Joann Kerperien

IMMUNOMODULATORY PROPERTIES OF DIETARY NON-DIGESTIBLE GALACTO-, FRUCTO- AND ACIDIC-OLIGOSACCHARIDES IN VACCINATION AND COW’S MILK ALLERGY JOANN KERPERIEN

Mucosal and systemic immune-regulation by fermentable ﬁ bers

IMMUNOMODULATORY PROPERTIES OF DIETARY NON-DIGESTIBLE GALACTO-, FRUCTO- AND ACIDIC- OLIGOSACCHARIDES IN VACCINATION AND COW’S MILK ALLERGY Mucosal and systemic immunoregulation by fermentable ﬁ bers

JoAnn Kerperien

IMMUNOMODULATORY PROPERTIES OF DIETARY NON-DIGESTIBLE GALACTO-, FRUCTO- AND ACIDIC-OLIGOSACCHARIDES IN VACCINATION AND COW’S MILK ALLERGY Mucosal and systemic immunoregulation by oligosaccharides

JoAnn Kerperien

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 1 ISBN: 978-90-393-7137-4 Thesis and Cover Design: Studio 0404, Nijmegen Printed by: Ipskamp Printing, Enschede, The Netherlands

© JoAnn Kerperien, 2019, Nijmegen, The Netherlands All rights reserved. No part of this thesis publication may be reproduced, stored in a retrieval system of any nature, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, included a complete or partial transcription, without the prior written permission of the author. Application for this should be addressed to the author. The research conducted in this thesis was financially supported by the Utrecht Institute for Pharmaceutical Sciences and Nutricia Research B.V.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 2 IMMUNOMODULATORY PROPERTIES OF DIETARY NON-DIGESTIBLE GALACTO-, FRUCTO- AND ACIDIC-OLIGOSACCHARIDES IN VACCINATION AND COW’S MILK ALLERGY Mucosal and systemic immunoregulation by oligosaccharides

Immunomodulatoire eigenschappen door dieetinterventie met onverteerbare galacto-, fructo- en zure- oligosachariden tijdens vaccinatie en bij koemelkallergie Systemische en mucosale immunoregulatie door oligosachariden (met een samenvatting in het Nederlands)

PROEFSCHRIFT

Ter verkrijging van de graad van doctor aan de Universiteit Utrecht op gezag van de rector magnificus, prof. Dr. H.R.B.M. Kummeling ingevolge het besluit van het college voor promoties in het openbaar te verdedigen op maandag 17 juni 2019 des middags te 2.30 uur

door

JoAnn Kerperien

Geboren op 23 juni 1980 te Borculo

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 3 Promotor: Prof. Dr. J. Garssen

Copromotor: Dr. L.E.M. Willemsen Dr. L.M.J. Knippels

The publication of this thesis was financially supported by: Nutricia Research, B.V.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 4 …. I can go the distance, and I’ll stay on track, no I won’t accept defeat, It’s an uphill slope, but I won’t lose hope, ‘till I go the distance…..

Go the distance, Hercules

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 5 530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 6 ‘Voor de wereld was jij maar iemand, voor mij was jij de wereld’ Mijn laatste belofte aan jou heb ik hiermee volbracht. Voor mijn allerliefste mama

Anke Kerperien 1960- 2015 †

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 7 530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 8 TABLE OF CONTENTS

Chapter 1 General introduction 11

Chapter 2 De velopment of the immune system; Early nutrition and 21 consequences for later life

Chapter 3 Non-diges tible oligosaccharides modulate intestinal 49 immune activation and suppress cow’s milk allergic symptoms

Chapter 4 IL10r or TGFβ neutralization abrogates the protective effect 67 of a specific non-digestible oligosaccharide mixture in cow’s milk allergic mice

Chapter 5 Alt erations in regulatory T cells induced by specific 89 oligosaccharides improve vaccine responsiveness in mice

Chapter 6 Diet ary vitamin D supplementation is ineffective in 107 preventing murine cow’s milk allergy, irrespective of the presence of non-digestible oligosaccharides

Chapter 7 Diet ary vitamin A supports a mixture of specific non- 133 digestible oligosaccharides in the prevention of cow’s milk allergic symptoms in mice

Chapter 8 General discussion 163

Abbreviations 181

Appendix Nederlandse samenvatting 189 Dankwoord 197 Curriculum Vitae 203 Publications 205

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 9 1 530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 10 General 1 introduction 530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 11 CHAPTER 1

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 12 GENERAL INTRODUCTION

GENERAL INTRODUCTION 1 The prevalence of allergic diseases has been rising in industrialized countries over the last decades (1). Often the first allergy to develop in early life is a food allergy. The risk of developing food allergy is increasing in the developed world and it has been suggested that food allergy is a disease of the more educated upper class (1). Furthermore, more children with food allergies are found in cities than in rural areas (1). Also there is a possible link between food allergy and the diminishing period of breast feeding, increased consumption of fat and n-6 over n-3 polyunsaturated fatty acids, low vitamin D levels and late introduction of allergenic foods (1). There are several well-known food allergies, such as allergy for egg, cow’s milk, peanut and also for wheat, fish, soy and several other food proteins. Cow’s milk allergy (CMA) is one of the first allergies to develop in life. One to three percent of children develop CMA and symptoms range from behavioral changes, to cutaneous, respiratory, conjunctival, gastrointestinal and cardiovascular reactions ranging from mild symptoms to anaphylaxis, from acute to delayed to chronic (2-4). This allergy which is characterized by acute allergic reactions upon oral ingestion of the culprit food can be IgE mediated or non-IgE mediated (4, 5). Severe allergic reaction to milk and high levels of milk specific IgE can predispose to persistence of the allergy (6, 7). For some allergies, like cow’s milk, egg, wheat or soy, most children will outgrow their allergy (3, 4, 6) although there are studies showing that CMA is getting more persistent (8, 9). Food allergy is not only a personal and social burden, it is also an economic burden (3). Over the recent years more hospitalizations to anaphylaxis from food allergies have been registered (3). Even those who outgrow their food allergy may be more predisposed to develop other allergies later in life. Although there is discussion for more specificity for the term atopic march, food allergy to peanut, egg and milk were significantly associated with the development of asthma and rhinitis (10, 11). So, it is important to find more preventive or treatment possibilities to stop this atopic expansion especially since currently no treatment for food allergy is available.

The maternal diet during pregnancy may influence the allergy status in the unborn children which may be maintained throughout life (12). Previously it was advised after birth that children at risk should avoid specific food particles, however this approach is being challenged. Recent studies show that early introduction rather than total avoidance of specific food proteins in the first year of life may be required to prevent development of food allergy (3). This was specifically shown for peanut allergy in the Learning Early About Peanut allergy (LEAP) study (13). In this study it was shown that children with severe eczema, egg allergy or both developed less peanut allergy when consuming products containing peanut (13). This was independent of the susceptibility to develop this allergy at the start of the study (13). Ingestion of baked milk products by children affected with CMA may accelerate and improve oral tolerance induction to milk (6). For now however the only safe option for CMA is total avoidance of the causative food proteins (3, 4, 14). Long lasting permanent tolerance induction via allergen specific

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 13 CHAPTER 1

immunotherapy is still a challenge and care should be taken since unwanted allergic side effects often occur. Hence, prevention of food allergy development is an important topic of investigation and dietary intervention with non-digestible oligosaccharides (NDO) alone or in combination with vitamin D and vitamin A may contribute to acquire natural oral tolerance to food proteins directly after birth and in the first year of life (12).

Oral tolerance is acquired in the gastro-intestinal tract through a state of non-responsiveness of the immune system to harmless food particles. An allergic reaction is triggered when the immune system incorrectly identifies a food particle as harmful. Knowledge about (oral) tolerance in general comes from murine experimental models (15). Specialized epithelial cells, for example microfold (M) cells, are responsible for capturing antigens from the intestine and shuttling them to the underlying dendritic cells (DC) in the Peyer’s Patches

(PP). In the intestine DC and fractalkine (CX3CR1) positive macrophages, sample antigens through the intestinal epithelial layer with cell protrusions, or they acquire particles via the epithelial layer through cell-cell contact and transfer these to immature DC (16). DC present the antigen to cells in the Lamina Propria (LP) or they can travel to the mesenteric lymph nodes (MLN) (15, 17). Both transforming growth factor beta (TGFβ) and interleukin (IL)10 are produced by DC and together with retinoic acid (RA) are held responsible for the development and maintenance of a tolerogenic environment in the MLN and LP of the small intestine (15, 16, 18-21). In this environment regulatory T cell (Treg) subsets are produced in the MLN and the activated Treg, expressing the necessary chemokine receptor CCR9 and integrin α4β7, migrate to the intestinal LP (16). Treg manage different cell populations by cell-cell contact through cytotoxic T-lymphocyte associated protein 4 (CTLA4), lymphocyte activation gene 3 (LAG3) or programmed cell death protein 1 (PD1) and by cytokine production of TGFβ, IL10 and IL35. These cytokines exert their tolerogenic responsibility via inducing B cell class switching to IgA, supporting Treg survival, inhibiting pro-inflammatory cell types and restricting the production of Th cell types (15, 21). The Th1 cell population is responsible for clearance of bacteria and protozoa, the Th2 cell population is responsible for extracellular parasites and Th17 is responsible for the clearance of pathogens under normal conditions. The disbalance between Treg/ Th1/ Th2 populations in favor of Th2 is responsible for the cellular cascade contributing to allergic sensitization (15). Th2 cells activate B cells to proliferate, produce allergen specific IgE and upregulate this IgE production. Subsequently IgE binds to type I and II Fcε receptors on allergic effector cells such as mast cells and basophils. Upon re-exposure to this allergen, allergen specific IgE binds to basophils, which release IL4 and IL13 and these cytokines binds to mast cells, which in turn will degranulate and release several factors like histamine, prostaglandins, proteases and tumor necrosis factor alfa (TNFα) resulting in local and/or systemic allergic reactions (Figure 1). NDO, present in human milk, have a beneficiary influence on intestinal gut microbiota and on the immune system. Glycans, resembling some of the functional aspects of human milk

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 14 GENERAL INTRODUCTION

oligosaccharides, are isolated from vegetable sources or enzymatically produced from lactose and used in cow’s milk based infant milk formulas. Since cow’s milk contains a very low 1 amount of NDO compared to human milk these beneficiary structures are added to mimic a more natural setting. The beneficiary effect of plant derived Fructo-oligosaccharides (FOS) was recognized in the mid-eighties (22), the beneficiary effect of lactose derived Galacto- oligosaccharides (GOS) in the nineties (23) and the 9:1 combination of short chain (sc)GOS and long chain (lc)FOS (GF) provided an improved stool consistency in infants and supported the growth of bifidobacteria and lactobacilli similar to those in breastfed infants (24). Some ten years later pectin-derived Acidic-oligosaccharides (pAOS) were combined with neutral scGOS and lcFOS (GFA) and this mixture was shown to beneficially modify the intestinal microbiota and to suppress the development of atopic dermatitis in infants comparable to breastfed infants (25). A diverse population of intestinal microbiota are held responsible for lower basal levels of IgE and a diminished sensitization to food allergens (16, 26). Metabolites, like butyrate from dietary fibers, fermented by the microbiota can even induce CD103+ DC in vitro and Treg in the colon in vivo (16, 27). The combination of GF or GFA suppressed airway inflammation and hyperresponsiveness in an ovalbumin (OVA) induced murine asthma model and GFA enhanced the Th1 dependent vaccination response (28, 29). Schouten et al. showed that GFA partially prevented the development of CMA in mice (30). GFA also induced whey specific CD25+ regulatory T cells in these CMA donor mice, which were crucially involved in the suppression of CMA in recipient mice (30). In this thesis the mechanisms by which GFA exerts its beneficial effects were further explored to have a better understanding of their contribution to the development of immune tolerance for harmless cow’s milk proteins.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 15 CHAPTER 1

Figure 1. The intestinal immune system. In the intestine, food proteins, beneficial as well as harmful bacteria, secretory IgA and dietary components, like GFA, are present. The intestinal epithelial lining contains different specialized cells, among them is the Microfold (M) cell, which allows transport of microbes or particles into the sub epithelial dome (SD) within the Peyer’s Patches (PP). Directly underneath the epithelial cells in the Lamina Propria (LP) and the PP immune cells are present. Dendritic cells (DC) will sample antigens and will either remain in the PP or travel to the mesenteric lymph nodes (MLN) where immune responses are generated. In the MLN naïve T cells (nT) will be instructed and travel towards the LP or PP via the bloodstream (schematic overview in yellow box). In case of allergic sensitisation (1a) T helper (Th) 2 cells will signal for B cells (B) to produce IgE. IgE will opsonize mast cells (M) and when IgE cross-linking occurs upon a second exposure to the allergen, mast cells will degranulate (1b), which will initiate the allergic response by releasing factors like histamine and TNFα. Under physiological conditions (2), regulatory T cells (Treg) will actively control the Th, B and DC response to sustain a non-responsive state towards harmless proteins like food proteins. Black arrows indicate migration towards MLN, green arrows indicate conversion of naïve T cell into Treg or Th effector cells (Th2 in case of allergic sensitization), dotted green arrows indicate suppressive effect mediated by direct cell contact or soluble mediators like interleukin 10 (IL10) and transforming growth factor beta (TGFβ).

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 16 GENERAL INTRODUCTION

THESIS OUTLINE 1 The aim of this thesis was to identify factors involved in the preventive effect of dietary NDO in CMA development in a mouse model for CMA. The main questions are: 1) Do NDO mixtures GFA, GF or the single components differentially affect CMA development and what are the underlying mechanisms? 2) Do soluble mediators IL10 or TGFβ contribute to the protective effect of GFA in CMA development? 3) Do GFA affect the vaccination response? 4) Do immunoregulatory vitamin A or D further enhance the preventive effect of GFA in CMA development?

In the current chapter a general outline of this thesis is presented and Chapter 2 provides a general overview of beneficiary factors required for healthy immune development and oral tolerance induction versus allergy development. In Chapter 3 the immune modulating properties of prebiotic non-digestible dietary components scGOS, lcFOS and pAOS were tested separately and compared to the combinations of GF or GFA in the CMA mouse model. Both combinations did have a preventive effect on the onset of CMA, but different underlying immune effects were observed. Since GFA contains both acidic as well as neutral oligosaccharides in the same ratio as can be found for these kind of glycans in human milk, GFA was further studied. The GFA mixture appeared to enhance the mRNA expression levels of Tbet (marker for Th1), IL10 and TGFβ (regulatory cytokines). Previous results showed the capacity of GFA to induce functional specific Treg which was further investigated in the current thesis. Chapter 4 describes the preventive effect of GFA as dietary supplementation during the development of CMA in the mouse model and studies the underlying mechanism by means of neutralization of the IL10 receptor or TGFβ separately using antibodies. Both IL10r or TGFβ blocking abrogated the preventive effect of GFA on CMA completely. In Chapter 5 it is shown that GFA improved the vaccination status of mice via enhancing the Th1 response. Not only NDO influence the immune system early in life, also certain vitamins play a crucial role in the establishment of a healthy immune balance. Due to the beneficiary influence of vitamin A and vitamin D on the immune system of the foetus and after birth, we evaluated their role in combination with the preventive effects of GFA in the CMA model. Chapter 6 describes the study in which mice were fed normal vitamin D levels compared to mice fed pellets containing no vitamin D, or a fourfold of normal vitamin D levels with or without GFA. However, no beneficiary effect of extra supplementation of vitamin D on the preventive effect of GFA in CMA development was observed in these mice. For vitamin A, we also used the preventive study set up. Mice were fed pellets containing no vitamin A, normal levels of vitamin A and twice the amount of normal vitamin A with or without GFA as described in Chapter 7. The extra vitamin A did support the preventive effect of GFA on CMA development. In Chapter 8 a summarizing discussion is presented of the findings in this thesis considering the current knowledge.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 17 CHAPTER 1

REFERENCES

1. Fiocchi A, Dahdah L, Fierro V, et al. Food 10. Hill DA, Spergel JM. The atopic march: allergy trends at the crossing among socio- Critical evidence and clinical relevance. Ann economics, history and geography. Curr Allergy Asthma Immunol. 2018;120(2):131-7. Opin Allergy Clin Immunol. 2018;18(3):271-6. 11. Busse WW. The atopic march: Fact or 2. Deschildre A, Lejeune S. How to cope with folklore? Ann Allergy Asthma Immunol. food allergy symptoms? Curr Opin Allergy 2018;120(2):116-8. Clin Immunol. 2018;18(3):234-42. 12. Skypala IJ, McKenzie R. Nutritional Issues 3. Yu W, Freeland DMH, Nadeau KC. Food in Food Allergy. Clin Rev Allergy Immunol. allergy: immune mechanisms, diagnosis 2018. and immunotherapy. Nat Rev Immunol. 13. Du Toit G, Roberts G, Sayre PH, et al. 2016;16(12):751-65. Randomized trial of peanut consumption in 4. Mousan G, Kamat D. Cow’s Milk infants at risk for peanut allergy. N Engl J Protein Allergy. Clin Pediatr (Phila). Med. 2015;372(9):803-13. 2016;55(11):1054-63. 14. Pecora V, Mennini M, Calandrelli V, et al. 5. Venter C, Brown T, Meyer R, et al. Better How to actively treat food allergy. Curr Opin recognition, diagnosis and management Allergy Clin Immunol. 2018;18(3):248-57. of non-IgE-mediated cow’s milk allergy in 15. Wambre E, Jeong D. Oral Tolerance infancy: iMAP-an international interpretation Development and Maintenance. Immunol of the MAP (Milk Allergy in Primary Care) Allergy Clin North Am. 2018;38(1):27-37. guideline. Clin Transl Allergy. 2017;7:26. 16. Tordesillas L, Berin MC. Mechanisms of Oral 6. Dahdah L, Pecora V, Riccardi C, et al. How Tolerance. Clin Rev Allergy Immunol. 2018. to predict and improve prognosis of food 17. du Pre MF, Samsom JN. Adaptive T-cell allergy. Curr Opin Allergy Clin Immunol. responses regulating oral tolerance to 2018;18(3):228-33. protein antigen. Allergy. 2011;66(4):478-90. 7. Koike Y, Sato S, Yanagida N, et al. Predictors 18. Coombes JL, Siddiqui KR, Arancibia- of Persistent Milk Allergy in Children: A Carcamo CV, et al. A functionally specialized Retrospective Cohort Study. Int Arch Allergy population of mucosal CD103+ DCs induces Immunol. 2018;175(3):177-80. Foxp3+ regulatory T cells via a TGF-beta and 8. Skripak JM, Matsui EC, Mudd K, et al. retinoic acid-dependent mechanism. J Exp The natural history of IgE-mediated cow’s Med. 2007;204(8):1757-64. milk allergy. J Allergy Clin Immunol. 19. Shiokawa A, Kotaki R, Takano T, et al. 2007;120(5):1172-7. Mesenteric lymph node CD11b(-) CD103(+) 9. Wood RA, Sicherer SH, Vickery BP, et PD-L1(High) dendritic cells highly al. The natural history of milk allergy in induce regulatory T cells. Immunology. an observational cohort. J Allergy Clin 2017;152(1):52-64. Immunol. 2013;131(3):805-12. 20. Boucard-Jourdin M, Kugler D, Endale Ahanda ML, et al. beta8 Integrin Expression

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 18 GENERAL INTRODUCTION

and Activation of TGF-beta by Intestinal a murine influenza vaccination model. Int Dendritic Cells Are Determined by Both Immunopharmacol. 2006;6(8):1277-86. 1 Tissue Microenvironment and Cell Lineage. 29. Vos AP, van Esch BC, Stahl B, et al. J Immunol. 2016;197(5):1968-78. Dietary supplementation with specific 21. Girard-Madoux MJ, Ober-Blobaum JL, oligosaccharide mixtures decreases Costes LM, et al. IL-10 control of CD11c+ parameters of allergic asthma in mice. Int myeloid cells is essential to maintain Immunopharmacol. 2007;7(12):1582-7. immune homeostasis in the small and large 30. Schouten B, van Esch BC, Hofman GA, et intestine. Oncotarget. 2016;7(22):32015-30. al. Oligosaccharide-induced whey-specific 22. Mitsuoka T, Hidaka H, Eida T. Effect of fructo- CD25(+) regulatory T-cells are involved in oligosaccharides on intestinal microflora. the suppression of cow milk allergy in mice. Nahrung. 1987;31(5-6):427-36. J Nutr. 2010;140(4):835-41. 23. Djouzi Z, Andrieux C. Compared effects of three oligosaccharides on metabolism of intestinal microflora in rats inoculated with a human faecal flora. Br J Nutr. 1997;78(2):313-24. 24. Moro G, Minoli I, Mosca M, et al. Dosage- related bifidogenic effects of galacto- and fructooligosaccharides in formula-fed term infants. J Pediatr Gastroenterol Nutr. 2002;34(3):291-5. 25. Fanaro S, Jelinek J, Stahl B, et al. Acidic oligosaccharides from pectin hydrolysate as new component for infant formulae: effect on intestinal flora, stool characteristics, and pH. J Pediatr Gastroenterol Nutr. 2005;41(2):186-90. 26. Ho HE, Bunyavanich S. Role of the Microbiome in Food Allergy. Curr Allergy Asthma Rep. 2018;18(4):27. 27. Qiang Y, Xu J, Yan C, et al. Butyrate and retinoic acid imprint mucosal-like dendritic cell development synergistically from bone marrow cells. Clin Exp Immunol. 2017;189(3):290-7. 28. Vos AP, Haarman M, Buco A, et al. A specific prebiotic oligosaccharide mixture stimulates delayed-type hypersensitivity in

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 19 2 530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 20 Development of the immune system; Early nutrition and consequences for later life

JoAnn Kerperien1, Bastiaan Schouten2, Günther Boehm2,3, Linette E.M. Willemsen1, Johan Garssen1,2, Léon M.J. Knippels1,2 and Belinda van ’t Land2,4

1 Division of Pharmacology, Department of Pharmaceutical Sciences, Faculty of Science, Utrecht University, the Netherlands 2 Nutricia Research B.V., Utrecht, The Netherlands 3 Sophia Children’s Hospital, Erasmus University Rotterdam, The Netherlands 4 Wilhelmina Children’s Hospital, University Medical Center, Utrecht, The Netherlands

Part of this chapter is published in: ‘Recent advances in Immunology to Target Cancer, Infl ammation and Infections’ 2011, 2 ISBN: 978-953-51-0592-3 530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 21 CHAPTER 2

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 22 IMMUNE DEVELOPMENT

INTRODUCTION

The immunological interaction between mother and foetus during pregnancy causes the foetal immune system to avoid excessive and destructive immunological reactions. This particular physiologic situation coexists with an immature immune system, which makes the infant very vulnerable for infections and susceptible to the development of immune system related 2 disorders. At birth, the immune system of the infant is particularly characterized by a not fully developed non-specific immune system. In addition, a suppressed capacity of antigen-specific T cells, a deletion of activated T cells, and the presence of high amounts of regulatory T cells (Treg) hamper proper immune responsiveness. During the first months of life the antigen- specific immune response must be developed in parallel to the maintenance of immune tolerance against compounds commonly found in the environment of mother and infant. There is evidence that disturbances of these complex developmental processes will have impact on the function of the immune system during lifetime causing immunological disorders such as allergy and autoimmunity.

Human milk contains several immunological active compounds which protect the infant from infection. Many of them, such as antibodies, are individually adapted to the maternal environment which is similar to the environment of the infant. This environment provides individual protection to the infant. Apart from this protection activated immediately after birth, human milk modulates also the described immunological developmental processes. Although the mechanisms of this modulation are not fully understood there is evidence that human milk can transfer “immunological memory of the mother” to the infant. This concept of the role of human milk underlines the importance of quality of nutrition during first months of life for total development of the immune system. Individual human milk analyses will provide insight in components that are important modulators. Immunologic active peptides, polyunsaturated fatty acids (PUFA), several glycolipids and non-digestible oligosaccharides (NDO) have already been identified as such modulators. The interaction of these active components with different parts of the immune system is very complex allowing a graduate and balanced development of the immune system.

One problem of studies in animals and humans is the fact that no single biomarker exists which describes the developmental status of the immune system entirely. The question which biomarkers are relevant is still a matter of intensive research. There is evidence that many “classical” biomarkers are not useful, because these are only sensitive in case the immune system is imbalanced but insensitive during normal development. Consequently, research to identify relevant biomarkers characterizing healthy immune development is strongly required. First results are promising indicating that the quality of nutrition early in life might support development of the immune system for lifetime. Acceptance of such a concept might provide

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 23 CHAPTER 2

opportunities for new ways of primary prevention of immune related diseases later in life. This chapter will summarise the newest and some specific insights of the mechanisms and impact of nutrition on the development of the immune system early in life.

INFLUENCES PRIOR LIFE

The nutritional status in early life has an important influence on human immune development, for example, a positive association is clearly observed between birth weight and antibody response to certain vaccines later in life (2). The precise relationship between nutritional exposures during critical periods of development and later immune function warrants further investigation. The early postnatal environment is a vital determinant of adult health. Environmental exposure, like bacteriological or nutritional modulation of the evolving intestine in utero and during early infancy has impact on immune function throughout life (3, 4). A concept like this (illustrated in Figure 1) will provide opportunities for primary prevention from immune related diseases later in life.

Figure 1. Disturbances of immune developmental processes will have impact to the function of the immune system during life causing immunological disorders such as allergy and autoimmunity. Although there are many questions still open first results are indicating that feeding early in life might support the development of the immune system for lifetime.

To understand the impact of nutrition on immune development early in life it is of key importance to know which steps in immune development are subject to change and depend on specific nutrition. During embryogenesis, stem cells start to differentiate into specific progenitor stem cells, creating a pool of more specific and less totipotent stem cells. Hematopoietic stem cells (HSC) are the progenitor cells for our whole immune system (Figure 2). Identification of the first HSC is still difficult, because these regions don’t contain many HSC and unique markers are lacking (5). After these few first HSC have colonized the human foetal liver, these cells expand and will relocate under influence of adhesion molecules and chemo-attractants to thymus,

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 24 IMMUNE DEVELOPMENT

spleen and bone marrow (6-8). The bone marrow will start to produce immune cells from the hematopoietic lineage at four to five months of gestation in human pregnancy. These foetal HSC will already produce a certain subset of γδ T cells, B lymphocyte 1a cells and macrophages before mobilization to their adult niche for lineage commitment (7, 9). Upon stimulation with ‘early acting cytokines’ a HSC will proliferate and differentiate into a multipotent progenitor or a lymphoid primed multipotent progenitor which subsequently give rise to a common lymphoid 2 precursor cell depending on their surroundings (10, 11).

BOX 1 A human embryo consists two weeks after fertilization of three layers, called ectoderm, endoderm and mesoderm. Ectoderm forms the nervous system and the exterior, endoderm forms among others the gastrointestinal tract and the mesoderm forms for example connective tissue and the cells of the immune system. In the fourth week of embryogenesis the flat tissue folds lateral and folds from head to toe, via which the endoderm is enclosed by the ectoderm and a foregut, midgut and hindgut are being formed, surrounded by mesoderm. At the end of the fourth week the liver starts growing also. From one of the pouches in the foregut, the thymus will be formed between third and fourth month during pregnancy. The loop of the midgut remains in contact with the yolk sac via the vitelline duct. In the same period that the definitive duodenum is formed, the midgut elongates and the hindgut becomes enlarged. When these changes in the pre-intestinal tract occur, the midgut migrates to the umbilical cord and returns into the embryo before the fourth month. During retraction the gut rotates to its final position. Throughout the whole gut enlargement of the surface is initiated via formation of crypts, villi and microvilli. The villi and microvilli start to develop after 9 to 10 weeks and at a later stage the colon loses its villi. After birth the intestine needs approximately one week to organize the lymph nodes with specific T- and B-cell regions and twenty-six weeks to acquire a tight epithelial barrier.

LYMPH NODE FORMATION

The lymphatic system originates mostly from the endothelial lining from the venous wall and some small organ specific lymph node structures come from an unidentified source (12). First Prospero homeobox 1, vascular endothelial cell growth factor receptor (VEGFr)3 and Podoplanin start lymphatic development (12, 13). Retinoic acid (RA), the active form of vitamin A (VitA), is needed to expand lymphatic precursor cells and also to initiate lymph node development (13). Retinyl hydrogenase (RALDH)2 converts retinaldehyde to RA and during lymph node development RALDH2 is one of the enzymes essential in initial cross talk

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 25 CHAPTER 2

between different cell types (14, 15). In absence of this enzyme proper lymphoid follicles are not developed. RA is important during embryogenesis, when axial patterning and organ formation take place (15, 16). RA is also involved in nerve development and expression of RALDH is found in nerve fibers near the lymph node origin, where RA can influence stromal cells to produce specific chemokines. Because RA is produced and used for different developmental processes, other factors should be present to direct lymph node development. One of these factors, C-X-C motif chemokine ligand (CXCL)13 is produced by mesenchymal stromal cells (MSC), and possibly produced by surrounding nerves (17, 18). Then lymphoid tissue inducer cells (also known as innate lymphoid cells) are attracted via C-X-C motif receptor CXCR)5 (17, 19). Subsequently these inducer cells activate lymphatic endothelial cells which retain cells needed for lymph node formation (17). After a small cluster of pre-T- and -B cells is formed, more cells arrive to this specific site for example via C-C motif chemokine receptor (CCR)7 and C-C motif chemokine ligand (CCL)19 and CCL21(18). When there are enough cells in the cluster, differentiation into lymph nodes occurs. Lymph nodes in the intestine are essential for priming of the whole immune system; therefore detailed development of the intestinal tract and intestinal immune tissues are described in Box 1.

Lymphotoxin α1β2, Receptor activator of nuclear factor kappa b (RANK) ligand (RANKl) and VEGF are important for the development of secondary lymphoid structures like the mesenteric lymph nodes (MLN) and Peyer’s patches (PP). MLN are the collecting lymph nodes for small intestine PP are the local small intestinal lymph nodes and Iliac lymph nodes are the collecting lymph nodes for the large intestine. These lymph structures develop very similar but there are some differences. For the development of PP cluster of differentiation (CD)11c+ dendritic like cells are required (20). In addition there are differences in essential growth or transcription factors which are necessary to develop lymph nodes or PP, like interleukin (IL)7, RA receptor-related orphan receptor (Ror)γt and RANK (20, 21). Using mice deficient for IL7 or Kit or both mouse models showed that IL7 is important for lymph node anlagen but not for PP development whereas both Kit and Il7 are important for MLN anlagen (21). Loss of RANKL results in abnormal or no lymph nodes, but small PP are still formed (20). Research with vitamin D (VitD) receptor (VDR) and 25-hydroxy vitamin D-1α-hydrolase (CYP27B1) knockout mice showed abnormal lymph node development due to the presence of more mature dendritic cells (DC) and increased DC trafficking (22).

Other smaller lymph structures, called tertiary lymphoid organs, can form in the intestine directly after birth (20). These lymph structures were not found in mice when the RA receptor was inhibited directly after birth (19). The influence of VitA and VitD on the development of lymphoid structures is a strong indication that nutritional components are of key importance at the base of infant’s immune system.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 26 IMMUNE DEVELOPMENT

Figure 2. Haematopoiesis/ immune cell development; From the hematopoietic stem cell (HSC) the different immune cell develop. MPP, multipotent progenitor; DC, dendritic cell; pDC, plasmacytoid DC; cDC, conventional DC; CD, cluster of differentiation; CCR2, C-C motif chemokine receptor type 2; CLP, common lymphoid progenitor; imm, immature; mat, mature; Breg, regulatory B cell; Ig, immunoglobulin; Th, T helper cell; Thf, T helper follicle cell; Treg, regulatory T cell; TR1, regulatory T cell type 1; IL, interleukin; TGFβ, transforming growth factor β.

IMMUNE CELL DEVELOPMENT

There is evidence that immunological cell types that arise before HSC differentiate into multipotent progenitor cells, can prosper after birth (7). These cell types, certain macrophages but possible also B1a B cells and γδ T cells, probably differentiate directly from MSC (7). The foetal lymphoid progenitor cells (LPC) differ from the adult LPCs. The foetal LPCs have the ability not only to differentiate into all lymphoid cell types but also into macrophages. There is still debate whether HSC differentiate step by step towards lineage commitment and loose multipotent characteristics or if HSC comprise transient multipotent stem cells which gradually acquire transcriptomic lineage priming in a continuous flow towards their final cell type fate and still retain some progenitor characteristics (23, 24). It is still uncertain if adult common lymphoid progenitor (CLP) are restricted to the division in myeloid and lymphoid progenitors or whether it is a gradual loss in lineage commitment (24, 25). The HSC survival and proliferation depends on many factors, recently the positive impact of the active form of VitD on HSC via CXCL8 in zebrafish was shown (26).

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 27 CHAPTER 2

T cell lineage maturation is extensively studied and straightforward after these cells become committed. During this process pro-T cells are able to orchestrate their commitment via surrounding signals and specific transcription factors but there is still debate about how gradual this commitment occurs (11, 23) Lineage specific T cell development and selection occurs in the thymus and is dependent on the Notch signalling pathway (11, 27). When Notch signalling is inhibited most lymphoid progenitor cells in the thymus will become B cells (27). Under normal circumstances a small portion of highly active B cells develop in the thymus, probably to regulate T cell development (27). Only naïve T cells, intermediately recognizing the Major Histocompatibility Complex (MHC)-self-peptide and expressing CD4 or CD8, are allowed to migrate to their target organ (28). CD4+ T cells form a subset of T cells called T helper (Th) cells and these cell types help to activate immature B cells and cytotoxic CD8+ T cells. Th cells need binding of CD4 and CD28 to become activated Th cells and dependent on co-stimulatory cytokines and chemokines, CD4+ T cells mature into different kinds of effector T cells, memory T cells or Treg cells [Box 02]. Many factors are important for the development of both B and T cells, IL7 is of uttermost importance for both lymphocyte lineages (29). Recently a human MSC - HSC coculture system was developed to specify necessary factors for development, surprisingly Fms like tyrosine kinase 3 (Flt3) was essential for human B cell development and IL7 was not (25). However, IL7 helps establish B cells, producing IgM, in the marginal zone of the spleen (30). Directly after birth maternal IL7 increases T cell production in the thymus and supports survival of T cells in other lymphoid structures in offspring, underlining the importance of IL7 during lymphogenesis (31).

BOX 2 Naïve CD4+ T cells are activated upon encountering an antigen presenting cell expressing MHC class II and the co-stimulatory signal CD28. They mature into effector, memory or regulatory T cells dependent on which cytokines are produced by their APC. Already in 1986 two T cell subsets were discovered by Mosmann and Cofmann and now many more a well characterized. For Th1 Tbet and Stat1 are the initiating transcription factors, producing IFNγ. Then Stat4 amplifies the Th1 response. Th2 differentiation is activated by IL4, which induces Stat6. Stat6 in turn will activate the transcription factor Gata3. For Th17 Rorγt is crucial, this is induced by TGFβ with IL6, subsequently Stat3 and IRF4 are expressed. Tregs are formed when RA is present activating the transcription factors Foxp3 and Stat5. After an infection is cleared most cells die by apoptosis, some cells become memory cells to establish a faster response upon re-infection (1).

The thymus is critically sensitive to malnutrition, with protein nutrition deficiency causing atrophy of the thymus (32). This suggests that the thymus is a putative target for early-life

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 28 IMMUNE DEVELOPMENT

programming effects. Most immune defence mechanisms are impaired in malnutrition, even in moderate nutritional deficiency. Protein-energy malnutrition is accompanied by deficiencies of micronutrients such as vitamins A, E, B6, C, or folate, zinc, iron, copper, and selenium. Rapid proliferating T cells are especially affected by the lack of essential nutrients. Severe and chronic malnutrition may even lead to thymus atrophy affecting the basis of our immune system. The potential of adding certain nutrients, like vitamin C, D, or E, at levels above recommended 2 dietary allowances (RDA) to the diet may improve immune function, is subject to increasing research.

The B cell lineage develops from HSC at the same sites as T cell development occurs (25). A major difference in lymphocyte development is that B cell development starts earlier, already in the pre-aorta gonad mesonephros region, which is also the most potent site for B cell development in the foetus (33). B cell lineage commitment occurs post-natally in the bone marrow, and immature B cells travel towards secondary lymphoid organs. When immature B cells are activated they start to produce immunoglobulins (Ig), soluble and cell surface products to neutralize pathogens. Follicular B2 cells respond to microbial infections, they will present processed bacterial peptide to their CD4+ T cell partner. Because of this interaction B cells will undergo isotype switching and mature into plasma cells secreting antibodies to clear the infection. Some of these mature cells become memory B cells in peripheral lymphoid organs. Marginal B cells respond to bacterial polysaccharides without T cell stimulation. Recently it has been found that regulatory B cells (Breg) can negatively influence the immune response via IL10, IL35 and transforming growth factor beta (TGFβ) secretion (34, 35). Immature and mature B cells and plasmablasts can differentiate into Breg via toll like receptors (TLR) or CD40 activation (34). These Breg can suppress inflammation via IL10 in an arthritis, colitis and experimental autoimmune encephalitis model (34). Also there are T cell independent B cell activators as mentioned above, but there is also T cell dependent B cell activation. At first the B cells will start with producing IgM and IgD. Upon stimulation they can switch towards an IgA, IgE or IgG, depending on their surroundings and seem to be influenced by dietary components as discussed later. Th1 type of CD4+ T cells can help B redirect to IgG2a in mice and IgG2 in humans. Th2 type of environment can induce switching towards IgE and IgG1 in mice and IgE and IgG4 in humans. At first the B cells will start with producing IgM and IgD. Upon stimulation they can switch towards an IgA, IgE or IgG, depending on their surroundings and seem to be influenced by dietary components as discussed later. Neonatal B cells are capable of switching to IgG1 and IgG3 during the first 2 years of life, but the switch to IgG2 and IgG3 is inadequate during this period. To compensate the lack of protection in foetus and new-born, microbe- specific maternal IgG antibodies move across the placental barrier to provide some vital protection. During the last trimester of pregnancy IgGs are transferred intrauterine to the infant, because new-borns memory T cells capable of generating IgG and isotype switching aren’t present yet. IgG1 and IgG4 are most effectively transported across the placenta compared to

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 29 CHAPTER 2

IgG3 and IgG2. Transfer of maternal antigen-specific IgG regulates the development of allergic airway inflammation early in life in a neonatal Fc region-dependent manner (36). This active transfer of IgGs from mother to child starts at week 17 and continues until birth. Moreover, just before birth, IgG levels of prenatal infants are even higher than levels present in the mother. The transfer and amount of pathogen specific IgGs is dependent on vaccinations and diseases the mother acquired during life. Many more factors influence the IgG transportation processes which are described in a review (37). IgA, IgD, IgM are the only known antibodies acquired after birth via breast feeding, covering the lack time during increasing antibody productions of infant’s immune system itself. When the infant can produce the different B cell antibodies, the activation, isotype switching and survival of B cells is for example under influence of B cell activating factor, TGFß1, IL6, IL7 and IL10 produced by PP stromal cells or mast cells (38, 39). Not much is known about the involvement of Ig free light chains (IgfLC) in humoral immune response early in life. However, the release of antigen specific IgfLC by B cells/ plasma cells like for example IgE may have implications for the health status of the new-born (40, 41). Furthermore, compromised immune status may result in enhanced production and secretion of IgfLC at the cost of other Ig (42).

Whereas T and B cell lineage experienced an in-depth research into their origin, the origin and differentiation of DC is not clear. When HSC differentiate into a multipotent progenitor they lose lymphoid potential. Then direct commitment to the DC lineage is a possibility or each progenitor lose their ability to differentiate into each myeloid cell type in a linear manner (43, 44). There is discussion about the existence of a common myeloid progenitor but it is not clear which specific myeloid cell types can evolve (11, 43, 44). There are three DC subtypes, conventional DC (cDC), plasmacytoid DC (pDC) and monocyte derived DC (mDC), whereas the cDC and pDC share a common progenitor. Microbial exposure is one of the key developing factors for DC (45) and is important in de novo generation of DC postnatally upon infection (43). Ftl3 is necessary for general development and expansion of pDC and cDC, for cDC development RA and granulocyte macrophage colony stimulating factor (GM-CSF) are also needed (46). For survival specific cytokines produced in the niche for DC are needed, like GM-CSF for cDC in the intestine and lymphotoxin α1β2 for cDC in the spleen (44). The first macrophages that colonize the intestine come from the embryonic yolk sac, these are replaced continuously after birth by bone marrow monocytes (47). The intestinal lamina propria (LP) contains mostly fractalkine positive macrophages sample the gut lumen with protrusions across the epithelial barrier to remove bacteria locally, they support the epithelial barrier function and can support Treg expansion (47, 48). The LP also has some CD103+ cDC which can have protrusions across the barrier (48). Besides the non-migrating DC population in the PP, there is a majority of DC which will migrate to specific sites after encountering an antigen. For example the intestinal LP derived CD103+CD11b-, CD103+CD11b+ and CD103-CD11b+ cDC subtype will migrate towards the MLN when they express CCR7 and the CD103+ DC is known for its capability to elicit a

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 30 IMMUNE DEVELOPMENT

regulatory response, in the presence of RA and TGFβ (47, 48). Recently these CD103/CD11b DC subtypes were subdivided with programmed death ligand 1 (PD-L1) marker (49).The CD11b- CD103+PD-L1high subgroup was especially effective in inducing Foxp3+ Treg in the presence of TGFβ (49). Upon VitA deficiency the CD103+CD11b+ cDC were shown to be reduced in the LP of the small intestine and colon (47). However, if other factors are present a Th1, Th2 or Th17 response can be induced by DC. Almost all intestinal DC information comes from mouse 2 models, but the DC population characterised for CD103 and CD11b are also found in humans (44). The activated T cell in the MLN will migrate to the site of inflammation for instance in the intestine to initiate a proper immune response (Figure 3). Intestinal development is already influenced during pregnancy by the amniotic fluid as intestinal epithelial cells of the foetus can react to components in the amniotic fluid by different receptor expression (50). During gestation the gut can readily react to its micro-environment, but still much is unknown about the influence of maternal status on infantile gut development.

Figure 3. Immune cell plasticity. Upon antigen encounter and T cell activation, naïve CD4+/ CD8- cells mature into several T helper (Th) subsets. Several factors including cytokines, described above in grey, determine the type of immune response, including Th1, Th2, Th17 and Treg. For example Th1 cell differentiation interleukin (IL)12, interferon gamma (IFNγ) and receptor activator of nuclear factor kappa B ligand (Rankl) are necessary (normal lines indicate stimulation). These cytokines help upregulate specific transcription factors like T box transcription factor (Tbet) and signal transducer and activator of transcription (Stat)1 and -4. These transcription factors activate certain genes to produce other cytokines like IL2,-10, -12, IFNγ and tumor necrosis factor alfa (TNFα) or specific receptors like c-c motif chemokine receptor type 5 (CCR5) and c-x-c motif chemokine receptor type 3 (CXCR3). These produced cytokines help inhibit the differentiation of other Th subsets (dotted lines indicate inhibition). RAR-related orphan receptor gamma t, Rorγt; retinoic acid, RA; forkhead box p3, Foxp3; cluster of differentiation, CD; programmed cell death protein 1 receptor, PD1; transforming growth factor beta, TGFβ.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 31 CHAPTER 2

An increasing number of studies have identified interesting links between early nutrition, epigenetic processes and disease development later life (4, 51). As the plasticity of growing and developing tissues shapes, the base of the responses to later challenges is established, therefore the exposures during early life may be critical. Folate deficiency during pregnancy is associated with increased risk for aberrant reprogramming of deoxyribonucleic acid (DNA) methylation inducing neural tube defects. Dietary folate intake can restore these deficiencies and neural tube defects. Folate is however not the only determinant of DNA methylation. Other methyl donor nutrients like betaine, vitamins B2, B6 and B12, and methionine, and choline, can also change DNA methylation status and therefore have an impact on development early in life (4, 52).

POSTNATAL IMMUNE DEVELOPMENT IN THE GUT

After birth a new-born changes from a sterile environment of the uterus into a world of constant challenges. In the first months of life new-borns are more dependent on their innate immunity and Ig derived from the mother as described before. IgG is already acquired intrauterine, but IgM and IgA are ingested via human milk to have an early defence against pathogens. Triggering the innate immunity repeatedly to microbial challenges after birth helps establishing a normal Th1 response postnatal. But the immune response is still less adequate compared to the adult response, possibly due to a general lower number of immune cells present. To build up an effective and balanced immune system interaction with germs and bacteria is needed (53). It is important that the immune balance is established properly because a deregulation early in life could trigger an adverse reaction leading to an allergic reaction. The gut is the main processing site of food and, here immune tolerance against new food particles is established. But an improper chemical barrier and weak mucosal barrier integrity will complicate acquiring immune tolerance. There are some factors positively influencing development of the gut and immune system during pregnancy and after birth, of which only few of them will be discussed in detail below.

INFLUENCE OF VITAMINS DURING AND AFTER PREGNANCY

VitA has been recognized for its importance in lymphoid structure formation during embryogenesis (13, 15). Sources of VitA are carrots, liver, sweet potatoes and butter. For example VitA supplementation before, during and after pregnancy using restricted dosages, results in long term positive effects on lung function which could be measured nine to thirteen years later (54). Shortly after the recognition of VitA essence during pregnancy, the teratogenic effect of excess VitA was shown. Malformations of the central nervous system, the eye and

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 32 IMMUNE DEVELOPMENT

the thymus were found due to excessive use of VitA (55). The teratogenic danger of VitA can be induced via extra prolonged ingestion of this vitamin because it is stored in fat cells in the liver. It was already mentioned that RA could increase the progression from HSC into myeloid cell lineage. When RA is lacking from the diet CD4+ and CD8+ T cells, IgA+ B cells and CD103+ DC aren’t found in the LP of the intestine, whereas CD4+ T cells were still present in lung tissue (47, 56). Stromal cells of intestinal epithelial layer and draining lymph nodes and DC and 2 macrophages in the LP induce high levels of RALDH expression postnatally. For this process the presence of VitA is crucial and there is also possible regulation from TLR signalling and cytokines (57, 58). They compared MLN with other lymph nodes at different postnatal weeks and found an increased expression of RALDH only in the MLN residing DC, almost exclusively in CD103+ MHC-II+ CD11c+ DC cells. VitA deficiency reduces this RALDH expression significantly (57). In addition, RA skews DC towards tolerance induction via instruction of Foxp3 Treg in the mother during pregnancy and in the infant just after birth in the presence of IL10 (59, 60). RA+ DC can also induce activated Th cells instead of Treg dependent on RA concentration and locally produced cytokines (58). VitA is also necessary for the intestinal homing properties of activated T cells, Treg and certain B cells. In vitro it was shown that naïve CD4+ cells induce gut homing receptors when they are activated by CD3 and CD28 together with RA, this combination simultaneously down regulates skin homing receptors (56). In vitro DC from the gut associated lymphoid tissue could also induce activated IgA secreting B cells more effectively when RA was present in the medium (61, 62). Not only RA is important for IgA secreting B cells, IL5 or IL6 are essential for IgA class switching and together with TGFβ RA increases IgA class switching and secretion in the B1 subpopulation (58, 63). There is also a positive feedback loop, IL6 stimulates IgA induction by DC and IL6 is induced by RA. In vivo lack of RA signalling abrogates antigen specific IgA responses in B cells after oral immunization (64). Synergistic effect on IgA induction could be obtained by the aforementioned cytokines and therefore immune protection of the infant could be achieved (60, 61). This indicates that also during early infancy the level of VitA is of importance for proper immune functioning, which may have profound consequences later in life.

An additional vitamin of interest for immune development is VitD. Worldwide there is no consensus on the healthy levels of VitD intake before and during pregnancy. Already during the first trimester of pregnancy VitD could be of importance for the development of the foetus.

CYP27B1, which converts VitD into its more active metabolite 1,25 (OH)2D3, can be detected in the placenta. There is some speculation about VitD deficiency and reduced fertility, this fertility problem may be caused by a reduced down regulation of the Th1 response by inadequate levels of VitD as well as circumstantial evidence pointing at a relation between gestational diabetes and preeclampsia in correlation to circulating VitD levels (65). In addition, VitD is important for the clearance of certain infections (66).

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 33 CHAPTER 2

VitD is linked to monocyte activation via activation of TLR2 and TLR1 (67). IL4 and interferon (IFN)γ are involved in the regulation of VitD expression (68). IL4 promotes expression of CYP24A1, which down regulates active VitD, and IFNγ positively influences the expression of CYP24B1 in monocytes. VitD positively influences the antimicrobial capacities of human macrophages and monocytes (69). It is suggested that VitD could skew the balance between Th1 and Th2 towards a Th2 phenotype directly or indirectly via monocyte activation, which however may be much more complex in vivo (70). Low VitD levels during pregnancy are suggested to be related in the development of food allergy (71). DC, B cells, T cells and activated macrophages have high levels of VitD converting enzymes and express the VDR in the nucleus. VitD can inhibit the development of immature DC via down regulation of the expression of co-stimulatory molecules (69). These VitD+ DC promote tolerance via upregulation of Treg and down regulation of T cell activation (69, 72). The Th cell population are also directly regulated by VitD, Th1 and Th17 are inhibited and for Th2 outcome of regulation probably depends on surrounding cytokines (69). Further down this signalling cascade is the indirect influence of VitD on B cells via Th cell regulation. VitD exerts a direct influence on B cells via favouring a regulatory phenotype expressing IL10 and interfering with plasma cell differentiation in vitro (69).

There is still debate about the amount of healthy VitD intake during pregnancy (73, 74). Indeed some studies show that VitD intake is associated with diminished wheeze, asthma and eczema in offspring (73). VitD influences lymphocytes, but to what extent this affects lymphocytes function later in life remains to be established. Increasing evidence from observational studies in infants at older ages, indicate that VitD insufficiency and deficiency might increase the risk of chronic diseases such as type 1 diabetes and multiple sclerosis. However, clear randomized trials on this association need to be conducted to confirm these findings. In the last decades, observations accumulate that VitD deficiency leads to more often and more serious respiratory infections than in individuals with sufficient VitD plasma levels. This illustrates the importance of the nutritional status early in life, to set the proper immune balance (75).

HUMAN MILK

It is well established that breastfeeding reduces the incidence of gastrointestinal and non-enteric infections in infants, due to its antimicrobial activity against several viruses, bacteria, and protozoa (76). In addition, it was shown that infants, breastfed for more than 4 months experienced significant reduced incidences of respiratory tract infection requiring hospitalization, as compared to infants who were not breastfed (77). Moreover, other studies showed that breastfeeding provides protection against urinary tract infections and otitis media and it reduces the development of inflammatory conditions like allergy (78), Crohn’s disease and ulcerative colitis (79). Apparently breast fed infants have a larger thymus, which could

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 34 IMMUNE DEVELOPMENT

possibly be regulated via IL7 (80). This could be beneficial because a larger thymus can have improved function during adolescence. This, moreover, emphasizes the diversity of activity and active components present in human breast milk. Although it is clear that allergy development is influenced by breast milk as well as atopy related disorders, still some controversy exists regarding the beneficial length of breastfeeding (81). The protective effects of human breast milk seem to persist at least during the first decade of life. 2

Prenatally the infant is supplied with maternal Ig, for the first protection against infections. It takes at least one year before an infant can produce about 60% of its IgG levels on its own. After birth the child can be supplied with essential IgG, IgM and IgA via human milk. For example, IgA is necessary as the first line of defence against microorganisms. This Ig also controls commensal (also called beneficial) bacteria all without activation of an inflammatory response. Because infants do not have an optimized Treg response and no memory cells, B cells cannot be directed to produce the right amount of Ig antibodies quickly as a first defence. IgM helps to eliminate the pathogen before IgGs are produced. As earlier mentioned IgM is normally produced by naive B cells before isotype switching occurs. The B cell response needs time after birth to be fully functional, so acquiring this component via human milk strengthens the first line of defence. Not only antibodies are factors present in human milk, other immune modulating components are present, including cytokines, NDO and PUFA, which are discussed in more detail below.

CYTOKINES

TGFβ and IL10 are held responsible for the induction of oral tolerance in the intestine (82). They educate the immune system locally not to respond to harmless antigens. In human milk TGFβ2 is the predominant isoform of the three existing mammalian isoforms. Addition of TGFβ2 to infant formula, can skew the Th2 allergic effector response towards Th1 in rat pups exposed to β-lactoglobulin, a cow’s milk allergy protein (83). In the adult immune system not TGFβ2 but TGFβ1 is the major player to establish tolerance to food. To guarantee the uptake of TGFβ in general, TGFβ receptors are abundantly expressed in the neonatal intestine, even the soluble TGFβ2 receptor is present in breast milk. TGFβ helps forming the intestinal epithelial barrier via exerting its effect on the reinforcement between the epithelial cells (84-86). The production of IgA is also positively influenced by TGFβ (87, 88). However it was found that TGFβ is not essential for tolerance induction when milk born-IgG antigen immune complexes are present (89). Moreover it is known that high levels of TGFβ together with RA skew the immune system towards a regulatory suppressive function, a low dose of TGFβ combined with IL6 or IL21 or IL23 will result in inflammatory Th17 activation instead of Treg upregulation (58, 90).

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 35 CHAPTER 2

IL10 is found in human milk, although a recent study found that IL10 is not always detectable in human milk (91). In the specifi c human milk samples where IL10 was not detectable, the highest concentrations of IgA were found (91). If there is a defect in either IL10 or the IL10 receptor, in mice as well as in humans, spontaneous infl ammation of the intestine occurs (92). IL10 is produced by leukocytes, natural killer cells, T cells, cells, monocytes, macrophages, dendritic cells, mast cells and some epithelial cells (92). Murine intestinal phagocytes, characterized high + by CX3CR1 CD11b , produce high levels of IL10, stimulate IgA secretion and limit intestinal infl ammation by enhancing the expansion of Foxp3+ Treg (93). Recently Barman et al. found the human phagocytic counterpart as CD163high CD160high cells (93). Foxp3+ Treg and type 1 Treg (Tr1) also produce IL10 to inhibit Th cell induction and maintain intestinal tolerance in the intestine (94). IL10 and TGFβ, both produced by Foxp3+ Treg, were found to be important in the prevention for or in oral immunotherapy against cow’s milk allergy (95, 96). However, IL10 was also found to enhance mast cell expansion, activity and survival in an ovalbumin allergy food model (97). IL10 also has a stimulatory effect on B cells and cytotoxic T cells (94). This indicates that these immune modulating components have pleiotropic functions, that are dependent on their time of expression, location, other cytokines or specifi c immune cells present.

NON-DIGESTIBLE OLIGOSACCHARIDES

There are different types of soluble dietary fi bres e.g. (hemi)cellulose, lignin, β-glucans, pectins, gums, inulin and oligofructose. In addition different NDO with specifi c properties are obtained or manufactured from natural sources (98). Human milk contains approximately 7-12 g/L oligosaccharides (99, 100). At least 130 different NDO have been isolated from human milk and the two main categories are neutral and acidic oligosaccharides (101, 102). These oligosaccharides are non-digestible carbohydrates that have many different properties and are believed to act on the microbiota in the gut (100, 103-106). Due to physicochemical properties of non-digestible carbohydrates the absorption of minerals and fecal consistency improves. Some of these have specifi c properties and can be used as prebiotics as a dietary supplement. Prebiotics are defi ned as “a selectively fermented ingredient that allows specifi c changes, both in the composition and/or activity in the gastrointestinal microfl ora that confers benefi ts upon host wellbeing and health” (107). Prebiotics enhance defensive mechanisms of the host by stimulation of growth of bifi dobacteria and lactobacilli. As the intestinal microbiota plays a critical role in the establishment and maintenance of healthy immune responses, the delayed colonisation of the infant gut with commensal bacteria are suggested to be a risk factor for the development of immune mediated chronic disorders such as allergic and autoimmune diseases. Short-chain fatty acids, released by these bacteria upon fermentation of prebiotics, are essential nutrients for intestinal epithelial cells and support gut function (108-110). In vivo and in vitro studies have shown benefi cial effects of prebiotics on the innate as well as the

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 36 IMMUNE DEVELOPMENT

adaptive immune system (111). Short-chain Galacto-oligosaccharides (scGOS), long-chain Fructo-oligosaccharides (lcFOS) and pectin-derived Acidic-oligosaccharides (pAOS) are some examples of NDO that mimic the functionality and molecular size distribution of human milk oligosaccharides (Figure 4).

Figure 4. Molecular structures of A) short-chain galacto-oligosaccharides (scGOS), B) long-chain fructo- oligosaccharides (lcFOS) and C) pectin-derived Acidic-oligosaccharides (pAOS).

The effects of scGOS and lcFOS (9:1) have been studied in a murine vaccination model (112), in an allergic asthma model (113) and a cow’s milk allergy model (114). The response to vaccination of mice fed the scGOS/lcFOS diet was signifi cant enhanced, as well as the fecal bifi dobacteria and lactobacilli proportions. pAOS enhanced the murine vaccination response and the combination (scGOS/lcFOS/pAOS) was even more effective (115). To stimulate the entire microbiota a great variation in the oligosaccharide structures will exert this effect, as human milk oligosaccharides comprise of many different oligosaccharides. The pAOS are not only another group of oligosaccharides, they act very specifi cally to their acidity. Based on this they are able to interact with surfaces and might prevent the adhesion of pathogens on the intestinal epithelium. In infants pAOS showed no difference in stool characteristics, pH, growth, crying, vomiting and regurgitation patterns as compared to control formula. In addition pAOS alone

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 37 CHAPTER 2

did not affect intestinal microecology (116). Furthermore, systemic Th1 dependent immune responses were enhanced using the prebiotics without inducing autoimmunity, as Th1 is low in newborn infants. In addition, in a murine model for cow’s milk allergy dietary intervention with scGOS/lcFOS with or without pAOS showed significant decrease of the allergic response and increased specific IgG2a levels (114, 117, 118).

Dietary intervention in mice with scGOS/lcFOS/pAOS reduces the development of an acute allergic response upon antigen challenge, although specific Ig levels remain high.Ex vivo depletion of CD25+ Treg abrogated the diminished acute allergic response, combined with adoptive transfer studies, imply crucial involvement of antigen specific CD25+ Treg cells in the suppression of the allergic effector response (119-121). Additionally, clinical trials have been performed with the scGOS/lcFOS mixture in children at high risk for allergies. A reduction in the incidence of atopic dermatitis (AD) (122) and the incidence of allergic manifestations during the first 6 months of life (123) was observed, furthermore this reduction lasted at least until 2 years of age (124).

The scGOS/lcFOS/pAOS mixture could reduce the incidence of AD in healthy not at risk children in a multicenter trial (125). In another clinical study, using the scGOS/lcFOS mixture, fecal secretory IgA was increased in healthy infants (126). Recently oligosaccharides had no preventive effect when supplied with an hydrolysed formula to children with AD (127). There is cumulative evidence in healthy infants that prebiotic mixtures might beneficially affect the host in both Th1 as well as Th2 prone settings as it might prevent food allergy (Th2 driven) and enhances the vaccination response (Th1 driven). Although it is believed that prebiotics exert their effect via stimulation of growth of selective bacterial species that beneficially improve host health, there is debate about this mechanism and there are potentially microbiota-independent mechanisms as well (100, 111, 128). Epithelial cells can transport scGOS, lcFOS and pAOS across from apical to the basolateral side (129), this illustrates that besides a prebiotic immune modulating effect the oligosaccharides also come in direct contact with immune cells themselves, making it possible to act directly on the immune cells (130-133).

POLY UNSATURATED FATTY ACIDS

Omega-3 and omega-6 PUFA are essential for humans and have to be provided via the diet especially found in seafood. PUFA are incorporated into the cellular membrane and are eicosanoid precursors hereby affecting the immune response. In this regard n-3 long chain (LC)-PUFA eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) are regarded to be anti-inflammatory while n-6 LCPUFA, like linoleic and arachidonic acid, can boost inflammatory responses. Arachidonic acid, for example can be converted into 2 series prostaglandins, these

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 38 IMMUNE DEVELOPMENT

are able to increase IL4 and decrease IFNγ, and 4 series of leukotrienes, which are able to induce endothelial permeability and production of inflammatory cytokines (134).

In westernized countries the n-3 fatty acid ingestion is no longer favoured over n-6 PUFA which may have implications for immune homeostasis. It has been stated that the original diet of the human race had an n-6 : n-3 ratio of approximately 1:1 and that this has changed over the 2 past decades to a ratio of at least 15:1 (135). This shift is suggested to be explanatory for the increased incidence of cardiovascular diseases, chronic inflammatory diseases, obesity and also allergic diseases. Therefore, pre-clinical as well as clinical trials are performed in order to investigate this hypothesis. It has been reviewed that n-3 LCPUFA intake by the mother reduces the incidence of allergic disorders, e.g. AD, less sensitization to egg in the clinical trials in children (136, 137). Two studies by Dunstan et al. show reduced sensitization to egg and less AD in offspring when pregnant woman were supplemented with fish oil, which contains EPA and DHA (138, 139). They found a positive influence of n-3 LCPUFA on Tgfβ mRNA levels in maternal peripheral blood and cord blood. This could imply a regulatory function of n-3 LCPUFA, but mRNA still needs to be processed towards an active component. Recently, it was shown that maternal n-3 LCPUFA intake decreased the risk of food allergy and IgE-associated eczema in children at risk for allergy (140). This shift in the decrease of allergy can be explained since n-3 LCPUFA is incorporated in the membrane of immune cells at the cost of n-6 LCPUFA as arachidonic acid, skewing towards a less inflammatory cytokine surrounding. Intake of fatty fish once a week for 6 months after birth reduces asthma and wheeze up to 4 years later (141). In contrast, in a clinical study by Almqvist et al., (142) it is shown that supplementation of n-3 PUFA, starting at a maximum of six months of age, did not prevent children with a family history of asthma from developing atopy, eczema nor asthma at the age of 5 years. Hence discrepancy on effects of n-3 PUFA in prevention of allergic disease exists. There are also discrepancies between studies in the preparation of fish oil, influencing EPA and DHA content (143). Of concern is the dosage in human studies, this is often much lower than used in animal studies, which makes it difficult to extrapolate the outcome and could explain differences observed (144). However, interventions using dietary factors like LCPUFAs are under-explored and that there is a need for additional research (144). One of the strategies that are proposed is the use of selected LCPUFA in the formula feeding of young children at high risk for allergies (145). Another strategy could be to supplement a novel LCPUFA during pregnancy; There is a possible link between maternal LCPUFA status and preterm birth (146). But the effect of LCPUFA on preterm infants did not show clear benefits nor harm when formula milk was supplemented with LCPUFA (147). LCPUFA maybe has its largest positive effect on an infant health status when primary immune responses are still developing. In addition novel synthetic LCPUFA may reveal to be potent in the reduction of the allergic burden (143).

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 39 CHAPTER 2

CONSEQUENCES LATER IN LIFE

During the different phases of life, several nutritional factors influence our immune system and immune responsiveness in collaboration with endogenous immune modulating mediators like humoral factors, lipids and oligosaccharides. A vast amount of literature is available on the role that nutrients and human milk (as reviewed in this article), have on the development of the immune system; a lot less is known about the exact requirements in the phases thereafter, in toddlers, during adolescence, and in the later stages of life. Moreover, as no single biomarker exists which is able to determine a proper functioning immune system, it is almost impossible to describe completely the developmental status of the immune system and the important influences of nutrition. Although each phase in life puts specific requirements on nutrition, no clear statement can be made based on literature as to what the exact dietary requirements are to fully support the immune system during these stages in life. But food as a beneficial dietary component is currently under very active scientific investigation, so more information about nutrition will be available soon.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 40 IMMUNE DEVELOPMENT

REFERENCES

1. Fang D, Zhu J. Dynamic balance between 11. Rothenberg EV, Kueh HY, Yui MA, Zhang JA. master transcription factors determines Hematopoiesis and T-cell specification as a the fates and functions of CD4 T cell and model developmental system. Immunol Rev. innate lymphoid cell subsets. J Exp Med. 2016;271(1):72-97. 2017;214(7):1861-76. 12. Escobedo N, Oliver G. Lymphangiogenesis: 2. McDade TW, Beck MA, Kuzawa C, Adair Origin, Specification, and Cell Fate 2 LS. Prenatal undernutrition, postnatal Determination. Annu Rev Cell Dev Biol. environments, and antibody response to 2016;32:677-91. vaccination in adolescence. Am J Clin Nutr. 13. Betterman KL, Harvey NL. The lymphatic 2001;74(4):543-8. vasculature: development and role in shaping 3. Walker RW, Clemente JC, Peter I, Loos immunity. Immunol Rev. 2016;271(1):276-92. RJF. The prenatal gut microbiome: are we 14. van de Pavert SA, Mebius RE. New insights colonized with bacteria in utero? Pediatr into the development of lymphoid tissues. Obes. 2017;12 Suppl 1:3-17. Nat Rev Immunol. 2010;10(9):664-74. 4. Berti C, Agostoni C, Davanzo R, Hypponen 15. Metzler MA, Sandell LL. Enzymatic E, Isolauri E, Meltzer HM, et al. Early-life Metabolism of Vitamin A in Developing nutritional exposures and lifelong health: Vertebrate Embryos. Nutrients. 2016;8(12). immediate and long-lasting impacts of 16. Campo-Paysaa F, Marletaz F, Laudet V, probiotics, vitamin D, and breastfeeding. Schubert M. Retinoic acid signaling in Nutr Rev. 2017;75(2):83-97. development: tissue-specific functions 5. Medvinsky A, Rybtsov S, Taoudi S. Embryonic and evolutionary origins. Genesis. origin of the adult hematopoietic system: 2008;46(11):640-56. advances and questions. Development. 17. Onder L, Morbe U, Pikor N, Novkovic M, 2011;138(6):1017-31. Cheng HW, Hehlgans T, et al. Lymphatic 6. Mazo IB, Massberg S, von Andrian UH. Endothelial Cells Control Initiation of Hematopoietic stem and progenitor Lymph Node Organogenesis. Immunity. cell trafficking. Trends Immunol. 2017;47(1):80-92 e4. 2011;32(10):493-503. 18. Brendolan A, Caamano JH. Mesenchymal 7. Hadland B, Yoshimoto M. Many layers of cell differentiation during lymph node embryonic hematopoiesis: new insights organogenesis. Front Immunol. 2012;3:381. into B-cell ontogeny and the origin of 19. Goverse G, Labao-Almeida C, Ferreira hematopoietic stem cells. Exp Hematol. M, Molenaar R, Wahlen S, Konijn T, et 2018;60:1-9. al. Vitamin A Controls the Presence of 8. Ivanovs A, Rybtsov S, Ng ES, Stanley RORgamma+ Innate Lymphoid Cells and EG, Elefanty AG, Medvinsky A. Human Lymphoid Tissue in the Small Intestine. J haematopoietic stem cell development: Immunol. 2016;196(12):5148-55. from the embryo to the dish. Development. 20. Buettner M, Lochner M. Development and 2017;144(13):2323-37. Function of Secondary and Tertiary Lymphoid 9. Geissmann F, Mass E. A stratified myeloid Organs in the Small Intestine and the Colon. system, the challenge of understanding Front Immunol. 2016;7:342. macrophage diversity. Semin Immunol. 21. Chappaz S, Gartner C, Rodewald HR, Finke 2015;27(6):353-6. D. Kit ligand and Il7 differentially regulate 10. Grassinger J, Haylock DN, Williams B, Peyer’s patch and lymph node development. Olsen GH, Nilsson SK. Phenotypically J Immunol. 2010;185(6):3514-9. identical hemopoietic stem cells isolated 22. Hewison M. Vitamin D and innate and adaptive from different regions of bone marrow immunity. Vitam Horm. 2011;86:23-62. have different biologic potential. Blood. 2010;116(17):3185-96. 23. Velten L, Haas SF, Raffel S, Blaszkiewicz S, Islam S, Hennig BP, et al. Human

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 41 CHAPTER 2

haematopoietic stem cell lineage 34. Rosser EC, Mauri C. Regulatory B cells: commitment is a continuous process. Nat origin, phenotype, and function. Immunity. Cell Biol. 2017;19(4):271-81. 2015;42(4):607-12. 24. Chi AW, Bell JJ, Zlotoff DA, Bhandoola A. 35. Chen X, Hao S, Zhao Z, Liu J, Shao Q, Wang Untangling the T branch of the hematopoiesis F, et al. Interleukin 35: Inhibitory regulator in tree. Curr Opin Immunol. 2009;21(2):121-6. monocyte-derived dendritic cell maturation 25. Ichii M, Oritani K, Kanakura Y. Early B and activation. Cytokine. 2018;108:43-52. lymphocyte development: Similarities and 36. Nakata K, Kobayashi K, Ishikawa Y, differences in human and mouse. World J Yamamoto M, Funada Y, Kotani Y, et al. Stem Cells. 2014;6(4):421-31. The transfer of maternal antigen-specific 26. Cortes M, Chen MJ, Stachura DL, Liu SY, IgG regulates the development of allergic Kwan W, Wright F, et al. Developmental airway inflammation early in life in an FcRn- Vitamin D Availability Impacts Hematopoietic dependent manner. Biochem Biophys Res Stem Cell Production. Cell Rep. Commun. 2010;395(2):238-43. 2016;17(2):458-68. 37. Chucri TM, Monteiro JM, Lima AR, Salvadori 27. Xiao S, Zhang W, Manley NR. Thymic B cell ML, Kfoury JR, Jr., Miglino MA. A review of development is controlled by the B potential immune transfer by the placenta. J Reprod of progenitors via both hematopoietic- Immunol. 2010;87(1-2):14-20. intrinsic and thymic microenvironment- 38. Vale AM, Schroeder HW, Jr. Clinical intrinsic regulatory mechanisms. PLoS One. consequences of defects in B-cell 2018;13(2):e0193189. development. J Allergy Clin Immunol. 28. Singer A, Adoro S, Park JH. Lineage fate 2010;125(4):778-87. and intense debate: myths, models and 39. Merluzzi S, Frossi B, Gri G, Parusso S, mechanisms of CD4- versus CD8-lineage Tripodo C, Pucillo C. Mast cells enhance choice. Nat Rev Immunol. 2008;8(10):788- proliferation of B lymphocytes and drive 801. their differentiation toward IgA-secreting 29. Kang J, Der SD. Cytokine functions in the plasma cells. Blood. 2010;115(14):2810-7. formative stages of a lymphocyte’s life. Curr 40. Redegeld FA, van der Heijden MW, Kool M, Opin Immunol. 2004;16(2):180-90. Heijdra BM, Garssen J, Kraneveld AD, et 30. Willems L, Li S, Rutgeerts O, Lenaerts C, al. Immunoglobulin-free light chains elicit Waer M, Billiau AD. IL-7 is required for immediate hypersensitivity-like responses. the development of the intrinsic function Nat Med. 2002;8(7):694-701. of marginal zone B cells and the marginal 41. Schouten B, van Esch BC, van Thuijl zone microenvironment. J Immunol. AO, Blokhuis BR, Groot Kormelink T, 2011;187(7):3587-94. Hofman GA, et al. Contribution of IgE and 31. Aspinall R, Prentice AM, Ngom PT. immunoglobulin free light chain in the Interleukin 7 from maternal milk crosses allergic reaction to cow’s milk proteins. J the intestinal barrier and modulates T-cell Allergy Clin Immunol. 2010;125(6):1308-14. development in offspring. PLoS One. 42. van Esch BC, Schouten B, Blokhuis 2011;6(6):e20812. BR, Hofman GA, Boon L, Garssen J, 32. Savino W, Dardenne M. Nutritional et al. Depletion of CD4+CD25+ T cells imbalances and infections affect the thymus: switches the whey-allergic response from consequences on T-cell-mediated immune immunoglobulin E- to immunoglobulin free responses. Proc Nutr Soc. 2010;69(4):636- light chain-dependent. Clin Exp Allergy. 43. 2010;40(9):1414-21. 33. Dorshkind K, Montecino-Rodriguez E. Fetal 43. Yanez A, Coetzee SG, Olsson A, Muench DE, B-cell lymphopoiesis and the emergence Berman BP, Hazelett DJ, et al. Granulocyte- of B-1-cell potential. Nat Rev Immunol. Monocyte Progenitors and Monocyte- 2007;7(3):213-9. Dendritic Cell Progenitors Independently Produce Functionally Distinct Monocytes. Immunity. 2017;47(5):890-902 e4.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 42 IMMUNE DEVELOPMENT

44. Sichien D, Lambrecht BN, Guilliams M, 55. Gutierrez-Mazariegos J, Theodosiou M, Scott CL. Development of conventional Campo-Paysaa F, Schubert M. Vitamin A: A dendritic cells: from common bone marrow multifunctional tool for development. Semin progenitors to multiple subsets in peripheral Cell Dev Biol. 2011;22(6):603-10. tissues. Mucosal Immunol. 2017;10(4):831- 56. Iwata M. Retinoic acid production by 44. intestinal dendritic cells and its role in T-cell 45. Plantinga TS, van Maren WW, van trafficking. Semin Immunol. 2009;21(1):8- Bergenhenegouwen J, Hameetman M, 13. 2 Nierkens S, Jacobs C, et al. Differential 57. Molenaar R, Knippenberg M, Goverse Toll-like receptor recognition and induction G, Olivier BJ, de Vos AF, O’Toole T, et al. of cytokine profile by Bifidobacterium breve Expression of retinaldehyde dehydrogenase and Lactobacillus strains of probiotics. Clin enzymes in mucosal dendritic cells and Vaccine Immunol. 2011;18(4):621-8. gut-draining lymph node stromal cells is 46. Zeng R, Bscheider M, Lahl K, Lee M, controlled by dietary vitamin A. J Immunol. Butcher EC. Generation and transcriptional 2011;186(4):1934-42. programming of intestinal dendritic cells: 58. Erkelens MN, Mebius RE. Retinoic Acid and essential role of retinoic acid. Mucosal Immune Homeostasis: A Balancing Act. Immunol. 2016;9(1):183-93. Trends Immunol. 2017;38(3):168-80. 47. Gross M, Salame TM, Jung S. Guardians of 59. Coombes JL, Siddiqui KR, Arancibia- the Gut - Murine Intestinal Macrophages and Carcamo CV, Hall J, Sun CM, Belkaid Y, et Dendritic Cells. Front Immunol. 2015;6:254. al. A functionally specialized population 48. Cerovic V, Bain CC, Mowat AM, Milling SW. of mucosal CD103+ DCs induces Foxp3+ Intestinal macrophages and dendritic cells: regulatory T cells via a TGF-beta and retinoic what’s the difference? Trends Immunol. acid-dependent mechanism. J Exp Med. 2014;35(6):270-7. 2007;204(8):1757-64. 49. Shiokawa A, Kotaki R, Takano T, Nakajima- 60. Duriancik DM, Lackey DE, Hoag KA. Vitamin Adachi H, Hachimura S. Mesenteric lymph A as a regulator of antigen presenting cells. node CD11b(-) CD103(+) PD-L1(High) J Nutr. 2010;140(8):1395-9. dendritic cells highly induce regulatory T 61. Mora JR, von Andrian UH. Role of retinoic cells. Immunology. 2017;152(1):52-64. acid in the imprinting of gut-homing 50. Drozdowski LA, Clandinin T, Thomson AB. IgA-secreting cells. Semin Immunol. Ontogeny, growth and development of the 2009;21(1):28-35. small intestine: Understanding pediatric 62. Gloudemans AK, Plantinga M, Guilliams M, gastroenterology. World J Gastroenterol. Willart MA, Ozir-Fazalalikhan A, van der Ham 2010;16(7):787-99. A, et al. The mucosal adjuvant cholera toxin 51. Boehm G, Moro G. Structural and functional B instructs non-mucosal dendritic cells to aspects of prebiotics used in infant nutrition. promote IgA production via retinoic acid and J Nutr. 2008;138(9):1818S-28S. TGF-beta. PLoS One. 2013;8(3):e59822. 52. Niculescu MD, Lupu DS. Nutritional 63. Roy B, Brennecke AM, Agarwal S, Krey M, influence on epigenetics and effects on Duber S, Weiss S. An intrinsic propensity of longevity. Curr Opin Clin Nutr Metab Care. murine peritoneal B1b cells to switch to IgA 2010;14(1):35-40. in presence of TGF-beta and retinoic acid. 53. van’t Land B, Schijf MA, Martin R, Garssen PLoS One. 2013;8(12):e82121. J, van Bleek GM. Influencing mucosal 64. Pantazi E, Marks E, Stolarczyk E, Lycke N, homeostasis and immune responsiveness: Noelle RJ, Elgueta R. Cutting Edge: Retinoic The impact of nutrition and pharmaceuticals. Acid Signaling in B Cells Is Essential for Oral Eur J Pharmacol. 2011;668 Suppl 1:S101-7. Immunization and Microflora Composition. J 54. Checkley W, West KP, Jr., Wise RA, Baldwin Immunol. 2015;195(4):1368-71. MR, Wu L, LeClerq SC, et al. Maternal vitamin 65. Dror DK, Allen LH. Vitamin D inadequacy A supplementation and lung function in in pregnancy: biology, outcomes, and offspring. N Engl J Med. 2010;362(19):1784- interventions. Nutr Rev. 2010;68(8):465-77. 94.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 43 CHAPTER 2

66. Hoyer-Hansen M, Nordbrandt SP, Jaattela 78. Fiocchi A, Martelli A, De Chiara A, Moro G, M. Autophagy as a basis for the health- Warm A, Terracciano L. Primary dietary promoting effects of vitamin D. Trends Mol prevention of food allergy. Ann Allergy Med. 2010;16(7):295-302. Asthma Immunol. 2003;91(1):3-12; quiz -5, 67. Lagishetty V, Liu NQ, Hewison M. Vitamin D 91. metabolism and innate immunity. Mol Cell 79. Hanson LA. Session 1: Feeding and infant Endocrinol. 2011;347(1-2):97-105. development breast-feeding and immune 68. Edfeldt K, Liu PT, Chun R, Fabri M, Schenk function. Proc Nutr Soc. 2007;66(3):384-96. M, Wheelwright M, et al. T-cell cytokines 80. Vieira Borba V, Sharif K, Shoenfeld differentially control human monocyte Y. Breastfeeding and autoimmunity: antimicrobial responses by regulating Programing health from the beginning. Am vitamin D metabolism. Proc Natl Acad Sci U J Reprod Immunol. 2018;79(1). S A. 2010;107(52):22593-8. 81. van Odijk J, Kull I, Borres MP, Brandtzaeg P, 69. Vanherwegen AS, Gysemans C, Mathieu C. Edberg U, Hanson LA, et al. Breastfeeding Vitamin D endocrinology on the cross-road and allergic disease: a multidisciplinary between immunity and metabolism. Mol review of the literature (1966-2001) on Cell Endocrinol. 2017;453:52-67. the mode of early feeding in infancy and 70. Adams JS, Hewison M. Unexpected its impact on later atopic manifestations. actions of vitamin D: new perspectives Allergy. 2003;58(9):833-43. on the regulation of innate and adaptive 82. du Pre MF, Samsom JN. Adaptive T-cell immunity. Nat Clin Pract Endocrinol Metab. responses regulating oral tolerance to 2008;4(2):80-90. protein antigen. Allergy. 2010;66(4):478-90. 71. Nwaru BI, Ahonen S, Kaila M, Erkkola M, 83. Penttila IA. Milk-derived transforming Haapala AM, Kronberg-Kippila C, et al. growth factor-beta and the infant immune Maternal diet during pregnancy and allergic response. J Pediatr. 2009;156(2 Suppl):S21-5. sensitization in the offspring by 5 yrs of age: 84. Hering NA, Schulzke JD. Therapeutic options a prospective cohort study. Pediatr Allergy to modulate barrier defects in inflammatory Immunol. 2010;21(1 Pt 1):29-37. bowel disease. Dig Dis. 2009;27(4):450-4. 72. Bscheider M, Butcher EC. Vitamin D 85. van’t Land B, Meijer HP, Frerichs J, Koetsier immunoregulation through dendritic cells. M, Jager D, Smeets RL, et al. Transforming Immunology. 2016;148(3):227-36. Growth Factor-beta2 protects the small 73. West CE, Videky DJ, Prescott SL. Role of diet intestine during methotrexate treatment in in the development of immune tolerance in rats possibly by reducing stem cell cycling. the context of allergic disease. Curr Opin Br J Cancer. 2002;87(1):113-8. Pediatr. 2010;22(5):635-41. 86. Malhotra N, Kang J. SMAD regulatory 74. Bouillon R. Comparative analysis of networks construct a balanced immune nutritional guidelines for vitamin D. Nat Rev system. Immunology. 2013;139(1):1-10. Endocrinol. 2017;13(8):466-79. 87. Borsutzky S, Cazac BB, Roes J, Guzman 75. Karatekin G, Kaya A, Salihoglu O, Balci H, CA. TGF-beta receptor signaling is critical Nuhoglu A. Association of subclinical vitamin for mucosal IgA responses. J Immunol. D deficiency in newborns with acute lower 2004;173(5):3305-9. respiratory infection and their mothers. Eur 88. Cazac BB, Roes J. TGF-beta receptor J Clin Nutr. 2009;63(4):473-7. controls B cell responsiveness and induction 76. Chirico G, Marzollo R, Cortinovis S, Fonte of IgA in vivo. Immunity. 2000;13(4):443-51. C, Gasparoni A. Antiinfective properties of 89. Verhasselt V, Milcent V, Cazareth J, Kanda A, human milk. J Nutr. 2008;138(9):1801S-6S. Fleury S, Dombrowicz D, et al. Breast milk- 77. Bachrach VR, Schwarz E, Bachrach LR. mediated transfer of an antigen induces Breastfeeding and the risk of hospitalization tolerance and protection from allergic for respiratory disease in infancy: a meta- asthma. Nat Med. 2008;14(2):170-5. analysis. Arch Pediatr Adolesc Med. 2003;157(3):237-43.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 44 IMMUNE DEVELOPMENT

90. Konkel JE, Chen W. Balancing acts: the role 100. Akkerman R, Faas MM, de Vos P. Non- of TGF-beta in the mucosal immune system. digestible carbohydrates in infant formula as Trends Mol Med. 2011;17(11):668-76. substitution for human milk oligosaccharide 91. Ruiz L, Espinosa-Martos I, Garcia-Carral C, functions: Effects on microbiota and gut Manzano S, McGuire MK, Meehan CL, et al. maturation. Crit Rev Food Sci Nutr. 2018:1- What’s Normal? Immune Profiling of Human 12. Milk from Healthy Women Living in Different 101. Greer FR, Sicherer SH, Burks AW. Effects Geographical and Socioeconomic Settings. of early nutritional interventions on the Front Immunol. 2017;8:696. development of atopic disease in infants 2 92. Shouval DS, Ouahed J, Biswas A, Goettel and children: the role of maternal dietary JA, Horwitz BH, Klein C, et al. Interleukin restriction, breastfeeding, timing of 10 receptor signaling: master regulator of introduction of complementary foods, intestinal mucosal homeostasis in mice and and hydrolyzed formulas. Pediatrics. humans. Adv Immunol. 2014;122:177-210. 2008;121(1):183-91. 93. Barman S, Kayama H, Okuzaki D, Ogino T, 102. Lara-Villoslada F, Olivares M, Sierra S, Osawa H, Matsuno H, et al. Identification of Rodriguez JM, Boza J, Xaus J. Beneficial a human intestinal myeloid cell subset that effects of probiotic bacteria isolated from regulates gut homeostasis. Int Immunol. breast milk. Br J Nutr. 2007;98 Suppl 1:S96- 2016;28(11):533-45. 100. 94. Geginat J, Larghi P, Paroni M, Nizzoli G, 103. Halken S, Hansen KS, Jacobsen HP, Penatti A, Pagani M, et al. The light and the Estmann A, Faelling AE, Hansen LG, et dark sides of Interleukin-10 in immune- al. Comparison of a partially hydrolyzed mediated diseases and cancer. Cytokine infant formula with two extensively Growth Factor Rev. 2016;30:87-93. hydrolyzed formulas for allergy prevention: a prospective, randomized study. Pediatr 95. Kerperien J, Veening-Griffioen D, ehkampW Allergy Immunol. 2000;11(3):149-61. T, van Esch B, Hofman GA, Cornelissen P, et al. IL-10r or TGF-beta Neutralization 104. Hill DJ, Murch SH, Rafferty K, Wallis P, Abrogates the Protective Effect of a Specific Green CJ. The efficacy of amino acid-based Nondigestible Oligosaccharide Mixture in formulas in relieving the symptoms of cow’s Cow-Milk-Allergic Mice. J Nutr. 2018. milk allergy: a systematic review. Clin Exp Allergy. 2007;37(6):808-22. 96. Smaldini PL, Orsini Delgado ML, Fossati CA, Docena GH. Orally-Induced Intestinal CD4+ 105. Host A, Koletzko B, Dreborg S, Muraro A, CD25+ FoxP3+ Treg Controlled Undesired Wahn U, Aggett P, et al. Dietary products Responses towards Oral Antigens and used in infants for treatment and prevention Effectively Dampened Food Allergic of food allergy. Joint Statement of the Reactions. PLoS One. 2015;10(10):e0141116. European Society for Paediatric Allergology and Clinical Immunology (ESPACI) 97. Polukort SH, Rovatti J, Carlson L, Committee on Hypoallergenic Formulas Thompson C, Ser-Dolansky J, Kinney and the European Society for Paediatric SR, et al. IL-10 Enhances IgE-Mediated Gastroenterology, Hepatology and Nutrition Mast Cell Responses and Is Essential for (ESPGHAN) Committee on Nutrition. Arch the Development of Experimental Food Dis Child. 1999;81(1):80-4. Allergy in IL-10-Deficient Mice. J Immunol. 2016;196(12):4865-76. 106. Niggemann B, Beyer K. Diagnosis of food allergy in children: toward a standardization 98. Calder PC, Krauss-Etschmann S, de Jong of food challenge. J Pediatr Gastroenterol EC, Dupont C, Frick JS, Frokiaer H, et al. Nutr. 2007;45(4):399-404. Early nutrition and immunity - progress and perspectives. Br J Nutr. 2006;96(4):774-90. 107. Host A, Halken S. Hypoallergenic formulas- -when, to whom and how long: after more 99. Hoppu U, Kalliomaki M, Laiho K, Isolauri E. than 15 years we know the right indication! Breast milk--immunomodulatory signals Allergy. 2004;59 Suppl 78:45-52. against allergic diseases. Allergy. 2001;56 Suppl 67:23-6. 108. Boehm G, Stahl B. Oligosaccharides from milk. J Nutr. 2007;137(3 Suppl 2):847S-9S.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 45 CHAPTER 2

109. Gibson GR, Roberfroid MB. Dietary the suppression of cow milk allergy in mice. modulation of the human colonic microbiota: J Nutr. 2010;140(4):835-41. introducing the concept of prebiotics. J Nutr. 120. Schouten B, van Esch BC, Hofman GA, 1995;125(6):1401-12. de Kivit S, Boon L, Knippels LM, et al. A 110. Nakhla T, Fu D, Zopf D, Brodsky NL, Hurt H. potential role for CD25+ regulatory T-cells Neutral oligosaccharide content of preterm in the protection against casein allergy by human milk. Br J Nutr. 1999;82(5):361-7. dietary non-digestible carbohydrates. Br J 111. Boehm G, Fanaro S, Jelinek J, Stahl B, Marini Nutr. 2011:1-10. A. Prebiotic concept for infant nutrition. Acta 121. van’t Land B, Schijf M, van Esch BC, van Paediatr Suppl. 2003;91(441):64-7. Bergenhenegouwen J, Bastiaans J, Schouten 112. Vos AP, Haarman M, Buco A, Govers M, B, et al. Regulatory T-cells have a prominent Knol J, Garssen J, et al. A specific prebiotic role in the immune modulated vaccine oligosaccharide mixture stimulates delayed- response by specific oligosaccharides. type hypersensitivity in a murine influenza Vaccine. 2010;28(35):5711-7. vaccination model. Int Immunopharmacol. 122. Moro G, Arslanoglu S, Stahl B, Jelinek J, 2006;6(8):1277-86. Wahn U, Boehm G. A mixture of prebiotic 113. Vos AP, van Esch BC, Stahl B, M’Rabet oligosaccharides reduces the incidence of L, Folkerts G, Nijkamp FP, et al. atopic dermatitis during the first six months Dietary supplementation with specific of age. Arch Dis Child. 2006;91(10):814-9. oligosaccharide mixtures decreases 123. Arslanoglu S, Moro GE, Boehm G. parameters of allergic asthma in mice. Int Early supplementation of prebiotic Immunopharmacol. 2007;7(12):1582-7. oligosaccharides protects formula-fed 114. Schouten B, van Esch BC, Hofman GA, van infants against infections during the first 6 Doorn SA, Knol J, Nauta AJ, et al. Cow milk months of life. J Nutr. 2007;137(11):2420-4. allergy symptoms are reduced in mice fed 124. Arslanoglu S, Moro GE, Schmitt J, Tandoi dietary synbiotics during oral sensitization L, Rizzardi S, Boehm G. Early dietary with whey. J Nutr. 2009;139(7):1398-403. intervention with a mixture of prebiotic 115. Vos AP, Haarman M, van Ginkel JW, oligosaccharides reduces the incidence Knol J, Garssen J, Stahl B, et al. Dietary of allergic manifestations and infections supplementation of neutral and acidic during the first two years of life. J Nutr. oligosaccharides enhances Th1-dependent 2008;138(6):1091-5. vaccination responses in mice. Pediatr 125. Gruber C, van Stuijvenberg M, Mosca F, Moro Allergy Immunol. 2007;18(4):304-12. G, Chirico G, Braegger CP, et al. Reduced 116. Fanaro S, Jelinek J, Stahl B, Boehm G, occurrence of early atopic dermatitis Kock R, Vigi V. Acidic oligosaccharides from because of immunoactive prebiotics among pectin hydrolysate as new component for low-atopy-risk infants. J Allergy Clin infant formulae: effect on intestinal flora, Immunol.126(4):791-7. stool characteristics, and pH. J Pediatr 126. Bakker-Zierikzee AM, Tol EA, Kroes H, Gastroenterol Nutr. 2005;41(2):186-90. Alles MS, Kok FJ, Bindels JG. Faecal SIgA 117. Meyer PD. Nondigestible oligosaccharides as secretion in infants fed on pre- or probiotic dietary fiber. J AOAC Int. 2004;87(3):718-26. infant formula. Pediatr Allergy Immunol. 2006;17(2):134-40. 118. Kerperien J, Jeurink PV, Wehkamp T, van der Veer A, van de Kant HJ, Hofman GA, et al. 127. Boyle RJ, Tang ML, Chiang WC, Chua Non-digestible oligosaccharides modulate MC, Ismail I, Nauta A, et al. Prebiotic- intestinal immune activation and suppress supplemented partially hydrolysed cow’s cow’s milk allergic symptoms. Pediatr milk formula for the prevention of eczema Allergy Immunol. 2014;25(8):747-54. in high-risk infants: a randomized controlled trial. Allergy. 2016;71(5):701-10. 119. Schouten B, van Esch BC, Hofman GA, Boon L, Knippels LM, Willemsen LE, et al. 128. Vos AP, M’Rabet L, Stahl B, Boehm G, Oligosaccharide-induced whey-specific Garssen J. Immune-modulatory effects CD25(+) regulatory T-cells are involved in and potential working mechanisms of orally applied nondigestible carbohydrates. Crit Rev Immunol. 2007;27(2):97-140.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 46 IMMUNE DEVELOPMENT

129. Eiwegger T, Stahl B, Haidl P, Schmitt health. Prostaglandins Leukot Essent Fatty J, Boehm G, Dehlink E, et al. Prebiotic Acids. 2017;126:98-104. oligosaccharides: in vitro evidence for 138. Dunstan JA, Mori TA, Barden A, Beilin gastrointestinal epithelial transfer and LJ, Taylor AL, Holt PG, et al. Fish oil immunomodulatory properties. Pediatr supplementation in pregnancy modifies Allergy Immunol. 2010;21(8):1179-88. neonatal allergen-specific immune 130. Wu RY, Maattanen P, Napper S, Scruten responses and clinical outcomes in E, Li B, Koike Y, et al. Non-digestible infants at high risk of atopy: a randomized, oligosaccharides directly regulate host controlled trial. J Allergy Clin Immunol. 2 kinome to modulate host inflammatory 2003;112(6):1178-84. responses without alterations in the gut 139. Dunstan JA, Roper J, Mitoulas L, Hartmann microbiota. Microbiome. 2017;5(1):135. PE, Simmer K, Prescott SL. The effect 131. Akbari P, Fink-Gremmels J, Willems of supplementation with fish oil during R, Difilippo E, Schols HA, Schoterman pregnancy on breast milk immunoglobulin MHC, et al. Characterizing microbiota- A, soluble CD14, cytokine levels and independent effects of oligosaccharides fatty acid composition. Clin Exp Allergy. on intestinal epithelial cells: insight 2004;34(8):1237-42. into the role of structure and size : 140. Furuhjelm C, Warstedt K, Larsson J, Structure-activity relationships of non- Fredriksson M, Bottcher MF, Falth- digestible oligosaccharides. Eur J Nutr. Magnusson K, et al. Fish oil supplementation 2017;56(5):1919-30. in pregnancy and lactation may decrease 132. Lehmann S, Hiller J, van Bergenhenegouwen the risk of infant allergy. Acta Paediatr. J, Knippels LM, Garssen J, Traidl- 2009;98(9):1461-7. Hoffmann C. In Vitro Evidence for Immune- 141. Papamichael MM, Shrestha SK, Itsiopoulos Modulatory Properties of Non-Digestible C, Erbas B. The role of fish intake on asthma Oligosaccharides: Direct Effect on Human in children: A meta-analysis of observational Monocyte Derived Dendritic Cells. PLoS studies. Pediatr Allergy Immunol. 2018. One. 2015;10(7):e0132304. 142. Almqvist C, Garden F, Xuan W, Mihrshahi 133. Hayen SM, den Hartog Jager CF, Knulst S, Leeder SR, Oddy W, et al. Omega-3 and AC, Knol EF, Garssen J, Willemsen LEM, omega-6 fatty acid exposure from early et al. Non-Digestible Oligosaccharides Can life does not affect atopy and asthma Suppress Basophil Degranulation in Whole at age 5 years. J Allergy Clin Immunol. Blood of Peanut-Allergic Patients. Front 2007;119(6):1438-44. Immunol. 2018;9:1265. 143. Prescott SL, Calder PC. N-3 polyunsaturated 134. Moreno JJ. Differential effects of fatty acids and allergic disease. Curr Opin arachidonic and eicosapentaenoic Acid- Clin Nutr Metab Care. 2004;7(2):123-9. derived eicosanoids on polymorphonuclear 144. Miles EA, Calder PC. Can Early Omega-3 transmigration across endothelial Fatty Acid Exposure Reduce Risk of cell cultures. J Pharmacol Exp Ther. Childhood Allergic Disease? Nutrients. 2009;331(3):1111-7. 2017;9(7). 135. Simopoulos AP. The importance of the 145. Vanderhoof JA. Hypoallergenicity and effects omega-6/omega-3 fatty acid ratio in on growth and tolerance of a new amino cardiovascular disease and other chronic acid-based formula with DHA and ARA. J diseases. Exp Biol Med (Maywood). Pediatr Gastroenterol Nutr. 2008;47 Suppl 2008;233(6):674-88. 2:S60-1. 136. Blumer N, Pfefferle PI, Renz H. 146. De Giuseppe R, Roggi C, Cena H. n-3 LC- Development of mucosal immune function PUFA supplementation: effects on infant in the intrauterine and early postnatal and maternal outcomes. Eur J Nutr. environment. Curr Opin Gastroenterol. 2014;53(5):1147-54. 2007;23(6):655-60. 147. Moon K, Rao SC, Schulzke SM, Patole SK, 137. Elliott E, Hanson CK, Anderson-Berry Simmer K. Longchain polyunsaturated AL, Nordgren TM. The role of specialized fatty acid supplementation in preterm pro-resolving mediators in maternal-fetal infants. Cochrane Database Syst Rev. 2016;12:CD000375.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 47 3 530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 48 Non-digestible oligosaccharides modulate intestinal immune activation and suppress cow’s milk allergic symptoms

JoAnn Kerperien1, Prescilla V. Jeurink1,2, Tjalling Wehkamp2, Anniek van der Veer1, Henk van de Kant1, Gerard Hofman1, Betty C.A.M. van Esch1,2, Johan Garssen1,2, Linette E.M. Willemsen1, Léon M.J. Knippels1,2

1 Division of Pharmacology, Department of Pharmaceutical Sciences, Faculty of Science, Utrecht University, the Netherlands 2 Nutricia Research, B.V. Utrecht, the Netherlands

This chapter is published in: 3 Pediatric Allergy and Immunology, 2014, 8:747-754 530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 49 CHAPTER 3

ABSTRACT

Background: Cow’s milk allergy is a common food allergy in childhood and no effective preventive or curative treatment is available. This study aims at comparing single oligosaccharides short- chain Galacto- (scGOS), long-chain Fructo- (lcFOS) or pectin-derived Acidic-oligosaccharides (pAOS) and/or mixtures of scGOS/lcFOS (GF) or scGOS/lcFOS/pAOS (GFA) to prevent or treat food allergy.

Methods: In the preventive protocol, C3H/HeOuJ mice were fed diets containing single oligosaccharides or mixtures GF or GFA throughout the study protocol. In the treatment protocol, GF or GFA were provided for four weeks starting after the last sensitization. The acute allergic skin response and anaphylaxis scores were determined and after oral challenge whey-specific immunoglobulins were measured and qPCR for T cell markers and Foxp3 counts using immunohistochemistry were performed on the small intestine and colon.

Results: Only in the preventive setting, the GF or GFA mixture but not the single oligosaccharides, reduced the allergic skin response and whey-IgG1 levels in whey-sensitized mice, compared to the control diet. Both GF and GFA increased the number of Foxp3+ cells in the proximal small intestine of whey- compared to sham-sensitized mice. Expression of Th2 and Th17 mRNA markers increased in the middle part of the small intestine of whey-sensitized mice which was prevented by GF. By contrast, GFA enhanced Tbet (Th1), Il10 and Tgfβ mRNA expression compared to GF which was maintained in the distal small intestine and/or colon.

Conclusion: Dietary supplementation with scGOS/lcFOS+/-pAOS during sensitization, effectively reduces allergic symptoms but differentially affects mucosal immune activation in whey sensitized mice.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 50 GF VERSUS GFA IN CMA

INTRODUCTION

The incidence of allergic diseases is increasing in westernized countries, rising to almost 20% in the United States over the last decade (1). One of the first allergies to develop in life is cow’s milk allergy (CMA). Current protocols for prevention or treatment of CMA are insufficient and novel strategies need to be developed. One approach is the use of specific dietary components that may aid to prevent or diminish atopic diseases rising in early infancy (2-4). Specific non-digestible oligosaccharides (NDO), also known as prebiotics, display some functional aspects of oligosaccharides in human milk. Neutral human milk NDO as well as the manufactured 9:1 mixture of short-chain Galacto- and long-chain Fructo-oligosaccharides (scGOS/lcFOS; GF) have been shown to affect the immune 3 system (2, 5). Since human milk also contains acidic oligosaccharides, pectin-derived Acidic oligosaccharides (pAOS) may be added to these neutral oligosaccharides in a 9:1:2 ratio (scGOS/ lcFOS/pAOS; GFA) to approach the ratios found in human milk (3).

GF and/or GFA can positively influence colonization of the gut with beneficial microbes such as bifidobacteria and lactobacilli in humans and mice, and have impact on the gut-associated lymphoid tissue and the systemic immune system (6-10). In addition, GFA reduced the incidence of atopic dermatitis in a general infant population and effectively modified the immune response in different mouse models (9, 11-13). In a vaccination model it was shown that these NDO enhanced the systemic T helper cell (Th)1 response and in an ovalbumin induced experimental asthma model they enhanced the Th1 over Th2 balance (12, 13). Indicating these NDO can skew the immune response away from the allergic phenotype.

Using an animal model for CMA, it was shown that adoptive transfer of splenocytes of whey- or casein-sensitized mice fed GFA prevented the occurrence of allergic symptoms in naïve recipients upon sensitization. This effect was generated by functional regulatory T cells (Treg) since cluster of differentiation (CD)25+ Treg depletion abrogated the protective effect (9, 14). Treg are competent suppressors of the effector cell population and act via cell-cell contact or cytokine secretion like interleukin (IL)10 and transforming growth factor beta (TGFβ), thereby contributing to oral tolerance. Besides systemic installation of functional Treg, feeding GFA to CMA mice enhanced the frequency of Th1 cells in the mesenteric lymph nodes (MLN), while for Th2 cells this tended to decline (9). Local factors in the MLN like retinoic acid (RA), TGFβ and IL10 are important for acquiring oral tolerance (15). After T cells are activated in the MLN they acquire integrin α4β7 surface expression which allows homing of these cells to the lamina propria (LP) of the small intestine or proximal colon, both drained by the MLN (16). A subset of MLN instructed cells, including Forkhead Box P3 (Foxp3)+ Treg, expresses c-c motif chemokine receptor (CCR)9 and selectively homes to the small intestine (16, 17). In the colon local factors like cytotoxic T lymphocyte antigen-(CTLA)4 can instruct Foxp3+ Treg formation from CD4+Foxp3- precursors (18).

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 51 CHAPTER 3

Since currently the effect of single components scGOS, lcFOS, or pAOS within the mixtures of GF or GFA is unknown, in the present study these were compared in the prevention and/or treatment of CMA in mice orally sensitized with whey. Furthermore, the effect of GF compared to GFA on mucosal immune activation was studied in specific parts of the intestine.

MATERIALS AND METHODS

Chemicals Whey protein, cholera toxin (CT), Phosphate Buffer Saline (PBS) and all biotin-labelled rat anti- mouse IgE, IgG1 and IgG2a, mouse mast cell protease 1 (mMCP1) (also known as MCPT1) enzyme-linked immunosorbent assay (ELISA) were used as described previously (19).

Diet Cow’s milk-protein free, AIN-93G-based diets were composed and mixed with NDO by Research Diet Services (Wijk bij Duurstede, The Netherlands). ScGOS (Vivinal scGOS: Friesland Campina Domo, Amersfoort, The Netherlands), lcFOS (Orafti, Wijchen, The Netherlands,) and pAOS (Sudsucker, Mannheim, Germany) were added as described previously (14). Combinations were scGOS/lcFOS (GF 9:1; 0.8 w/w%) and scGOS/lcFOS (9:1;0.8 w/w%)/pAOS (2;0.2w/w%) (GFA 9:1:2; 1 w/w%). Single components were added in the same amount compared to the combined GFA diet: scGOS (0.75 w/w%), scFOS (0.08 w/w%) and pAOS (0.17 w/w%).

Mouse model Three- to four-week-old specific pathogen-free female C3H/HeOuJ mice were purchased from Charles River Laboratories (France). The care and use of animals was performed in accordance with the guidelines of the Dutch Committee of Animal Experiments (number 2010.II.06.113).

Preventive protocol The preventive diet was given two weeks prior to and during sensitization of the mice (n=6 per group; Figure 1). One week after the last sensitization, mice were challenged intradermally (i.d.) in both ears with 20μL whey (0.5 g/L PBS) to determine anaphylaxis and the acute allergic skin response as described previously (14). Anaphylaxis was scored according to Li et al. (20). Eighteen hours before sacrifice, mice were challenged orally intra gastrical (i.g.) with 0.5 mL homogenized whey (200 g/L PBS). Blood was collected in minicollect Z serum Sep vials (Greiner Bio-one, Alphen a/d Rijn, the Netherlands) and centrifuged at 14,000g for 15 minutes. Serum was stored at -20°C for measurements of mMCP1, whey-specific IgE, -IgG1, and -IgG2a as described previously (19).

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 52 GF VERSUS GFA IN CMA

Preventive protocol

Control diet or single / mixtures of non-digestible oligosaccharide diet

-14 -7 0 7 33 28 21 14 34 days Oral challenge, Sacrifice Sensitization by gavage with CT (sham) or whey+CT Acute allergic skin response: i.d. with whey in ear

days 0 7 21 14 28 35 42 70 72

control diet control diet or mixtures of non- digestible oligosaccharide diet Treatment protocol 3 Figure 1. Overview of the preventive and treatment protocol. CT: cholera toxin; i.d.: intra-dermal.

Treatment protocol Mice (n=8 per group) were sensitized for six consecutive weeks i.g. with CT (20 μg/mL PBS) as an adjuvant with or without 0.5 mL homogenized whey (40 mg/mL PBS). All groups received control diet until i.d. challenge in one ear at day 42. Subsequently, mice were subjected to a control, GF or GFA diet. After the dietary treatment of four weeks, all mice received another i.d. challenge in the other ear (Figure 1). One hour before sacrifi ce, mice were i.g. challenged with 0.5 mL homogenized whey (200 g/L PBS).

Immunohistochemistry Immunohistochemical Foxp3 staining was performed according to previously described protocol by van den Elsen et al. (21).

qPCR One centimetre of four specifi c mouse intestinal parts (indicated in Table 1C) was stored in RNAlater™ (Qiagen GmbH, Hilden, Germany) at 4°C until further processing. Messenger RNA was extracted using Roche mRNA extraction kit (Roche, Mannheim, Germany, 11787896001) and cDNA was made using the reverse transcriptase system (Promega, Madison, WI, USA) according to Garcia-Vallejo et al. (22). Quantitive real time PCR was performed on a CFX96 real time PCR detection system (Bio Rad, Veenendaal, the Netherlands) using iQ SYBR green supermix (Bio Rad, Veenendaal, the Netherlands). Validated primers for Foxp3, T box transcription factor (Tbet), Gata3, RAR-related orphan receptor gamma (Rorγ), Tgfβ1, Il10, interferon gamma (Ifnγ), Il4 and Il17A were purchased from SA bioscience (Qiagen, German Town, MD, USA). Housekeeping gene Ribosomal protein S13 (Rps13) was obtained from Isogen Life Science (De Meern, the Netherlands): forward 5′-GTCCGAAAGCACCTTGAGAG-3′ and reverse 5′-AGCAGAGGCTGTGGATGACT-3’. The mRNA level was calculated with CFX Manager software (version 1.6) and corrected for the expression of Rps13 with 100*2^(Rps13-gene of

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 53 CHAPTER 3

interest (g.o.i.)). The data in Table 1 were generated with the calculated values, relative values of the g.o.i. of whey-sensitized mice were divided by the mean of the g.o.i. of sham-sensitized mice fed the control diet. Levels <1 indicate a decrease and >1 an increase in the relative expression of the specific gene compared to control diet-fed sham-sensitized mice.

Statistics The acute skin response (mean ± S.E.M), mMCP1 and immunoglobulin data were analysed using One-way ANOVA and post-hoc Bonferroni multiple comparison test using GraphPad Prism software (version 5). Immunoglobulin and mMCP1 data were log transformed prior to analyses to normalize the data distribution and presented in Tukey box-and-whiskers plots. The anaphylaxis score was evaluated with a Kruskal-Wallis and post-hoc Mann-Whitney U-test. P<0.05 was considered significant.

RESULTS

Acute allergic skin response and anaphylactic symptom scores After the whey-challenge in the ear, the acute allergic skin response and anaphylactic symptom scores were determined. In whey-sensitized mice fed the control diet or different NDO diets, the acute allergic skin response was increased compared to sham sensitized mice (Figure 2A). The acute allergic skin response was reduced in mice fed the mixture of GF or GFA compared to whey-sensitized mice fed the control diet (Figure 2A). The anaphylactic symptom scores showed the same pattern (Figure 2B).

Since the GF or GFA diet reduced the acute allergic skin response in the preventive setting, these mixtures were also used in a treatment setting. The oligosaccharide diets did not reduce the acute allergic skin response compared to mice fed a control diet during treatment (Figure 2C). The anaphylactic symptom scores were only increased in the whey-sensitized mice fed the control diet compared to sham-sensitized mice (Figure 2D).

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 54 GF VERSUS GFA IN CMA

Figure 2. The acute allergic skin response and anaphylactic symptom scores in whey-sensitized mice fed a scGOS/lcFOS (GF) or scGOS/lcFOS/pAOS (GFA) diet were reduced in a preventive (A, B), but not in a treatment setting (C, D). The acute allergic skin response one hour after intra-dermal injection (i.d.) with whey in the ear, n=5 or 6, mean ± SEM. The measured ear thickness before i.d. injection is subtracted from the measured thickness one hour after the i.d. injection (A). The anaphylactic symptom scores of sham- sensitized mice on control diet and whey-sensitized mice fed a control- or non-digestible oligosaccharide diet (Kruskal Wallis) (B). One hour after i.d. injection with whey in the ear, the acute allergic skin response was measured after 4 weeks of dietary treatment; n=7 or 8, mean ± SEM (C). Anaphylactic symptom score was determined one hour after the i.d. injection (Kruskal Wallis) (D). Asterix indicates a significant difference compared to sham-sensitized mice, ***p<0.001, **p<0.01, *p<0.05. Hash tag indicates significance compared to whey-sensitized mice fed a control diet, ##p<0.01, #p<0.05. scGOS or G; short- chain Galacto-ligosaccharides, lcFOS or F; long-chain Fructo-oligosaccharides, pAOS or A; pectin-derived Acidic-oligosaccharides.

Whey-specific immunoglobulins and mMCP1 in serum Whey-specific IgE was increased in whey-sensitized mice fed a control or GFA diet compared to sham-sensitized mice (Figure 3A). The GF or GFA diets did not affect whey-specific IgE levels compared to whey-sensitized mice fed the control diet. However, whey-specific IgG1

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 55 CHAPTER 3

was increased in whey-sensitized mice fed the control diet compared to sham-sensitized mice, whereas both the GF and GFA diet lowered these levels (Figure 3B). mMCP1 was only increased in whey-sensitized mice fed the control diet compared to sham-sensitized mice (Figure 3C). Whey-specific IgG1 correlated positively with mMCP1 (Figure 3D).

Figure 3. Whey-specific immunoglobulins and mouse mast cell protease 1 (mMCP1) levels measured in serum of mice (preventive protocol). Whey-specific-IgE (AU) (A), whey-specific-IgG1 (AU x1000) (B) and mMCP1 (ng/mL) (C) were measured. Whey-specific IgG1 serum levels correlated positive with mMCP1 concentrations (Pearson correlation) (D). Data represent n= 5 to 6 per group, Tukey box-and-whiskers. Asterix indicates a significant difference compared to sham sensitized mice, ***p<0.001, **p<0.01, *p<0.05. Hash tag indicates significance compared to whey sensitized mice fed a control diet, ##p<0.01, #p<0.05. GF or GFA; short-chain Galacto-oligosaccharides (G), long-chain Fructo-oligosaccharides (F), pectin-derived Acidic-oligosaccharides (A).

Foxp3+ positive cells in different parts of the intestinal mucosa To determine whether Foxp3+ cell counts were altered in the intestinal LP of the small intestine or colon, immunohistochemistry was used. Eighteen hours after the oral challenge, the number of Foxp3+ cells in the proximal part of the small intestine was increased by the GF diet and the GFA diet showed the same tendency in whey-sensitized mice compared to sham-sensitized mice fed the control diet (Figure 4A-D).

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 56 Non‐digestible oligosaccharides modulate intestinal immune activation and suppress cow's milk allergic symptoms

GF VERSUS GFA IN CMA

A B C D

Figure 4. Analysis of Foxp3+ cells in the intestine by immunohistochemistry (preventive protocol). Foxp3+ cells were stained in the lamina propria of the first part of the small intestine (A, B) and colon (C, D). Data 3 represent n= 3 to 4 per group, mean ± SEM. Asterix indicate a significant difference compared to sham- sensitized mice, **p<0.01, $ p=0.076. GF or GFA; short-chain Galacto-oligosaccharides (G), long-chain Fructo-oligosaccharides (F), pectin-derived Acidic-oligosaccharidess (A).

qPCR analysis of specific markers in the intestine To study the effect of the dietary non-digestible oligosaccharides on intestinal immune polarization, qPCR analysis was performed on markers for T cell subtypes and regulatory Pediatric Allergy and Immunology, Volume: 25, Issue: 8, Pages: 747-754, First published: 19 November 2014, DOI: (10.1111/pai.12311) cytokines in specific parts of the intestine; 1 cm proximal-, middle-, distal small intestine and colon (Table 1C). Only in the middle part of the small intestine significant differences in gene expression was observed in whey-sensitized mice compared to sham-control mice fed the control diet (Figure 5), relative expression of Gata3 (Th2), Rorγ (Th17) and Foxp3 (Treg) transcription factors to housekeeping gene Rps13 were increased. Dietary intervention with the non-digestible oligosaccharide mixtures during whey-sensitization did not affect relative Foxp3 expression in this part of the intestine but remarkable differences between the GF and GFA interventions were observed for Gata3 (Th2), Tbet (Th1), Rorγ (Th17), Tgfβ and Il10 (regulatory cytokines).

The GF diet significantly down-regulated the relative expression ofGata3 and Rorγ compared to whey-sensitized mice fed the control diet (Figure 5A, C). The GFA diet significantly increased the relative expression of Tbet, Gata3 and Il10 compared to control diet fed sham-sensitized mice and Tbet and Il10 compared to whey sensitized mice fed the GF mixture, with the same tendency for Tgfβ (Figure 5B, E, D).

In Table 1, analysis of other parts of the intestine is shown expressed as the relative increase for the Rps13 normalized g.o.i. of whey-sensitized mice fed the different diets compared to sham-sensitized mice fed the control diet. The GF diet increased relative expression of Foxp3 in the proximal small intestine of whey-sensitized mice, while decreasing Gata3, Rorγ and Il17A in the middle small intestine compared to the relative expression of whey-sensitized

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 57 CHAPTER 3

mice fed the control diet (Table 1A, B). By contrast, the GFA diet significantly increased relative expression of Tbet, Il4 and Il17A compared to the relative expression of whey-sensitized mice fed the control diet in the proximal small intestine, and Tbet, Ifnγ, Gata3, Il4, Rorγ and Il17A in the distal small intestine and colon (Table 1A). This coincided with an increase of Tgfβ and Il10 in the proximal ileum, Il10 in the middle small intestine and regulatory Foxp3, Tgfβ and Il10 in the distal small intestine and colon (Table 1A, B).

Figure 5. Whey-allergic mice fed GF or GFA showed a differential increase of specific markers in the middle part of the small intestine (preventive protocol). The mRNA of transcription factors representing Th2 (Gata3) (A), Th1 (Tbet) (B), Th17 (Rorγ) (C), and regulatory cytokines Tgfβ (D) and Il10 (E) and Treg transcription factor Foxp3 (F) were evaluated in 1 cm of the middle small intestine of C3H/HeOuJ mice fed control, GF or GFA diet during sensitization. Data represent n= 3 to 6 per group, mean ± SEM. Asterix indicates a significant difference compared to sham with **p<0.01, *p<0.05. Dollar sign indicates significance compared to whey-sensitized mice fed GF diet, $p<0.05. GF or GFA; short-chain Galacto- oligosaccharides (G), long-chain Fructo-oligosaccharides (F), pectin-derived Acidic-oligosaccharides (A).

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 58 GF VERSUS GFA IN CMA

C ileum colon ileum colon C

------

TH17 Rorg 0.87 +/ 2.13 +/ 2.98 =/ 6.57 +/ IL17a 3.22 +/ 1.09 +/ 1.91 +/ 4.70 +/ 0.46 0.95 1.22 ## 2.99 $ 1.06 $ 0.48 0.89 $$ 3.09 * 6.57 +/- 2.99 $ 4.70 +/- 3.09 $ 2.98 +/- 1.22 ## 1.91 +/- 0.89 $$ 2.13 +/- 0.95 1.09 +/- 0.48 Rorg 0.87 +/- 0.46 IL17a 3.22 +/- 1.06 $$ red to

- -

------* $ $

$$ $$ *

18.49 +/- 10.59 $ 17.60 +/- 10.37 $ 7.27 +/- 2.88 $$ 7.21 +/- 2.78 $$ 3.49 +/- 1.66 3.42 +/- 1.55 Gata3 3.51 +/- 0.76 IL4 3.19 +/- 0.92 Th2 Gata3 3.51 +/ 3.49 +/ 7.27 +/ 18.49 +/ IL4 3.19 +/ 3.42 +/ 7.21 +/ 17.60 +/ 0.76 1.66 2.88 10.59 0.92 1.55 2.78 10.37

$ - - - - # -

- - -

* # *

$$ $$ GFA

18.19 +/- 13.13 12.02 +/- 7.76 # 9.80 +/- 4.46 $$ 4.92 +/- 3.43 5.14 +/- 3.69 #p=0.051 2.00 +/- 1.21 3.11 +/- 1.04 $$ 2.05 +/- 1.46 Th1T bet Th2 TH17 IFNg IFNg 2.05 +/ 2.00 +/ 4.92 +/ 12.02 +/ 1.46 1.21 3.43 7.76 Th1 T bet 3.11 +/ 5.14 +/ 9.80 +/ 18.19 +/ 1.04 3.69 4.46 13.13

- 3

- - -

$$ $$ $$ #

15.57 +/- 7.72 $$ 6.20 +/- 2.61 $$ 3.83 +/- 1.80 # 1.48 +/- 0.45 $ ## $$ $ L10 I 1.48 +/ 3.83 +/ 6.20 +/ 15.57 +/ 0.45 1.80 2.61 7.72 Hash tag indicates significance compa

0.26 1.17 1.65 4.50

------

$$ $$

9.66 +/- 4.50 7.31 +/- 4.64 # 4.18 +/- 1.65 $ 4.66 +/- 1.98 $$ 3.13 +/- 1.17 2.78 +/- 1.74 1.08 +/- 0.26 ### 1.71 +/- 0.62 $ Whey sensitized mice fed fed sensitized mice Whey (GFA) scGOS/lcFOS/pAOS Treg Foxp3 TGFb IL10 , distal ileum and proximal colon calculated as factor - (IL4) and Th17 in black (IL17) showed as factor Treg Foxp3 1.08 +/ 3.13 +/ 4.18 +/ 9.66 +/ TGFb 1.71 +/ 2.78 +/ 4.66 +/ 7.31 +/ 0.62 1.74 1.98 4.64 * , middle

------

0.55 +/- 0.40 0.19 +/- 0.14 0.25 +/- 0.12 0.07 +/- 0.05 0.88 +/- 0.61 0.19 +/- 0.12 $$ Rorg 0.75 +/- 0.16 IL17a 1.08 +/- 0.08

* p<0.05, *** p<0.001. TH17 Rorg 0.75 +/ 0.88 +/ 0.25 +/ 0.55 +/ IL17a 1.08 +/ 0.19 +/ 0.07 +/ 0.19 +/ 0.16 0.61 * 0.12 0.40 0.08 0.12 $$ 0.05 0.14 , * and histology (dark grey). All numbers are shown as factor increase/

n (IFNγ), Th2 in red 1.01 +/- 0.86 1.39 +/- 0.87 0.68 +/- 0.48 0.58 +/- 0.34 1.13 +/- 0.83 1.33 +/- 0.81 Gata3 2.32 +/- 0.71 IL4 2.51 +/- 0.62 ------n (C) proximal

Th2 Gata3 2.32 +/ 1.13 +/ 0.68 +/ 1.01 +/ IL4 2.51 +/ 1.33 +/ 0.58 +/ 1.39 +/ 0.71 0.83 0.48 0.86 0.62 0.81 0.34 0.87 fed control diet

0.35 +/- 0.45 0.60 +/- 0.41 0.31 +/- 0.26 0.23 +/- 0.16 0.45 +/- 0.37 0.40 +/- 0.29 0.55 +/- 0.23 0.93 +/- 0.16 Th1T bet Th2 TH17 IFNg ------

GF *** Th1 T bet 0.55 +/ 0.45 +/ 0.31 +/ 0.35 +/ IFNg 0.93 +/ 0.40 +/ 0.23 +/ 0.60 +/ 0.23 0.37 0.26 0.45 0.16 0.29 0.16 0.41 0.75 +/- 0.74 0.27 +/- 0.17 0.67 +/- 0.32 0.56 +/- 0.18 sensitized mice

- - - -

2.53 +/- 2.63 0.20 +/- 0.12 1.19 +/- 0.83 0.19 +/- 0.12 4.05 +/- 2.90 0.49 +/- 0.59 2.20 +/- 0.35 0.47 +/- 0.18 Whey sensitized mice fed fed sensitized mice Whey (GF) scGOS/lcFOS Treg Foxp3 TGFb IL10 ***

IL10 0.56 +/ 0.67 +/ 0.27 +/ 0.75 +/ 0.18 0.32 0.17 0.74 . A; Treg in blue (Foxp3), Th1 in green (Tbet), Th2 in red (Gata3) and Th17 in black (Rorγ) showed as factor 0.35 2.90 0.83 2.63

------

2.73 +/- 1.49 0.33 +/- 0.35 1.53 +/- 1.04 0.20 +/- 0.16 2.66 +/- 0.77 0.39 +/- 0.13 Rorg 0.96 +/- 0.14 IL17a 0.83 +/- 0.62 Treg Foxp3 2.20 +/ 4.05 +/ 1.19 +/ 2.53 +/ TGFb 0.47 +/ 0.49 +/ 0.19 +/ 0.20 +/ 0.18 0.59 0.12 0.12 es significant difference compared to whey and to GF, $ p<0.05, $$ p<0.01.

------

4.02 +/- 4.04 3.24 +/- 3.42 2.57 +/- 1.31 1.82 +/- 1.04 3.67 +/- 0.89 2.59 +/- 1.36 Gata3 3.01 +/- 1.06 IL4 1.41 +/- 0.57

TH17 Rorg 0.96 +/ 2.66 +/ 1.53 +/ 2.73 +/ IL17a 0.83 +/ 0.39 +/ 0.20 +/ 0.33 +/ 0.14 0.77 1.04 1.49 0.62 0.13 0.16 0.35 cant difference compared to whey

1.87 +/- 2.28 3.92 +/- 4.69 1.32 +/- 0.65 1.44 +/- 0.93 2.50 +/- 0.53 0.94 +/- 0.29 0.87 +/- 0.19 2.05 +/- 2.25 Th1T bet Th2 TH17 IFNg ------

1. Dollar sign indicat Th2 Gata3 3.01 +/ 3.67 +/ 2.57 +/ 4.02 +/ IL4 1.41 +/ 0.57 2.59 +/ 1.82 +/ 1.04 3.24 +/ 1.06 0.89 1.31 4.04 1.36 3.42 <0.0

p 1.86 +/- 1.76 1.28 +/- 0.78 2.09 +/- 1.24 0.71 +/- 0.33 ------

Whey Th1 T bet 0.87 +/ 2.50 +/ 1.32 +/ 1.87 +/ IFNg 2.05 +/ 0.94 +/ 1.44 +/ 3.92 +/ 0.19 0.53 0.65 2.28 2.25 0.29 0.93 4.69 Whey sensitized mice fed fed sensitized mice Whey diet control 3.22 +/- 3.60 2.11 +/- 2.51 2.00 +/- 1.20 1.11 +/- 0.70 4.76 +/- 1.37 2.09 +/- 0.86 1.14 +/- 0.24 0.88 +/- 0.34 Treg Foxp3 TGFb IL10 # p<0.05, SEM. Asterix indicates a signifi - - - - -

IL10 0.71 +/ 2.09 +/ 1.28 +/ 1.86 +/ 1.76 GF diet, 0.33 1.24 0.78 0.24 1.37 1.20 3.60

------

red to sham +/ B Colon Colon Distal small Distal intestine small Distal intestine Middle small Middle intestine small Middle intestine Proximal small Proximal intestine small Proximal intestine A Treg Foxp3 1.14 +/ 4.76 +/ 2.00 +/ 3.22 +/ TGFb 0.88 +/ 2.09 +/ 1.11 +/ 2.11 +/ 0.34 0.86 0.70 2.51 59

Table 1; Relative transcription factors expression (A) for Treg, Th1, Th2, and Th17 and (B) associated cytokines measured i increased compared to sham sensitized ,ice fed a control diet in the preventive protocol increased, or decreased, compared to sham sensitized mice. B; regulatory cytokines for Treg in blue (TGFβ, IL10), Th1 in gree increased, or decreased, compared to sham sensitized mice. C; Schematic overview of the intestine used for qPCR (light grey) decrease compa whey sensitized mice on A Proximal ileum Middle ileum Distal ileum Colon B Proximal ileum Middle ileum Distal ileum Colon

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 59 CHAPTER 3

Table 1. Relative transcription factor expression (A) for Treg, Th1, Th2, and Th17 markers and (B) associated cytokines measured in (C) proximal-, middle-, and distal small intestine and proximal colon. Calculated as factor increase or decrease compared to sham-sensitized mice fed a control diet in the preventive protocol; Treg in blue (Foxp3), Th1 in green (Tbet), Th2 in red (Gata3) and Th17 in black (Rorγ) (A). Regulatory cytokines in blue (Tgfβ, Il10), Th1 (Ifnγ) cytokine in green, Th2 (Il4) cytokine in red and Th17 (Il17) in black (B). Schematic overview of parts of the intestine used for qPCR (light grey) and histology (dark grey) (C). All numbers represent a factor increase or decrease compared to sham +/- SEM. Asterix indicate a significant difference compared to whey-sensitized mice fed control diet, * p<0.05,** p<0.01, *** p<0.001. Hash tag indicates significance compared to whey-sensitized mice fed the GF or GFA diet, # p<0.05, ## p<0.01, ### p<0.001. Dollar sign indicates significant difference compared to whey sensitized mice fed control diet and to whey sensitized mice fed GF, $ p<0.05, $$ p<0.01, $$$ p< 0.001. GF or GFA; short-chain Galacto-oligosaccharides (G), long-chain Fructo-oligosaccharides (F), pectin-derived Acidic- oligosaccharides (A).

DISCUSSION

In the current study preventive administration of the single components scGOS, lcFOS or pAOS did not reduce the clinical symptoms in a mouse model for CMA while the mixtures did. Although scGOS and other NDO may provide protection against infectious agents, atopic dermatitis and stimulate the growth of beneficial bacterial (2, 4, 11, 23, 24). The single components were used in the same amount as present in the mixture which resembles the natural composition of neutral versus acidic oligosaccharides present in human milk. However, increased dosages of the single components may be beneficial as was shown for example for FOS (25-27). Although the mixtures of oligosaccharides were effective in allergy prevention, which is in line with recent studies in human and mice (4, 9, 11-14), they lacked effectiveness in the used treatment protocol. Future studies are needed to assess whether these dietary components are able to enhance efficacy and/or safety of allergen-specific immunotherapy.

As summarized by Berin and Sampson (28), Peyer’s Patches are dispensable for the development of tolerance, whereas MLN are required. Previously, we showed the importance of Treg in the protective effect of dietary oligosaccharides (9), and found synbiotics to induce Th1 and Treg cell differentiation in the MLN of CMA mice (29). Furthermore, GFA enhanced the Th1 prone vaccination response (12). These findings indicate NDO to skew away from the allergic phenotype. Since cells originating from the MLN home to the LP of the small intestine and colon, where they can exert their effector function, effects of GF and GFA on mucosal immune activation was studied. In particular in the middle part of the small intestine, increased Th2 (Gata3), Th17 (Rorγ) and Treg (Foxp3) type mRNA expression was observed in the whey- versus sham-sensitized mice. Studies of Zuercher et al. (30) and Wang et al. (31) using murine models for food allergy also show increased mRNA expression of Th2 and/or Th17 markers in the small intestine.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 60 GF VERSUS GFA IN CMA

GF and GFA were both effective in reducing the allergic symptoms and enhancing the number of Foxp3+ cells in the small intestine, it remains to be established whether the oligosaccharide diets exert their effect during the process of sensitization and/or whether Foxp3+ cell numbers rise as a consequence of oral allergen challenge. In addition, we cannot exclude the possibility that the rise in Foxp3+ cells is a consequence of an inflammatory response to whey rather than a protective response. However, in previous studies we showed CD25+ regulatory T cells to be responsible for the protective effect of the GFA diet. Splenocytes of whey or casein sensitized mice fed the GFA diet could transfer allergy protection when transferred to naive recipient mice prior to sensitization. This protective effect was abrogated when CD25+Foxp3+CD4+ regulatory T cells were depleted prior to transfer (9, 32). 3

A differential effect on mucosal immune activation was observed. In terms of controlling allergic sensitization GF may actually reduce Th2 prone immune activation hereby dampening allergic sensitization. By contrast, GFA may alter the immune balance by enhancing Th1 and regulatory T cell responses capable of counter acting the Th2 driven process of allergic sensitization. In addition, regulatory T cells, TGFβ and IL10 have been shown to modulate effector cell degranulation (33-35). Hence, GF, but in particular GFA may suppress allergic symptoms by enhancing regulatory T cells and regulatory cytokines. Future studies are needed to determine whether IL10 and TGFβ contribute to the protective effect of the GFA diet.

One of the known functions of pAOS is that it can be fermented by luminal bacteria and acts as receptor analogue to inhibit adhesion of pathogens (36, 37). However, currently it is unknown whether addition of pAOS to GF results in alterations of the microflora which could contribute to differential immune activation of the intestinal mucosa. Alternatively pAOS may have interfered directly with the immune cells as was described previously (37, 38).

In summary, dietary intervention with mixtures of GF or GFA, but not its single components, reduced CMA symptoms in a preventive setting. Both mixtures enhanced the number of Foxp3+ cells in the intestinal LP of whey-sensitized mice, although they differentially regulated mucosal immune activation. The GF diet selectively down-regulated Th2 and Th17 activation in the middle part of the small intestine, while GFA induced an overall intestinal immune activation with regulatory component, both in association with reduced allergic outcome.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 61 CHAPTER 3

REFERENCES

1. Branum AM, Lukacs SL. Food allergy among allergy symptoms are reduced in mice fed children in the United States. Pediatrics. dietary synbiotics during oral sensitization 2009;124(6):1549-55. with whey. J Nutr. 2009;139(7):1398-403. 2. Arslanoglu S, Moro GE, Schmitt J, Tandoi 11. Gruber C, van Stuijvenberg M, Mosca F, Moro L, Rizzardi S, Boehm G. Early dietary G, Chirico G, Braegger CP, et al. Reduced intervention with a mixture of prebiotic occurrence of early atopic dermatitis oligosaccharides reduces the incidence because of immunoactive prebiotics among of allergic manifestations and infections low-atopy-risk infants. J Allergy Clin during the first two years of life. J Nutr. Immunol. 2010;126(4):791-7. 2008;138(6):1091-5. 12. Vos AP, Haarman M, van Ginkel JW, 3. Fanaro S, Jelinek J, Stahl B, Boehm G, Knol J, Garssen J, Stahl B, et al. Dietary Kock R, Vigi V. Acidic oligosaccharides from supplementation of neutral and acidic pectin hydrolysate as new component for oligosaccharides enhances Th1-dependent infant formulae: effect on intestinal flora, vaccination responses in mice. Pediatr stool characteristics, and pH. J Pediatr Allergy Immunol. 2007;18(4):304-12. Gastroenterol Nutr. 2005;41(2):186-90. 13. Vos AP, van Esch BC, Stahl B, M’Rabet 4. Moro G, Arslanoglu S, Stahl B, Jelinek J, L, Folkerts G, Nijkamp FP, et al. Wahn U, Boehm G. A mixture of prebiotic Dietary supplementation with specific oligosaccharides reduces the incidence of oligosaccharide mixtures decreases atopic dermatitis during the first six months parameters of allergic asthma in mice. Int of age. Arch Dis Child. 2006;91(10):814-9. Immunopharmacol. 2007;7(12):1582-7. 5. Schouten B, Van Esch BC, Kormelink 14. Schouten B, van Esch BC, Hofman GA, TG, Moro GE, Arslanoglu S, Boehm G, Boon L, Knippels LM, Willemsen LE, et al. et al. Non-digestible oligosaccharides Oligosaccharide-induced whey-specific reduce immunoglobulin free light-chain CD25(+) regulatory T-cells are involved in the concentrations in infants at risk for allergy. suppression of cow milk allergy in mice. J Pediatr Allergy Immunol. 2011;22(5):537-42. Nutr. 2010;140(4):835-41. 6. Knol J, Scholtens P, Kafka C, Steenbakkers 15. Barnes MJ, Powrie F. Regulatory T cells J, Gro S, Helm K, et al. Colon microflora reinforce intestinal homeostasis. Immunity. in infants fed formula with galacto- and 2009;31(3):401-11. fructo-oligosaccharides: more like breast- 16. Johansson-Lindbom B, Agace WW. fed infants. J Pediatr Gastroenterol Nutr. Generation of gut-homing T cells and their 2005;40(1):36-42. localization to the small intestinal mucosa. 7. Roberfroid M, Gibson GR, Hoyles L, Immunological reviews. 2007;215:226-42. McCartney AL, Rastall R, Rowland I, et al. 17. Jaensson E, Uronen-Hansson H, Pabst Prebiotic effects: metabolic and health O, Eksteen B, Tian J, Coombes JL, et al. benefits. Br J Nutr. 2010;104 Suppl 2:S1-63. Small intestinal CD103+ dendritic cells 8. Roberfroid MB. Health benefits of non- display unique functional properties that digestible oligosaccharides. Advances are conserved between mice and humans. in experimental medicine and biology. The Journal of experimental medicine. 1997;427:211-9. 2008;205(9):2139-49. 9. Schouten B, van Esch BC, Hofman GA, 18. Barnes MJ, Griseri T, Johnson AM, Young W, de Kivit S, Boon L, Knippels LM, et al. A Powrie F, Izcue A. CTLA-4 promotes Foxp3 potential role for CD25+ regulatory T-cells induction and regulatory T cell accumulation in the protection against casein allergy by in the intestinal lamina propria. Mucosal dietary non-digestible carbohydrates. Br J immunology. 2013;6(2):324-34. Nutr. 2012;107(1):96-105. 19. Schouten B, van Esch BC, Hofman GA, van 10. Schouten B, van Esch BC, Hofman GA, van den Elsen LW, Willemsen LE, Garssen J. Doorn SA, Knol J, Nauta AJ, et al. Cow milk Acute allergic skin reactions and intestinal

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 62 GF VERSUS GFA IN CMA

contractility changes in mice orally 28. Berin MC, Sampson HA. Mucosal sensitized against casein or whey. Int Arch immunology of food allergy. Current biology Allergy Immunol. 2008;147(2):125-34. : CB. 2013;23(9):R389-400. 20. Li XM, Schofield BH, Huang CK, Kleiner 29. de Kivit S, Saeland E, Kraneveld AD, van de GI, Sampson HA. A murine model of IgE- Kant HJ, Schouten B, van Esch BC, et al. mediated cow’s milk hypersensitivity. J Galectin-9 induced by dietary synbiotics is Allergy Clin Immunol. 1999;103(2 Pt 1):206- involved in suppression of allergic symptoms 14. in mice and humans. Allergy. 2012;67(3):343- 21. van den Elsen LW, van Esch BC, Hofman 52. GA, Kant J, van de Heijning BJ, Garssen J, 30. Zuercher AW, Weiss M, Holvoet S, Moser et al. Dietary long chain n-3 polyunsaturated M, Moussu H, van Overtvelt L, et al. fatty acids prevent allergic sensitization Lactococcus lactis NCC 2287 alleviates food to cow’s milk protein in mice. Clinical and allergic manifestations in sensitized mice experimental allergy : journal of the British by reducing IL-13 expression specifically 3 Society for Allergy and Clinical Immunology. in the ileum. Clinical & developmental 2013;43(7):798-810. immunology. 2012;2012:485750. 22. Garcia-Vallejo JJ, Van Het Hof B, Robben J, 31. Wang M, Takeda K, Shiraishi Y, Okamoto Van Wijk JA, Van Die I, Joziasse DH, et al. M, Dakhama A, Joetham A, et al. Peanut- Approach for defining endogenous reference induced intestinal allergy is mediated genes in gene expression experiments. Anal through a mast cell-IgE-FcepsilonRI- Biochem. 2004;329(2):293-9. IL-13 pathway. J Allergy Clin Immunol. 23. Kukkonen K, Savilahti E, Haahtela T, 2010;126(2):306-16, 16 e1-12. Juntunen-Backman K, Korpela R, Poussa 32. van Esch BC, Schouten B, Blokhuis T, et al. Probiotics and prebiotic galacto- BR, Hofman GA, Boon L, Garssen J, oligosaccharides in the prevention of et al. Depletion of CD4+CD25+ T cells allergic diseases: a randomized, double- switches the whey-allergic response from blind, placebo-controlled trial. J Allergy Clin immunoglobulin E- to immunoglobulin free Immunol. 2007;119(1):192-8. light chain-dependent. Clin Exp Allergy. 24. Sangwan V, Tomar SK, Singh RR, Singh 2010;40(9):1414-21. AK, Ali B. Galactooligosaccharides: novel 33. Gomez G, Ramirez CD, Rivera J, Patel M, components of designer foods. J Food Sci. Norozian F, Wright HV, et al. TGF-beta 1 2011;76(4):R103-11. inhibits mast cell Fc epsilon RI expression. J 25. Fujiwara R, Sasajima N, Takemura N, Immunol. 2005;174(10):5987-93. Ozawa K, Nagasaka Y, Okubo T, et al. 34. Kashyap M, Thornton AM, Norton SK, 2,4-Dinitrofluorobenzene-induced contact Barnstein B, Macey M, Brenzovich J, et hypersensitivity response in NC/Nga al. Cutting edge: CD4 T cell-mast cell mice fed fructo-oligosaccharide. Journal interactions alter IgE receptor expression of nutritional science and vitaminology. and signaling. J Immunol. 2008;180(4):2039- 2010;56(4):260-5. 43. 26. Inoue R, Tsukahara T, Ueno K, Kitabayashi 35. Kennedy Norton S, Barnstein B, Brenzovich T, Ushida K. Possible link of a compositional J, Bailey DP, Kashyap M, Speiran K, et al. change in intestinal microbiota with the anti- IL-10 suppresses mast cell IgE receptor allergic effect of fructo-oligosaccharides in expression and signaling in vitro and in vivo. NC/jic mice. Bioscience, biotechnology, and J Immunol. 2008;180(5):2848-54. biochemistry. 2010;74(9):1947-50. 36. Boehm G, Moro G. Structural and functional 27. Yasuda A, Inoue K, Sanbongi C, aspects of prebiotics used in infant nutrition. Yanagisawa R, Ichinose T, Tanaka M, J Nutr. 2008;138(9):1818S-28S. et al. Dietary supplementation with 37. Knaup B, Kempf M, Fuchs J, Valotis A, fructooligosaccharides attenuates allergic Kahle K, Oehme A, et al. Model experiments peritonitis in mice. Biochem Biophys Res mimicking the human intestinal transit Commun. 2012;422(4):546-50. and metabolism of D-galacturonic acid

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 63 CHAPTER 3

and amidated pectin. Mol Nutr Food Res. 2008;52(7):840-8. 38. Eiwegger T, Stahl B, Haidl P, Schmitt J, Boehm G, Dehlink E, et al. Prebiotic oligosaccharides: in vitro evidence for gastrointestinal epithelial transfer and immunomodulatory properties. Pediatr Allergy Immunol. 2010;21(8):1179-88.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 64 GF VERSUS GFA IN CMA

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 65 4 530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 66 IL10r or TGFβ neutralization abrogates the protective effect of a speciﬁ c non-digestible oligosaccharide mixture in cow’s milk allergic mice

JoAnn Kerperien1, Désirée Veening-Griffi oen1,2, Tjalling Wehkamp2, Betty C.A.M. van Esch1,2, Gerard A. Hofman1, Paquita Cornelissen1, Louis Boon3, Prescilla V. Jeurink1,2, Johan Garssen1,2, Léon M.J. Knippels1,2, Linette E.M. Willemsen1

1 Division of Pharmacology, Department of Pharmaceutical Sciences, Faculty of Science, Utrecht University, the Netherlands 2 Nutricia Research B.V., Utrecht, the Netherlands 3 Bioceros B.V., Utrecht, the Netherlands

This chapter is published in: 4 Journal of Nutrition, 2018, 148:1372-1379 530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 67 CHAPTER 4

ABSTRACT

Background: Dietary non-digestible short-chain Galacto-, long-chain Fructo- and pectin- derived Acidic-oligosaccharides (GFA) lower the effector response in cow’s milk allergic (CMA) mice and Foxp3+ regulatory T-cells (Treg) were shown to contribute to this.

Objective: The aim of this study was to assess the contribution of IL10 and TGFβ in the protective effect of the GFA diet in CMA mice.

Methods: Three-four weeks old female C3H/HeOuJ mice were orally sensitized with cholera toxin (Sham) or whey and cholera toxin (Whey) once a week for five consecutive weeks and challenged with whey one week later. The mice were fed a control or 1% GFA (9:2:1) (Whey+GFA) diet starting two weeks prior to the first sensitization. In a second experiment the mice were also injected with αIL10r, αTGFβ or isotype control antibodies 24 hours prior to each sensitization. The acute allergic skin response, anaphylaxis score, whey-specific IgE, mucosal mast cell protease 1 (mMCP1), intestinal Foxp3, Il10 and Tgfβ mRNA expression and Treg frequency in the mesenteric lymph nodes (MLN) were determined.

Results: In Whey+GFA mice intestinal Il10, Tgfβ or Foxp3 mRNA expression was 2-10 times higher (p<0.05), and the MLN Treg frequency was 25% higher compared to Whey mice (p<0.05). The acute allergic skin response was 50% lower in Whey+GFA compared to Whey mice (p<0.01), and IL10r or TGFβ neutralizing antibodies prevented this protective effect (p<0.001). The Whey mice had higher serum mMCP1 concentrations and whey-IgE levels compared to Sham mice (p<0.01), whereas these were not higher in Whey+GFA mice and neutralizing antibodies partially interfered with these responses.

Conclusions: Dietary GFA enhance Treg frequency in the MLN and mucosal IL10 and TGFβ transcription, while suppressing the allergic effector response. Neutralizing antibodies showed that the allergy protective effect of the GFA diet was mediated by IL10 and TGFβ in CMA mice.

Abbreviations Sham Sham sensitized mice fed the control diet Whey Whey sensitized mice fed the control diet Whey+GFA Whey sensitized mice fed the GFA diet Whey+isotype Whey sensitized mice fed the control diet treated with isotype Whey+GFA+isotype Whey sensitized mice fed the GFA diet treated with isotype Whey+αIL10r Whey sensitized mice fed the control diet treated with αIL10r Whey+GFA+αIL10r Whey sensitized mice fed the GFA diet treated with αIL10r Whey+αTGFβ Whey sensitized mice fed the control diet treated with αTGFβ Whey+GFA+αTGFβ Whey sensitized mice fed the GFA diet treated with αTGFβ

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 68 IL10 AND TGFβ WITH GFA IN CMA

INTRODUCTION

Cow’s milk allergy (CMA) is one of the most common food allergies among young children, affecting 1-3% of infants worldwide (1-3). Most of these children acquire tolerance against cow’s milk proteins between the age of 3-5 (2, 4, 5). However, children who have or had CMA are predisposed to develop other food allergies or asthma later in life (6, 7). There are many different approaches to support tolerance induction, which includes the use of specific dietary components to prevent the development of the allergic response to cow’s milk protein. The effectiveness of a mixture of short-chain Galacto-, long-chain Fructo- and pectin-derived Acidic-oligosaccharides (GFA) has been studied in mice sensitized for cow’s milk protein whey (1, 4). These oligosaccharides structurally and functionally mimic specific aspects of oligosaccharides present in human milk and can have a positive effect on the immune system and the growth of beneficial bacteria in the intestine. They can partially prevent CMA symptoms in mice and were shown to reduce the risk of developing atopic dermatitis in infants (8-11). 4

Acquiring oral tolerance occurs through deletion, anergy and/or active suppression of T cells (7, 12). Dendritic cells (DC) take up a food particle and present it to naive T cells in the Peyer’s Patches (PP) or mesenteric lymph nodes (MLN), from where the activated T cells home back to the lamina propria (LP) in the small intestine. Under normal circumstances anergy or deletion results in silencing or loss of T cells having recognition for epitopes from the food protein. Alternatively, various regulatory T cells (Treg) develop and these specific Treg are important to acquire oral tolerance since they induce hypo responsiveness via active suppression (12-15). If acquiring tolerance is disturbed, it can lead to an allergic reaction. Hadis et al. showed that deleting Forkhead Box P3 (Foxp3)+ Treg breaks tolerance in an ovalbumin (OVA)-mouse model (16). This break in tolerance was also observed when Treg could not home towards the gut (16). In addition, transfer of splenic Treg from CMA mice fed a specific oligosaccharide diet was shown to protect control diet fed recipient mice from developing allergic symptoms when injected prior to sensitization with whey protein (17). This implicates an important role for this cell type in acquiring tolerance to food proteins and its involvement in allergy protection induced by dietary oligosaccharides.

When DC skew naïve T cells towards a regulatory phenotype to induce active peripheral tolerance, cytokines interleukin (IL)10 and transforming growth factor beta (TGFβ) play an important role (18). DC produce TGFβ in the MLN and together with retinoic acid (RA) induce Foxp3+ Treg (19, 20). Furthermore, Treg secrete TGFβ as well as IL10 to inhibit proliferation and activation of other immune cells (20). Both IL10 and TGFβ can inhibit the development and cytokine production of T helper cells and cytokine production from mast cells (21). For a more complete overview of the functions of IL10 and TGFβ the reader is referred to Taylor et al. (22) and du Pré and Samson (23).

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 69 CHAPTER 4

To establish the role of IL10 and TGFβ in tolerance versus food allergy various studies were performed. It was shown that cultured splenocytes of orally tolerized mice produce significantly higher levels of IL10 than splenocytes of allergic mice (24). Furthermore, injecting IL10 before applying a specific contact allergen inhibited the allergic response (25). For TGFβ it was shown that adding TGFβ to formula milk led to a beneficial Th1 immune profile in a CMA rat model during suckling and even after weaning when the rats were re-challenged (26). In addition, it was revealed that orally administrated TGFβ in an OVA-allergic mouse model is active in the intestine and enhances the induction of oral tolerance (27). Also orally administrated TGFβ after weaning could prolong beneficial effects of breast milk in OVA-challenged mice (28). Hence, there are several indications that IL10 and TGFβ are important in tolerance induction against food particles.

Currently it is not known whether the preventive effect of non-digestible oligosaccharides in CMA occurs through indirect interactions with the intestinal microbiome, direct cell interactions with epithelial or immune cells and/ or interactions via soluble factors like IL10 and/ or TGFβ locally in the intestine. From a previous experiment using GFA in a similar setting we have a strong indication that GFA could act via either IL10 or TGFβ (29). In the latter study higher intestinal Il10 or Tgfβ mRNA expression was observed in the GFA fed allergic mice.

The aim of the current study was to assess the contribution of IL10 and TGFβ to the protective effect of the GFA diet in CMA. Therefore TGFβ or the receptor of IL10 (IL10r) was neutralized via specific antibody treatment prior to each oral sensitization to determine if this could abrogate the protective effect of the GFA diet.

MATERIALS AND METHODS

Diet Semi-synthetic cow’s milk protein free AIN-93G-based diet (milk proteins were replaced with soy proteins) was composed and mixed with isocaloric supplementation of non-digestible oligosaccharides by Research Diet Services (Wijk bij Duurstede, The Netherlands). One percent of a mixture of scGOS, lcFOS and pAOS (GFA; 75.0%: 16.7%: 8.3% respectively) was added to the diet (17).

Animal model Three to four- week-old specific pathogen-free female C3H/HeOuJ mice, bred for at least two generations on a cow’s milk free diet, were purchased from Charles River Laboratories (Saint Germain Nuelles, France). Mice were fed the control diet or the GFA diet starting directly at arrival for two weeks prior to and during oral sensitization using cholera toxin and whey protein

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 70 IL10 AND TGFβ WITH GFA IN CMA

(WPC60; Milei, Leutkirch, Germany). Mice were orally sensitized via gavage once a week for five consecutive times and at day 33 the mice were anesthetized and the acute allergic skin response and anaphylactic symptom scores were measured 30 minutes after intradermal whey challenge in the ear as described previously (8). Mice were given an oral challenge at day 34 and were killed 18 hours later via terminal bleeding under isoflurane/ air anaesthesia followed by cervical dislocation. Blood was collected and serum was stored at -20ºC until measurement of mouse mast cell protease 1 (mMCP1) (enzyme-linked immunosorbent assay (ELISA) from Ebiosciences, San Diego, California, USA) and whey-specific immunoglobulin E (IgE) as described previously (30). Animal procedures were approved by an independent ethics committee for animal experimentation (Animal Ethics Committee of Utrecht University, Utrecht, The Netherlands) and complied with the principles of good laboratory animal care following the European Directive for the protection of animals used for scientific purposes. The group size in experiment 1 was calculated using 2*(POWER((Za+Zb)/2))*POWER(variation/2))/ POWER(difference/2), with Za=1.96 and Zb=1.28, using 17% for variation and 25% for the 4 difference. The statistical power was calculated based on the expected result for the acute skin response. In the experimental setup for experiment 1 (Figure 2) no antibody treatment was used and power calculations which would allow significant differences between the groups were accepted at N=6.

Experiment 2 was performed according to the same protocol, however in addition in this experiment mice were injected intraperitoneally (i.p.) with either Rat IgG serum isotype control (Sigma-Aldrich, Zwijndrecht, the Netherlands), or αIL10 receptor antibody (Biolegend, San Diego, CA USA) or an αTGFβ antibody (200 μgram per mouse), 24 hours prior to each oral sensitization (Figure 1). The TGF-B hybridoma (clone 1D11) was cultured in IMDM containing 1% of FCS and gentamycin. Clarified supernatant was used to purify the antibody using affinity chromatography. The antibody was sterile preserved in PBS. For these studies (Figures 3-7, OSM supplemental Figure 1), N=10 mice per group (depending on the treatment) were used since the statistical power was calculated based on the same expected difference in the acute skin response, but due to the use of antibodies a larger variation was accounted for. Due to the study size this study was performed in two cohorts which alternated one week and data were combined. For the acute skin response an additional cohort of n=4-6 was analysed and added to the dataset (Figure 1). All mice were housed under SPF conditions, fed ad libitum with unlimited access to water. Mice were housed with n= 6-10 per cage, cages were enriched with shelter and nesting material and kept at reversed light/dark cycle.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 71 CHAPTER 4

Control diet or GFA non-digestible oligosaccharide diet

-14 -7 0 7 14 21 28 33 34/35 days

Oral challenge/ Sensitization by oral gavage killed with CT (sham) or whey+CT

Acute allergic skin response: intradermal with whey in ear Intraperitoneal injection with isotype, αIL10r or αTGFβ

Figure 1. Schematic overview of the experimental designs. The first experiment (data Figure 2 and 3) was performed without injections of antibodies, while the second experiment (data Figure 4, 5, 6, and 7 and supplementary Figure 1) included antibody injections. The intraperitoneal injection protocol is shown in the grey box written in black. CT, cholera toxin; GFA, short-chain Galacto-oligosaccharides/ long-chain Fructo-oligosaccharides/ pectin-derived Acidic-oligosaccharides; Il10, interleukin 10; Tgfβ, transforming growth factor beta; α, anti.

Flow Cytometry MLN were removed, placed on ice in RPMI1640/Penicillin-Streptomysin1% (PS), processed through a 70 nm filter to form single cell suspension and blocked for 30 minutes in PBS containing 1% BSA and 5% heat inactivated fetal bovine serum (FBS). One million cells per well were incubated at 4º C for 30 minutes with CD8a-APC-Cy7, CD11c-PerCP-Cy5.5 and CD25-Pe- Cy7 were used from BD biosciences (San Jose, CA, USA) and CD4-PerCP-Cy5.5, CD103-APC and Foxp3-APC from Ebiosciences (San Diego CA, USA) or matching isotype controls. Extracellular stained cells were fixed using 2% paraformaldehyde and intracellular staining was performed according to the manufacturer’s instructions (eBiosciences, Breda, the Netherlands). The flow cytometry data are shown as a percentage of the control data (Sham).

Immunohistochemistry Swiss rolls of the proximal small intestine were fixed in neutral 10% formalin for at least 24 hours and paraffin sections were embedded using a LeicaTP1020. Tissue processing and Foxp3 staining was performed according to van den Elsen et al. (31). Foxp3 positive cells were counted only in completely attached villi and crypts, and the number per villus/crypt or crypt unit was calculated.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 72 IL10 AND TGFβ WITH GFA IN CMA

qPCR One centimetre of the proximal and distal small intestine (s.i.) and colon was collected shortly after killing and stored in RNAlater™ (Qiagen GmbH, Hilden, Germany) at 4°C until further processing, described previously by Kerperien et al. (29). Validated primers for Ribosomal protein S13 (Rps13), Foxp3, Tgfβ1 and Il10, were purchased from SAbioscience (Qiagen, German Town, MD, USA). mRNA levels were calculated with CFX Manager software (version 1.6) and corrected for the expression of Rps13 with 100 x 2(Rps13-gene of interest) as described previously (32).

Statistics For experiment 1, a multiple comparison test of the whole data set was performed using a one- way ANOVA and Bonferroni post hoc test to correct for multiple comparisons (Graphpad Prism software (version 6)). For experiment 2, all data except for the anaphylaxis score were analysed with the One-way ANOVA and Bonferroni post hoc test with pre-selected pairs (Graphpad Prism software (version 6)). The pre-selected pairs were: Sham vs all other groups, Whey vs 4 Whey-GFA or Whey-αIL10r or Whey-αTGFβ; Whey-GFA versus Whey-GFA-αIL10r or Whey- GFA-αTGFβ; Whey-αIL10r versus Whey-GFA-αIL10r or Whey-αTGFβ; Whey-GFA-αIL10r versus Whey-GFA-αTGFβ; and Whey-αTGFβ versus Whey-GFA-αTGFβ. If required, LOG transformation was used to normalize data distribution. The anaphylaxis score (non-parametrical data, scores are defined step-wise from 0-4) was evaluated with the Kruskal-Wallis and Dunn’s post hoc test. P-values < 0.05 were considered significant and data is shown as mean ± SEM.

RESULTS

Acute allergic skin response and intestinal qPCR analysis (Experiment 1) The acute allergic skin response was determined in experiment 1 in sham sensitized mice fed the control diet (Sham), whey sensitized mice fed the control diet (Whey) and whey sensitized mice fed the GFA diet (Whey+GFA). The significantly higher acute skin response in the Whey compared to Sham group (p<0.0001) was reduced in the Whey+GFA group (p<0.0001) (Figure 2). In the proximal part of the small intestine, only in Whey+GFA mice, the relative expression of Il10 and Tgfβ mRNA was significantly higher compared to Sham mice (p<0.05) (Figure 3A- C). In the distal ileum and colon Foxp3, Il10, and Tgfβ mRNA were significantly higher in the Whey+GFA compared to the Whey and/or Sham group (p<0.05) (Figure 3D-I). One animal was lost during the study and some intestinal samples did not yield enough RNA, therefore the sample size became N= 4 to 6.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 73 CHAPTER 4

Figure 2. Acute allergic skin response in sham sensitized mice fed a control diet or in whey sensitized mice fed a control or a GFA diet (experiment 1). The acute allergic skin response was measured one hour after intra-dermal (i.d.) injection with whey in the ears. Values are means ± SEM, n = 5 – 6. Labelled means without a common letter differ, P<0.05. Δ Indicates ear thickness before i.d. injection deducted from thickness after i.d. injection. GFA, short-chain Galacto-oligosaccharides/ long-chain Fructo- oligosaccharides/ pectin-derived Acidic-oligosaccharides.

Figure 3. mRNA markers in the intestine in sham sensitized mice fed a control diet or whey sensitized mice fed a control or a GFA diet (experiment 1). In the proximal small intestine (A-C), distal small intestine (D-F), and proximal colon (G-I) the relative expression of consecutively transcription factor Foxp3, Il10 and Tgfβ relative to Rps13 is shown. Values are means ± SEM, n = 3 - 6. Labelled means without a common letter differ, P<0.05. S.I., small intestine; Foxp3, forkhead box P3; Il10, interleukin 10; Tgfβ, transforming growth factor beta; Rps13, ribosomal protein S13; GFA, short-chain Galacto-oligosaccharides/ long-chain Fructo-oligosaccharides/ pectin-derived Acidic-oligosaccharides.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 74 IL10 AND TGFβ WITH GFA IN CMA

Contribution IL10 and TGFβ to the allergy protective effect of GFA (Experiment 2)

In the next experiment the effects of IL10r or TGFβ neutralizing antibodies were studied and Sham mice were injected with an isotype antibody (Sham+isotype). Mice fed allergic control diet treated with isotype (Whey+isotype) showed a significantly higher acute allergic skin response compared to Sham+isotype mice (p<0.001), which was suppressed by the GFA diet (p<0.05) (Figure 4A). The significantly lower acute allergic skin response in mice fed the GFA diet was abolished when Whey+GFA mice were treated with either αIL10r (Whey+GFA+αIL10r) or αTGFβ (Whey+GFA+αTGFβ) (p<0.001) (Figure 4A). For the anaphylactic symptom score only Whey+isotype mice showed more shock compared to Sham+isotype mice (p<0.01) (Figure 4B). The anaphylactic symptom score did not differ between the Whey+GFA+isotype mice and Whey+GFA+αIL10r or Whey+GFA+αTGFβ mice. In five mice no ear thickness measurement could be obtained (1 in Whey, 1 in Whey+GFA, 2 in Whey+αTGFβ, 1 in Whey+GFA+αTGFβ) these were removed from the analysis. 4

Figure 4. Clinical symptoms in sham and whey sensitized mice fed a control or GFA diet treated with isotype control or αIL10r or αTGFβ antibody (experiment 2). A) The acute allergic skin response was measured one hour after intra-dermal (i.d.) injection with whey in the ears. Δ Indicates ear thickness before i.d. injection deducted from thickness after i.d. injection. Values are means ± SEM, n=10 - 16, compiled from three cohorts and data were analysed with a one-way ANOVA and Bonferroni post hoc with pre-selected pairs.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 75 CHAPTER 4

B) The anaphylactic symptom scores of sham-sensitized mice fed control diet and whey-sensitized mice fed control- or GFA diet treated with an isotype, αIL10r or αTGFβ antibody, n = 10. Data were analysed with the Kruskal-Wallis and Dunn’s post hoc test. * Different from Sham, P < 0.05, # different from Whey+GFA, P < 0.05. GFA, short-chain Galacto-oligosaccharides/ long-chain Fructo-oligosaccharides/ pectin-derived Acidic-oligosaccharides; IL10, interleukin 10; TGFβ, transforming growth factor beta.

Mouse mast cell protease-1 and whey specific IgE (Experiment 2) In the Whey+isotype mice serum mMCP1 concentrations were higher compared to the Sham+isotype group (p<0.01) (Figure 5A). Serum mMCP1 was not higher in the Whey+GFA mice, while it was higher in the Whey+GFA mice treated with αIL10r or αTGFβ compared to Sham+isotype mice (p<0.01) (Figure 5A). Whey-specific IgE was higher in Whey+isotype mice (p<0.01) and not in Whey+GFA+isotype mice (Figure 5B). However, in the Whey+GFA+αTGFβ mice whey-specific IgE levels tended to be significantly higher compared to Sham+isotype mice (p=0.06) (Figure 5B). The results for whey-specific IgE remain inconclusive in Whey+GFA+αIL10r mice. In five mMCP1 serum samples (3 in Whey, 2 in Whey+GFA, 3 in Whey+αTGFβ) and two whey-IgE serum samples (1 in Whey, 1 in Whey+αTGFβ) of whey sensitized mice the ELISA signal remained below detection and therefore these were removed from the analysis.

CD103+ dendritic cells and regulatory T cells in the mesenteric lymph nodes (Experiment 2) The relative percentage of CD11c+CD103+ DC in Whey+αIL10r or Whey+GFA+αIL10r mice was higher compared to this DC subset in Whey+isotype mice in the MLN (Figure 6A). The relative percentage of CD4+CD25+Foxp3+ Treg was significantly higher in Whey+GFA+isotype compared to Sham+isotype mice (p<0.05) (Figure 6B). The Treg in Whey+αIL10r or Whey+GFA+αIL10r mice was significantly higher compared to Sham+isotype mice (p<0.01). For both the CD103+ DC and the Foxp3+ Treg flow cytometry dot plots are shown for the Whey+isotype and the Whey+αIL10r group (Supplemental Figure 1).

Foxp3+ cells in the lamina propria of the small intestine (Experiment 2) After DC have activated T cells in the MLN, these T cells can be instructed to migrate towards the intestinal LP, therefore Foxp3+ cells were stained in the intestinal LP (Figure 7). In the proximal part of the small intestine of the Whey+GFA+isotype mice the Foxp3+ counts were not significantly higher than Sham+isotype mice. However, significantly greater numbers of Foxp3+ cells were counted in both the Whey+αIL10r as well as the Whey+GFA+αIL10r group compared to Sham+isotype mice (p<0.05) (Figure 7).

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 76 IL10 AND TGFβ WITH GFA IN CMA

Figure 5. mMCP1 and whey-specific IgE measured in serum of sham and whey sensitized mice fed a control or GFA diet with isotype control or αIL10r or αTGFβ antibody (experiment 2). A) mMCP1 (ng/ mL) and B) whey-specific-IgE (AU) were measured in serum. Values are means ± SEM, n= 7-10. Data were analyzed with a one-way ANOVA and Bonferroni post hoc test with pre-selected pairs, if required, LOG transformation was used to normalize data distribution. The circles represent outliers. * Different from sham P < 0.05. GFA, short-chain Galacto-oligosaccharides/ long-chain Fructo-oligosaccharides/ pectin-derived Acidic-oligosaccharired; mMCP1, mouse mast cell protease 1; IgE, immunoglobulin E; AU, arbitrary units; IL10, interleukin 10; TGFβ, transforming growth factor beta.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 77 CHAPTER 4

Figure 6. Percentage of CD103+ dendritic cells and Foxp3+CD25+ regulatory T cells in the MLN of sham and whey sensitized mice fed a control or GFA diet treated with isotype control or αIL10r or αTGFβ antibody (experiment 2). MLN were stained for A) DC (CD11c-PERCP-Cy5.5+ and CD103-APC+) and B) Treg cells (CD4-PERCP-Cy5.5+ CD8a-APC-Cy7-, Foxp3-APC+ CD25-PE-Cy7+). In fi gure 5A and 5B percentages relative to control are shown. Results are shown in percentage relative to mean of sham sensitized mice fed control diet injected with the isotype antibody which was set to 100%. Values are means ± SEM, n= 9-10. and were analysed with a one-way ANOVA and Bonferroni post hoc test with pre-selected pairs. *Different from Sham, P < 0.05, & different from Whey+αIL10r, P < 0.05, % different from Whey+GFA+αIL10r, P < 0.05. MLN, mesenteric lymph nodes; GFA, short-chain Galacto-oligosaccharides/ long-chain Fructo- oligosaccharides/ pectin-derived Acidic-oligosaccharides; IL10, interleukin 10; TGFβ, transforming growth factor beta; Foxp3, Forkhead Box P3; CD, cluster of Differentiation. Representative fl ow cytometric plots are shown with actual percentages in the plots in Supplemental Figure 1.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 78 IL10 AND TGFβ WITH GFA IN CMA

A 8

& & % * * 6 c e l l s + 4 F o x p 3 2 n(villus xcrypt) in ileum 0

t r e a t m e n t sham whey whey whey whey whey whey

d i e t - - GFA - GFA - GFA

a n t i b o d y i s o t y p e a I L 1 0 r a TGF b

B Whey sensitized + IgG isotype C Whey sensizited + αIL10r 4

Figure 7. Foxp3+ cells in the proximal part of the small intestine (experiment 2). A) IHC staining of Foxp3+ cells in the fi rst part of the intestine of sham and whey sensitized mice fed a control or GFA diet and treated with isotype control or αIL10r or αTGFβ antibody, B) in whey sensitized mice fed control diet treated with isotype and C) in whey sensitized mice fed control diet treated with αIL10r. Values are mean ± SEM, n = 3-4, represented as Foxp3+ n/(villus x crypt). * Different from Sham, P < 0.05, & different from Whey+αIL10r, P < 0.05, % different from Whey+GFA+αIL10r, P < 0.05. GFA, short-chain Galacto-oligosaccharides, long-chain Fructo-oligosaccharides, pectin-derived Acidic-oligosaccharides; IL10, interleukin 10; TGFβ, transforming growth factor beta; Foxp3, Forkhead Box P3; arrowhead indicates positive intracellular staining for Foxp3.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 79 CHAPTER 4

DISCUSSION

In the current study the contribution of IL10 and TGFβ in the protective effect of dietary intervention with GFA in CMA prevention was investigated. We and others already have shown that GFA can effectively reduce allergic symptoms (8, 29, 33, 34). Since it was also shown that Treg play a role in this CMA protective effect, we aimed to study the contribution of soluble factors IL10 and TGFβ. IL10 and/or TGFβ support the development of Treg, are produced by Treg and contribute to their suppressive effect (17, 35). In humans affected with CMA lower levels of IL10 and/or TGFβ are produced by lymphocytes isolated from peripheral blood after in vitro stimulation with cow’s milk (36-38).

In our study, dietary intervention with GFA reduced the acute allergic skin response, while increasing Foxp3+ Treg in the MLN. By means of qPCR it was observed that GFA enhanced levels of IL10 and TGFβ along the intestinal tract of the CMA mice. IL10 and TGFβ are regulatory cytokines that are important for oral tolerance induction and proper immune function. The protective effect of GFA on the acute allergic skin response in the CMA model was abolished upon injection with an αIL10r or αTGFβ antibody (p<0.001) and a similar pattern was shown for the anaphylactic symptom score although statistical significance was not attained. Anti-IL10r or -TGFβ treatment did not negatively affect the acute allergic skin response, whey specific IgE levels or the anaphylactic symptom scores in control diet fed CMA mice.

IL10 and TGFβ are amongst others derived from Treg, and the frequency of Foxp3+ Treg in the MLN of GFA fed CMA mice was enhanced. However, the αIL10r treatment itself resulted in a higher percentage of Foxp3+ Treg in the MLN and proximal part of the small intestine in both the control as well as GFA fed CMA mice (p<0.05). This unexpected finding may have several explanations. In previous studies it was shown that Treg derived from αIL10 antibody exposed co-cultures of DC and Treg or Treg derived from IL10-/- KO mice could both not suppress effector T cell activation (39). Hence, Treg that develop in absence of IL10 were phenotypically normal but had lost their suppressive function. However, Treg are also known to develop in the absence of IL10 which is dependent on TGFβ (40). Indeed, in αIL10r treated mice the population of CD103+ DC reached a higher percentage and TGFβ is known to contribute to the development of these cells which are involved in the generation of Foxp3+ Treg (19).

Beyond the possibility that the Treg have lost their suppressive function upon αIL10r treatment, it also may have blocked the IL10r on the target effector cells making them non-responsive for the effect of IL10. Indeed, mucosal mast cell derived mMCP1 was higher in whey sensitized mice fed the control diet either or not treated with an αIL10r and in allergic mice fed the GFA diet and treated with αIL10r antibody (p<0.05). However, mMCP1 concentrations were not higher in mice fed GFA. This suggests that IL10 produced by cells generated by the GFA diet

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 80 IL10 AND TGFβ WITH GFA IN CMA

could be involved in suppression of mast cell degranulation. IL10 is known to suppress murine and human mast cell IgE receptor expression and signalling in vitro and in vivo in mice (41). Hence, IL10 is suggested to be functionally involved in the protective effect of the GFA diet and may suppress effector cell degranulation.

The αTGFβ antibody abolished the protective effect of the GFA diet on the acute allergic skin response in CMA mice (p<0.001). CMA mice fed the GFA diet had low whey-IgE and mMCP1, while neutralization of TGFβ in these mice resulted in significantly higher mMCP1 and a same tendency was shown for whey-IgE comparable to those of control diet fed whey allergic mice (p<0.05). Thus, TGFβ may be involved in the protective effect of the GFA diet on whey sensitization and mast cell degranulation as well. However, this result remains inconclusive since αTGFβ treatment in control diet fed whey sensitized mice also resulted in low whey-IgE and mMCP1 levels. 4 The acute allergic skin response reflects effector cell degranulation, such as mast cells in the ear. The effect of TGFβ on mast cells depends on timing and other cytokines present. TGFβ can act as chemoattractant for mast cells and enhance extracellular release of mMCP1 together with IL3 and IL9, but TGFβ can also inhibit mast cell activation via attenuation of FcεR receptor expression (42-44). Recently TGFβ isoforms -1 and -3 were shown to have suppressive effects on class switching and affinity maturation for IgG and IgA isoforms in human B cells (45). Hence the aTGFβ treatment may have affected the production of IgE directly because TGFβ is known to significantly reduce STAT6 activation in the nucleus of B cells and STAT6 is critical for IgE class switching in B cells (46, 47). This may explain the dual outcome of the findings. The αTGFβ antibody effectively inhibited the protective effect of the GFA diet on acute allergic symptoms, indicating that TGFβ is possibly involved via inhibition of IgE production or mast cell degranulation. By contrast, TGFβ neutralization itself may have contributed to reduced mast cell recruitment and could have led to suppressed whey-IgE generation (48).

Neutralization of the IL10r may have implications for the capacity of TGFβ to instruct Treg, and vice versa neutralization of TGFβ may affect the generation of IL10 secreting Treg. According to Maynard et al. in the small intestine Foxp3-IL10+ Treg cells were most prevalent and were most similar to type 1 Treg (TR1) cells (40). This regulatory subset is dependent on TGFβ for their development. They also show that all IL10+ Treg subsets are dependent on TGFβ for their induction and/ or maintenance (40). In addition to modifying mast cell function, TGFβ is also controlling Treg maintenance and is therefore alike IL10 important to acquire oral tolerance in CMA.

Mice fed the GFA diet had higher Il10 and Tgfβ mRNA levels in intestinal tissue which may associated with the protective effect since neutralization of either one of these mediators results

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 81 CHAPTER 4

in total loss of the protective effect. Currently it is not certain whether GFA enhances intestinal Il10 and Tgfβ mRNA expression directly or indirectly. Eiwegger et al. showed that there is in vitro evidence of epithelial transport of oligosaccharides and that Acidic-oligosaccharides induce IL10 (49). According to Lehmann et al. GF can directly influence DC to produce IL10 and to up regulate functional Treg (50). These are indications that also GFA may regulate DC function directly. However GFA may also modulate the intestinal microbiome hereby contributing to allergy protection.

In summary, in allergic mice fed the GFA diet the percentage of Treg in the MLN and mucosal Il10 and Tgfβ mRNA expression was higher compared to the control diet. Furthermore, the protective effect of the dietary intervention with GFA on allergic symptoms in the murine CMA model can be abrogated via αIL10r or αTGFβ treatment. Considering that Treg are involved in allergy protection (16, 17), this preventive effect of the GFA diet may involve Treg derived IL10 and TGFβ.

Acknowledgements: J.K., L.M.J.K., L.E.M.W. designed research; J.K., D.V.G., T.W., B.C.A.M.E., G.A.H., P.C., conducted research; L.B. provided essential material; J.K., D.V.G., P.C., L.E.M.W. analyzed data or performed statistical analysis, J.K., D.V.G., J.G., L.M.J.K., L.E.M.W. wrote paper or critically read the paper; J.K. and L.E.M.W. had primary responsibility for final content. All authors have read and approved the final manuscript.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 82 IL10 AND TGFβ WITH GFA IN CMA

REFERENCES

1. Huang F, Kim JS. IgE-mediated cow’s milk 11. Nauta AJ, Garssen J. Evidence-based allergy in children. Current allergy and benefits of specific mixtures of non- asthma reports. 2012;12(6):630-40. digestible oligosaccharides on the immune 2. Sanchez-Garcia S, Cipriani F, Ricci G. system. Carbohydr Polym. 2013;93(1):263-5. Food Allergy in childhood: phenotypes, 12. Lepski S, Brockmeyer J. Impact of dietary prevention and treatment. Pediatric allergy factors and food processing on food allergy. and immunology : official publication of the Molecular nutrition & food research. European Society of Pediatric Allergy and 2013;57(1):145-52. Immunology. 2015;26(8):711-20. 13. Akdis M, Verhagen J, Taylor A, et al. Immune 3. Du Toit G, Foong RX, Lack G. The role of responses in healthy and allergic individuals dietary interventions in the prevention of IgE- are characterized by a fine balance between mediated food allergy in children. Pediatric allergen-specific T regulatory 1 and T helper allergy and immunology : official publication 2 cells. J Exp Med. 2004;199(11):1567-75. of the European Society of Pediatric Allergy 14. Vickery BP, Scurlock AM, Jones SM, et al. and Immunology. 2017;28(3):222-9. Mechanisms of immune tolerance relevant 4. Brozek JL, Terracciano L, Hsu J, et al. Oral to food allergy. The Journal of allergy and immunotherapy for IgE-mediated cow’s clinical immunology. 2011;127(3):576-84; 4 milk allergy: a systematic review and meta- quiz 85-6. analysis. Clinical and experimental allergy : 15. Pabst O, Mowat AM. Oral tolerance to food journal of the British Society for Allergy and protein. Mucosal immunology. 2012;5(3):232- Clinical Immunology. 2012;42(3):363-74. 9. 5. Fiocchi A, Schunemann HJ, Brozek J, et al. 16. Hadis U, Wahl B, Schulz O, et al. Intestinal Diagnosis and Rationale for Action Against tolerance requires gut homing and Cow’s Milk Allergy (DRACMA): a summary expansion of FoxP3+ regulatory T cells in the report. J Allergy Clin Immunol.126(6):1119-28 lamina propria. Immunity. 2011;34(2):237- e12. 46. 6. Host A. Frequency of cow’s milk allergy 17. Schouten B, van Esch BC, Hofman GA, et in childhood. Annals of allergy, asthma & al. Oligosaccharide-induced whey-specific immunology : official publication of the CD25(+) regulatory T-cells are involved in American College of Allergy, Asthma, & the suppression of cow milk allergy in mice. Immunology. 2002;89(6 Suppl 1):33-7. J Nutr. 2010;140(4):835-41. 7. Brandtzaeg P. Food allergy: separating 18. Soyer OU, Akdis M, Ring J, et al. Mechanisms the science from the mythology. Nature of peripheral tolerance to allergens. Allergy. reviews Gastroenterology & hepatology. 2013;68(2):161-70. 2010;7(7):380-400. 19. Coombes JL, Siddiqui KR, Arancibia- 8. Schouten B, van Esch BC, Hofman GA, Carcamo CV, et al. A functionally specialized et al. Cow milk allergy symptoms are population of mucosal CD103+ DCs induces reduced in mice fed dietary synbiotics Foxp3+ regulatory T cells via a TGF-beta and during oral sensitization with whey. J Nutr. retinoic acid-dependent mechanism. J Exp 2009;139(7):1398-403. Med. 2007;204(8):1757-64. 9. Roberfroid M, Gibson GR, Hoyles L, et al. 20. Berin MC, Sampson HA. Mucosal Prebiotic effects: metabolic and health immunology of food allergy. Current biology benefits. Br J Nutr. 2010;104 Suppl 2:S1-63. : CB. 2013;23(9):R389-400. 10. Knol J, Scholtens P, Kafka C, et al. Colon 21. Marshall JS, Leal-Berumen I, Nielsen microflora in infants fed formula with L, et al. Interleukin (IL)-10 inhibits long- galacto- and fructo-oligosaccharides: term IL-6 production but not preformed more like breast-fed infants. Journal of mediator release from rat peritoneal mast pediatric gastroenterology and nutrition. cells. The Journal of clinical investigation. 2005;40(1):36-42. 1996;97(4):1122-8.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 83 CHAPTER 4

22. Taylor A, Verhagen J, Blaser K, et al. experimental allergy : journal of the British Mechanisms of immune suppression by Society for Allergy and Clinical Immunology. interleukin-10 and transforming growth 2013;43(7):798-810. factor-beta: the role of T regulatory cells. 32. Garcia-Vallejo JJ, Van Het Hof B, Robben Immunology. 2006;117(4):433-42. J, et al. Approach for defining endogenous 23. du Pre MF, Samsom JN. Adaptive T-cell reference genes in gene expression responses regulating oral tolerance to experiments. Analytical biochemistry. protein antigen. Allergy. 2011;66(4):478-90. 2004;329(2):293-9. 24. Shandilya UK, Kapila R, Singh S, et al. 33. de Kivit S, Saeland E, Kraneveld AD, et al. Induction of immune tolerance to caseins Galectin-9 induced by dietary synbiotics is and whey proteins by oral intubation in mouse involved in suppression of allergic symptoms allergy model. Journal of animal physiology in mice and humans. Allergy. 2012;67(3):343- and animal nutrition. 2014;98(3):467-75. 52. 25. Enk AH, Saloga J, Becker D, et al. Induction 34. Moro G, Arslanoglu S, Stahl B, et al. A of hapten-specific tolerance by interleukin mixture of prebiotic oligosaccharides 10 in vivo. The Journal of experimental reduces the incidence of atopic dermatitis medicine. 1994;179(4):1397-402. during the first six months of age. Arch Dis 26. Penttila I. Effects of transforming growth Child. 2006;91(10):814-9. factor-beta and formula feeding on systemic 35. van den Elsen LW, Meulenbroek LA, van Esch immune responses to dietary beta- BC, et al. CD25+ regulatory T cells transfer lactoglobulin in allergy-prone rats. Pediatric n-3 long chain polyunsaturated fatty acids- research. 2006;59(5):650-5. induced tolerance in mice allergic to cow’s 27. Ando T, Hatsushika K, Wako M, et al. Orally milk protein. Allergy. 2013;68(12):1562-70. administered TGF-beta is biologically active 36. Savilahti EM, Savilahti E. Development of in the intestinal mucosa and enhances natural tolerance and induced desensitization oral tolerance. J Allergy Clin Immunol. in cow’s milk allergy. Pediatric allergy and 2007;120(4):916-23. immunology : official publication of the 28. Rekima A, Macchiaverni P, Turfkruyer European Society of Pediatric Allergy and M, et al. Long-term reduction in food Immunology. 2013;24(2):114-21. allergy susceptibility in mice by combining 37. Pérez-Machado MA, Ashwood P, Thomson breastfeeding-induced tolerance and TGF- MA, et al. Reduced transforming growth beta-enriched formula after weaning. factor-β1-producing T cells in the Clinical and experimental allergy : journal duodenal mucosa of children with food of the British Society for Allergy and Clinical allergy. European journal of immunology. Immunology. 2017;47(4):565-76. 2003;33(8):2307-15. 29. Kerperien J, Jeurink PV, Wehkamp T, et al. 38. Beyer K, Castro R, Birnbaum A, et al. Human Non-digestible oligosaccharides modulate milk–specific mucosal lymphocytes of the intestinal immune activation and suppress gastrointestinal tract display a TH2 cytokine cow’s milk allergic symptoms. Pediatric profile. Journal of Allergy and Clinical allergy and immunology : official publication Immunology. 2002;109(4):707-13. of the European Society of Pediatric Allergy 39. Chattopadhyay G, Shevach EM. Antigen- and Immunology. 2014;25(8):747-54. specific induced T regulatory cells impair 30. Schouten B, van Esch BC, Hofman GA, dendritic cell function via an IL-10/MARCH1- et al. Acute allergic skin reactions and dependent mechanism. J Immunol. intestinal contractility changes in mice 2013;191(12):5875-84. orally sensitized against casein or whey. 40. Maynard CL, Harrington LE, Janowski KM, et International archives of allergy and al. Regulatory T cells expressing interleukin immunology. 2008;147(2):125-34. 10 develop from Foxp3+ and Foxp3- 31. van den Elsen LW, van Esch BC, Hofman GA, precursor cells in the absence of interleukin et al. Dietary long chain n-3 polyunsaturated 10. Nature immunology. 2007;8(9):931-41. fatty acids prevent allergic sensitization 41. Kennedy Norton S, Barnstein B, Brenzovich to cow’s milk protein in mice. Clinical and J, et al. IL-10 suppresses mast cell IgE

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 84 IL10 AND TGFβ WITH GFA IN CMA

receptor expression and signaling in vitro and in vivo. J Immunol. 2008;180(5):2848-54. 42. Li MO, Wan YY, Sanjabi S, et al. Transforming growth factor-beta regulation of immune responses. Annual review of immunology. 2006;24:99-146. 43. Yamazaki S, Nakano N, Honjo A, et al. The Transcription Factor Ehf Is Involved in TGF- beta-Induced Suppression of FcepsilonRI and c-Kit Expression and FcepsilonRI- Mediated Activation in Mast Cells. J Immunol. 2015;195(7):3427-35. 44. Miller HR, Wright SH, Knight PA, et al. A novel function for transforming growth factor-beta1: upregulation of the expression and the IgE-independent extracellular release of a mucosal mast cell granule- specific beta-chymase, mouse mast cell protease-1. Blood. 1999;93(10):3473-86. 4 45. Tsuchida Y, Sumitomo S, Ishigaki K, et al. TGF-beta3 Inhibits Antibody Production by Human B Cells. PLoS One. 2017;12(1):e0169646. 46. Sugai M, Gonda H, Kusunoki T, et al. Essential role of Id2 in negative regulation of IgE class switching. Nature immunology. 2003;4(1):25-30. 47. Okamura T, Sumitomo S, Morita K, et al. TGF-beta3-expressing CD4+CD25(-)LAG3+ regulatory T cells control humoral immune responses. Nat Commun. 2015;6:6329. 48. Halova I, Draberova L, Draber P. Mast cell chemotaxis - chemoattractants and signaling pathways. Frontiers in immunology. 2012;3:119. 49. Eiwegger T, Stahl B, Haidl P, et al. Prebiotic oligosaccharides: in vitro evidence for gastrointestinal epithelial transfer and immunomodulatory properties. Pediatr Allergy Immunol. 2010;21(8):1179-88. 50. Lehmann S, Hiller J, van Bergenhenegouwen J, et al. In Vitro Evidence for Immune- Modulatory Properties of Non-Digestible Oligosaccharides: Direct Effect on Human Monocyte Derived Dendritic Cells. PloS one. 2015;10(7):e0132304.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 85 CHAPTER 4

Supplemental Figure 1. Dot plots of the gating strategy for the CD103+ dendritic cell (DC) population and Foxp3+ CD25+ regulatory T cell population in the MLN (experiment 2). A) After gating a live cell population, the next gate was set for the CD11c+ cell population as shown for pooled MLN cells with the isotype staining and specific CD11c-PERCP-Cy5.5+ antibody, and B) in MLN of whey allergic mice fed control diet and injected with isotype control (whey) or αIL10r (whey + αIL10r). Within the CD11c+ gate the CD103- APC+ gate was set, the isotype control is shown in A). Isotype staining was performed on pooled MLN cells from all animals. Positive cell staining signal in isotype was deducted from specific staining. C) Treg cells were stained as CD4-PERCP-Cy5.5+, CD8a-APC-Cy7-, Foxp3-APC+, CD25-PE-Cy7+. Within the live cell population, the CD4+ CD8a- T-cells were gated and the isotype staining controls are shown in pooled MLN cells. D) In MLN of whey allergic mice fed control diet and injected with isotype control (whey) or αIL10r (whey + αIL10r) within the CD4+CD8a- cell population the Foxp3+ CD25+ gate was set, excluding the Foxp3+ CD25- cell types. Isotype staining was performed on pooled MLN cells of all animals and shown in C). Positive cell staining signal in isotype was deducted from specific staining.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 86 IL10 AND TGFβ WITH GFA IN CMA

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 87 5 530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 88 Alterations in regulatory T cells induced by speciﬁ c oligosaccharides improve vaccine responsiveness in mice

Marcel A. Schijf1,2, JoAnn Kerperien3, Jacqueline Bastiaans2,3, Kirsten Szklany2, Jenny Meerding1, Gerard Hofman3, Louis Boon4, Femke van Wijk1, Johan Garssen2,3, Belinda van’t Land1,2

1 Department of Pediatric Immunology, Wilhelmina Children’s Hospital, University Medical Center, Utrecht 2 Nutricia Research B.V., Utrecht, the Netherlands 3 Division of Pharmacology, Department of Pharmaceutical Sciences, Faculty of Science, Utrecht University, the Netherlands 4 Bioceros B.V., Utrecht, the Netherlands

This chapter is published in: 5 PLoS One. 2013 Sep 20;8(9):e75148 530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 89 CHAPTER 5

ABSTRACT

Prophylactic vaccinations are generally performed to protect naïve individuals with or without suppressed immune responsiveness. In a mouse model for Influenza vaccinations the specific alterations of CD4+CD25+Foxp3+ regulatory T cells (Treg) in the immune modulation induced by orally supplied oligosaccharides containing scGOS/lcFOS/pAOS (GFA) was assessed. This dietary intervention increased vaccine specific DTH responses. In addition, a significant increased percentage of Tbet+ (Th1) activated CD69+CD4+ T cells (p<0.001) and reduced percentage of Gata3+ (Th2) activated CD69+CD4+T cells (p<0.001) was detected in the mesenteric lymph nodes (MLN) of mice receiving scGOS/lcFOS/pAOS compared to control mice. Although no difference in the number or percentage of Treg (CD4+Foxp3+) could be determined after GFA intervention, the percentage of CXCR3+/Tbet+ (Th1-Treg) was significantly reduced (p<0.05) in mice receiving GFA as compared to mice receiving placebo diets. Moreover, although no absolute difference in suppressive capacity could be detected, an alteration in cytokine profile suggests a regulatory T cell shift towards a reducing Th1 suppression profile, supporting an improved vaccination response.

Conclusion: These data are indicative for improved vaccine responsiveness due to reduced Th1 suppressive capacity in the Treg population of mice fed the oligosaccharide specific diet, showing compartmentalization within the Treg population. The modulation of Treg to control immune responses provides an additional arm of intervention using alternative strategies possibly leading to the development of improved vaccines.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 90 GFA AND TREG DURING VACCINATION

INTRODUCTION

The induction of proper immune responsiveness to vaccinations shortly after birth or in immune compromised individuals is challenging. The highly protective environment and need to avoid immunological interactions of the fetus against the mother seem to be the main reason for this “physiological” immaturity of the immune system in new-born infants.

Regulatory T cells (Treg) are particularly abundant and potent during pregnancy and at birth and inhibit excessive immune responses whereby ultimately the maintenance of peripheral T cell tolerance and pathogen clearance are key (1-4). Moreover, the signalling within the intestine through pathogen recognition receptors like the toll like receptor (TLR) family for instance is crucial for the generation of effective immunity. Illustrative for this is that TLR2 influences the function of Treg (5) and establishes a direct link between the intestinal microbiota and the control of immune responses through Treg (6). Treg are important cells involved in immune regulation and play an increasing role in many immune related disorders of which many are found to be related to disturbed Treg function. For instance in human immunodeficiency virus (HIV) disease progression, the up regulation of Forkhead Box p3 (Foxp3) expression in cluster of differentiation (CD)4+ T cells seems to be a marker of disease severity (7) and also 5 in HAART-therapy treatment the Treg activity is associated with persistently reduced CD4+ T cell counts during antiretroviral therapy (8). Even, in chronic diseases like asthma, allergy, cancer an altered Treg number or function has been described. This is exemplified by the increased number of Treg in children with eosinophilic esophagitis and explicit role of Treg in tumor immunity (9). In addition, related towards acute infections and innate immunity, an important role for Treg is the suppression of innate immune pathology during influenza A virus infection (10). Foxp3+CD4+ Treg limit pulmonary immune pathology by modulating the CD8+ T cell responses during respiratory syncytial virus infection (11, 12). These findings contribute to the notification, that next to immune response induction, a regulated suppression is essential for maintaining proper immune balance.

The diversity in immune responses evoked upon pathogen recognition may require several subsets of CD4+Foxp3+ Treg to maintain proper immune homeostasis. There have been important functions reported for T box transcription factor (Tbet) and Interferon regulatory factor 4 (IRF4) in Treg demonstrating that the suppression ability requires the expression of transcription factors typically associated with the effector T cell (Teff) function at place (13). IRF4 is a decisive factor during T helper cell (Th)17 development by influencing the balance of Foxp3, retinoid-related orphan receptor (Ror)α, and Rorγt (14). In response to interferon (IFN)γ, the Treg are found to up regulate Tbet which promotes the expression chemokine receptor c-x-c motif chemokine receptor (CXCR)3, and Tbet+ Treg accumulate at sites of Th1 cell–mediated inflammation. Tbet expression is required for the homeostasis and function of

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 91 CHAPTER 5

Treg cells during type 1 inflammation (15). This subscribes the hypothesis that specific subsets of Foxp3+ Treg develop for the suppression of Th1 responses in vivo (13). Moreover, besides strong stimulation of Teff cells a modulation of Treg to control immune responses provide an additional arm of intervention in the development of improved vaccines.

Early in life establishment of immune responsiveness is influenced by several factors, including nutrition. Breast milk contains several interesting immune modulating components with specific modulating potentials, which are known to have a clear role in immune mediated disease resistance later in life (16). Specific oligosaccharides are known to modulate immune responses, as they can improve the immune balance in infants, resulting in lower incidence of infections and simultaneously can have an impact on allergy related symptoms (17). Prebiotic oligosaccharides can have a direct effect via activation or inhibition of cellular receptors on immune competent cells (18) and may act indirectly through microbiota-dependent mechanisms (i.e. rebalancing microbiota composition in the gut) (19) The pre- and probiotic concept is based on the fact that our microbiota are considered to contribute to induction and maintenance of immune homeostasis possibly via CD25+ Treg (20). More specifically, it was found recently that Treg play a fundamental role in the immune modulation induced by the supplementation of these specific oligosaccharides (21). The exact underlying mechanism by which prebiotic oligosaccharides induce immune modulation effects however remains to be elucidated and is subject of current investigation. Given the importance of Treg in maintenance of immune homeostasis and vaccine efficiency in the host, the specific role of Treg in the immune-modulating effects of dietary supplementation with a unique mixture of prebiotic oligosaccharides short-chain Galacto-oligosaccharides (scGOS)/ long-chain Fructo- oligosaccharides (lcFOS)/ pectin-derived Acidic-oligosaccharides (pAOS) (combined GFA) in a vaccination model has been investigated.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 92 GFA AND TREG DURING VACCINATION

MATERIALS AND METHODS

Animals and diets Eight-week-old old male specific pathogen-free inbred C57BL/6J mice and C57BL/6- Foxp3tm1Flv/J were obtained from Charles River (Someren, the Netherlands) and housed under standard housing conditions with a twelve hours dark and light cycle. All animals had free access to tap water and the semi-purified AIN-93G diet (Research Diet Services, Wijk bij Duurstede, the Netherlands), with or without oligosaccharide mixture consisting of three different prebiotic oligosaccharide materials, i.e. scGOS (Borculo Domo, Zwolle, 45% scGOS), lcFOS (Orafti, Wijchen, 100% lcFOS) and pAOS (Sudzucker, Mannheim, 85% galacturonic acid). The oligosaccharides were mixed in a ratio of 9:1:10 based on carbohydrate purity. Although other combinations of these specific prebiotic oligosaccharides are known to be effective as well, it was this specific ratio within currently used mouse vaccination model which gives the largest immune modulation, detected by a delayed-type hypersensitivity (DTH) increase at the time and was therefore used for these mechanistically studies. A small negative control group of animals (n=3) was included only to show specificity of the vaccination procedure. Therefore, this group was not used for any statistical comparison to the supplemented groups (n=10 per group). Fourteen days prior to the first vaccination dietary supplementation started which was 5 maintained during the entire experimental procedure. The study protocol was reviewed and approved by the Animal Experimental Committee of the Utrecht University (permit number 2011.II.06.102).

Vaccination protocol All mice except the small negative control group received primary vaccination (day 1) and a booster vaccination (day 21) with a human influenza subunit vaccine consisting of haemagglutinin proteins of 3 different influenza strains (Influvac®, Solvay Pharmaceuticals, Weesp, the Netherlands). Vaccinations were performed by subcutaneous injection of vaccine (30 μg/ml per subunit) in a total volume of 100 μL. The negative control group received concurrent injections with PBS in a total volume of 100 μL. Vaccine-specific DTH reactions were induced 7 days after the last vaccination, by subcutaneous injection of 25 μL Influvac (30 μg/mL per haemagglutinin subunit) into the ear pinnae of one ear. For control, the other ear was injected with 25 μL PBS. Ear thickness was measured in duplicate before challenge, and 24 hours thereafter, with a digital micrometre (Mitutoyo Digimatic 293561, Veenendaal, the Netherlands). The influvac specificity of the DTH response was calculated by subtracting the basal ear thickness from the value at 24 hours after challenge and was corrected for the control swelling.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 93 CHAPTER 5

Flowcytometric analysis Splenocytes and cells from the mesenteric lymph nodes (MLN) were isolated by gently pressing the organs through nylon mesh filters (Falcon cell strainer, Becton Dickinson, Alphen a/d Rijn, the Netherlands). After erythrocyte lysis (spleens only), a total of 1x106 cells were washed with PBS containing 1% FBS and incubated for 15 minutes with an anti-mouse CD16/CD32 antibody (BD Pharmingen cat# 553142) blocking the Fc receptors. The cells were then stained with different combinations of anti-CD4-FITC (BD Pharmingen cat# 553046), anti-CD4-PE (BD Pharmingen cat# 553049), CD4-PeCy5 (BD Pharmingen cat# 553654), anti-CD25-PE (Beckman Coulter 732091), anti-CD69-APC (eBioscience cat# 17-0691), anti-CXCR3-APC (eBioscience cat# 17-1831), anti-CD103-PE (eBioscience cat# 12-1031), anti GITR-PE (eBioscience cat# 12-5874), or with anti CTLA-4-PE (eBioscience cat#12-1522) for 30 minutes. Intracellular stainings were performed according manufacturer’s protocol, (eBioscience, Foxp3 staining set, Bio connect, The Netherlands). For intracellular staining the antibodies anti-Foxp3-FITC (eBioscience cat# 11-5773), anti-Gata-3-PE (eBioscience cat# 12-9966), anti-T-bet-PerCP-Cy5 (eBioscience cat# 45-5825) were used in combination with above mentioned surface markers. Matching Isotype controls were used for all staining to minimize the influence of nonspecific binding, and proper gate setting. All staining procedures were performed on ice and protected from light. In total a minimum of 50.000 cells were counted and analyses were performed using FACSCanto II and FACSDiva software (BD Biosciences).

Suppression assay After labelling of MLN and spleen single cell suspensions obtained from C57BL/6-Foxp3tm1Flv/J mice (Charles river) with CD4-FITC and CD25-APC the CD4+CD25+mRFP+ Treg were isolated by flow cytometry using a FACS Aria cell sorter. Sorted Treg were cultured together with 2μ M pacific blue succinimidyl ester (PBSE Invitrogen) labeled total spleen cells (20.000 per well (Teff)) in ratios as indicated in the graphs. Cells were stimulated with 1ug/ml anti-CD3 (BD Pharmingen clone 145-2c11) for 96 hours. at 37°C. Cell culture supernatants were collected and cytokines were measured according to manufacturer’s protocol using Bio-Plex Pro Mouse Cytokine Grp I panel 23-Plex (BioRad). For cell division analysis using PBSE positivity, the cells were stained with CD4-FITC (BD Pharmingen cat# 553046), CD8-APC (BD Pharmingen cat# 553932) and cell division was measured using flowcytometry.

Statistical analysis All statistical calculations were performed using SPSS version 12.0.1 software and Graph Path prism 4.03 software. Statistical differences between test and control groups were analysed by one-way ANOVA followed by multi-group comparison analysis using the Bonferroni test; otherwise an un-paired two-sided t-test was used. Correlations were identified using Spearman and Pearson tests. All values are presented as mean ± SEM. P-values <0.05 were considered significant.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 94 GFA AND TREG DURING VACCINATION

RESULTS

Oligosaccharide induced Tbet / Gata3 differentiation increase the Th1 responsiveness Immune modulation effects of the dietary intervention were analysed by measuring antigen specific DTH responses, representing an in vivo parameter for Th1 type of cellular immunity. In mice receiving GFA, influenza vaccination results significantly (p<0.001) increased influenza-specific DTH responses compared to mice receiving placebo diets (Figure 1) which are comparable with earlier described immune modulation capacities of orally supplied oligosaccharides (22). In addition, the expression of Tbet and Gata3 in the activated (CD69+) CD4+ T-cell population was assessed both in spleen (data not shown) and MLN cells (Figure 2). Neither in the total CD4+ T cell (Figure 2B), nor in the percentage of activated CD69+CD4+ T cell populations (Figure 2C) is a significant change seen between the dietary interventions Interestingly a statistical significant increase in percentage of Tbet positive CD4+CD69+ T cells (p<0.001) could be detected in MLN of mice receiving GFA compared to mice receiving placebo diet (25.41 ± 1.85 placebo vs 41.19 ± 1.92 GFA % of CD4+CD69+ T cells (mean +/- SE)) as depicted in Figure 2D. In addition, a statistical significant reduction (p<0.001) was found in the percentage of Gata3 expressing activated CD69+CD4+ T cells (20.30 ± 1.28 placebo vs 10.66 ± 1.13 GFA % (mean +/- SE)) as depicted in Figure 2E. 5

Figure 1. Dietary intervention alters Flu-specific DTH responses. C57Bl/6J mice (n = 10 per group) received a sub-maximal vaccination and dietary intervention with or without scGOS/lcFOS/pAOS (GFA) during the entire vaccination procedure. Antigen specific DTH responses were measured by ear swelling (24 hours post-antigen injection) and corrected for background (PBS) swelling. A non-vaccinated group (n = 3) was used as control (NV). Lines represent median with interquartile range of the DTH responses from individual mice per group (as indicated through separate dots). The statistical differences are indicated in the graph. DTH, delayed-type hypersensitivity; GFA, short-chain Galacto-oligosaccharides/ long-chain Fructo-oligosaccharides/ pectin-derived Acidic-oligosaccharides; Il10, interleukin 10; Tgfβ, transforming growth factor beta

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 95 CHAPTER 5

Figure 2. Activated CD4+ T cells are modulated towards Th1 type of immune responsiveness. MLN cells were isolated from mice (n = 7 per group) receiving either placebo or GFA and were labelled with CD4/ CD69/Tbet/Gata3 flowcytometric analysis. The characterization of different cell populations is indicated in the gating strategy (A). Lines represent mean CD4+ T cells % (B), activated CD69+CD4+ T cells % (C) % of Tbet+ activated T cells (D) and % of Gata3+ activated T cells (E). In addition, individual measurements are indicated through separate dots. Data presented is representative for 3 individual experiments. Statistically significant differences between the groups are indicated in the graphs. CD, cluster of differentiation; GFA, short-chain Galacto-oligosaccharides/ long-chain Fructo-oligosaccharides/ pectin-derived Acidic- oligosaccharides.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 96 GFA AND TREG DURING VACCINATION

Moreover, a significant (p<0.001) positive linear correlation (r2 = 0,54) was detected between the percentage of activated Tbet+ T cells and in vivo DTH response as well as a significant (p<0.01) negative correlation (r2 = 0,47) between the percentage of activated Gata3+ T cells and DTH response (Figure 3A, B). These data combined are indicative for improved Th1 type of immune responsiveness in C57BL/6J mice towards a fixed antigen dose due to specific dietary oligosaccharides.

Figure 3. Correlation between DTH response and percentage of Tbet+, Gata3+ of activated CD4 T cells in MLN. From individual mice the DTH values were correlated to the percentage of Tbet + (A) or percentage Gata3+ (B) of the activated CD4+ T cells irrespective of dietary intervention, using Pearson and Spearman correlation tests. Mean correlation (line) and 95% CI (dashed) are indicated in the graphs next to the individual data points. With r2 = 0.5438 and r2 = 0.4653 for Tbet and Gata3 respectively these percentages of activated T cells correlate significantly p<0.001 and p<0.01 respectively. CD, cluster of differentiation; DTH, delayed-type hypersensitivity.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 97 CHAPTER 5

Changes within Treg population are in line with increased Th1 responsiveness If the development of Treg are in line with the developed immune response as postulated by Barnes et al. (13), than alterations in immune response should be accompanied with changes in Treg population accordingly. In order to test this hypothesis for the GFA induced Th1 responsiveness, the Treg population was analysed using flowcytometry with surface staining of CXCR3 as additional marker besides Tbet expression (Th1 polarization) (Figure 4A). In mice receiving GFA the percentage of CD4+Foxp3+ T cells did not change in the MLN (Figure 4B) or in the spleen (data not shown). Strikingly however, the percentage of Treg were significantly lower (p<0.05) in the MLN of mice receiving GFA compared to mice receiving placebo diets (2.99 ± 0.53 placebo vs 1.06 ± 0.11 GFA % of Treg (mean +/- SE)) (Figure 4C). Treg (Foxp3+CD4+) cells which are positive for both CXCR3 and Tbet are hypothesized to down regulate an increased Th1 response. Therefore these data are indicative for a reduced Th1 suppressive capacity in the Treg population of mice fed the GFA diet. Although the changed percentages are small some strongly significant immune modulatory changes in Treg population are detected due to specific oligosaccharides in the diet, showing their functional capacity as evidenced by increased DTH responses.

Finally the suppressive capacity of isolated Treg from spleen as well as MLN was analysed. From C57BL/6-Foxp3tm1Flv/J mice the regulatory T cells (characterized as CD4+CD25+mRFP+) were cell sorted and cultured together with PBSE labelled Teff cells for 96 hours. Cell division was observed after CD3 stimulation for 96 hours. Although Treg isolated from both the spleen as well as the MLN suppressed the CD3 induced T cell proliferation, in none of the ratios tested a difference could be observed between the Treg isolated from the mice on the different diets (data not shown). In addition to an overall reduction in cytokines produced with increasing percentage of Treg, a small change in some cytokines was detected between the diets. As shown in Figure 5 no difference in IL2 (proliferation) could be detected, but a significant (p<0.05) increased IFNγ (Th1) as well as significant (p<0.05) reduced IL17 (Th17) response was detected in spleen cell populations suppressed by Treg from mice receiving the specific oligosaccharides. The production of IL13 (Th2), IL10 as well as IL6 (inflammatory), were suppressed with increasing amounts of Treg, but not different between the diets. These changes are in line with improved Th1 type of immune responsiveness in C57BL/6J mice towards a fixed antigen dose due to specific dietary oligosaccharides.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 98 GFA AND TREG DURING VACCINATION

Figure 4. Treg population in MLN is influenced by the dietary intervention. Cells from MLN were isolated from mice (n = 7 per group) receiving either placebo or scGOS/lcFOS/pAOS and were labelled with CD4/ Foxp3 in combination with CXCR3, Tbet, for flowcytometric analysis (A). Lines represent mean percentages of Treg in total (CD4+Foxp3+ T cells) (B), and sub-populations of Treg including % of CXCR3+/Tbet+ Treg (C). In addition, individual measurements are indicated through separate dots. Data presented is representative for 3 individual experiments. Statistically significant differences between the groups are indicated in the graphs. Ssc, side scatter; Foxp3, Forkhead Box P3; CXCR3, c-x-c motif chemokine receptor 3, CD, cluster of differentiation; GFA, short-chain Galacto-oligosaccharides/ long-chain Fructo-oligosaccharides/ pectin- derived Acidic-oligosaccharides.

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 99 CHAPTER 5

Figure 5. Suppression method but not capacity of Treg is influenced by dietary intervention. CD4+CD25+mRFP+ regulatory T cells (Treg) were isolated from MLN of the C57BL/6-Foxp3tm1Flv/J mice (n=4) using FACS Aria cell sorter and cultured with T effector cells (Teff) cells in ratio of 1/1, 1/2, 1/4 (Treg /Teff) for 96 hours after stimulation with α-CD3. Cell culture supernatant contained different cytokines (B) including (IL2, IFNγ , IL13, IL10 and IL17A). Only the significant differences between the diets were indicated using * for p<0.05. Ssc, side scatter; fsc, forward scatter; Foxp3, Forkhead Box P3; IL, interleukin; IFNγ, interferon gamma; CD, cluster of differentiation; GFA, short-chain Galacto-oligosaccharides/ long-chain Fructo- oligosaccharides/ pectin-derived Acidic-oligosaccharides

100

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 100 GFA AND TREG DURING VACCINATION

DISCUSSION

A better understanding of the mechanism by which vaccine induced immune responses can be increased using alternative strategies will improve vaccine development. In our current study we show that specific modulation of the immune response by specific oligosaccharide containing diet results in alteration in Treg population. More specifically, a reduced percentage of Tbet+ Treg were induced in mice fed GFA diet during vaccination, resulting in increased vaccine responsiveness. Although the involvement of Treg in dietary immune modulations has been indicated before (21, 23), this is the first study showing that an alteration of activated CD4+ T cells (increased Tbet (Th1) and reduced Gata3 (Th2) is accompanied with a reduced population Treg expressing Tbet. Moreover, although suppressive capacity does not seem to be altered, the Treg seem to suppress through a different and to hereunto unknown mechanism.

Recent observations have challenged the notification of one stable Treg sub lineage. Some reports suggest that there are multiple, functional Treg subsets. One of these subsets expresses Tbet and has shown to be specifically adapted for the suppression of Th1 responses (15). In addition, another Treg subset expressing IRF4 was identified, which is essential for Th2 differentiation. The absence of IRF4+ Treg even resulted in spontaneous Th2 mediated 5 inflammation, suggesting the necessity of the IRF4+ Treg to control Th2 type of responses (14). Furthermore, a third Treg subtype expressing STAT3 seemed required suppressing Th17 responses (24). Therefore as the plasticity of T cell population seems to require plasticity of Treg to control excessive immune responses, this provides an additional arm of intervention. Within our studies we clearly show increased vaccination responsiveness, by dietary intervention. Both the increased DTH response as well as increased percentage of Tbet expressing T cells is indicative for improved Th1 responsiveness, which is accompanied by alterations within the Treg population.

Previously, we observed a systemic decrease in Treg numbers in lung, spleen, and mesenteric lymph nodes at day 4 after infection in FI-RSV-vaccinated mice receiving the GFA diet. Treg have previously been found to regulate RSV-specific primary immune responses. Within this study the expression of granzyme B (GzmB) in Treg locally in the lung was shown to be involved in the immune-regulatory function of Treg in the primary RSV infection model (25). Although the percentage of Treg did not differ, a significantly decreased absolute number of Treg was detected in the BAL fluid at day 4 after challenge in FI-RSV-vaccinated mice receiving the GFA diet. This indicates that the dietary intervention with GFA has an effect on the function of regulatory immune cells which correlates with altered immune responses.

The changes detected in the Treg population are multiple, they include a reduced percentage of CXCR3+/Tbet+ Treg, GITR+CXCR3+ Treg (data not shown) in the MLN of mice receiving GFA

101

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 101 CHAPTER 5

diet compared to mice fed control diet. Tbet controls migration to inflammatory sites through CXCR3 up regulation as CXCR3 is a direct transcript of Tbet (26). Thus, Tbet may therefore play an important role in the direction and regulation of regulatory T cells. This indicates that it is not the percentage of Treg in total, but merely the functionality which changes due to dietary intervention. Treg also express receptors for inflammatory chemokines c-c motif chemokine receptor (CCR)4, CCR9, and CXCR3, integrin’s, and tissue-homing receptors like CD103 (27). Indeed, it was demonstrated that CD103+ Treg are attracted and retained in inflamed tissues where they may exert their suppressive function (28). Functional compartmentalization of Treg is linked to the expression of different phenotypes, with Treg found in tissues expressing CD103, IL10, IL2R, and CCR5.

In conclusion; this study shows that with dietary intervention using specific oligosaccharides improved vaccine responsiveness can be induced due to reduced Th1 suppressive capacity in the Treg population of mice. The better understanding of the mechanism by which vaccine induced immune responses can be amplified using alternative strategies will improve vaccine development.

102

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 102 GFA AND TREG DURING VACCINATION

REFERENCES

1. Belkaid Y, Rouse BT. Natural regulatory T 9. Nishikawa H, Sakaguchi S. Regulatory cells in infectious disease. Nat Immunol. T cells in tumor immunity. Int J 2005;6(4):353-60. Cancer.127(4):759-67. 2. Marchant A, Goldman M. T cell-mediated 10. Antunes I, Kassiotis G. Suppression of immune responses in human newborns: Innate Immune Pathology by Regulatory T ready to learn? Clin Exp Immunol. Cells During Influenza A Virus Infection of 2005;141(1):10-8. Immunodeficient Mice. J Virol. 3. Schonland SO, Zimmer JK, Lopez- 11. Brandenburg AH, Kleinjan A, van’t Land B, Benitez CM, et al. Homeostatic control et al. Type 1-like immune response is found of T-cell generation in neonates. Blood. in children with respiratory syncytial virus 2003;102(4):1428-34. infection regardless of clinical severity. J 4. Tilburgs T, Roelen DL, van der Mast BJ, et Med Virol. 2000;62(2):267-77. al. Evidence for a selective migration of 12. Fulton RB, Meyerholz DK, Varga SM. Foxp3+ fetus-specific CD4+CD25bright regulatory CD4 regulatory T cells limit pulmonary T cells from the peripheral blood to the immunopathology by modulating the CD8 decidua in human pregnancy. J Immunol. T cell response during respiratory syncytial 5 2008;180(8):5737-45. virus infection. J Immunol.185(4):2382-92. 5. Sutmuller RP, den Brok MH, Kramer M, et 13. Barnes MJ, Powrie F. Hybrid Treg cells: steel al. Toll-like receptor 2 controls expansion frames and plastic exteriors. Nat Immunol. and function of regulatory T cells. J Clin 2009;10(6):563-4. Invest. 2006;116(2):485-94. 14. Zheng Y, Chaudhry A, Kas A, et al. Regulatory 6. Van’t Land B, Schijf MA, Martin R, et al. T-cell suppressor program co-opts Influencing mucosal homeostasis and transcription factor IRF4 to control T(H)2 immune responsiveness: The impact responses. Nature. 2009;458(7236):351-6. of nutrition and pharmaceuticals. Eur J 15. Koch MA, Tucker-Heard G, Perdue NR, et Pharmacol. 2011. al. The transcription factor Tbet controls 7. Suchard MS, Mayne E, Green VA, et al. FOXP3 regulatory T cell homeostasis and function expression is upregulated in CD4T cells in during type 1 inflammation. Nat Immunol. progressive HIV-1 infection and is a marker 2009;10(6):595-602. of disease severity. PLoS One.5(7):e11762. 16. Van’t Land B, Boehm G, Garssen J. Breast 8. Piconi S, Trabattoni D, Gori A, et al. Immune milk: components with immune modulating activation, apoptosis, and Treg activity are potential and their possible role in immune associated with persistently reduced CD4+ mediated disease resistance. Watson RR, T-cell counts during antiretroviral therapy. Zibadi S, Preedy VR, editors2011. Aids.24(13):1991-2000. 17. Arslanoglu S, Moro GE, Schmitt J, et al. Early dietary intervention with a mixture of prebiotic oligosaccharides reduces the

103

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 103 CHAPTER 5

incidence of allergic manifestations and syncytial virus infection model. J Virol. infections during the first two years of life. J 2012;86(21):11472-82. Nutr. 2008;138(6):1091-5. 26. Lord GM, Rao RM, Choe H, et al. Tbet is 18. Vos A, M’Rabet L, Stahl B, et al. Immune- required for optimal proinflammatory CD4+ modulatory effects and potential working T-cell trafficking. Blood. 2005;106(10):3432- mechanisms of orally applied nondigestible 9. carbohydrates. Crit Rev Immunol. 27. Sather BD, Treuting P, Perdue N, et 2007;27(2):97-140. al. Altering the distribution of Foxp3(+) 19. Gori A, Rizzardini G, Van’t Land B, et al. regulatory T cells results in tissue- Specific prebiotics modulate gut microbiota specific inflammatory disease. J Exp Med. and immune activation in HAART-naive HIV- 2007;204(6):1335-47. infected adults: results of the “COPA” pilot 28. Suffia I, Reckling SK, Salay G, et al. A oler for randomized trial. Mucosal Immunol. 2011. CD103 in the retention of CD4+CD25+ Treg 20. Ishikawa H, Tanaka K, Maeda Y, et al. Effect and control of Leishmania major infection. J of intestinal microbiota on the induction of Immunol. 2005;174(9):5444-55. regulatory CD25+ CD4+ T cells. Clin Exp Immunol. 2008;153(1):127-35. 21. Van’t Land B, Schijf M, van Esch BC, et al. Regulatory T-cells have a prominent role in the immune modulated vaccine response by specific oligosaccharides. Vaccine. 2010;28(35):5711-7. 22. Vos AP, Haarman M, van Ginkel JW, et al. Dietary supplementation of neutral and acidic oligosaccharides enhances Th1- dependent vaccination responses in mice. Pediatr Allergy Immunol. 2007;18(4):304-12. 23. Schouten B, van Esch BC, Hofman GA, et al. A potential role for CD25+ regulatory T-cells in the protection against casein allergy by dietary non-digestible carbohydrates. Br J Nutr.2011:1-10. 24. Chaudhry A, Rudra D, Treuting P, et al. CD4+ regulatory T cells control TH17 responses in a Stat3-dependent manner. Science. 2009;326(5955):986-91. 25. Schijf MA, Kruijsen D, Bastiaans J, et al. Specific dietary oligosaccharides increase Th1 responses in a mouse respiratory

104

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 104 GFA AND TREG DURING VACCINATION

105

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 105 6 530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 106 Dietary vitamin D supplementation is ineffective in preventing murine cow’s milk allergy, irrespective of the presence of non-digestible oligosaccharides

JoAnn Kerperien1*, Désirée Veening-Griffi oen1,2*, Anna Oja1, Tjalling Wehkamp2, Prescilla V. Jeurink1,2, Johan Garssen1,2, Leon M.J. Knippels1,2, Linette E.M. Willemsen1

ABSTRACT

Background: Cow’s milk allergy (CMA) is one of the most common food allergies especially early in life. A mixture of non-digestible short-chain Galacto-oligosaccharides (scGOS), long- chain Fructo-oligosaccharides (lcFOS), and pectin-derived Acidic-oligosaccharides (pAOS) (GFA) may reduce allergy development and allergic symptoms in murine CMA. Recently vitamin D (VitD) has been suggested to have beneficial effects in reducing allergy as well. In this study the effect on allergy prevention using the combination of these two different immune modulating pathways was investigated.

Methods: Female C3H/HeOuJ mice were fed a control or GFA containing diet with depleted, standard (1000 IU/kg) or supplemented (5000 IU/kg) VitD for two weeks before and during whey sensitization (n=10-15). Mice were sensitized five times intragastrically with PBS as a control, whey as cow’s milk allergen and/or cholera toxin (CT) as adjuvant on a weekly interval. One week after the last sensitization mice were intradermally challenged in both ear pinnae and orally with whey. After 18h terminal blood samples, mesenteric lymph nodes (MLN) and spleens were collected. Whey-specific IgE, IgG1 and IgG2a levels, and mucosal mast cell protease1 (mMCP1) concentrations were measured by means of standardized ELISA. T cell subsets and dendritic cells were studied using flow cytometry.

Results: CMA mice fed the GFA diet supplemented with VitD (GFA VitD+) significantly decreased the acute allergic skin response of whey sensitized mice when compared to the CMA mice fed VitD (VitD+) group indicating the potency of these oligosaccharides to suppress allergic symptoms (p<0.05). The effect of GFA was not improved by extra VitD supplementation even though the CMA mice fed the GFA VitD+ diet had a significantly increased percentage of CD103+ DCs compared to the VitD+ group (p<0.05). The VitD deprivated mice showed a high percentage of severe shock and many reached the Humane Endpoint, therefore these groups were not further analysed.

Conclusions: High dose vitamin D supplementation in mice does not protect against CMA development in presence or absence of GFA.

Abbreviation list sham sham sensitized mice fed control diet allergic mice whey sensitized mice fed control diet allergic mice fed VitD+ whey sensitized mice fed control diet with vitamin D supplementation allergic mice fed GFA whey sensitized mice fed scGOS/lcFOS/pAOS diet allergic mice fed VitD+/GFA whey sensitized mice fed scGOS/lcFOS/pAOS diet with vitamin D supplementation

108

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 108 VITD AND GFA IN CMA

INTRODUCTION

Food allergy occurs when the immune system incorrectly identifies a food protein as a threat and triggers an immune response, causing sensitization and upon re-exposure, allergic symptoms. Cow’s milk allergy (CMA) is one of the most common food allergies in children (1-3% is affected), symptoms can range from gastrointestinal to dermatological to respiratory manifestations and in the worst case might cause life-threatening anaphylactic shock reactions (1, 2).

Non-digestible oligosaccharides (NDO) resembling structural and functional aspects of human milk oligosaccharides have been developed, such as short-chain Galacto- oligosaccharides (scGOS), long-chain Fructo-oligosaccharides (lcFOS), and pectin-derived Acidic-oligosaccharides (pAOS) (GFA). A human study has shown positive effects of GFA in allergic disease prevention, particularly in the prevention of atopic dermatitis in infants (3). However, a recent publication indicated no preventive effect of atopic dermatitis when high risk children were supplied with an hydrolyzed formula supplemented with the oligosaccharides GFA (4). In a CMA mouse model, supplementation of GFA to the diet was found to decrease allergic symptoms (3, 5-7). Furthermore, GFA may suppress allergy development by promoting regulatory T cell (Treg) function (8).

Vitamin D (VitD) is obtained via the diet or synthesized in the skin upon UV light exposure. VitD is known for its importance in bone mineralisation. However, recently VitD has gained attention 6 in terms of its possible role in allergic diseases as well. VitD activating enzymes were found for example in, epithelial cells, T cells, dendritic cells (DC) and activated macrophages (9, 10). According to multiple studies a substantial portion of the world population has low serum VitD levels. Major risk factors for the low or too low VitD status are aging, female gender, living at higher latitudes, winter season, darker skin pigmentation, obesity and/ or low dietary intake (9, 11, 12). Due to low VitD levels found in people living further from the equator and the coinciding increase in food allergy prevalence, this suggests an implication of the influence of VitD on food allergy development (9, 11, 13).

Activation of VitD occurs in two steps, first in the liver it is modified to a prohormone followed by conversion to the biological active hormone in the kidneys (14). As mentioned above VitD activation also occurs in structural cells and cells of the immune system (9, 10). After VitD is enzymatically activated it will bind to the nuclear vitamin D receptor (VDR), the VDR will form a homodimer or a heterodimer with retinoic x receptor (RXR) bound to retinoic acid (RA) to activate specific genes in the nucleus. The VDR/RXR complex binds with high affinity to specific genes to up or downregulate transcription. For more details concerning receptor functions and binding sites the reader is referred to Christakos et al. (15).

109

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 109 CHAPTER 6

VitD signalling leads to tolerogenic immature interleukin (IL)10 producing DC which favours Treg upregulation and reduces T cell activation (10, 16, 17). Simultaneously DC interacting with VitD have a better migratory pattern towards lymph nodes and Peyer’s Patches (PP) (17). Not only is T cell activation reduced by VitD exposed DC, also VitD has a direct influence on T cells. VitD can directly inhibit IL2, interferon gamma (IFNγ) and IL17 production for T helper (Th)1 and Th17 cells, and upregulates IL10 (10, 18, 19). On the other hand, when VitD and transforming growth factor beta (TGFβ) were added to human CD4+ T cells in a cell proliferation assay, more CD25+Foxp3+ cell expansion and more IL2 production was measured (20). For the regulation of Th2 cytokines IL4 and IL13 by VitD, data remain inconclusive (9, 15, 18-21). VitD can also enhance antimicrobial mechanisms via increase of various antimicrobial peptides, strengthening beneficial microbiota, as well as improve barrier function (16, 21-24). In a colitis mouse model it was shown that providing VitD can still positively regulate beneficiary bacteria even when the vitamin D receptor is knocked out (25). Furthermore VitD has an important role in keeping the epithelial barrier intact in the gastro intestinal tract via maintaining tight junctions (16).

Due to differential design and the small number of studies with well-defined clinical parameters to determine VitD effects, the relationship between VitD and food allergy is unclear (19, 21, 26). There are studies suggesting a role for VitD not only in the DC - T cell interaction, but also a regulatory role in B cell - mast cell interaction (10, 27). VitD stabilizes mast cell activation and upon dietary VitD supplementation, immunoglobulin (Ig)E mediated mast cell activation of ovalbumin (OVA) treated mice is inhibited (28). In another study using OVA allergic mice an association between VitD deficiency and exacerbation of allergic symptoms for OVA specific– IgE and -IgG1, diarrhoea and Il4 mRNA levels in mesenteric lymph nodes (MLN) was shown (10, 13). In OVA allergic children who developed natural tolerance to egg, increased serum VitD levels were associated with changes in innate immune profiles (29). Low VitD status could possibly have a links to IgE mediated food allergy.

A recent study in a small cohort of egg allergic children suggested that VitD insufficiency did not correlate with food allergy (30). However, VitD insufficiency in Asian children was found to correlate with a risk for milk sensitization at the age of two (31). To further study the contribution of VitD depletion or supplementation in allergy development, the current study investigates the role of VitD in the prevention of CMA in the presence or absence of dietary GFA supplementation in mice.

110

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 110 VITD AND GFA IN CMA

MATERIALS AND METHODS

Diets The control diet, supplied by Research Diet Services, Wijk bij Duurstede, the Netherlands, is a semi-synthetic cow’s milk free AIN93G chow in which casein and whey are replaced by soy proteins (32). The diets contain no VitD, ‘standard’ 1000 IU vitamin D/kg while the VitD supplemented diet contains 5000 IU vitamin D/kg. For deprivation dose, it must be noted that this started upon arrival when mice were 4 weeks of age. Before that, mice received normal VitD amounts from their mothers and standard chow without casein and whey. For the supplementation dose we chose 5 times supplementation compared to control, because no direct adverse effects have been found in mice using this approach as shown by Agrawal et al. (33). The prebiotic diet is a control diet with standard or high VitD and additional isocaloric supplementation of 1% GFA in a 9:1:2 ratio; scGOS (Vivinal, GOS, Friesland Campina Domo, Zwolle, The Netherlands), lcFOS (Raftiline HP, Orafti, Wijchen, The Netherlands), pAOS (Südzucher, Mannheim, Germany).

Animals Upon arrival, 4 weeks old (>11 g bodyweight) specific pathogen-free female C3H/HeOuJ mice (Charles River, Sulzfeld, Germany) were fed either control diet, diet supplemented with GFA, diet supplemented with VitD or diet supplemented with VitD and GFA throughout the study (Figure 1). Food and tap water were available ad libitum. Mice (n=10 – 15 per group) were random allocated to a cage and group housed (5/cage), in Makrolon type III cages per 6 treatment, with 9kGy irradiated sawdust bedding (Lignocel 9s, J. Rettenmaier & Söhne GmbH, Germany), 2 red-transparent polycarbonate cages as environmental enrichment, and no nesting-material due to the potential effect on allergy-parameters (34) at the animal facility (Intravacc, Bilthoven, the Netherlands). Mice were on light/dark cycle of 12h/12h and 65-70 % relative humidity. Animal procedures, housing and care were performed in accordance with the EU-guidelines (2010/63/EU) and approved by a licensed Animal Ethics Committee (DEC- Consult, Soest, the Netherlands).

111

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 111 CHAPTER 6

Oral sensitization and challenge of mice After 2 weeks, mice were sensitized weekly for five times, with 10μg Cholera Toxin (CT), (List Biological Laboratories, USA) or 20 mg whey (sweet whey protein concentrate 60 (sWPC60), Milei, Germany) + 10μg CT per 500μl DPBS (Life Technologies, Inc., Invitrogen, USA) per mouse per oral gavage using a blunt needle. After the third sensitization, temperature transponders (IPTT-300, BMDS, USA) were injected subcutaneously into the mice under isoflurane. At day 35 following initial sensitization, whey protein was intradermally (i.d.) injected (10 μg whey product/20 μl PBS/ear) in the ear pinnae and the corresponding delta (Δ) ear swelling at 1 hour was measured as read-out for the local activation of mast cells. One day after the ear challenge, 18 hours before termination, the mice were challenged intragastrically (i.g.) with 50 mg whey product/500μl DPBS to determine the mast cell degranulation. Finally, at day 37, blood and MLN were isolated under terminal anaesthesia and stored for further analysis (Figure 1).

Figure 1. Design of experiment. GFA, short-chain Galacto-oligosaccharides (scGOS), long-chain Fructo- oligosaccharides (lcFOS), and pectin-derived acidic oligosaccharides (pAOS); CT, cholera toxin; i.d., intradermal.

Evaluation of the skin allergic response To measure the acute allergic skin response, mice were challenged intradermally in the ear pinnae with 10 μg whey protein per ear. Ear thickness was recorded before and 1 hour after the intradermal challenge using a digital micrometre (Mitutoyo, Veenendaal, the Netherlands) and the acute skin response was calculated as Δ = ear thickness at 1 hour − basal ear thickness and is expressed as delta micrometre. The body temperature and the anaphylactic shock symptoms were scored according to the decision table as shown (Table 1) which is adapted from the method previously described by Van Esch et al. (5).

Whey-specific Immunoglobulins and mouse mast cell protease-1 in serum Serum whey-specific immunoglobulins were quantified by means of an enzyme-linked immunosorbent assay (ELISA) as previously described with few modifications (35). Briefly, high-binding 96 well plates (Costar® Assay Plate, Corning, USA) were coated with whey (20 mg/L) overnight, washed and blocked with 5% bovine serum albumin (BSA, Sigma-Aldrich, St. Louis, USA). Three different dilutions of the samples and standards were added and incubated

112

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 112 VITD AND GFA IN CMA

Table 1. Decision tree on the monitoring of wellbeing during the development of anaphylactic shock following intradermal challenge adapted from Van Esch et al. (5). The severity of the anaphylactic shock as well as the body temperature as monitored at level of behaviour and mobility every 15 minutes following intradermal challenge. Anaphylactic shock score Body temperature > 32°C. Body temperature ≤32°C. 0 No action Heating pad no subjective symptoms mouse is likely to recover unlikely to occur within 60 min. mouse is likely to recover within 60 min. repositioning to home cage when body temperature > 35°C.

1 No action Heating pad scratching around nose and/or mouse is likely to recover unlikely to occur mouth within 60 min. mouse is likely to recover within 60 min. repositioning to home cage when body temperature > 35°C.

2 No action Heating pad swollen eyes and/or mouth, mouse is likely to recover mouse is likely to recover within 60 piloerection, reduced mobility, within 60 min. min. increased breath frequency repositioning to home cage when body temperature > 35°C.

3 No action Heating pad shortness of breath and/or unlikely to occur mouse is likely to recover within 60 increased breath frequency, mouse is likely to recover min. bluish color around mouth and within 60 min. repositioning to home cage when tail, further reduced/painful body temperature > 35°C. mobility 6

4 Humane Endpoint Humane Endpoint no mobility following stimulation, unlikely to occur mouse not likely to recover and is convulsions mouse not likely to recover actively euthanized and is actively euthanized

for two hours and washed. This was followed by application of biotin-labelled rat anti- mouse IgE, IgG1 or IgG2a (1 mg/L; Becton Dickinson, Belgium) and incubation for one hour. The plates were washed, incubated for one hour with streptavidin peroxide (SPO, Sanquin, The Netherlands) in the dark, washed, and incubated with 3, 3′, 5, 5′-tetramethylbenzidine 1-step substrate (TMB, Thermo Scientific) for up to 15 minutes. The reaction was stopped with 10% sulfuric acid (Sigma-Aldrich, Steinheim, Germany), followed by measurement at 450 nm with microplate reader (BioTek, Powerwave HT, Vermont, USA). The mouse mast cell 1 (mMCP1) concentrations were determined using a Mouse MCPT1 ELISA Ready-SET-Go!® kit (eBioscience, San Diego, USA), using a protocol obtained from the manufacturer.

113

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 113 CHAPTER 6

Isolation of lymphocytes To remove red blood cells, spleens were lysed according to Kostadinova et al. (36). Single cell suspensions of MLNs and spleens were made with a 70 μm cell strainers (Thermo Fisher Scientific, Amsterdam, the Netherlands) and then resuspended in RPMI 1640, 10% foetal bovine serum and penicillin (100 U/mL)/streptomycin (100 μg/mL). The cells were quantified using a Coulter Z1 particle counter (Beckmann Coulter, Brea, USA).

Flow cytometry of immune cells One million cells per well were added to polypropylene V-bottom 96 well plates (BD Biosciences, Germany) and blocked with rat anti-mouse CD16/32 (BD Biosciences, Germany) for 20 minutes at 4°C to block non-specific binding sites.

For surface staining the cells were incubated with the different antibodies of the staining panels for 30 minutes at 4°C. After the antibody incubation, the cells were fixed using BD cell fix (Becton Dickinson, Belgium). The Th1/Th2, Th17/Treg staining contained intracellular markers (Tbet, Gata3, Rorγt and Foxp3), therefore directly after the staining procedure for extracellular markers, these cells were permeabilized using a fixation/ permeabilization buffer (eBioscience, San Diego, USA) overnight. The next day, the cells were blocked again and incubated with the intracellular marker antibodies for 30 minutes at 4°C. Then the cells were fixed with the BD cell fix. Fluorescence Minus Ones (FMOs) and unstained cells were used as controls. In addition, compensation staining with UltraComp beads (eBioscience, San Diego, USA) was performed to correct for the spectral overlap, which occurs in multi- color staining. Antibodies used were: CD8a-APC-Cy7, CD11c-PerCP-Cy5.5 and CD25-Pe-Cy7 from BD biosciences (San Jose, CA, USA) and CD4-PerCP-Cy5.5, CD69-Pe-Cy7, CD11b-PE, CD103-APC, CX3CR1-FITC, B220-APC, IL10-PE, IFNγ-APC, IL4-Pe-Cy7, IL17α–FITC, Rorγ-PE, Gata3-PE, Tbet-APC, and Foxp3-APC from eBioscience (San Diego CA, USA).The analysis of the stained cells was performed using FACS Canto II cytometer (BD Biosciences, USA) and FACS Diva software (BD Biosciences, Germany).

Statistical Analysis The data are represented either as the mean, mean ± SEM or Tukey box-and-whiskers plots. Not normal distributed data was LOG transformed. Statistical analysis was performed using GraphPad Prism 7 (GraphPad Software, San Diego, CA, USA). One-way ANOVA and Bonferroni multiple comparisons post-test comparing selected groups were used to analyse the data. Selected groups are sham vs. allergic mice, allergic mice vs. allergic mice fed VitD+, allergic mice vs. allergic mice fed GFA, allergic mice fed VitD+ vs. allergic mice fed VitD+/GFA and allergic mice fed GFA vs. allergic mice fed VitD+/GFA. Kruskal-Wallis was used to analyse the anaphylactic shock score, followed by Dunn’s multiple comparison post-test, because the shock score data is not normally distributed. P-value < 0.05 was considered significant.

114

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 114 VITD AND GFA IN CMA

RESULTS

Allergic clinical symptoms To measure clinical symptoms the acute allergic skin response, anaphylaxis (Table 1), and body temperature were measured. The acute allergic skin response, measured as delta (Δ) ear swelling, and anaphylactic symptom score were significantly increased in whey sensitized mice (allergic mice) compared to sham sensitized mice fed control diet (sham) (p<0.01) (Figure 2A, 2B). VitD supplementation in whey sensitized mice fed the GFA diet (allergic mice fed VitD+/GFA) significantly decreased the allergic skin response compared to the VitD supplemented mice fed the control diet (allergic mice fed VitD+) (p<0.05). Nine mice, fed food deprived of VitD, reached the human endpoint during the 60 minutes establishment of the clinical allergic symptoms which was calculated as a loss of 38-40% in the VitD deprived group versus 22-30% in the CMA mice fed the standard VitD levels (Table 2). These VitD deprived mice were not included in the rest of the data.

The anaphylactic shock score, measured at 30 minutes, of the groups fed the GFA diet with or without VitD were not significantly different from sham (Figure 2B). Furthermore, the body temperature of the mice was monitored during the acute allergic skin response to determine systemic shock symptoms (Figure 2C). No significant differences were observed between groups with regard to body temperature. No significant differences were measured in bodyweight (data not shown).

Table 2. Total amount of mice per group suffering of anaphylactic shock symptoms during one hour after the intradermal ear challenge. Vitamin D Amount in chow; intake GFA % Anaphylactic % Humane Endpoint shock (mice/total) (mice/mice with shock) - 100 (15/15) 40 (6/15) Deprivation 0 IU/kg; 0 IU/day + 89 (8/9) 38 (3/8) - 60 (6/10) 30 (2/6) Normal 1000 IU/kg1; ≈ 5 IU/day + 90 (9/10) 22 (2/9) - 100 (10/10) 10 (1/10) Supplementation 5000 IU/kg2; ≈ 25 IU/day + 80 (8/10) 25 (2/8) 1. Reeves, P. G. & Suppl, M. Symposium : Animal Diets for Nutritional and Toxicological Research Components of the AIN-93 Diets as Improvements in the AIN-76A Diet 1 , 2. 838–841 (1997) 2. Agrawal, T., Gupta, G. K. & Agrawal, D. K. Vitamin D supplementation reduces airway hyperresponsiveness and allergic airway inflammation in a murine model. Clin. Exp. Allergy 43, 672–83 (2013)

115

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 115 CHAPTER 6

Figure 2. Clinical symptoms in sham and whey sensitized mice fed a control or GFA diet with standard (s) or extra (+) vitamin D. The acute allergic skin response (A) was measured one hour after intra-dermal (i.d.) injection. Values are means ± SEM, n=8-10, a one-way ANOVA and Bonferroni post hoc was used. Δ indicates ear thickness before i.d. injection deducted from thickness after i.d. injection. The anaphylactic symptom scores (B) of sham-sensitized mice fed control diet and whey-sensitized mice fed control- or GFA diet with standard (s) or extra (+) vitamin D, n = 8-10. Data is calculated with the Kruskal-Wallis and Dunn’s post hoc test with pre-selected pairs. The body temperature (C) was measured 30 minutes after i.d. injection via transponder read out of sham-sensitized mice fed control diet and whey-sensitized mice fed

116

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 116 VITD AND GFA IN CMA

control- or GFA diet with standard (s) or extra (+) vitamin D. Values are means ± SEM, n=8-10, a one-way ANOVA and Bonferroni post hoc was used to calculate the dataset. Asterix indicates a significant difference compared to sham sensitized mice, **** p<0.0001, ** p<0.01. Hash tag indicates significance compared to whey sensitized mice fed control diet, # p<0.05. GFA, short-chain Galacto-oligosaccharides/ long-chain Fructo-oligosaccharides/ pectin-derived Acidic-oligosaccharides; s, standard 1000 IU/kg vitamin D; +, 5000 IU/kg vitamin D.

Mouse mast cell protease and whey specific immunoglobulins To investigate whether the Th2 and Th1 type humoral responses are related to the differences in acute allergic skin responses of the different groups, whey-specific IgE, IgG1 and IgG2a were measured. The whey-specific IgE and IgG1 levels were significantly increased in the allergic mice compared to sham (p<0.01) (Figure 3A, B). However, the whey-specific-IgE and -IgG1 levels were not affected by the GFA diet with or without VitD supplementation. The mMCP1 concentration, which is a mediator released by mucosal mast cells upon degranulation, and whey-specific IgG2a levels were not affected by treatment or diet (Supplementary Figure 1).

T cells and dendritic cells The effect of the diets on Th1 (CD4+Tbet+), Th2 (CD4+Gata3+) Treg (CD4+CD25+Foxp3+), CD11c+CD103+ DC and plasmacytoid dendritic cells (pDC) (CD11clow B220+) was also determined in spleen and MLN. No significant differences were observed in the percentages of Th1 and Th2 in the spleen (Figure 4A) and MLN (Figure 4B) and no significant differences were shown in pDC and Treg cells for both the spleen (Figure 5A) and MLN (Figure 5B). There was a significantly 6 higher percentage of CD11c+CD103+ DC in the allergic mice fed VitD+/GFA group compared to the allergic mice fed VitD+ (p<0.05) (Figure 5B).

117

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 117 CHAPTER 6

Figure 3. Whey-specific immunoglobulin E and immunoglobulin G1 measured in serum of mice. Whey- specific-IgE (AU) (A) and Whey-specific-IgG1 (AU) (B) were measured in serum from sham-sensitized mice fed control diet and whey-sensitized mice fed control- or GFA diet with standard (s) or extra (+) vitamin D. Values are means ± SEM, n= 8-10. Data was calculated with a one-way ANOVA and Bonferroni post hoc test, if required, LOG transformation was used to normalize data distribution. Asterix indicates a significant difference compared to sham sensitized mice, **** p<0.0001, *** p<0.001. GFA, short-chain Galacto-oligosaccharides/ long-chain Fructo-oligosaccharides/ pectin-derived Acidic-oligosaccharides; IgE, immunoglobulin E; IgG1, immunoglobulin G1; AU, arbitrary units; s, standard1000 IU/kg vitamin D; +, 5000 IU/kg vitamin D.

118

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 118 VITD AND GFA IN CMA

Figure 4. T helper cell population, Th1 and Th2, measured in the spleen (A) and MLN (B). Spleen (A) were removed after sacrifice of mice and stained for CD4-PERCP-Cy5.5+ CD8α-APC-Cy7- then two subpopulations were gated, one for Th1 the intracellular Tbet-APC+ and the other for Th2 intracellular + + Gata3-PE . MLN (B) were removed after sacrifice of mice and stained for CD4-PERCP-Cy5.5 CD8α-APC- 6 Cy7- then for CD69-Pe-Cy7+ and subsequently for two subpopulations, one for Th1 the intracellular Tbet- APC+ and the other for Th2 intracellular Gata3-PE+. Representative flow cytometric plots are shown in Supplemental Figures 2 (spleen) and 3 (MLN). Results are shown in percentage increase or decrease compared to the mean of sham sensitized mice fed control diet. Values are means ± SEM, n= 8-10 and were calculated with a one-way ANOVA and Bonferroni post hoc test. CD, cluster of differentiation; GFA, short-chain Galacto-oligosaccharides/ long-chain Fructo-oligosaccharides/ pectin-derived Acidic- oligosaccharides; s, standard 1000 IU/kg vitamin D; +, 5000 IU/kg vitamin D.

119

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 119 CHAPTER 6

Figure 5. Plasmacytoid and CD103+ dendritic cell population and Foxp3+CD25+ regulatory T cell population measured in the spleen (A) and MLN (B). Spleen (A) was stained for plasmacytoid DC (pDC); B220-APC- Cy7+, CD11c-PERCP-Cy5.5+ CD11b-PE-, and for Treg cells CD4-PERCP-Cy5.5+ CD8a-APC-Cy7-, Foxp3- APC+ CD25-PE-Cy7+. MLN (B) were removed after sacrifice of mice and stained for pDC; B220-APC-Cy7+, CD11c-PERCP-Cy5.5+ CD11b-PE-, for migratory DC; CD11c-PERCP-Cy5.5+, CD103-APC+ , CX3CR1-FITC- and CD11b-PE+, CD8α-APC-Cy7- and for Treg cells CD4-PERCP-Cy5.5+ CD8a-APC-Cy7-, Foxp3-APC+ CD25- PE-Cy7+. Values are means ± SEM, n= 3-10 and were calculated with a one-way ANOVA and Bonferroni post hoc test. For the MLN for the CD103 as well as the Treg population only data of one cohort were available. Representative flow cytometric plots are shown in Supplemental Figure 4. Asterix indicates a significant difference compared to sham sensitized mice, # p<0.05. CD, cluster of differentiation; GFA, short-chain Galacto-oligosaccharides/ long-chain Fructo-oligosaccharides/ pectin-derived Acidic-oligosaccharides; s, standard 1000 IU/kg vitamin D; +, 5000 IU/kg vitamin D.

120

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 120 VITD AND GFA IN CMA

DISCUSSION

VitD is best known for its role in bone metabolism but it also modulates the function of the innate and adaptive immune system. In the current study it was investigated whether VitD deprivation or supplementation could affect CMA development in presence or absence of dietary NDO mixture GFA.

Normal VitD levels are for 90% acquired via synthesis in the skin via conversion of cholesterol to the active form of VitD and 10% is acquired via dietary intake in humans. The advised human VitD levels are optimized for bone health, it is unknown if these levels are also adequate for optimizing immune properties. Furthermore, VitD has a “U” shaped effectiveness curve, too little or too much is not desirable (9). In mice not much is known regarding their acquisition and processing of VitD. VitD dosages applied in standard mouse chow are extrapolated from studies in rats (37). The dosage used in this study, five times higher compared to control, was shown not to cause adverse effects in mice (33). For mice, just as for humans, VitD is an essential vitamin for many body functions. All mice in our study were weened by their mothers with normal VitD status so in their first four weeks of life they had access to VitD supplies via their mother. So only after weening, during the study protocol the mice were fed VitD deprived or supplemented food.

Previously an allergy preventive effect of GFA was shown, however in this study GFA alone had no significant suppressive effect on allergic symptoms in cow’s milk allergic mice (5, 8, 38, 39). 6 VitD supplementation did not have a significant effect on the clinical symptoms such as the acute allergic skin response and anaphylactic shock. However, the percentage of mice that were actively euthanized due to severe shock upon intradermal whey injection was highest in the cow’s milk allergic mice deprived from VitD fed with or without GFA. This indicates that VitD levels below the standard levels in chow may negatively affect the immune status of the mice. Indeed, an adequate VitD status is required otherwise allergic symptoms can be more severe. This is in line with studies of Matsui et al. where VitD deprived mice showed worse OVA allergy via increased OVA specific immunoglobulin levels (13). In a C57BL/6 VitD deficient mouse model it was shown that eosinophilic cells activate spontaneously and subsequently release inflammatory mediators which also cause intestinal epithelial barrier dysfunction (40). Furthermore Liu et al. showed positive inhibitory effects of VitD supplementation on allergic symptoms in an OVA BALB/c mouse model, however in these studies histamine and tumor necrosis factor-α were measured in serum as a reflection of mast cell activation (28). In the current study no significant difference in whey specific IgE was observed.

For VitD to become effective in the nucleus it first needs to bind to the VDR, and consecutively it will form a dimer with the activated vitamin A (VitA) receptor, RXR (16). Future studies may aim

121

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 121 CHAPTER 6

to combine VitD and VitA supplementation which may further improve the protective effect of GFA in CMA. Because VitA also has a beneficiary effect on the immune system and it uses the same receptors for binding to DNA as VitD. This additionally, may imply that a combination of these two vitamins may affect the allergic outcome in the CMA model. This is also suggested by Ricco and Rosanno who call VitD and VitA dietary anti-inflammatory molecules (41). However, Scheider et al. show that the interplay between VitD and VitA in the immune response is likely to be context and cell specific, because RXR can also form a dimer with other vitamin A receptors, therefore both vitamins could also compete for the same activating receptors (17). This was found for example in vitro for intestinal T cells were RA mediated α4β7/ CCR9 homing was reversed by VitD (42).

VitD supplementation in presence or absence of GFA has no significant effect on activated Th2 or Th1 cells in this model. No difference in T helper cells between the sham and CMA mice was observed. In a previous study, it was shown that CMA mice fed GFA had a higher increase of Th1 Tbet mRNA compared to Th2 Gata3 mRNA (38). The diets did not affect the percentage of CD25+Foxp3+ Treg cells and pDC in this study. However, the percentage of CD11c+CD11b+CD103+ DC was significantly increased in the allergic VitD+/GFA group when compared to the allergic mice fed VitD+. This subtype of DC may have tolerogenic properties and support the generation of functional Treg, which may be capable of suppressing mast cell degranulation and subsequently decrease the acute allergic skin response (43-45). VitD indeed has been shown to regulate different dermal and epidermal DC subsets and even to induce TGFβ-dependent Foxp3+ Treg or IL10-dependent IL10+ Treg through the modulation of DC function (20, 46). However, in the current study intracellular IL10 levels in DC in the MLN or spleen remained below detection (data not shown). Although we did not observe any significant differences in pDC, another study suggests that VitD can regulate pDC function (47). In the latter study bone marrow derived DC were used and a tumor mouse model was used to multiply pDC in lymphoid organs, which is quite different compared to our food allergy model (47). Migration of DC to the MLN occurs between one and 72 hours after allergen ingestion, while we examined the MLN eighteen hours after oral challenge. This time point may be more appropriate for analysis of T cells in the lamina propria that have migrated from the MLN. In addition, under influence of VitD, DC may be triggered not only to migrate towards the lymph nodes, but also to non-draining lymph nodes and PP (17) which may also be of interest for further studies.

Still very little is known about the effect of VitD status on food allergy development. 38-40% of cow’s milk allergic mice fed the VitD deprived diet had very severe anaphylactic shock manifestations while this was 22-30% in cow’s milk allergic mice fed standard VitD levels, indicating that VitD deprivation may enhance allergic symptom severity. The data presented in this study showed that extra supplementation of VitD did not protect against CMA development in mice and was ineffective in supporting the allergy protective effect of GFA.

122

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 122 VITD AND GFA IN CMA

REFERENCES

1. Longo G, Berti I, Burks AW, Krauss B, Barbi 12. Hoyos-Bachiloglu R, Morales PS, Cerda E. IgE-mediated food allergy in children. J, Talesnik E, Gonzalez G, Camargo CA, Lancet. 2013;382(9905):1656-64. Jr., et al. Higher latitude and lower solar 2. Huang F, Kim JS. IgE-mediated cow’s milk radiation influence on anaphylaxis in Chilean allergy in children. Current allergy and children. Pediatric allergy and immunology : asthma reports. 2012;12(6):630-40. official publication of the European Society of Pediatric Allergy and Immunology. 3. Jeurink PV, van Esch BC, Rijnierse A, Garssen 2014;25(4):338-43. J, Knippels LM. Mechanisms underlying immune effects of dietary oligosaccharides. 13. Matsui T, Yamashita H, Saneyasu KI, Tanaka Am J Clin Nutr. 2013;98(2):572S-7S. H, Ito K, Inagaki N. Vitamin D deficiency exacerbates sensitization and allergic 4. Boyle RJ, Tang ML, Chiang WC, Chua diarrhea in a murine food allergy model. MC, Ismail I, Nauta A, et al. Prebiotic- Allergol Int. 2017. supplemented partially hydrolysed cow’s milk formula for the prevention of eczema 14. Adorini L, Penna G. Induction of tolerogenic in high-risk infants: a randomized controlled dendritic cells by vitamin D receptor agonists. trial. Allergy. 2016;71(5):701-10. Handb Exp Pharmacol. 2009(188):251-73. 5. Schouten B, van Esch BC, Hofman GA, van 15. Christakos S, Dhawan P, Verstuyf A, Doorn SA, Knol J, Nauta AJ, et al. Cow milk Verlinden L, Carmeliet G. Vitamin D: allergy symptoms are reduced in mice fed Metabolism, Molecular Mechanism of dietary synbiotics during oral sensitization Action, and Pleiotropic Effects. Physiol Rev. with whey. J Nutr. 2009;139(7):1398-403. 2016;96(1):365-408. 6. Nauta AJ, Garssen J. Evidence-based 16. Barbachano A, Fernandez-Barral A, Ferrer- benefits of specific mixtures of non- Mayorga G, Costales-Carrera A, Larriba MJ, digestible oligosaccharides on the immune Munoz A. The endocrine vitamin D system in system. Carbohydr Polym. 2013;93(1):263-5. the gut. Mol Cell Endocrinol. 2017;453:79- 87. 7. Roberfroid M, Gibson GR, Hoyles L, McCartney AL, Rastall R, Rowland I, et al. 17. Bscheider M, Butcher EC. Vitamin D 6 Prebiotic effects: metabolic and health immunoregulation through dendritic cells. benefits. Br J Nutr. 2010;104 Suppl 2:S1-63. Immunology. 2016;148(3):227-36. 8. Schouten B, van Esch BC, Hofman GA, 18. Cantorna MT, Snyder L, Lin YD, Yang L. Boon L, Knippels LM, Willemsen LE, et al. Vitamin D and 1,25(OH)2D regulation of T Oligosaccharide-induced whey-specific cells. Nutrients. 2015;7(4):3011-21. CD25(+) regulatory T-cells are involved in 19. Peters RL, Neeland MR, Allen KJ. Primary the suppression of cow milk allergy in mice. Prevention of Food Allergy. Current allergy J Nutr. 2010;140(4):835-41. and asthma reports. 2017;17(8):52. 9. Reinholz M, Ruzicka T, Schauber J. Vitamin 20. Chambers ES, Suwannasaen D, Mann EH, D and its role in allergic disease. Clinical and Urry Z, Richards DF, Lertmemongkolchai experimental allergy : journal of the British G, et al. 1alpha,25-dihydroxyvitamin D3 Society for Allergy and Clinical Immunology. in combination with transforming growth 2012;42(6):817-26. factor-beta increases the frequency 10. Hufnagl K, Jensen-Jarolim E. Vitamin A of Foxp3(+) regulatory T cells through and D in allergy: from experimental animal preferential expansion and usage of models and cellular studies to human interleukin-2. Immunology. 2014;143(1):52- disease. Allergo J Int. 2018;27(3):72-8. 60. 11. Suaini NH, Zhang Y, Vuillermin PJ, Allen 21. Muehleisen B, Gallo RL. Vitamin D in allergic KJ, Harrison LC. Immune Modulation by disease: shedding light on a complex Vitamin D and Its Relevance to Food Allergy. problem. The Journal of allergy and clinical Nutrients. 2015;7(8):6088-108. immunology. 2013;131(2):324-9.

123

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 123 CHAPTER 6

22. Dimitrov V, White JH. Vitamin D signaling in 32. Reeves PG. Components of the AIN-93 diets intestinal innate immunity and homeostasis. as improvements in the AIN-76A diet. The Mol Cell Endocrinol. 2017;453:68-78. Journal of nutrition. 1997;127(5 Suppl):838S- 23. Wang J, Thingholm LB, Skieceviciene 41S. J, Rausch P, Kummen M, Hov JR, et al. 33. Agrawal T, Gupta GK, Agrawal DK. Vitamin Genome-wide association analysis identifies D supplementation reduces airway variation in vitamin D receptor and other hyperresponsiveness and allergic airway host factors influencing the gut microbiota. inflammation in a murine model. Clinical Nature genetics. 2016;48(11):1396-406. and experimental allergy : journal of the 24. Bashir M, Prietl B, Tauschmann M, Mautner British Society for Allergy and Clinical SI, Kump PK, Treiber G, et al. Effects of high Immunology. 2013;43(6):672-83. doses of vitamin D3 on mucosa-associated 34. Pasalic I, Bosnjak B, Tkalcevic VI, Jaran gut microbiome vary between regions of the DS, Javorscak Z, Markovic D, et al. Cage human gastrointestinal tract. Eur J Nutr. enrichment with paper tissue, but not plastic 2016;55(4):1479-89. tunnels, increases variability in mouse 25. Ooi JH, Li Y, Rogers CJ, Cantorna MT. model of asthma. Lab Anim. 2011;45(2):121- Vitamin D regulates the gut microbiome and 3. protects mice from dextran sodium sulfate- 35. van Esch BC, Schouten B, de Kivit S, Hofman induced colitis. The Journal of nutrition. GA, Knippels LM, Willemsen LE, et al. Oral 2013;143(10):1679-86. tolerance induction by partially hydrolyzed 26. Zakariaeeabkoo R, Allen KJ, Koplin JJ, whey protein in mice is associated with Vuillermin P, Greaves RF. Are vitamins A enhanced numbers of Foxp3+ regulatory and D important in the development of food T-cells in the mesenteric lymph nodes. allergy and how are they best measured? Pediatric allergy and immunology : official Clin Biochem. 2014;47(9):804-11. publication of the European Society of Pediatric Allergy and Immunology. 27. James J, Weaver V, Cantorna MT. Control of 2011;22(8):820-6. Circulating IgE by the Vitamin D Receptor In Vivo Involves B Cell Intrinsic and Extrinsic 36. Kostadinova AI, Meulenbroek LA, van Esch Mechanisms. Journal of immunology. BC, Hofman GA, Garssen J, Willemsen 2017;198(3):1164-71. LE, et al. A Specific Mixture of Fructo- Oligosaccharides and Bifidobacterium 28. Liu ZQ, Li XX, Qiu SQ, Yu Y, Li MG, Yang LT, breve M-16V Facilitates Partial Non- et al. Vitamin D contributes to mast cell Responsiveness to Whey Protein in Mice stabilization. Allergy. 2017;72(8):1184-92. Orally Exposed to beta-Lactoglobulin- 29. Neeland MR, Koplin JJ, Dang TD, Dharmage Derived Peptides. Front Immunol. SC, Tang ML, Prescott SL, et al. Early life 2016;7:673. innate immune signatures of persistent food 37. Benevenga N.J. CC, Eckhert C.D., Fahey allergy. The Journal of allergy and clinical G.C., Greger J.L., Keen C.L., Knapka J.J., immunology. 2017. Magalhaes H., Otfedal O.T., Reeves P.G., 30. Molloy J, Koplin JJ, Allen KJ, Tang MLK, Anderson Shaw H., Smith J.E. Steele R.D. Collier F, Carlin JB, et al. Vitamin D . In: Nutrition NRCUSoLA, editor. Nutrient insufficiency in the first 6 months of infancy Requirements of Laboratory Animals: and challenge-proven IgE-mediated food Fourth Revised Edition, 1995. Washington allergy at 1 year of age: a case-cohort study. (DC): the National Academy of Sciences; Allergy. 2017;72(8):1222-31. 1995. 31. Chiu CY, Yao TC, Chen SH, Tsai MH, Tu YL, 38. Kerperien J, Jeurink PV, Wehkamp T, van Hua MC, et al. Low cord blood vitamin D der Veer A, van de Kant HJ, Hofman GA, et al. levels are associated with increased milk Non-digestible oligosaccharides modulate sensitization in early childhood. Pediatric intestinal immune activation and suppress allergy and immunology : official publication cow’s milk allergic symptoms. Pediatric of the European Society of Pediatric Allergy allergy and immunology : official publication and Immunology. 2014;25(8):767-72. of the European Society of Pediatric Allergy and Immunology. 2014;25(8):747-54.

124

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 124 VITD AND GFA IN CMA

39. Eiwegger T, Stahl B, Haidl P, Schmitt J, Boehm G, Dehlink E, et al. Prebiotic oligosaccharides: in vitro evidence for gastrointestinal epithelial transfer and immunomodulatory properties. Pediatr Allergy Immunol. 2010;21(8):1179-88. 40. Lu H, Xie RD, Lin R, Zhang C, Xiao XJ, Li LJ, et al. Vitamin D-deficiency induces eosinophil spontaneous activation. Cell Immunol. 2017;322:56-63. 41. Riccio P, Rossano R. Diet, Gut Microbiota, and Vitamins D + A in Multiple Sclerosis. Neurotherapeutics. 2017. 42. Sigmundsdottir H, Pan J, Debes GF, Alt C, Habtezion A, Soler D, et al. DCs metabolize sunlight-induced vitamin D3 to ‘program’ T cell attraction to the epidermal chemokine CCL27. Nature immunology. 2007;8(3):285- 93. 43. Fujimoto K, Karuppuchamy T, Takemura N, Shimohigoshi M, Machida T, Haseda Y, et al. A new subset of CD103+CD8alpha+ dendritic cells in the small intestine expresses TLR3, TLR7, and TLR9 and induces Th1 response and CTL activity. Journal of immunology. 2011;186(11):6287-95. 44. Matteoli G, Mazzini E, Iliev ID, Mileti E, Fallarino F, Puccetti P, et al. Gut CD103+ dendritic cells express indoleamine 2,3-dioxygenase which influences T regulatory/T effector cell balance and oral 6 tolerance induction. Gut. 2010;59(5):595- 604. 45. Smit JJ, Bol-Schoenmakers M, Hassing I, Fiechter D, Boon L, Bleumink R, et al. The role of intestinal dendritic cells subsets in the establishment of food allergy. Clinical and experimental allergy : journal of the British Society for Allergy and Clinical Immunology. 2011;41(6):890-8. 46. van der Aar AM, Sibiryak DS, Bakdash G, van Capel TM, van der Kleij HP, Opstelten DJ, et al. Vitamin D3 targets epidermal and dermal dendritic cells for induction of distinct regulatory T cells. J Allergy Clin Immunol. 2011;127(6):1532-40 e7. 47. Karthaus N, van Spriel AB, Looman MWG, Chen S, Spilgies LM, Lieben L, et al. Vitamin D controls murine and human plasmacytoid dendritic cell function. The Journal of investigative dermatology. 2014;134(5):1255- 64.

125

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 125 CHAPTER 6

SUPPLEMENTAL FIGURES

Supplemental Figure 1. Mouse mast cell protease 1 (mMCP1) and immunoglobulin G2 measured in serum of mice. Mouse mast cell protease 1 (pg/ml) (A) and whey-specific-IgG2 (AU) (B) were measured in serum from sacrificed sham-sensitized mice fed control diet and whey-sensitized mice fed control- or GFA diet with standard (s) or extra (+) vitamin D. Values are means ± SEM, n= 8-10. Data was calculated with a one-way ANOVA and Bonferroni post hoc test, if required, LOG transformation was used to normalize data distribution. GFA, short-chain Galacto-oligosaccharides/ long-chain Fructo-oligosaccharides/ pectin- derived Acidic-oligosaccharides; mMCP-1, mouse mast cell protease 1; IgG, immunoglobulin G2; AU, arbitrary units; s, standard 1000 IU/kg vitamin D; +, 5000 IU/kg vitamin D.

126

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 126 VITD AND GFA IN CMA

Chapter 6 VitD and GFA in CMA

single stained single stained CD4-PerCP-Cy5.5 CD8α-APC-Cy7

CD8α+/CD4++ isotypes CD8α+ + isotypes

single stained Tbet-APC cells in single stained Gata3-PE cells in spleen cell population spleen cell population

6 CD4-PerC-Cy5.5 stained cells + CD4-PerC-Cy5.5 stained cells + isotypes in spleen cell population isotypes in CD69+ spleen cell population

585 Supplemental Figure 2. Gating strategy for Figure 4A. Staining was used for CD4-PERCP-Cy5.5+, CD8α- APC-Cy7-, Gata3-PE+ and Tbet-APC+ CD4+ cell population were gated using a plot separating CD4+ from CD8α+ cells. For the Th1 and Th2 cells analysis was set up the same for spleen and MLN, for the spleen the CD69 staining gave no positive signal within the sample population, thus CD4+ Tbet/Gata3+ staining was used. First single stained dot plots are shown for CD4-PERCP-Cy5.5+ and CD8α-APC-Cy7+, then an isotype staining with CD4-PERCP-Cy5.5+ /CD8 -APC-Cy7+ and an isotype staining with only CD8 -APC-Cy7+. The α 155 α specifi c Tbet-APC+ and Gata3-PE+ populations are shown within CD4-PERCP-Cy5.5 gate. Due to diffi cult separation of Tbet-APC+ and Gata3-PE+ populations in a dot plot, histograms were used. Next histograms show specifi c staining and gate placement for Tbet-APC+ and Gata3-PE+ within total spleen population as well as CD4-PERCP-Cy5.5+.

127

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 127 CHAPTER 6 Chapter 6 VitD and GFA in CMA

596

597 single stained single stained CD4-PerCP-Cy5.5 CD8α-APC-Cy7 598

599

600

601 + + + CD8α /CD4 + isotypes CD8α + isotypes

602

603

604

single stained Tbet-APC cells 605 single stained Gata3-PE cells in MLN cell population in MLN cell population

606

607

608

609 CD4-PerC-Cy5.5 stained cells + isotypes CD4-PerC-Cy5.5 stained cells + in CD69+in MLN cellpopulation 610 isotypes in MLN cell population

611

612

613

614 + + + + CD69+ isotype vs single stained CD69 Gata3 CD69 Tbet ( 400 counts) + in MLN cell population. 615 CD69 in MLN cell population

616

617

618

157

128

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 128 VITD AND GFA IN CMA

Supplemental Figure 3. Staining was used for CD4-PERCP-Cy5.5+, CD8α-APC-Cy7-, CD69-Pe-Cy7+, Gata3-PE+ and Tbet-APC+. CD4+ cell population were gated using a plot separating CD4+ from CD8α+ cells. For the Th1 and Th2 cells analysis was set up the same for spleen and MLN, for the MLN CD4+/CD8α-, CD69+ Tbet/Gata3+ staining was used. First single stained dot plots are shown for CD4-PERCP-Cy5.5+ and CD8α-APC-Cy7+, then an isotype staining with CD4-PERCP-Cy5.5+ /CD8α-APC-Cy7+ and an isotype staining with only CD8α-APC-Cy7+. Due to diffi cult separation of Tbet-APC+ and Gata3-PE+ populations in a dot plot, histograms were used. Next histograms show specifi c staining and gate placement for Tbet-APC+ and Gata3-PE+in the CD4-PERCP-Cy5.5+ population and the CD69-Pe-Cy7+ population. Then specifi c staining is shown for CD69-Pe-Cy7+ in the total MLN population in which the Tbet/ Gata3 staining can be detected.

129

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 129 CHAPTER 6

Chapter 6 VitD and GFA in CMA

low CD11c

B220 iso CD11c

iso Foxp3

CD4 CD25

low CD11c

B220 iso CD11c CD11b CD103 iso iso

CD11c CX3CR1 CD8α

iso Foxp3

CD4 CD25 579

154

130

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 130 VITD AND GFA IN CMA

Supplemental Figure 4. Gating strategy for Figure 5 is shown in representative fl ow cytometric plots, only positive staining is shown in representative plots using arrows. Staining for plasmacytoid DC (pDC); B220-APC-Cy7+, CD11c-PERCP-Cy5.5+ CD11b-PE-, and for Treg cells CD4-PERCP-Cy5.5+ CD8a-APC-Cy7-, Foxp3-APC+ CD25-PE-Cy7+. For pDC population gating strategy was B220+ then a subpopulation of CD11c low within the CD11c+ /CD11b- cell population. The isotype control is shown for B220+/CD11c- (iso). For Treg fi rst the CD4+CD8α- cell population was selected, then this subpopulation was selected for Foxp3+CD25+. The isotype control is shown for CD4+/Foxp3+/CD25- (iso). In the MLN the migratory DC is strategy contains two subpopulations, fi rst CD11c+, then CD103+/CX3CR1-, subsequently CD11b+/ CD8α-. Isotypes are shown for CD11c+/CD103- and CD11c+/CD103+/ CD11b-.

131

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 131 7 530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 132 Dietary vitamin A supports a mixture of speciﬁ c non-digestible oligosaccharides in the prevention of cow’s milk allergic symptoms in mice

JoAnn Kerperien1*, Anna Oja1, Désirée Veening-Griffi oen1,2*, Tjalling Wehkamp2, Prescilla V. Jeurink1,2, Johan Garssen1,2, Léon M.J. Knippels1,2, Linette E.M. Willemsen1

ABSTRACT

Background: Cow’s milk allergy (CMA) affects three percent of children worldwide. A mixture of non-digestible short-chain Galacto-oligosaccharides (scGOS), long-chain Fructo- oligosaccharides (lcFOS), and pectin-derived Acidic-oligosaccharides (pAOS) (GFA) can have a beneficiary effect on allergy development and a preventive effect on allergy symptoms in a CMA mouse model. Vitamin A (VitA) is known to be an important immune regulatory component and may contribute to allergy prevention. In this study the VitA effect with or without GFA on CMA development was investigated in mice.

Methods: Female four weeks old C3H/HeOuJ mice (n=10-15/ group) fed a control or GFA diet with depleted, standard (4000 IU/kg) or supplemented (8000 IU/kg) VitA (VitA+) were weekly sensitized intragastrical with PBS or whey with cholera toxin for 5 weeks. One week after the last sensitization mice were intradermally challenged in both ear pinnae and orally challenged with whey. Terminal blood samples, spleens and mesenteric lymph nodes (MLN) were collected 18 hours after oral challenge. Whey-specific IgE, IgG2a and mucosal mast cell protease 1 (mMCP1) concentrations were measured in serum using ELISA. Intestinal Il4, Il5, Il10, Ifnγ, Foxp3 and Aldh1a2 mRNA was evaluated with qPCR. T cell subsets and dendritic cells from spleen and MLN were studied using flow cytometry.

Results: Dietary VitA+ nor GFA supplementation alone were able to reduce the allergic skin response in the current study, while combined VitA+ supplementation with GFA (VitA+/GFA) did when compared to allergic mice fed VitA+ alone. The VitA depleted mice showed the same allergic patterns as the allergic mice fed control diet, so this group was not further analyzed. The intestinal samples from allergic mice fed VitA+/GFA showed modulation of Il4, Il5, Ifnγ and Foxp3 mRNA expression (P < 0.05). Cytokine expression of IL4 from T cells and IFNγ from splenic phagocytes was affected (P < 0.05).

Conclusions: Dietary supplementation with VitA supports GFA to reduce CMA symptom development in a mouse model for allergic sensitization.

Abbreviation list sham sham sensitized mice fed control diet allergic mice whey sensitized mice fed control diet allergic mice fed VitA+ whey sensitized mice fed control diet with vitamin A supplementation allergic mice fed GFA whey sensitized mice fed scGOS/lcFOS/pAOS diet allergic mice fed VitA+/GFA whey sensitized mice fed scGOS/lcFOS/pAOS diet with vitamin A supplementation

134

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 134 VITA AND GFA IN CMA

INTRODUCTION

One of the most common food allergies is cow’s milk allergy (CMA) (1-3). CMA affects 2-3% of infants worldwide and typically 90% of these infants outgrow their allergy by developing oral tolerance (1-3). Upon allergy persistence the only form of treatment is avoiding the causative allergen completely, although allergen specific immunotherapies for food allergy are under development (1, 3).

An healthy immune balance, a tightly regulated intestinal epithelial barrier function, and a well- balanced intestinal microbiome are required to acquire oral tolerance and avoid an allergic reaction towards food proteins (1). Due to the positive effect on microbiome development of oligosaccharides in human milk, a mixture of lactose derived non-digestible short-chain Galacto-oligosaccharides (scGOS), inulin derived long-chain Fructo-oligosaccharides (lcFOS), and pectin-derived Acidic-oligosaccharides (pAOS) (GFA) was composed to be added to formula milk (4, 5). scGOS and lcFOS are neutral oligosaccharides and pAOS is an acidic oligosaccharide. Combining these three oligosaccharides in a 9:1:2 ratio mimics the natural balance of oligosaccharide structures in human milk and have a positive effect on bifidobacteria and lactobacilli development and in atopic dermatitis prevention in children in the general population (5-8). Although in a recent study no effects were observed for the prevention of eczema using GFA in the first year of life in high risk infants (13). Addition of GFA to the diet did decrease allergic symptoms in a murine model for CMA (9).

Vitamin A (VitA) is essential for growth and development, preservation of vision and for mucosal immunity. Both VitA deficiency as well as excess of VitA can lead to severe problems, like blindness or toxicity (10-12). VitA is fat soluble and accumulates in fat tissue and the liver, therefore an optimized dose of this vitamin is necessary to avoid toxicity (10-12). Via several steps, VitA is metabolized to its active form retinoic acid (RA) (12-14). During the last step 7 the enzyme retinal aldehyde dehydrogenase (ALDH) forms RA, which amongst others can be detected in epithelial, stromal or dendritic cells (DC) in the gut (15). Upon activation RA will increase ALDH activity creating a positive feedback loop and excessive RA will eventually end this positive feedback (13, 14, 16). The ALDH activity in intestinal DC is restricted to CD103+ DC only (13, 16, 17).

After migration of these RA+CD103+ DC to the mesenteric lymph nodes (MLN), these cells instruct naïve T cells. Next production of RA by MLN stromal cells is also necessary to induce gut homing T cells (11, 13). When RA is produced in the presence of interleukin (IL)10 and/ or transforming growth factor (TGFβ) in the MLN, this will lead to a better T cell migration via upregulation of integrin α4β7 and gut homing chemokine receptor c-c motif chemokine receptor (CCR)9 on the T-cells and also to a diverse subset of regulatory T cells (Treg) (11, 13, 15, 16, 18, 19).

135

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 135 CHAPTER 7

Treg induction and homing to the intestinal wall is crucial for the development of oral tolerance to food proteins. This tolerance induction is not solely dependent on RA, also TGFβ or IL10 are required to induce Foxp3+ Treg or other types of Treg (15). Allergic symptoms were suppressed for peanut or ovalbumin (OVA) allergens when allergic BALB/c or C57/BL6 mice were injected with allergen pulsed autologous RA-DC (20). Using these RA-DC in vitro in a co-culture with splenocytes from OVA sensitized mice, the RA-DC induced a novel Foxp3- CD25+ LAG3+ CD49b- regulatory cell type via IL27, and these novel Treg suppressed the Th2 cell type population (20).

RA does not only influence Treg, it also influences other T cell subsets. Studies investigating the role of VitA in infectious disease models have shown that VitA deficient animals have diminished Th1 and Th17 responses (14, 21). When an IL15 transgenic mice, in which IL15 is important for T cell activation and proliferation is locally elevated in the lamina propria (LP) and MLN, were made allergic to OVA and fed RA, the Th1 immune response was induced in cooperation with IL12 and these mice had no detectable Foxp3+ Treg (22). Conversely, Th2 cell differentiation can be promoted by RA as well through promotion of increased IL4 production (23). For the effect of RA on the development of T cells in general, the concentration of RA and the mediators present in the micro-environment are very important (13, 15).

RA also exhibits inhibitory effects on B cell proliferation, immunoglobulin (Ig)A and IgE production (15, 18). In an OVA or peanut allergy mouse model in which allergic mice were injected with DC treated in vitro with RA, clinical symptoms like anaphylaxis, mast cell activation, serum specific IgE and IgG1, and Th2 responses were reduced compared to non- treated controls (20). When 9-cis-RA was provided together with an OVA challenge in a BALB/c OVA allergy mouse model a reduction of OVA-IgE, was found (24). In murine models for asthma lower VitA levels led to increased asthma symptoms and Th2 responses (15). The RA signalling pathway may interact with components in its micro-environment, like cytokines and other dietary components, so depending on these other signalling molecules RA may skew the balance towards a tolerogenic or inflammatory phenotype in vivo.

The role of VitA in the prevention of CMA remains to be further explored. In the current study dietary VitA depletion and supplementation after weening were compared to feeding diets with normal VitA levels with or without GFA in a C3H/HeOuJ mouse model for CMA to investigate it’s preventive effect on CMA.

136

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 136 VITA AND GFA IN CMA

MATERIALS AND METHODS

Diets The semi-synthetic cow’s milk free AIN93G chow (Research Diet Services, Wijk bij Duurstede, the Netherlands) is used as control diet (25). In all diets casein and whey were replaced by soy protein. The diets were composed without VitA (VitA), with ‘standard’ 4000 IU VitA/kg (s) (according to AIN93G formulation), or with 8000 IU VitA/kg (VitA+). For the latter supplementation dose we chose twice the standard amount of VitA, Molenaar et al. did not report any adverse effects for this dose and time frame (26). Before start of dietary intervention, mice were born and weened by their mothers which received standard chow (without whey or casein) containing the standard VitA dose of 4000 IU/kg. Beyond VitA diets were either or not supplemented with a mixture of oligosaccharides. One percent iso-caloric supplementation of scGOS (Vivinal GOS, Friesland Campina Domo, Zwolle, The Netherlands), lcFOS (Raftiline HP, Orafti, Wijchen, The Netherlands), and pAOS (Südzucher, Mannheim, Germany) in a 9:1:2 ratio was added at the expense of cellulose and lactose.

Animals Four weeks old and body weight > 11 gram, specific pathogen-free female C3H/HeOuJ mice (Charles River, Sulzfeld, Germany) were housed under conventional housing at the animal facility (Intravacc, Bilthoven, the Netherlands). Mice were held in a light/dark cycle of 12h/ 12h, controlled 65-70% relative humidity and 22 ± 2°C. temperature with ad libitum access to tap water and pelleted food. Upon arrival, mice were fed the specific diets which was continued throughout the study: control diet (s), VitA+ diet and VitA- diet, diet supplemented with GFA (GFA) and diet supplemented with GFA and VitA (VitA+/GFA) or without VitA (VitA-/GFA) (Figure 1). Mice (n=10-15) were housed in groups and kept in two cohorts (5/cage) in Makrolon III-H cages. For cage enrichment mice were provided with two red-transparent polycarbonate igloo’s (no nesting-material due to the potential effect on allergy-parameters (27)) on 9kGy irradiated sawdust bedding (Lignocel 9s, J. Rettenmaier & Söhne GmbH, Germany). Subsequently animal 7 procedures, housing and care were performed in accordance with the EU-guidelines (2010/63/ EU) and approved by a licensed Animal Ethics Committee (DEC-Consult, Soest, the Netherlands).

Figure 1. Experimental setting. GFA, short-chain Galacto-oligosaccharides (scGOS), long-chain Fructo- oligosaccharides (lcFOS), and pectin-derived Acidic-oligosaccharides (pAOS); CT, cholera toxin; i.d., intradermal.

137

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 137 CHAPTER 7

Oral sensitization Following 13 days of pre-treatment with the diets, mice were sensitized weekly for five times by means of oral gavage, with 10μg Cholera Toxin (CT, List Biological Laboratories, USA) or 10μg CT/ 20 mg whey protein concentrate (sweet whey protein concentrate 60 (sWPC60), Milei, Germany) per 500μl DPBS (Life Technologies, Inc., Invitrogen, USA) per oral gavage using a blunt needle (stainless steel) (Kent Scientific Corporation, USA). Temperature transponders (IPTT-300, BMDS, USA) were injected subcutaneously under isoflurane gas anaesthesia, between the 3rd and 4th sensitization (Figure 1).

Allergic skin response To measure the acute allergic skin response whey protein was injected intradermally (i.d.) (10 μg whey /20 μl PBS) in the ear pinnae at day 35. The acute allergic skin response was determined as delta (Δ) ear swelling (expressed as Δ μm) by subtraction the ear thickness measured just before i.d. injection from the ear thickness measured one hour after i.d. challenge using a digital micrometre (Mitutoyo, Veenendaal, the Netherlands). The body temperature and the anaphylactic shock symptoms were scored according to the method previously described by van Esch et al. (28) and adapted to a decision table where body temperature and shock score are both assessed by Kerperien et al. (29).

Oral challenge Eighteen hours before the end of the study, mice were challenged intragastrically (i.g). with 50 mg whey/500μl DPBS to activate the mucosal gastrointestinal immune response. At day 37 under terminal isoflurane gas anaesthesia, blood and MLN were isolated and stored for further analysis.

Whey-specific immunoglobulins and mucosal mast cell protease-1 Whey-specific immunoglobulins and mucosal mast cell protease-1 (mMCP1) were determined by means of an enzyme-linked immunosorbent assay (ELISA) previously described (30) and serum mMCP1 was quantified using a ELISA Ready-SET-Go!® kit (eBioscience, San Diego, USA) according to manufacturer’s protocol.

qPCR One centimetre of the intestinal mid small intestine was collected after sacrifice and stored in RNAlater™ (Qiagen GmbH, Hilden, Germany) at 4°C until further processing, described previously by Kerperien et al. (31). mRNA levels were calculated with CFX Manager software (version 1.6) and corrected for the expression of Ribosomal protein S13 (Rps13) with 100 x 2^(Rps13-gene of interest) as described previously (32). Validated primers for Rps13, Il4, Il5, Il10, Ifnγ, Foxp3, and Aldh1a2, were purchased from SAbioscience (Qiagen, German Town, MD, USA).

138

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 138 VITA AND GFA IN CMA

Flow cytometry of immune cells The splenocyte suspension was lysed to remove red blood cells as described previously (33). MLN and spleen cell suspensions were made with a 70 μm cell strainers (Thermo Fisher Scientific, Amsterdam, the Netherlands), resuspended in RPMI 1640, 10% foetal bovine serum and penicillin (100 U/mL)/streptomycin (100 μg/mL) and quantified using a Coulter Z1 particle counter (Beckmann Coulter, Brea, USA). One million cells per well were added to polypropylene V-bottom 96 well plates (BD Biosciences, Germany) and blocked with rat anti- mouse CD16/32 (BD Biosciences, Germany) for 20 minutes at 4°C to block non-specific binding sites. For surface staining the cells were incubated with the different antibodies of the staining panels for 30 minutes at 4°C. After the antibody incubation, the cells were fixed using BD cell fix (Becton Dickinson, Belgium). Directly after the staining procedure for extracellular markers, these cells were permeabilized using a fixation/permeabilization buffer (eBioscience, San Diego, USA) overnight. The next day, the cells were blocked again and incubated with the intracellular marker antibodies for 30 minutes at 4°C. Then the cells were fixed with the BD cell fix. Fluorescence Minus Ones (FMOs) and unstained cells were used as controls. In addition, compensation staining with UltraComp beads (eBioscience, San Diego, USA) was performed to correct for the spectral overlap, which occurs in multi-colour staining. Cell surface staining antibodies used were: CD8a-APC-Cy7, CD11c-PerCP-Cy5.5 and CD25-Pe- Cy7 from eBiosciences (San Jose, CA, USA) and CD4-PerCP-Cy5.5, CD69-Pe-Cy7, CD11b-PE, CD103-APC, CX3CR1-FITC, B220-APC from eBiosciences (San Diego CA, USA). Cytokines and intracellular staining for IL10-PE, IFNγ-APC, IL4-Pe-Cy7, IL17α-FITC, Rorγ-PE, Gata3-PE, Tbet-APC, and Foxp3-APC came from Ebiosciences (San Diego CA, USA). The analysis of the stained cells was performed using FACS Canto II cytometer (BD Biosciences, USA) and FACS Diva software (BD Biosciences, Germany).

Statistical Analysis To determine the sample size historical data with sWPC60 as whey source and ear swelling 7 of allergic mice (123 ± 20 μm) as primary outcome parameter were used. The estimated effect size was set at 20% with 80% power and an α of 5%. Statistical analysis was performed using GraphPad Prism 7 (GraphPad Software, San Diego, CA, USA). Not normal distributed numerical data was LOG transformed. The data are represented either as the mean, mean ± SEM or Tukey box-and-whiskers plots. One-way ANOVA and post hoc Bonferroni multiple comparisons test with selected groups were used to analyse the data. Selected groups are sham vs. allergic mice (as model control), allergic mice vs. allergic mice fed VitA+, allergic mice vs. allergic mice fed GFA, allergic mice fed VitA+ vs. allergic mice fed VitA+/GFA and allergic mice fed GFA vs. allergic mice fed VitA+/GFA. Kruskal-Wallis was used to analyse the anaphylactic shock score (categorial data), followed by Dunn’s multiple comparison post-test. P-value < 0.05 was considered significant.

139

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 139 CHAPTER 7

RESULTS

Allergic symptoms The allergic symptoms measured were the acute allergic skin response, anaphylaxis (Table 1) and body temperature. The acute allergic skin response, was significantly different between the sham sensitized mice fed control diet (sham) and whey sensitized mice fed control diet (allergic mice) (p<0.0001) (Figure 2A). In addition, all other treatment groups had a significantly higher acute allergic skin response than the sham (p<0.0001) (Figure 2A). Only whey sensitized mice supplemented with the combination of VitA and GFA (allergic mice fed VitA+/GFA) had a significant lower acute allergic skin response compared to whey sensitized mice supplemented with VitA (allergic mice fed VitA+) (p<0.05). Measured at thirty minutes, the anaphylactic symptom score was significantly higher in the allergic mice, allergic mice fed VitA+ and allergic mice fed GFA compared to the sham (p<0.001), while this was not significant for the allergic mice fed VitA+/GFA (Figure 2B). There were no significant differences between groups in body temperature at 30 minutes (Figure 2C). No difference was observed with the groups fed the control diet compared to allergic mice fed the VitA- diet with or without GFA apart from the point that the shock scores of two mice of these groups were so severe that they reached the humane endpoint (Table 1).

Table 1. Anaphylactic shock symptoms at one hour after the intradermal ear challenge of all mice. Data are represented in percentages per treated group. Vitamin A Amount in chow; intake GFA % Anaphylactic shock % Humane Endpoint (mice/total) (mice/mice with shock) - 80 (12/15) 8 (1/15) Deprivation 0 IU/kg; 0 IU/day + 70 (7/10) 14 (1/7) - 90 (9/10) 0 (0/9) Normal 4000 IU/kg1; ≈ 20 IU/day + 80 (8/10) 0 (0/8) - 80 (8/10) 0 (0/8) Supplementation 8000 IU/kg2; ≈ 40 IU/day + 40 (4/10) 0 (0/4)

1. Reeves, P. G. & Suppl, M. Symposium : Animal Diets for Nutritional and Toxicological Research Components of the AIN-93 Diets as Improvements in the AIN-76A Diet 1 , 2. 838–841 (1997) 2. Molenaar R, Knippenberg M, Goverse G, Olivier BJ, de Vos AF, O’Toole T, et al. Expression of retinaldehyde dehydrogenase enzymes in mucosal dendritic cells and gut-draining lymph node stromal cells is controlled by dietary vitamin A. J Immunol. 2011;186(4):1934-42.

140

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 140 VITA AND GFA IN CMA

Figure 2. Clinical symptoms in control and whey sensitized mice fed VitA and GFA. A; The acute allergic skin response measured one hour after intra-dermal injection. Values are means ± SEM, n=8-10, a one- way ANOVA and Bonferroni post hoc was used. Δ μm Indicates ear thickness before i.d. injection deducted from thickness after i.d. injection. B; The anaphylactic symptom scores measured at thirty minutes after challenge. Values are means ± SEM, n = 8-10. Data is calculated with the Kruskal-Wallis and Dunn’s post hoc test. C; The body temperature was measured 30 minutes after i.d. injection via transponder read out. Values are means ± SEM, n=8-10, a one-way ANOVA and Bonferroni post hoc was used. Asterix indicates a significant difference compared to sham, **** p< 0.0001, *** p< 0.001, * p< 0.05. The ‘and’ sign indicates a significant difference compared to the VitA+ group. GFA, short-chain Galacto-oligosaccharides/ long- chain Fructo-oligosaccharides/ pectin-derived Acidic-oligosaccharides; s, standard 4000 IU/kg vitamin A; +, 8000 IU/kg vitamin A.

141

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 141 CHAPTER 7

Allergic mediators in serum mMCP1 and whey-specific IgE and IgG2a were measured in serum, to study mucosal mast cell degranulation and the humoral responses in relation to the acute allergic skin response. All allergic mice showed a clear induction of whey-specific IgE which was significantly higher in all dietary intervention groups compared to the sham (p<0.0001) (Figure 3A). Whey-specific IgG2a and mMCP1 were only significantly increased in the allergic mice fed GFA compared to sham (p<0.05), although all groups showed a similar pattern (Figure 3 B, C).

Real time quantitative PCR To measure markers of local T helper cell responses, mRNA levels for Ifnγ (Th1), Il4 and Il5 (Th2), Il10 and Foxp3 (Treg) and Aldh1a2 (RA converting enzyme) were measured in the mid small intestine. For Ifnγ, Il4, Il5 and Foxp3 mRNA a significant increase was observed in allergic mice fed VitA+/GFA compared to allergic mice fed VitA+ (p<0.05) (Figure 4A-C, E) and Il10 showed the same tendency (Figure 4D). Ifnγ and Foxp3 mRNA expression in allergic mice fed VitA+/GFA also increased compared to sham (p<0.05) (Figure 4C, E). mRNA Il4 and Foxp3 mRNA was decreased in allergic mice fed VitA+ compared to allergic mice (p<0.05) (Figure 4A, E). Aldh1a2 remained unaffected (Figure 4F).

Dendritic cells and T cells The possible effect of dietary VitA+ and GFA on specific T cells and DC subsets in the spleen as well as in the MLN was also determined. First Th2 (CD4+CD8α-Gata3+), Th1 (CD4+CD8α-Tbet+), Th17 (CD4+CD8α-Rorγ+) and Treg (CD4+CD8α-Foxp3+) subsets in the spleen were analysed. Due to the study size the experiment was conducted in two cohorts, only a valuable signal in both cohorts for Th1 (Tbet) and Treg (Foxp3) was detected in the spleen (Figure 5B, D). Th2, Th1, Th17, and Treg were also analysed in MLN, but no differences between the groups were found (Supplemental Figure 1). In the spleen intracellular cytokine expression of IFNγ and IL4 was measured in CD4+CD11c- Th-cells (Figure 6A, B) or CD11c+CD4- phagocytes (Figure 6C, D). The percentage CD4+CD11c-IL4+ Th cells in the allergic mice was lower than sham (p<0.05) (Figure 6B).The CD4+CD11c-IL4+ Th cells from allergic mice fed VitA+ and allergic mice fed GFA were significant higher than the allergic mice (p<0.05) (Figure 6B).Compared to sham, allergic mice fed VitA+/GFA showed an increase in CD11c+CD4-IFNγ+ phagocytes (p=0.064) (Figure 6C). Different DC play a role in T cell development so the plasmacytoid DC (pDC; B220+) and the migratory DC (CD103+CX3CR1-) subsets were analysed in the MLN (Supplemental Figure 2A) and spleen (Supplemental Figure 2B). These DC subsets may be able to skew T cells towards a regulatory phenotype or enhance immunity, so also Tr1 (CD4+Foxp3+CXCR3+) was evaluated in the MLN (Supplemental Figure 2A) and spleen (Supplemental Figure 2B). For both the DC populations and the Tr1 population in MLN and spleen no significant differences were observed.

142

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 142 VITA AND GFA IN CMA

Figure 3. Whey-specific IgE, IgG2a and mMCP-1 measured in serum of mice at day 37. A; Whey-specific- IgE (AU), B; Whey-specific-IgG2a (AU) and mMCP1 (ng/ml) were measured in serum. Values are means ± SEM, n= 8-10. Data was calculated with a one-way ANOVA and Bonferroni post hoc test, if required, LOG transformation was used to normalize data distribution. Asterix indicates a significant difference compared to sham, **** p< 0.0001, * p< 0.05. GFA, short-chain Galacto-oligosaccharides/ long-chain Fructo-oligosaccharides/ pectin-derived Acidic-oligosaccharides; mMCP1, mouse mast cell protease 1; AU, arbitrary units; s, standard 4000 IU/kg vitamin A; +, 8000 IU/kg vitamin A.

143

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 143 CHAPTER 7

Figure 4. mRNA markers in the intestine in sham or whey sensitized mice fed a control, VitA or GFA diet. In the mid small intestine the relative expression of transcription factor Il4, Il5, Ifnγ, Il10, Foxp3 and Aldh1a2 relative to Rps13 is shown. Values are means ± SEM, n = 6-10. Asterix indicates a significant difference compared to sham, *** p < 0.001, ** p < 0.01, * p < 0.05. msi; mid small intestine, Il4, interleukin 4; Il5, interleukin 5; Ifnγ, interferon gamma; Il10, interleukin 10; Foxp3, forkhead box P3; Aldh1a2, aldehydedehydrogenase 1a2; Rps13, ribosomal protein S13; GFA, short-chain Galacto-oligosaccharides/ long-chain Fructo-oligosaccharides/ ppectin-derived Acidic-oligosaccharides; s, standard 4000 IU/kg vitamin A; +, 8000 IU/kg vitamin A.

144

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 144 VITA AND GFA IN CMA

Figure 5. T cell population, Th1, Th2, Th17 and Treg, measured in the spleen. Spleens were removed after sacrifice of mice and stained for CD4-PERCP-Cy5.5+ CD8α-APC-Cy7- then subpopulations were gated, (A) one for Th2 intracellular Gata3-PE+, (B) another for Th1 the intracellular Tbet-APC+, (C) a third for Th17 intracellular Rorγ-PE+ and (D) the last for Treg intracellular Foxp3-APC+. Representative flow cytometric plots are shown in Supplemental Figure 3 and 4. Values are means ± SEM, n= 3-10 and were calculated with a one-way ANOVA and Bonferroni post hoc test. CD, cluster of differentiation; GFA, short-chain Galacto-oligosaccharides/ long-chain Fructo-oligosaccharides/ pectin-derived Acidic-oligosaccharides; s, standard 2000 IU/kg vitamin A; +, 8000 IU/kg vitamin A.

145

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 145 CHAPTER 7

Figure 6. Cytokine production in CD4+ T cell and CD11c+ phagocyte cell population in the spleen. Spleens were removed after sacrifice of mice and stained for CD4-APC-Cy7+ / CD11c- PERCP-Cy5.5- then subpopulations were gated for IFNγ-APC (A) and IL4-PE-Cy7 (B). CD11c-PERCP-Cy5.5+ was gated separately in a histogram, then subpopulations were gated for IFNγ-APC (C) and IL4-PE-Cy7 (D). Representative flow cytometric plots are shown in Supplemental Figure 5. Values are means ± SEM, n= 4-10 and were calculated with a one-way ANOVA and Bonferroni post hoc test. Asterix indicates a significant difference compared to sham, *** p< 0.001, ** p< 0.01, * p< 0.05. CD, cluster of differentiation; GFA, short-chain Galacto-oligosaccharides/ long-chain Fructo-oligosaccharides/ pectin-derived Acidic- oligosaccharides; s, standard 2000 IU/kg vitamin A; +, 8000 IU/kg vitamin A.

DISCUSSION

The active metabolite of VitA, RA, is essential in cellular signalling in growth and development and is an active participant in immune modulation (34). In this study the effects of VitA deprivation or supplementation with or without GFA on the development of murine CMA were examined.

Allergic symptoms were not affected by VitA deprivation or supplementation only. The mice used were born from and weened by mothers with a normal VitA status and this VitA is stored in fat tissue. These mice had a normal VitA status upon arrival, although the diet was depleted for VitA two weeks before sensitization, the mice may still have had sufficient levels for proper physiological function. Hence it remains to be investigated what would be the effects of

146

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 146 VITA AND GFA IN CMA

prolonged VitA deprivation on CMA development. VitA+ supplementation only did not affect the allergy development in mice.

In contrast with our findings, there are some food allergy studies showing positive effects of VitA supplementation on oral tolerance development. In neonatal mice it was shown that VitA levels at one week of age are three times lower than at three weeks of age, and this corresponds with induction of oral tolerance to OVA after challenge which is present at three weeks and not in the first week (35). When one-week old mice were supplemented with extra VitA, these mice showed the same level of oral tolerance for OVA as three week old mice (35). For the influence of VitA on murine allergic asthma there are conflicting results. VitA deficiency could lead to exacerbated OVA induced lung inflammation where others reported a reduction of asthma symptoms (15). In a human setting, excess of VitA increased the risk of asthma development and aerosol administration of VitA to asthmatic children was of no benefit, however increased VitA serum levels were positively correlated with improved lung function in children with stable asthma (15, 36, 37).

In previous murine studies GFA alone could partially prevent the development of CMA allergic symptoms (28, 31, 38, 39), while in the current study GFA did not show this regulation. One of the reasons that GFA did not show a significant preventive effect on the outcome of CMA could be that our study was performed at a different animal facility compared to the previous animal experiments (8, 28, 31). For the C3H/HeOuJ female CMA model, clinical symptoms like the acute allergic skin response, anaphylaxis and the generation of whey specific Ig’s differ between animal facilities (40). This difference in outcome could possibly be attributed to differences in the composition of the intestinal microbiota, which is influenced by housing conditions at the breeding facility and housing conditions at the animal facility where the experiments are conducted (41). Different species of microbiota may use GFA distinctively, and subsequently modify the immune response in a different manner. Interestingly, the allergic symptoms (acute 7 skin response) were significantly lower in allergic mice fed VitA+/GFA compared to allergic mice fed VitA+, while GFA alone was not effective. Furthermore, the allergic mice fed VitA+/GFA also tended to have reduced anaphylactic symptoms at 30 minutes compared to allergic mice fed the control diet, implying an interaction between the effect of GFA and VitA+.

RA has a very important role in inducing oral tolerance via skewing of the CD103+ DC towards a regulatory phenotype (13, 17). This results in more Treg and less activated T helper cell populations. We evaluated markers of different T cell subsets in the mid small intestine with qPCR. We observed that several immune markers (Il4, Il5, Ifnγ, Il10 and Foxp3 mRNA) were increased in the allergic mice fed VitA+/GFA compared to sham. Previously, this general increase of these mRNA markers in the intestine by GFA alone was observed when the preventive effect of GF versus GFA was evaluated in a CMA mouse model (31). In current study allergic mice

147

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 147 CHAPTER 7

fed VitA+/GFA also affected immune markers in relation to a lower acute skin response. This may imply that the action of GFA requires a sufficient amount of available VitA. Indeed, the necessary availability of VitA to support dietary fibre in the generation of functional Treg and to establish oral tolerance was elegantly shown by Tan et al. (42).

Alike the microbiota also VitA levels or ALDH expression levels may differ between cohorts of mice, hence this may also have contributed to differences in results using GFA. ALDH is required to convert VitA in its active form retinoic acid and its expression and activity may be modulated by environmental triggers. In the current study Aldh1a2 mRNA expression in the mid ileum was not affected, however it is now known that most ALDH1a2 activity is found in the first part of the intestine (43). In addition, for all three ALDH1a isoforms expression differs in the small or large intestine, the exact location within the intestine and the expression differs between laboratory mouse strains BALB/c and C57BL6 (43). Hence it could be interesting to further study Aldh mRNA expression levels in other parts of the intestine (44).

DC traffic from the mucosa to the MLN and instruct T cells that migrate towards the intestine and spleen. Therefore, we assessed the MLN and spleen, for Th1, Th2, Th17, Treg and pDC and CD103+ DC. Unexpectedly, the percentage of splenic CD4+CD11c-IL4+ Th2 cells was found to be lower in the allergic mice compared to sham. One explanation could be that this Th2 frequency is lower in this late part of the sensitization phase and that it was increased in an earlier phase contributing to IgE isotype switching in B cells. Another explanation could be that these CD4+CD11c-IL4+ Th2 cells are re-distributed to the lamina propria of the intestine.

Interestingly, a distinct population of phagocytes producing IL4 or IFNγ was identified. Not much is known about these types of cells which include DC. However, IL4 can lead to IL4 production in DC which possibly acts in an autocrine manner (45). For IFNγ there are some papers suggesting a role for IFNγ as an autocrine regulator for DC function as well. Exogenous IL12 or CD40 ligation combined with high levels of IL18 could induce IFNγ production in immature BALB/c derived DC in vitro (46). One study shows that human monocytes derived from buffy coats pre-treated with IFNγ become regulatory DC (47). Subsequently these IFNγ+ DC produced more IL10 upon stimulation, and expressed more inhibitory molecules and downregulated T cell migration in vitro (47). This could suggest that the IFNγ+ phagocytes observed in the spleen could be a more regulatory DC phenotype and the frequency of these cells tended to increase in the mice fed VitA+/GFA.

No difference in the Foxp3+ cell type was observed in the MLN, only significantly moreFoxp3 mRNA was expressed when allergic mice were fed VitA+/GFA. The kinetics and location may influence the Treg result. However, also other Treg and regulatory mediators may be involved in the CMA protective effect of VitA+/GFA. Dawicki et al. found in two murine allergy models

148

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 148 VITA AND GFA IN CMA

another set of Treg, of which the Foxp3-CD25+LAG3+CD49b- is driven by RA, which could suppress allergic responses to oral allergen challenges to OVA and peanut (20). In our study LAG3 or CD49b was not analysed and therefore we cannot deduct the possible effect on the Foxp3-CD25+ regulatory cell population. However, in previous studies GFA was shown to require IL10 and TGFβ to protect against development of symptoms of CMA, and indeed IL10 mRNA expression was increased in allergic mice fed VitA+/GFA TGFβ mRNA was not determined). Hence VitA+ may have potentiated the mechanism by which GFA is able to enhance IL10, hereby contributing to its protective effect. This is in line with our observations in the CMA model were the preventive effect of GFA on CMA was supressed via blockage of IL10 receptor or TGFβ (30). Previous studies using the CMA model showed that CD25+ Treg from allergic mice fed GFA could transfer tolerance to naïve mice (38). This could involve Foxp3 or LAG3 Treg, however this is currently unknown. There are indications that regulation of food allergy by RA may act via Treg subsets which can be either Foxp3+ or Foxp3- (9, 13, 15, 17, 20, 42).

The effect of VitA on the intestinal immune system is part of extensive research but there are still many things unknown. In this study VitA supplementation or GFA alone did not lead to a suppression of CMA symptoms in mice, while VitA+/GFA combined did. This indicates that VitA may support the biological function of GFA and may help to reduce CMA development when added together with GFA.

149

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 149 CHAPTER 7

REFERENCES

1. Nowak-Wegrzyn A, Szajewska H, Lack Stage. Arch Immunol Ther Exp (Warsz). G. Food allergy and the gut. Nature 2015;63(4):251-68. reviews Gastroenterology & hepatology. 12. Theodosiou M, Laudet V, Schubert M. From 2017;14(4):241-57. carrot to clinic: an overview of the retinoic 2. Du Toit G, Foong RM, Lack G. Prevention acid signaling pathway. Cell Mol Life Sci. of food allergy - Early dietary interventions. 2010;67(9):1423-45. Allergol Int. 2016;65(4):370-7. 13. Erkelens MN, Mebius RE. Retinoic Acid and 3. Burks AW, Sampson HA, Plaut M, Lack G, Immune Homeostasis: A Balancing Act. Akdis CA. Treatment for food allergy. The Trends Immunol. 2017;38(3):168-80. Journal of allergy and clinical immunology. 14. Hall JA, Grainger JR, Spencer SP, Belkaid 2018;141(1):1-9. Y. The role of retinoic acid in tolerance and 4. Le Doare K, Holder B, Bassett A, Pannaraj immunity. Immunity. 2011;35(1):13-22. PS. Mother’s Milk: A Purposeful Contribution 15. Hufnagl K, Jensen-Jarolim E. Vitamin A to the Development of the Infant Microbiota and D in allergy: from experimental animal and Immunity. Frontiers in immunology. models and cellular studies to human 2018;9:361. disease. Allergo J Int. 2018;27(3):72-8. 5. Louis P, Flint HJ, Michel C. How to 16. Faria AM, Gomes-Santos AC, Goncalves JL, Manipulate the Microbiota: Prebiotics. Moreira TG, Medeiros SR, Dourado LP, et al. Advances in experimental medicine and Food components and the immune system: biology. 2016;902:119-42. from tonic agents to allergens. Frontiers in 6. Wopereis H, Sim K, Shaw A, Warner JO, Knol immunology. 2013;4:102. J, Kroll JS. Intestinal microbiota in infants at 17. Coombes JL, Siddiqui KR, Arancibia- high risk for allergy: Effects of prebiotics and Carcamo CV, Hall J, Sun CM, Belkaid Y, et role in eczema development. The Journal of al. A functionally specialized population allergy and clinical immunology. 2017. of mucosal CD103+ DCs induces Foxp3+ 7. Moro G, Arslanoglu S, Stahl B, Jelinek J, regulatory T cells via a TGF-beta and retinoic Wahn U, Boehm G. A mixture of prebiotic acid-dependent mechanism. J Exp Med. oligosaccharides reduces the incidence of 2007;204(8):1757-64. atopic dermatitis during the first six months 18. Iwata M, Hirakiyama A, Eshima Y, Kagechika of age. Arch Dis Child. 2006;91(10):814-9. H, Kato C, Song SY. Retinoic acid imprints 8. Schouten B, Van Esch BC, Kormelink gut-homing specificity on T cells. Immunity. TG, Moro GE, Arslanoglu S, Boehm G, 2004;21(4):527-38. et al. Non-digestible oligosaccharides 19. Seo GY, Lee JM, Jang YS, Kang SG, Yoon reduce immunoglobulin free light-chain SI, Ko HJ, et al. Mechanism underlying the concentrations in infants at risk for allergy. suppressor activity of retinoic acid on IL4- Pediatr Allergy Immunol. 2011;22(5):537-42. induced IgE synthesis and its physiological 9. Schouten B, van Esch BC, Hofman GA, implication. Cell Immunol. 2017;322:49-55. Boon L, Knippels LM, Willemsen LE, et al. 20. Dawicki W, Li C, Town J, Zhang X, Gordon Oligosaccharide-induced whey-specific JR. Therapeutic reversal of food allergen CD25(+) regulatory T-cells are involved in sensitivity by mature retinoic acid- the suppression of cow milk allergy in mice. differentiated dendritic cell induction of J Nutr. 2010;140(4):835-41. LAG3(+)CD49b(-)Foxp3(-) regulatory T cells. 10. Czarnewski P, Das S, Parigi SM, Villablanca J Allergy Clin Immunol. 2017;139(5):1608-20 EJ. Retinoic Acid and Its Role in Modulating e3. Intestinal Innate Immunity. Nutrients. 21. Brown CC, Esterhazy D, Sarde A, London M, 2017;9(1). Pullabhatla V, Osma-Garcia I, et al. Retinoic 11. Brugman S, Perdijk O, van Neerven acid is essential for Th1 cell lineage stability RJ, Savelkoul HF. Mucosal Immune and prevents transition to a Th17 cell Development in Early Life: Setting the program. Immunity. 2015;42(3):499-511.

150

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 150 VITA AND GFA IN CMA

22. DePaolo RW, Abadie V, Tang F, Fehlner- Non-digestible oligosaccharides modulate Peach H, Hall JA, Wang W, et al. Co-adjuvant intestinal immune activation and suppress effects of retinoic acid and IL-15 induce cow’s milk allergic symptoms. Pediatr inflammatory immunity to dietary antigens. Allergy Immunol. 2014;25(8):747-54. Nature. 2011;471(7337):220-4. 32. Garcia-Vallejo JJ, Van Het Hof B, Robben J, 23. Goverse G, Molenaar R, Macia L, Tan Van Wijk JA, Van Die I, Joziasse DH, et al. J, Erkelens MN, Konijn T, et al. Diet- Approach for defining endogenous reference Derived Short Chain Fatty Acids Stimulate genes in gene expression experiments. Intestinal Epithelial Cells To Induce Mucosal Analytical biochemistry. 2004;329(2):293-9. Tolerogenic Dendritic Cells. Journal of 33. Kostadinova AI, Meulenbroek LA, van Esch immunology. 2017;198(5):2172-81. BC, Hofman GA, Garssen J, Willemsen 24. Heine G, Hollstein T, Treptow S, Radbruch LE, et al. A Specific Mixture of Fructo- A, Worm M. 9-cis retinoic acid modulates Oligosaccharides and Bifidobacterium the type I allergic immune response. The breve M-16V Facilitates Partial Non- Journal of allergy and clinical immunology. Responsiveness to Whey Protein in Mice 2017. Orally Exposed to beta-Lactoglobulin- 25. Reeves PG. Components of the AIN-93 diets Derived Peptides. Front Immunol. as improvements in the AIN-76A diet. The 2016;7:673. Journal of nutrition. 1997;127(5 Suppl):838S- 34. Das BC, Thapa P, Karki R, Das S, Mahapatra 41S. S, Liu TC, et al. Retinoic acid signaling 26. Molenaar R, Knippenberg M, Goverse pathways in development and diseases. G, Olivier BJ, de Vos AF, O’Toole T, et al. Bioorg Med Chem. 2014;22(2):673-83. Expression of retinaldehyde dehydrogenase 35. Turfkruyer M, Rekima A, Macchiaverni P, Le enzymes in mucosal dendritic cells and Bourhis L, Muncan V, van den Brink GR, et al. gut-draining lymph node stromal cells is Oral tolerance is inefficient in neonatal mice controlled by dietary vitamin A. J Immunol. due to a physiological vitamin A deficiency. 2011;186(4):1934-42. Mucosal Immunol. 2016;9(2):479-91. 27. Pasalic I, Bosnjak B, Tkalcevic VI, Jaran 36. Parr CL, Magnus MC, Karlstad O, Holvik DS, Javorscak Z, Markovic D, et al. Cage K, Lund-Blix NA, Haugen M, et al. Vitamin enrichment with paper tissue, but not plastic A and D intake in pregnancy, infant tunnels, increases variability in mouse supplementation, and asthma development: model of asthma. Lab Anim. 2011;45(2):121-3. the Norwegian Mother and Child Cohort. Am 28. Schouten B, van Esch BC, Hofman GA, van J Clin Nutr. 2018;107(5):789-98. Doorn SA, Knol J, Nauta AJ, et al. Cow milk 37. Bai YJ, Dai RJ. Serum levels of vitamin A allergy symptoms are reduced in mice fed and 25-hydroxyvitamin D3 (25OHD3) as dietary synbiotics during oral sensitization reflectors of pulmonary function and quality 7 with whey. J Nutr. 2009;139(7):1398-403. of life (QOL) in children with stable asthma: 29. Kerperien J, Veening-Griffioen D, O-ja A, A case-control study. Medicine (Baltimore). Wehkamp T, Jeurink PV, Garssen J, et 2018;97(7):e9830. al. Dietary vitamin D supplementation is 38. Schouten B, van Esch BC, Hofman GA, ineffective in preventing murine cow’s milk de Kivit S, Boon L, Knippels LM, et al. A allergy, irrespective of the presence of non- potential role for CD25+ regulatory T-cells digestible oligosaccharides. manuscript in in the protection against casein allergy by preparation. dietary non-digestible carbohydrates. Br J 30. Kerperien J, Veening-Griffioen D, Wehkamp Nutr. 2012;107(1):96-105. T, van Esch B, Hofman GA, Cornelissen P, 39. de Kivit S, Kostadinova AI, Kerperien J, et al. IL-10r or TGF-beta Neutralization Morgan ME, Muruzabal VA, Hofman GA, et Abrogates the Protective Effect of a Specific al. Dietary, nondigestible oligosaccharides Nondigestible Oligosaccharide Mixture in and Bifidobacterium breve M-16V suppress Cow-Milk-Allergic Mice. J Nutr. 2018. allergic inflammation in intestine via 31. Kerperien J, Jeurink PV, Wehkamp T, van targeting dendritic cell maturation. J Leukoc der Veer A, van de Kant HJ, Hofman GA, et al. Biol. 2017;102(1):105-15.

151

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 151 CHAPTER 7

40. van Esch BC, van Bilsen JH, Jeurink PV, Garssen J, Penninks AH, Smit JJ, et al. Interlaboratory evaluation of a cow’s milk allergy mouse model to assess the allergenicity of hydrolysed cow’s milk based infant formulas. Toxicol Lett. 2013;220(1):95- 102. 41. Thoene-Reineke C, Fischer A, Friese C, Briesemeister D, Gobel UB, Kammertoens T, et al. Composition of intestinal microbiota in immune-deficient mice kept in three different housing conditions. PLoS One. 2014;9(11):e113406. 42. Tan J, McKenzie C, Vuillermin PJ, Goverse G, Vinuesa CG, Mebius RE, et al. Dietary Fiber and Bacterial SCFA Enhance Oral Tolerance and Protect against Food Allergy through Diverse Cellular Pathways. Cell Rep. 2016;15(12):2809-24. 43. Goverse G, Olivier BJ, Molenaar R, Knippenberg M, Greuter M, Konijn T, et al. Vitamin A metabolism and mucosal immune function are distinct between BALB/c and C57BL/6 mice. Eur J Immunol. 2015;45(1):89-100. 44. Sellers RS, Clifford CB, Treuting PM, Brayton C. Immunological variation between inbred laboratory mouse strains: points to consider in phenotyping genetically immunomodified mice. Vet Pathol. 2012;49(1):32-43. 45. Maroof A, Penny M, Kingston R, Murray C, Islam S, Bedford PA, et al. Interleukin-4 can induce interleukin-4 production in dendritic cells. Immunology. 2006;117(2):271-9. 46. Stober D, Schirmbeck R, Reimann J. IL- 12/IL-18-dependent IFN-gamma release by murine dendritic cells. J Immunol. 2001;167(2):957-65. 47. Svajger U, Obermajer N, Jeras M. IFN- gamma-rich environment programs dendritic cells toward silencing of cytotoxic immune responses. J Leukoc Biol. 2014;95(1):33-46.

152

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 152 VITA AND GFA IN CMA

SUPPLEMENTAL FIGURES

Supplemental figure 1. T cell population, Th1, Th2, Th17 and Treg, measured in the MLN. MLN were removed after sacrifice of mice and stained for CD4-PERCP-Cy5.5+ CD8α-APC-Cy7- then subpopulations were gated, (A) one for Th2 intracellular Gata3-PE+, (B) another for Th1 the intracellular Tbet-APC+, (C) a third for Th17 intracellular Rorγ-PE+ and (D) the last for Treg intracellular Foxp3-APC+. Values are means ± SEM, n= 4-10 and were calculated with a one-way ANOVA and Bonferroni post hoc test. MLN, mesenteric lymph nodes; CD, cluster of differentiation; GFA, short-chain Galacto-oligosaccharides/ long- chain Fructo-oligosaccharides/ pectin-derived Acidic-oligosaccharides; s, standard 2000 IU/kg vitamin A; +, 8000 IU/kg vitamin A.

153

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 153 CHAPTER 7

Supplemental fi gure 2. pDC and CD103+DC population and Foxp3+ CXCR3+ Treg cell population measured in the MLN (A) and spleen (B). MLN (A) was stained for pDC; B220-APC-Cy7+, CD11c-PERCP-Cy5.5+ CD11b-PE-, for migratory DC; CD11c-PERCP-Cy5.5+ CD103-APC+ CX3CR1-FITC- and for Tr1 cells CD4- PERCP-Cy5.5+ CD8a-APC-Cy7-, Foxp3-APC+ CXCR3-FITC+ . Spleens (B) were removed after sacrifi ce of mice and stained for pDC; B220-APC-Cy7+, CD11c-PERCP-Cy5.5+ CD11b-PE-, for migratory DC; CD11c- PERCP-Cy5.5+, CD103-APC+ , CX3CR1-FITC- and for Tr1 cells CD4-PERCP-Cy5.5+ CD8a-APC-Cy7-, Foxp3- APC+ CXCR3-FITC+. Values are means ± SEM, n= 7-10 and were calculated with a one-way ANOVA and Bonferroni post hoc test. MLN, mesenteric lymph nodes; CD, cluster of differentiation; GFA, short-chain Galacto-oligosaccharides/ long-chain Fructo-oligosaccharides/ pectin-derived Acidic-oligosaccharides; s, standard 4000 IU/kg vitamin A; +, 8000 IU/kg vitamin A.

154

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 154 VITA AND GFA IN CMA

Chapter 7 VitA and GFA in CMA Cy7 - APC - FMO without CD8 PE - without Gata3 without FMO withisotypes + -5 Cy5 - PerCP - 7 CD4 Sample

598 599 Supplemental figure 3. Gating strategy for Figure 6A, B. Upon live cell gating, the CD4+ cell population

185 155

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 155 CHAPTER 7

Supplemental figure 3. Gating strategy for Figure 6A, B. Upon live cell gating, the CD4+ cell population was gated using a plot separating CD4+ from CD8α+ cells based on the staining using CD4-PERCP-Cy5.5+, CD8α-APC-Cy7-, Gata3-PE+ and Tbet-APC+. For the Th1 and Th2 cells analysis the CD69 staining gave no positive signal within the sample population, thus CD4+ Tbet/ Gata3+ staining was used. First Fluorescence Minus One (FMO) stained dot plot is shown for CD4-PERCP-Cy5.5+ Tbet-APC+ and Gata3-PE+ without CD8α-APC-Cy7, then only the FMO’s without Tbet-APC+/ Gata3-PE and the FMO Tbet-APC/ Gata3-PE+ is shown. Subsequently, an isotype staining for CD8α-APC-Cy7/ Tbet-APC / Gata3-PE with CD4-PERCP- Cy5.5+. Then a sample staining is shown for the specific Tbet-APC+ and Gata3-PE+ populations within CD4- PERCP-Cy5.5+/ CD8α-APC-Cy7-. Due to difficult separation of Tbet-APC+ and Gata3-PE+ populations in a dot plot, histograms were used.

156

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 156 VITA AND GFA IN CMA

Chapter 7 VitA and GFA in CMA - APC - FMO without CD8 PE - γ FMO without Foxp3-APC without FMO FMO without Ror isotypes with + -5 Cy5 -

PerCP 7 - CD4 Sample

157

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 157 CHAPTER 7

Supplemental figure 4. Gating strategy for Figure 6C, D. After the live cell gating, the CD4+ cell population was gated using a plot separating CD4+ from CD8α+ cells based on the staining using CD4-PERCP-Cy5.5+, CD8α-APC-Cy7-, Rorγ-PE+ (p8) and Foxp3-APC+ (p6)/ IL17-FITC- (p7). First Fluorescence Minus One (FMO) stained dot plot is shown for CD4-PERCP-Cy5.5+, Rorγ-PE+, Foxp3-APC+ and IL17-FITC+ without CD8α- APC-Cy7. Subsequently only the FMO for Foxp3-APC+/ IL17-FITC+ / Rorγ-PE and the FMO Foxp3-APC/ IL17-FITC+ / Rorγ-PE+ is shown. Then an isotype staining for CD8α-APC-Cy7/ Foxp3-APC/ IL17-FITC / Rorγ-PE with CD4-PERCP-Cy5.5+. Last a sample staining is shown for the specific CD4-PERCP-Cy5.5+/ CD8α-APC-Cy7-, Foxp3-APC+ / IL17-FITC+ and Rorγ-PE+. Due to difficult separation of IL17-FITC+ and Rorγ- PE+ populations in a dot plot, an histogram was used for Rorγ-PE+.

158

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 158 VITA AND GFA IN CMA

Chapter 7 VitA and GFA in CMA - Cy7 - FMO without CD4 APC - CD11c Cy5.5Cy7 - FMO without PerCP APC Cy7 γ - PE FMO IFNγ-APC without IFN FMO without IL4-PE-Cy7

+ CD11c-PerCP-Cy5-5 with isotypes CD4-APCCy7+ with isotypes

Sample - CD11c + CD4 - CD4 + CD11c

620 621 Supplemental Figure 5. Gating strategy for Figure 7 is shown in representative flow cytometric plots. 622 Staining for DC CD11c-PERCP-Cy5.5+ and for T cells CD4-APC-Cy7+ and cytokines IFNγ-APC+ and 623 IL4-PE-Cy7- gate. T cell population was separated from DC in a CD4-APC-Cy7+/CD11c-PERCP-

159 189

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 159 CHAPTER 7

Supplemental Figure 5. Gating strategy for Figure 7 is shown in representative flow cytometric plots. Staining for DC CD11c-PERCP-Cy5.5+ and for T cells CD4-APC-Cy7+ and cytokines IFNγ-APC+ and IL4-PE- Cy7- gate. T cell population was separated from DC in a CD4-APC-Cy7+/CD11c-PERCP-Cy5.5- dot plot. First the Fluoride Minus One (FMO) without CD11c-PERCP-Cy5.5 and with CD4-APC-Cy7+/ IFNγ-APC+/ IL4-PE- Cy7+. Secondly the FMO without CD4- APC-Cy7 and with CD11c-PERCP-Cy5.5 +/ IFNγ-APC+/ IL4-PE-Cy7+. Third the FMO for IL4 with IFNγ-APC+/ IL4-PE-Cy7 and last the FMO for IFNγ with IFNγ-APC/ IL4-PE-Cy7+ are shown. The CD4-APC-Cy7+ cells with isotypes for all other staining and the CD11c-PERCP-Cy5.5+ with all isotypes for all other staining are shown. Due to unclear CD11-PERCP-Cy5.5+ population only the CD4- APC-Cy7+/ CD11c-PERCP-Cy5.5- population was selected in the combinatory staining, and for DC selection a histogram was used. This DC population selection was verified not to be CD4-APC-Cy7+ (data not shown). Then a sample population is shown for CD4-APC-Cy7+/ CD11c-PERCP-Cy5.5-/ IFNγ-APC+ or IL4-PE-Cy7+ and for CD11c-PERCP-Cy5.5+/ IFNγ-APC+ or IL4-PE-Cy7+.

160

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 160 VITA AND GFA IN CMA

161

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 161 8 530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 162 General 8 discussion 530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 163 CHAPTER 8

164

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 164 GENERAL DISCUSSION

GENERAL DISCUSSION

Food allergy is becoming more prevalent in the western world and the developing world (1). Cow’s milk allergy (CMA) is one of the first food allergies to develop in life and all children, even those who outgrow CMA (2), are predisposed to develop other allergies later in life (3). So it is important to develop a treatment but also preventive strategies for CMA (4).

Exposure to microbes and dietary components after birth is an important period for immune development and maturation. Under healthy conditions the mucosal immune system renders a non-responsive state, called oral tolerance, against i.e. harmless food proteins. Breast feeding supports the maturation of the infants immune system and oral tolerance induction. One of the many beneficiary components in human milk that contribute to the development of adequate immune function and the microbiome are non-digestible oligosaccharides (NDO). NDO function in several manners. They can support the composition and function of the microbiome, since specific bacteria can selectively ferment NDO in the colon, and the microbiota is known to be involved in immune maturation (5, 6). In addition, NDO can directly influence on the function of intestinal epithelial cells which can regulate immune function or alternatively they can become available in the bloodstream and directly interact with immune cells (7, 8). The NDO content in cow’s milk, a prominent source for infant nutrition, is very low therefore oligosaccharides mimicking some features and functions of human milk NDO are used to complement formula milk. These NDO are isolated from other food sources or enzymatically produced. Short- chain lactose derived Galacto-oligosaccharides (scGOS) (9), long-chain plant derived Fructo- oligosaccharides (lcFOS) (10) and pectin-derived Acidic oligosaccharide (pAOS) (11) are some of these NDO. Combinations of these oligosaccharides, abbreviated as GF or GFA, were used both in clinical and preclinical studies to assess their immunomodulatory properties and underlying mechanisms are mentioned hereafter. In a murine model GFA could enhance the vaccination response via upregulation of the T helper (Th)1 cell population (12-14) and in a murine model for allergic asthma both GF and GFA could decrease the manifestation of experimental allergic asthma (13). Furthermore, both GF and GFA could reduce clinical symptoms in cow’s milk allergic mice (15, 16). It was shown that when isolated T cells from GFA fed cow’s milk allergic mice were transferred to naïve recipient mice, the latter were protected from the development 8 of CMA (16). In human studies using infants with high or low risk for atopy, GF and GFA were also effective in the prevention of atopic dermatitis (17-19). When GF was fed the first six months to healthy term infants at risk for atopy, this could reduce cumulative allergic manifestations and atopic dermatitis (17). This reduction was even present five years after the intervention with GF (17). GFA could reduce atopic dermatitis in low atopy risk infants up to the first birthday (19). GFA could not prevent eczema in combination with partly hydrolysed whey in high risk infants although cow’s milk specific immunoglobulin (Ig) G1 was reduced (20). However, the combination of GFA with partially hydrolysed protein formula fed to infants could induce a

165

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 165 CHAPTER 8

gut microbiota more similar to breast fed infants (21). GFA results show immunomodulatory properties through the whole intestinal mucosa in CMA mice, more in depth research can help understand this mechanism.

In this thesis the preventive effect of NDO on CMA development is further investigated. The preventive effect of the single components, scGOS, lcFOS and pAOS were compared to two combinations of these NDO, GF and GFA in a mouse model for CMA (Chapter 3). Previously GFA was found to positively influence the regulatory T cells (Treg) which resulted in the suppression of CMA in mice (16). We investigated whether the soluble factors interleukin (IL)10 and transforming growth factor (TGF) β contributed to this protective effect via in vivo exposure to IL10 receptor and TGFβ neutralizing antibodies (Chapter 4). Both cytokines are known to play a role in Treg development (22). A murine vaccination model was used to further investigate the immunomodulatory capacity of GFA. GFA was added to the diet to assess if GFA supports the vaccination response and which specific cell types are influenced by GFA (Chapter 5).

Other dietary components, like vitamins are known to influence the immune system as well (23, 24). Vitamins A (VitA) and D (VitD) interact with regulatory components of the immune system like TGFβ and IL10 (24-26). We assessed the possible additional influence of VitD and VitA on the preventive effect of GFA in the CMA model (Chapter 6 and 7). All in vivo studies were carried out to gain a more complete understanding of the supporting role that GFA can have on immune regulation. This chapter contains a synopsis of our main results, discussed in light of the current knowledge.

Single components scGOS, lcFOS, pAOS and combinations thereof Previous studies have shown that both GF and GFA were effective in preventing CMA symptoms in a female C3H/HeOuJ mouse model for CMA, while the single components showed less pronounced effects (15, 16). However, in other animals models, the single components lcFOS and scGOS, present in different amounts as used in the studies described in the current thesis, could affect egg allergy (27) or asthma (28) respectively and an addition of five percent of pAOS to a normal diet could improve the outcome of a lung infection in a mouse model (29, 30). In our study using the CMA mouse model, GF and GFA could reduce the allergic skin response and whey specific IgG1, which is in line with previous data published by Schoutenet. al (15, 16, 31, 32). Other percentages or combinations can also be effective to reduce allergic symptoms in mouse models for peanut (33), wheat (34), egg (35, 36) or asthma (37) for example.

There are approximately 150 different types of NDO in human milk (38). Furthermore, there is difference in human milk composition between individual mothers and difference according to the infants’ age. NDOs like scGOS, lcFOS and pAOS modulate the microbiota composition and activity in the intestine (29, 39), but they may also directly affect cells of the intestinal mucosa or

166

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 166 GENERAL DISCUSSION

even in small amounts enter the bloodstream (40). In vitro evidence shows that scGOS, lcFOS and pAOS can be transported through an intestinal Caco-2 cell layer (41). In the small intestine GOS could upregulate mucin expression, these proteins are secreted by epithelial cells in the small intestine to improve barrier function (40). In vivo human NDO have been found in blood and urine which suggests that these structures can pass through the intestinal epithelial cell layer intact (40). The amount and composition of GF and GFA that were used in the pre-clinical studies, mimic some aspects of the natural NDO composition in human milk. Our data are in line with the current scientific knowledge of NDO and show the importance of combining NDO in priming the immune system towards a less allergic setting in the CMA model.

GF versus GFA in modulation of immune markers in the small intestine NDO pass the small intestine and are known to serve as fermentable material for the commensal bacteria present throughout the intestine. In the colon NDO have probably the most influence, since the largest abundancy of bacteria is in the colon and short chain fatty acids like butyrate are produced upon fermentation (7). We evaluated the association between the preventive effect of GF and GFA in CMA and the effect on several immune markers in the small intestine. GF could downregulate Gata3 (Th2) and RAR related orphan receptor gamma (Rorγ) (Th17) mRNA and GFA upregulated T box transcription factor (Tbet) (Th1) and IL10 (Treg) mRNA in the small intestine. Both GF and GFA could increase Forkhead Box P3 (Foxp3) at mRNA and protein levels in the small intestine. Beyond the regulation of Foxp3, GF and GFA also were able to modulate specific immune pathways in association with suppression of clinical symptoms in CMA.

Human milk derived NDO have immunomodulatory properties since they were found to enhance interferon gamma (IFNγ) and IL10 levels in cord blood, although this did not occur when the combination of GF or single pAOS was used (41). Schouten et al. already showed the importance of isolated splenic cluster of differentiation (CD)25+ T cells, which include Treg, in the protection for CMA (16). Recently it was also shown that macromolecules in the diet induce Foxp3+ Treg in the small intestine (42). GF were shown to enhance IL10 in monocyte derived dendritic cells (DC) isolated from human blood (8). Subsequently these GF treated DC could in turn increase CD4+Foxp3+ T cells (8). Also GF or scFOS/lcFOS could enhance IFNγ 8 and IL10 in a ttrans well co culture model with a HT29 cell line and PBMC from both healthy individuals or peanut allergic patients (43). This shows that dietary components, like NDO, may also have direct effects in the small intestine or colon of the mice alike what is shown in vitro using human cells. These data also support our findings of GFA as well as GF as regulatory components modulating immune markers in the small intestine. However, since the small intestine also contains bacteria, these bacteria and/ or their fermentation products may also have contributed to the effect on mucosal immune markers.

167

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 167 CHAPTER 8

The knowledge about the beneficial functions of NDO is improving but not a lot research has been done for the combination with pAOS. Acidic type oligosaccharide pAOS, may have an added immunomodulatory effect compared to using only both neutral oligosaccharides. The differences in mucosal immune modulation between GF, which downregulates allergic T cell markers and GFA that upregulates Th1 and regulatory markers, should be further explored to gain a better understanding for further use of either of these combinations in the prevention of CMA.

Contribution of IL10r and TGFβ in the CMA protective effect of GFA GFA significantly lowered the allergic reaction to cow’s milk, while neutralizingα -TGFβ or α-IL10r abolished the preventive effect of GFA on CMA symptoms. Mice fed GFA and treated with α-IL10r or α-TGFβ antibody also showed significantly more mouse mast cell protease (mMCP)1.

Both TGFβ and IL10 are involved in the immune regulatory pathway, TGFβ is necessary to establish Treg, and IL10 sustains the Treg population after initiation (44). Both cytokines are also produced by Tregs in order to regulate the DC, T and B cell population by dampening their activity (44). In addition, the function of effector cells such as mast cells is modulated by TGFβ and IL10. That both antibodies could independently abrogate the preventive effect of GFA on CMA symptoms was clearly shown (Chapter 3). We already found GFA to enhance mucosal Tgfβ and Il10 at the mRNA level using qPCR. Only, strands of mRNA need to be translated into a protein and subsequently proteins need to be folded to become functional. TGFβ and IL10r neutralization indeed confirmed that GFA induces protective immunomodulatory properties which are mediated via IL10 or TGFβ in vivo. It would be interesting to determine whether this preventive effect of GFA and its protective mechanism which is mediated by IL10 or TGFβ is induced via direct epithelial cell and/or immune cell stimulation or indirect via modulation of the intestinal microbiome, to determine this these studies could be performed in germ free mice.

In an allergic reaction mast cells respond to the release of IgE, which stimulates degranulation of mast cells. In these granules proteins are stored to induce an allergic reaction, e.g. histamine and Tumor Necrosis Factor (TNF)α, chemoattractant and costimulatory molecules to enhance the Th2 cell response (45). IL10 is produced by DC to suppress allergic responses (46) and this IL10 also downregulates IgE production by B cells (47). When IL10 and TGFβ are present in a B cell rich environment, these B cells produce IgA and not IgE. IL10 can induce mast cell proliferation and differentiation but IL10 can also interfere with mast cell activation on multiple levels. It can inhibit signal transducer and activator of transcription factor (Stat)3 expression and FcεRI expression (45). Subsequently IL10 can alter signalling molecules which are otherwise activated after IgE binding to mast cells (48) and IL10 can make mast cells more

168

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 168 GENERAL DISCUSSION

susceptible to FAS induced apoptosis (49). The inhibition of mast cells at several levels also occurs when TGFβ is present, but it is dependent on the microenvironment (49). To establish whether GFA interferes directly with mast cell activation, peritoneal mast cells from mice fed a control or GFA diet with or without IL10r or TGFβ antibodies could be isolated and activation upon whey stimulation measured. With this experiment it may be possible to further establish the influence of TGFβ or IL10 induced by GFA in CMA mice on mast cells.

Intestinal Treg cells are essential to acquire oral tolerance to food (44, 50). Schouten et al. found the CD25+CD4+ T cells to be of importance for the establishment of tolerance in CMA (16). Treg are formed from naïve CD4+ T cells in the mesenteric lymph nodes (MLN), this instruction is done by tolerogenic DC. This process in the MLN is dependent on cell contact and several soluble factors, one of them being TGFβ and the other retinoic acid (RA) (44). IL10 is not essential to form CD103+ DC to instruct Treg in the MLN, which could explain the elevated CD103+ DC population found in the IL10r antibody treated group. IL10 is still necessary for the maintenance and expansion of the Treg population in the intestinal lamina propria (LP). Still we found significant more Foxp3+CD4+ T cells in the intestinal LP when mice were treated with IL10r antibody. Possibly these Treg have lost their functional capacity to suppress effector cells. Although we did not use an assay to test functionality of these cells, others already showed that induced Treg from IL10-/- knockout mice could not suppress DC or effector T cells (51). It is probable that our Treg also lost their functional capacities. To be sure this occurs these Foxp3+CD4+ Treg should be analysed to show if these cells still function.

Also a role for the regulatory T cell type 1 (Tr1) in allergy prevention is suggested (50). These Tr1 are mainly effective via the secretion of IL10. The Tr1 cell population was investigated in healthy and in peanut allergic humans (52). Both groups produced Tr1 cells, but the Tr1 from peanut allergic patients was functionally defective in suppressing Th2 cells in vitro (52). These results underline the importance of IL10 in allergy suppression.

There are multiple processes in which the antibodies against IL-10r and TGFβ could have intervened with the beneficiary effects of GFA in the CMA model, but that IL10 and TGFβ are involved in the prevention of allergic symptoms mediated by the GFA diet has become clear 8 from these studies.

GFA and vaccination In a mouse model for influenza vaccination the immunomodulatory support of dietary GFA was assessed. We have shown an improved vaccination response in association with an increase in the Th1 cell frequency and a decrease in the Th2 cell and CD4+Foxp3+CXCR3+Tbet+ Th1-Treg cell population. This is in line with previous results showing that GFA is capable of supporting the immunological response in an influenza vaccination model (14). GFA reduced the Th1-Treg

169

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 169 CHAPTER 8

population in the MLN, however the splenic CD4+CD25+ T cells from GFA treated vaccinated mice could induce IFNγ and reduce IL17 ex vivo in a suppression assay. Also in a piglet study NDO (scFOS) could induce a better vaccine response measured as increased sIgA vaccine response and IFNγ secretion (53), when at least present in the diet of the piglet mothers, supporting our data that NDO like GFA can enhance IFNγ induction. Beyond improving the vaccination response, in a suckling rat pup model GFA was able to protect these rats from a Rota virus infection (54). GFA could ameliorate clinical symptoms and modulate the immune response contributing to a faster resolution of the Rota virus (54). In neonatal mice, the influence of the CD25+ cell or Treg population on influenza clearance was investigated (55). When the CD25+ and/or Treg population was either inhibited via anti-CD25 antibodies or when scurfy neonatal mice (mice lacking Foxp3+ Treg cells) were used, the neonatal mice had significantly higher viral loads than controls indicating that CD25+ positive cells and/ or Treg are essential for viral clearance (55). Another study used different vaccines in humans and in a mouse model to evaluate the Treg population (56). Influenza vaccinations of healthy humans resulted in minor changes in both naïve (CD45RA+) and activated (CD45RA-) Treg populations in blood (56). Spleen and draining lymph nodes from mice were assessed after vaccination and three days after vaccination the percentage of Treg was reduced in the draining lymph nodes while seven days after vaccination an increase of Treg was measured in the spleen (56). We did not find any difference in splenic Treg cell population seven days after vaccination. De Wolf et al. (56) only investigated Treg for CD4+Foxp3+CD25+ markers but they did not look at expression markers for Tbet or CXCR3. Typically in our studies, the increase in vaccination response was associated with effect on Treg, but a decrease in the Th1-Treg and increase in Th1 population which could explain the observed differences.

Recently it was shown that Th1-Treg expressing Tbet are highly stable and are essential for inhibition of Th1 (57). However, Tbet+ Th1-Treg can also contribute to immune tolerance, since when missing in male or female type 1 diabetic mice the disease severity increased rapidly, underlining their importance in immune suppression (58). These Th1-Treg can also be found in intestinal Peyer’s Patches (PP) (59). Since we show that GFA can decrease these Th1-Treg in the C3H/HeOuJ influenza mouse model in association with a better Th1 type vaccination response and reduced Th2 response, it would be of interest to study these Th1-Treg cells in the cow’s milk allergic model as well.

Vitamin D VitD deficiency in certain regions of the world has been indicated to have a possible connection with increase in allergy in the same regions (60). Already established is the influence of VitD in certain DC subsets to induce tolerance (61). In the CMA model it was studied if either VitD depletion or supplementation would affect allergy development compared to the standard dose of VitD in the diets. However, VitD depletion tended to enhance the risk of severe

170

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 170 GENERAL DISCUSSION

allergic symptoms and extra VitD supplementation did not reduce allergy development. Only supplementation with VitD combined with GFA added to the diet in a preventive setting reduced allergic symptoms compared to the group supplemented with VitD alone. However, this was similar to the effect of GFA which already tended to lower allergic symptoms. Despite showing no clinical benefits, the VitD supplementation combined with GFA did induce significant more CD11c+CD103+ DC compared to allergic mice fed the VitD supplement only.

The VitD deprived group showed a severe reaction to the cow’s milk protein challenge and consequently more mice needed to be euthanized. This is an indication that a critical VitD level is required for keeping the immune system in control. Severe mast cell and/or basophil degranulation could lead to increased allergic symptoms. As a reflection of mucosal mast cell degranulation mMCP1 is measured in the cow’s milk allergic mice. However, in this experiment serum parameter mMCP1 could be measured but the levels remained low and no significant increase was observed in cow’s milk allergic mice (data not shown). In an Ovalbumin (OVA) induced allergy model in C3H/HeJ and BALB/c mice, several mMCP’s, to evaluate mast cell subset activity during allergy, and immunoglobulins were measured in serum of blood (62). mMCP1, -2, -4, -5, -6, -7 and 8 could not be detected with qPCR in the intestine of OVA allergic C3H/HeJ mice (62). In our C3H/HeOuJ CMA mouse model mMCP1 can be detected in the colon and also after degranulation in serum (63).For the OVA allergic C3H/HeJ mice serum mMCP1 could be significantly increased via oral challenge, serum mMCP7 showed a significant increase only in the symptomatic group (62). Plasma histamine and serum mMCP7 showed a correlation with anaphylactic symptoms measured as a drop in body temperature and diarrhoea (62). Correlations between mMCP7 and mMCP1, mMCP7 and OVA specific IgE were found, there was no correlation between mMCP1 and OVA specific IgE (62). mMCP2 showed promising results using qPCR in the small intestine, but no commercial antibodies are available yet (62). mMCP2 and -7 could be new factors to analyse the allergic outcome in serum of C3H/HeOuJ mice.

In 2017 a cross sectional study in a small cohort was done to asses if cow’s milk allergic children, two years or younger, were at risk for VitD deficiency (64). A higher frequency of inadequate levels of VitD was observed in Brazilian children suffering from immediate and late 8 onset CMA (64), however a number of limitations were addressed and suggested this result should be verified in a larger cohort. Another study in nine month old children showed, one to four days after the onset of CMA manifestation, an increase in IL4+ T cells which correlated with decreased serum levels of VitD (65). A third study aimed to assess the VitD levels in Finnish school children and found 16% VitD insufficiency in the general population (66). This insufficiency level was significantly lower in children with CMA compared to children without CMA (66). One limitation to these three studies is that it is not known what the VitD levels were at birth. Another limitation to VitD research is the difference in what the chosen cut-off levels

171

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 171 CHAPTER 8

are between VitD deficiency and normal VitD levels. It remains to be questioned whether the VitD deficiency was a reason for CMA to develop in these children or if the CMA can cause VitD deficiency. Molly et al. found that the VitD status at birth was not related to egg allergy at birth nor one year later (67). On the contrary a correlation was found between egg allergic children who develop natural tolerance to egg later in life and increasing VitD serum levels (68). Some studies find a possible correlation between allergy and VitD, but there is very little direct evidence in humans for a link between VitD supplementation and the risk to develop (food) allergy (69-71).

In a mouse model for OVA induced allergy VitD deficiency significantly enhances the allergic reactions (72, 73). The mice in the study from Matsui et al. (73) were kept on a VitD deficient diet for 25 days before treatment and the other study by Wu et al. (72) deprived mice of VitD during gestation and used their offspring. Their offspring was also on a deprived diet, to study the effect of VitD deficiency. Hence, although our data indicate more severe symptoms in CMA when mice were fed a VitD deficient diet, this could not be proven significantly. However, prolonged VitD deprivation may be needed to show the relevance of sufficient levels of VitD for CMA prevention.

Although also no protective of VitD supplementation was found in CMA and it did not improve the protective effect of GFA in CMA, combined supplementation if VitD and GFA did affect the DC population. The VitD receptor is found in many immune cells, such as neutrophils, B cells, CD4 and CD8 T cells, macrophages, mast cells and DC. VitD can suppress the cytokine production, the expression of MHCII and co-stimulatory molecules in DC in general (74, 75). It was shown in mice that VitD inhibits the IL12, IFNγ and TNFα production by DC in vitro, these cytokines are important for the induction of a Th1 response (75). Furthermore, human Langerhans cells, exposed to VitD, suppressed Th2 type chemokines (75). Bakdash et al. found that VitD physiological precursor cholecalciferol weakly primed DC to induce Treg (76). VitD precursors calcitriol and calcidiol both had similar effects of DC priming towards immunosuppression and subsequently on the positive correlation of IL10 producing Treg (76). Additionally, the calcidiol DC primed Treg exhibited sustained IFNγ production (76). Recently it was also shown that VitD is required for the maintenance of mast cell stability (77). The activated VDR inhibited the intrinsic signaling cascade for greater mast cell activation (77). Calcitriol can even block mast cell degranulation (77).

VitD is involved in many different immune processes, the influence of VitD will probably be dose dependent and it may act different in human versus mice (78). Although VitD supplementation did not have a beneficial effect either or not combined with GFA in the prevention of CMA in mice, it would still be interesting to study if VitD and GFA could affect CMA in mice with longstanding low levels of VitD.

172

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 172 GENERAL DISCUSSION

Vitamin A VitA can also modulate the (intestinal) immune system in its active form, RA. Intestinal CD103+ DC are the only intestinal DC that can generate RA via aldehyde dehydrogenase (ALDH). RA is very important in the generation of Treg, the first step is to induce tolerogenic CD103+ DC, secondly in creating a tolerogenic micro-environment in the MLN so CD103+ DC will drive the T cell to a regulatory phenotype (79). This Treg response can downregulate the Th response in the intestinal LP. Thus we investigated if supplementation of VitA could support the preventive effect of GFA on CMA. Only when VitA was combined with GFA (VitA+/GFA) a reduction of the allergic response was measured. Both components singularly could not reduce CMA symptoms in this experiment. Expression of Il4, Il5, Ifnγ and Foxp3 mRNA in the small intestine from cow’s milk allergic mice fed VitA+/GFA was affected and a difference in expression of splenic IL4 and IFNγ cytokines was detected.

VitA is an essential vitamin during growth and development, total depletion of this vitamin is not feasible because RA is an essential signalling molecule during embryogenesis. VitA is stored in fat and in the liver, thus it will take a long time to deplete. Our mice were depleted for VitA for two weeks after weening before mice became allergic, but the exact VitA depletion status is not certain since we did not measure RA serum levels. Neonatal mice deficient in VitA had impairment of tolerance induction to OVA (80). When VitA deprived dams were supplied with extra VitA, their one week old weening new-born mice showed oral tolerance levels the same as three week old mice when challenged with OVA (80). VitA can also improve intestinal functions via upregulation of tight junction proteins, restore damaged villi and reduce diarrhoea severity for example (81). There has to be some precaution with VitA supplementation, because VitA can accumulate to toxic levels (82). To generate a more robust VitA depleted model, offspring of VitA deprived dams could be used to assess the influence of VitA supplementation on the development of CMA. Subsequently GFA can be added to the diet of the VitA deprived offspring to explore the immunomodulatory role of GFA in CMA development in absence of VitA compared to supplementation with VitA alone or the combination of VitA and GFA.

The preventive effect of GFA on CMA was not shown in this experiment. It is already shown that results for CMA in the C3H/HeOuJ mouse model can differ, depending on the amount of 8 whey used, the specific whey source (unpublished data), the laboratory (83), and in housing conditions (83-85). When CMA clinical symptoms, like ear swelling or shock, are established and prevented via GF, GFA or partly hydrolysed whey, mMCP1 and whey specific -IgE -IgG1, -IgG2a levels do not always correspond with the outcome of the clinical symptoms (15, 31, 86). These studies show that also the symptom lowering effect of dietary components like GFA may differ between studies. This indicates that environmental factors may affect the protective capacity of the diet. These effects may for example relate to changes in the microbial

173

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 173 CHAPTER 8

composition or may be related to the availability or dose of other factors, like certain nutrients, and of which VitA may be an example.

In line with previous findings (Chapter 3) increased expression ofIfn γ, Foxp3 and ALDH1a2 mRNA is found in the small intestine, only if allergic mice were fed GFA and VitA. For ALDH expression it is shown that expression differs throughout the intestine and between mouse strains (87). To further determine the influence of GFA and/ or VitA on ALDH, its expression should be evaluated in more intestinal locations not only using qPCR but also via immunohistochemistry. This will lead to a better understanding of the influence of GFA in combination with VitA in the development of CMA.

Overall conclusion Both GF and GFA are effective in reducing the cow’s milk allergic symptoms in a preventive setting, skewing the immune balance in a specific manner. GFA can skew the T cell balance in a CMA model towards Treg to diminish allergic symptoms (16), and TGFβ and Il10 are involved in this protective effect. Although the function of Treg is improved by GFA in the CMA model, a subtype of Th1-Treg cells is downregulated in a murine vaccination model upon dietary intervention with GFA, while the Th1 over Th2 balance is increased. VitD supplementation does not add to the preventive effect of GFA in CMA development. VitA supplementation on the other hand appears to support GFA in its capacity to reduce CMA symptoms. The combination of GFA, mimicking some aspects if the NDO composition in human milk, is effective in reducing CMA symptoms. Further exploration of the effect of GFA with or without VitA and its effect on the mucosal and systemic immune response, will improve our understanding of the regulatory mechanisms that are affected by this dietary intervention and help to reduce CMA symptoms.

174

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 174 GENERAL DISCUSSION

REFERENCES

1. Fiocchi A, Dahdah L, Fierro V, Artesani with a human faecal flora. Br J Nutr. MC, Valluzzi R. Food allergy trends at the 1997;78(2):313-24. crossing among socio-economics, history 11. Fanaro S, Jelinek J, Stahl B, Boehm G, and geography. Curr Opin Allergy Clin Kock R, Vigi V. Acidic oligosaccharides from Immunol. 2018;18(3):271-6. pectin hydrolysate as new component for 2. Fiocchi A, Schunemann HJ, Brozek J, Restani infant formulae: effect on intestinal flora, P, Beyer K, Troncone R, et al. Diagnosis and stool characteristics, and pH. J Pediatr Rationale for Action Against Cow’s Milk Gastroenterol Nutr. 2005;41(2):186-90. Allergy (DRACMA): a summary report. J 12. Vos AP, Haarman M, Buco A, Govers M, Allergy Clin Immunol. 2010;126(6):1119-28 Knol J, Garssen J, et al. A specific prebiotic e12. oligosaccharide mixture stimulates delayed- 3. Brandtzaeg P. Food allergy: separating type hypersensitivity in a murine influenza the science from the mythology. Nat Rev vaccination model. Int Immunopharmacol. Gastroenterol Hepatol. 2010;7(7):380-400. 2006;6(8):1277-86. 4. Dahdah L, Pecora V, Riccardi C, Fierro V, 13. Vos AP, van Esch BC, Stahl B, M’Rabet Valluzzi R, Mennini M. How to predict and L, Folkerts G, Nijkamp FP, et al. improve prognosis of food allergy. Curr Opin Dietary supplementation with specific Allergy Clin Immunol. 2018;18(3):228-33. oligosaccharide mixtures decreases 5. Candy DCA, Van Ampting MTJ, Oude Nijhuis parameters of allergic asthma in mice. Int MM, Wopereis H, Butt AM, Peroni DG, et al. Immunopharmacol. 2007;7(12):1582-7. A synbiotic-containing amino-acid-based 14. Vos AP, Haarman M, van Ginkel JW, formula improves gut microbiota in non- Knol J, Garssen J, Stahl B, et al. Dietary IgE-mediated allergic infants. Pediatr Res. supplementation of neutral and acidic 2018;83(3):677-86. oligosaccharides enhances Th1-dependent 6. Veereman-Wauters G, Staelens S, Van de vaccination responses in mice. Pediatr Broek H, Plaskie K, Wesling F, Roger LC, et Allergy Immunol. 2007;18(4):304-12. al. Physiological and bifidogenic effects of 15. Schouten B, van Esch BC, Hofman GA, van prebiotic supplements in infant formulae. J Doorn SA, Knol J, Nauta AJ, et al. Cow milk Pediatr Gastroenterol Nutr. 2011;52(6):763- allergy symptoms are reduced in mice fed 71. dietary synbiotics during oral sensitization 7. Akkerman R, Faas MM, de Vos P. Non- with whey. J Nutr. 2009;139(7):1398-403. digestible carbohydrates in infant formula as 16. Schouten B, van Esch BC, Hofman GA, substitution for human milk oligosaccharide Boon L, Knippels LM, Willemsen LE, et al. functions: Effects on microbiota and gut Oligosaccharide-induced whey-specific maturation. Crit Rev Food Sci Nutr. 2018:1- CD25(+) regulatory T-cells are involved in 12. the suppression of cow milk allergy in mice. 8. Lehmann S, Hiller J, van Bergenhenegouwen J Nutr. 2010;140(4):835-41. J, Knippels LM, Garssen J, Traidl- 17. Arslanoglu S, Moro GE, Boehm G, Wienz Hoffmann C. In Vitro Evidence for Immune- F, Stahl B, Bertino E. Early neutral 8 Modulatory Properties of Non-Digestible prebiotic oligosaccharide supplementation Oligosaccharides: Direct Effect on Human reduces the incidence of some allergic Monocyte Derived Dendritic Cells. PLoS manifestations in the first 5 years of life. One. 2015;10(7):e0132304. J Biol Regul Homeost Agents. 2012;26(3 9. Mitsuoka T, Hidaka H, Eida T. Effect of fructo- Suppl):49-59. oligosaccharides on intestinal microflora. 18. Arslanoglu S, Moro GE, Schmitt J, Tandoi Nahrung. 1987;31(5-6):427-36. L, Rizzardi S, Boehm G. Early dietary 10. Djouzi Z, Andrieux C. Compared effects intervention with a mixture of prebiotic of three oligosaccharides on metabolism oligosaccharides reduces the incidence of intestinal microflora in rats inoculated of allergic manifestations and infections

175

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 175 CHAPTER 8

during the first two years of life. J Nutr. 28. Verheijden KA, Willemsen LE, Braber S, 2008;138(6):1091-5. Leusink-Muis T, Delsing DJ, Garssen J, 19. Gruber C, van Stuijvenberg M, Mosca F, Moro et al. Dietary galacto-oligosaccharides G, Chirico G, Braegger CP, et al. Reduced prevent airway eosinophilia and occurrence of early atopic dermatitis hyperresponsiveness in a murine house because of immunoactive prebiotics among dust mite-induced asthma model. Respir low-atopy-risk infants. J Allergy Clin Res. 2015;16:17. Immunol. 2010;126(4):791-7. 29. Bernard H, Desseyn JL, Bartke N, Kleinjans 20. Boyle RJ, Tang ML, Chiang WC, Chua L, Stahl B, Belzer C, et al. Dietary pectin- MC, Ismail I, Nauta A, et al. Prebiotic- derived acidic oligosaccharides improve supplemented partially hydrolysed cow’s the pulmonary bacterial clearance of milk formula for the prevention of eczema Pseudomonas aeruginosa lung infection in in high-risk infants: a randomized controlled mice by modulating intestinal microbiota trial. Allergy. 2016;71(5):701-10. and immunity. J Infect Dis. 2015;211(1):156- 65. 21. Wopereis H, Sim K, Shaw A, Warner JO, Knol J, Kroll JS. Intestinal microbiota in infants at 30. Bernard H, Desseyn JL, Gottrand F, Stahl high risk for allergy: Effects of prebiotics and B, Bartke N, Husson MO. Pectin-Derived role in eczema development. J Allergy Clin Acidic Oligosaccharides Improve the Immunol. 2018;141(4):1334-42 e5. Outcome of Pseudomonas aeruginosa Lung Infection in C57BL/6 Mice. PLoS One. 22. Wambre E, Jeong D. Oral Tolerance 2015;10(11):e0139686. Development and Maintenance. Immunol Allergy Clin North Am. 2018;38(1):27-37. 31. Schouten B, van Esch BC, Hofman GA, de Kivit S, Boon L, Knippels LM, et al. A 23. Maggini S, Pierre A, Calder PC. Immune potential role for CD25+ regulatory T-cells Function and Micronutrient Requirements in the protection against casein allergy by Change over the Life Course. Nutrients. dietary non-digestible carbohydrates. Br J 2018;10(10). Nutr. 2012;107(1):96-105. 24. Hosomi K, Kunisawa J. The Specific 32. Schouten B, Van Esch BC, Kormelink Roles of Vitamins in the Regulation of TG, Moro GE, Arslanoglu S, Boehm G, Immunosurveillance and Maintenance et al. Non-digestible oligosaccharides of Immunologic Homeostasis in the Gut. reduce immunoglobulin free light-chain Immune Netw. 2017;17(1):13-9. concentrations in infants at risk for allergy. 25. van der Aar AM, Sibiryak DS, Bakdash G, van Pediatr Allergy Immunol. 2011;22(5):537-42. Capel TM, van der Kleij HP, Opstelten DJ, et 33. Wagenaar L, Bol-Schoenmakers M, al. Vitamin D3 targets epidermal and dermal Giustarini G, Vonk MM, van Esch B, Knippels dendritic cells for induction of distinct LMJ, et al. Dietary Supplementation with regulatory T cells. J Allergy Clin Immunol. Nondigestible Oligosaccharides Reduces 2011;127(6):1532-40 e7. Allergic Symptoms and Supports Low 26. Coombes JL, Siddiqui KR, Arancibia- Dose Oral Immunotherapy in a Peanut Carcamo CV, Hall J, Sun CM, Belkaid Y, et Allergy Mouse Model. Mol Nutr Food Res. al. A functionally specialized population 2018;62(20):e1800369. of mucosal CD103+ DCs induces Foxp3+ 34. Bouchaud G, Castan L, Chesne J, Braza regulatory T cells via a TGF-beta and retinoic F, Aubert P, Neunlist M, et al. Maternal acid-dependent mechanism. J Exp Med. exposure to GOS/inulin mixture prevents 2007;204(8):1757-64. food allergies and promotes tolerance in 27. Tsuda M, Arakawa H, Ishii N, Ubukata C, offspring in mice. Allergy. 2016;71(1):68-76. Michimori M, Noda M, et al. Dietary Fructo- 35. de Kivit S, Kostadinova AI, Kerperien J, Oligosaccharides Attenuate Early Activation Morgan ME, Muruzabal VA, Hofman GA, et of CD4+ T Cells Which Produce both Th1 and al. Dietary, nondigestible oligosaccharides Th2 Cytokines in the Intestinal Lymphoid and Bifidobacterium breve M-16V suppress Tissues of a Murine Food Allergy Model. Int allergic inflammation in intestine via Arch Allergy Immunol. 2017;174(3-4):121- targeting dendritic cell maturation. J Leukoc 32. Biol. 2017;102(1):105-15.

176

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 176 GENERAL DISCUSSION

36. Hogenkamp A, Knippels LM, Garssen J, 46. Schulke S. Induction of Interleukin-10 van Esch BC. Supplementation of Mice with Producing Dendritic Cells As a Tool to Specific Nondigestible Oligosaccharides Suppress Allergen-Specific T Helper 2 during Pregnancy or Lactation Leads to Responses. Front Immunol. 2018;9:455. Diminished Sensitization and Allergy in the 47. Akdis CA, Akdis M. Mechanisms of immune Female Offspring. J Nutr. 2015;145(5):996- tolerance to allergens: role of IL-10 and 1002. Tregs. J Clin Invest. 2014;124(11):4678-80. 37. Hogenkamp A, Thijssen S, van Vlies N, 48. Kennedy Norton S, Barnstein B, Brenzovich Garssen J. Supplementing pregnant mice J, Bailey DP, Kashyap M, Speiran K, et al. with a specific mixture of nondigestible IL-10 suppresses mast cell IgE receptor oligosaccharides reduces symptoms of expression and signaling in vitro and in vivo. allergic asthma in male offspring. J Nutr. J Immunol. 2008;180(5):2848-54. 2015;145(3):640-6. 49. Caslin HL, Kiwanuka KN, Haque TT, 38. Thurl S, Munzert M, Boehm G, Matthews Taruselli MT, MacKnight HP, Paranjape C, Stahl B. Systematic review of the A, et al. Controlling Mast Cell Activation concentrations of oligosaccharides in and Homeostasis: Work Influenced by Bill human milk. Nutr Rev. 2017;75(11):920-33. Paul That Continues Today. Front Immunol. 39. Chung WS, Walker AW, Louis P, Parkhill J, 2018;9:868. Vermeiren J, Bosscher D, et al. Modulation 50. Palomares O, Martin-Fontecha M, Lauener of the human gut microbiota by dietary R, Traidl-Hoffmann C, Cavkaytar O, Akdis fibres occurs at the species level. BMC Biol. M, et al. Regulatory T cells and immune 2016;14:3. regulation of allergic diseases: roles 40. Goehring KC, Kennedy AD, Prieto PA, of IL-10 and TGF-beta. Genes Immun. Buck RH. Direct evidence for the presence 2014;15(8):511-20. of human milk oligosaccharides in the 51. Chattopadhyay G, Shevach EM. Antigen- circulation of breastfed infants. PLoS One. specific induced T regulatory cells impair 2014;9(7):e101692. dendritic cell function via an IL-10/MARCH1- 41. Eiwegger T, Stahl B, Haidl P, Schmitt dependent mechanism. J Immunol. J, Boehm G, Dehlink E, et al. Prebiotic 2013;191(12):5875-84. oligosaccharides: in vitro evidence for 52. Pellerin L, Jenks JA, Chinthrajah S, gastrointestinal epithelial transfer and Dominguez T, Block W, Zhou X, et al. immunomodulatory properties. Pediatr Peanut-specific type 1 regulatory T cells Allergy Immunol. 2010;21(8):1179-88. induced in vitro from allergic subjects 42. Kim KS, Hong SW, Han D, Yi J, Jung J, are functionally impaired. J Allergy Clin Yang BG, et al. Dietary antigens limit Immunol. 2018;141(1):202-13 e8. mucosal immunity by inducing regulatory 53. Le Bourgot C, Le Normand L, Formal M, T cells in the small intestine. Science. Respondek F, Blat S, Apper E, et al. Maternal 2016;351(6275):858-63. short-chain fructo-oligosaccharide 43. Hayen SM, Otten HG, Overbeek SA, Knulst supplementation increases intestinal AC, Garssen J, Willemsen LEM. Exposure cytokine secretion, goblet cell number, of Intestinal Epithelial Cells to Short- and butyrate concentration and Lawsonia Long-Chain Fructo-Oligosaccharides and intracellularis humoral vaccine response in 8 CpG Oligodeoxynucleotides Enhances weaned pigs. Br J Nutr. 2017;117(1):83-92. Peanut-Specific T Helper 1 Polarization. 54. Rigo-Adrover M, Perez-Berezo T, Ramos- Front Immunol. 2018;9:923. Romero S, van Limpt K, Knipping K, Garssen 44. Tordesillas L, Berin MC. Mechanisms of Oral J, et al. A fermented milk concentrate and Tolerance. Clin Rev Allergy Immunol. 2018. a combination of short-chain galacto- 45. Elieh Ali Komi D, Grauwet K. Role of Mast oligosaccharides/long-chain fructo- Cells in Regulation of T Cell Responses in oligosaccharides/pectin-derived acidic Experimental and Clinical Settings. Clin Rev oligosaccharides protect suckling rats Allergy Immunol. 2018;54(3):432-45. from rotavirus gastroenteritis. Br J Nutr. 2017;117(2):209-17.

177

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 177 CHAPTER 8

55. Oliphant S, Lines JL, Hollifield ML, Garvy BA. 66. Rosendahl J, Fogelholm M, Pelkonen A, Regulatory T Cells Are Critical for Clearing Makela MJ, Makitie O, Erkkola M. A History Influenza A Virus in Neonatal Mice. Viral of Cow’s Milk Allergy Is Associated with Immunol. 2015;28(10):580-9. Lower Vitamin D Status in Schoolchildren. 56. de Wolf A, van Aalst S, Ludwig IS, Bodinham Horm Res Paediatr. 2017;88(3-4):244-50. CL, Lewis DJ, van der Zee R, et al. Regulatory 67. Molloy J, Koplin JJ, Allen KJ, Tang MLK, T cell frequencies and phenotypes Collier F, Carlin JB, et al. Vitamin D following anti-viral vaccination. PLoS One. insufficiency in the first 6 months of infancy 2017;12(6):e0179942. and challenge-proven IgE-mediated food 57. Levine AG, Mendoza A, Hemmers S, Moltedo allergy at 1 year of age: a case-cohort study. B, Niec RE, Schizas M, et al. Stability and Allergy. 2017;72(8):1222-31. function of regulatory T cells expressing 68. Neeland MR, Koplin JJ, Dang TD, Dharmage the transcription factor T-bet. Nature. SC, Tang ML, Prescott SL, et al. Early life 2017;546(7658):421-5. innate immune signatures of persistent 58. Tan TG, Mathis D, Benoist C. Singular role food allergy. J Allergy Clin Immunol. for T-BET+CXCR3+ regulatory T cells in 2018;142(3):857-64 e3. protection from autoimmune diabetes. Proc 69. Yepes-Nunez JJ, Brozek JL, Fiocchi A, Natl Acad Sci U S A. 2016;113(49):14103-8. Pawankar R, Cuello-Garcia C, Zhang Y, et 59. Huang CH, Wang CC, Lin YC, Hori M, Jan TR. al. Vitamin D supplementation in primary Oral administration with diosgenin enhances allergy prevention: Systematic review of the induction of intestinal T helper 1-like randomized and non-randomized studies. regulatory T cells in a murine model of food Allergy. 2018;73(1):37-49. allergy. Int Immunopharmacol. 2017;42:59- 70. Hennessy A, Hourihane JO, Malvisi L, Irvine 66. AD, Kenny LC, Murray DM, et al. Antenatal 60. Suaini NH, Zhang Y, Vuillermin PJ, Allen vitamin D exposure and childhood eczema, KJ, Harrison LC. Immune Modulation by food allergy, asthma and allergic rhinitis at Vitamin D and Its Relevance to Food Allergy. 2 and 5 years of age in the atopic disease- Nutrients. 2015;7(8):6088-108. specific Cork BASELINE Birth Cohort Study. Allergy. 2018. 61. Adorini L, Penna G. Induction of tolerogenic dendritic cells by vitamin D receptor agonists. 71. Poole A, Song Y, Brown H, Hart PH, Zhang Handb Exp Pharmacol. 2009(188):251-73. GB. Cellular and molecular mechanisms of vitamin D in food allergy. J Cell Mol Med. 62. Benede S, Berin MC. Mast cell heterogeneity 2018;22(7):3270-7. underlies different manifestations of food allergy in mice. PLoS One. 72. Wu J, Zhong Y, Shen X, Yang K, Cai W. 2018;13(1):e0190453. Maternal and early-life vitamin D deficiency enhances allergic reaction in an ovalbumin- 63. Schouten B, van Esch BC, Hofman GA, van sensitized BALB/c mouse model. Food Nutr den Elsen LW, Willemsen LE, Garssen J. Res. 2018;62. Acute allergic skin reactions and intestinal contractility changes in mice orally 73. Matsui T, Yamashita H, Saneyasu KI, Tanaka sensitized against casein or whey. Int Arch H, Ito K, Inagaki N. Vitamin D deficiency Allergy Immunol. 2008;147(2):125-34. exacerbates sensitization and allergic diarrhea in a murine food allergy model. 64. Silva CM, Silva SAD, Antunes MMC, Silva G, Allergol Int. 2018;67(2):289-91. Sarinho ESC, Brandt KG. Do infants with cow’s milk protein allergy have inadequate 74. Corripio-Miyar Y, Mellanby RJ, Morrison levels of vitamin D? J Pediatr (Rio J). K, McNeilly TN. 1,25-Dihydroxyvitamin D3 2017;93(6):632-8. modulates the phenotype and function of Monocyte derived dendritic cells in cattle. 65. Perezabad L, Lopez-Abente J, Alonso- BMC Vet Res. 2017;13(1):390. Lebrero E, Seoane E, Pion M, Correa-Rocha R. The establishment of cow’s milk protein 75. Barragan M, Good M, Kolls JK. Regulation allergy in infants is related with a deficit of Dendritic Cell Function by Vitamin D. of regulatory T cells (Treg) and vitamin D. Nutrients. 2015;7(9):8127-51. Pediatr Res. 2017;81(5):722-30.

178

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 178 GENERAL DISCUSSION

76. Bakdash G, van Capel TM, Mason LM, enhance regulatory T and B cell frequencies. Kapsenberg ML, de Jong EC. Vitamin D3 Mol Nutr Food Res. 2017;61(11). metabolite calcidiol primes human dendritic 87. Goverse G, Olivier BJ, Molenaar R, cells to promote the development of Knippenberg M, Greuter M, Konijn T, et immunomodulatory IL-10-producing T cells. al. Vitamin A metabolism and mucosal Vaccine. 2014;32(47):6294-302. immune function are distinct between 77. Caccamo D, Ricca S, Curro M, Ientile R. BALB/c and C57BL/6 mice. Eur J Immunol. Health Risks of Hypovitaminosis D: A Review 2015;45(1):89-100. of New Molecular Insights. Int J Mol Sci. 2018;19(3). 78. Bscheider M, Butcher EC. Vitamin D immunoregulation through dendritic cells. Immunology. 2016;148(3):227-36. 79. Erkelens MN, Mebius RE. Retinoic Acid and Immune Homeostasis: A Balancing Act. Trends Immunol. 2017;38(3):168-80. 80. Turfkruyer M, Rekima A, Macchiaverni P, Le Bourhis L, Muncan V, van den Brink GR, et al. Oral tolerance is inefficient in neonatal mice due to a physiological vitamin A deficiency. Mucosal Immunol. 2016;9(2):479-91. 81. Xiao L, Cui T, Liu S, Chen B, Wang Y, Yang T, et al. Vitamin A supplementation improves the intestinal mucosal barrier and facilitates the expression of tight junction proteins in rats with diarrhea. Nutrition. 2019;57:97- 108. 82. Penniston KL, Tanumihardjo SA. The acute and chronic toxic effects of vitamin A. Am J Clin Nutr. 2006;83(2):191-201. 83. van Esch BC, van Bilsen JH, Jeurink PV, Garssen J, Penninks AH, Smit JJ, et al. Interlaboratory evaluation of a cow’s milk allergy mouse model to assess the allergenicity of hydrolysed cow’s milk based infant formulas. Toxicol Lett. 2013;220(1):95- 102. 84. Chang HY, Mitzner W, Watson J. Variation in airway responsiveness of male C57BL/6 mice from 5 vendors. J Am Assoc Lab Anim Sci. 2012;51(4):401-6. 8 85. Sbierski-Kind J, Kath J, Brachs S, Streitz M, von Herrath MG, Kuhl AA, et al. Distinct Housing Conditions Reveal a Major Impact of Adaptive Immunity on the Course of Obesity- Induced Type 2 Diabetes. Front Immunol. 2018;9:1069. 86. Kiewiet MBG, van Esch B, Garssen J, Faas MM, de Vos P. Partially hydrolyzed whey proteins prevent clinical symptoms in a cow’s milk allergy mouse model and

179

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 179 530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 180 Abbreviations

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 181 182

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 182 ABREVIATIONS

ABBREVIATIONS

AD atopic dermatitis ALDH aldehyde dehydrogenase Breg regulatory B cell CCL c-c motif chemokine ligand CCR c-c motif chemokine receptor CD cluster of differentiation CMA cow’s milk allergy CT cholera toxin CTLA4 cytotoxic T lymphocyte antigen 4 CXCL c-x-c motif chemokine ligand CXCR c-x-c motif chemokine receptor

CYP27B1 25-hydroxyvitamin D3 1-alpha-hydroxylase DC dendritic cells DHA docosahexaenoic acid DNA deoxyribonucleic acid DTH delayed-type hypersensitivity ELISA enzyme-linked immunosorbent assay EPA eicosapentaenoic acid FBS foetal bovine serum FMO fluorescent minus one Foxp3 forkhead box P3 GF scGOS/lcFOS GFA scGOS/lcFOS/pAOS GM-CSF granulocyte-macrophage colony-stimulating factor g.o.i. gene of interest GzMB granzyme B HSC hematopoietic stem cell HIV human immunodeficiency virus i.d. intradermal IFN interferon i.g. intragastrical Ig immunoglobulin Igflc immunoglobulin free light chain IL interleukin i.p. intraperitoneally IRF interferon response factor lcFOS long chain Fructo-oligosaccharides

183

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 183 LP lamina propria LPC lymphoid progenitor cell MHC major histocompatibility complex MLN mesenteric lymph nodes mMCP-1 mouse mast cell protease 1 MSC mesenchymal stem cell NDO non-digestible oligosaccharides NV non-vaccinated OVA ovalbumin pAOS pectin-derived Acidic-oligosaccharides PBS phosphate-buffered saline PP Peyers patches PS penicillin-streptomycin PUFA polyunsaturated fatty acid RA retinoic acid RALDH retinaldehyde dehydrogenase RANK receptor activator of nuclear factor kappa-B RANKl receptor activator of nuclear factor kappa-B ligand Rorγ RAR-related orphan receptor gamma RPS13 ribosomal protein S13 RXR retinoid X receptor scGOS short chain Galacto-oligosaccharides s.i. small intestine sWPC60 sweet whey protein concentrate 60 Tbet T-box transcription factor Teff effector T cells Th T helper cells TGFβ transforming growth factor beta TLR toll like receptor TNF tumor necrosis factor alpha Tr1 regulatory T cell type 1 Treg regulatory T cells VDR vitamin D receptor VEGFr vascular endothelial growth factor receptor VitA vitamin A VitD vitamin D

184

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 184 ABREVIATIONS

185

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 185 530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 186 Appendix Nederlandse samenvatting Dankwoord Curriculum Vitae List of publications

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 187 APPENDIX

188

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 188 NEDERLANDSE SAMENVATTING

NEDERLANDSE SAMENVATTING

Voedselallergie, de darm en het afweersysteem van de baby Voedselallergie komt steeds vaker voor en hierbij treden allergische reacties vaak op tegen eiwitten in kippenei, koemelk, pinda, noten, vis, schaal- en schelpdieren of soja. Voedselallergie is een verkeerde reactie van het afweersysteem van het lichaam tegen een ongevaarlijk voedseleiwit. In totaal heeft vier tot acht procent van de kinderen en twee tot drie procent van de volwassenen een voedselallergie De allergische reactie kan onder andere bestaan uit jeuk, eczeem, benauwdheid, misselijkheid, diarree of obstipatie, en in het meest ernstige geval een anafylactische shock. Koemelkallergie is een van de eerste vormen van voedselallergie en deze kan zich al op jonge leeftijd ontwikkelen. Koemelkallergie, gediagnostiseerd en bevestigd door een arts, komt bij ongeveer twee tot drie procent van de kinderen voor. Ongeveer 80 procent van de koemelkallergische kinderen groeit vanzelf over de allergie heen.

Het maag-darmkanaal wordt bekleed door een laagje cellen, de epitheelcellen. Vlak onder het epitheel bevinden zich de afweercellen van de darm. Vlak na de geboorte is zowel de rijping van darm als de ontwikkeling van het afweersysteem nog niet volledig en moet er een stabiele en diverse samenstelling van de darmbacteriën gevormd worden. De eerste paar weken tot maanden is het darmepitheel nog wat meer doorlaatbaar voor deeltjes die in de darm aanwezig zijn dan later in het leven. Zo is er in deze periode meer kans dat er voedseleiwitten door het epitheel gaan en in contact komen met de afweercellen. Over het algemeen zullen de afweercellen in de darm hier niet op reageren met een ontstekingsreactie en het voedseleiwit zien als ongevaarlijk. Zo wordt je tolerant voor een ongevaarlijk voedseleiwit. Soms kan er echter een speciaal type ontstekingsreactie ontwikkeld worden tegen het eiwit, het afweersysteem is immers nog in opleiding. Zo kan er een allergie tegen koemelkeiwitten ontstaan. Bij koemelkallergie is de balans in het afweersysteem, ook wel immuunsysteem genoemd, dus verstoord. Het immuunsysteem is echter nog beïnvloedbaar door prikkels van buitenaf, bijvoorbeeld van de bacteriën die aanwezig zijn aan de binnenkant van de darm. Deze bacteriën kunnen er mede voor zorgen dat de balans zich herstelt en dat er alsnog tolerantie wordt opgebouwd voor het voedseleiwit. De meeste koemelkallergische kinderen groeien gelukkig in de eerste levensjaren over hun allergie heen, maar ze blijven wel gevoeliger voor het ontwikkelen van andere allergieën zoals astma. Het tegengaan van het ontwikkelen van koemelkallergie in jonge kinderen is dan ook van groot belang.

Niet-verteerbare oligosachariden en het tegengaan van koemelkallergie De Wereldgezondheidsorganisatie (WHO) raadt aan om tenminste 6 maanden borstvoeding te geven. In borstvoeding zitten namelijk ontelbare verschillende soorten stoffen die de gezondheid van de baby ondersteunen. Niet-verteerbare oligosachariden zijn ook onderdeel van de borstvoeding en deze zijn niet alleen belangrijk voor het stimuleren van groei van gunstige

189

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 189 APPENDIX

bacteriën in de darm, maar ook voor de ondersteuning van het immuunsysteem. Op basis van eigenschappen van deze moedermelksuikers zijn een aantal suikerstructuren ontwikkeld die kunnen worden toegevoegd aan poedermelk voor baby’s. Uit eerder onderzoek is gebleken dat deze oligosachariden het ontwikkelen van koemelkallergie kan tegengaan in muizen. Van een mix van drie van deze niet-verteerbare suikerstructuren, namelijk neutrale korte keten Galacto-oligosachariden (scGOS), neutrale lange keten Fructo-oligosachariden (lcFOS) en zure oligosachariden afgeleid van pectine (pAOS) (GFA), is in dit proefschrift bestudeerd hoe ze koemelkallergie kunnen verminderen in een muismodel en of ze de vaccinatie tegen influenza kunnen ondersteunen.

Muismodel Vier weken jonge muizen werden in groepen verdeeld en kregen een controle dieet of een dieet met daaraan toegevoegd 1% GFA. Hierna werden de muizen op een kunstmatige manier via de darm allergisch gemaakt tegen het koemelkeiwit wei. Dit vond elke week eenmaal plaats, gedurende vijf weken. Na deze vijf weken werd bestudeerd of de muizen koemelkallergie hadden ontwikkeld door middel van het blootstellen van de huid van de oren aan koemelkeiwit (een huidpriktest). Een acute allergische reactie tegen dit eiwit kenmerkt zich namelijk als een zwelling van het oor binnen een uur na de huidprik. Een dag na de huidpriktest kregen de muizen een provocatie met het koemelkeiwit wei door toediening aan het maagdarmkanaal, waarna de dieren werden opgeofferd, er bloed afgenomen werd en verschillende organen verwijderd om verdere allergische parameters te kunnen meten (Hoofdstuk 3, figuur 1).

Werking van oligosachariden In hoofdstuk 3 van dit proefschrift zijn de allergie beschermende effecten van de drie suikerstructuren los (scGOS, lcFOS, pAOS) en/of in combinaties (GF, GFA) in een preventief of therapeutisch model getest (hoofdstuk 3). Preventief houdt in dat de muizen de dieetinterventie al kregen twee weken voordat ze allergisch werden gemaakt. Bij de therapievorm kregen de muizen de diëten nadat ze allergisch waren gemaakt (Hoofdstuk 3, Figuur 1). De losse oligosacchariden, in de dosering zoals aanwezig in de GFA mix, hadden geen preventief effect op de uitkomst van de allergische huidreactie, de anafylactische shock symptomen of de hoeveelheid wei-specifiek immunoglobuline E (IgE) in het serum van bloed. De combinaties GF en GFA hadden wel een preventief effect en verminderden het ontwikkelen van koemelkallergie en dit was gekoppeld aan het beïnvloeden van parameters in het immuunsysteem in de darm. GF of GFA hadden echter in het therapeutische model geen effect op de allergische reactie.

Het immuunsysteem Het immuunsysteem bevat verschillende typen witte bloedcellen die betrokken zijn bij de aangeboren of verworven afweerreactie. Het aangeboren immuunsysteem (aspecifiek) heeft geen selectieve herkenning voor bepaalde eiwitten, maar reageert meer algemeen

190

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 190 NEDERLANDSE SAMENVATTING

op immuunstimulerende componenten zoals bijvoorbeeld virale of bacteriële deeltjes en toxines via patroon herkenningsreceptoren en bestaat uit dendritische cellen, macrofagen, mestcellen, granulocyten, natural killer cellen en daarbij helpt ook het complement systeem. Het verworven immuunsysteem (specifiek of adaptief) reageert specifiek op bepaalde eiwitten van ziekteverwekkers of tumorcellen en bouwt een geheugen op voor de herkenning van deze specifieke eiwitten. Bij een tweede blootstelling van dezelfde ziekteverwekker wordt er snel gereageerd met een ontstekingsreactie specifiek gericht tegen deze ziekteverwekker. Dit systeem bestaat uit verschillende soorten T-helper cellen, B-cellen, regulatoire T- en B-cellen, cytotoxische T-cellen en specifieke T- en B- geheugencellen. Om allergisch te worden vindt er een eerste reactie plaats en wordt er IgE gevormd tegen een eiwit (allergeen) wat aan effectorcellen zoals mestcellen bindt (sensibilisatie) en pas bij het tweede contact met het allergeen zal een allergische reactie optreden doordat het allergeen het IgE bindt en de mestcel activeert, waardoor er onder andere histamine wordt vrijgezet. Voor sensibilisatie zal een dendritische cel het wei-eiwit oppikken en meenemen via de lymfekanalen naar de verzamel lymfeklier in de darm, de mesenteriale lymfeklier. De dendritische cel vertelt een naïeve T-cel, die het allergeen dat aangeboden wordt door de dendritische cel herkent, hoe deze zich verder moet ontwikkelen. In het geval van sensibilisatie zal deze T-cel een T-helper-2 cel worden. De geactiveerde T-helper-2 cel zal een antilichaam producerende B-cel, die herkenning heeft voor dit allergeen, instrueren dat het wei eiwit gevaarlijk is, waardoor deze wei-specifieke immunoglobulines gaat maken en verder uitrijpt tot een IgE producerende plasmacel. Dit IgE zal vervolgens aan mestcellen binden. Bij een tweede contact met het wei eiwit zal het wei het specifieke IgE samen op effectorcellen zoals mestcellen binden en de mestcel zal dan degranuleren, wat inhoudt dat specifieke stoffen uit de cel vrij worden gegeven. Hierbij komt bijvoorbeeld histamine vrij waardoor er of lokaal of een systemische allergische reactie optreedt (Hoofdstuk 1, Figuur 1). Lokale allergische reacties zoals jeuk, darmklachten, misselijkheid of een extreem sterke allergische reactie met anafylactische shock kunnen zo dus plaatsvinden.

Oligosachariden beïnvloeden T-cellen Zowel GF als GFA konden preventief de koemelkallergische reactie verminderen, alleen was het effect op immuunmarkers in de darm verschillend. GF onderdrukte de transcriptiefactoren (eiwitten die specifieke stukjes DNA herkennen en gebruiken) van T-helper-2 cellen en T-helper-17 cellen, terwijl GFA transcriptiefactoren van T-helper-1 cellen en de regulatoire stof ‘interleukine-10’ verhoogde. Zowel GF als GFA zorgden voor een verhoogde transcriptie van Foxp3, een transcriptiefactor voor regulatoire T-cellen. De regulatoire T-cellen kunnen direct via cel-cel contact en indirect via het vrijmaken van regulatoire stoffen (interleukine-10 en TGFβ) T-helper cellen onderdrukken om zodoende een allergische ontstekingsreactie te verminderen. Ook kunnen regulatoire T-cellen activatie van mestcellen onderdrukken waardoor er minder allergische symptomen optreden. GF en GFA hadden niet alleen effect op

191

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 191 APPENDIX

immuunmarkers van de T-cel populatie, er werd ook minder wei-specifiek immunoglobuline (Ig)G1 in het serum van het bloed gemeten en een minder sterke oorzwelling in de huidpriktest. De muizen met GF of GFA in het dieet hadden dus ook minder sterke allergische symptomen dan de muizen op controle dieet.

Cytokines ‘interleukine 10’ en ‘transforming growth factor bèta’ Voor GFA is verder onderzoek gedaan naar het mechanisme achter het preventieve dieet effect op de koemelk allergische symptomen. In hetzelfde preventieve muismodel, beschreven hierboven, werd twee keer een antilichaam toegediend tegen de ‘interleukine (IL)10 receptor’ of ‘transforming growth factor bèta’ (TGFβ) (Hoofdstuk 4, Figuur 1). Een antilichaam blokkeert de werking van één receptor of stof voor een bepaalde tijd. Cytokines IL10 en TGFβ zijn betrokken bij de vorming van regulatoire T-cellen en bij de onderdrukking van T-helper cellen, B-cellen en mestcellen. De allergische huidreactie die was verminderd in muizen die het GFA-dieet kregen was door de behandeling met anti-IL10 receptor en anti-TGFβ antilichamen weer net zo hoog als de muizen die het controle dieet kregen. IL10 en TGFβ bleken dus betrokken bij de beschermende werking van het GFA-dieet. Opmerkelijk was dat in de groep muizen die gevoerd werd met het GFA-dieet en behandeld met het anti-IL10-receptor antilichaam, wel veel regulatoire T-cellen aangemaakt werden. Wij hebben dit niet verder onderzocht maar anderen hebben gevonden dat regulatoire T-cellen die behandeld werden met anti-IL10- receptor antilichamen wel gemaakt konden worden maar niet functioneerden.

GFA ondersteunt de immuunreactie tijdens vaccinatie Het preventieve dieet met GFA kon de koemelk allergische symptomen verminderen door invloed uit te oefenen op de regulatoire T-cellen en regulatoire cytokines en transcriptiefactoren van T-helper-1 cellen in de darm. In hoofdstuk 5 is daarom de invloed van een dieet met GFA op de vaccinatie reactie tegen een griepvaccin in een muismodel bestudeerd. T-helper cellen zijn bij vaccinaties namelijk belangrijk voor het opbouwen van de specifieke afweerreactie tegen de ziekteverwekker waartegen gevaccineerd wordt. GFA ondersteunde het verworven immuunsysteem voor griepvaccinatie door het aantal T-helper-1 cellen te verhogen en het aantal T-helper-2 cellen te verlagen. Dit kwam doordat het GFA-dieet een specifieke regulatoire-Th1-cel, die ervoor zorgt dat er niet veel T-helper-1 cellen worden gemaakt, onderdrukte. Dus de rem op de ontwikkeling van de T-helper-1 cellen werd stilgezet, zodat de T-helper-1 populatie kon groeien, wat beter was voor de werkzaamheid van de vaccinatie.

Vitamines A en D Het is bekend dat gezonde voeding voor een sterk immuunsysteem kan zorgen. Sommige voedingstoffen, zoals vitamines, worden essentieel genoemd, omdat het lichaam ze niet, in beperkte mate of alleen onder bepaalde omstandigheden/ voorwaarden zelf kan aanmaken. Vitamine A en D zijn twee vitamines waarvan bekend is dat ze een grote invloed kunnen uitoefenen

192

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 192 NEDERLANDSE SAMENVATTING

op het immuunsysteem. Uit eerder onderzoek is gebleken dat vitamine A de dendritische cel in de mesenteriale lymfeklier beïnvloedt om regulatoire T-cellen te vormen. Hierdoor kan vitamine A mogelijk helpen om allergie te voorkomen en een gezond immuunsysteem in stand te houden. Ook zijn er aanwijzingen dat een vitamine D tekort samenhangt met de ontwikkeling van allergie en zijn er op verschillende soorten immuuncellen vitamine D receptoren gevonden. Om die redenen hebben we onderzocht wat het effect van zowel vitamine A als vitamine D op koemelkallergie is en of deze vitamines de preventieve werking van het GFA-dieet in het koemelkallergie muismodel ondersteunen.

Zowel een hogere dosering van vitamine A als vitamine D alleen had geen effect op koemelkallergie ontwikkeling. Een gemiddelde of hoge dosering vitamine D had ook geen invloed op de vermindering van koemelkallergische symptomen door het GFA-dieet. Wel werd er in de groep muizen die het dieet met GFA en een hoge dosering vitamine D kregen, een verhoging van het percentage dendritische cellen gemeten die mogelijk regulatoire T-cellen kunnen vormen. Ondanks deze verhoging in dendritische cellen werd er geen versterking van het beschermende effect van GFA gemeten met de huidpriktest of in het bloed. Vitamine A ondersteunde daarentegen wel het beschermende effect van het GFA-dieet op de koemelk allergische symptomen. In de groep muizen die GFA en de hoge dosering vitamine A in het dieet kregen, werd een verminderde allergische huidreactie gemeten in tegenstelling tot de muizen die alleen extra vitamine A of GFA kregen. Uit deze studie is echter niet duidelijk naar voren gekomen hoe vitamine A samen werkt met GFA. Ook bleek het allergie beschermende effect van GFA in deze studie in tegenstelling tot eerdere studies niet op te treden. Er zou dus verder onderzoek gedaan moeten worden naar de allergie beschermende werking van een combinatie van een dieet met oligosachariden en een verrijking van het vitamine A gehalte.

Conclusie De niet-verteerbare oligosachariden in de GFA-mix hebben een positieve invloed op koemelkallergie preventie in het muismodel, doordat ze allergische symptomen verminderen. GFA had een effect op immuunmarkers in de darm door onder andere de expressie van cytokines ‘interleukine 10’ en ‘TGFβ’ en van transcriptie factoren voor regulatoire T-cellen en T-helper-1 cellen te verhogen. ‘IL10’ en ‘TGFβ’ bleken essentieel bij de beschermende werking van het GFA-dieet. Ook ondersteunde het GFA-dieet de vaccinatie respons door de regulatoire- Th1 cellen te verlagen, waardoor er meer T-helper-1 cellen ontstonden. Het effect van GFA op koemelkallergie werd ondersteund door aanwezigheid van extra vitamine A. Er zal verder onderzoek gedaan moeten worden om te bepalen of GFA ook de allergische symptomen via cel-cel contact tussen regulatoire T-cellen en allergische effector cellen kan beïnvloeden. Vervolgens kan verder bepaald worden in welke cellen GFA de productie van ‘IL10’ en ‘TGFβ’ verhoogt om uiteindelijk te leiden tot het verminderen van de koemelk allergische symptomen.

193

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 193 APPENDIX

Door een beter inzicht in de beschermende werking van GFA tegen het ontwikkelen van koemelkallergie, kunnen in de toekomst kinderen met een verhoogd risico op koemelkallergie beter worden geholpen.

194

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 194 NEDERLANDSE SAMENVATTING

195

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 195 APPENDIX

196

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 196 DANKWOORD

DANKWOORD

Dan is mijn promotie bijna klaar en schrijf je aan het belangrijkste en meest gelezen stuk van het hele proefschrift, het dankwoord. Ik wil iedereen die dit leest bedanken, zo vergeet ik niemand, die op wat voor manier dan ook mij in de afgelopen negen (!) jaar geholpen heeft. Laten we bij het begin beginnen. De eerste die ik heb ontmoet zijn Linette en Léon. Ik weet nog goed hoe ik in het Went gebouw door allerlei gangetjes naar een kamertje moest waar jullie met zijn twee in een kamertje vol met wetenschappelijke zooi achter een tafeltje zaten voor mijn interview. En dat we samen na negen jaar zover zijn gekomen vind ik zo tof. Het heeft tijd en geduld gekost om verschillende redenen, maar hier zijn we dan. Jullie via deze weg bedanken, daar doe ik jullie mee te kort. Jullie zijn allebei anders en beide hebben jullie op je eigen manier mij zoveel geleerd. Als ik dit allemaal had geweten aan het begin van mijn promotie, dan had ik nog steeds met niemand anders deze promotie willen doen.

Linette, ik was niet zo goed in de puntjes op de i, maar jij hebt altijd weer tijd voor me genomen om me daar mee te helpen. Of om door te vragen zodat ik het toch weer scherp voor ogen had waar ik mee bezig was, ook toen je tijdelijk wat minder ging werken. De eerste versie van mijn eerste paper, kreeg ik helemaal rood (verbeterd) terug. Dat was even slikken, maar door er samen stap voor stap doorheen te lopen, en mij uit te leggen waarom iets ander moest, dàt was een leermoment voor mij. Enorm respect voor jou, ik zou namelijk nooit zoveel geduld met mezelf hebben gehad. En ik was niet de enige PhD die van alles van je wilde, je had er meer, en toch wist je tijd vrij te maken. Ook heb je me (denk ik) meer dan eens in de afgelopen jaren verdedigd zodat ik mijn promotie toch af kon maken. Even een laptop regelen die het wel weer doet met alle data, ‘even’ een rustige werkplek voor me regelen i.v.m. de gevolgen van een auto-ongeluk. Je bent iemand die niet snel opgeeft en voor de volle 200% achter iemand gaat staan. Dank je daarvoor, dat betekent heel veel voor mij.

Léon, als ik aan je denk dan schieten me te veel dingen te binnen om hier neer te zetten. Ik weet nog goed dat jij ontzettend nieuwsgierig was naar de uitkomst van mijn ‘kleurenschema’. Hoe blauw jij en ik absoluut niet zijn in dat schema (maar wel een beetje aangeleerd hebben), en dat ik bij jou ‘mijn stoom’ kon afblazen zonder dat je er raar van stond te kijken. Zelfs niet met wat meer persoonlijkere onderwerpen. Niet alleen met wetenschappelijke inzichten wist je me te verrassen, ook tijdens een congres in Amsterdam heb ik een hoop geleerd. Vooral de avond zal ik niet vergeten, één en al gezelligheid om met jou op pad te zijn. Alleen hoe jij de volgende ochtend dan weer in mijn ogen fris en fruitig met de wetenschap bezig kon zijn, dat is mij nog steeds een raadsel. Ook jij bent druk en hebt meerdere collega’s die je aanstuurt en nooit heb ik het gevoel gehad dat je geen tijd voor me had. Daar drinken we straks wat op.

197

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 197 APPENDIX

En na dat eerste gesprek mocht ik bij Johan komen praten. Inspirerend indrukwekkend vond ik je in dat eerste gesprek. En een beetje zenuwachtig werd ik ook wel van je omdat je zoveel wist (en weet). Jouw deur staat altijd open, en zoals je laatst nog herhaalde, met tissues voor de AIO die het nodig heeft. Waarop jij me nog zei: “Jou heb ik hier nog nooit met tissues zien zitten.” Klopt, maar ik heb je kamer wel eens gebruikt om ‘even’ uit te rusten met mijn ogen dicht toen ik zwanger was. Als er iemand van hot naar her vliegt ben jij het, maar toch ook altijd bereikbaar. Na ons eerste gesprek zijn er vele meer gevolgd, en niet alleen over een snelle werk update die je soms wilde. Met jou kun je over een breed scala aan onderwerpen praten en discussiëren, dat maakt je zo ontzettend toegankelijk als promotor. Je gaf me het gevoel welkom te zijn, en dat het telt dat je er bent, of ik nu bij Nutricia zat of op de Universiteit. Zolang ik het werk maar gedaan kreeg, uiteraard. Onlangs na het printen (en uitdelen) van mijn A4 versie van mijn proefschrift, kreeg ik een appje van je: “het ziet er heel goed uit”. Dank je Johan dat je mijn promotor bent geweest.

Er zijn zoveel mensen waar ik mee gewerkt heb! Waar ben je zonder iemand die alles weet, eerst waren daar Els (UU) en Saskia (Nutricia). Ze wisten eigenlijk alles, vooral ook waar Johan was. Later kwamen Yvonne (Nutricia), die ook steeds meer voor mij in Léons agenda naar ruimte moest zoeken en Lidija (UU), zonder jouw Word/Acrobat kwaliteiten was ik gek geworden door de laatste veranderde loodjes in het MyPhD-systeem. Bastiaan, ik zette jouw project voort en in het begin kon ik altijd bij je terecht voor elke vraag die ik had. Ik kwam ook op jouw plekje te zitten in het Went gebouw in Utrecht. Op de kamer bij Daphne en Tom; vol, knus en supergezellig. In het begin hebben Betty en Alma me veel op weg geholpen. Later binnen Nutricia waren Anneke en Prescilla mijn vraagbaak, maar ik kon ongevraagd ook bij Bea terecht. Je maakte altijd tijd voor me Bea, en je wist me doen te geloven dat ik dit (lees promoveren) gewoon kon. Ik werkte dus ook bij Nutricia in Wageningen. De ene keer zat ik bij Paul, Jeroen en Lieke dan weer bij Laura, Tjalling, Désirée en Nienke of bij Jacqueline en Nicole en soms ergens anders. De dagen vlogen voor mij voorbij. Ik wil jullie niet alleen bedanken voor jullie hulp ook bedankt voor alle gezelligheid. Later in de nieuwbouw (David de Wied) heb ik ook overal gezeten, niet geheel vrijwillig. Ik zat eerst bij Suzan, Tom en Caroline, op de open vide en dichtbij het lab. Maar vanwege een ongeluk had ik een rustigere werkplek nodig en ik heb tijdens Astrids verlof haar plekje warm gehouden in het hoekje van de tweede verdieping. Over verlof gesproken, veel van de gesprekken gingen niet alleen over de werk- inhoud, ook over vrouwenzaken. Tjalling heeft bij Nutricia wel wat voor zijn kiezen gekregen denk ik, zeker met al die dames die wel van een gesprek hielden. Tjalling je bent gelukkig nog wel met me naar het kerstdiner geweest!

Zoveel mensen, zoveel verhalen. Sander, je hebt me weer op weg geholpen na mijn ongeluk door toch het geduld te nemen en me dingen twee (of soms drie) keer uit te leggen. Ik kletste net zo graag met alle andere collega’s van de UU, eerst plakte in nog wel eens vast in de kamer

198

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 198 DANKWOORD

van Astrid en Naomi in het Went of bij Hans in de dierenstallen. Er is ook veel gespreksstof en koffie doorheen gegaan in de koffiehoek van het Wied met Carmen en Astrid. Het is ook wel makkelijk wanneer je dezelfde kant op reist, dan lijkt het net alsof je in vijf minuten op je bestemming bent. Als ik in het Goffertpark kom dan denk ik altijd aan Marco (en dan natuurlijk ook aan Liz) en soms nog aan FACS-perikelen. Of nu als ik door Nijmegen Noord fiets waar Laura heeft gewoond, toen ik er de weg echt niet vinden kon. Ik weet nog dat Lieke en ik bij jou in de auto zaten en er kinderliedjes op stonden “Wacht maar af,” zei je, “die kennen jullie straks ook allemaal.” Inderdaad heeft Laura gelijk gekregen, en ook alle nieuwe liedjes van Kinderen voor Kinderen.

Ik heb ook een aantal stagiaires begeleid. Annika, jij was mijn proefkonijn, mijn allereerste stagiaire. Samen met Henk heb je voor mij hele mooie histologie plaatjes gemaakt voor mijn eerste artikel. Stiekem wilde ik jou houden Annika, zo goed heb je gewerkt. Suzan en Lotte, ook jullie hebben tijdens je bachelor stage ontzettend veel histologie gedaan. Daar is de basis gelegd voor een stuk in mijn tweede artikel. Ik heb jullie koffiemok nog steeds. Paquita heeft in haar eindstage het plaatje voor het tweede artikel afgemaakt voor histologie. Henk, je hebt me daar echt enorm mee geholpen bij het histologie gedeelte. Anna, you became my last intern during my second pregnancy. After a short introduction you did two giant experiments by yourself (under supervision of Désirée). And you wrote an excellent report. You became a PhD yourself, and I wish you a good scientific career. Alle stagiaires ontzettend bedankt voor jullie hulp!

Ik wil ook kort nog zeggen dat ik heel blij ben dat ik bij Jeroen van Leeuwen en Inèz Meijer stage heb gelopen op de FNWI van de Radboud Universiteit. Naast dat Inèz en ik vriendinnetjes zijn geworden, heb ik van jullie geleerd dat onderzoek doen heel leuk is. Alleen wil ik nooit meer western blotjes gieten, ik bestel ze nu liever kant en klaar. Ook wil ik alle vrienden van mij bedanken die in de afgelopen jaren gevraagd hebben hoe het met mij en mijn promotie ging. Een dagje op pad, vaak samen met de kinderen, was een aangename afwisseling en afleiding van het promoveren. Jullie steun vind ik echt supertof!

Mijn paranimf Désirée, bij jou zat ik, denk ik, het meest in de auto en ik heb veel met je samen gewerkt. Met de kleurtjes-trainingsdag zaten we voor het eerst naast elkaar. Bij mij viel het kwartje iets later, we kenden elkaar van Organon. Gedurende die periode zijn we steeds meer samen gaan werken. Zonder jouw expertise, kunde en inzet waren mijn laatste twee proeven nooit meer van de grond gekomen. En nu in mijn 0-uren-contract-jaren, laptop meenemen, geschreven artikelen doorlezen, proefschrift mee helpen printen, je hebt het (voor mij) er zo even bij gedaan. Ondertussen ben je gewisseld van Nutricia naar een promotieplaats op de universiteit, en dan ook nog eens uitdagend in drie jaar tijd. Als iemand het kan, dan ben jij het. Ik ben zo blij met jou!

199

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 199 APPENDIX

Mijn andere paranimf Daniëlle, mijn vriendinnetje, mijn zus. Deze pagina’s gevuld met mijn promotie onderwerp, dat is een schijntje vergeleken met de pagina’s die ik kan vullen over ons. Wat ben ik blij dat we in Noordwijk hecht zijn geworden en dat al zo lang (ik ga geen jaartallen noemen). Door dik en dun sta je altijd voor me klaar, bij jou ben ik ook thuis. Laten we vooral samen heel oud en grijs worden. Maar eerst dit feestje vieren, lief leven samen!

Thuis, dat is een makkelijk bruggetje. Ik ken Roy net iets minder lang dan mijn zus, maar ook van de middelbare school. Samen hebben we veel meegemaakt en menig hobbel overwonnen, mooie hobbels en minder mooie hobbels, maar wel samen. En dat heeft geleid tot Nikki en Isa. De allermooiste, slimste, knapste, liefste, ondeugendste fantastische kleine meisjes. Om verwarring te voorkomen, ik heb gepubliceerd onder mijn meisjesnaam Kerperien, maar ik ben ook mrs. Rouwhorst.

Dan kom ik uit bij mensen die ik al ken vanaf dat ik geboren ben. Mijn opa, geboren in 1928, gaat het nog meemaken dat ik promoveer. Mocht ik ooit zo oud worden, dan hoop ik ook met zoveel levenslust en zo positief in het leven te blijven staan als hij. Mijn oom Manfred met jou heb ik getennist, eerst ballen oprapen, toen bij jou lessen volgen en daarna samen competitie spelen. Tegenwoordig staat er ook altijd een kop koffie klaar. Ik ben blij met jou als mijn oom. Hoe ga ik jou omschrijven, toma Trudy? Ook al woon je in Rotterdam, ik kom al sinds dat ik me kan herinneren logeren bij jou. Je bent er altijd voor me gedurende mijn hele leven, en dat mag zeker nog heel lang blijven. Als ik een tweede mam mocht kiezen, dan ben jij dat. Ik ga mijn kinderen net zoveel bij jou laten logeren als dat ik dat mocht! Gedurende allerlei praktische en gezellige zaken was jij er altijd voor mij, mijn adoptie-papa. Mooi verhaal is dat om te vertellen! Nu ben je Opa-Willem-Twee, dat moet je nog lang blijven hoor!

Naast mijn mam zijn er een aantal mensen niet meer bij die er toch hadden moeten zijn. Doodgaan hoort bij het leven, maar het went niet. Oma, Oma Ita, Truus, Bennie en Hennie, jullie hadden dit eigenlijk zelf moeten kunnen meemaken en lezen. Mama, mijn laatste belofte aan jou heb ik ingelost. Op je sterfbed moest ik twee dingen beloven/doen, één ervan was mijn promotie afmaken. Jij bent en blijft mijn mama. Je hebt me altijd gestimuleerd te leren, meegenomen op culturele uitstapjes van jongs af aan en altijd samen huiswerk gemaakt als ik er niet uit kwam. Je deed het allemaal alleen, met en voor mij (alhoewel waarschijnlijk wel vaak met hulp van 1 opa en 3 oma’s). We lijken op elkaar dus ik ga er maar vanuit dat ik niet altijd de makkelijkste ben geweest. Maar je stond altijd voor me klaar, altijd. Ik heb je nog meegenomen naar Istanboel en Dublin, ik had congresdingen en jij had vakantie. ’s Ochtends zaten we wel samen aan het ontbijt en ’s avonds samen aan het avondeten. Jouw laatste anderhalf jaar ben je de beste oma voor Nikki en Isa geweest die ik me had kunnen wensen. Jij leerde Nikki dat de gele M voor ijsjes stond. Je hebt de eerste stapjes van Isa gezien. Je gaf niet op, ook niet

200

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 200 DANKWOORD

toen je ziek werd en zelfs niet toen je op het laatst morfine kreeg. Wat vind ik het knap dat je zo gevochten hebt, je zou nu zo trots op me zijn geweest. Zonder jou had ik hier nu niet gestaan, ik ben nu trots voor ons allebei.

201

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 201 APPENDIX

202

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 202 CURRICULUM VITAE

CURRICULUM VITAE

JoAnn Kerperien was born on June 23rd 1980 in Borculo, the Netherlands. She graduated from her secondary school, pre-university education, in 1998 at the RKSG Marianum in Groenlo. In the same year she started her study Physical Educational Teacher (ALO) at the Hanzehogeschool in Groningen. In her final year she did her teacher’s internship at 2 different elementary schools in Enschede, which was continued in a full time position at five different elementary schools in Enschede. After two years she decided to make a career change and started the bachelor Biology at the Radboud University in Nijmegen in 2004. She finished her bachelor in 2007 and continued her study with the research master Medical Biology. The first internship was conducted at the department of Cell Biology (FNWI) at the Radboud University under supervision of Inèz Meijer and Dr. Jeroen E.M. van Leeuwen for nine months. She contributed to the research into the interaction of UBPY and 14-3-3 in the control of regulating the EFGR signalling cascade. Her second internship was at Organon/ Schering-Plough in Oss under supervision of Dr. Ir. M.J.C. van Lierop for six months. The internship concerned the research of the apoptotic effects of selective non steroidal immunomodulators on T cells compared to the effects of dexamethason and prednison, under a confidentiality agreement. She received her master’s degree in Medical Biology in June 2010.

In April 2010 she started her PhD research project in the research group of Prof. dr. Johan Garssen under the direct supervision of Dr. Linette E.M. Willemsen and Dr. Léon M.J. Knippels. She studied the mechanism of protection by specific oligosaccharides on cow’s milk allergic symptoms and showed the involvement of IL10 and TGFβ in this preventive effect. She also evaluated the effect of these oligosaccharides in a vaccination model. Finally, the involvement of vitamins A and D in the protective effect of GFA on cow’s milk allergy was assessed. The results of this work are described in this thesis. JoAnn lives in Nijmegen with her husband Roy Rouwhorst and her two children Nikki (2013) and Isa (2014).

203

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 203 APPENDIX

204

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 204 PUBLICATIONS

PUBLICATIONS

• IL-10 Receptor or TGF-β neutralization abrogates the protective effect of a specific nondigestible oligosaccharide mixture in cow-milk-allergic mice. Kerperien J., Veening-Griffioen D., Wehkamp T., van Esch B.C.A.M., Hofman G.A., Cornelissen P., Boon L., Jeurink P.P., Garssen J., Knippels L.M.J., Willemsen L.E.M. 2018 Journal of Nutrition 148 (8): 1372-1379 • Dietary, nondigestible oligosaccharides and Bifidobacterium breve M-16V suppress allergic inflammation in intestine via targeting dendritic cell maturation De Kivit S., Kostadinova A.I., Kerperien J., Morgan M.E., Muruzabal V.A., Hofman G.A., Knippels L.M.J., Kraneveld A.D., Garssen J., Willemsen L.E.M. 2017 Journal of Leukocyte Biology 102 (1): 105-115 • Galectin-9 produced by intestinal epithelial cells enhances aldehyde dehydrogenase activity in dendritic cells in a PI3K- and p38-dependent manner De Kivit S., Kostadinova A.I., Kerperien J., Muruzabal A.V., Morgan M.E., Knippels L.M.J., Kraneveld A.D., Garssen J., Willemsen L.E.M. 2017 Journal of innate immunology 9 (6): 609-620 • Non-digestible oligosaccharides modulate intestinal immune activation and suppress cow’s milks allergic symptoms Kerperien J., Jeurink P.V., Wehkamp T., van der Veer A., van de Kant H.J., Hofman G.A., van Esch B.C.A.M., Garssen J., Willemsen L.E.M., Knippels L.M.J. 2014 Pediatric Allergy and Immunology 25 (8): 747-54 • Alterations in regulatory T cells induced by specific oligosaccharides improve vaccine responsiveness in mice Schijf M.A., Kerperien J., Bastiaans J., Szklany K., Meerding J., Hofman G., Boon L., van Wijk F., Garssen J., Van’t Land B. 2013 PLoS One 8 (9): e75148 • The Usp8 deubiquitination enzyme is post-translationally modified by tyrosine and serine phosphorylation Meijer I.M., Kerperien J., Sotoca A.M., van Zoelen E.J., van Leeuwen J.E. 2013 Cell Signalling 25 (4): 919-30 • Recent advances in immunology to target cancer inflammation and infections Chapter 13: Development of the immune system – early nutrition and consequences for later life JoAnn Kerperien, Bastiaan Schouten, Günther Boehm, Linette E.M. Willemsen, Johan Garssen, Léon M.J. Knippels and Belinda van’t Land 2012, ISBN: 978-953-51-0592-3

205

530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 205 530928-L-bw-Kerperien Processed on: 7-5-2019 PDF page: 206

IMMUNOMODULATORY PROPERTIES OF DIETARY NON-DIGESTIBLE GALACTO-, FRUCTO- AND ACIDIC-OLIGOSACCHARIDES IN VACCINATION AND COW’S MILK ALLERGY JOANN KERPERIEN

Mucosal and systemic immune-regulation by fermentable ﬁ bers

JoAnn Kerperien