The intestinal microbiome of infants with cow's milk-induced FPIES is enriched in taxa and genes of enterobacteria

Received: 20 May 2024 | Accepted: 7 August 2024 DOI: 10.1002/jpn3.12356 ORIGINAL ARTICLE Gastroenterology The intestinal microbiome of infants with cow's milk‐induced FPIES is enriched in taxa and genes of enterobacteria Ana M. Castro 1 |Carlos Sabater 1 |Sandra Navarro 2 |Silvia Rodriguez 3 | Cristina Molinos 4 |Santiago Jiménez 5 |Angela Claver 6 |Beatriz Espin 7 | Gloria Domínguez 8 |Cristóbal Coronel 9 |Paula Toyos 5 | Isabel Gutiérrez‐Díaz 1 |Lydia Sariego 1 |Porfirio Fernández 5 |David Perez 3 | Abelardo Margolles 1 |Juan J. Díaz 5 |Susana Delgado 1 1 MicroHealth Group, Instituto de Productos Lácteos de Asturias‐Consejo Superior de Investigaciones Científicas (IPLA‐CSIC)/Instituto Biosanitario del Principado de Asturias (ISPA), Villaviciosa, Asturias, Spain 2 Primary Care Center Teatinos‐Corredoria, Oviedo, Asturias, Spain 3 Paediatrics Service, Hospital Universitario de San Agustín, Avilés, Asturias, Spain 4 Paediatrics Department, Hospital Universitario de Cabueñes, Gijón, Asturias, Spain 5 Paediatric Group, ISPA, Oviedo, Asturias, Spain 6 Allergology, Hospital Universitario Dexeus, Barcelona, Spain 7 Paediatric Gastroenterology Unit, Hospital Universitario Virgen del Rocío de Sevilla, Sevilla, Spain 8 Gastroenterology and Nutrition Section, Hospital Infantil Universitario Niño Jesús, Madrid, Spain 9 Primary Care Center Amante Laffón, Sevilla, Spain Correspondence Juan J. Díaz, Hospital Universitario Central de Asturias, Avda. de Roma s/n, 33011 Oviedo, Asturias, Spain. Email: [email protected] Funding information “Sociedad Española de Gastroenterología Hepatología y Nutrición Pediátrica,”edition 2020; MCIN/AEI/10.13039/501100011033, grant FJC2021‐047052‐I; Instituto de Investigación Sanitaria del Principado de Asturias and Fundació Banc Sabadell; The Spanish Ministry of Science and Innovation through the project MICROALERGYMILK (PID2019‐104546RB‐I00/AEI/10.13039/) and FJC2019‐042125‐I; CSIC's Global Health Platform (PTI Salud Global). Abstract Objectives: Food protein‐induced enterocolitis syndrome (FPIES) is a severe type of non‐IgE (immunoglobulin E)‐mediated (NIM) food allergy, with cow's milk (CM) being the most common offending food. The relationship between the gut microbiota and its metabolites with the inflammatory process in infants with CM FPIES is unknown, although evidence suggests a microbial dysbiosis in NIM patients. This study was performed to contribute to the knowledge of the interaction between the gut microbiota and its derived metabolites with the local immune system in feces of infants with CM FPIES at diagnosis. Methods: Twelve infants with CM FPIES and a matched healthy control group were recruited and the gut microbiota was investigated by 16S amplicon and shotgun sequencing. Fatty acids (FAs) were measured by gas chromatography, while immune factors were determined by enzyme‐linked immunosorbent assay and Luminex technology. Results: A specific pattern of microbiota in the gut of CM FPIES patients was found, characterized by a high abundance of enterobacteria. Also, an intense excretion of FAs in the feces of these infants was observed. Furthermore, correlations were found between fecal bifidobacteria and immune factors. J Pediatr Gastroenterol Nutr. 2024;79:841–849. wileyonlinelibrary.com/journal/jpn3 | 841 This is an open access article under the terms of the Creative Commons Attribution‐NonCommercial‐NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non‐commercial and no modifications or adaptations are made. © 2024 The Author(s). Journal of Pediatric Gastroenterology and Nutrition published by Wiley Periodicals LLC on behalf of European Society for Pediatric Gastroenterology, Hepatology, and Nutrition and North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition. Conclusion: These fecal determinations may be useful to gain insight into the pathophysiology of this syndrome and should be taken in consideration for future studies of FPIES patients. KEYWORDS dysbiosis, food hypersensitivity, food protein‐induced enterocolitis syndrome, SCFAs 1|INTRODUCTION Food protein‐induced enterocolitis syndrome (FPIES) is a rare form of non‐IgE (immunoglobulin E)‐mediated (NIM) food allergy affecting less than 1% of infants and children. 1 In FPIES, the ingestion of a triggering food elicits a reaction that typically appears from 1 to 4 h after that intake. There are geographical differences regarding the offending food, but cow's milk (CM) is the main food responsible for FPIES throughout the world. 2 CM FPIES may have two different clinical presentations: acute and chronic. Although FPIES is a NIM form of CM protein allergy (CMPA), sometimes positive specific IgE to CM proteins could be detected at diagnosis or at follow‐up (atypical FPIES). Diagnosis is established based on clinical criteria, 3 and only in doubtful cases, an oral food challenge might be performed in a hospital setting, as severe reactions might appear, even with small doses of CM. The pathophysiology of the syndrome is not fully understood, but involvement of the immune and neuroendocrine systems, has been hypothesized. 4 Several inflammatory mediators have been studied in peripheral blood and in feces of children with FPIES, 5,6 but no diagnostic marker has been discovered to date. Besides, it is well known that the interaction between gut microbiota and the immune system is a key factor in allergy and tolerance development. 7 Data on the relationship between gut microbiota and FPIES are scarce. 8 The objective of this study was to analyze the gut microbiome and its relationship with intestinal immune mediators in infants with CM FPIES. 2|METHODS 2.1 |Subject recruitment and sample collection The study sample included a group of 12 patients with CM FPIES recruited at the time of diagnosis at different health centers in Spain (Table S1). International consensus guidelines for the diagnosis and management of FPIES were used for patient inclusion. 3 Infants with acute FPIES presented with vomiting What is Known •Food protein‐induced enterocolitis syndrome (FPIES) is a severe type of non‐immunoglobulin E‐mediated food allergy, with cow's milk (CM) being the most common offending food. •FPIES pathology in not fully understood, but gut dysbiosis has been considered a possible pathogenic factor involved. What is New •Infants with CM FPIES patients show aspecific pattern ofgut microbiota characterised by a high abundance of enterobacteria. •Moreover, these infants show an intense excretion of fatty acids in the stools. 842 | CASTRO ET AL. 15364801, 2024, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jpn3.12356 by Readcube (Labtiva Inc.), Wiley Online Library on [10/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License in the first 4 h after the ingestion of CM, without associated skin or respiratory symptoms, and with at least three of the minor criteria indicated in the guidelines. Those with chronic FPIES presented with different gastrointestinal symptoms that improved or resolved with the elimination of CM proteins from the diet, and presented an acute FPIES reaction when the CM was reintroduced. A control group of fourteen 5‐month‐old infants was recruited by pediatricians in primary care centers. To be eligible for inclusion, participants should not have taken any antibiotics or probiotics in the 2 weeks before the start of the study. Clinical data were collected from the patient's medical history anonymously by the treating pediatrician. Fresh stool samples were collected at the time of diagnosis, in a special container (GutAlive ® ; MicroViable Therapeutics) that allows the transport of the samples in anaerobic conditions at room temperature. 9 Samples were processed in a biosafety cabinet within the first 24 h after deposition. The study was conducted in accordance with applicable local and legal regulations and following internationally accepted ethical standards. In addition, signed individual informed consent were obtained from all families participating in the study. The study was approved by the “Comité de Ética de la Investigación del Principado de Asturias”(Reference number 343/19). 2.2 |Intestinal microbiota analysis Fecal samples (1 g) were homogenized in 9 mL of phosphate‐buffered saline solution (PBS) and used for deoxyribonucleic acid (DNA) extraction according to the DNA extraction protocol Q 10 using the QIAamp DNA Stool Mini Kit (Qiagen) with some modifications. These modifications were mainly in the lysis steps using a Fastprep FP24 homogenizer (Biomedicals). In this case, the cycles lasted 45 s, leaving the samples on ice for 5 min between each treatment. Quantification of the extracted DNA was performed using the Qubit dsDNA BR Assay Kit (Thermo Fisher Scientific). 2.3 |Analysis of 16S ribosomal RNA gene amplicons sequencing The 16S recombinant DNA (rDNA) was amplified from the DNA of the samples according to Milani et al. 11 and 250 bp paired‐end sequences were obtained using an Illumina MiSeq System (Illumina) at the spin‐off of the University of Parma Genprobio srl. Sequences were processed using the Quantitative Insights Into Microbial Ecology software suite and assigned to the lowest possible taxonomic rank considered, using the SILVA database v. 132 as reference. The raw sequences data were deposited in the Sequence Read Archive (SRA) of the National Center for Biotechnology Information (NCBI) (https://www.ncbi.nlm.nih.gov/sra) under bioproject code PRJNA1036152 (sequence library identifer SAMN38115097–SAMN38115122). 2.4 |Analysis of shotgun metagenomics sequencing Total DNA extracted was submitted to an external sequencing service (www.BaseClear.com). Paired‐end sequence reads (2 × 150 bp) showing an average minimum of 25 million reads per sample were generated using an Illumina NovaSeq system. Contaminant low‐quality reads were removed using Kneaddata (v0.7.4) and Trimmomatic (v0.39) software. An assembly‐free analysis of decontaminated reads was carried out following MetaPhlAn 3.0 (v3.0.4) and HUMAnN 3.0 (v3.0.0) pipelines to perform taxonomic and functional analysis, respectively. 12,13 Then, several statistical methods implemented on R (v.4.2.3) were computed to analyze the results. Composition barplots were calculated and generated using “Phyloseq”and “Microbiome”R packages. 14,15 To investigate statistically significant differences (p< 0.05 and corrected p values considering a false discovery rate [FDR] of 0.25 [q< 0.25]) between controls and FPIES patients at all taxonomic levels, MaAsLin2 16 and several microbiota‐ specific statistical methods (ANCOM, LEfSe, and DESeq2) implemented in “microbiomeMarker”R package were computed. 17–20 The raw sequences data were deposited in the SRA of the NCBI under bioproject code PRJNA1037775 (sequence library identifer SAMN38197428–SAMN38197447). 2.5 |Intestinal inflamation markers determination Fecal calprotectin (FC) was analyzed using the CALPROLAB™Kit (Calpro) following the instructions for making an enzyme‐linked immunosorbent assay. The final absorbance (450 nm) data were measured in a Modulus Microplate Photometer (Turner BioSystems). The profile of cytokines, chemokines, and growth factors excreted in the fecal water was analyzed using the Bio‐Plex Pro Human Cytokine 27‐plex Assay Kit (Bio‐Rad Laboratories Inc.) in a Bio‐plex 200 system instrument (Bio‐Rad). For that, 1 g of fecal samples was suspended in 9 mL of PBS. After homogenization, the samples were centrifuged (13,000 rpm, 15 min, 4°C) and the supernatants were collected and stored frozen (−20°C) until analysis. Determinations in fecal samples were performed in duplicate according to the manufacturer's protocol and calibration curves (five parameter logistic) were constructed for each analyte using duplicate values CASTRO ET AL. | 843 15364801, 2024, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jpn3.12356 by Readcube (Labtiva Inc.), Wiley Online Library on [10/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License for each known concentration and the Bio‐Plex Manager 6.2 software (Bio‐Rad). 2.6 |Quantification of fatty acids (FAs) in feces For the determination of FAs, a 1:2 dilution of feces (1 g) in PBS was used. 0.1 mL of this dilution was supplemented with 50 μLof2‐ethyl butyric acid (Sigma‐Aldrich), as internal standard (1.05 mg/mL in methanol) and acidified with 50 μL of 20% formic acid (vol/vol). The acidic solution was then extracted with 450 µL of methanol and centrifuged for 10 min at 16,200g. The supernatants were stored at −20°C until analysis in a gas chromatography (GC) apparatus composed of a 6890 GC injection module (Agilent Technologies) with a HP‐FFAP (30 m × 0.250 mm × 0.25 µm) column (Agilent Technologies). The chromatographic system was equipped with a flame ionization detector. Data acquisition and processing were performed using ChemStation Agilent software (Agilent Technologies). All samples were analyzed in duplicate and FAs were quantified as previously described. 21 2.7 |Statistical analysis Statistical analyses were performed using IBM SPSS Statistics v.28.0.1 (IBM) and RStudio software version 4.3.0 (R Foundation for Statistical Computing). Normality was checked using the Kolmogorov–Smirnov test. The nonparametric Mann–Whitney Utest was used to examine differences in continuous variables of general characteristic and p< 0.05 were considered significant. A statistical significance of p<0.05 was also considered for both 16S and metagenomic sequencing analyses. In addition, statistical results from shotgun metagenomic analysis were adjusted considering a FDR of 0.25 (q< 0.25). α‐Diversity estimators and general statistical analysis of microbiota composition were calculated using the Microbiome R package. 15 Additionally, β‐diversity analysis of microbial communities was conducted in accordance with the Bray–Curtis dissimilarity method, 22 implemented in “Phyloseq”R package, 14 with the adonis2 function (vegan), employing permutational multivariate analysis of variance (PERMANOVA). The Origin Pro 2021 software (OriginLab) was also used for graph design, and GraphPad Prism 9 (GraphPad Software Inc.) to calculate the curve of 4‐point sigmoidal fit used in the determination of FC. Statistical correlations (p< 0.05) between microbial composition, FAs, and immunity factors were calculated using base R functions and expressed as Pearson correlation coefficients. 3|RESULTS 3.1 |Participant characteristics and clinical data Information about the age at diagnosis, gender, mode of delivery, and type of feeding before diagnosis was recorded (Table S1). None of the studied infants were born preterm (taking into account prematurity at less than 37 weeks' gestation). No significant differences were found between patients and controls with respect to gender, mode of delivery, and type of feeding. While the mean age of the patients was 4 ± 2 months, the mean age of the controls was 5 ± 1 months (p= 0.04). 3.2 |Microbiome First, we sequenced amplicons of 16S rDNA from fecal samples of the 12 FPIES recruited patients and the14age‐matched controls. Analysis of α‐diversity estimated by Shannon and Simpson indexes revealed no statistically significant differences (p>0.05) between patients and controls (data not shown), however β‐diversity analysis performed by the Bray– Curtis dissimilarity method showed statistically significant differences (p< 0.05) between both groups when PERMANOVA was applied (Figure S1). Additionally, analysis at the compositional level revealed that patients had a lower abundance of Actinomycetota members than controls (p= 0.05). Within this phylum, sequences assigned to the family Bifidobacteriaceae stood out, being less abundant in infants with FPIES (Table S2). On the other hand, representatives of the family Enterobacteriaceae were found in higher percentages in these patients compared to controls although the differences were not statistically significant (p= 0.08). To further investigate the observed differences in the gut microbiome at both taxonomic and functional levels, we performed shotgun metagenomics sequencing on samples from half of the infants with FPIES (six samples) and the control group (14 samples). The results were consistent with those observed by 16S sequencing and showed differences between both groups at phylum, family, and genus level. Actinomycetota (Figure 1A), Bifidobacteriaceae (Figure 1B)andBifidobacterium (Figure 1C) were significantly more abundant in the control group compared to FPIES patients (p<0.05 and q< 0.25). In contrast, Pseudomonadota (Figure 1A), Enterobacteriaceae (Figure 1B), Klebsiella, and Escherichia (Figure 1C) showed significantly higher (p< 0.05 and q< 0.25) abundances in patients compared to control infants. Regarding the functional analysis of metagenomes, gene count for the control group and FPIES 844 | CASTRO ET AL. 15364801, 2024, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jpn3.12356 by Readcube (Labtiva Inc.), Wiley Online Library on [10/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License patients was 17,829 ± 5775 and 17,086 ± 2547, respectively. The statistical analysis of functional profiles showed microbial gene families and metabolic pathways with highest abundances (Table 1)in metagenomes from the control group, corresponding with Bifidobacterium bifidum and Bifidobacterium longum species. On the other hand, patients with FPIES have a large number of gene families and metabolic pathways of Enterobacteriaceae (709 families and 21 pathways), specifically of the species Escherichia coli and Klebsiella pneumoniae,presenting abundances significantly higher than those of the control group (p<0.05 and q<0.25). These pathways comprise a wide variety of functions involved in carbohydrate, aminoacid and ribonucleotide metabolism (Table S3). Of note that among the carbohydrate degradation pathways those of D‐fructuronate and D‐galactarate were predominant in FPIES infants, whereas those of sucrose degradation were more abundant in controls. 3.3 |Immunological and inflammatory markers Regarding the levels of FC, higher values were observed in patients: 292.88 µg/g (72.31–825.93) feces than in controls 92.83 µg/g (47.01–267.69) feces, although the differences were not statistically significant (p=0.06) (Figure S2). Of the 27 immune factors examined, including cytokines, chemokines, and growth factors excreted in fecal waters, only five mediators were detected in half of the subjects (n> 13) or in half of one of the two groups (n> 6 for patients and n> 7 for controls) (Table S4). These analytes included interleukins (ILs) 4 and 10, the IL‐1 receptor antagonist protein (IL‐1ra), the cytokine interferon gamma inducible protein‐10 (IP‐10), and the platelet‐ derived growth factor BB (PDGF‐bb). The levels of IL1‐ra, IP‐10, and PDGF‐bb which presented a significant reduction in the patient group compared to the healthy group (p= 0.01) are showed in Figure 2. For the other immune compounds quantified in feces there were no statistical differences between the infants of both groups. 3.4 |Fecal FAs The concentration of fecal FAs, and specifically acetic acid, was higher in patients (p< 0.001) (Figure 3A). It was also observed that the excretion of FAs derived from FIGURE 1 Taxonomic clades identified at phylum (A), family (B), and genus (C) level in the fecal microbiota of controls and patients. Data are expressed as abundance percentages (%). TABLE 1 Number of microbial gene families and metabolic pathways showing highest abundances in controls and patients (p< 0.05 and q< 0.25). Microbial gene families showing higher abundances in control group Microbial gene families showing higher abundances in patients Taxa Frequency Taxa Frequency Bifidobacterium bifidum 2447 E. coli 689 Bifidobacterium longum 255 Klebsiella pneumoniae 20 Escherichia coli 7B. longum 3 Eggerthella lenta 2Enterococcus faecalis 2 Total 2711 Total 714 Microbial metabolic pathways showing higher abundances in control group Microbial metabolic pathways showing higher abundances in patients Taxa Frequency Taxa Frequency B. bifidum 48 E. coli 11 E. lenta 7K. pneumoniae 10 E. coli 1 Total 21 Total 56 Note: These gene families and metabolic pathways are summarized by bacterial species. CASTRO ET AL. | 845 15364801, 2024, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jpn3.12356 by Readcube (Labtiva Inc.), Wiley Online Library on [10/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License protein degradation, branched chain FAs (BCFAs) that includes isovaleric and isobutyric acids, were significantly higher in patients compared to controls (p= 0.04) (Figure 3B). 3.5 |Association between metagenomes and fecal parameters Correlation analysis was performed to investigate possible associations between the abundance of gut microorganisms and the other fecal biochemical parameters measured. As shown in Figure S3,moderate positive associations were found between bifidobacteria and the levels of the factors IL‐1ra, IP‐10, and PDGF‐bb (Pearson correlation coefficient values of 0.5). In contrast, these immune compounds were negatively correlated with E. coli. In addition, correlation analysis revealed strong positive associations between Enterobacteriaceae and Klebsiella abundance and propionic acid (Pearson correlation coefficient values of 0.8), while Actinomycetota showed negative correlations with propionic and butyric acids (Figure S3). No significant associations were found for acetic acid levels. 4|DISCUSSION The incompletely understood pathophysiology of FPIES, the lack of diagnostic biomarkers, and the poor knowledge of its natural history have recently made this disease an important target for research. 23 There are reports suggesting that microbiota may be involved in the pathogenesis of this syndrome. 24 However, studies are really scarce and the gut microbiota of this type of NIM food allergy remains poorly characterized with few authors using omics techniques to explore it. 25 Recently, differences in gut microbiota composition in patients with pediatric FPIES respect to healthy control infants have been reported by sequencing of 16S rDNA amplicons 8 though the studied cases were not induced by CM proteins. In fact, to our knowledge, there are no previous research papers on shotgun sequencing of the gut microbiome incasesofCMFPIES.Inourstudy,weusedboth16S rDNA amplicons and in depth shotgun sequencing. By both techniques, a different microbiota structure between FIGURE 2 Relevant immunological markers in feces. Data are represented in a box‐and‐whisker plot. Mann–Whitney Utest were used to evaluate differences in concentrations among groups. Significant differences for *p< 0.05 and **p< 0.01. IL‐1ra, interleukin‐1 receptor antagonist protein; IP‐10, interferon gamma inducible protein‐10; PDGF‐bb, platelet‐derived growth factor BB. FIGURE 3 Fatty acids in fecal samples. Box‐and‐whisker plot depicting concentrations (in µg/g) of major short chain fatty acids (A) and branched chain fatty acids (B) in fecal samples of the two groups of study. Mann–Whitney Utest was used. Significance was set at p< 0.05. 846 | CASTRO ET AL. 15364801, 2024, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jpn3.12356 by Readcube (Labtiva Inc.), Wiley Online Library on [10/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License patients and controls was observed, with differences in the distribution of bacterial taxa. Mainly, it was noted a depletion in bifidobacteria in CM FPIES compared to controls, and an enrichment of enterobacteria, in particular the enteric pathogens Klebsiella and E. coli, in the gut microbiome of these patients. These are facultative anaerobic microorganisms belonging to the phylum of Pseudomonadota (earlier known as Proteobacteria) potentially harmful to the colonocytes. 26 Indeed, a dysbiotic expansion of this phylum in the gut has been proposed as a potential diagnostic microbial signature of epithelial dysfunction. 27 An increased abundance of Proteobacteria has been observed in humans with severe intestinal inflammation, including patients with inflammatory bowel disease, 28 necrotizing enterocolitis, 29 or even FPIES, which is consistent with our findings. 30 In addition, in our work the fecal excretion of FAs produced by the gut microbiota was quantified. Short‐chain fatty acids (SCFAs), mainly acetic, propionic, and butyric acid are produced as result of bacterial fermentation in the colon, and they have been associated with gut health and exert beneficial effects on the host at various levels. 31,32 Gut microbiota composition and fecal butyrate levels in children affected by NIM‐CMPA were evaluated by Berni Canani et al. 33 These authors observed that these children had a significantly lower fecal butyrate concentration compared to healthy controls. In our previous work with NIM‐CMPA cases, we did not find such differences for the main SCFAs between controls and patients, however, the BCFAs were significantly higher in patients. 34 In all these studies, the average age of the infants was higher and they were not FPIES patients. Apart from these, we have not found any other work on this type of food allergy in which the levels of butyrate or other SCFAs were measured directly in feces. In the present work, the excretion of total FAs and BCFAs was indeed increased. In particular, acetic, which is the most abundant SCFAs in the colon, 32 was 10 times higher in FPIES respect to controls. It has been described that colonocytes rapidly absorb an estimated 95%–99% of microbial produced FAs, while the remaining 5% are secreted in the feces. 31 We hypothesized that the amount detected in feces in patients could not represent that produced by the microbiota in the lumen, but rather a result of colonocyte dysfunction associated to the shift in the obligate to facultative anaerobic ratio observed in these infants with FPIES. 27 The high levels detected in feces of these FAs could be due to a reduced absorption and an alteration of colonocyte metabolism according to the “oxygen hypothesis.” 35 Future studies are needed to confirm and understand this phenomenon. Ideally, the determination of SCFAs in serum might give some clues about circulating levels and reduced absorption, although it is difficult to obtain this type of samples in young infants, almost all from controls. Another point that can give some consistence to our theory is the fact that we did not observe any association between fecal excretion of acetic acid (higher in patients) and abundance of bifidobacteria (higher in controls), which are the main bacterial producer of this compound in the infant gut. 32 Nevertheless, for propionic acid, we found a correlation with abundance of enterobacteria, for which a propanediol pathway for the conversion of deoxy‐sugars to propionate has been described, 32,36 and accordingly, metabolic genes related to fermentation to propionate (MetaCyc Pathway; PWY‐7013) were revealed from the metagenomic functional analysis to be associated to enterobacteria in some of these patients (data not shown). In our study, we combined metagenomic sequencing with the quantification of fecal immune factors looking for correlations, since the involvement of the immune system is poorly understood, and gastrointestinal immune events may be relevant in this syndrome where the gastrointestinal tract is the main system affected. In this sense, identification of new fecal biomarkers will be optimal to improve diagnosis and management of CM FPIES occurring in small babies, especially for the noninvasive collection requirements. 37 Among fecal biomarkers, the measurement of FC is considered useful for intestinal mucosal inflammation. Reference values are clearly defined in adults, however no clear cut‐offs have been convincingly established in children, 38 with the studied cases of NIM‐CMPA reporting no differences respect to healthy infants. 34,39 In our study of CM FPIES, no statistically significant differences were found, probably due to the dispersion of the data, but a tendency to a higher concentration was observed in patients with a mean value double than in controls and above 250 μg/g. For other cytokines and local immune mediators, we observed a significant reduction in the fecal concentrations in patients. In particular, PDGF‐bb, a platelet‐derived factor involved in vascular remodeling and angiogenesis, IP‐10, and the IL1‐ra. This suggests to us that their low levels in the feces of CM FPIES infants could be related to a disrupted colonic homeostasis. 5|CONCLUSION In support of our theory, a positive correlation between these immune factors and the levels of the beneficial bifidobacteria was observed in our study. It is clear that several cellular elements and cytokines may be involved, 23 although the exact molecular mechanism and interactions with commensal bacteria remain to be elucidated. By using different approaches, the experimental data obtained revealed potentially pathogenic taxa capable of damaging intestinal colonocytes and disrupting intestinal immune homeostasis. This should be taken into consideration for future studies, ideally with a high number of patients, to confirm the findings. We are conscious that the reduced sample size is a limitation, but this is the first study CASTRO ET AL. | 847 15364801, 2024, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jpn3.12356 by Readcube (Labtiva Inc.), Wiley Online Library on [10/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License exclusively in CM FPIES cases. Although diet is a factor that can influence the microbiota, and the FPIES patients in our work were on different types of feeding (breastfeeding, mixed, or bottle feeding), we did not find statistical differences in dietary habits with respect to the age‐matched control group of infants. However, it is true that for further studies with large cohorts of patients dietary differences in feeding should be taken into account. ACKNOWLEDGMENTS The authors would like to thank Ignacio Carbajal and Agueda Garcia for providing samples of healthy infants from the Primary Care Centers “La Eria”and “Vallobin‐La Florida”in Asturias. The technical support of Lydia Sariego is also acknowledged. Juan J. Díaz received a research grant from “Sociedad Española de Gastroenterología Hepatología y Nutrición Pediátrica,”edition 2020. Isabel Gutiérrez‐Díaz has a grant FJC2021‐047052‐Ifinanced by MCIN/AEI/10.13039/501100011033. Ana M. Castro acknowledges her predoctoral research contract funded by the Instituto de Investigación Sanitaria del Principado de Asturias and Fundació Banc Sabadell. The work was partially supported by The Spanish Ministry of Science and Innovation through the project MICROALERGYMILK (PID2019‐104546RB‐I00/AEI/10.13039/) and FJC2019‐ 042125‐I. Susana Delgado is part of the CSIC's Global Health Platform (PTI Salud Global). CONFLICT OF INTEREST STATEMENT Susana Delgado and Abelardo Margolles are Scientific Founder and Member of the Scientific Advisory Board of MicroViable Therapeutics S.L. The remaining authors declare no conflict of interest. DATA AVAILABILITY STATEMENT Sequencing files and metadata were deposited in the Sequence Read Archive of the National Center for Biotechnology Information under bioprojects codes PRJNA1036152 for amplicons sequences and PRJNA1037775 for shotgun sequences. ORCID Juan J. Díaz http://orcid.org/0000-0003-0962-8403 REFERENCES 1. Nowak‐Wegrzyn A, Berin MC, Mehr S. Food protein‐induced enterocolitis syndrome. J Allergy Clin Immunol Pract. 2020;8(1):24‐35. doi:10.1016/j.jaip.2019.08.020 2. Prattico C, Mulé P, Ben‐Shoshan M. A systematic review of food protein‐induced enterocolitis syndrome. Int Arch Allergy Immunol. 2023;184(6):567‐575. doi:10.1159/000529138 3. Nowak‐Węgrzyn A, Chehade M, Groetch ME, et al. International consensus guidelines for the diagnosis and management of food protein‐induced enterocolitis syndrome: executive summary‐ workgroup report of the adverse reactions to foods committee, American Academy of Allergy, Asthma & Immunology. JAllergy Clin Immunol. 2017;139(4):1111‐1126. doi:10.1016/j.jaci.2016. 12.966 4. Mathew M, Leeds S, Nowak‐Węgrzyn A. Recent update in food Protein‐Induced enterocolitis syndrome: pathophysiology, diagnosis, and management. Allergy Asthma Immunol Res. 2022;14(6):587‐603. doi:10.4168/aair.2022.14.6.587 5. Hamano S, Yamamoto A, Fukuhara D, Yan K. Serum thymus and activation‐regulated chemokine level as a potential biomarker for food protein‐induced enterocolitis syndrome. Pediatr Allergy Immunol. 2019;30(3):387‐389. doi:10.1111/ pai.13030 6. Wada T, Toma T, Muraoka M, Matsuda Y, Yachie A. Elevation of fecal eosinophil‐derived neurotoxin in infants with food protein‐induced enterocolitis syndrome. Pediatr Allergy Immunol. 2014;25(6):617‐619. doi:10.1111/pai.12254 7. Bunyavanich S, Berin MC. Food allergy and the microbiome: current understandings and future directions. J Allergy Clin Immunol. 2019;144(6):1468‐1477. doi:10.1016/j.jaci.2019. 10.019 8. Caparrós E, Cenit MC, Muriel J, et al. Intestinal microbiota is modified in pediatric food protein–induced enterocolitis syndrome. J Allergy Clin Immunol Glob. 2022;1(4):217‐224. doi:10. 1016/j.jacig.2022.07.005 9. Martínez N, Hidalgo‐Cantabrana C, Delgado S, Margolles A, Sánchez B. Filling the gap between collection, transport and storage of the human gut microbiota. Sci Rep. 2019;9:8327. doi:10.1038/s41598-019-44888-8 10. Costea PI, Zeller G, Sunagawa S, et al. Towards standards for human fecal sample processing in metagenomic studies. Nat Biotechnol. 2017;35:1069‐1076. doi:10.1038/nbt.3960 11. Milani C, Hevia A, Foroni E, et al. Assessing the fecal microbiota: an optimized ion torrent 16S rRNA gene‐based analysis protocol. PLoS One. 2013;8:e68739. doi:10.1371/journal.pone. 0068739 12. Franzosa EA, McIver LJ, Rahnavard G, et al. Species‐level functional profiling of metagenomes and metatranscriptomes. Nature Methods. 2018;15(11):962‐968. doi:10.1038/s41592018-0176-y 13. Truong DT, Tett A, Pasolli E, Huttenhower C, Segata N. Microbial strain‐level population structure and genetic diversity from metagenomes. Genome Res. 2017;27(4):626‐638. doi:10. 1101/gr.216242.116 14. McMurdie PJ, Holmes S. Phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data. PLoS One. 2013;8(4):e61217. doi:10.1371/journal.pone.0061217 15. Lahti L, Shetty S. Tools for microbiome analysis in R. Version 2.1.24. 2017. Accessed November 1, 2023. https://microbiome. github.io/tutorials/ 16. Mallick H, Rahnavard A, McIver LJ, et al. Multivariable association discovery in population‐scale meta‐omics studies. PLoS Comput Biol. 2021;17(11):e1009442. doi:10.1371/journal.pcbi. 1009442 17. Cao Y, Dong Q, Wang D, Zhang P, Liu Y, Niu C. MicrobiomeMarker: an R/Bioconductor package for microbiome marker identification and visualization. Bioinformatics. 2022; 38(16):4027‐4029. doi:10.1093/bioinformatics/btac438 18. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA‐seq data with DESeq2. Genome Biol. 2014;15:550. doi:10.1186/s13059-014-0550-8 19. Mandal S, Van Treuren W, White RA, et al. Analysis of composition of microbiomes: a novel method for studying microbial composition. Microb Ecol Health Dis. 2015;26(1):27663. doi:10. 3402/mehd.v26.27663 20. Segata N, Izard J, Waldron L, et al. Metagenomic biomarker discovery and explanation. Genome Biol. 2011;12(6):R60. doi:10.1186/gb-2011-12-6-r60 21. Salazar N, Gueimonde M, Hernandez‐Barranco A, Ruas‐ Madiedo P, de los Reyes‐Gavilan CG. Exopolysaccharides produced by intestinal Bifidobacterium strains act as 848 | CASTRO ET AL. 15364801, 2024, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jpn3.12356 by Readcube (Labtiva Inc.), Wiley Online Library on [10/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License fermentable substrates for human intestinal bacteria. Appl Environ Microbiol. 2008;74:4737‐4745. doi:10.1128/AEM. 00325-08 22. Lozupone C, Knight R. UniFrac: a new phylogenetic method for comparing microbial communities. Appl Environ Microbiol. 2005;71(12):8228‐8235. doi:10.1128/AEM.71.12.8228-8235.2005 23. Calvani M, Anania C, Bianchi A, et al. Update on food protein‐ induced enterocolitis syndrome (FPIES). Acta Biomed. 2021;92(S7):e2021518. doi:10.23750/abm.v92iS7.12394 24. Mennini M, Fierro V, Di Nardo G, Pecora V, Fiocchi A. Microbiota in non‐IgE‐mediated food allergy. Curr Opin Allergy Clin Immunol. 2020;20:323‐328. doi:10.1097/ACI.0000000000000644 25. Zubeldia‐Varela E, Barker‐Tejeda TC, Blanco‐Pérez F, Infante S, Zubeldia JM, Pérez‐Gordo M. Non‐IgE‐mediated gastrointestinal food protein‐induced allergic disorders. Clinical perspectives and analytical approaches. Foods. 2021;10(11):2662. doi:10.3390/ foods10112662 26. Litvak Y, Byndloss MX, Bäumler AJ. Colonocyte metabolism shapes the gut microbiota. Science. 2018;362(6418):eaat9076. doi:10.1126/science.aat9076 27. Litvak Y, Byndloss MX, Tsolis RM, Bäumler AJ. Dysbiotic proteobacteria expansion: a microbial signature of epithelial dysfunction. Curr Opin Microbiol. 2017;39:1‐6. doi:10.1016/j.mib.2017.07.003 28. Morgan XC, Tickle TL, Sokol H, et al. Dysfunction of the intestinal microbiome in inflammatory bowel disease and treatment. Genome Biol. 2012;13:R79. doi:10.1186/gb-2012-13-9-r79 29. Normann E, Fahlén A, Engstrand L, Lilja HE. Intestinal microbial profiles in extremely preterm infants with and without necrotizing enterocolitis. Acta Paediatr. 2013;102(2):129‐136. doi:10.1111/apa.12059 30. Boyer J, Scuderi V. P504 comparison of the gut microbiome between food protein‐induced enterocolitis sydrome (FPIES) infants and allergy‐free infants. Ann Allergy Asthma Immunol. 2017;119(5):e3. doi:10.1016/j.anai.2017.09.070 31. den Besten G, van Eunen K, Groen AK, Venema K, Reijngoud DJ, Bakker BM. The role of short‐chain fatty acids in the interplay between diet, gut microbiota, and host energy metabolism. JLipid Res. 2013;54(9):2325‐2340. doi:10.1194/jlr.R036012 32. Ríos‐Covián D, Ruas‐Madiedo P, Margolles A, Gueimonde M, de los Reyes‐Gavilán CG, Salazar N. Intestinal short chain fatty acids and their link with diet and human health. Front Microbiol. 2016;7:185. doi:10.3389/fmicb.2016.00185 33. Berni Canani R, De Filippis F, Nocerino R, et al. Gut microbiota composition and butyrate production in children affected by non‐IgE‐mediated cow's milk allergy. Sci Rep. 2018;8:12500. doi:10.1038/s41598-018-30428-3 34. Díaz M, Guadamuro L, Espinosa‐Martos I, et al. Microbiota and derived parameters in fecal samples of infants with non‐IgE cow's milk protein allergy under a restricted diet. Nutrients. 2018;10(10):1481. doi:10.3390/nu10101481 35. Rigottier‐Gois L. Dysbiosis in inflammatory bowel diseases: the oxygen hypothesis. ISME J. 2013;7(7):1256‐1261. doi:10.1038/ ismej.2013.80 36. Reichardt N, Duncan SH, Young P, et al. Phylogenetic distribution of three pathways for propionate production within the human gut microbiota. ISME J. 2014;8(6):1323‐1335. doi:10. 1038/ismej.2014.14 37. Zubeldia‐Varela E, Barber D, Barbas C, Perez‐Gordo M, Rojo D. Sample pre‐treatment procedures for the omics analysis of human gut microbiota: turning points, tips and tricks for gene sequencing and metabolomics. JPharm Biomed Anal. 2020;191:113592. doi:10.1016/j.jpba.2020. 113592 38. Li F, Ma J, Geng S, et al. Fecal calprotectin concentrations in healthy children aged 1‐18 months. PLoS One. 2015;10(3): e0119574. doi:10.1371/journal.pone.0119574 39. Ataee P, Zoghali M, Nikkhoo B, et al. Diagnostic value of fecal calprotectin in response to mother's diet in breast‐fed infants with cow's milk allergy colitis. Iran J Ped. 2018;28(4):e66172. doi:10.5812/ijp.66172 SUPPORTING INFORMATION Additional supporting information can be found online in the Supporting Information section at the end of this article. How to cite this article: Castro AM, Sabater C, Navarro S, et al. The intestinal microbiome of infants with cow's milk‐induced FPIES is enriched in taxa and genes of enterobacteria. J Pediatr Gastroenterol Nutr. 2024;79:841‐849. doi:10.1002/jpn3.12356 CASTRO ET AL. | 849 15364801, 2024, 4, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/jpn3.12356 by Readcube (Labtiva Inc.), Wiley Online Library on [10/01/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License