Society: AGA
Microbiome science in the gut has focused on the study of colonic, and more frequently, fecal communities. While the upper gastrointestinal tract has significantly lower bacterial density than the colon, it still harbors a dynamic microbial community. Comparatively little is known about microbiota structure and function in the oral cavity, gastric and small intestine. This is counterintuitive as the upper GI tract is the place for the establishment of oral tolerance, a main site for carbohydrate, lipid and protein digestion and metabolism. This process releases active metabolites that will directly impact host tissues. Indeed, emerging studies in model diseases such as celiac disease and environmental enteropathy, that primarily affect the small intestine, have demonstrated precise diet-opportunistic pathogen interactions that may affect disease onset and severity. More importantly, understanding of these mechanisms can inform the development of targeted microbial therapeutics, that could also benefit other inflammatory disorders.
Objective
Non-steroidal anti-inflammatory drugs (NSAIDs), including aspirin, are widely used clinically. However, there are many side effects. In addition to gastric and duodenal ulcers, they can especially cause small intestinal injury, with an accidence rate as high as 70%. Using metagenomic sequencing, we identified the reduction of Lactobacillus reuteri in patients with NSAID-induced small intestinal injury. We aimed to evaluate the potential treatment role of L. reuteri in small intestinal injury.
Method
The potential treatment role of L. reuteri was assessed in murine models of indomethacin. Goblet cell line LS174T and mouse small intestinal organoids were respectively co-cultured with cultured with L. reuteri or Escherichia coli K12 (E. coli K12) culture-supernatant to evaluate the expression of mucin 2 (MUC2) and related biomarkers of intestinal secretory lineage cell. Gut microbiota was assessed by 16S ribosomal RNA sequencing. The effective molecule produced from L. reuteri was identified by liquid chromatography mass spectrometry (LC-MS/MS) and targeted mass spectrometry.
Result
L. reuteri significantly alleviated mucosal injury and inflammatory cell infiltration compared with E. coli K12 and phosphate-buffered saline in indomethacin treated mice. Ileocecal microbial profiling revealed enrichment of probiotics and reduction of pathogenic bacteria in L. reuteri-treated mice. Culturing LS174T cell line with L. reuteri culture-supernatant (0.5%,1%,5%) concentration-dependently promoted the expression of MUC2 and ATOH1/SPDEF. Besides. reuteri culture-supernatant simulates the differentiation of intestinal secretory cells including goblet and Paneth cell in mouse small intestinal organoids and mice. Only L. reuteri with fraction size <3kDa suppressed proliferation in LS174T cell. Using LC-MS/MS, enrichment of indole-3-lactic acid (ILA) was identified in both L. reuteri culture-supernatant and gut of L. reuteri-treated mice. Moreover, ILA promoted expression of MUC2 in vitro and alleviated the indomethacin-induced small intestinal injury in mice.
Conclusion
L. reuteri protects against indomethacin-induced small intestinal injury by producing protective metabolite which promotes intestinal secretory cell differentiation and protect mucosal barrier.
Small intestinal bacterial overgrowth (SIBO) is characterized by excessive small bowel coliforms, and associated with symptoms such as diarrhea, bloating, and abdominal pain. Recent duodenal 16S rRNA sequencing narrowed the microbes in SIBO to two main genera, Escherichia and Klebsiella, but the specific species/strains were undetermined. Here, we characterize for the first time the specific bacterial species/strains associated with SIBO and its symptoms via shotgun sequencing of the duodenal luminal microbiome.
Methods: Subjects undergoing esophagogastroduodenoscopy without colon prep were prospectively recruited and completed a GI symptoms severity questionnaire. Duodenal luminal aspirates were obtained using a double lumen protected catheter and plated on MacConkey agar (MAC). SIBO was defined as ≥103 CFU/mL. Aspirate DNAs were extracted and shotgun and 16S sequencing were performed on NovaSeq and MiSeq Illumina platforms, respectively. Analyses were carried out using CLC Genomics Microbial Modules. Associations were determined using Spearman tests.
Results: 21 non-SIBO subjects (12 female (F), age=57±12 yrs, BMI=24.7±5.6 kg/m2) and 17 SIBO subjects (11 F; age=61±16 yrs; BMI=26.5±6.6 kg/m2) (P=NS) were included. Duodenal microbial profiles were significantly different between groups (Bray-Curtis adj-p-value<0.05), regardless of sequencing technique. Shotgun sequencing showed that specific species/strains of Escherichia and Klebsiella (both family Enterobacteriaceae) were greatly increased in SIBO vs. non-SIBO subjects, including E. coli (fold change (FC)=6.9, adj-p=6.46E-5), K. pneumoniae (FC=3.79, adj-p=2.40E-3), and K. aerogenes (FC=5.09, adj-p=2.63E-6), as well as Enterobacter spp (FC=5.83, adj-p=1.03E-3). Remarkably, Escherichia and Klebsiella species/strains accounted for 40.24% of all duodenal microbes in SIBO subjects, vs. 5.6% in non-SIBO (Fig 1). Increased abundance of these specific microbes was associated with significantly decreased duodenal microbial diversity (559 taxa in non-SIBO vs 476 taxa in SIBO, Fig 1). Strains of E. coli sharing genomic similarities with BL21(DE3) and K-12 were only present in the duodenum of subjects with SIBO and correlated with abdominal pain (r=0.284, p-value=0.049; r=0.337, p=0.02, respectively) and diarrhea (r=0.321, p=0.03; r=0.298, p=0.04, respectively); a specific strain of K. aerogenes sharing genomic similarities with KCTC 2190, correlated with abdominal pain (r=0.374, p=0.013), diarrhea (r=0.305, p=0.038), and excess of gas (r=0.286, p=0.048) (Table 1).
Conclusion: This first luminal shotgun sequencing of the small bowel in SIBO now clearly shows that a very small and specific group of Escherichia and Klebsiella species and strains profoundly dominate the microbiome in SIBO. Further, specific strains of E. coli and Klebsiella aerogenes correlate directly with the severity of common SIBO symptoms.

Figure 1. Duodenal Escherichia and Klebsiella species average abundance distribution in subjects without SIBO (A) and with SIBO (B) detected by shotgun sequencing of duodenal luminal samples.
Table 1. Microbial correlations with severity of GI symptoms.