Secondary Logo

Journal Logo

Original Article: Gastroenterology: Inflammatory Bowel Disease

Gut Microenvironment and Bacterial Invasion in Paediatric Inflammatory Bowel Diseases

Zaidi, Deenaz∗,†; Huynh, Hien Q.; Carroll, Matthew W.; Mandal, Rupasri; Wishart, David S.; Wine, Eytan∗,†,§

Author Information
Journal of Pediatric Gastroenterology and Nutrition: November 2020 - Volume 71 - Issue 5 - p 624-632
doi: 10.1097/MPG.0000000000002848

Abstract

See “The Gut Microbiome and the Triple Environmental Hit Concept of Inflammatory Bowel Disease Pathogenesis” by Kellermayer and Zilbauer on page 589.

What Is Known/What Is New

What Is Known

  • Gut homeostasis and immune regulation is disrupted in inflammatory bowel diseases.
  • Microbial dysbiosis and inflammation are hallmarks of inflammatory bowel diseases.

What Is New

  • Alterations in gut luminal environment in inflammatory bowel diseases affect microbial virulence.
  • NMR successfully differentiates intestinal aspirate metabolites amongst noninflammatory bowel diseases and inflammatory bowel disease patients.
  • Luminal succinate is associated with E coli invasion and enhances bacterial invasion in vitro.

An infographic is available for this article at:https://links.lww.com/MPG/B885.

Inflammatory bowel diseases (IBD), encompassing Crohn disease (CD) and ulcerative colitis (UC) are chronic disorders with increasing incidence, especially in children (1). Gut metabolism in healthy individuals is in a state of homeostasis, with a well-balanced relationship between resident microbes and the host. Alterations in microbial composition and function (simplistically defined as dysbiosis) are associated with increased epithelial permeability and stimulation of the immune system and have been clearly established as hallmarks of IBD (2). Example for the perturbed host-microbe balance in IBD are the presence of lower butyrate-producing Roseburia hominis and a reduction of short-chain fatty acids (SCFA) observed in UC (3). Gut metabolites reflect the outcomes of cellular processes reflecting host-microbial interactions, and can be used as a ‘forensic’ tool to investigate the pathogenesis of intestinal diseases.

We hypothesized that because of the imbalanced homeostasis in IBD, the gut luminal microenvironment differs between individuals with or without IBD, and that this could alter the virulence capacity of bacteria.

MATERIALS AND METHODS

Patients

The patient cohort included in this study (previously described) (4,5) was constituted of paediatric patients who were scheduled to have gastroscopies and ileocolonoscopies at the Stollery Children's Hospital, University of Alberta. Ethics approval from the University of Alberta Research Ethics Board was obtained before recruitment. The control (non-IBD) group included children who had gastrointestinal symptoms, such as abdominal pain and diarrhoea but had completely normal endoscopic and histological mucosal findings. IBD patients included newly diagnosed and known CD and UC patients, following the Porto criteria (6). All patients were directed to maintain a clear fluid diet and ingest Picosalax (sodium picosulphate with magnesium citrate, for bowel cleansing) the day before the procedure and were fasted for at least 6 hours. The procedures were performed by paediatric gastroenterologists under general anaesthesia, administered using propofol and fentanyl by paediatric anaesthesiologists.

Collection of Intestinal Washes

During endoscopy, 50 mL of normal saline (NS), injected through the endoscopy working channel, were sprayed against the duodenal or TI mucosa and aspirates were collected via a suction trap (15–30 mL), before taking biopsies. A portion of the aspirates was cultured on agar plates (LB [Luria-Bertani], MRS [de Man, Rogosa, and Sharpe], and McConkey) at 37 °C, under aerobic and microaerophilic conditions for 24–48 hours and the remainder was immediately frozen in aliquots at −80 °C for in vitro experiments and NMR metabolomics. Bacterial colonies grown were isolated and frozen at −80 °C in 50% LB-glycerol. Three bacterial colonies from each plate were selected based on differences in morphology and biobanked for in vitro experiments.

Cell Culture

Colorectal epithelial cell lines, T84 and Caco-2, were used as a model for bacteria invasion. Cells were obtained from ATCC (American Type Culture Collection, Manassas, VA) and cultured in Dulbecco's Modified Eagle's Medium with 10% fetal bovine serum (Catalogue number F8192, Sigma-Aldrich, Oakville, ON) and 1% penicillin/streptomycin (Catalogue number P4333, Sigma-Aldrich) at 37 °C in 5% CO2. For in vitro infection, cells were grown in 24-well plates (seeding density of 1 × 106 cells/cm2) to confluence.

Identification of Bacteria Isolated From Patients

To characterize the bacterial isolates from intestinal aspirates of patients, frozen bacteria were cultured in either LB, MRS, or McConkey broths overnight at 37 °C and were subjected to bacterial DNA extraction, using the DNeasy Blood and Tissue kit (Qiagen, Mississauga, ON). The 16S rRNA gene of each bacterial isolate was amplified by PCR using universal primers (Forward primer: GAGTTTGATYMTGGCTCAG, reverse primer: ACTACYNGGGTATTAAKCC). Individual isolates were identified with Sanger DNA sequencing (Department of Biological Sciences, University of Alberta, Canada).

In Vitro Invasion Potential of Patient Microbes

Bacterial isolates were cultured overnight in LB broth at 37 °C under aerobic conditions. Gentamicin protection assays were conducted on either T84 or Caco-2 cells with these bacterial isolates. Strains were cultured in LB broth overnight at 37 °C in aerobic conditions and inoculated with T-84 or Caco-2 cells at multiplicity of infection (MOI) of 1 : 20. Wells were washed thrice with PBS, antibody free medium (800 μL/well) was added, and then wells were infected with bacteria at an MOI of 1 : 20 and incubated for 2 hours at 37 °C. After the incubation, 100 μL from each well was plated on agar plates, to quantify bacterial growth. Wells were washed 3 times with PBS and gentamicin (100 μg/mL; does not penetrate epithelial cells so only kills extracellular bacteria) was added in antibiotic-free medium (1 mL/well) and incubated for 2 hours at 37 °C to quantify bacterial invasion. Suspensions of 100 μL from each well were plated after the 2-hour gentamicin treatment on agar plates to confirm that bacteria were killed by gentamicin. Wells were washed thrice with PBS and 100 μL of 1% Triton-X were added to each well to lyse the cells and left at room temperature for 15 minutes; cell lysis allows for quantification of invasive bacteria.

After serial dilution, 100 μL from each well were plated on agar plates after further serial dilutions and placed in aerobic growth conditions for 24 hours. Bacterial colonies grown on the plates were counted and the number was multiplied by the dilution factor to reach the number of colony-forming unit (CFU)/well.

Effect of Intestinal Aspirates on Invasion of Known Eschrichia coli Strains

Gentamicin assays were conducted using 3 defined E coli strains: enterohemorrhagic E coli (EHEC), strain CL56 (O157:H7), LF82 (the reference strain of adherent-invasive E coli associated with Crohn disease) (7), or HB101 (nonpathogenic E coli strain) at an MOI of 1:20, as described above with the following modifications. This was done to assess whether aspirates from the gut can impact the ability of known bacteria to invade cells. Each infected well contained 200 μL of TI aspirate from either non-IBD, CD, or UC patients, or normal saline (as a no-aspirate control). Aspirates were syringe filtered (0.2 μm) to remove food, cell debris, and bacteria present in the aspirates. Absence of bacteria was confirmed by plating the syringe-filtered aspirates on agar plates and incubating at 37 °C for 48 hours. Wells containing media only (without bacteria) were used as negative controls.

Metabolomic Analysis of Intestinal Aspirates

Metabolomic analysis was conducted on intestinal aspirates to determine differences in gut metabolites between non-IBD and IBD cases using NMR spectroscopy, in order to correlate these differences with the effects of aspirates on bacteria. After thawing and centrifuging the intestinal aspirates, 285 μL aliquots were removed and placed in 1.5 mL Eppendorf tubes, followed by the addition of 65 μL of a standard buffer solution (54% D2O: 46%; 1.75 mmol/L KH2PO4; pH 7.0 v/v containing 5.84 mM DSS [2,2-dimethyl-2-silcepentane-5-sulphonate], and 0.1% NaN3 in H2O). The sample (350 μL) was then transferred to a 3 mm SampleJet NMR tube for subsequent spectral analysis. For data collection, 1H-NMR spectra were collected on a 700 MHz Avance III (Bruker) spectrometer equipped with a 5 mm HCN Z-gradient pulsed-field gradient (PFG) cryoprobe. 1H-NMR spectra were acquired at 25 °C, using the first transient of the NOESY presaturation pulse sequence (noesy1dpr), chosen for its high degree of quantitative accuracy (8,9). All free induction decays were zero-filled to 250 K data points. The singlet produced by the DSS methyl groups was used as an internal standard for chemical shift referencing (set to 0 ppm). For quantification, all 1H-NMR spectra were processed and analyzed using the Chenomx NMR Suite Professional Software package version 8.1 (Chenomx Inc, Edmonton, AB). The Chenomx NMR Suite software allows for qualitative and quantitative analysis of an NMR spectrum by manually fitting spectral signatures from an internal database to the spectrum. Specifically, the spectral fitting for metabolite was done using the standard Chenomx 700 MHz metabolite library. Most of the visible peaks were annotated with a compound name. Each spectrum was processed and analyzed by at least 2 NMR spectroscopists to minimize compound misidentification and misquantification. Each metabolite was scaled as the percentage of total metabolite concentrations. One metabolite, isopropanol, represented more than 50% of the total metabolites in many samples, and is likely a breakdown product of lubricant gel used during endoscopy; it was excluded from analyses.

Effect of Identified Metabolites on In Vitro Invasion of Known Escherichia coli Strains

For the in vitro experiments, the above-mentioned gentamicin assay was conducted on the E coli strains, EHEC, and LF82, with or without the addition of succinate (50 mmol/L), acetate (25 mmol/L), or formate (30 ml/L) in order to validate the role of these specific metabolites.

Statistical Analysis

Data were analyzed using Graph Pad Prism (Graph Pad Software, San Diego, CA). Comparison between the cohort groups was done using ANOVA. Correlation analysis was done using Spearman's correlation coefficient. Statistical significance was determined as P < 0.05.

For metabolomic analysis, to identify metabolites that were significantly different amongst non-IBD and IBD patients, fold change analysis, t-tests, and volcano plots were used. P values <0.05 were considered statistically significant. All data for metabolomics were analyzed using Metaboanalyst 3.0 (www.metaboanalyst.ca) and were normalized using log transformation, and autoscaled using data-scaling algorithm before conducting statistical analysis.

RESULTS

Patients

Thirty-one patients were recruited; complete data are available for 29 patients, including 10 non-IBD controls, 9 CD, and 10 UC patients. Two patients were excluded because of technical challenges with NMR analysis. Six CD patients were newly diagnosed and 3 were previously diagnosed. All UC patients were previously diagnosed and in remission, confirmed by both endoscopy and histology. Patient characteristics are described in Supplementary Table 1 (Supplemental Digital Content, https://links.lww.com/MPG/B886).

Characterization of Bacterial Isolates

A total of 100 bacterial isolates were characterized by sequencing from non-IBD and IBD patients.

Similar Invasion of Intestinal Epithelial Cells by Bacterial Isolates from Noninflammatory Bowel Disease and Inflammatory Bowel Disease Patients

Invasion of bacteria (all E coli), isolated from non-IBD, CD, and UC patients into T84 and Caco-2 cells using gentamicin protection assays showed no difference in the invasion capacity of bacteria from the different groups (Fig. 1A: Non-IBD: 983.1 ± 309.5 [mean ± SEM; standard error of the mean]; CD: 1889 ± 1280; UC: 482.7 ± 118.3 CFU/mL; ANOVA: P > 0.05, N = 14 non-IBD, 8 CD, and 9 UC bacterial isolates, respectively; T84 cell lines). Invasion of the positive control LF82 was higher than seen with bacteria from all patient groups (LF82: 48359 ± 37797, Student's t-test, P < 0.05). Similar findings were observed in Caco-2 cells (Fig. 1B).

F1
FIGURE 1:
Similar epithelial cell invasion of intestinal bacteria isolated from the terminal ileum of noninflammatory bowel disease and inflammatory bowel disease patients. (A) There was no difference in the invasion using gentamicin protection assay of bacterial isolates from the TI mucosal surface from non-IBD, CD, and UC patients into T-84 cells. There was, however, a significant difference (P < 0.05) between bacteria from all patient groups and E coli, strain LF82, used as a positive control for invasion (T84 cells. N = 13 for non-IBD, 9 CD, and 10 UC). (B) Similar results to T-84 cells were obtained when bacterial isolates were tested on Caco-2 cells. CD = Crohn disease; IBD = inflammatory bowel disease; UC = ulcerative colitis.

Effects of the Intestinal Microenvironment on Bacterial Invasion

Invasion of EHEC and LF82 increased when Caco-2 cells were incubated with duodenal and TI aspirates from CD and UC patients (Fig. 2A and B). Incubation of EHEC with duodenal aspirates of UC patients resulted in increased invasion compared with incubation with non-IBD duodenum aspirates (Fig. 2C: non-IBD: 920.6 ± 693.6; UC: 4410 ± 1494, CFU/mL; ANOVA: P < 0.05). Invasion of LF82 increased when incubated with aspirates from CD patients, compared with non-IBD cases (non-IBD: 206.7 ± 118.8, CD: 5193 ± 917.5 CFU/mL; ANOVA, P < 0.05) in T84 cells. Interestingly, incubation of LF82 bacteria with aspirates from non-IBD patients caused decreased invasion compared with when they were incubated with normal saline (non-IBD: 206.7 ± 118.8, NS: 1978 ± 248.8, P < 0.05). EHEC invasion in the presence of TI aspirates from UC patients increased significantly compared with those incubated with aspirates from non-IBD cases (Fig. 2D: Non-IBD: 2263 ± 663.9; UC: 72115 ± 4525 CFU/mL; ANOVA: P < 0.05). The invasion potential of LF82 increased in the presence of TI aspirates from CD patients (Fig. 2D: non-IBD 1038 ± 173.2, CD: 4741 ± 2754 CFU/mL; ANOVA: P < 0.05, N = 9 non-IBD, CD, and UC patients).

F2
FIGURE 2:
Luminal aspirates from the duodenum and terminal ileum of inflammatory bowel disease patients increase invasion of bacteria into Caco-2 (A: duodenum; B: TI) and T-84 cells (C: duodenum; D: TI). Gentamicin assays on 3 known E coli strains with various pathogenic potential were conducted to test the effects of the intestinal aspirates on bacterial invasion. The invasion potential of both EHEC and LF82 increased when Caco-2 cells were incubated with duodenal (A) and TI (B) aspirates from CD and UC patients. (C) Incubation with aspirates from the duodenum of UC patients resulted in increased invasion of EHEC as compared with incubation with non-IBD duodenum aspirates. Aspirates from CD patients increased invasion of LF82 compared with non-IBD cases. (D) EHEC invasion in the presence of TI aspirates from UC patients increased significantly compared with incubation with aspirates from non-IBD cases. The invasion potential of LF82 increased in the presence of TI aspirates from CD patients (ANOVA: P < 0.05; n = 9 non-IBD, CD, and UC patients). CD = Crohn disease; EHEC = enterohemorrhagic Escherichia coli; IBD = inflammatory bowel disease; TI = terminal ileum; UC = ulcerative colitis.

Metabolomic Analysis of Intestinal Aspirates

NMR-based metabolomics of intestinal aspirates detected, on average, 40 metabolites in each sample (Fig. 3 ). Although all 29 TI samples were analyzed (Fig. 3 A), 16 of the duodenal ones (Fig. 3 B) could not be analyzed because of high salt content and sample loss for optimization of this technique.

F3
FIGURE 3:
Global profiling of metabolites in luminal washes from the duodenum and terminal ileum of noninflammatory bowel disease and inflammatory bowel disease patients. Analysis of intestinal aspirates with NMR revealed the concentrations of specific metabolites present in noninflammatory bowel disease, CD, and UC patients. Heat-map of metabolites present in the duodenal aspirates (A) and TI aspirates (B) of patients represent the metabolite variation amongst the groups. The heatmap is log2-based, with dark blue colour representing the lowest and dark red the highest concentration of metabolites. CD = Crohn disease; NMR = nuclear magnetic resonance spectroscopy; TI = terminal ileum; UC = ulcerative colitis.
F4
FIGURE 3 (Continued):
Global profiling of metabolites in luminal washes from the duodenum and terminal ileum of noninflammatory bowel disease and inflammatory bowel disease patients. Analysis of intestinal aspirates with NMR revealed the concentrations of specific metabolites present in noninflammatory bowel disease, CD, and UC patients. Heat-map of metabolites present in the duodenal aspirates (A) and TI aspirates (B) of patients represent the metabolite variation amongst the groups. The heatmap is log2-based, with dark blue colour representing the lowest and dark red the highest concentration of metabolites. CD = Crohn disease; NMR = nuclear magnetic resonance spectroscopy; TI = terminal ileum; UC = ulcerative colitis.

Volcano analysis identified 7 metabolites in the TI that were significantly different between non-IBD and UC patients, and t-test identified 1 additional metabolite. KEGG Pathway analysis of these metabolites showed their association with arginine, proline, glutathione, carbohydrate, lipid, serine, and threonine metabolism (Supplementary Table 2, Supplemental Digital Content, https://links.lww.com/MPG/B886, fold change threshold >2, P < 0.05). Four additional metabolites were identified, including succinate, that is used by EHEC under gluconeogenic conditions (10) but were not statistically different. Serine was the only metabolite significantly different between CD and non-IBD (Supplementary Table 2, Supplemental Digital Content, https://links.lww.com/MPG/B886 Lysine (UC vs non-IBD) and valerate (CD vs non-IBD) were the only statistically significantly different metabolites in the duodenum, likely because of smaller patient numbers (Supplementary Table 3, Supplemental Digital Content, https://links.lww.com/MPG/B886).

Escherichia coli Invasion Correlates Positively With Luminal Succinate

Correlation of the metabolites present in the TI aspirates with E coli invasion showed that higher invasion of both LF82 and EHEC was associated with high levels of succinate in UC patients (Fig. 4 A: Spearman ’r’ coefficient: 0.8, P < 0.05).

F5
FIGURE 4:
Succinate correlates with and reproduces intestinal aspirate effect on Escherichia coli invasion in vitro and enterohemorrhagic Escherichia coli invasion correlates negatively with acetate and formate in noninflammatory bowel disease patients: (A) invasion of E coli strains LF82 and EHEC positively correlated with the concentration of succinate in UC patients; Spearman ’r’ coefficient: 0.7, P < 0.05. (B and C) Invasion of EHEC (P < 0.05). (D and E) Invasion of LF82 (P < 0.05). Gentamicin assay was conducted on the E coli strains, EHEC and LF82 with the addition of succinate (50 mmol/L), and incubation was conducted for 2 hours at 37 °C. (F) Acetate negatively correlates with epithelial Invasion of EHEC in non-IBD patients; Spearman ’r’ coefficient: −0.7, P < 0.05. (G) Invasion of EHEC is decreased in the presence of high levels of formate in non-IBD patients; Spearman ’r’ coefficient: 0.9, P < 0.05. (H and I) The invasion of EHEC and LF82 is unaltered in the presence of acetate in T84 cells (P > 0.05). (J and K) The invasion of EHEC and LF82 is unaltered in the presence of formate in T84 cells (P > 0.05). Gentamicin assay was conducted on the E coli strains, EHEC and LF82 with the addition of acetate (25 mmol/L) or formate (30 mm/L) and incubation was conducted for 2 hours at 37 °C. EHEC = enterohemorrhagic Escherichia coli; UC = ulcerative colitis.
F6
FIGURE 4 (Continued):
Succinate correlates with and reproduces intestinal aspirate effect on Escherichia coli invasion in vitro and enterohemorrhagic Escherichia coli invasion correlates negatively with acetate and formate in noninflammatory bowel disease patients: (A) invasion of E coli strains LF82 and EHEC positively correlated with the concentration of succinate in UC patients; Spearman ’r’ coefficient: 0.7, P < 0.05. (B and C) Invasion of EHEC (P < 0.05). (D and E) Invasion of LF82 (P < 0.05). Gentamicin assay was conducted on the E coli strains, EHEC and LF82 with the addition of succinate (50 mmol/L), and incubation was conducted for 2 hours at 37 °C. (F) Acetate negatively correlates with epithelial Invasion of EHEC in non-IBD patients; Spearman ’r’ coefficient: −0.7, P < 0.05. (G) Invasion of EHEC is decreased in the presence of high levels of formate in non-IBD patients; Spearman ’r’ coefficient: 0.9, P < 0.05. (H and I) The invasion of EHEC and LF82 is unaltered in the presence of acetate in T84 cells (P > 0.05). (J and K) The invasion of EHEC and LF82 is unaltered in the presence of formate in T84 cells (P > 0.05). Gentamicin assay was conducted on the E coli strains, EHEC and LF82 with the addition of acetate (25 mmol/L) or formate (30 mm/L) and incubation was conducted for 2 hours at 37 °C. EHEC = enterohemorrhagic Escherichia coli; UC = ulcerative colitis.

Addition of Succinate Increases Escherichia coli Invasion into T84 Cells and enterohemorrhagic Escherichia coli Invasion Correlates Negatively With Acetate and Formate in Noninflammatory Bowel Disease Patients

Invasion assays on T84 and Caco-2 cells, incubating EHEC and LF82 strains with succinate, with or without aspirates from TI of UC patients showed increased invasion in the presence of succinate (Fig. 4 B--E; P < 0.05). The addition of succinate resulted in similar invasion to that seen with the aspirate alone, suggesting that succinate could explain and recapitulate the majority of this effect. Furthermore, succinate appeared to have an additive effect to the aspirate (Fig. 4 B and E).

Correlation of metabolites present in the aspirates with E coli invasion in each group showed that lower invasion of EHEC was associated with acetate and formate in non-IBD patients (Fig. 4 F and G: Spearman's ’r’ coefficient: −0.7, −0.9, P < 0.05), Incubation of EHEC with aspirates from non-IBD patients with acetate and formate did not, however, cause a reduction in the invasion of the bacteria (Fig. 4 H--K; P > 0.05).

DISCUSSION

The aim of this study was to assess host-microbial interactions in IBD by examining the location where this interface takes place—the gut mucosal surface. We analyzed whether microbes isolated from intestinal washes from non-IBD and IBD patients differ in their invasive capabilities in vitro, when removed from the gut, and did not find a difference. Therefore, even if microbes in patients with IBD are more virulent, virulence is lost after they are removed from the gut environment.

We then explored potential factors in the gut mucosal interface that could possibly enhance bacterial virulence by using laboratory strains of E coli with variable virulent potential (commensal, pathobiont, and pathogen). We found that invasion of EHEC (pathogen), when incubated with intestinal aspirates from either the duodenum or TI of UC patients, increased significantly as compared with incubation with aspirates from non-IBD patients using 2 different epithelial cell lines. CD is a heterogeneous disease and can occur in any part of the gastrointestinal tract. Incubation of aspirates from CD patients caused an increase in the invasion of AIEC, strain LF82. Thus, our results show that the gut microenvironment and local changes can affect how bacteria behave.

Certain metabolites can be biomarkers of immunometabolic changes in diseases, including succinate (11,12) and citrate (13). Mass-spectrometry (14), NMR (15), and liquid chromatography mass spectrometry (LC)-MS/MS are used to assess the gut metabolome (16). In IBD patients, metabolomic analysis has been mostly conducted on urine or stool samples (17–20). We have analyzed, for the first time to our knowledge, metabolites in samples that are in close proximity to the actual disease site—luminal aspirates. Using NMR-based metabolomics, we identified a potential driver of microbial virulence, succinate. This metabolite plays a significant role in inflammation, by stimulating dendritic cells (12,21). Succinate is used by EHEC in gluconeogenic conditions and expression of E coli genes associated with formation of attaching and effacing lesions is increased in the presence of succinate (10). Our data show a correlation between the cell invasion capacity of E coli strains, EHEC and LF82, and succinate, further validated by exogenously adding succinate to our in vitro experimental model. The direct mechanistic effects of succinate on the virulence of LF82, however, remain to be explored. We acknowledge that although 16S sequencing and metabolomics have shown interesting results related to host-microbial changes in IBD, the results can be strengthened by using Next Generation Sequencing in future studies (22).

Our study has several limitations. Firstly, we were not able to incubate anaerobic bacteria with epithelial cells (low oxygen killed the epithelial cells; high oxygen killed the anaerobic bacteria) and, therefore, chose to focus on E coli strains. We do recognize that most of the luminal bacteria are anaerobic but argue that the mucosal surface has higher oxygen concentrations, especially in IBD (23), making our model more relevant, but still not ideal. Given that we needed to culture bacteria that we could regrow and use for infections, we needed to accept this growth bias. Another potential challenge was that aspirates from IBD patients could have adverse effects on cells in culture; therefore, cell lines were incubated with aspirates only to see if the cells are destroyed by the aspirates; however, this was not the case. A limitation regarding the inclusion of the control group was that although completely healthy children would have been better controls, they cannot ethically be offered endoscopy. Also, the effects of medications on the gut microenvironment and microbes cannot be ignored. We ensured that none of our patients were on antibiotics for the last 3 months. We realize that the presence of diarrhoea in controls who, although had normal endoscopies, may have influenced the results as the aetiology of the diarrhoea was unknown. Importantly, given the small sample size is, future studies encompassing larger sample sizes are required to reproduce our findings. In that sense, our study should be considered as a proof-of-concept study given the small sample size; however, similar sample sizes have provided important insight into host-microbe interactions in the setting of the human gut and IBD (24,25).

Our study implies that alterations in the gut microenvironment in IBD can affect bacterial invasion capacity. The therapeutic potential of our research requires better understanding of the mechanisms by which metabolites alter the immune system and bacteria, which could be utilized to develop drugs, or even targeted nutrition that would reduce bacterial virulence, in the setting of IBD.

Acknowledgments

We thank our patients, their families and the hospital staff for their assistance. Special thanks to Cheryl Kluthe, RN, Leanne Shirton, RN, and Pavel Medvedev, EPIC research coordinator.

REFERENCES

1. Benchimol EI, Manuel DG, Guttmann A, et al. Changing age demographics of inflammatory bowel disease in Ontario, Canada. Inflamm Bowel Dis 2014; 20:1761–1769.
2. Matsuoka K, Kanai T. The gut microbiota and inflammatory bowel disease. Semin Immunopathol 2015; 37:47–55.
3. Machiels K, Joossens M, Sabino J, et al. A decrease of the butyrate-producing species Roseburia hominis and Faecalibacterium prausnitzii defines dysbiosis in patients with ulcerative colitis. Gut 2013; 63:1275–1283.
4. Zaidi D, Jorgensen M, Huynh HQ, et al. Increased epithelial gap density in the non-inflamed duodenum of children with inflammatory bowel diseases. J Pediatr Gastroenterol Nutr 2016; 63:644–650.
5. Zaidi D, Churchill L, Huynh HQ, et al. Capillary flow rates in the duodenum of pediatric ulcerative colitis patients are increased and unrelated to inflammation. J Pediatr Gastroenterol Nutr 2017; 65:306–310.
6. Levine A, Koletzko S, Turner D, et al. European Society of Pediatric Gastroenterology, Hepatology, and Nutrition. ESPGHAN revised porto criteria for the diagnosis of inflammatory bowel disease in children and adolescents. J Pediatr Gastroenterol Nutrn 2014; 58:795–806.
7. Darfeuille-Michaud A. Adherent-invasive Escherichia coli: a putative new E. coli pathotype associated with Crohn's disease. Int J Med Microbiol 2002; 292:185–193.
8. Saude EJ, Slupsky CM, Sykes BD. Optimization of NMR analysis of biological fluids for quantitative accuracy. Metabolomics 2006; 2:113–123.
9. Xia J, Sinelnikov IV, Han B, et al. MetaboAnalyst 3.0-making metabolomics more meaningful. Nucleic Acids Res 2015; 43:W251–W257.
10. Njoroge JW, Nguyen Y, Curtis MM, et al. Virulence meets metabolism: Cra and KdpE gene regulation in enterohemorrhagic Escherichia coli. MBio 2012; 3:e00280-12.
11. Sadagopan N, Li W, Roberds SL, et al. Circulating succinate is elevated in rodent models of hypertension and metabolic disease. Am J Hypertens 2007; 20:1209–1215.
12. Mills E, O’Neill LAJ. Succinate: a metabolic signal in inflammation. Trends Cell Biol 2014; 24:313–320.
13. Kelly B, O’Neill LA. Metabolic reprogramming in macrophages and dendritic cells in innate immunity. Cell Res 2015; 25:771–784.
14. Ahmed I, Greenwood R, Costello B, et al. Investigation of faecal volatile organic metabolites as novel diagnostic biomarkers in inflammatory bowel disease. Aliment Pharmacol Ther 2016; 43:596–611.
15. Wissenbach DK, Oliphant K, Rolle-Kampczyk U, et al. Optimization of metabolomics of defined in vitro gut microbial ecosystems. Int J Med Microbiol 2016; 306:280–289.
16. Yau YY, Leong RWL, Shin S, et al. Bimodal plasma metabolomics strategy identifies novel inflammatory metabolites in inflammatory bowel diseases. Discov Med 2014; 18:113–124.
17. Jansson J, Willing B, Lucio M, et al. Metabolomics reveals metabolic biomarkers of Crohn's disease. PLoS One 2009; 4:e6386.
18. Marchesi JR, Holmes E, Khan F, et al. Rapid and noninvasive metabonomic characterization of inflammatory bowel disease. J Proteome Res 2007; 6:546–551.
19. Karu N, Deng L, Slae M, et al. A review on human fecal metabolomics: methods, applications and the human fecal metabolome database. Anal Chim Acta 2018; 1030:1–24.
20. Alghamdi A, Gerasimidis K, Blackburn G, et al. Untargeted metabolomics of extracts from faecal samples demonstrates distinct differences between paediatric Crohn's disease patients and healthy controls but no significant changes resulting from exclusive enteral nutrition treatment. Metabolites 2018; 8:821–18.
21. Haas R, Smith J, Rocher-Ros V, et al. Lactate regulates metabolic and proinflammatory circuits in control of T cell migration and effector functions. PLoS Biol 2015; 13:e10022021–24.
22. Lloyd-Price J, Arze C, Ananthakrishnan AN, et al. Multi-omics of the gut microbial ecosystem in inflammatory bowel diseases. Nature 2019; 569:655–662.
23. Albenberg L, Esipova TV, Judge CP, et al. Correlation between intraluminal oxygen gradient and radial partitioning of intestinal microbiota. Gastroenterology 2014; 147:1055.e8–1063.e8.
24. Ferrer-Picón E, Dotti I, Corraliza AM, et al. Intestinal inflammation modulates the epithelial response to butyrate in patients with inflammatory bowel disease. Inflamm Bowel Dis 2020; 26:43.
25. Deehan EC, Yang C, Perez-Muñoz ME, et al. Precision microbiome modulation with discrete dietary fiber structures directs short-chain fatty acid production. Cell Host Microbe 2020; 27:389.e6–404.e6.
Keywords:

bacterial invasion; Crohn disease; metabolites; succinate; ulcerative colitis

Supplemental Digital Content

Copyright © 2020 by European Society for Pediatric Gastroenterology, Hepatology, and Nutrition and North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition