Third generation sequencing analysis detects significant differences in duodenal microbiome composition between functional dyspepsia patients and control subjects.

Background: Functional dyspepsia (FD) is a multifactorial disorder as its development may be based on several different pathophysiological mechanisms. Interaction of gut microbiome with the host has been proposed as a potential mechanism involved in the disease's pathogenesis.

Aim/methods: We aimed to characterize microbiome profiling on duodenal luminal content (DLC) of FD patients and compare it to that of controls (CG) and patients with irritable bowel syndrome (IBS). Outpatients fulfilling Rome IV criteria for FD, IBS, and control group (CG) underwent upper gastrointestinal endoscopy and 2 cc of duodenal aspirate (3rd - 4th part) was aspirated in sterile traps. Duodenal microbiome was assessed after DNA extraction and 16S gene-based sequencing on Oxford Nanopore MinION followed by EPI2ME analysis (ONT/Metrich-ore Ltd). Bioanalysis of the microbiome (alpha-, beta-diversity, comparisons of relative abundances for all taxonomic ranks) was implemented in Python. Multiple group means comparisons were performed with one-way Analysis of Variance (ANOVA) and Kruskal-Wallis test with Tuckey's and Dunn's post hoc tests respectively, in case of significance (P-value <0.05).

Results: 20 subjects with FD (8 females; age 49.9 ± 13.5 yrs.), 20 with IBS (14 females; age 57.6 ± 14.8 yrs.) and 10 CG (6 females; age 49.2 ± 13.8 yrs.) had their DLC analyzed. The α-diversity index of subjects with FD was significantly lower compared to controls (Shannon's index, p = 0.0218) and similar to that of patients with IBS. Principal Coordinate Analysis (PCoA) generated from species relative abundances (beta-diversity) showed no difference in the DLC profile of subjects with FD and IBS when compared to controls (p = 0.513). Compared to controls, the relative abundance (RA) of Chloroflexota phylum was lower in subjects with FD (p = 0.017) and IBS (p = 0.026), respectively. Additionally, the RA of the Rhodothermota and Thermotogota phyla was lower in FD (p = 0.017 and p = 0.018, respectively) but not in IBS patients (p = 0.15 and p = 0.06, respectively) compared to controls. Interestingly, the RA of specific taxa from Chloroflexota, Rhodothermota and Thermotogota phyla were consistently lower in subjects with FD when compared to CG but similar to IBS, during analysis of all the subsequent major ranks of taxonomy. At the class level, there were significant differences in Syntrophobacteria, Acidithiobacillia, Cytophagia and Flavobacteriia between the FD and CG groups (p < 0.05), but no such difference between FD and IBS was found. Finally, multiple significant differences at the order, family, genus and species level between the FD and CG groups were also detected. A positive relationship between the RA of Streptococcus and those from genus Granulicatella was observed both in FD (p = 0.014) and IBS (p = 0.014) patients.

Conclusion & inferences: The microbiome profiling from duodenal luminal content of FD patients is significantly different to that of controls, including lower microflora diversity, different microflora structure/composition and specific taxa. Similar differences in the DLC between FD and IBS patients were not evident.