What Is Known
- Pathophysiology of postoperative enterocolitis and bowel functional impairments in patients with Hirschsprung disease (HD) are not fully known, but associated disturbances of intestinal microbiota have recently been reported.
- Sustaining a balanced intestinal microbial community is critical for maintaining intestinal health.
What Is New
- Hirschsprung disease patients showed decreased intestinal microbial richness with expansion of Enterobacteria, which associated with a history of enterocolitis episodes and loose stools.
- Intestinal microbiota in Hirschsprung disease patients did not undergo normal age-related maturation.
- Pathologic alterations of the intestinal microbiota may contribute to postoperative bowel dysfunction in Hirschsprung disease patients.
Hirschsprung disease (HD) is characterized by the absence of intrinsic parasympathetic ganglia (aganglionosis) in varying length of the bowel, resulting in functional obstruction (1). The incidence is 1 to 2 per 10,000 births (2,3). Despite resection of the aganglionic bowel segment during infancy in most cases (4), significant postoperative morbidity commonly occurs due to Hirschsprung-associated enterocolitis (HAEC) (5–7) and other disturbances in bowel function. Although the etiology of these has remained unclear (6,8–10), a role dysbiosis of the gut microbiota has been recently proposed (11–13). Dysbiosis comprises a disruption of the normal microbiota composition and function, including a significant reduction in diversity and an abnormal expansion of new bacterial groups of commensal species (13,14). Gut micronial dysbiosis has already been associated with a number of pathological conditions including inflammatory bowel disease (14,15) and a role in HAEC has been implicated (16,17).
Food-allergies, particularly cow's milk allergy, have also been associated with an increased prevalence of HAEC (6) and postoperative bowel symptoms in HD (18–20). Connections between the LCT genotype C/C (21), that leads to lactase deficiency (22), and HAEC alongside other bowel symptoms in HD has, however, not been explored. Moreover, no studies exist of fecal calprotectin (FC) levels, which are a sensitive marker of intestinal inflammation (23,24), among patients with HD.
This study aimed to characterize the microbiota profiles of patients with HD and to evaluate the microbiota composition in relation to postoperative bowel symptoms including HAEC. To further understand the pathophysiological basis of functional gastrointestinal problems in HD, we investigated FC levels and prevalence of the LCT genotype C/C in these patients.
All patients with HD treated at our tertiary referral center between 1987 and 2011 were identified from records and contacted by post by an independent investigator who had not been involved in their care between 2012 and 2013. They were invited to answer a detailed questionnaire concerning bowel symptoms and medications. On a voluntary basis and after written informed consent, they were also asked to participate on a clinical follow-up, which included giving blood samples for the genotyping of the brush border lactase (LCT) gene, and stool samples for DNA isolation and calprotectin measurement. Patient characteristics, operative treatment, hospital attendances for enterocolitis, and medication history were obtained retrospectively from records. The ethics committee of the Helsinki University Hospital approved the research protocol.
All patients had undergone endorectal pull-through as the definitive surgical treatment. Endorectal pull-through was performed completely transanally or in combination with transabdominal colonic mobilization via laparotomy or laparoscopy. Transanal mucosectomy was commenced 5 mm above the dentate line, proceeding to full-thickness dissection approximately 3 cm thereafter (25). Total colonic aganglionosis was managed by restorative proctocolectomy with a short J-pouch ileoanal anastomosis with a protective temporary ileostomy (26). The same surgical team of consultant pediatric colorectal surgeons operated and followed the patients up until adulthood.
Definitions of Clinical Symptoms
Soiling was defined as small amounts of fecal staining on underclothing, which defined to be frequent as occurring at least once a week. Enterocolitis was defined on an intention-to-treat basis in the presence of clinical symptoms including abdominal distension, diarrhea, and vomiting, and, in some cases, bloody stools and/or fever (27). Enterocolitis was considered recurrent or persistent, when symptoms lead to 2 or more successive treatment episodes. Normal defecation frequency was defined as stooling every other day to twice per day (28).
Postoperative Bowel Function and Medication
Patients without major cognitive impairment, an antegrade continence enema (ACE) or enterostomy (n = 28), answered questions regarding soiling, fecal accidents and constipation from the Rintala Bowel Function Score (BFS) (28) that were assessed with reference to the microbiota composition. Patients were enquired about their current use of antibiotics due to enterocolitis and medication for bowel function. Additionally, stool consistency (formed/loose or liquid) was enquired from all patients without an ACE or enterostomy (n = 34; including 6 with cognitive impairment) or their caregivers.
Fecal DNA Extraction and Sequencing
The patients with an ACE (n = 3; bowel emptying with washouts) and a proximal small bowel enterostomy (n = 1) were excluded from the microbiome analysis due to the invalid fecal sample from dilutional effects. Fecal samples were collected and frozen at −20°C within 4 hours of collection and stored at −70°C until analysis. Bacterial DNA was extracted from the fecal samples using the repeated bead beating method (29), and 16S rDNA amplicon sequencing was used to analyze the microbiota composition. The V3-V4 variable region was sequenced using Illumina HiSeq using the forward primer CTACGGGNGGCWGCAG and the reverse primer GACTACHVGGGTATCTAATCC. The analysis in this study is, however, based on V3 region only (see below).
The DNA amplicon sequences were processed using the R package mare (30). Only the forward reads were analyzed. The reads were trimmed to 150 bases and reads representing <0.001% of all unique reads were discarded as potentially erroneous. For taxonomic annotation, the reads were not clustered into operational taxonomic units (OTUs) but, instead, mapped to the Silva reference database for taxonomic annotation, as OTU clustering would cause unnecessary errors to the taxonomic annotation. After annotation, the abundances reads were summarized at different taxonomic levels. OTU clustering, using USEARCH (31), was done only for the richness analysis, as it is not possible to estimate richness based only on the number of unique reads or taxonomic classification. Furthermore, taxonomic classification of 16S reads at the species level is not reliable. Therefore, microbial richness was defined as the number of species-like OTUs (97% similarity) present in the sample.
Analysis of Carbohydrate-active Enzyme Levels and Carbohydrate Degradation Potential
To further evaluate the function of the gut microbiome, carbohydrate-active enzymes and carbohydrate degradation potential using the CAZy database (32). Carbohydrate active enzyme abundances were predicted based on the observed taxonomic profiles.
The level of FC was measured in a routine clinical laboratory using a quantitative, enzyme-linked immunoassay (PhiCal Test, Calpro AS, Oslo, Norway; NovaTec Immunodiagnostica, Dietzenbach, GmBH, Germany) in routine analyses of the central laboratory services of the Helsinki University Hospital. A level of 100 μg/g was taken as the upper limit of normal (15,33).
Brush Border Lactase Genotype
For the lactase gene LCT genotype (LCT-13910), genomic DNA was extracted from blood samples and genotyped using standard methods (34). The LCT genotype C/C results in lactase deficiency and lactose intolerance (22).
Of a previously collected cohort of 236 Finnish children aged 2 to 7 years without serious chronic illnesses (35), 141 children had donated fecal samples, which served as control data for intestinal microbiota composition. Two independent samples per healthy control were analyzed, resulting in 282 control samples. The control samples were analyzed using a phylogenetic microarray, HITChip. Additionally, a small subset of the control samples was sequenced using the same methods as with the patient samples. To remove potential technical bias, the microarray data were standardized using the sequenced control samples as the reference. The standardization was done as previously described (36).
Statistical analyses were conducted in R, using the package mare, including tools from packages vegan (37), metacoder (38), sp (39), MASS (40), and nlme (41). Data were presented as frequencies or medians ± range unless otherwise stated. Taxa occurring in less than 30% of the samples, that is, taxa occurring in 10 samples or less, were excluded from the taxon-specific analyses. However, the full control data was derived from a microarray experiment and did not contain zeros, and, thus, the tests comparing the patients and the control had no omitted taxa. Analysis of the microbiota composition and relative abundances of bacterial taxa were examined by age, stool consistency, history of antibiotic treatment, symptoms of HAEC, LCT genotype and FC as appropriate, and associations between them were tested using negative binomial models. All analyses, except the patient-control comparisons, were adjusted for patient age, level of aganglionosis, and a history of HAEC. If these models did not fit the data due to heteroscedasticity, general least squares models were used. P < 0.005 were considered significant.
Of 89 patients responding to the survey questionnaire, 48 patients (54%) participated in the clinical follow-up of whom 38 (43%) gave stool and 37 (41%) blood samples for laboratory analysis. The baseline patient characteristics and surgical details of participants are presented in Table 1. After excluding patients with ACE (n = 3) or small bowel enterostomy (n = 1), 34 stool samples remained for the microbiota analysis, including 28 patients without major cognitive impairment who had also participated in the functional outcomes survey.
Bowel Functional Outcomes
Any fecal soiling was reported by 50% (n = 19/28), and in 33% it occurred frequently. Stooling frequency was abnormally increased in 36% (n = 10/28), and 56% (n = 19/34) experienced loose or liquid stools. No participants reported constipation. Six patients (n = 6/28; 21%) currently used medication for bowel function, including loperamide (n = 4), psyllium (n = 1), and macrogols combined with weekly enema of docusate sodium and sorbitol solution (n = 1) due to frequent soiling.
Two patients (n = 2/34; 6%) had experienced at least 1 episode of enterocolitis before definitive surgery, 77% of patients (n = 26/34) had experienced at least 1 episode after definitive surgery, and 53% (n = 18/34) had a history of recurrent or persistent HAEC. All patients with recurrent HAEC had received multiple courses of oral antibiotics, and 83% of them (n = 15/18) had previously or currently used prophylactic oral antibiotics long term. At the time of the survey, 2 patients (n = 2/34; 6%) were currently taking prophylactic oral metronidazole, but none reported having HAEC at the time. Additionally, 4 patients (12%) were taking regular probiotics.
Fecal Calprotectin Level
The median FC was 9 (range 0–65) μg/g among patients at the time of sampling and therefore within the normal range.
Brush Border Lactase Genotype
Blood samples were available for LCT genotyping from 35/38 patients (1 refusal, 2 missing samples). The LCT genotype C/C was detected in 6 participants (n = 6/35; 17%). No associations were found between the C/C genotype and HAEC, fecal soiling, increased stooling frequency or abnormal stool consistency (P < 0.05 for all).
Intestinal Microbiota Composition
Compared to healthy controls, patients with HD had significantly increased relative abundances of Escherichia, Pseudomonas, Dialister, Actinomyces, Bacilli, and Prevotella (shown in red in Fig. 1) and decreased abundances of most normally abundant taxa, including Bacteroidales, Ruminococcaceae, and Lachnospiraceae (shown in blue in Fig. 1; P < 0.005 [FDR-corrected P-value <0.01 for all]). The levels of major bacteria against different variables are listed in the supplementary tables (Supplemental Digital Content 1–4, http://links.lww.com/MPG/B359, http://links.lww.com/MPG/B360, http://links.lww.com/MPG/B361, http://links.lww.com/MPG/B362). Patients with HD had a predicted lower abundances of carbohydrate-active enzymes compared with controls, based on the taxonomic data (P < 0.005).
Effects of age on Microbiota Composition
Among patients with HD, the microbiota composition showed some changes with age: the abundance of Christensenella decreased, and the abundance of Clostridia, Asaccharobacter, Collinsella, and Lactococcus increased (Supplemental Digital Content 2, http://links.lww.com/MPG/B360); P < 0.005), but the DNA richness stayed constant up to the age of 25 years (P = 0.72).
Intestinal Microbiota and Recurrent Enterocolitis
Patients with a history of recurrent HAEC had a significantly lower microbial richness (mean 112 OTUs) than the patients without a history of recurrent HAEC [(mean 150 OTUs); P = 0.004]. Abundances of Proteobacteria and Lactobacillus were significantly increased, and abundances of Clostridia and Prevotella were significantly decreased in patients with a history of recurrent HAEC (p < 0.005 for all; Fig. 2) Moreover, regarding some taxa, the patients with a history of recurrent HAEC had significantly differing abundances compared both to healthy controls and patients without HAEC (P < 0.05). Lactococcus and Escherichia were dramatically increased among the patients with HAEC, but only modestly in the patients without HAEC (P < 0.05). Conversely, the abundance of Oscillospira and Holdemania were significantly decreased in the patients with a history of HAEC compared with those without it (P < 0.05). Furthermore, a few taxa, such as Roseburia, were equally low in both patient groups compared with the healthy controls (Fig. 3). The predicted enzyme-levels for carbohydrate degradation were decreased (P < 0.005). The microbiota compositions of both of these patient groups, however, clearly differed from the controls, and we acknowledged that the changes of the microbiota in the patients with a history of HAEC are closely associated to the use of oral antibiotics (Supplemental Digital Content 5, http://links.lww.com/MPG/B363).
Intestinal Microbiota and Bowel Functional Impairment
Patients with normal stool consistency had a lower abundance of Proteobacteria and higher abundances of Actinomyces compared to those with abnormal stool consistency (P < 0.005) (Supplemental Digital Content 6, http://links.lww.com/MPG/B364). Additionally, patients with frequent soiling had decreased abundance of Clostridia in relation to controls (P < 0.005) No associations were found between increased stooling frequency and the microbiota composition (P > 0.005).
Antibiotic usage, which is common in patients with HD, can modify the microbiota composition significantly (13,14). In children, a single course of oral antibiotics has shown to induce long-term, but largely reversible changes in the microbiota, including decrease in the abundances of Firmicutes and Actinobacteria and increase in the abundances of Bacteroidetes and Proteobacteria (13,14). Established dysbiosis from repeated courses of oral antibiotics can lead to permanent disturbances in the metabolic and immunological health (13,42). In this study, all patients with a history of recurrent HAEC had taken multiple courses of oral antibiotics during childhood, and most also had a history of taking long-term prophylactic antibiotics. Another important finding was that the patients with HD had an immature microbiome composition, as indicated by decreased microbial richness and high abundances of Bacilli and Enterobacteria. These bacteria are normally present at high abundance in infants, but not in adults (43). The immature type of microbiota composition could result from HD itself (44), antibiotics or dietary alterations. As microbial richness is essential for the stability and resilience of the intestinal microbiota (45,46), the lack of maturation and loss of richness found in the patients with HD could increase the vulnerability of the gut microbiome to colonization by pathogenic bacteria and permit pathological outgrowth of selected taxa (45,46). Our previous work has shown that the tendency to HAEC symptoms decreases over time (47). Therefore, most antibiotic treatments are received by our patients during early childhood around the time that normal microbiome maturation should be taking place.
Among patients with recurrent HAEC, particularly the expansion of Proteobacteria was an important finding. Proteobacteria have been previously implicated in the pathogenesis of HAEC (11,16) and other inflammatory and diarrheal gastrointestinal diseases including IBD (48–52). Proteobacteria are gram-negative opportunistic pathogens that produce potent inflammatory compounds, lipopolysaccharides (48). In healthy individuals, Proteobacteria comprise only a small proportion (1%–2%) of the microbiota, and expanded abundances have been suggested to be a diagnostic microbial signature of epithelial dysfunction, dysbiosis and risk for disease development (48–50). Consistent with this finding, patients with loose stools at the time of survey were also found to have increased abundances of Proteobacteria in their stool. Furthermore, the patients with frequent fecal soiling had less Clostridia in their stool. Although transit time is, however, known to determine the gut microbial habitat as it affects the nutrient and water absorption along the intestinal wall (51,52), we did not find an association between microbiota composition or increased stooling frequency. This could be due to the low sample size of our cohort and large variation caused by other factors, such as antibiotic use history.
The anaerobic bacteria of the colon normally ferment complex dietary carbohydrates to short-chain fatty acids (49). Short-chain fatty acids may have a marked nutritional and functional value to the gut epithelium (49), and alterations in their levels may therefore impact on the bowel function. In this study, patients with HD had reduced levels of carbohydrate-active enzymes in relation to controls, which may be a consequence of the low abundances of Clostridia and Bacteroidia, which are the dominant carbohydrate-degrading taxa. The expansion of Proteobacteria observed is likely to be another important etiologic factor, as Proteobacteria participate very little in the digestion and fermentation of carbohydrates. The degradation of complex carbohydrates is one of the most important digestive functions that the microbiota perform for the human host, who intrinsically has a very limited repertoire of carbohydrate-active enzymes. This can serve to explain why patients with HD commonly report gastrointestinal disturbances akin to lactose intolerance, although the actual prevalence of lactase deficiency (genotype C/C) was only 17% in our study and no correlation with bowel symptoms was found. No patients reported HAEC or active bowel inflammation at the time of calprotectin measurement, and consistent with this FC levels were within the normal range in all cases.
A limitation of this study is clearly the small number of patients. Also, it cannot be shown from the data whether HD itself causes intrinsic disturbances to the gut microbiome because most patients in this series had received antibiotics at the time of study. Additionally, we found that the patients without a history of recurrent HAEC had a microbiota that in some respects resembled more that of healthy controls compared to the patients with a history of recurrent HAEC. Further prospective evaluation is needed to establish the natural history and maturation of the gut microbiome in HD, the impact of this on HAEC and to re-evaluate the rationality of antibiotic treatment and prophylaxis of HAEC.
1. Obermayr F, Hotta R, Enomoto H, et al. Development and developmental disorders of the enteric nervous system. Nat Rev Gastroenterol Hepatol
2. Bradnock TJ, Knight M, Kenny S, et al. Hirshcsprung's disease in the UK and Ireland: incidence and anomalies. Arch Dis Child
3. Löf Granström A, Svenningsson A, Hagel E, et al. Maternal risk factors and perinatal characteristics for Hirschsprung disease. Pediatrics
2016; 138: doi: 10.1542/peds.2015-4608.
4. Dasgupta R, Langer JC. Transanal pull-through for Hirschsprung disease. Semin Pediatr Surg
5. Langer JC, Durrant AC, de la Torre L, et al. One-stage transanal Soave pullthrough for Hirschsprung disease. Ann Surg
6. Bjornland K, Pakarinen MP, Stenstrom P, et al. A Nordic multicentre survey of long-term bowel function after transanal endorectal pull-through in 200 patients with rectosigmoid Hirschsprung disease. J Pediatr Surg 2017; 52:1458–1464.
7. Catto-Smith AG, Trajanovska M, Taylor RG. Long-term continence after surgery for Hirschsprung's disease. J Gastoenterol Hepatol
8. Ieiri S, Nakatsuji T, Akiyoshi J, et al. Long-term outcomes and the quality of life of HD in adolescent who have reached 18 years or older—a 47-year single-institute experience. J Pediatr Surg
9. Stensrud KJ, Emblem R, Bjornland K. Anal endosonography and bowel function in patients undergoing different types of endorectal pull-through procedures for Hirschsprung disease. J Pediatr Surg
10. Neuvonen MI, Kyrklund K, Rintala RJ, et al. Bowel function and quality of life after transanal endorectal pull-through for Hirschsprung disease: controlled outcomes up to adulthood. Ann Surg
11. Frykman PK, Nordensköld A, Kawaguchi A, et al. Characterization of bacterial and fungal microbiome in children with Hirschsprung disease with and without a history of enterocolitis: a multicentre study. PLoS One
2015; 10: e0124172.
12. Yan Z, Poroyko V, Gu S, et al. Characterization of the intestinal microbiome of Hirschsprung's disease with and without enterocolitis. Biochem Biophys Res Commun
13. Korpela K, Salonen A, Virta LJ, et al. Intestinal microbiome is related to lifetime antibiotic use in Finnish pre-school children. Nat Commun
14. Weiss GA, Hennet T. Mechanisms and consequences of intestinal dysbiosis. Cell Mol Life Sci 2017; 74:2959–2977.
15. Kolho K-L, Korpela K, Jaakkola T, et al. Fecal microbiota in pediatric inflammatory bowel disease and its relation to inflammation. Am J Gastroenterol
16. Li Y, Poroyoko V, Yan Z, et al. Characterization of intestinal microbiomes of Hirschsprung's disease patients with or without enterocolitis using Illumina-MiSeq High-throughput sequencing. PLOS One 2017; 11:e0162079.
17. Jiao CL, Chen XY, Feng JX. Novel insight into the pathogenesis of Hirschsprung’-associated enterocolitis. Chin Med J (Engl)
18. Umeda S, Kawahara H, Yoneda A, et al. Impact of cow's milk allergy on enterocolitis associated with Hirschsprung's disease. Pediatr Surg Int
19. Katz Y, Goldenberg MR, Rajuan N, et al. The prevalence and natural course of food protein induced enterocolitis syndrome to cow's milk: a large-scale, prospective population based study. J Allergy Clin Immunol
2011; 127:647–653. e1-3.
20. Mehr S, Kakakios A, Frith K, et al. Food protein-induced enterocolitis syndrome: 16-year experience. Pediatrics
21. Rasinperä H, Savilahti E, Enattah NS, et al. A genetic test which can be used to diagnose adult-type hypolactasia in children. Gut
22. Hubacek JA, Adamkova V, Sedova L, et al. Frequency of adult type-associated lactase persistence LCT-13910C/T genotypes in the Czech/Slav and Chech Roma/Gypsy populations. Genet Mol Biol
23. Herrera OR, Christensen ML, Helms RA. Calprotectin: clinical applications in pediatrics. J Pediatr Pharmacol Ther
24. Albanna EA, Ahmed HS, Awad HA. Stool calprotectin in necrotizing enterocolitis. J Clin Neonatal
25. Rintala RJ. Transanal coloanal pull-through with a short muscular cuff for classic Hirschsprung's disease. Eur J Pediatr Surg
26. Hukkinen M, Koivusalo A, Rintala RJ, et al. Restorative proctocolectomy with J-pouch ileoanal anastomosis for total colonic aganglionosis among neonates and infants. J Pediatr Surg
27. Frykman PK, Kim S, Wester T, et al. Critical evaluation of the Hirschsprung-associated enterocolitis (HAEC) score: a multicentre study of 116 children with Hirschsprung disease. J Pediatr Surg
28. Rintala RJ, Lindahl H. Is normal bowel function possible after repair of intermediate and high anorectal malformations? J Pediatr Surg
29. Salonen A, Nikkilä J, Jalanka-Tuovinen J, et al. Comparative analysis of fecal DNA extraction methods with phylogenetic microarray: effective recovery of bacterial and archaeal DNA using mechanical cell lysis. J Microbiol Methods
30. Korpela K (2016). Mare: microbiota analysis in R easily. R package version 1.0. Available at: doi:10.5281/zenodo.50310 and Github.com/katrikorpela/mare.
31. Edgar RC. Search and clustering orders of magnitude faster than BLAST. Bioinformatics
32. Lombard V, Golaconda Ramulu H, Drula E, et al. The carbohydrate-active enzymes database (CAZy) in 2013. Nucleic Acids Res
2014; 42 (Database issue):D490–D495.
33. Berni Canani R, Rapacciuolo L, Romano MT, et al. Diagnostic value of faecal calprotectin in paediatric gastroenterology clinical practice. Dig Liver Dis
34. Mattar R, Monteiro Mdo S, Villares CA, et al. Single nucleotide polymorphism C/T(-13910), located upstream of the lactase gene, associated with adult-type hypolactasia: validation for clinical practise. Clin Biochem
35. Korpela K, Salonen A, Virta LJ, et al. Lactobacillus rhamnosus GG intake modifies preschool children's intestinal microbiota, alleviates penicillin-associated changes, and reduces antibiotic use. PLoS One
36. Korpela K, Flint HJ, Johnstone AM, et al. Gut microbiota signatures predict host and microbiota responses to dietary interventions in obese individuals. PLoS One
37. Oksanen J, Guillaume Blanchet F, Friendly M, et al. Vegan: Community Ecology Package. R package version 2.4-0. Published: May 2018. Available at: http://CRAN.R-project.org/package=vegan
38. Foster ZSL, Sharpton TJ, Grünwald NJ. Metacoder: an R package for visualization and manipulation of community taxonomic diversity data. PLoS Comput Biol
39. Pebesma EJ. Edzer & Bivand, Roger. Classes and methods for spatial data in R. R News. Published: February 2005. Available at: http://cran.r-project.org/doc/Rnews/
40. Venables WN, Ripley BD. Modern Applied Statistics With S. 4th ed.New York: Springer-Verlag; 2002.
41. Pinheiro J, Bates D, DebRoy S, et al. nlme: Linear and Nonlinear Mixed Effects Models. R package version 3.1-128. Published: April 2018. Available at: http://CRAN.R-project.org/package=nlme
42. Virta L, Auvinen A, Helenius H, et al. Association of repeated exposure to antibiotics with the development of pediatric Crohn's disease—a nationwide, register-based Finnish case-control study. Am J Epidemiol
43. Yatsunenko T, Rey FE, Manary MJ, et al. Human gut microbiome viewed across age and geography. Nature
44. Ward NL, Pieretti A, Dowd SE, et al. Intestinal aganglionosis is associated with early and sustained disruption of the colonic microbiome. Neurogastroenterol Motil
45. Honda K, Littman DR. The microbiota in adaptive immune homeostasis and disease. Nature
46. Fukuda S, Toh H, Hase K, et al. Bifidobacteria can protect from enteropathogenic infection through production of acetate. Nature
47. Neuvonen MI, Kyrklund K, Rintala RJ, et al. Bowel function and quality of life after transanal endorectal pull-through for Hirschsprung Disease: controlled outcomes up to adulthood. Ann Surg
48. Shin NR, Whon TW, Bae JW. Proteobacteria: microbial signature of dysbiosis in gut microbiota. Trends Biotechnol
49. Morowitz MJ, Carlisle E, Alverdy JC. Contributions of intestinal bacteria to nutrition and metabolism in the critically ill. Surg Clin North Am
50. Litvak Y, Byndloss MX, Tsolis RM, et al. Dysbiotic Proteobacteria expansion: a microbial signature of epithelial dysfunction. Curr Opin Microbiol
51. Vandeputte D, Falony G, Vieira-Silva S, et al. Stool consistency
is strongly associated with gut microbiota richness and composition, enterotypes and bacterial growth rates. Gut
52. Gorkiewicz G, Thallinger GG, Trajanoski S, et al. Alterations in the colonic microbiota in response to osmotic diarrhea. PLoS One