J Neurogastroenterol Motil 2024; 30(3): 332-342  https://doi.org/10.5056/jnm22191
Impact of Esophageal Motility on Microbiome Alterations in Symptomatic Gastroesophageal Reflux Disease Patients With Negative Endoscopy: Exploring the Role of Ineffective Esophageal Motility and Contraction Reserve
Ming-Wun Wong,1,2 I-Hsuan Lo,3 Wei-Kai Wu,3,4,5 Po-Yu Liu,3 Yu-Tang Yang,3 Chun-Yao Chen,6 Ming-Shiang Wu,3,5 Sunny H Wong,7 Wei-Yi Lei,1 Chih-Hsun Yi,1 Tso-Tsai Liu,1 Jui-Sheng Hung,1 Shu-Wei Liang,1 C Prakash Gyawali,8 and Chien-Lin Chen1,9*
1Department of Medicine, Hualien Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation and Tzu Chi University, Hualien, Taiwan; 2School of Post-Baccalaureate Chinese Medicine, Tzu Chi University, Hualien, Taiwan; 3Department of Internal Medicine, National Taiwan University College of Medicine, Taipei, Taiwan; Departments of 4Medical Research and 5Internal Medicine, National Taiwan University Hospital, Taipei, Taiwan; 6Department of Life Sciences, Tzu Chi University, Hualien, Taiwan; 7Lee Kong Chian School of Medicine, Nanyang Technological University, Singapore; 8Division of Gastroenterology, Washington University School of Medicine, St Louis, MI, USA; and 9Institute of Medical Sciences, Tzu Chi University, Hualien, Taiwan
Correspondence to: *Chien-Lin Chen, MD, PhD
Department of Medicine, Hualien Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation and Tzu Chi University, 707, Section 3, Chung-Yang Road, Hualien 970, Taiwan
Tel: +886-3-8561825, E-mail: harry.clchen@msa.hinet.net
Ming-Wun Wong and I-Hsuan Lo equally contributed to this study.
Received: November 10, 2022; Revised: March 20, 2023; Accepted: May 1, 2023; Published online: July 30, 2024
© The Korean Society of Neurogastroenterology and Motility. All rights reserved.

cc This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Abstract
Background/Aims
Ineffective esophageal motility (IEM) is common in patients with gastroesophageal reflux disease (GERD) and can be associated with poor esophageal contraction reserve on multiple rapid swallows. Alterations in the esophageal microbiome have been reported in GERD, but the relationship to presence or absence of contraction reserve in IEM patients has not been evaluated. We aim to investigate whether contraction reserve influences esophageal microbiome alterations in patients with GERD and IEM.
Methods
We prospectively enrolled GERD patients with normal endoscopy and evaluated esophageal motility and contraction reserve with multiple rapid swallows during high-resolution manometry. The esophageal mucosa was biopsied for DNA extraction and 16S ribosomal RNA gene V3-V4 (Illumina)/full-length (Pacbio) amplicon sequencing analysis.
Results
Among the 56 recruited patients, 20 had normal motility (NM), 19 had IEM with contraction reserve (IEM-R), and 17 had IEM without contraction reserve (IEM-NR). Esophageal microbiome analysis showed a significant decrease in microbial richness in patients with IEM-NR when compared to NM. The beta diversity revealed different microbiome profiles between patients with NM or IEM-R and IEM-NR (P = 0.037). Several esophageal bacterial taxa were characteristic in patients with IEM-NR, including reduced Prevotella spp. and Veillonella dispar, and enriched Fusobacterium nucleatum. In a microbiome-based random forest model for predicting IEM-NR, an area under the receiver operating characteristic curve of 0.81 was yielded.
Conclusions
In symptomatic GERD patients with normal endoscopic findings, the esophageal microbiome differs based on contraction reserve among IEM. Absent contraction reserve appears to alter the physiology and microbiota of the esophagus.
Keywords: Contraction reserve; Esophageal motility disorders; High-resolution manometry; Ineffective esophageal motility; Microbiota
Introduction

Esophageal motor function contributes to the pathophysiology of gastroesophageal reflux disease (GERD).1 Ineffective esophageal motility (IEM), identified using high-resolution manometry (HRM), may contribute to impaired clearance of refluxate from the esophagus.1-3 Although the prevalence of IEM increases with the severity of GERD,4 causal relationships are complicated since IEM can occur in asymptomatic healthy individuals.5 For further characterization of IEM, a provocative test termed multiple rapid swallows (MRS) is used to assess esophageal contraction reserve during HRM.6 The vigor of the contraction reserve induced by MRS inversely correlates with esophageal acid exposure, especially in patients with IEM.7,8

The esophagus is hosted by microbial species similar to that seen in the oral cavity because of its anatomic continuity to the mouth, and its role in transferring food and saliva from the mouth to the stomach.9 Changes in the esophageal microbiome have been reported with gastric reflux.10 In the healthy esophagus, Streptococcus is consistently reported to be a dominant genus, together with Prevotella, Veillonella, and Neisseria.9 Y.T.Y. et al analyzed 34 esophageal biopsy samples and classified the esophageal microbiome into 2 types using the 16S ribosomal RNA (rRNA) gene sequencing. Type 1 was dominated by the genus Streptococcus and was deemed phenotypically normal. By contrast, type 2 included a greater proportion of gram-negative bacteria and was correlated with severer GERD phenotypes, such as erosive esophagitis and Barrett’s esophagus.11 In addition, patients with esophageal adenocarcinoma were also reported to have esophageal microbiota different from healthy controls.12,13

Esophageal motor function and esophageal microbiome alteration have each been linked to GERD,1 but their relationship in patients with reflux symptoms has not been studied in detail.9,13 In this study, we tested the hypothesis that abnormal esophageal motor function contributes to the altered esophageal microbiome in patients with reflux symptoms. Specifically, by comparing the esophageal microbiome among GERD patients with normal motility (NM), IEM with contraction reserve (IEM-R), and IEM without contraction reserve (IEM-NR), we aim to explore the role of esophageal microbiota in esophagus hypomotility, especially in the presence or absence of contraction reserve.

Materials and Methods

Participants

Adult patients (age 18-70 years) having a history of exclusively typical GERD symptoms (heartburn and/or acid regurgitation) for at least 6 months were prospectively enrolled from the Gastrointestinal Outpatient Department of Hualien Tzu Chi Hospital. Exclusion criteria consisted of recent use of acid-suppressive therapy, antibiotics, histamine-2 receptor antagonists (H2RAs), or nonsteroidal anti-inflammatory drug (NSAID) in the past 6 months; any major organ disease (including scleroderma), systemic disease (including diabetes and hypertension), irritable bowel syndrome, functional dyspepsia, or cancer. The study was approved by the Research Ethics Committee of Hualien Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation, Hualien, Taiwan (IRB109-210-A). Written informed consent was obtained from each participant. The study was performed in accordance with the principles of the Declaration of Helsinki.

Protocol

All participants underwent esophagogastroduodenoscopy (EGD) prior to initiation of acid-suppressive therapy, and those who had erosive esophagitis, Barrett’s esophagus, esophageal neoplasm, or hiatal hernia were excluded for preventing their influences on esophageal microbiota.12-15 Microbiome samples were obtained by biopsies at the distal esophagus 5 cm above the esophagogastric junction during EGD. We performed esophageal biopsies solely at the mucosal layer, which theoretically should not affect esophageal motility. However, to minimize the possibility of any impact and to ensure consistency, we obtained biopsy samples from each patient at the same anatomical height, thereby maintaining a uniform effect across all participants. We performed HRM 1 day after EGD, participants were categorized into study groups based on esophageal motility on HRM: NM, IEM-R, and IEM-NR.

High-resolution Manometry

HRM was performed by using a catheter with 22 water-perfused sensors (MMS, HRM, Enschede, the Netherlands) in patients under supine position after an overnight fast. The catheter was zeroed to atmospheric pressure, inserted transnasally into the esophagogastric lumen and then positioned with at least 3 distal sensors in the stomach.

Primary peristalsis was assessed using 10 swallows of 5 mL water at 30-second intervals. Esophageal smooth muscle contractility was measured using the distal contractile integral (DCI) which consists of the amplitude, duration, and length of distal esophageal contraction extending from the proximal pressure trough to the lower esophageal sphincter (LES). Esophageal contractility was characterized using DCI value as hypercontractile (> 8000 mmHg∙s∙cm), normal (450-8000 mmHg∙s∙cm), and ineffective (< 450 mmHg∙s∙cm).16 Provocative testing with 5 MRS sequences was undertaken by steadily injecting water into the mouth through a syringe. Each successful MRS sequence requires at least 4 rapid 2 mL water swallows with an interval ≤ 4 seconds between swallows. Intact contraction reserve was defined as a ratio greater than 1 between the post-MRS DCI and the average DCI for 10 standard single swallows (SS) (ie, MRS/SS DCI > 1).8

Multichannel Intraluminal Impedance pH

Following the HRIM studies, part of study patients underwent multichannel intraluminal impedance pH (MII-pH) testing as per standard clinical protocol. A multichannel intraluminal impedance pH catheter (MMS, HRM) was placed transnasally, with positioning based on the manometrically identified LES. The catheter consisted of 6 impedance sensors positioned 3, 5, 7, 9, 15, and 17 cm from the LES and pH sensors positioned 5 cm above and 10 cm below the LES. MII-pH testing was used to assess mean nocturnal baseline impedance (MNBI) and acid exposure time (AET) in this study. MNBI was calculated by averaging baseline impedance values at 3 cm above the upper border of the LES during stable nocturnal 10-minute periods at 1 AM, 2 AM, and 3 AM.17 Total AET, defined as the fraction of time the distal esophagus is exposed to a pH < 4.0, was extracted from pH-impedance studies and considered pathological when > 6%. A positive symptom association probability (SAP) was defined as > 95%, corresponding to P < 0.05. Based on AET and SAP, patients were stratified into non-erosive reflux disease (NERD) (AET > 6%), reflux hypersensitivity (AET < 4%, positive SAP), and functional heartburn (AET < 4%, negative SAP).2

Microbial Profiling

The genomic DNA from the biopsied specimen was extracted using the column-based method (QIAamp PowerFecal DNA Kit; Qiagen, Hilden, Germany). DNA concentration was determined by Qubit 4.0 Fluorometer (Thermo Scientific, Waltham, MA, USA) and adjusted to 1 ng/uL for the following process. The genomic DNA of all 56 samples were used for V3-V4 amplicon sequencing, and the remaining genomic DNA from 52 samples were used for full-length amplicon sequencing.

16S ribosomal RNA gene V3-V4 amplicon sequencing

Polymerase chain reaction (PCR) targeting the 16S rRNA V3-V4 region was performed with the primer pair 341F/805R, using KAPA HiFi HotStart ReadyMix (Roche, Munich, Germany) with Illumina adaptor sequences on their 5’ end. DNA was recovered from excised gels after electrophoresis of the PCR products using an EasyPure PCR/Gel extraction kit (Bioman Scientific Co, New Taipei City, Taiwan).

The DNA library was constructed from extracted DNA using a Nextera XT index kit (Illumina, San Diego, CA, USA). The resulting DNA was then purified using AMPure XP beads (Beckman Coulter, Brea, CA, USA). A Quant-Ti dsDNA Assay Kit (Thermo Fisher Scientific Inc, Waltham, MA, USA) and an Agilent Technologies 2100 Bioanalyzer (Santa Clara, CA, USA) were used to quantify DNA concentration (adjusted to 10 nM for sequencing). The samples were sequenced in MiSeq (Illumina), and the obtained reads were processed with Quantitative Insights Into Microbial Ecology 2 (QIIME2).18 The DADA2’ plugin was used to identify amplicon sequence variants (ASVs) from demultiplexed sequence files. The Human Oral Microbiome Database (eHOMD) was used for taxonomy assignment.19 Read count of each sample was rarefied to the minimal counts among all samples (15 928 reads). The ASVs were collapsed into the phylogenetic levels for downstream analyses.

16S ribosomal RNA gene full-length amplicon sequencing

The full-length 16S rRNA genes (V1-V9 regions) were amplified according to the Amplification of Full-Length 16S Gene with Barcoded Primers for Multiplexed SMRTbell Library Preparation and Sequencing Procedure (Pacbio, Menlo Park, CA, USA). Each primer is designed to contain a 5’ buffer sequence (GCATC) with a 5’ phosphate modification, a 16-base barcode, and the degenerate 16S gene-specific forward or reverse primer sequences (Forward: 5’-Phos/GCATC-16-base barcode–AGRGTTYGATYMTGGCTCAG-3’, Reverse: 5’-Phos/GCATC-16-base barcode–RGYTACCTTGTTACGACTT-3’). In brief, 2 ng of gDNA was used for the PCR reaction carried out with KAPA HiFi HotStart ReadyMix (Roche). The PCR products were monitored on 1% agarose gel. Samples with a bright main strip around 1500 base pairs were chosen and purified by using the AMPure PB Beads for the following library preparation.

The library preparation was performed according to the Amplification of Full-Length 16S Gene with Barcoded Primers for Multiplexed SMRTbell Library Preparation and Sequencing Procedure (Pacbio). The SMRTbell library was incubated with sequencing primer v4 and sequel II Binding Kit 2.1 for the primer annealing and polymerase binding. At last, sequencing was performed in the circular consensus sequence mode on a PacBio Sequel IIe instrument to generate the HiFi reads with Predicted Accuracy (Phred scale) = 30. After demultiplexing, the circular consensus sequence were further processed with DADA2 to identify ASVs. The QIIME2 algorithm was employed for taxonomy assignment by alignment against the eHOMD. Read count of each sample was rarefied to the minimal counts among all samples (1278 reads). The ASVs were collapsed into phylogenetic levels for downstream analyses.

Re-analysis of Published Microbiota Data

To verify the esophageal microbiota in our study is representative of esophageal microbiome profile, we compared our sequences with datasets by Li et al20 (BioProject accession number PRJNA628659) and Lopetuso et al21 (PRJNA553177). Li et al20 included patients with esophageal squamous cell carcinoma and esophagogastric junction cancer from China, whereas Lopetuso et al21 included patients with esophageal adenocarcinoma and Barrett esophagus from Italy; both included healthy controls. The downloaded 16S rDNA sequences were trimmed to fragments from bases 341 to 805 to be comparable with the sequences in the present study.

Statistical Methods

The richness of microbiota species (alpha diversity) was assessed by observed species, Shannon index, Simpson index, and Chao1 index. Compositional differences (beta diversity) were compared by Bray-Curtis dissimilarity and plotted with principal coordinate analysis and nonparametric multidimensional scaling. An analysis of similarities (ANOSIM) test was used to test heterogeneity between groups. All numerical analyses were tested using two-tailed Student’s t tests, Mann-Whitney U tests, or ANOVA tests as deemed appropriate under a significance level of P < 0.05. All statistics were analyzed using R software version 3.6.3 or GraphPad Prism (version 9). Finally, a microbiome-based random forest classifier was built with random Forest R package.

Data Availability

16S rRNA gene V3-V4 amplicon sequencing data have been deposited on NCBI public repository (BioProject #PRJNA746612) and 16S rRNA gene full-length amplicon sequencing data have been deposited on NCBI public repository (BioProject #PRJNA831571).

Results

Clinical Characteristics of the Enrolled Gastroesophageal Reflux Disease Patients

A total of 78 participants met the criteria of chronic typical GERD symptoms after excluding recent use of proton pump inhibitor (n = 27), H2RA (n = 32), antibiotics (n = 4), or NSAID (n = 7) were enrolled. After receiving EGD, 10 patients with erosive esophagitis, 5 patients with hiatal hernia, and 7 patients with erosive esophagitis as well as hiatal hernia were excluded. Finally, 56 patients with reflux symptoms and normal EGD were included (mean age 48.6 ± 12.0 years, 30 females). Characterized by HRM metrics and MRS response, 20 patients had NM, 19 patients were grouped as IEM-R, and 17 patients were grouped as IEM-NR (Fig. 1A). Patients with IEM-NR exhibited higher body mass index compared to patients with IEM-R (P = 0.032). Furthermore, patients with NR demonstrated greater DCI values than both patients with IEM-R and those with IEM-NR (P < 0.001). There were no significant differences in age, gender distribution, gastroesophageal reflux disease questionnaire scores, and 4-second integrated relaxation pressure among the 3 patient groups (P > 0.05) (Supplementary Table).

Figure 1. Study flow chart and comparative analysis of esophageal microbiota in normal and ineffective esophageal motility groups. (A) Study flow chart. (B) The distribution of phylum, family, and genus of the esophageal microbiota among normal motility (NM), ineffective esophageal motility with contraction reserve (IEM-R), and ineffective esophageal motility without contraction reserve (IEM-NR) groups. GERD, gastroesophageal reflux disease; PPI, proton pump inhibitor; H2RA, histamine-2 receptor antagonist; HRM, high-resolution manometry; EGD, esophagogastroduodenoscopy.

Esophageal Microbiome Pattern in Gastroesophageal Reflux Disease Patients

Among the 56 esophageal biopsy samples, 4668 amplicon sequence variants were identified and grouped into 212 species, 136 genera, 94 families, and 14 phyla. Streptococcus, Prevotella, Neisseria, Veillonella, Haemophilus, Alloprevotella, Porphyromonas, and Fusobacterium were the most abundant genera (Fig. 1B). Streptococcus constituted a predominant proportion of the esophageal microbiota in the NM, IEM-R, and IEM-NR groups, which were consistent with the type 1 microbiome of the normal esophageal mucosa. We then explored the similarity between the esophageal microbiome in this study with 2 previously published datasets,14,21 and found our samples were within the spectrum of microbiota-derived from healthy esophageal tissues of these 2 studies (Supplementary Fig. 1). Of note, a trend of increased Streptococcus-to-Prevotella ratio was observed in the IEM groups despite not reaching statistical significance.

Microbiome in Different Esophageal Motility Phenotypes

Microbiota species richness (alpha diversity) and compositional differences (beta diversity) were compared among the NM, IEM-R, and IEM-NR groups. The observed species and Chao1 index representing the richness of microbial community were significantly decreased in the IEM-NR group as compared with the NM group (P = 0.040, ANNOVA + Tukey test) (Fig. 2). On the other hand, the evenness-weighted Shannon and Simpson index showed a decreased trend in the IEM-NR group.

Figure 2. Biodiversity analysis for the esophageal microbiome of patients with normal motility (NM), ineffective esophageal motility with contraction reserve (IEM-R), and ineffective esophageal motility without contraction reserve (IEM-NR). Alpha diversity indexes indicating microbial richness, such as observed species and Chao1 index, were significantly decreased in the IEM-NR groups when compared with the NM group (P = 0.040, Tukey test).

For the comparison of beta diversity, the principal coordinate analysis showed extensive overlapping of microbial composition and no significant difference among the 3 groups (ANOSIM, P = 0.196) (Fig. 3A). The microbiome compositions were also not different between the NM and IEM (IEM-R and IEM-NR) groups (ANOSIM, P = 0.771) (Fig. 3B). Of note, a significant difference in microbiome profile was observed between patients with NM or IEM-R and those with IEM-NR (ANOSIM, P = 0.037) (Fig. 3C).

Figure 3. Esophageal microbiome profile comparison between the normal motility (NM), ineffective esophageal motility with contraction reserve (IEM-R), and ineffective esophageal motility without contraction reserve (IEM-NR) groups by using principal coordinate analysis (PCoA). (A) No significant difference in esophageal microbiome profiles was found among the 3 groups. (B) Unremarkable difference in esophageal microbiome profiles between patients with NM and patients with IEM. (C) A significant microbiome compositional difference was observed between patients with NM or IEM-R and patients with IEM-NR (analysis of similarities [ANOSIM] P = 0.037).

We then performed multivariate sparse partial least-squares discriminant analysis (sPLS-DA) to identify bacterial species contributing to the distinction between the study groups (Fig. 4A). The abundance of microbial taxa decreased in patients with IEM-NR, including Actinomyces spp., Peptidiphaga spp., Rothia mucilaginosa, Prevotella spp., and Veillonella dispar, which contributed to the separation along with component 1 (Fig. 4B). A heatmap plotted by significantly different species observed between NM or IEM-R and IEM-NR groups (Mann-Whitney U test, P < 0.05) showed multiple esophageal bacterial species decreased markedly in the IEM-NR group (Fig. 5). Notably, several differential taxa, including Actinomyces spp._HMT_169, Peptidiphaga spp._HMT_183, R. mucilaginosa, Prevotella denticola, Kingella spp._HMT_932, and Prevotella pallens were markedly reduced in patients without a contraction reserve and were consistent with the result of sPLS-DA.

Figure 4. Multivariate sparse partial least-squares discriminant analysis (sPLS-DA) and bacterial species contributions in esophageal motility groups. (A) sPLS-DA for identifying variances responsible for the distinction between the study groups. (B) Specific bacterial species contribute to the separation along with component 1. NM, normal motility; IEM-R, ineffective esophageal motility with contraction reserve; IEM-NR, ineffective esophageal motility without contraction reserve.

Figure 5. A heatmap diagram showing the abundance of esophageal bacterial species differs significantly between patients with normal motility (NM) or ineffective esophageal motility with contraction reserve (IEM-R) and patients with ineffective esophageal motility without contraction reserve (IEM-NR). Gray scale represented log-transformed relative abundance normalized into Z-scores by row. The P-values indicate the significance of differential abundance of microbial taxa between patients with NM or IEM-R and patients with IEM-NR by using Mann-Whitney U test.

Moreover, 27 patients of this study received MII-pH testing and parameters such as AET and MNBI were obtained. The MNBI values for patients with NM and those with IEM-R were both higher than the values observed in patients with IEM-NR (P = 0.015). The proportions of GERD phenotypes and AET values showed no significant differences among the 3 patient groups (P > 0.05) (Supplementary Table). Regarding the phenotypes, 5 patients were NERD, 12 patients were reflux hypersensitivity, and 10 patients were functional heartburn. There were no significant difference of microbiota species richness and compositional differences among phenotypes. Among the identified 212 esophageal bacterial species, 3 taxa including Parvimonas micra, Gemella morbillorum, and Schaalia spp._HMT_172 were identified to be correlated positively with AET and negatively with MNBI (Supplementary Fig. 2).

Microbiome-based Predicting Model for the Absence of Contractile Reserve

Finally, a microbiome-based random forest model was built to determine if the distinctive esophageal bacterial taxa could be used to predict patients with IEM-NR. The model yielded an area under the receiver operating characteristic curve (AUROC) of 0.81 (95% CI, 0.68-0.93) in differentiating GERD patients with and without a contraction reserve (Fig. 6A). Moreover, we further performed 16S rRNA full-length (FL) amplicon sequencing (Pacbio HiFi reads) by using residual extracted DNA from the biopsied esophagus specimen (n = 52; NM:19, IEM-R:17, IEM-NR:16) to verify the results observed in the 16S rRNA V3-V4 amplicon sequencing (Illumina Miseq). In brief, the results of 16S FL sequencing were compatible with the findings of 16S V3-V4 sequencing. Some distinctive microbial features decreased in IEM-NR were noted, including Prevotella pallens, Prevotella melaninogenica, V. dispar, and Neisseria spp. Notably, Fusobacterium spp._HMT_203 was significantly enriched in the IEM-NR group (false discovery rate [FDR] < 0.01), which was not detected by the 16S V3-V4 sequencing. We then BLAST the ASVs of Fusobacterium spp._HMT_203 against the NCBI database and found the sequences belonged to F. nucleatum (99% identity with the ATCC 25 586). With the distinctive esophageal bacterial taxa identified by the 16S FL sequencing, a FL-based random forest classifier was established for predicting an absent contraction reserve and an AUROC of 0.82 (95% CI, 0.7-0.94) was obtained (Fig. 6B).

Figure 6. Microbiome-based random forest model and differential bacterial species in esophageal motility groups. (A) A microbiome-based random forest model was built with 20 selected 16S V3-V4 microbial features significantly different between normal motility (NM) or ineffective esophageal motility with contraction reserve (IEM-R) and ineffective esophageal motility without contraction reserve (IEM-NR), and an area under the receiver operating characteristic curve (AUROC) of 0.81 (95% CI, 0.68-0.93) for predicting IEM-NR was obtained. (B) Differential bacterial species between NM or IEM-R and IEM-NR identified by using 16S full-length (FL) sequencing showed consistent results with that found by using 16S V3-V4 sequencing. Gray scale represented log-transformed relative abundance normalized into Z-scores by row. The P-values indicate the significance of differential abundance of microbial taxa between patients with NM or IEM-R and patients with IEM-NR by using Mann-Whitney U test. A microbiome-based random forest model was built with seven selected 16S FL microbial features yielded an AUROC of 0.82 (95% CI, 0.7-0.94) for predicting IEM-NR. AUC, area under the curve; PCoA, principal coordinate analysis.
Discussion

This study demonstrated that impairment of esophageal contraction reserve contributes to esophageal microbiome alterations in patients with reflux symptoms and IEM in the setting of normal endoscopy. Patients who had IEM with absent contraction reserve also exhibited a marked decrease of observed bacterial species (alpha diversity), which included several taxa of Prevotella spp. and V. dispar. Based on these findings, we developed a potential microbiome-based random forest model for predicting the absence of esophageal contraction reserve. In addition, we used the state-of-the-art long-read sequencing technique for 16S FL esophageal microbiome profiling and further verified the results found in 16S V3-V4 sequencing. Of note, F. nucleatum was observed to be significantly increased in patients without a contraction reserve (FDR < 0.01).

Streptococcus, Prevotella, Veillonella, Neisseria, and Hemophilus have been consistently reported to colonize the healthy esophageal mucosa in both culture-dependent and culture-independent studies,9 which was also evident in the current study. Reflux-related disease, such as esophagitis, Barrett’s esophagus, and esophageal adenocarcinoma, was also known to be an important cause of altered esophageal microbiota.9 To date, however, the microbiota had not been explored in nonerosive reflux disease, reflux hypersensitivity, or functional heartburn, which may range in severity from a healthy esophagus to esophageal mucosal injury at endoscopy. The results of the present study demonstrate that the esophageal microbiota in patients with chronic reflux symptoms and normal endoscopy findings are similar to that of healthy controls (Streptococcus-predominant type 1 microbiota).11 In turn, this suggests that a shift in esophageal microbiota colonization may reflect GERD-related mucosal injury seen at endoscopy.

In the present study, a significant decrease in the richness of esophageal microbiota was observed in the IEM-NR group, rather than the IEM-R group, when compared with patients having normal esophageal motility. Since the contraction reserve is the compulsive compensatory function of esophagus with ineffective motility, it plays a critical role in protecting the esophageal mucosa from prolonged exposure to acid reflux in the IEM patients. The impaired contraction reserve may inhibit the growth of acid-sensitive esophageal microbes by increasing esophageal acid exposure time. Besides, the impaired compression forces along the esophagus in IEM patients without contraction reserve may reduce the adhesion of saliva microbes to the esophageal mucosa. Thus, the presence or absence of esophageal contraction reserve could be a key environmental factor that affects the biodiversity of esophageal microbial community.

Among the 27 patients who received MII-pH testing, P. micra and G. morbillorum were strongly associated with increased esophageal acid exposure and decreased mucosal integrity. Interestingly, fecal P. micra and G. morbillorum had been reliably associated with human colorectal cancer22 and saliva P. micra was also reported significantly increased with pathologic stage of oral squamous cell carcinoma.23 Whether these microbes are linked to the pathogenesis of esophageal disorders requires further investigation.

For the microbial composition, our analysis demonstrated different microbiome profiles between those with and without contraction reserve rather than those with and without IEM. These results further provide evidence that esophageal contraction reserve during MRS may uncover important pathophysiological differences relative to GERD in patients with IEM.5 Among the various bacterial species decreased in patients without esophageal contraction reserve, the decrease in P. pallens, P. melaninogenica, and V. dispar were constantly observed with different analytical approaches, including the 16S FL sequencing, and might become potential biomarkers in those who are absent in esophageal contraction reserve. Interestingly, F. nucleatum was identified to be significantly enriched in the IEM-NR group (FDR < 0.01) by analyzing the 16S FL sequences. As one of the periodontal bacteria, F. nucleatum was considered as an opportunistic pathogen and has been causally correlated with an increased risk of colorectal cancer and esophageal cancer through its oncogenic and pro-inflammatory properties.24,25 Despite an increased abundance of F. nucleatum in patients with IEM-NR, however, whether there is a causal role of F. nucleatum for an absent esophageal contraction reserve still requires to be investigated by further study.

Furthermore, we established a microbiome-based random forest model (AUROC = 0.81) that could predict the absence of esophageal contraction reserve based on the changes in the microbiota. These findings reinforce the view that a distinct pattern of esophageal microbiota is present in patients with reflux symptoms who have normal endoscopic findings and that these changes are significantly associated with a change in the esophageal contraction reserve.

The novelty and significance of this study is that it is the first to explore the impact of esophageal motor function on the esophageal microbiota. Specifically, esophageal contraction reserve was shown to exert more contribution to esophageal microbiome alteration than IEM per se. This supports the notion that IEM is a manometric rather than a clinical diagnosis and that evaluating esophageal contraction reserve using MRS further identifies the severity of esophageal hypomotility and IEM, which are part of GERD related pathophysiology.5,8 Besides, this is the first study that applies the technique of 16S FL sequencing in esophageal microbiota research. The FL sequence-based random forest modeling by 7 selected features shows a comparable prediction accuracy (AUROC = 0.82) with the V3-V4 sequence-based modeling by 20 selected features. This is probably because the 16S FL sequences have a higher resolution in assigning bacterial taxa. Besides, the differential taxa at species level obtained from LEfSe were highly consistent with differential species mentioned above in both 16S V3-V4 and 16S FL amplicon sequencing datasets (Supplementary Fig. 3). These consistent findings observed in both 16S V3-V4 and 16S FL microbiome analysis further strengthen the significance and robustness of this study.

While our study primarily focused on the relationship between esophageal motility and the esophageal microbiome in patients with GERD symptoms, it is important to consider the possibility of a bidirectional relationship between these 2 factors. The findings may have implications for the treatment of GERD. A deeper understanding of the esophageal microbiome and its association with esophageal motility dysfunction could potentially help in developing targeted therapeutic strategies. However, it is crucial to first establish the causality between these factors. For instance, if alterations in the esophageal microbiome are found to play a significant role in GERD pathogenesis and a causal relationship is established, probiotics or other microbiome-modifying therapies could be considered as a part of a comprehensive treatment plan. Additionally, identifying specific motility patterns that contribute to microbiome alterations may guide the development of tailored therapies to address underlying motility issues in GERD patients. The concept of treating GERD symptoms by adjusting the microbiota to further improve esophageal motility function could be an innovative approach to managing this condition. It is essential to note that further research is necessary to confirm these associations and establish the efficacy of such therapeutic approaches. Future studies could explore the direct impact of esophageal motility-targeted interventions on the esophageal microbiome and GERD symptom improvement, while also focusing on elucidating the causal relationships involved and investigating the potential benefits of microbiota modulation in improving esophageal motility. In particular, understanding the potential bidirectional relationship between esophageal motility and the esophageal microbiome will be crucial to advancing our knowledge of these interactions and their implications for GERD treatment.

This study has several limitations that we acknowledge. First, we did not perform MII-pH testing to further phenotype all patients with reflux symptoms, which may have a risk of type II error to conclude no significant difference of microbiota species richness and compositional differences among phenotypes. However, the primary aim of this exploratory research was to determine relationships, if any, between esophageal motor function and the microbiota in symptomatic patients with normal endoscopy. Now that we have established the presence of such relationships, future studies are warranted to evaluate esophageal microbiota patterns in GERD phenotypes determined by reflux monitoring. Second, we did not evaluate gene expression in the esophageal epithelium for host-microbe interactions, and there was no further causal interpretation between patients with and without esophageal contraction reserve. Third, the altered esophageal microbiome could be induced by liquid retention rather than esophageal hypomotility; however, since there is no available data from patients with achalasia or scleroderma for comparison, it remains uncertain whether the microbiome changes in patients with IEM-NR is due to lack of motility or liquid retention. Finally, this work did not enroll patients with absent contractility, which may represent more severe hypomotility. If microbiota patterns between patients with absent contractility and IEM patients with absent contraction reserve are similar, it may provide additional support to current findings.26,27

In conclusion, esophageal contraction reserve rather than IEM appears to be significantly associated with altered richness and composition of esophageal microbiome in patients with chronic reflux symptoms and normal endoscopy findings. Specific distinctive bacteria taxa include reduced Prevotella spp. and V. dispar, and enriched F. nucleatum in patients with absent esophageal contraction reserve. The microbiome-based random forest models established by both 16S V3-V4 and 16S FL sequences can predict absent contraction reserve. Further studies are needed to determine the interactions among esophageal motility, mucosa integrity, reflux burden, and microbiota colonization.

Supplementary Materials

Note: To access the supplementary table and figures mentioned in this article, visit the online version of Journal of Neurogastroenterology and Motility at http://www.jnmjournal.org/, and at https://doi.org/10.5056/jnm22191.

Financial support

This study was supported by a grant (TCRD110-23) from Hualien Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation, Hualien, Taiwan, a grant (TCMF-A 107-02-03) from Buddhist Tzu Chi Medical Foundation, Hualien, Taiwan, and a grant (MOST110-2628-B-002-046) from Ministry of Science and Technology, Taiwan. We would like to acknowledge the service provided by the Medical Microbiota Center of the First Core Laboratory, National Taiwan University College of Medicine. We thank BIOTOOLS Co., Ltd. for providing service of 16S FL library preparation and sequencing with PacBio HiFi reads.

Conflicts of interest

None.

Author contributions

Study concept and protocol design: Ming-Wun Wong and Chien-Lin Chen; patient recruitment and questionnaire interview and interpretation: Ming-Wun Wong, Wei-Yi Lei, Chih-Hsun Yi, Tso-Tsai Liu, Jui-Sheng Hung, Shu-Wei Liang, and Chien-Lin Chen; sample preparation and genomic DNA extraction: Yu-Tang Yang and Wei-Kai Wu; bioinformatics and statistical analysis: I-Hsuan Lo, Wei-Kai Wu, Chun-Yao Chen, and Po-Yu Liu; manuscript drafting: Ming-Wun Wong, Wei-Kai Wu, Sunny H Wong, C Prakash Gyawali, and Chien-Lin Chen; and manuscript revision: I-Hsuan Lo, Wei-Kai Wu, Ming-Wun Wong, Ming-Shiang Wu, Sunny H Wong, C Prakash Gyawali, and Chien-Lin Chen. All authors have approved the final version of the manuscript.

References
  1. Tack J, Pandolfino JE. Pathophysiology of gastroesophageal reflux disease. Gastroenterology 2018;154:277-288.
    Pubmed CrossRef
  2. Gyawali CP, Kahrilas PJ, Savarino E, et al. Modern diagnosis of GERD: the Lyon Consensus. Gut 2018;67:1351-1362.
    Pubmed KoreaMed CrossRef
  3. Savarino E, Giacchino M, Savarino V. Dysmotility and reflux disease. Curr Opin Otolaryngol Head Neck Surg 2013;21:548-556.
    Pubmed CrossRef
  4. Triadafilopoulos G, Tandon A, Shetler KP, Clarke J. Clinical and pH study characteristics in reflux patients with and without ineffective oesophageal motility (IEM). BMJ Open Gastroenterol 2016;3:e000126.
    Pubmed KoreaMed CrossRef
  5. Savarino E, Gemignani L, Pohl D, et al. Oesophageal motility and bolus transit abnormalities increase in parallel with the severity of gastro-oesophageal reflux disease. Aliment Pharmacol Ther 2011;34:476-486.
    Pubmed CrossRef
  6. Gyawali CP, Sifrim D, Carlson DA, et al. Ineffective esophageal motility: concepts, future directions, and conclusions from the Stanford 2018 symposium. Neurogastroenterol Motil 2019;31:e13584.
    Pubmed KoreaMed CrossRef
  7. Martinucci I, Savarino EV, Pandolfino JE, et al. Vigor of peristalsis during multiple rapid swallows is inversely correlated with acid exposure time in patients with NERD. Neurogastroenterol Motil 2016;28:243-250.
    Pubmed CrossRef
  8. Shaker A, Stoikes N, Drapekin J, Kushnir V, Brunt LM, Gyawali CP. Multiple rapid swallow responses during esophageal high-resolution manometry reflect esophageal body peristaltic reserve. Am J Gastroenterol 2013;108:1706-1712.
    Pubmed KoreaMed CrossRef
  9. Quader F, Rogers B, Sievers T, et al. Contraction reserve with ineffective esophageal motility on esophageal high-resolution manometry is associated with lower acid exposure times compared with absent contraction reserve. Am J Gastroenterol 2020;115:1981-1988.
    Pubmed CrossRef
  10. Di Pilato V, Freschi G, Ringressi MN, Pallecchi L, Rossolini GM, Bechi P. The esophageal microbiota in health and disease. Ann N Y Acad Sci 2016;1381:21-33.
    Pubmed CrossRef
  11. Dong L, Yin J, Zhao J, et al. Microbial similarity and preference for specific sites in healthy oral cavity and esophagus. Front Microbiol 2018;9:1603.
    Pubmed KoreaMed CrossRef
  12. Lv J, Guo L, Liu JJ, Zhao HP, Zhang J, Wang JH. Alteration of the esophageal microbiota in barrett's esophagus and esophageal adenocarcinoma. World J Gastroenterol 2019;25:2149-2161.
    Pubmed KoreaMed CrossRef
  13. Yang L, Lu X, Nossa CW, Francois F, Peek RM, Pei Z. Inflammation and intestinal metaplasia of the distal esophagus are associated with alterations in the microbiome. Gastroenterology 2009;137:588-597.
    Pubmed KoreaMed CrossRef
  14. Elliott DRF, Walker AW, O'Donovan M, Parkhill J, Fitzgerald RC. A non-endoscopic device to sample the oesophageal microbiota: a case-control study. Lancet Gastroenterol Hepatol 2017;2:32-42.
    Pubmed CrossRef
  15. Gall A, Fero J, McCoy C, et al. Bacterial composition of the human upper gastrointestinal tract microbiome is dynamic and associated with genomic instability in a barrett's esophagus cohort. PloS One 2015;10:e0129055.
    Pubmed KoreaMed CrossRef
  16. Kahrilas PJ, Bredenoord AJ, Fox M, et al. The Chicago classification of esophageal motility disorders, v3.0. Neurogastroenterol Motil 2015;27:160-174.
    Pubmed KoreaMed CrossRef
  17. Martinucci I, de Bortoli N, Savarino E, et al. Esophageal baseline impedance levels in patients with pathophysiological characteristics of functional heartburn. Neurogastroenterol Motil 2014;26:546-555.
    Pubmed CrossRef
  18. Bolyen E, Rideout JR, Dillon MR, et al. Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. Nat Biotechnol 2019;37:852-857.
    Pubmed KoreaMed CrossRef
  19. Escapa IF, Chen T, Huang Y, Gajare P, Dewhirst FE, Lemon KP. New insights into human nostril microbiome from the expanded human oral microbiome database (ehomd): a resource for the microbiome of the human aerodigestive tract. mSystems 2018;3:e00187-18.
    Pubmed KoreaMed CrossRef
  20. Li D, He R, Hou G, et al. Characterization of the esophageal microbiota and prediction of the metabolic pathways involved in esophageal cancer. Front Cell Infect Microbiol 2020;10:268.
    Pubmed KoreaMed CrossRef
  21. Lopetuso LR, Severgnini M, Pecere S, et al. Esophageal microbiome signature in patients with barrett's esophagus and esophageal adenocarcinoma. PloS One 2020;15:e0231789.
    Pubmed KoreaMed CrossRef
  22. Drewes JL, White JR, Dejea CM, et al. High-resolution bacterial 16S rRNA gene profile meta-analysis and biofilm status reveal common colorectal cancer consortia. NPJ Biofilms Microbiomes 2017;3:34.
    Pubmed KoreaMed CrossRef
  23. Yang L, Zheng C, Chen Y, Ying H. FLO genes family and transcription factor MIG1 regulate Saccharomyces cerevisiae biofilm formation during immobilized fermentation. Front Microbiol 2018;9:1860.
    Pubmed KoreaMed CrossRef
  24. Brennan CA, Garrett WS. Fusobacterium nucleatum - symbiont, opportunist and oncobacterium. Nat Rev Microbiol 2019;17:156-166.
    Pubmed KoreaMed CrossRef
  25. Idrissi Janati A, Karp I, Von Renteln D, et al. Investigation of Fusobacterium Nucleatum in saliva and colorectal mucosa: a pilot study. Sci Rep 2022;12:5622.
    Pubmed KoreaMed CrossRef
  26. Rengarajan A, Bolkhir A, Gor P, Wang D, Munigala S, Gyawali CP. Esophagogastric junction and esophageal body contraction metrics on high-resolution manometry predict esophageal acid burden. Neurogastroenterol Motil 2018;30:e13267.
    Pubmed CrossRef
  27. Wong MW, Hung JS, Liu TT, et al. Esophageal acid burden in reflux patients with normal endoscopy: does esophageal peristalsis matter?. J Formos Med Assoc 2022;121(1 Pt 2):388-394.
    Pubmed CrossRef


This Article

e-submission

Archives

Aims and Scope