Abstract
Environmental enteric dysfunction, an asymptomatic condition characterized by inflammation of the small bowel mucosa, villous atrophy, malabsorption, and increased intestinal permeability, is a major contributor to childhood stunting in low-income countries. Here we report the relationship of increased intestinal permeability with serum metabolites in 315 children without acute malnutrition, aged 12–59 months, in rural Malawi. Increased gut permeability was associated with significant differences in circulating metabolites that included lower serum phosphatidylcholines, sphingomyelins, tryptophan, ornithine, and citrulline, and elevated serum glutamate, taurine, and serotonin. Our findings suggest that environmental enteric dysfunction is characterized by alterations in important metabolites involved in growth and differentiation and gut function and integrity.
Similar content being viewed by others
Introduction
Environmental enteric dysfunction (EED) is highly prevalent in children younger than five years of age in rural sub-Saharan Africa, South America, and South Asia1,2. EED is strongly associated with childhood stunting1,2,3,4. Longitudinal studies of young children in The Gambia showed that >40% of linear growth failure is attributable to EED4. Stunting was associated with increased intestinal permeability but not with diarrhea history5. Close contact with animals, and more frequent geophagy are significant risk factors for EED in young children1,2,6. Chronic fecal-oral transmission of pathogenic bacteria, viruses, and parasites is considered a key factor underlying EED1,2. The pathological findings of EED include villous shortening, crypt hyperplasia, and lymphocytic infiltration7. With the breakdown of the intestinal barrier, there is translocation of bacteria and microbial products into the circulation, leading to chronic immune stimulation and the characteristic histopathology of EED1,2. Accordingly, immunohistochemical studies show T-cell infiltration and elevated proinflammatory cytokines such as tumor necrosis factor-α and interferon-γ in the lamina propria7.
Increased gut permeability could potentially affect the absorption and metabolism of amino acids, proteins, lipids, carbohydrates, and other nutrients1. However, this hypothesis has not been adequately investigated, and the underlying biological mechanisms by which EED could affect child health and linear growth are incompletely understood. This may explain why the many clinical trials that have been conducted with the aims of reducing gut permeability and improving linear growth among children in low-income countries have generally failed to both reduce gut permeability and improve child growth2. A better understanding of the basic biological mechanisms by which EED adversely impacts child health and growth may lead to more effective interventions against EED to reduce childhood stunting. We investigated the hypothesis that increased gut permeability is associated with concentrations of amino acids, glycerophospholipids, and sphingolipids by using a targeted metabolomics approach to characterize serum metabolites in children with and without EED in rural Malawi. We present data on young children from rural Malawi. Our results highlight specific metabolic alterations associated with increased gut permeability in young children with EED that are likely to affect growth and development.
Results
Study setting
We examined the relationship between gut permeability and serum metabolites in 315 children without acute malnutrition, aged 12–59 months, from six villages in rural Malawi. The villages where the children live are comprised largely of subsistence farm families who grow and consume maize as the staple crop. Water is acquired from boreholes or nearby streams. Homes are mostly of mud and thatch. Defecation generally occurs outdoors or in pit latrines. There is frequent contact between children and domestic animals such as chickens and goats. Electricity, running water, and other utilities are invariably absent in the homes. The environment of these rural villages in Malawi is a typical setting for EED. The characteristics of the children with and without EED in the study and demographic features of their household environment are shown in Table 1. There were nearly an equal number of boys and girls. The lactulose:mannitol (L:M) ratio, assessed in the dual-sugar absorption test, was used as the measure of gut permeability and intestinal absorption8. A higher L:M ratio indicates greater gut permeability. Children with an L:M ratio ≥ 0.15 were defined as having EED.9 Serum metabolites were measured using liquid chromatography-tandem mass spectrometry10. The 139 metabolites analyzed included 22 amino acids, 3 biogenic amines, 6 amino acid metabolites, 15 sphingolipids, 8 acylcarnitines, and 85 glycerophospholipids.
Serum metabolites differ by gut permeability
The relationship of gut permeability with all serum metabolites is summarized in a volcano plot in Fig. 1. There were 14 serum metabolites that were significantly related with gut permeability. Serum tryptophan, ornithine, citrulline, four lysophosphatidylcholines (lysoPC a C16:0, lysoPC a C18:0, lysoPC a C18:2, lysoPC a C20:4), two acyl-alkyl-phosphatidylcholines (PC ae C40:1, PC ae C44:3), and two sphingomyelins (SM C16:0, SM C16:1) were negatively correlated with gut permeability (L:M ratio). The general negative correlation of the different serum phosphatidylcholines and sphingomyelins with gut permeability as displayed on the left side of the volcano plot is quite evident. Serum glutamate, serotonin, and taurine were positively correlated with gut permeability. Tryptophan that is converted to serotonin or to kynurenine is no longer available for protein synthesis, thus, ratios of these two metabolites with tryptophan were calculated. The adjusted Spearman correlations between gut permeability and the serotonin/tryptophan and kynurenine/tryptophan ratio were 0.27 (P = 7.34 × 10−7) and 0.25 (P = 6.44 × 10−6), respectively. A summary heatmap of the top fifty serum metabolites by gut permeability and their clusters is shown in Fig. 2. We further investigated potential nonlinear relationships of the metabolites that were significantly correlated with gut permeability. Scatterplots and fitted regression curves (with 95% confidence intervals) using linear mixed-effects models to address within-village clustering with natural cubic splines (3 knots) are shown for the significant metabolites in Fig. 3 and for the serotonin/tryptophan and kynurenine/tryptophan ratios in Fig. 4. While nonlinearity is evident for some metabolites, inflection tends to occur at values of the L:M ratio where there are few data points and therefore uncertainty in estimation.
Broken lines indicate 95% confidence intervals. Data were modeled using linear mixed effects models that include age and sex and village as a random intercept. Correlation coefficients are shown. Abbreviations for lipid nomenclature are described in the methods section. Amino acid abbreviations: Trp (tryptophan), Orn (ornithine), Glu (glutamate), Cit (citrulline).
The age-, sex-, and village-adjusted Spearman correlations between gut permeability and serum metabolites are presented for amino acids, amino acid metabolites, and biogenic amines (Supplementary Table 1), sphingolipids and acylcarnitines (Supplementary Table 2), and glycerophospholipids (Supplementary Table 3). The correlations between gut permeability and serum metabolites ranged between −0.214 (tryptophan) to 0.155 (serotonin). More detailed heatmaps of serum amino acids, biogenic amines, amino acid metabolites, sphingolipids, acylcarnitines, and glycerophospholipids by gut permeability are shown in Supplementary Figs 1 and 2. Exploratory analyses showed that additional adjustment for height-for-age Z-score had little effect upon the relationship of the estimated correlations of serum metabolites with gut permeability. The relationships between urinary mannitol and metabolites were also explored. Only serum taurine and glutamate were associated with urinary mannitol (r = −0.173, P = 0.002; r = −0.157, P = 0.006, respectively). Unadjusted median (interquartile range) of serum metabolite concentrations in children with and without EED are shown for amino acids, biogenic amines, amino acid metabolites, sphingolipids, acylcarnitines, and glycerophospholipids in Supplementary Tables 4–6.
Multiple machine-learning classification algorithms were used to assess the ability of the 14 metabolites significantly correlated with L:M ratio to classify EED status (EED vs. no EED). We used the Super Learner11 algorithm to fit, internally cross-validate, and combine the results into a single classifier. Nine different classifiers were trained on a subset of the data and fit using data omitted for training via leave 10% out internal cross-validation. Next, a weighted-average of the estimates was computed to minimize mean-square error, and the weights themselves were estimated via leave 10% out internal cross-validation (See Supplementary Appendix for additional details). Five algorithms had positive weights (ridge regression12, generalized boosted models13, random forests14, multivariate adaptive regression splines implemented using the earth() function in R software version 3.2.015,16, and Bayes generalized linear models). Supplementary Figure 3 and Supplementary Appendix Table 1 show that the metabolites have a modest ability to discriminate between children with and without EED (cross-validated Super Learner cross-validated area under the ROC curve 0.679, 95% confidence interval 0.610, 0.747). However, the final cross-validated Super Learner model was well calibrated showing little evidence of lack of fit (P = 0.70).
Discussion
The present study shows that children with higher gut permeability, as measured by the L:M ratio, have alterations in serum metabolites that include lower serum phosphatidylcholines, lysophosphatidylcholines, and sphinogomyelins, lower serum tryptophan, ornithine, and citrulline, and higher serum glutamate, taurine, and serotonin. This is the first study to apply a targeted metabolomics approach to gain insight into the pathophysiology of EED. The results suggest that there may be abnormalities in several diverse metabolic pathways in children with increased gut permeability, including the CDP-choline or “Kennedy pathway” for synthesis of phosphatidylcholines, the synthesis of lysophosphatidylcholines and sphingomyelins, the tryptophan-serotonin pathway, the kynurenine pathway, and citrulline synthesis. Perturbations in these pathways and imbalance of metabolites could have potential adverse consequences for child health and growth. The individual pathways are discussed below.
Phosphatidylcholines, the dominant phospholipids in cell membranes and in the circulation, play a major role in cell proliferation, differentiation, and growth17. Phosphatidylcholines are essential for assembly and secretion of very low-density lipoproteins and high-density lipoproteins by the liver18. Phosphatidylcholines are the main active component of pulmonary surfactant and are vital to pulmonary function19. Phosphatidylcholines and lysophosphatidylcholines comprise >90% of the lipids in the protective mucus layer of the gastrointestinal tract20. Phosphatidylcholines are synthesized via the CDP-choline pathway in three enzymatic steps from dietary choline17. Phosphatidylcholines can also be synthesized via the phosphatidylethanolamine N-methyltransferase (PEMT) pathway. However, since the PEMT pathway accounts for ~30% of hepatic synthesis of phosphatidylcholines, it is only quantitatively important in the liver and cannot meet the body’s needs. Thus, choline is considered an essential nutrient21. The availability of phosphatidylcholine affects lysophosphatidylcholine, which is produced from the hydrolysis of phosphatidylcholine. Lysophosphatidylcholines are important constituents of biological membranes and bioactive phospholipids involved in inflammation22.
Since sphingomyelins are the dominant sphingolipids in membranes of mammalian cells, especially the outer leaflet, the endocytic recycling compartment, and the trans Golgi network, they may affect a large number of biological mechanisms. For example, sphingomyelins serve as a precursor to bioactive sphingolipids, are building blocks of lipid rafts, regulate plasma membrane homeostasis, and are involved in T cell activation and differentiation23,24,25. Sphingomyelins play a key role in the myelination of the central nervous system during the development of infants and young children26.
Tryptophan is an essential amino acid for protein synthesis as well as a precursor to the neurotransmitter serotonin. The major route for tryptophan metabolism is the kynurenine pathway in which tryptophan is oxidized by cleavage of the indole ring27. Tryptophan that is metabolized to serotonin or the kynurenine pathway is lost to protein synthesis27. Plasma tryptophan is decreased during inflammation or infection due to increased tryptophan oxidation27. The present study shows that the serotonin/tryptophan and kynurenine/tryptophan ratios are significantly correlated with gut permeability. Limitation of tryptophan availability for protein synthesis could potentially have adverse consequences for child growth. Dietary tryptophan limitation in Sprague Dawley rats resulted in severe growth restriction of young animals28. About 5% of dietary tryptophan is degraded to indole and other derivatives by bacteria in the gut lumen29. Possible mechanisms by which EED could affect tryptophan metabolism is through increased tryptophan degradation by altered gut microbiota, or through alterations in tryptophan transport by amino acid transporters29 in damaged enterocytes.
In contrast to the other amino acids that circulate in free form, about 80–90% of tryptophan is bound to albumin in the circulation, while the remaining free form can be transported across the blood-brain barrier27. Serotonin does not cross the blood-brain barrier, thus, tryptophan can give rise to two separate pools of serotonin in the brain and in the periphery, where it is mostly found in the gastrointestinal mucosa29. A limitation of the present study is that serum albumin was not measured. About 95% of total serotonin content in the human body is present in the gut, primarily in enterochromaffin cells of the intestinal mucosa30. Enterochromaffin cells act as local sensory transducers that respond to various chemical and mechanical stimuli in the gut31. Serotonin is synthesized from tryptophan through hydroxylation by tryptophan hydroxylase (TPH) followed by decarboxylation by aromatic acid decarboxylase (AADC). Serotonin mediates gastrointestinal functions that include peristalsis, secretion, vasodilation, and perception of pain or nausea31. Serotonin and one if its receptors, 5-HT3, are implicated as modulators of gastrointestinal inflammation in inflammatory bowel disease31, and the association with increased gut permeability suggests it may play a role in EED. Serotonin is removed from the intestinal mucosa by the serotonin-selective reuptake transporter (SERT). The underlying cause of elevated serum serotonin in children with EED is unclear. Possible biological mechanisms include upregulation of enzymes involved in serotonin synthesis, TPH and AADC, or a decrease in SERT expression during EED.
Serum glutamate concentrations were elevated in children with increased gut permeability. Glutamate, a non-essential amino acid, plays many important roles, such as providing a link between the tricarboxylic acid cycle and urea cycle in energy generation, serving in hepatic amino acid catabolism, being a precursor to glutathione, and acting as the primary fast excitatory neurotransmitter32,33. Glutamate, a main constituent of dietary protein, is a major oxidative fuel for the small intestine and is almost completely metabolized by the gut on first pass, presumably in the entrocytes32. Glutamate is also used for protein synthesis in the intestinal mucosa and can be used by enterocytes to produce aspartate, alanine, proline, ornithine, and citrulline34. Whether elevated serum glutamate arises from decreased utilization of glutamate by the gut, increased catabolism of amino acids such as arginine, ornithine, proline, histidine and glutamine, or other mechanisms in children with increased gut permeability is not known.
Taurine is an abundant free amino acid that plays diverse biological roles in membrane stabilization, cell volume regulation, mitochondrial function, growth and development, lung function, the conjugation of bile acids, and as an antioxidant35,36. Taurine is synthesized via the cysteine sulfinic acid pathway. The biological mechanisms that contribute to elevated serum taurine concentrations in children with EED are unclear.
Serum citrulline is considered to reflect total small bowel enterocyte mass since circulating citrulline is mainly produced by enterocytes37,38. Low circulating citrulline occurs in patients with short bowel syndrome, villous atrophy states, and Crohn’s disease38. Thus, serum citrulline has been proposed as a potential marker for EED2. Citrulline is synthesized in enterocytes via the conversion of glutamine or arginine to ornithine39. Citrulline is then produced from the conversion of ornithine by ornithine transcarbamylase38. The association of both low serum citrulline and ornithine with increased gut permeability may reflect reduced small bowel enterocyte mass.
The present study suggests several serum metabolites as candidate biomarkers for EED: citrulline, ornithine, glutamate, taurine, serotonin, tryptophan, serotonin/tryptophan ratio, kynurenine/tryptophan ratio, and phosphatidylcholines and sphingomyelins. However, the correlations between specific serum metabolites and gut permeability were modest, which may limit their use as biomarkers. Other blood biomarkers have been proposed as marker for gut permeability, including zonulin, anti-endotoxin core antibody, and soluble CD142. Further studies are needed to determine the utility of selected serum metabolites as biomarkers of EED, including consideration of the specificity of the markers.
The present study is limited in that the findings cannot necessarily be generalized to all settings where children are at risk for EED. There may be differences in diet and environment that could potentially modify the relationship between increased gut permeability and circulating metabolites. Urinary lactulose and mannitol in the present study were measured using high performance liquid chromatography (HPLC). A recent analysis suggests that LC-MS/MS platforms using multiple reaction monitoring for measurement of urinary lactulose and mannitol is more accurate, although measurements using either HPLC or LC-MS/MS are highly correlated40. Another limitation is that other markers for EED were not measured in this study.
In conclusion, increased gut permeability is associated with alterations in important metabolites involving growth and differentiation and gut function and integrity. Further studies are needed to elucidate the relationship between metabolic alterations in EED and specific biological mechanisms, including insufficient dietary choline and tryptophan, the effects of increased gut permeability on metabolic pathways involving glutamine, serotonin, and taurine, and to corroborate the relationship of gut permeability with serum citrulline, ornithine, and serotonin/tryptophan and kynurenine/tryptophan ratios.
Methods
Study design and participants
We examined the relationship of EED with serum metabolites in a cross-sectional study of 315 children, aged 12–59 months, seen in six villages (Masika, Makhwira, Mitondo, Mibiza, Chamba, and Mayaka) in rural southern Malawi in 2011. Children were eligible for the study if they had no evidence of kwashiorkor, congenital or chronic disease, caretaker-reported diarrhea, or were under treatment for malnutrition. Children had anthropometry conducted by experienced field workers. Weight was measured to the nearest 5 g using a digital scale (Seca 344, Chino, CA). Height was measured to the nearest 0.1 cm using a rigid height board (Seca 417). The dual sugar permeability test was used as the standard non-invasive measure of gut integrity, as described in detail elsewhere3. The urinary L:M ratio was used as the measure of intestinal mucosal permeability8. Lactulose and mannitol were measured using high-performance liquid chromatography3. Chichewa-speaking Malawian research nurses obtained written and oral informed consent from each child’s caretaker before enrollment in the study. Community consent for the study also was obtained from the village chief and local health officials. The protocol for this study was approved by the College of Medicine Research Ethics Committee of the University of Malawi, the Human Research Protection Office of Washington University in St. Louis, and the Johns Hopkins School of Medicine Institutional Review Board. The methods were carried out in accordance with the approved guidelines.
Measurement of serum metabolites
Venous blood was drawn by study nurses and doctors. Serum samples were processed, aliquoted, and snap frozen in liquid nitrogen in cryovials within 4 h of blood drawing. Cryovials were transferred to storage at −80 °C. Serum metabolites were measured using liquid chromatography tandem mass spectrometry (LC-MS/MS). Metabolites were extracted and concentrations are measured using the AbsoluteIDQ p180 kit (Biocrates Life Sciences AG, Innsbruck, Austria) following the manufacturers protocol for a 5500 QTrap (Sciex, Framingham, MA) mass spectrometer equipped with an electrospray ionization source, a Shimadzu CBM-20A command module, LC-20AB pump, and a Shimadzu SIL-20AC-HT autosampler, a CTO-10Ac column oven heater, and running with Analyst 1.5.2 software, as described in detail elsewhere10. Briefly, 10 μL of serum was pipetted onto a 96 well Biocrates kit. The samples were dried at room temperature (RT) for 30 min. 50 μL of 5% PITC reagent was added and incubated for 20 min and the plate was dried under nitrogen for 1 h. 300 μL of 5 mM ammonium acetate in methanol was added and incubated at RT on a shaker (450 rpm) for 30 min. The plate was centrifuged at 500 × g for 2 min and labeled; 50 μL of each sample was transferred to a 96 deep well LC plate, and 10 uL of each sample was transferred to the 96 deep well FIA plate. To the LC plate, 450 μl of 40% methanol (in HPLC grade water) was added. To the FIA plate, 490 μL of FIA running solvent was added. 10 μL was injected onto the Eclipse XDB C18, 3.5 μm, 3.0 × 100 mm with a Phenomenex C18 Security Guard Cartridge, 3.0 mm ID. The mobile phase consisted of solvent A (water containing 0.2% formic acid) and solvent B (acetonitrile containing 0.2% formic acid), with the following gradient: 0–0.5 min 0% B, 5.5 min: 95% B; 6.5 min: 95% B; 7.0 min: 0% B; 9.5 min: 0% B. LC plate evaluation of the samples was carried out using the MetIDQ software. The FIA plate was run with 20 μL injection directly onto the MS at a flow of 30 μL/min with water/acetonitrile (1:1) containing 0.2% formic acid as the mobile phase, with the following flow rate program: 0–1.6 min: 30 μL/min; 2.4 min: 200 μL/min; 2.80 min: 200 μL/min and 3.00 min: 30 μL/min. Concentrations were calculated using the Analyst/MetIDQ software and reported in μmol/L. The method measured 140 metabolites, including 22 amino acids, 3 biogenic amines, 6 amino acid metabolites, 15 sphingolipids, 8 acylcarnitines, and 85 glycerophospholipids (lyso-, diacyl-, and acyl-alkyl phosphatidylcholines). Glycerophospholipids are differentiated on the basis of ester and ether bonds in the glycerol moiety. Diacyl or “aa” indicates that fatty acids are bound with ester bonds at the sn-1 and sn-2 positions on the glycerol backbone. Acyl-alkyl or “ae” indicates that the fatty acid at the sn-1 position is bound with an ether bond. The total number of carbon atoms and double bonds in fatty acid chains is represented by “C x:y”, where x denotes the number of carbons and y denotes the number of double bonds. Phosphatidylcholine (PC), lysophosphatidylcholine (lysoPC), and sphingomyelin (SM), and hydroxysphingomyelin (SM [OH]) are used as abbreviations. Standard amino acid abbreviations are used in all figures. Acylcarnitines were designated with abbreviations: carnitine (C0), acetylcarnitine (C2), proprionylcarnitine (C3), butyrylcarnitine (C4), hydroxybutyrylcarnitine (C4-OH [C3-DC]), hexadecanoylcarnitine (C16), octadecenoylcarnitine (C18), and octadecadienylcarnitine (C18:1). Dimethylarginines were designated with abbreviations: total dimethylarginine (TDMA), symmetric dimethylarginine (SDMA), and asymmetric dimethylarginine (ADMA). The MS spectra were evaluated using Analyst/MetIDQ (Biocrates) software. Human serum samples spiked with standard metabolites were used to monitor the reproducibility of the assay. Serum metabolites that were below the limit of quantification were excluded from the data analyses. The inter-assay and intra-assay coefficients of variation ranged from 5% to 15% for nearly all analytes.
Statistical analysis
Exploratory data analyses using histograms and boxplots were used to examine the distribution of the L:M ratio and serum metabolites. The L:M ratio is a measure of gut permeability. Spearman correlations of each metabolite with L:M ratio were estimated with and without adjustment for age, gender, and village. A false discovery rate approach was used to correct for multiple testing; the bootstrap method was used to compute q-values41,42,43. This analysis was carried out using the pcor.test() function in the ppcor package in R software version 3.2.0. Classification modeling using nine machine-learning algorithms was used to examine the ability of the 14 metabolites that were significantly correlated with L:M ratio to classify EED status (L:M ≥ 0.15 versus L:M < 0.15). SuperLearner11, an ensembling approach to machine learning, implemented multiple individual machine-learning algorithms (e.g., logistic regression). The cross-validated results from the individual machine-learning algorithms were combined via weighted average to minimize a cross-validated mean squared error (and potentially fit the data better than each individual algorithm). Details of SuperLearner, including computation of estimates, cross-validation, and metrics to assess discrimination and calibration, are described in detail in the Supplementary Appendix. Classification modeling was carried out using the SuperLearner() and CV.SuperLearner() functions in the SuperLearner package in R software version 3.2.044.
Additional Information
How to cite this article: Semba, R. D. et al. Metabolic alterations in children with environmental enteric dysfunction. Sci. Rep. 6, 28009; doi: 10.1038/srep28009 (2016).
References
Keusch, G. T. et al. Environmental enteric dysfunction: pathogenesis, diagnosis, and clinical consequences. Clin. Infect. Dis. 59 (suppl 4), S207–212 (2014).
Crane, R. J., Jones, K. D. J. & Berkley, J. A. Environmental enteric dysfunction: an overview. Food Nutr. Bull. 36, S76–87 (2015).
Weisz, A. J. et al. Abnormal gut integrity is associated with reduced linear growth in rural Malawian children. J. Pediatr. Gastroenterol. Nutr. 55, 747–750 (2012).
Lunn, P. G., Northrop-Clewes, C. A. & Downes, R. M. Intestinal permeability, mucosal injury, and growth faltering in Gambian infants. Lancet 338, 907–910 (1991).
Lunn, P. G., Northrop-Clewes, C. A. & Downes, R. M. Recent developments in the nutritional management of diarrhoea. 2. Chronic diarrhoea and malnutrition in The Gambia: studies on intestinal permeability. Trans. R. Soc. Trop. Med. Hyg. 85, 8–11 (1991).
George, C. M. et al. Fecal markers of environmental enteropathy are associated with animal exposure and caregiver hygiene in Bangladesh. Am. J. Trop. Med. Hyg. 93, 269–275 (2015).
Campbell, D. I. et al. Chronic T cell-mediated enteropathy in rural west African children: relationship with nutritional status and small bowel function. Pediatr. Res. 54, 306–311 (2003).
Denno, D. M. et al. Use of the lactulose to mannitol ratio to evaluate childhood environmental enteric dysfunction: a systematic review. Clin. Infect. Dis. 59 (suppl 4), S213–219 (2014).
Lord, R. S. & Bralley, J. A. (eds.) Laboratory Evaluations for Integrative and Functional Medicine. Metametrix Institute, Duluth, Georgia (2008).
Schmerler, D. et al. Targeted metabolomics for discrimination of systemic inflammatory disorders in critically ill patients. J. Lipid Res. 53, 1369–1375 (2012).
van der Laan, M. J., Polley, E. C. & Hubbard, A. E. Super learner. Stat. Appl. Genet. Mol. Biol. 6, Article 25 (2007).
Friedman, J., Hastie, T. & Tibshirani, R. Regularization paths for generalized linear models via coordinate descent. J. Stat. Software 33, 1–22 (2010).
Friedman, J. H. Greedy function approximation: a gradient boosting machine. Ann. Stat. 29, 1189–1232 (2001).
Breiman, L. Random forests. Machine Learning 45, 5–32 (2001).
Milborrow, S. Earth: Multivariate Adaptive Regression Spline Models, Package Version 4.4-3. Vienna, Austria: R Foundation for Statistical Computing (2015).
Friedman, J. H. Multivariate adaptive regression splines (with discussion) Ann. Stat. 19, 1–141 (1991).
Fagone, P. & Jackowski, S. Phosphatidylcholine and the CDP-choline cycle. Biochim. Biophys. Acta 1831, 523–532 (2013).
Cole, L. K., Vance, J. E. & Vance, D. E. Phosphatidylcholine biosynthesis and lipoprotein metabolism. Biochim. Biophys. Acta 1821, 754–761 (2012).
Goss, V., Hunt, A. N. & Postle, A. D. Regulation of lung surfactant phospholipid synthesis and metabolism. Biochim. Biophys. Acta 1831, 448–458 (2013).
Stremmel, W. et al. Mucosal protection by phosphatidylcholine. Dig. Dis. 30 (suppl 3), 85–91 (2012).
Food and Nutrition Board, Institute of Medicine. Dietary Reference Intakes: Thiamin, Riboflavin, Niacin, Vitamin B-6, Vitamin B-12, Panthothenic Acid, Biotin, and Choline. National Academy of Sciences. Washington, D.C. Food and Nutrition Board, Institute of Medicine National Academic Press, 390–422 (1998).
Grzelczyk, A. & Gendaszewska-Darmach, E. Novel bioactive glycerol-based lysophospholipids: new data–new insight into their function. Biochimie 95, 667–679 (2013).
Slotte, J. P. Biological functions of sphingomyelins. Prog. Lipid Res. 2013, 52, 424–437 (2013).
Slotte, J. P. Molecular properties of various structurally defined sphingomyelins—correlation of structure with function. Prog. Lipid Res. 52, 206–219 (2013).
Beyersdorf, N. & Müller, N. Sphingomyelin breakdown in T cells: role in activation, effector functions and immunoregulation. Biol. Chem. 396, 749–758 (2015).
Kinney, H. C. et al. Myelination in the developing human brain: biochemical correlates. Neurochem. Res. 19, 983–996 (1994).
Le Floc’h, N., Otten. W. & Merlot, E. Tryptophan metabolism, from nutrition to potential therapeutic applications. Amino Acids 41, 1195–1205 (2011).
Imbesi, R., Mazzone, V. & Castrogiovanni, P. Is tryptophan ‘more’ essential than other essential aminoacids in development? A morphologic study. Anat. Histol. Embryol. 38, 361–369 (2009).
El-Merahbi, R., Löffler, M., Mayer, A. & Sumara, G. The roles of peripheral serotonin in metabolic homeostasis. FEBS Lett. 589, 1728–1734 (2015).
Keszthelyi, D., Troost, F. J. & Masclee, A. A. M. Understanding the role of tryptophan and serotonin metabolism in gastrointestinal function. Neurogastroenterol. Motil. 21, 1239–1249 (2009).
Mawe, G. M. & Hoffman, J. M. Serotonin signalling in the gut–functions, dysfunctions and therapeutic targets. Nat. Rev. Gastroenterol. Hepatol. 10, 473–486 (2013).
Burrin, D. G. & Stoll, B. Metabolic fate and function of dietary glutamate in the gut. Am. J. Clin. Nutr. 90 (suppl), 850S–856S (2009).
Brosnan, M.E. & Brosnan, J. T. Hepatic glutamate metabolism: a tale of 2 hepatocytes. Am. J. Clin. Nutr. 90 (suppl), 857S–861S (2009).
Blachier, F., Boutry, C., Bos, C. & Tomé, D. Metabolism and functions of L-glutamate in the epithelial cells of the small and large intestines. Am. J. Clin. Nutr. 90 (suppl), 814S–821S (2009).
De Luca, A., Pierno, S. & Camerino, D. C. Taurine: the appeal of a safe amino acid for skeletal muscle disorders. J. Transl. Med. 13, 243 (2015).
Lambert, I. H., Kristensen, D. M., Holm, J.B. & Mortensen, O. H. Physiological role of taurine–from organism to organelle. Acta. Physiol. (Oxf.) 213, 191–212 (2015).
Jianfeng, G. et al. Serum citrulline is a simple quantitative marker for small intestinal enterocytes mass and absorption function in short bowel patients. J. Surg. Res. 127, 177–182 (2005).
Crenn, P., Messing, B. & Cynober, L. Citrulline as a biomarker of intestinal failure due to enterocyte mass reduction. Clin. Nutr. 27, 328–339 (2008).
Marini, J. C. Arginine and ornithine are the main precursors for citrulline synthesis in mice. J. Nutr. 142, 572–580 (2012).
Lee, G. O. et al. Lactulose: mannitol diagnostic test by HPLC and LC-MSMS platforms: considerations for field studies of intestinal barrier function and environmental enteropathy. J. Pediatr. Gastroenterol. Nutr. 59, 544–550 (2014).
Storey, J. D. A direct approach to false discovery rates. J. Roy. Stat. Soc. Ser. B. 64, 479–498 (2002).
Storey, J. D., Taylor, J. E. & Siegmund, D. Strong control, conservative point estimation, and simultaneous conservative consistency of false discovery rates: a unified approach. J. Roy. Stat. Soc. Ser. B. 66, 187–205 (2004).
Rachakonda, V. et al. Serum metabolomic profiling in acute alcoholic hepatitis identifies multiple dysregulated pathways. PLoS One 9, e113860 (2014).
Polley, E. C. & van der Laan, M. J. SuperLearner: Super Learner Prediction, Package Version 2.0-15. Vienna, Austria: R Foundation for Statistical Computing [https://cran.r-project.org/web/packages/SuperLearner/SuperLearner.pdf] (Accessed November 25, 2015) (2015).
Acknowledgements
This work was supported by the Children’s Discovery Institute, the Hickey Family Foundation, the National Institutes of Health R01 AG27012 and the Intramural Branch of the National Institute on Aging, Baltimore, Maryland.
Author information
Authors and Affiliations
Contributions
R.D.S., M.S., I.T., M.I.O., K.K., L.F. and M.J.M. were responsible for the study design, statistical analyses, data interpretation, and writing the manuscript. I.T., K.M.M. and M.J.M. implemented the study in rural Malawi and collected samples and data. R.M., M.K. and M.I.O. contributed towards laboratory analyses of serum metabolites and urinary sugars. All authors contributed to writing and approving the final manuscript.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing financial interests.
Supplementary information
Rights and permissions
This work is licensed under a Creative Commons Attribution 4.0 International License. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in the credit line; if the material is not included under the Creative Commons license, users will need to obtain permission from the license holder to reproduce the material. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/
About this article
Cite this article
Semba, R., Shardell, M., Trehan, I. et al. Metabolic alterations in children with environmental enteric dysfunction. Sci Rep 6, 28009 (2016). https://doi.org/10.1038/srep28009
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/srep28009
- Springer Nature Limited
This article is cited by
-
Factors associated with stunting: gut inflammation and child and maternal-related contributors among under-five children in Hawassa City, Sidama Region, Ethiopia
BMC Nutrition (2023)
-
Paracellular permeability and tight junction regulation in gut health and disease
Nature Reviews Gastroenterology & Hepatology (2023)
-
Citrulline and kynurenine to tryptophan ratio: potential EED (environmental enteric dysfunction) biomarkers in acute watery diarrhea among children in Bangladesh
Scientific Reports (2023)
-
Study on follicular fluid metabolomics components at different ages based on lipid metabolism
Reproductive Biology and Endocrinology (2020)
-
Lactoferrin and lysozyme to reduce environmental enteric dysfunction and stunting in Malawian children: study protocol for a randomized controlled trial
Trials (2017)