EXPERIMENTAL IMMUNOLOGY doi: 10.1111/sji.12271 ..................................................................................................................................................................
Characterization of the Immunogenicity and Allergenicity of Two Cow’s Milk Hydrolysates – A Study in Brown Norway Rats K. L. Bøgh*, V. Barkholt† & C. B. Madsen*
Abstract *Division of Toxicology and Risk Assessment, National Food Institute, Technical University of Denmark, Søborg, Denmark; †Department of Systems Biology, Enzyme and Protein Chemistry, Technical University of Denmark, Kgs. Lyngby, Denmark
Received 21 August 2014; Accepted in revised form 23 December 2014 Correspondence to: K. L. Bøgh, Division of Toxicology and Risk Assessment, National Food Institute, Technical University of Denmark, Mørkhøj Bygade 19, DK-2860 Søborg, Denmark. E-mail: [email protected]
Hypoallergenic infant formulas based on hydrolysed milk proteins are used in the diet for cow’s milk allergic infants. For a preclinical evaluation of the immunogenicity and allergenicity of new protein ingredients for such hypoallergenic infant formulas as well as for the investigation of which characteristics of hydrolysates that contribute to allergenicity, in vivo models are valuable tools. In this study, we examine the immunogenicity and allergenicity of two hydrolysates in a Brown Norway (BN) rat model, using i.p. dosing, which allows for the use of small quantities. Intact BLG, hydrolysed BLG and a hydrolysed whey product suitable for use in extensively hydrolysed formulas were thoroughly characterized for protein chemical features and administered to BN rats by i.p. immunization with or without adjuvant. Sera were analysed for specific IgG and IgE for evaluation of sensitizing capacity, immunogenicity and antibody-binding capacity. For evaluation of eliciting capacity a skin test was performed. The study showed that the hydrolysates had no residual allergenicity, lacking the capacity to sensitize and elicit reactions in the BN rats. Dosing with or without adjuvant induced a large difference in immunogenicity. Only antibodies from rats sensitized to intact BLG with adjuvant were able to bind the hydrolysates, and the whey-based hydrolysate only showed immunogenicity when dosed with adjuvant. This study showed that hydrolysates can be evaluated by an i.p. animal model, but that the choice of in vitro tests used for evaluation of antibody responses may greatly influence the result as well as may the use of adjuvant.
Introduction Cow’s milk is the most common cause of food allergy in young children. Cow’s milk allergy (CMA) affects around 2.5% of the young children [1, 2], of whom 80–85% fortunately outgrow their allergy by the age of 3 years [3, 4]. Strict avoidance of cow’s milk is the management of choice; however, for infants and young children, hypoallergenic cow’s milk formulas based on hydrolysed whey or caseins are available. Both the casein and the whey fraction of cow’s milk hold proteins known to be allergenic. blactoglobulin (BLG), also designated Bos d 5, is the most abundant allergen in whey, where it constitutes approximately 50% of the protein content [5]. Cow’s milk-based infant formulas based on hydrolysates are designated either extensively hydrolysed formulas (eHFs) or partially hydrolysed formulas (pHFs), according to the degree of protein hydrolysis [6] and the molecular weight (MW) of the present peptides [7]. Although there is
274
no international consensus for the definition of eHF and pHF [6, 7], eHFs generally consist of peptides of which more than 95% have a MW of < 3 kDa, while pHFs generally contain peptides where a higher percentage have a MW between 3 and 10 kDa and a much lower percentage have a MW below the 3 kDa [8]. When eHFs are developed, the goal is to degrade proteins to such an extent that all potential allergenicity is lost. However, this may likely result in a concomitant loss of immunogenicity, preventing the immune system from developing tolerance to the milk proteins. In contrast, pHFs are designed to minimize the sensitizing capacity, though at the same time maintaining peptides with sizes large enough to be recognized by the immune system for induction of tolerance [9]. While eHFs are considered efficient for prevention of allergic reactions, in the management of the allergy in infants with a diagnosed CMA, pHFs are on the other hand suggested to be useful for the prevention of de novo sensitization, in the management of the allergy in
Ó 2015 John Wiley & Sons Ltd
K. L. Bøgh et al. Characterization of Milk Hydrolysates 275 ..................................................................................................................................................................
infants with a high risk for developing CMA [9, 10]. These considerations are supported by animal studies, showing a general lower allergenicity of eHF compared to pHF [11, 12] as well as indicating that pHFs are able to induce oral tolerance, whereas eHFs are not [13–15]. Nevertheless, Crittenden and Bennett [9] accentuate that no clinical studies confirm such animal studies. In general, clinical studies emphasize that eHFs should be the first choice for prevention of symptoms, as eHFs have been estimated to cause allergic reaction in approximately 5–10% of infants with CMA, while pHFs have been estimated to cause allergic reaction in between 33% and 50% of cow’s milk allergic infants [10, 16]. However, the degree of hydrolysis does not uniformly affect the allergenic potency in different cow’s milk allergic patients [17]. Further the residual allergenicity may in addition be influenced by the protein source [18]. A protein chemical characterization combined with an evaluation of residual immunogenicity and allergenicity may be of great value for the initial screening of newly developed cow’s milk-based hydrolysates for hypoallergenic infant formulas. Such assessment should at first rest on studies showing efficacy in animals and only then on carefully conducted clinical testing [19]. The aim of this study was to set up a preclinical series of analyses useful for a thorough evaluation of hydrolysed cow’s milk proteins for use in infant formulas in a Brown Norway (BN) rat model using intraperitoneal (i.p.) immunization. Two hydrolysates differing in the source material, a hydrolysed whey product suitable for use in eHF as well as a product based on hydrolysed purified BLG, were examined. These products were characterized for residual intact proteins, peptide size distribution profile, state of peptide aggregation and evaluated for (1) sensitizing capacity, (2) immunogenicity, (3) antibody-binding capacity and (4) eliciting capacity.
Materials and methods Cow’s milk protein products. A pilot batch of purified BLG and a batch of an extensively hydrolysed whey protein hydrolysate (PEPTIGEN IF-3080) suitable as a protein source in hypoallergenic infant formulas were kindly delivered by Arla Foods Ingredients (Videbæk, Denmark). From the purified BLG, a pilot batch of extensively hydrolysed BLG was made and delivered from Arla Foods Ingredients using the same hydrolysis protocol as the one used to make the extensively hydrolysed whey product. Preparation details of protein hydrolysates cannot be presented for proprietary reasons. Hydrolysed BLG and hydrolysed whey differed on the starting material, where hydrolysed BLG consisted of > 95% pure BLG, while hydrolysed whey was a mixture of proteins, consisting of approximately 50% BLG, 20% a-lactalbumin, 18% casein macro-peptides, 4% bovine serum albumin, 4%
Ó 2015 John Wiley & Sons Ltd
immunoglobulins and 4% of minor proteins. The endotoxin content was tested by Lonza endotoxin testing service (Verviers, Belgium) to be 75%) of the peptides were between 0.5 and 1.5 kDa, corresponding to peptides constituted of around 4–14 amino acid residues. For peptide aggregation profiles, GPC was performed and revealed that the peptides did aggregate to larger complexes, where > 40% of the peptides in hydrolysed BLG aggregated into complexes of up to 25 kDa, and > 55% of the peptides in hydrolysed whey aggregated into complexes of up to 20 kDa (Fig. 3). Sensitizing capacity of cow’s milk protein hydrolysates
We have previously shown that gastroduodenal digest from the major peanut allergens Ara h 1, having similar peptide size distribution profile as well as aggregation profile as the described hydrolysates, retained the capacity to sensitize the BN rats, using the same dosing regimen [23]. Therefore, hydrolysates containing only small peptides may indeed retain the sensitizing capacity. Sera from individual BN rats immunized with either PBS (control), 200 lg of intact BLG, 200 lg of hydrolysed BLG or with Ó 2015 John Wiley & Sons Ltd
C
Figure 1 Analytical RP-HPLC for detection of intact proteins. Comparison of chromatography profiles for intact BLG (A), hydrolysed BLG (B) and hydrolysed whey (C). Profiles are shown with absorbance at 280 and 220 nm.
200 lg of hydrolysed whey either without (animal experiment 1) or with (animal experiment 2) the use of Al(OH)3 as adjuvant were analysed for specific antibody responses. For all groups of rats, IgG1, IgG2a and IgE specificities were examined against intact BLG, hydrolysed BLG as well as hydrolysed whey. Specific IgG1 and IgG2a results were more or less identical, for which reason only specific IgG1 is shown. The results showed that while intact BLG could induce specific IgG antibodies, both hydrolysed BLG and whey had no immunogenicity without the use of adjuvant (Fig. 4A). While use of adjuvant had no effect on the immunogenicity of the hydrolysed BLG, a small effect was shown on the immunogenicity of hydrolysed whey, where three rats now reacted with a specific IgG response, although it was only with a low titre. The use of adjuvant affected the immunogenicity of intact BLG, where not only a slight higher BLG-specific IgG response (~1 titre (Log2) value, corresponding to the double amount of specific IgG antibodies), but also a lower heterogeneity between
278 Characterization of Milk Hydrolysates K. L. Bøgh et al. .................................................................................................................................................................. A
A
B
B
C
Figure 2 Peptide size distribution profiles. Size frequency of the peptides constituting hydrolysed BLG (A) and hydrolysed whey (B), divided into intervals of 0.5 kDa.
individually rats were shown (Fig. 4B). From Fig. 4C,D, it can be seen that intact BLG had sensitizing capacity, being able to induce specific IgE antibodies, while the hydrolysates had no sensitizing capacity. The use of adjuvant resulted in a higher BLG-specific IgE response (~2 titre (Log2) value, corresponding to a four times higher specific IgE antibody level), but no apparent reduction in the heterogeneity between the individual rats. Antibody-binding capacity of cow’s milk protein hydrolysates
Even though hydrolysates may hold no sensitizing capacity, they may still retain the capacity to bind antibodies in an already sensitized patient, for which reason the antibody-binding capacity may also be important to evaluate. A comparison of Fig. 4 A and B shows the effect of the adjuvant. While IgG1 raised in rats immunized with intact BLG without adjuvant could only react with the intact BLG, IgG1 from rats immunized with intact BLG with the use of adjuvant could in addition react with both the hydrolysed BLG and the hydrolysed whey and both to a statistically significant degree. The same could not be demonstrated for IgE antibodies. It is though worth noticing that detection of specific IgE by means of indirect ELISA could prove false negative, because of the competition with other antibody isotypes of same specificity.
Figure 3 GPC profiles for detection of aggregation status. Comparison of chromatography graphs, performed under physiological conditions, of intact BLG (A), hydrolysed BLG (B) and hydrolysed whey (C) with absorbance at 280 and 220 nm. Standard MW markers for absorbance at weigh length at 280 as well as 220 nm are shown across the top of each graph.
Inhibitory ELISA was performed with sera raised against intact BLG with the use of adjuvant, where intact BLG competed with either intact BLG, hydrolysed BLG or hydrolysed whey for antibody binding. The competitive test showed that even though the hydrolysed BLG and whey possess antibody-binding capacity, the hydrolysates could not compete with intact BLG for binding to the IgG1 antibodies, even in the presence of very high concentrations (Fig. 5). This could reflect a much higher avidity of the antibody binding to intact BLG than to the hydrolysates. Eliciting capacity of cow’s milk protein hydrolysates
Because of the low sensitivity of indirect IgE ELISA and the aggregation status of peptides in the hydrolysates, it could prove important also to test the allergenicity of
Scandinavian Journal of Immunology, 2015, 81, 274–283
K. L. Bøgh et al. Characterization of Milk Hydrolysates 279 ..................................................................................................................................................................
Figure 4 Specific IgG1 and IgE antibody responses in sensitization studies. Comparison of the intact BLG (squares), hydrolysed BLG (triangle) and hydrolysed whey (circles) specific IgG1 (A and B) or IgE (C and D) level in groups of rats immunized with either PBS (control), 200 lg of intact BLG, 200 lg of hydrolysed BLG or 200 lg of hydrolysed whey without (A and C) or with (B and D) adjuvant. The specific antibody levels are expressed as the Log2 titre values to the dilution giving the absorbance corresponding to the cut-off level. Each symbol represents an individual rat, and the horizontal bars indicate the median value in each group of rats. Statistically significant differences between the control group and the indicated group of rats are indicated with asterisk(s): *P ≤ 0.001, ***P ≤ 0.05. Notice the different Y-axes between IgG1 and IgE graphs.
A
C
B
D
Figure 5 Comparison of the inhibitory capacity of intact BLG, hydrolysed BLG and hydrolysed whey. Sera from individual rats dosed with intact BLG with the use of adjuvant were preincubated with 10-fold dilutions of intact BLG, hydrolysed BLG or hydrolysed whey. Results are expressed as the percentage inhibitory capacity, when competing with intact BLG coated on plated for binding to IgG1 antibodies. Error bars in the inhibition curves represent SD.
hydrolysates in a much more physiologic and sensitive in vivo test. Therefore, to analyse the eliciting capacity of the hydrolysates and compare it with that of intact BLG, skin tests were performed in rats sensitized to BLG with or without the use of adjuvant (animal experiment 3). All four concentrations (Fig. 6A) of BLG used in the skin test were
Ó 2015 John Wiley & Sons Ltd
able to induce an elicitation response in rats sensitized with BLG. No apparent differences were evident between rats dosed with or without adjuvant, for which reason only a single photo is shown (Fig. 6B). In contrast, none of the concentrations (Fig. 6A) of hydrolysed BLG and hydrolysed whey used in the skin test were able to elicit a response in rats dosed with or without adjuvant (Fig. 6C). That only intact BLG and none of the hydrolysates could induce positive skin reactions was further confirmed in the skin tests where each of the rats immunized either without or with adjuvant was injected intradermally with both the intact BLG, hydrolysed BLG and hydrolysed whey (Fig. 6D). Only intact BLG was able to induce a degranulation response resulting in a colour reaction significantly greater than the one induced by the negative control solution (PBS). No obvious differences were evident between individual rats, for which reason only a single photo is shown for rats dosed without use of adjuvant (Fig. 6E) and a single photo is shown for rats dosed with adjuvant (Fig. 6F). Collectively, the skin tests revealed that neither hydrolysed BLG nor hydrolysed whey could elicit a response by cross-linking of specific IgE of effector cells.
Discussion Infant formulas based on hydrolysed cow’s milk proteins are used for management as well as prevention of CMA. This calls for knowledge about the residual degree of
280 Characterization of Milk Hydrolysates K. L. Bøgh et al. .................................................................................................................................................................. A
D
B
E
C
F
Figure 6 Skin test of BLG-sensitized rats. The eliciting capacity of intact BLG, hydrolysed BLG and hydrolysed whey was studied in a skin test. Six rats sensitized to BLG (three without and three with adjuvant) were used for test of the optimal antigen concentration for intradermal injections (A–C). (A) shows the intradermal injection scheme. One rat from each dosing regime was injected with the indicated concentration of intact BLG, hydrolysed BLG or hydrolysed whey. Photograph (B) represents the eliciting response for rats, immunized without or with adjuvant, and intradermally injected with BLG, while photograph (C) represents the eliciting response for rats, immunized without or with adjuvant, and intradermally injected with hydrolysed BLG or whey. Twelve rats sensitized to BLG (six without and six with adjuvant) were used for skin tests (D–F). (D) shows the intradermal injection scheme. All rats were injected intradermally with intact BLG, hydrolysed BLG and hydrolysed whey. Photograph (E) represents the eliciting response for rats immunized without the use of adjuvant, while photograph (F) represents the eliciting response for rats immunized with the use of adjuvant. PBS was used as negative control, and the mast cell degranulation product 48/80 was used as positive control.
allergenicity as well as tolerogenicity of such infant formulas. Presently, several infant formulas based on hydrolysates are considered appropriate for infants and small children with a diagnosed CMA. To claim these infant formulas as hypoallergenic, an appropriate combination of preclinical in vivo and in vitro testing is necessary, where a significant reduction in immunogenicity and allergenicity shall be demonstrated [19, 24–27]. In contrast, no recommendation exists for declaration of infant formulas as tolerogenic. In the present study, we have carefully examined the protein chemical characteristics of two hydrolysates based on different source materials: a whey protein hydrolysate suitable as a protein source in hypoallergenic eHF as well as a BLG-based hydrolysate made with an identical procedure as the whey product. We have further made a thorough assessment of the: (1) sensitizing capacity, (2) immunogenicity, (3) antibody-binding capacity and (4) eliciting capacity in a Brown Norway rat model using i.p. immunization. An i.p. immunization model allows for the study of the inherent capacity of the cow’s milk-based
hydrolysates to stimulate the immune system, avoiding further modification caused by the acidic and proteolytic environment in the gastrointestinal tract and allows for the use of only small amounts. However, the i.p. model is less physiological relevant compared to an oral animal model. An evaluation of residual sensitizing capacity is required for claiming a hydrolysed cow’s milk-based infant formula as hypoallergenic [27]. For ethical reasons, sensitization needs to be studied in animal models. This study showed that neither the hydrolysed BLG nor the hydrolysed whey, both products having a peptide size profile corresponding to eHFs [8], holds immunogenicity and thereby was not able to induce specific IgG when rats were immunized without the use of adjuvant. Thus, as expected they did also not show any residual sensitizing capacity. When immunizing the rats with the use of Al(OH)3 as adjuvant, still the hydrolysates showed no sensitizing capacity; however, a few rats immunized with the hydrolysed whey now showed a low level of specific IgG1, indicating that the hydrolysed whey contained a small amount of immunogenicity when administered with adjuvant. As no
Scandinavian Journal of Immunology, 2015, 81, 274–283
K. L. Bøgh et al. Characterization of Milk Hydrolysates 281 ..................................................................................................................................................................
immunogenicity was evident for hydrolysed BLG, which contains a small proportion of peptides with larger sizes as well as a larger proportion of peptides aggregated into complexes of larger sizes, than the hydrolysed whey, the differences in the inherent immunogenicity of the two hydrolysates must lie in the differences in source materials. The present results are in line with other studies, all evaluating the sensitizing capacity of eHF with the use of adjuvant, likewise showing a reduced or abrogated capacity to sensitize. For example, in a study by van Esch et al. [28], an eHF based on whey showed immunogenicity but no sensitizing capacity in C3H/HeOuJ mice dosed by gavage with the use of cholera toxin as adjuvant. The specific IgG1 response was markedly reduced compared to the level induced by intact whey. A sensitization study with Sprague Dawley rats immunized i.p. with cow’s milk-based infant formulas based on either intact proteins, pHF or eHF with the use of Al(OH)3 as adjuvant showed that the pHF as well as eHF was able to induce specific IgE as well as IgG, though to a degree which were 100 and 10,000 times lower, respectively, than the level induced by the formula based on intact proteins [11]. In another study, BALB/c mice were immunized i.p. with different pHFs and eHFs with Al(OH)3 as adjuvant [12]. This study showed that all three pHFs were able to induce specific IgE antibodies, while only one of two eHFs was able to induce specific IgE antibodies. These hydrolysates could all induce specific IgG1 antibodies, though in various magnitude, signifying the very different degree of immunogenic potency [12]. Collectively, these studies showed that eHFs evaluated in animal models seem to have a markedly reduced sensitizing capacity compared to pHF and intact milk proteins, if any at all, depending on the exact product and immunization regime. Besides a direct measure of the allergenicity and immunogenicity, it is recommended that effort should be made to analyse the amount of residual immunological recognizable peptides [19, 24, 26, 27] and for claiming infant formulas as having reduced allergenicity, the amount of immune-reactive proteins should be of < 1% of total nitrogen-containing substances [29]. Immunochemical analyses may, in a semi-quantitative way, estimate the reactivity of the hydrolysates with preformed antibodies [25], where the primary idea is to monitor and verify the disruption of specific epitopes caused by the enzymatic hydrolysis. Our study showed that IgG1 raised in rats immunized with BLG without the use of adjuvant could not react with the hydrolysates, indicating that the hydrolysates contained no immune-reactivity. However, IgG1 raised in rats dosed with BLG with the use of Al (OH)3 as adjuvant reacted with both hydrolysates and to similar degree in a statistically significant way. A plausible explanation for the observed differences in immunereactivity of the hydrolysates between rats immunized with and without the use of Al(OH)3 as adjuvant lies in the way Al(OH)3 influences how BLG is seen by the immune
Ó 2015 John Wiley & Sons Ltd
system. In a very recent study investigating the reactivity of antibodies raised against a-lactablumin (ALA) or BLG either with or without the use of Al(OH)3, we found that antibodies raised with the use of Al(OH)3 reacted differently with ALA or BLG than antibodies raised without the use of Al(OH)3 [K.L. Bøgh, M.S. Andreasen, C.B. Madsen, paper in preparation]. The antibodies raised with Al(OH)3 had a stronger binding to the denatured protein compared to antibodies raised without Al(OH)3. This suggests that adsorption of proteins to Al(OH)3 changes the structure of the proteins, probably by some kind of unfolding of the protein, causing availability of more linear epitopes and less conformational epitopes. The consequence is that the use of Al(OH)3 may highly influence the outcome of results, as the adjuvant may significantly affect the way in which the protein is presented to the immune system and thereby influence which epitopes the antibody response is raised against. By choosing to use Al(OH)3 as adjuvant, the rats might be more proned to react with linear epitopes and thereby give the impression of a higher immune-reactivity. Inhibitory ELISA is an assay often used to evaluate the residual immune-reactivity of cow’s milk formulas based on hydrolysates. To assess the competitive capabilities of the hydrolysates with that of intact BLG, an inhibitory ELISA was performed with individual sera from rats immunized with BLG together with Al(OH)3 as adjuvant. Even though the hydrolysates contained a statistically significant antibody binding, shown by a simple ELISA, they showed no immune-reactivity in the inhibitory ELISA. A likely explanation could be a higher binding strength between antibodies and BLG compared to antibodies and hydrolysates, resulting in the lack of competitive capabilities of the peptides in the hydrolysates in spite of the peptides being in great excess compared to intact BLG. As pointed out by Leary [25], results from inhibitory ELISAs may be affected by the functional avidity of the specific antibodies. Nevertheless, this study stresses that cautions should be taken when using inhibitory ELISA as the only method to evaluate the residual immune-reactivity of hydrolysates, as a false-negative result may be the result. However, the functional and thereby clinical relevance of the immunereactivity of the hydrolysates in the present study may on the other hand be questioned. Peptides in both hydrolysates did aggregate to complexes of larger sizes. Aggregation of peptides has previously been suggested to be of great importance for the inherent allergenicity of breakdown products from food allergens and may confer allergenic properties to mixtures of peptides originally thought to be too small in sizes to inhere sensitizing as well as eliciting capacity [20, 23, 30]. Likewise aggregation of the 2.8 kDa large allergen melittin from bee venom has been suggested to provide the allergenic properties of this very small protein [31–33]. The complexes in the two hydrolysates were large enough to contain two or more IgE-binding epitopes and thereby
282 Characterization of Milk Hydrolysates K. L. Bøgh et al. .................................................................................................................................................................. potentially large enough to elicit an allergic reaction. The potential IgE-binding capacity of the aggregates and the fact that ELISA may give false-negative IgE results because of low sensitivity [23], led us to perform skin test in rats immunized with BLG with or without adjuvant. Both hydrolysates showed no eliciting capacity. This is in line with other animal studies evaluating the eliciting capacity of eHF. In an in vitro rat mast cell secretion assay, both pHF and eHF showed a strongly reduced capacity to trigger release of the mast cells sensitized with anti-BLG antibodies compared to intact BLG, where pHF was more potent than the eHF [11]. Likewise van Esch et al. [34] showed that eHF and pHF had a statistically significant reduced capacity to induce an acute allergic skin response in C3H/HeOuJ mice sensitized to whey compared to intact whey, and eHF showed to be even less effective than pHF. Further, Niggemann et al. [12] reported that both pHF and eHF could not elicit reactions in cow’s milk-sensitized BALB/c mice. In conclusion, this study demonstrated that the hydrolysates contained no allergenicity, as the hydrolysis showed successful in eliminating the sensitizing, IgE-binding and eliciting capacity. To evaluate the applicability of cow’s milk-based hydrolysates for management of CMA in preclinical settings, it is important to choose an appropriate combination of tests during both initial research and development. This is a responsibility of the manufacturers [25]. With this study, we present a possible approach for such evaluation, where only small amounts are needed, pointing out that the way in which the in vivo animal experiment is conducted and the choice of in vitro tests for evaluation of responses may greatly influence the results and thereby the interpretation of the usefulness of the hydrolysates for management of CMA in infants.
Acknowledgment The authors wish to thank Nehad Moradian, Lene Hansen, Anne Blicher, Anne Ørngreen, Eva Ferdinansen, Elisa Navntoft, Eigil Frank, Kenneth Worm and Maja Danielsen for their excellent assistance in the laboratory and animal facilities. A special thanks to Dr. Kirsten Pilegaard for fruitful discussions. These studies were financially supported by FOOD Denmark Research School, Danish Dairy Research Foundation and Technical University of Denmark.
Conflict of interest The authors declare that they have no conflict of interest.
References 1 Sampson HA. Update on food allergy. J Allergy Clin Immunol 2004;113:805–19.
2 Kattan JD, Cocco RR, Jarvinen KM. Milk and soy allergy. Pediatr Clin North Am 2011;58:407–26. 3 Skripak JM, Matsui EC, Mudd K, Wood RA. The natural history of IgE-mediated cow’s milk allergy. J Allergy Clin Immunol 2007;120:1172–7. 4 Host A, Halken S. A prospective study of cow milk allergy in Danish infants during the first 3 years of life. Clinical course in relation to clinical and immunological type of hypersensitivity reaction. Allergy 1990;45:587–96. 5 Wal JM. Cow’s milk allergens. Allergy 1998;53:1013–22. 6 Host A, Koletzko B, Dreborg S et al. Dietary products used in infants for treatment and prevention of food allergy. Joint Statement of the European Society for Paediatric Allergology and Clinical Immunology (ESPACI) Committee on Hypoallergenic Formulas and the European Society for Paediatric Gastroenterology, Hepatology and Nutrition (ESPGHAN) Committee on Nutrition. Arch Dis Child 1999;81:80–4. 7 Bindels JG, Boerma JA. Hydrolysed cow’s milk formulae. Pediatr Allergy Immunol 1994;5:189–90. 8 Von Berg A. Modified proteins in allergy prevention. Nestle Nutr Workshop Ser Pediatr Program 2009;64:239–47. 9 Crittenden RG, Bennett LE. Cow’s milk allergy: a complex disorder. J Am Coll Nutr 2005;24:582S–91S. 10 Bahna SL. Hypoallergenic formulas: optimal choices for treatment versus prevention. Ann Allergy Asthma Immunol 2008;101:453–9. 11 Fritsche R, Bonzon M. Determination of cow milk formula allergenicity in the rat model by in vitro mast cell triggering and in vivo IgE induction. Int Arch Allergy Appl Immunol 1990;93:289–93. 12 Niggemann B, Nies H, Renz H, Herz U, Wahn U. Sensitizing capacity and residual allergenicity of hydrolyzed cow’s milk formulae: results from a murine model. Int Arch Allergy Immunol 2001;125:316– 21. 13 Fritsche R. Utility of animal models for evaluating hypoallergenicity. Mol Nutr Food Res 2009;53:979–83. 14 Fritsche R, Pahud JJ, Pecquet S, Pfeifer A. Induction of systemic immunologic tolerance to beta-lactoglobulin by oral administration of a whey protein hydrolysate. J Allergy Clin Immunol 1997;100:266– 73. 15 van Esch BC, Schouten B, de Kivit S et al. Oral tolerance induction by partially hydrolyzed whey protein in mice is associated with enhanced numbers of Foxp3+ regulatory T-cells in the mesenteric lymph nodes. Pediatr Allergy Immunol 2011;22:820–6. 16 Ragno V, Giampietro PG, Bruno G, Businco L. Allergenicity of milk protein hydrolysate formulae in children with cow’s milk allergy. Eur J Pediatr 1993;152:760–2. 17 Meulenbroek LA, Oliveira S, den Hartog Jager CF et al. The degree of whey hydrolysis does not uniformly affect in vitro basophil and T cell responses of cow’s milk-allergic patients. Clin Exp Allergy 2014;44:529–39. 18 Wahn U, Wahl R, Rugo E. Comparison of the residual allergenic activity of six different hydrolyzed protein formulas. J Pediatr 1992;121:S80–4. 19 Kleinman RE, Bahna S, Powell GF, Sampson HA. Use of infant formulas in infants with cow milk allergy. Pediatr Allergy Immunol 1991;2:146–55. 20 Bøgh KL, Barkholt V, Rigby NM, Mills ENC, Madsen CB. Digested Ara h 1 loses sensitizing capacity when separated into fractions. J Agric Food Chem 2012;60:2934–42. 21 Bøgh KL, Barkholt V, Madsen CB. The sensitising capacity of intact beta-lactoglobulin is reduced by co-administration with digested beta-lactoglobulin. Int Arch Allergy Immunol 2012;161:21–36. 22 Untersmayr E, Scholl I, Swoboda I et al. Antacid medication inhibits digestion of dietary proteins and causes food allergy: a fish allergy model in Balb/c mice. J Allergy Clin Immunol 2003;112:616–23. 23 Bøgh KL, Kroghsbo S, Dahl L et al. Digested Ara h 1 has sensitizing capacity in Brown Norway rats. Clin Exp Allergy 2009;39:1611–21.
Scandinavian Journal of Immunology, 2015, 81, 274–283
K. L. Bøgh et al. Characterization of Milk Hydrolysates 283 .................................................................................................................................................................. 24 American Academy of Pediatrics. Committee on nutrition. Hypoallergenic infant formulas. Pediatrics 2000;106:346–9. 25 Leary HL Jr. Nonclinical testing of formulas containing hydrolyzed milk protein. J Pediatr 1992;121:S42–6. 26 Gortler I, Urbanek R, Forster J. Characterization of antigens and allergens in hypo-allergenic infant formulae. Eur J Pediatr 1995;154:289–94. 27 Cordle CT, Duska-McEwen G, Janas LM, Malone WT, Hirsch MA. Evaluation of the immunogenicity of protein hydrolysate formulas using laboratory animal hyperimmunization. Pediatr Allergy Immunol 1994;5:14–19. 28 van Esch BC, van Bilsen JH, Jeurink PV 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;220:95–102. 29 Commission Directive 96/4/EC of 16 February 1996 amending Directive 91/321/EEC on infant formulae and follow-on formulae. Official Journal of the European Communities 1996.
Ó 2015 John Wiley & Sons Ltd
30 Eiwegger T, Rigby NM, Mondoulet L et al. Gastro-duodenal digestion products of the major peanut allergen Ara h 1 retain an allergenic potential. Clin Exp Allergy 2006;36:1281–8. 31 King TP, Coscia MR, Kochoumian L. Structure-immunogenicity relationship of a peptide allergen, melittin. In: Kraft D, Sehon A, eds. Molecular Biology and Immunology of Allergens, Vol. 11. Ann Arbor: CRC Press, 1993:11–19. 32 King TP, Coscia MR, Kochoumian L. Structure-immunogenicity relationship of melittin and its N-terminal truncated analogs. Biochemistry 1993;32:3506–10. 33 King TP, Wade D, Coscia MR, Mitchell S, Kochoumian L, Merrifield B. Structure-immunogenicity relationship of melittin, its transposed analogues, and D-melittin. J Immunol 1994;153:1124–31. 34 van Esch BC, Schouten B, Hofman GA et al. Acute allergic skin response as a new tool to evaluate the allergenicity of whey hydrolysates in a mouse model of orally induced cow’s milk allergy. Pediatr Allergy Immunol 2010;21:e780–6.
Characterization of the Immunogenicity and Allergenicity of Two Cow’s Milk Hydrolysates – A Study in Brown Norway Rats K. L. Bøgh*, V. Barkholt† & C. B. Madsen*
Abstract *Division of Toxicology and Risk Assessment, National Food Institute, Technical University of Denmark, Søborg, Denmark; †Department of Systems Biology, Enzyme and Protein Chemistry, Technical University of Denmark, Kgs. Lyngby, Denmark
Received 21 August 2014; Accepted in revised form 23 December 2014 Correspondence to: K. L. Bøgh, Division of Toxicology and Risk Assessment, National Food Institute, Technical University of Denmark, Mørkhøj Bygade 19, DK-2860 Søborg, Denmark. E-mail: [email protected]
Hypoallergenic infant formulas based on hydrolysed milk proteins are used in the diet for cow’s milk allergic infants. For a preclinical evaluation of the immunogenicity and allergenicity of new protein ingredients for such hypoallergenic infant formulas as well as for the investigation of which characteristics of hydrolysates that contribute to allergenicity, in vivo models are valuable tools. In this study, we examine the immunogenicity and allergenicity of two hydrolysates in a Brown Norway (BN) rat model, using i.p. dosing, which allows for the use of small quantities. Intact BLG, hydrolysed BLG and a hydrolysed whey product suitable for use in extensively hydrolysed formulas were thoroughly characterized for protein chemical features and administered to BN rats by i.p. immunization with or without adjuvant. Sera were analysed for specific IgG and IgE for evaluation of sensitizing capacity, immunogenicity and antibody-binding capacity. For evaluation of eliciting capacity a skin test was performed. The study showed that the hydrolysates had no residual allergenicity, lacking the capacity to sensitize and elicit reactions in the BN rats. Dosing with or without adjuvant induced a large difference in immunogenicity. Only antibodies from rats sensitized to intact BLG with adjuvant were able to bind the hydrolysates, and the whey-based hydrolysate only showed immunogenicity when dosed with adjuvant. This study showed that hydrolysates can be evaluated by an i.p. animal model, but that the choice of in vitro tests used for evaluation of antibody responses may greatly influence the result as well as may the use of adjuvant.
Introduction Cow’s milk is the most common cause of food allergy in young children. Cow’s milk allergy (CMA) affects around 2.5% of the young children [1, 2], of whom 80–85% fortunately outgrow their allergy by the age of 3 years [3, 4]. Strict avoidance of cow’s milk is the management of choice; however, for infants and young children, hypoallergenic cow’s milk formulas based on hydrolysed whey or caseins are available. Both the casein and the whey fraction of cow’s milk hold proteins known to be allergenic. blactoglobulin (BLG), also designated Bos d 5, is the most abundant allergen in whey, where it constitutes approximately 50% of the protein content [5]. Cow’s milk-based infant formulas based on hydrolysates are designated either extensively hydrolysed formulas (eHFs) or partially hydrolysed formulas (pHFs), according to the degree of protein hydrolysis [6] and the molecular weight (MW) of the present peptides [7]. Although there is
274
no international consensus for the definition of eHF and pHF [6, 7], eHFs generally consist of peptides of which more than 95% have a MW of < 3 kDa, while pHFs generally contain peptides where a higher percentage have a MW between 3 and 10 kDa and a much lower percentage have a MW below the 3 kDa [8]. When eHFs are developed, the goal is to degrade proteins to such an extent that all potential allergenicity is lost. However, this may likely result in a concomitant loss of immunogenicity, preventing the immune system from developing tolerance to the milk proteins. In contrast, pHFs are designed to minimize the sensitizing capacity, though at the same time maintaining peptides with sizes large enough to be recognized by the immune system for induction of tolerance [9]. While eHFs are considered efficient for prevention of allergic reactions, in the management of the allergy in infants with a diagnosed CMA, pHFs are on the other hand suggested to be useful for the prevention of de novo sensitization, in the management of the allergy in
Ó 2015 John Wiley & Sons Ltd
K. L. Bøgh et al. Characterization of Milk Hydrolysates 275 ..................................................................................................................................................................
infants with a high risk for developing CMA [9, 10]. These considerations are supported by animal studies, showing a general lower allergenicity of eHF compared to pHF [11, 12] as well as indicating that pHFs are able to induce oral tolerance, whereas eHFs are not [13–15]. Nevertheless, Crittenden and Bennett [9] accentuate that no clinical studies confirm such animal studies. In general, clinical studies emphasize that eHFs should be the first choice for prevention of symptoms, as eHFs have been estimated to cause allergic reaction in approximately 5–10% of infants with CMA, while pHFs have been estimated to cause allergic reaction in between 33% and 50% of cow’s milk allergic infants [10, 16]. However, the degree of hydrolysis does not uniformly affect the allergenic potency in different cow’s milk allergic patients [17]. Further the residual allergenicity may in addition be influenced by the protein source [18]. A protein chemical characterization combined with an evaluation of residual immunogenicity and allergenicity may be of great value for the initial screening of newly developed cow’s milk-based hydrolysates for hypoallergenic infant formulas. Such assessment should at first rest on studies showing efficacy in animals and only then on carefully conducted clinical testing [19]. The aim of this study was to set up a preclinical series of analyses useful for a thorough evaluation of hydrolysed cow’s milk proteins for use in infant formulas in a Brown Norway (BN) rat model using intraperitoneal (i.p.) immunization. Two hydrolysates differing in the source material, a hydrolysed whey product suitable for use in eHF as well as a product based on hydrolysed purified BLG, were examined. These products were characterized for residual intact proteins, peptide size distribution profile, state of peptide aggregation and evaluated for (1) sensitizing capacity, (2) immunogenicity, (3) antibody-binding capacity and (4) eliciting capacity.
Materials and methods Cow’s milk protein products. A pilot batch of purified BLG and a batch of an extensively hydrolysed whey protein hydrolysate (PEPTIGEN IF-3080) suitable as a protein source in hypoallergenic infant formulas were kindly delivered by Arla Foods Ingredients (Videbæk, Denmark). From the purified BLG, a pilot batch of extensively hydrolysed BLG was made and delivered from Arla Foods Ingredients using the same hydrolysis protocol as the one used to make the extensively hydrolysed whey product. Preparation details of protein hydrolysates cannot be presented for proprietary reasons. Hydrolysed BLG and hydrolysed whey differed on the starting material, where hydrolysed BLG consisted of > 95% pure BLG, while hydrolysed whey was a mixture of proteins, consisting of approximately 50% BLG, 20% a-lactalbumin, 18% casein macro-peptides, 4% bovine serum albumin, 4%
Ó 2015 John Wiley & Sons Ltd
immunoglobulins and 4% of minor proteins. The endotoxin content was tested by Lonza endotoxin testing service (Verviers, Belgium) to be 75%) of the peptides were between 0.5 and 1.5 kDa, corresponding to peptides constituted of around 4–14 amino acid residues. For peptide aggregation profiles, GPC was performed and revealed that the peptides did aggregate to larger complexes, where > 40% of the peptides in hydrolysed BLG aggregated into complexes of up to 25 kDa, and > 55% of the peptides in hydrolysed whey aggregated into complexes of up to 20 kDa (Fig. 3). Sensitizing capacity of cow’s milk protein hydrolysates
We have previously shown that gastroduodenal digest from the major peanut allergens Ara h 1, having similar peptide size distribution profile as well as aggregation profile as the described hydrolysates, retained the capacity to sensitize the BN rats, using the same dosing regimen [23]. Therefore, hydrolysates containing only small peptides may indeed retain the sensitizing capacity. Sera from individual BN rats immunized with either PBS (control), 200 lg of intact BLG, 200 lg of hydrolysed BLG or with Ó 2015 John Wiley & Sons Ltd
C
Figure 1 Analytical RP-HPLC for detection of intact proteins. Comparison of chromatography profiles for intact BLG (A), hydrolysed BLG (B) and hydrolysed whey (C). Profiles are shown with absorbance at 280 and 220 nm.
200 lg of hydrolysed whey either without (animal experiment 1) or with (animal experiment 2) the use of Al(OH)3 as adjuvant were analysed for specific antibody responses. For all groups of rats, IgG1, IgG2a and IgE specificities were examined against intact BLG, hydrolysed BLG as well as hydrolysed whey. Specific IgG1 and IgG2a results were more or less identical, for which reason only specific IgG1 is shown. The results showed that while intact BLG could induce specific IgG antibodies, both hydrolysed BLG and whey had no immunogenicity without the use of adjuvant (Fig. 4A). While use of adjuvant had no effect on the immunogenicity of the hydrolysed BLG, a small effect was shown on the immunogenicity of hydrolysed whey, where three rats now reacted with a specific IgG response, although it was only with a low titre. The use of adjuvant affected the immunogenicity of intact BLG, where not only a slight higher BLG-specific IgG response (~1 titre (Log2) value, corresponding to the double amount of specific IgG antibodies), but also a lower heterogeneity between
278 Characterization of Milk Hydrolysates K. L. Bøgh et al. .................................................................................................................................................................. A
A
B
B
C
Figure 2 Peptide size distribution profiles. Size frequency of the peptides constituting hydrolysed BLG (A) and hydrolysed whey (B), divided into intervals of 0.5 kDa.
individually rats were shown (Fig. 4B). From Fig. 4C,D, it can be seen that intact BLG had sensitizing capacity, being able to induce specific IgE antibodies, while the hydrolysates had no sensitizing capacity. The use of adjuvant resulted in a higher BLG-specific IgE response (~2 titre (Log2) value, corresponding to a four times higher specific IgE antibody level), but no apparent reduction in the heterogeneity between the individual rats. Antibody-binding capacity of cow’s milk protein hydrolysates
Even though hydrolysates may hold no sensitizing capacity, they may still retain the capacity to bind antibodies in an already sensitized patient, for which reason the antibody-binding capacity may also be important to evaluate. A comparison of Fig. 4 A and B shows the effect of the adjuvant. While IgG1 raised in rats immunized with intact BLG without adjuvant could only react with the intact BLG, IgG1 from rats immunized with intact BLG with the use of adjuvant could in addition react with both the hydrolysed BLG and the hydrolysed whey and both to a statistically significant degree. The same could not be demonstrated for IgE antibodies. It is though worth noticing that detection of specific IgE by means of indirect ELISA could prove false negative, because of the competition with other antibody isotypes of same specificity.
Figure 3 GPC profiles for detection of aggregation status. Comparison of chromatography graphs, performed under physiological conditions, of intact BLG (A), hydrolysed BLG (B) and hydrolysed whey (C) with absorbance at 280 and 220 nm. Standard MW markers for absorbance at weigh length at 280 as well as 220 nm are shown across the top of each graph.
Inhibitory ELISA was performed with sera raised against intact BLG with the use of adjuvant, where intact BLG competed with either intact BLG, hydrolysed BLG or hydrolysed whey for antibody binding. The competitive test showed that even though the hydrolysed BLG and whey possess antibody-binding capacity, the hydrolysates could not compete with intact BLG for binding to the IgG1 antibodies, even in the presence of very high concentrations (Fig. 5). This could reflect a much higher avidity of the antibody binding to intact BLG than to the hydrolysates. Eliciting capacity of cow’s milk protein hydrolysates
Because of the low sensitivity of indirect IgE ELISA and the aggregation status of peptides in the hydrolysates, it could prove important also to test the allergenicity of
Scandinavian Journal of Immunology, 2015, 81, 274–283
K. L. Bøgh et al. Characterization of Milk Hydrolysates 279 ..................................................................................................................................................................
Figure 4 Specific IgG1 and IgE antibody responses in sensitization studies. Comparison of the intact BLG (squares), hydrolysed BLG (triangle) and hydrolysed whey (circles) specific IgG1 (A and B) or IgE (C and D) level in groups of rats immunized with either PBS (control), 200 lg of intact BLG, 200 lg of hydrolysed BLG or 200 lg of hydrolysed whey without (A and C) or with (B and D) adjuvant. The specific antibody levels are expressed as the Log2 titre values to the dilution giving the absorbance corresponding to the cut-off level. Each symbol represents an individual rat, and the horizontal bars indicate the median value in each group of rats. Statistically significant differences between the control group and the indicated group of rats are indicated with asterisk(s): *P ≤ 0.001, ***P ≤ 0.05. Notice the different Y-axes between IgG1 and IgE graphs.
A
C
B
D
Figure 5 Comparison of the inhibitory capacity of intact BLG, hydrolysed BLG and hydrolysed whey. Sera from individual rats dosed with intact BLG with the use of adjuvant were preincubated with 10-fold dilutions of intact BLG, hydrolysed BLG or hydrolysed whey. Results are expressed as the percentage inhibitory capacity, when competing with intact BLG coated on plated for binding to IgG1 antibodies. Error bars in the inhibition curves represent SD.
hydrolysates in a much more physiologic and sensitive in vivo test. Therefore, to analyse the eliciting capacity of the hydrolysates and compare it with that of intact BLG, skin tests were performed in rats sensitized to BLG with or without the use of adjuvant (animal experiment 3). All four concentrations (Fig. 6A) of BLG used in the skin test were
Ó 2015 John Wiley & Sons Ltd
able to induce an elicitation response in rats sensitized with BLG. No apparent differences were evident between rats dosed with or without adjuvant, for which reason only a single photo is shown (Fig. 6B). In contrast, none of the concentrations (Fig. 6A) of hydrolysed BLG and hydrolysed whey used in the skin test were able to elicit a response in rats dosed with or without adjuvant (Fig. 6C). That only intact BLG and none of the hydrolysates could induce positive skin reactions was further confirmed in the skin tests where each of the rats immunized either without or with adjuvant was injected intradermally with both the intact BLG, hydrolysed BLG and hydrolysed whey (Fig. 6D). Only intact BLG was able to induce a degranulation response resulting in a colour reaction significantly greater than the one induced by the negative control solution (PBS). No obvious differences were evident between individual rats, for which reason only a single photo is shown for rats dosed without use of adjuvant (Fig. 6E) and a single photo is shown for rats dosed with adjuvant (Fig. 6F). Collectively, the skin tests revealed that neither hydrolysed BLG nor hydrolysed whey could elicit a response by cross-linking of specific IgE of effector cells.
Discussion Infant formulas based on hydrolysed cow’s milk proteins are used for management as well as prevention of CMA. This calls for knowledge about the residual degree of
280 Characterization of Milk Hydrolysates K. L. Bøgh et al. .................................................................................................................................................................. A
D
B
E
C
F
Figure 6 Skin test of BLG-sensitized rats. The eliciting capacity of intact BLG, hydrolysed BLG and hydrolysed whey was studied in a skin test. Six rats sensitized to BLG (three without and three with adjuvant) were used for test of the optimal antigen concentration for intradermal injections (A–C). (A) shows the intradermal injection scheme. One rat from each dosing regime was injected with the indicated concentration of intact BLG, hydrolysed BLG or hydrolysed whey. Photograph (B) represents the eliciting response for rats, immunized without or with adjuvant, and intradermally injected with BLG, while photograph (C) represents the eliciting response for rats, immunized without or with adjuvant, and intradermally injected with hydrolysed BLG or whey. Twelve rats sensitized to BLG (six without and six with adjuvant) were used for skin tests (D–F). (D) shows the intradermal injection scheme. All rats were injected intradermally with intact BLG, hydrolysed BLG and hydrolysed whey. Photograph (E) represents the eliciting response for rats immunized without the use of adjuvant, while photograph (F) represents the eliciting response for rats immunized with the use of adjuvant. PBS was used as negative control, and the mast cell degranulation product 48/80 was used as positive control.
allergenicity as well as tolerogenicity of such infant formulas. Presently, several infant formulas based on hydrolysates are considered appropriate for infants and small children with a diagnosed CMA. To claim these infant formulas as hypoallergenic, an appropriate combination of preclinical in vivo and in vitro testing is necessary, where a significant reduction in immunogenicity and allergenicity shall be demonstrated [19, 24–27]. In contrast, no recommendation exists for declaration of infant formulas as tolerogenic. In the present study, we have carefully examined the protein chemical characteristics of two hydrolysates based on different source materials: a whey protein hydrolysate suitable as a protein source in hypoallergenic eHF as well as a BLG-based hydrolysate made with an identical procedure as the whey product. We have further made a thorough assessment of the: (1) sensitizing capacity, (2) immunogenicity, (3) antibody-binding capacity and (4) eliciting capacity in a Brown Norway rat model using i.p. immunization. An i.p. immunization model allows for the study of the inherent capacity of the cow’s milk-based
hydrolysates to stimulate the immune system, avoiding further modification caused by the acidic and proteolytic environment in the gastrointestinal tract and allows for the use of only small amounts. However, the i.p. model is less physiological relevant compared to an oral animal model. An evaluation of residual sensitizing capacity is required for claiming a hydrolysed cow’s milk-based infant formula as hypoallergenic [27]. For ethical reasons, sensitization needs to be studied in animal models. This study showed that neither the hydrolysed BLG nor the hydrolysed whey, both products having a peptide size profile corresponding to eHFs [8], holds immunogenicity and thereby was not able to induce specific IgG when rats were immunized without the use of adjuvant. Thus, as expected they did also not show any residual sensitizing capacity. When immunizing the rats with the use of Al(OH)3 as adjuvant, still the hydrolysates showed no sensitizing capacity; however, a few rats immunized with the hydrolysed whey now showed a low level of specific IgG1, indicating that the hydrolysed whey contained a small amount of immunogenicity when administered with adjuvant. As no
Scandinavian Journal of Immunology, 2015, 81, 274–283
K. L. Bøgh et al. Characterization of Milk Hydrolysates 281 ..................................................................................................................................................................
immunogenicity was evident for hydrolysed BLG, which contains a small proportion of peptides with larger sizes as well as a larger proportion of peptides aggregated into complexes of larger sizes, than the hydrolysed whey, the differences in the inherent immunogenicity of the two hydrolysates must lie in the differences in source materials. The present results are in line with other studies, all evaluating the sensitizing capacity of eHF with the use of adjuvant, likewise showing a reduced or abrogated capacity to sensitize. For example, in a study by van Esch et al. [28], an eHF based on whey showed immunogenicity but no sensitizing capacity in C3H/HeOuJ mice dosed by gavage with the use of cholera toxin as adjuvant. The specific IgG1 response was markedly reduced compared to the level induced by intact whey. A sensitization study with Sprague Dawley rats immunized i.p. with cow’s milk-based infant formulas based on either intact proteins, pHF or eHF with the use of Al(OH)3 as adjuvant showed that the pHF as well as eHF was able to induce specific IgE as well as IgG, though to a degree which were 100 and 10,000 times lower, respectively, than the level induced by the formula based on intact proteins [11]. In another study, BALB/c mice were immunized i.p. with different pHFs and eHFs with Al(OH)3 as adjuvant [12]. This study showed that all three pHFs were able to induce specific IgE antibodies, while only one of two eHFs was able to induce specific IgE antibodies. These hydrolysates could all induce specific IgG1 antibodies, though in various magnitude, signifying the very different degree of immunogenic potency [12]. Collectively, these studies showed that eHFs evaluated in animal models seem to have a markedly reduced sensitizing capacity compared to pHF and intact milk proteins, if any at all, depending on the exact product and immunization regime. Besides a direct measure of the allergenicity and immunogenicity, it is recommended that effort should be made to analyse the amount of residual immunological recognizable peptides [19, 24, 26, 27] and for claiming infant formulas as having reduced allergenicity, the amount of immune-reactive proteins should be of < 1% of total nitrogen-containing substances [29]. Immunochemical analyses may, in a semi-quantitative way, estimate the reactivity of the hydrolysates with preformed antibodies [25], where the primary idea is to monitor and verify the disruption of specific epitopes caused by the enzymatic hydrolysis. Our study showed that IgG1 raised in rats immunized with BLG without the use of adjuvant could not react with the hydrolysates, indicating that the hydrolysates contained no immune-reactivity. However, IgG1 raised in rats dosed with BLG with the use of Al (OH)3 as adjuvant reacted with both hydrolysates and to similar degree in a statistically significant way. A plausible explanation for the observed differences in immunereactivity of the hydrolysates between rats immunized with and without the use of Al(OH)3 as adjuvant lies in the way Al(OH)3 influences how BLG is seen by the immune
Ó 2015 John Wiley & Sons Ltd
system. In a very recent study investigating the reactivity of antibodies raised against a-lactablumin (ALA) or BLG either with or without the use of Al(OH)3, we found that antibodies raised with the use of Al(OH)3 reacted differently with ALA or BLG than antibodies raised without the use of Al(OH)3 [K.L. Bøgh, M.S. Andreasen, C.B. Madsen, paper in preparation]. The antibodies raised with Al(OH)3 had a stronger binding to the denatured protein compared to antibodies raised without Al(OH)3. This suggests that adsorption of proteins to Al(OH)3 changes the structure of the proteins, probably by some kind of unfolding of the protein, causing availability of more linear epitopes and less conformational epitopes. The consequence is that the use of Al(OH)3 may highly influence the outcome of results, as the adjuvant may significantly affect the way in which the protein is presented to the immune system and thereby influence which epitopes the antibody response is raised against. By choosing to use Al(OH)3 as adjuvant, the rats might be more proned to react with linear epitopes and thereby give the impression of a higher immune-reactivity. Inhibitory ELISA is an assay often used to evaluate the residual immune-reactivity of cow’s milk formulas based on hydrolysates. To assess the competitive capabilities of the hydrolysates with that of intact BLG, an inhibitory ELISA was performed with individual sera from rats immunized with BLG together with Al(OH)3 as adjuvant. Even though the hydrolysates contained a statistically significant antibody binding, shown by a simple ELISA, they showed no immune-reactivity in the inhibitory ELISA. A likely explanation could be a higher binding strength between antibodies and BLG compared to antibodies and hydrolysates, resulting in the lack of competitive capabilities of the peptides in the hydrolysates in spite of the peptides being in great excess compared to intact BLG. As pointed out by Leary [25], results from inhibitory ELISAs may be affected by the functional avidity of the specific antibodies. Nevertheless, this study stresses that cautions should be taken when using inhibitory ELISA as the only method to evaluate the residual immune-reactivity of hydrolysates, as a false-negative result may be the result. However, the functional and thereby clinical relevance of the immunereactivity of the hydrolysates in the present study may on the other hand be questioned. Peptides in both hydrolysates did aggregate to complexes of larger sizes. Aggregation of peptides has previously been suggested to be of great importance for the inherent allergenicity of breakdown products from food allergens and may confer allergenic properties to mixtures of peptides originally thought to be too small in sizes to inhere sensitizing as well as eliciting capacity [20, 23, 30]. Likewise aggregation of the 2.8 kDa large allergen melittin from bee venom has been suggested to provide the allergenic properties of this very small protein [31–33]. The complexes in the two hydrolysates were large enough to contain two or more IgE-binding epitopes and thereby
282 Characterization of Milk Hydrolysates K. L. Bøgh et al. .................................................................................................................................................................. potentially large enough to elicit an allergic reaction. The potential IgE-binding capacity of the aggregates and the fact that ELISA may give false-negative IgE results because of low sensitivity [23], led us to perform skin test in rats immunized with BLG with or without adjuvant. Both hydrolysates showed no eliciting capacity. This is in line with other animal studies evaluating the eliciting capacity of eHF. In an in vitro rat mast cell secretion assay, both pHF and eHF showed a strongly reduced capacity to trigger release of the mast cells sensitized with anti-BLG antibodies compared to intact BLG, where pHF was more potent than the eHF [11]. Likewise van Esch et al. [34] showed that eHF and pHF had a statistically significant reduced capacity to induce an acute allergic skin response in C3H/HeOuJ mice sensitized to whey compared to intact whey, and eHF showed to be even less effective than pHF. Further, Niggemann et al. [12] reported that both pHF and eHF could not elicit reactions in cow’s milk-sensitized BALB/c mice. In conclusion, this study demonstrated that the hydrolysates contained no allergenicity, as the hydrolysis showed successful in eliminating the sensitizing, IgE-binding and eliciting capacity. To evaluate the applicability of cow’s milk-based hydrolysates for management of CMA in preclinical settings, it is important to choose an appropriate combination of tests during both initial research and development. This is a responsibility of the manufacturers [25]. With this study, we present a possible approach for such evaluation, where only small amounts are needed, pointing out that the way in which the in vivo animal experiment is conducted and the choice of in vitro tests for evaluation of responses may greatly influence the results and thereby the interpretation of the usefulness of the hydrolysates for management of CMA in infants.
Acknowledgment The authors wish to thank Nehad Moradian, Lene Hansen, Anne Blicher, Anne Ørngreen, Eva Ferdinansen, Elisa Navntoft, Eigil Frank, Kenneth Worm and Maja Danielsen for their excellent assistance in the laboratory and animal facilities. A special thanks to Dr. Kirsten Pilegaard for fruitful discussions. These studies were financially supported by FOOD Denmark Research School, Danish Dairy Research Foundation and Technical University of Denmark.
Conflict of interest The authors declare that they have no conflict of interest.
References 1 Sampson HA. Update on food allergy. J Allergy Clin Immunol 2004;113:805–19.
2 Kattan JD, Cocco RR, Jarvinen KM. Milk and soy allergy. Pediatr Clin North Am 2011;58:407–26. 3 Skripak JM, Matsui EC, Mudd K, Wood RA. The natural history of IgE-mediated cow’s milk allergy. J Allergy Clin Immunol 2007;120:1172–7. 4 Host A, Halken S. A prospective study of cow milk allergy in Danish infants during the first 3 years of life. Clinical course in relation to clinical and immunological type of hypersensitivity reaction. Allergy 1990;45:587–96. 5 Wal JM. Cow’s milk allergens. Allergy 1998;53:1013–22. 6 Host A, Koletzko B, Dreborg S et al. Dietary products used in infants for treatment and prevention of food allergy. Joint Statement of the European Society for Paediatric Allergology and Clinical Immunology (ESPACI) Committee on Hypoallergenic Formulas and the European Society for Paediatric Gastroenterology, Hepatology and Nutrition (ESPGHAN) Committee on Nutrition. Arch Dis Child 1999;81:80–4. 7 Bindels JG, Boerma JA. Hydrolysed cow’s milk formulae. Pediatr Allergy Immunol 1994;5:189–90. 8 Von Berg A. Modified proteins in allergy prevention. Nestle Nutr Workshop Ser Pediatr Program 2009;64:239–47. 9 Crittenden RG, Bennett LE. Cow’s milk allergy: a complex disorder. J Am Coll Nutr 2005;24:582S–91S. 10 Bahna SL. Hypoallergenic formulas: optimal choices for treatment versus prevention. Ann Allergy Asthma Immunol 2008;101:453–9. 11 Fritsche R, Bonzon M. Determination of cow milk formula allergenicity in the rat model by in vitro mast cell triggering and in vivo IgE induction. Int Arch Allergy Appl Immunol 1990;93:289–93. 12 Niggemann B, Nies H, Renz H, Herz U, Wahn U. Sensitizing capacity and residual allergenicity of hydrolyzed cow’s milk formulae: results from a murine model. Int Arch Allergy Immunol 2001;125:316– 21. 13 Fritsche R. Utility of animal models for evaluating hypoallergenicity. Mol Nutr Food Res 2009;53:979–83. 14 Fritsche R, Pahud JJ, Pecquet S, Pfeifer A. Induction of systemic immunologic tolerance to beta-lactoglobulin by oral administration of a whey protein hydrolysate. J Allergy Clin Immunol 1997;100:266– 73. 15 van Esch BC, Schouten B, de Kivit S et al. Oral tolerance induction by partially hydrolyzed whey protein in mice is associated with enhanced numbers of Foxp3+ regulatory T-cells in the mesenteric lymph nodes. Pediatr Allergy Immunol 2011;22:820–6. 16 Ragno V, Giampietro PG, Bruno G, Businco L. Allergenicity of milk protein hydrolysate formulae in children with cow’s milk allergy. Eur J Pediatr 1993;152:760–2. 17 Meulenbroek LA, Oliveira S, den Hartog Jager CF et al. The degree of whey hydrolysis does not uniformly affect in vitro basophil and T cell responses of cow’s milk-allergic patients. Clin Exp Allergy 2014;44:529–39. 18 Wahn U, Wahl R, Rugo E. Comparison of the residual allergenic activity of six different hydrolyzed protein formulas. J Pediatr 1992;121:S80–4. 19 Kleinman RE, Bahna S, Powell GF, Sampson HA. Use of infant formulas in infants with cow milk allergy. Pediatr Allergy Immunol 1991;2:146–55. 20 Bøgh KL, Barkholt V, Rigby NM, Mills ENC, Madsen CB. Digested Ara h 1 loses sensitizing capacity when separated into fractions. J Agric Food Chem 2012;60:2934–42. 21 Bøgh KL, Barkholt V, Madsen CB. The sensitising capacity of intact beta-lactoglobulin is reduced by co-administration with digested beta-lactoglobulin. Int Arch Allergy Immunol 2012;161:21–36. 22 Untersmayr E, Scholl I, Swoboda I et al. Antacid medication inhibits digestion of dietary proteins and causes food allergy: a fish allergy model in Balb/c mice. J Allergy Clin Immunol 2003;112:616–23. 23 Bøgh KL, Kroghsbo S, Dahl L et al. Digested Ara h 1 has sensitizing capacity in Brown Norway rats. Clin Exp Allergy 2009;39:1611–21.
Scandinavian Journal of Immunology, 2015, 81, 274–283
K. L. Bøgh et al. Characterization of Milk Hydrolysates 283 .................................................................................................................................................................. 24 American Academy of Pediatrics. Committee on nutrition. Hypoallergenic infant formulas. Pediatrics 2000;106:346–9. 25 Leary HL Jr. Nonclinical testing of formulas containing hydrolyzed milk protein. J Pediatr 1992;121:S42–6. 26 Gortler I, Urbanek R, Forster J. Characterization of antigens and allergens in hypo-allergenic infant formulae. Eur J Pediatr 1995;154:289–94. 27 Cordle CT, Duska-McEwen G, Janas LM, Malone WT, Hirsch MA. Evaluation of the immunogenicity of protein hydrolysate formulas using laboratory animal hyperimmunization. Pediatr Allergy Immunol 1994;5:14–19. 28 van Esch BC, van Bilsen JH, Jeurink PV 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;220:95–102. 29 Commission Directive 96/4/EC of 16 February 1996 amending Directive 91/321/EEC on infant formulae and follow-on formulae. Official Journal of the European Communities 1996.
Ó 2015 John Wiley & Sons Ltd
30 Eiwegger T, Rigby NM, Mondoulet L et al. Gastro-duodenal digestion products of the major peanut allergen Ara h 1 retain an allergenic potential. Clin Exp Allergy 2006;36:1281–8. 31 King TP, Coscia MR, Kochoumian L. Structure-immunogenicity relationship of a peptide allergen, melittin. In: Kraft D, Sehon A, eds. Molecular Biology and Immunology of Allergens, Vol. 11. Ann Arbor: CRC Press, 1993:11–19. 32 King TP, Coscia MR, Kochoumian L. Structure-immunogenicity relationship of melittin and its N-terminal truncated analogs. Biochemistry 1993;32:3506–10. 33 King TP, Wade D, Coscia MR, Mitchell S, Kochoumian L, Merrifield B. Structure-immunogenicity relationship of melittin, its transposed analogues, and D-melittin. J Immunol 1994;153:1124–31. 34 van Esch BC, Schouten B, Hofman GA et al. Acute allergic skin response as a new tool to evaluate the allergenicity of whey hydrolysates in a mouse model of orally induced cow’s milk allergy. Pediatr Allergy Immunol 2010;21:e780–6.
