Regulatory T Cells and T Helper Subsets in Breast-fed Infants With Hematochezia Caused by Allergic Colitis

Cseh, Áron*; Molnár, Kriszta; Pintér, Petra; Szalay, Balazs*; Szebeni, Beata*; Treszl, Andras*; Arató, Andras; Vásárhelyi, Barna*; Veres, Gabor

Journal of Pediatric Gastroenterology & Nutrition: November 2010 - Volume 51 - Issue 5 - p 675–677
doi: 10.1097/MPG.0b013e3181e85b22
Short Communications

The objective of our study was to investigate the prevalence of CD4+ T lymphocyte subsets and their commitment to TH1 or TH2 direction in 10 infants with allergic colitis (AC) and 10 healthy controls. Infants with AC presented with a higher ratio of naïve to memory cells, lower prevalence of activated CD4+CD25+ cells and FoxP3+ regulatory cells, and a shift to TH2 direction in balance compared with controls. These alterations are normalized upon cessation of AC symptoms on elemental L-amino acid formula. These findings suggest the importance of antigen exposure in AC in infancy.

*Research Group for Pediatrics and Nephrology, Semmelweis University and Hungarian Academy of Sciences, Hungary

First Department of Pediatrics, Semmelweis University, Budapest, Hungary.

Received 1 March, 2010

Accepted 13 May, 2010

Address correspondence and reprint requests to Gábor Veres, First Department of Pediatrics, Bókay u. 53. H-1083 Budapest, Hungary (e-mail:

The present work was supported by grants TÁMOP-4.2.2-08/1/KMR-2008-0004, OTKA-76316, OTKA-K81117 and ETT-028-02.

The authors report no conflicts of interest.

Article Outline

In developed countries allergic colitis (AC) is one of the most common causes of gastrointestinal bleeding (hematochezia, HC) in breast-fed infants (1). The mechanism of AC is suggested to be a hypersensitivity to allergens excreted with breast milk and is regarded as a form of food allergy. This notion is supported by the clinical experience that AC is resolved upon switching mothers and their babies to an elimination diet or elemental formula. Nevertheless, there are a significant number of infants with AC in whom exclusion of cow's milk and egg from maternal diet does not cease HC. In a recent study only 18% of infants with HC had established cow's-milk allergy (2).

There is a growing body of evidence that delayed maturation of gastrointestinal immune system plays a crucial role in the pathomechanism of food allergies (3). In accordance with hygiene hypothesis (ie, a suboptimal exposure of individuals to microorganisms) the adaptive immunity is dysfunctional in subjects with food allergies (4). The prevalence of effector cells and activated lymphocytes (5) and regulatory T cells (Tregs) is indeed decreased in peripheral blood samples of children with food allergy (6). Simultaneously with cell number alterations a shift from TH1 to TH2 cytokines is also present (7).

The aim of our exploratory study was to phenotype the subsets of CD4+ cells such as naïve/effector, early/late activated and regulatory T lymphocyte population, along with their TH1/TH2 commitment in AC.

Back to Top | Article Outline


During a 16-month period (September 2006–January 2008) at the Outpatient Clinic of the First Department of Pediatrics, Semmelweis University, 34 infants with HC (age, median [range] 5 months [4 days–7.5 months]) were evaluated. After exclusion of infection and fissures, adherence to maternal elimination diet (ie, exclusion of cow's milk and eggs from diet) was advised. After 1 month HC persisted in 11 infants who were still breast-fed; of those, 1 infant was diagnosed as having Crohn disease on performing colonoscopy. The remaining 10 infants with HC (age, median [range] 4.5 [1.0–6.5] months) were enrolled in the study (Fig. 1 and Table 1). In these patients, AC was established according to characteristic signs on colonoscopy (eg, lymphoid nodular hyperplasia or aphthous ulceration), elevated number of eosinophils in colonic biopsy specimens (>6/1 high power field), and cessation of HC after discontinuation of breast-feeding and the introduction of elemental L-amino acid formula (Neocate; SHS Int, Liverpool, UK) with no recurrence after long-term follow-up (13–24 months) (2,8).

Two milliliters of lithium-heparin anticoagulated blood was taken at the time of diagnosis. All 10 infants were switched to elemental L-amino acid–based formula or solids. The second blood sample was taken when symptoms resolved (after median [range] 2 [1–3] months). A fecal occult blood test (HSV10; Diagnosticum Zrt, Budapest, Hungary) was used to verify the complete absence of blood in stool. Ten age- and sex-matched and breast-fed matched healthy infants with functional abdominal pain were used as controls; blood sampling was done simultaneously with routine laboratory measurements. The institutional ethical committee approved our study; written parental informed consent was obtained.

Peripheral blood mononuclear cells were separated and stored at −80 °C until the measurements were taken (9). Intracellular FoxP3 assay (eBioscience, San Diego, CA) and the measurement and identification of cell surface markers (CD4, CD25, CD45RO, CD45RA, CD62L, CCR4, and CXCR3; BD Biosciences Pharmingen, San Diego, CA) were performed with a BD FACS Aria according to the manufacturer's protocols. We identified the subsets of T helpers (eg, CD4+), such as TH1 and TH2 committed lymphocytes (eg, CXCR3+ and CCR4+, respectively), naïve and effector/memory cells (eg, CD45RA+ and CD45RO+, respectively), early and late activation marker expressing (eg, CD25+ and CD62L+, respectively), and regulatory T lymphocytes (eg, CD4+FoxP3+).

The measurements of the serum levels of interferon (IFN)-γ and interleukin (IL)-4 were done with Bioplex Protein Array system (Bio-Rad, Hemel Hempstead, UK).

The Mann-Whitney test was applied for the comparison of data at the baseline from controls and infants with AC. Cell prevalence values at baseline and by the end of therapy were compared by paired Wilcoxon tests.

Back to Top | Article Outline


Our results are summarized in Table 2. At baseline the prevalence of Tregs was lower in infants with AC than in controls (P = 0.03). Simultaneously, the CD4+CD45RA+/CD4+CD45RO+ ratio was higher, and the prevalence of CD4+CD25+ cells was lower in infants with AC (P = 0.02 and P = 0.01, respectively). The CD4+CXCR3+/CD4+CXCR4+ and IFN-γ/IL-4 ratios were lower in patients with AC compared with the controls (P = 0.02 and P = 0.04, respectively). We also tested the alteration of these parameters after the resolution of AC. We found that the prevalence of Tregs increased to normal at this time (P = 0.02). Although other cell types tested did not change significantly during the therapy in comparison with baseline cell prevalence values, the CD4+CXCR3+/CD4+CXCR4+ and IFN-γ/IL-4 ratios and cell prevalence data measured at symptom cessation were not significantly different from those in the controls.

Back to Top | Article Outline


The present study revealed that AC is accompanied by low Treg numbers, an increased ratio of naïve/memory T cells, a decreased prevalence of activated T cells, and a shift of CD4+ T lymphocytes to TH2 direction. Later, with the cessation of AC signs and symptoms these alterations cease. These findings indicate that an immune dysfunction may contribute to this gastrointestinal disorder in infancy.

The pathomechanism of AC in healthy breast-fed infants is still unknown (10). Earlier studies revealed an increased prevalence of T cells in the colonic epithelium (11) and immunoglobulin deficiencies (12) in infants with AC. HC generally resolves after the elimination of cow's milk from the maternal diet in most breast-fed infants with AC. Although recent studies demonstrated that the majority of HC cases (up to 82%) are not linked to cow's-milk allergy (2), theoretically there is still a high chance to find AC among patients with HC. Therefore, the possible presence of food allergy in patients with HC is an important issue. A recent article showed, however, that the use of food allergy screening tests and measurement of specific serum IgE levels do not improve the diagnostic workup of this patient population (2). In our study population, none of the patients had cow's-milk or egg allergy, whereas typical macroscopic (lymphonodular hyperplasia with erythema, bleeding spot/aphthous lesions) and microscopic abnormalities (elevated number of eosinophils) and positive response after introduction of amino acid–based formula clearly showed AC.

In the present study we focused on a peculiar subpopulation of patients with HC in whom signs and symptoms persisted even after 1 month. This approach guarantees that patients with acute infection, fissures, and the like are excluded from our study. Furthermore, endoscopy also excluded infants with other rare conditions.

Recent reports suggest the paramount importance of Tregs in food allergies in children and adults (6). This specialized subgroup of CD4+ cells is of central importance in maintaining gastrointestinal tolerance to oral antigens. Interestingly, in our study the prevalence of lymphocytes expressing early and late activation markers (eg, CD25+ and CD62L+, respectively) and effector/memory cells (CD45RO+) is not high in spite of the low Treg (CD4+FoxP3+) prevalence. This is an apparent contradiction given the known suppressing action of Tregs on CD4 cells.

However, alterations in cell prevalence detected in AC fit well into the concept of hygiene hypothesis (13). Indeed, microbial antigens are powerful inducers of Treg generation, memory cells, and T lymphocyte activation. The simultaneous decrease of Treg, memory cell, and activated T cell prevalence in AC suggests that this disorder may be linked to a dysmaturation of adaptive immunity, possibly due to the suboptimal antigen exposure (4). It is worth assuming that this may be a player in immune imbalance characterized by a lower ratio of TH1/TH2 cells.

In addition, previous studies clearly showed that gastrointestinal flora and certain probiotic bacteria may have a beneficial effect in the treatment of gastrointestinal allergic conditions by inducing Tregs (14,15). Lactobacillus reuteri– and L casei–primed monocyte-derived dendritic cells were demonstrated to induce the development of Treg cells (14). The induced Treg cells increased the IL-10 level and were capable of inhibiting the proliferation of bystander T cells in an IL-10–dependent fashion. In a recent study (15) administration of probiotic strains induced Tregs. FoxP3+ cells were induced from CD25- cells and were able to suppress effector T cells.

When considering the significance of our observations we should take the limitations of our study into account. First, we identified Tregs with FoxP3 intracellular expression, because it is regarded to be the most reliable marker of Tregs (16). The use of other markers of Tregs such as CD127, CD4+CD25high would lead to some discordant results. Second, with the alteration of prevalence of CD4, the target cells of Tregs may support indirectly that altered Treg prevalence corresponds to altered suppressive action. This hypothesis should be reinforced by functional assays; however, due to the limited blood volume we could not obtain data regarding their functional characteristics in the present study. Third, we measured peripheral Treg numbers and their cellular network. Data obtained from previous studies on other types of bowel inflammation such as inflammatory bowel disease support the notion that peripheral and central Treg prevalence values may be discordant (17). Therefore, our results reflect the systemic alteration of immune cell prevalence values instead of local alterations. However, the lack of biopsy specimens from healthy control infants and from healed patients prevented us from testing local cellular milieu. Fourth, although we measured simultaneously increasing Treg prevalence with the normalization of TH1/TH2 ratio, it is not clear from our results whether these occur coincidentally or whether there is a real causative relation between these phenomena. Fifth, the low number of patients in the present study is a major limitation; however, the congruent results may form a basis for larger studies. If our results are reinforced by other groups, then there is also the possibility that they provide a new therapeutic approach. If AC is indeed an early form of allergy and its pathogenesis includes the germfree conditions in early infancy, then infants' exposure to microbial antigens (with oral vaccines or probiotics) in AC may be a reasonable approach.

Back to Top | Article Outline


The invaluable contribution of Edina Biro to the measurement of cytokine levels is highly appreciated.

Back to Top | Article Outline


1. Machida HM, Catto Smith AG, Gall DG, et al. Allergic colitis in infancy: clinical and pathologic aspects. J Pediatr Gastroenterol Nutr 1994; 19:22–26.
2. Arvola T, Ruuska T, Keränen J, et al. Rectal bleeding in infancy: clinical, allergological, and microbiological examination. Pediatrics 2006; 117:e760–e768.
3. Eigenmann PA. Mechanisms of food allergy. Pediatr Allergy Immunol 2009; 20:5–11.
4. Prioult G, Nagler-Anderson C. Mucosal immunity and allergic responses: lack of regulation and/or lack of microbial stimulation? Immunol Rev 2005; 206:204–218.
5. Beyer K, Renz H, Wahn U, et al. Changes in blood leukocyte distribution during double-blind, placebo-controlled food challenges in children with atopic dermatitis and suspected food allergy. Int Arch Allergy Immunol 1998; 116:110–115.
6. Karlsson MR, Rugtveit J, Brandtzaeg P. Allergen-responsive CD4+CD25+ regulatory T cells in children who have outgrown cow's milk allergy. J Exp Med 2004; 199:1679–1688.
7. Cox HE. Food allergy as seen by an allergist. J Pediatr Gastroenterol Nutr 2008; 47:S45–S48.
8. Ravelli A, Villanacci V, Chiappa S, et al. Dietary protein-induced proctocolitis in childhood. Am J Gastroenterol 2008; 103:2605–2612.
9. Svec P, Vásárhelyi B, Pászthy B, et al. Do regulatory T cells contribute to Th1 skewness in obesity? Exp Clin Endocrinol Diabetes 2007; 115:439–443.
10. Lake AM. Food-induced eosinophilic proctocolitis. J Pediatr Gastroenterol Nutr 2000; 30:S58–60.
11. Ormälä T, Rintala R, Savilahti E. T cells of the colonic mucosa in patients with infantile colitis. J Pediatr Gastroenterol Nutr 2001; 33:133–138.
12. Ojuawo A, St Louis D, Lindley KJ, et al. Non-infective colitis in infancy: evidence in favour of minor immunodeficiency in its pathogenesis. Arch Dis Child 1997; 76:345–348.
13. Romagnani S. The increased prevalence of allergy and the hygiene hypothesis: missing immune deviation, reduced immune suppression, or both? Immunology 2004; 112:352–363.
14. Smits HH, Engering A, van der Kleij D, et al. Selective probiotic bacteria induce IL-10-producing regulatory T cells in vitro by modulating dendritic cell function through dendritic cell-specific intercellular adhesion molecule 3-grabbing nonintegrin. J Allergy Clin Immunol 2005; 115:1260–1267.
15. de Roock S, van Elk M, van Dijk ME, et al. Lactic acid bacteria differ in their ability to induce functional regulatory T cells in humans. Clin Exp Allergy 2010; 40:103–110.
16. Hori S, Nomura T, Sakaguchi S. Control of regulatory T cell development by the transcription factor Foxp3. Science 2003; 299:1057–1061.
17. Himmel ME, Hardenberg G, Piccirillo CA, et al. The role of T-regulatory cells and Toll-like receptors in the pathogenesis of human inflammatory bowel disease. Immunology 2008; 125:145–153.

allergic colitis; food allergy; hematochezia; infants; lymphocyte activation

Copyright 2010 by ESPGHAN and NASPGHAN