clinical signs of Helicobacter pylori related GI disease) and exclusion criteria (no history of antimicrobial drug use or usage of proton-pump inhibitors for three months prior). Two mucosal tissue samples were collected from each participant and placed in sample collection tubes with Brucella broth (Becton Dickinson, Sparks, MD, NJ, USA) and phosphate-buffered saline solution.

Genomic DNA from human tissue biopsy samples was extracted using the QIAamp DNA Mini Kit (QIAGEN GmbH, Hilden, Germany). It was quantified and assessed for quality using agarose gel electrophoresis. A specific primer with a sequencing linker was created for the 16S rRNA V3 and V4 region using the KAPA HiFi HotStart PCR Kit (R&D Cape Town, South Africa) for 26 cycles. Amplified signals were checked on a 1.2 per cent agarose gel. Round 1 PCR amplicons were further amplified (10 cycles) to include Illumina sequencing barcoded adaptors. Round2 PCR amplicons (the sequencing libraries) were examined on a gel. Samples underwent paired-end sequencing on the Illumina MiSeq v3 600-cycle cartridge, focusing on the V3-V4 primer sequences quality bases. Reads were stitched using Fastq-join3 and used for microbiome analysis in the QIIME pipeline. The UCLUST5 method clustered query sequences against the Greengenes 16S rRNA database (v13.8), generating a biome file with taxonomies assigned by the RDP7 classifier at ≥97 per cent sequence similarity. Operational taxonomic units (OTUs) were identified using acquired reads and QIIME scripts (https://qiime2.org/). MicrobiomeAnalyst (https://www.microbiomeanalyst.ca/) was employed for random forest analysis and correlation network construction based on the Spearman correlation coefficient.

Microbial alpha diversity in human tissue biopsy samples was calculated using QIIME 1.7.0, represented by the ACE (Abundance Based Coverage Estimators) index, and rarefaction curves were plotted on R Package-vegan (Version 2.6-4). MicrobiomeAnalyst was utilized for alpha diversity, core microbiome, cluster study, random forest and Spearman rank correlation5. The PICRUSt method was employed to predict microbial community metabolic potential based on 16S rRNA gene data6 referencing the Kyoto Encyclopedia of Genes and Genomes (KEGG) database for functional pathways associated with the bacteriome of diseases7. Graphs displaying the relative abundance of KEGG pathways were generated using statistical analysis of taxonomic and functional profile8.

A total of 44 participants (26 were male and 18 were female) (mean age of 45.5±18.2 yr; Supplementary Table) were included in this study. Of these, those individuals with GERD (n=18) had a mean age of 43±15.8 yr, individuals with DU (n=12) had a mean age of 48.1±12.6 yr, those with gastritis (n=9) had a mean age of 52.5±23.9 yr and the chosen controls (n=5) were 35.4±9.8 yr old. The mean age of the males was 45.3±17.1, whereas for females, it was 45.8±20.8 yr. Control participants were individuals seeking treatment for conditions like liver disease who underwent endoscopic screening as part of their medical evaluation.

The rarefaction curve indicates sufficient sequence depth (

In our initial bacterial population analysis, we compared three groups: control vs. DU, control vs. gastritis and control vs. GERD. Across all three groups, four major phyla were consistently present: Campylobacterota, Firmicutes, Actinobacteria and Bacteroidota (3-71%). Campylobacterota was notably prominent in all disease groups, particularly in gastritis (70%). Firmicutes dominated in control but had lower presence in gastritis, followed by Bacteroidota (13-15%). Proteobacteria and Actinobacteria were prevalent in all disease groups, with gastritis showing a lower abundance of Actinobacteria (4%) compared to control, DU and GERD (Fig. 1A-i, B-i and C-i).

Close

Fig. 1

Composition of gastric microbiota. (A) Relative abundance between control and duodenal ulcer (i) at phylum level (ii) at genus level. (B) Relative abundance between control and gastritis (i) at phylum level (ii) at genus level. (C) Relative abundance between control and GERD (i) at phylum level (ii) at genus level. GERD, gastro-oesophageal reflux disease; DU, duodenal ulcer.

Export to PPT

The top 20 genera were observed, with Helicobacter being notably abundant in all disease groups (10-70%). Controls had higher Streptococcus levels (9-10%), except for DU (18%). Controls featured Catenibacterium (25%), Faecalibacterium (12-16%) and Dialister (5%) as the dominant genera. In disease categories, prevalent genera included Veillonella, Neisseria, Haemophilus, Rothia, Dialister, Gemella and Fusobacterium with relative abundances ranging from 0.5 to 6.75 per cent (Fig. 1A-ii, B-ii and C-ii).

Core bacterial phyla Firmicutes, Proteobacteria, Bacteroidota, Actinobacteria, Fusobacteriota and Campylobacterota were common across all groups. Among 43 core genera, Streptococcus (75%), Veillonella (59%), Staphylococcus (57%), Prevotella 7 (55%), Cutibacterium (52%) and Haemophilus (50%) dominated. In the control group, Streptococcus, Catenibacterium (80%), Veillonella, Rothia, Prevotella 7 (60%), respectively and Faecalibacterium (40%) were prevalent (

Genus-level analysis revealed positive correlations of Fusobacterium, Veillonella and Neisseria with disease groups, indicating their potential involvement in disease-related microbial dynamics. Conversely, Faecalibacterium, Catenibacterium and Dialister showed negative correlations, suggesting their roles in maintaining a balanced microbial profile (Fig. 2A). In this study, Faecalibacterium was found to have positive correlations with Catenibacterium, Dialister, Rothia and Gemella, while it showed a negative correlation with Fusobacterium (Fig. 2B). Dialister also had positive correlations with Faecalibacterium, Corynebacterium and Catenibacterium but a negative correlation with Fusobacterium (Fig. 2C). In addition, Fusobacterium was found to have positive correlations with Veillonella and Gemella (Fig. 2D).

Close

Fig. 2

Pattern search analysis: (A) Top 25 genus associated with the control and disease groups (B) Top 25 genus correlated with Faecalibacterium (C) Top 25 genus associated with Dialister (D) Genera associated with Fusobacterium.

Export to PPT

Single-factor statistical analysis identified five significant features (P ≤0.05) in taxon abundances: Fusobacterium (P=0.03), Catenibacterium (P=0.041), Faecalibacterium (P=0.044), Dialister (P=0.047) and Prevotella_7 (P=0.05). Faecalibacterium were more abundant in control and DU compared to gastritis and GERD (

The study utilized 16S rDNA sequencing data and Greengenes annotated OTUs to predict the metabolic capabilities of the bacteriome. Using KEGG Orthology, metagenome contributions were determined for control and disease categories. A total of 500 KO IDs were identified in both disease and control groups, revealing several significant KEGG metabolic pathways when comparing control to disease groups. Methane metabolism was significantly higher among all disease groups. In DU, five pathways were notably elevated (

This study underscores the link between GI microbiota and overall well-being, investigating beneficial bacteria and their connection to disease-associated microbes. Biopsy samples revealed varying species richness, with GERD having the highest, followed by gastritis and DU, aligning with earlier observations9. This study found that the phyla Bacteroidetes and Firmicutes were predominant in the GI microbiota of the controls similar to a previous study10. There are also reports on Firmicutes being higher in control group11 while Actinobacteria were lower in individuals with gastritis12. Faecalibacterium which belongs to phylum Firmicutes, known for its anti-H. pylori properties, was more abundant in control in the present study similar to a prior report3. The GC-associated bacteria (Veillonella and Fusobacterium) and Faecalibacterium have a negative correlation with one another, according to correlation analysis. This demonstrates how an increase in the abundance of Fusobacterium can result in a decrease in the number of Faecalibacterium, a type of beneficial bacterium that is supported by previous research13. Individuals with gastritis and GERD exhibited lower SCFA-producing bacteria, including Faecalibacterium, and reduced overall SCFA levels, making it a potential therapeutic and biomarker for GI diseases14. This study shows the importance of a balanced host-microorganism relationship for optimal health and disease prevention. It suggests investigating microbial imbalances, such as Faecalibacterium, Catenibacterium and Dialister, as potential markers for various GI disorders. In addition, certain microflora like Fusobacterium may serve as diagnostic indicators for GI diseases. Integrating Faecalibacterium as a biomarker could transform disease management, offering personalized interventions based on an individuals gut microbiome. The research hints at the impact of gastric microbiota on the nervous system, affecting brain development and function, with potential implications for preventing immune-related and neuropsychiatric disorders in the future15. However, limitations, such as a small sample size and lack of healthy controls, are acknowledged, emphasizing the need for larger, diverse studies to validate findings and explore these interactions further.