Ige and Igg Antibodies in Skin Allergy of the Horse Bettina Wagner, William Miller, Erin Morgan, Julia Hillegas, Hollis Erb, Wolfgang Leibold, Douglas Antczak

IgE and IgG antibodies in skin allergy of the horse Bettina Wagner, William Miller, Erin Morgan, Julia Hillegas, Hollis Erb, Wolfgang Leibold, Douglas Antczak

To cite this version:

Bettina Wagner, William Miller, Erin Morgan, Julia Hillegas, Hollis Erb, et al.. IgE and IgG antibodies in skin allergy of the horse. Veterinary Research, BioMed Central, 2006, 37 (6), pp.813-825. 10.1051/vetres:2006039. hal-00903061

HAL Id: hal-00903061 https://hal.archives-ouvertes.fr/hal-00903061 Submitted on 1 Jan 2006

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Vet. Res. 37 (2006) 813–825 813 c INRA, EDP Sciences, 2006 DOI: 10.1051/vetres:2006039 Original article

IgE and IgG antibodies in skin allergy of the horse

Bettina Wa*, William H. Mb,ErinE.Ma, Julia M. Ha, Hollis N. Ec,WolfgangLd, Douglas F. Aa

a Baker Institute for Animal Health, College of Veterinary Medicine, Cornell University, Ithaca, NY 14853, USA b Department of Clinical Sciences, College of Veterinary Medicine, Cornell University, Ithaca, NY 14853, USA c Department of Population Medicine and Diagnostic Sciences, College of Veterinary Medicine, Cornell University, Ithaca, NY 14853, USA d Institute of Immunology, University of Veterinary Medicine, 30173 Hannover, Germany

(Received 15 March 2006; accepted 5 June 2006)

Abstract – In horses, allergies have been characterized by clinical signs and/or intradermal (i.d.) allergen testing. Our aim was to find the first direct evidence that immunoglobulin E (IgE) mediates equine allergy. In addition, we tested the hypothesis that immediate skin reactions in horses can also be mediated by IgG. Anti-IgE affinity columns were used to purify IgE from serum of one healthy horse and three horses affected with summer eczema, an allergic dermatitis which is believed to be induced by Culicoides midges. A modified Prausnitz-Küstner experiment was performed in four clinical healthy horses by i.d. injection of the purified serum IgE antibodies. The following day, Culicoides allergen was injected at the same sites. Skin reactions were not observed in response to allergen alone, and in two horses after stimulation at any previous IgE injection site. However, the other two horses showed an immediate skin reaction at the previous injection sites of IgE obtained from allergic horses. In addition, purified monoclonal antibodies to various equine immunoglobulin isotypes were injected i.d. into six healthy horses. Immediate skin reactions were observed in response to anti-IgE (6/6 horses) and anti-IgG(T) injections (5/6 horses). The specificities of both antibodies for IgE and IgG(T), respectively, were confirmed by enzyme linked immunosorbent assays. The results provide the first direct evidence that IgE mediates classical Type-I allergy in horses and plays a major role in the pathogenesis of summer eczema. The data also suggest that IgG(T) can bind to skin mast cells and might contribute to clinical allergy.

horse / allergy / summer eczema / IgE / IgG

1. INTRODUCTION [19]. These include skin hypersensitiv- ities [4, 13, 27], urticaria [17], chronic obstructive pulmonary disease (recurrent Based on clinical signs and intrader- airway obstruction) [20, 24], and head- mal (i.d.) skin testing, several diseases shaking [17]. Epidemiological data about have been associated with Immunoglob- allergy in horses are rare. This most likely ulin E (IgE) mediated allergy in horses reﬂects the existing diﬃculties in pre- cisely diagnosing the presence of allergy. * Corresponding author: [email protected]

Article available at http://www.edpsciences.org/vetres or http://dx.doi.org/10.1051/vetres:2006039 814 B. Wagner et al.

Currently, the best allergy test available IgE antibodies. These antibodies are sub- is intradermal skin testing with allergen sequently bound by high-affinity FcεRI extracts. However, this procedure requires on cell membranes of mast cells and ba- expertise in technical performance and sophils, a process which is called “sen- interpretation of the results. Thus, it is sitization”. The IgE sensitized cells can mainly performed in some larger clinics be stimulated at each successive contact or university hospitals [14]. Because of with allergen, resulting in release of pre- economic, scheduling, or transportation is- formed inflammatory mediators, such as sues, allergies in the field are often diag- histamine, and induction of synthesis and nosed by clinical signs only. These horses release of leukotrienes, prostaglandins and generally do not contribute to statistical cytokines, all of which together initiate the analysis about the prevalence of allergy. inflammatory response and maintain the The most common skin hypersensitiv- production of allergen-specific IgE [11]. ity in horses is called “summer eczema” Although a role for IgE-mediated mast or “sweet itch”. This is a recurrent al- cell degranulation in summer eczema has lergic dermatitis that has been described been hypothesized by inducing immediate frequently in Icelandic horses, but it also skin reactions after intradermal Culicoides occurs in many other horse breeds, in- allergen exposure in affected horses [2, 4, cluding Thoroughbreds, Arabians, Warm- 13], the direct involvement of IgE has not bloods, Draft horses, Quarter horses and yet been demonstrated. Our aim was to ponies. Summer eczema has been found all provide the first formal evidence that IgE over the world and is believed to be in- mediates allergy in horses. We performed duced by allergens from the saliva of biting two experiments. First, we used a previ- midges of Culicoides spp. [2,4,8,9,13,15]. ously described method to purify IgE from The prevalence of the disease varies be- serum [30] and performed a Prausnitz- tween 3–72% [5,10,12]. Similar to human Küstner (P-K) experiment by transferring atopic dermatitis, summer eczema is not the allergic reaction to healthy animals life-threatening, but it induces high levels using IgE from horses affected with sum- of discomfort and massive scratching fol- mer eczema. Second, to determine the im- lowed by loss of hair and severe skin irrita- munoglobulin isotypes which can mediate tions. Current treatments include allergen mast cell degranulation in horses, we tested avoidance, anti-allergic medications, and monoclonal antibodies (mabs) to equine immunotherapy, also known as “hypo-” or IgE and IgG isotypes for their potential to “desensitization” [3]. The current therapies induce an immediate skin reaction in vivo. reduce clinical signs but do not cure the allergic disease. 2. MATERIALS AND METHODS IgE has a natural protective function as a defense against parasites. However, 2.1. Horses IgE can also mediate pathological reactions which manifest as allergy. IgE binds The recipient horses for the P-K ex- to high-affinity IgE-receptors (FcεRI) on periment and for intradermal (i.d.) injec- mast cells. This interaction between IgE tion of anti-equine IgE or IgG mabs were and the FcεRI plays a key role in allergic clinical healthy, adult horses (10–19 years inflammatory responses induced by mast of age) from the experimental herd at cell degranulation. Type-I hypersensitivi- the Baker Institute for Animal Health. ties are characterized by an initial contact Two geldings and four mares of different to the allergen, resulting in activation of breeds (Thoroughbred, Warmblood, Quar- B-cells and production of allergen-specific ter Horse) were used. The serum donors Skin allergy in horses 815 for the P-K experiment and the periph- run under non-reducing conditions. The eral blood donors for the histamine release proteins were either stained in the gel us- assay were adult female Icelandic horses ing Coomassie Brilliant Blue or blotted with or without clinical signs of summer to PVDF membranes (Bio-Rad, Hercules, eczema. All experimental procedures were CA, USA). The membranes were incu- approved by the Animal Care Committee bated with the following monoclonal anti- of Cornell University and were in accor- bodies to equine IgG: anti-IgGa (CVS45), dance with the guidelines established by anti-IgGb (CVS39), anti-IgGc (CVS52), the NIH. and anti-IgG(T) (CVS40) [16, 26]. The CVS antibodies were kindly provided by Dr P. Lunn (Colorado State University, 2.2. Purification of IgE and IgG from Fort Collins, CO, USA). serum

IgE and IgG were purified from equine 2.4. Prausnitz-Küstner (P-K) serum as previously described [30]. In experiment brief, the purification was performed using an ÄKTA FPLC instrument (Amersham Biosciences, Piscataway, NJ, USA) and The horses used as recipients in this ex- two affinity columns. First, a Protein G periment showed no clinical signs of skin affinity column (HiTrap Protein G, Amer- hypersensitivity over the past two sum- sham Biosciences, Piscataway, NJ, USA) mers, nor reaction to Culicoides allergen was used to deplete IgG from serum. Sec- when skin testing was performed. An area ond, the IgG depleted serum was passed at the lateral side of the neck was prepared over an anti-equine IgE affinity column for i.d. injection by clipping. A total of generated by coupling of anti-IgE134 mab 100 µL purified IgE (50 µg/mL) and IgG 1 to CNBr-activated sepharose 4B. Both the (300 µg/mL) were injected i.d. to four an- Protein G and anti-IgE columns were imals. The next day, allergen (Culicoides eluted using 0.1M glycine (pH 3.0) and extract; 1000 pnu/mL; Greer Laboratories, the eluates were subsequently neutralized Lenoir, NC, USA) was injected at the same in 1M Tris buffer (pH 8.0). All eluates were injection sites. Histamine (27.5 µg/mL) dialyzed against 0.9% (w/v) NaCl solution. The eluates obtained from the anti-IgE col- 1 Because of the believed role of IgE in umn were concentrated using Amicon Ul- allergic reactions, the IgG fractions were tra 15 concentrators with a 30kDa cut-off included as negative controls in the Prausnitz- (Millipore, Bedford, MA, USA). The puri- Küstner experiment. The purified IgG anti- fied IgG and IgE fractions were transferred bodies from adult horses represent the accu- passively to non-allergic horses in the P-K mulated immunological experience of these experiment. animals against a wide variety of antigens, in- cluding pathogens and vaccines. In contrast to IgG, the IgE antibody response is believed to be more restricted and is provoked mainly by 2.3. SDS-Page and western blotting allergens and also parasite infections. We used higher amounts of purified IgG than of IgE for SDS-Page and western blotting were intradermal injection to address the predicted performed as described [28]. In brief, variation of antibody specificities in these frac- 7.5% SDS gels were used to separate the tions. However, both purified fractions were serum fractions obtained during purifica- polyclonal and the concentrations of antigen- tion of IgE and IgG. All samples were specific antibodies were unknown. 816 B. Wagner et al. and saline served as positive and nega- clonal goat anti-horse IgG(H+L) in con- tive controls (both from Greer Laborato- centrations of 5 µg/mL. IgE was purified ries). The skin reactivity was read after from serum as described above and used 15–20 min. The reactions were visible as in concentrations of 400 ng/mL. IgG(T) wheals at the injection sites and were eval- was obtained from the serum flowthrough uated by inspection and palpation. A score after passing over the protein G and anti- (0–4) was recorded for the wheal sizes IgE columns and used in a dilution of at each i.d. injection site and photographs 1:200. The biotinylated antibodies anti- of the reactions were taken. The score of IgE134 or CVS40 were used for detection. 0 was always given for the saline con- Afterwards, streptavidin conjugated perox- trol and the score of 4 was applied to the idase and substrate solution were added to wheal of each histamine control. The re- the plates. All buffers, substrate solution, maining reactions were rated according to incubation steps and the measurement of the histamine control for each recipient the plates were used as described previ- horse. The evaluation of skin reactivity and ously [30]. The IgG(T) fraction was tested the scoring were performed blindly by the in a different ELISA using all anti-IgG and same person. Additional controls of the re- anti-IgE antibodies described above. The actions were performed at 4 and 24 h after IgG(T) fraction contained high amounts of allergen injection. IgG(T) as detected by the CVS40 antibody. No IgGa, IgGc or IgE and only low amounts of IgGb were detectable in the 2.5. Monoclonal antibodies used IgG(T) fraction (data not shown). in intradermal skin testing

The following anti-equine immunoglob- 2.7. Histamine release assay ulin mabs were used in vivo for i.d. injection: anti-IgE134 and anti-IgE176 recog- Histamine release was performed in nizing different epitopes of the IgE heavy whole blood samples. Heparinized equine chain [30], and the IgG-specific mabs anti- peripheral blood was obtained from four IgGa (CVS45), anti-IgG(T) (CVS40), and Icelandic horses affected with summer anti-IgGb (CVS39). A total of 100 µLof eczema. The plasma was removed by two each antibody dilution (50 µg/mL) was washes with phosphate buffered saline used per injection. All monoclonal anti- (PBS) and the cells were resuspended in re- bodies were mouse IgG1 isotypes. lease buffer (119 mmol/L NaCl, 5 mmol/L KCl, 0.4 mmol/LMgSO4, 25 mmol/L Pipes acid, 56 mmol/L glucose, 40 mmol/L 2.6. Specificity testing of mabs (ELISA) NaOH; and freshly added before use: 1 mmol/LCaCl2). A total of 500 µLofre- The mabs to equine IgE (anti-IgE134 suspended cells were used per sample. One and anti-IgE176) and the anti-IgG(T) an- of the samples from each horse was boiled tibody (CVS40) were tested by ELISA for 5 min and used to determine the max- for their specificity for IgE and IgG(T), imal release. The remaining samples were respectively. The ELISA was set up in dif- incubated for 1 h at 37 ◦C in release buffer ferent ways to test whether these three only (spontaneous release), or after adding mabs specifically detected equine IgE and Culicoides extract (15 µg/mL), or anti- IgG(T) or cross-reacted with the other iso- IgE134 or anti-IgE176 (cell culture super- type. The coating of the plates was per- natants diluted 1:2). After incubation, the formed with anti-IgE176, CVS40 or poly- cells were put on ice for 20 min. Then, all Skin allergy in horses 817 samples were spun down (700 g,10min, (Fig. 1A) were tested by ELISA and west- 4 ◦C) and the supernatants were collected. ern blotting for their immunoglobulin iso- The radioimmunoassay to measure the his- type content. The IgE fractions contained tamine concentration in the supernatants IgE, small contaminations of IgM, but no was described previously in detail [29]. IgG (data not shown). Different antibod- The evaluation was performed by setting ies to IgG were used for detection of the the maximal release sample for each horse IgG fractions from donor horses by west- as 100%. The relative histamine content in ern blotting (Fig. 1B). This indicated that the other samples was calculated according the purified IgG fractions contained IgGa, to the corresponding maximal release. This IgGb and IgGc. In contrast, IgG(T) was not relative evaluation allowed the comparison detectable in the IgG fractions. of samples from different animals, inde- pendent of the number of basophils and the histamine content per cell in samples from 3.2. Intradermal transfer of an allergic different horses. reaction by IgE (Prausnitz-Küstner reaction)

2.8. Statistics To provide direct evidence that IgE mediates allergic responses in the horse, Scores were recorded for immediate IgE fractions from four donor horses were skin reactivity at each reaction site. Me- transferred into the skin of four clinical dian scores were calculated across horses healthy animals (recipients). In addition to for identical intradermal injections. The the absence of clinical signs, the four re- histamine release data (relative percent- cipients were not sensitized to Culicoides. ages) were presented as minimum, me- This was tested by i.d. injection of Culi- dian, and maximal, because for several coides extract alone. On day 1, the recip- antibodies the SD was large relative to ients obtained purified IgG and IgE from the mean. P-values were calculated by the donor horses by i.d. injection. On the Wilcoxon signed-rank tests. next day, an allergen (Culicoides extract) was applied at the same skin locations 3. RESULTS of the recipients. After 15–20 min, the skin reaction was evaluated and a score 3.1. Immunoglobulin isotype was applied relative to the saline (score 0) composition of purified serum and histamine (score 4) controls. Allergic IgG and IgE skin reactions were detected in two of the four healthy recipient horses after trans- IgE and IgG from the serum of four fer of purified IgE (Fig. 2). These animals Icelandic horses (donors) were purified. showed an immediate skin reaction after Three horses were affected with skin hy- allergen injection in 4 of 6 injection sites persensitivity (summer eczema) and re- of IgE from allergic horses (scores 2 or 3). acted to Culicoides extract. The fourth No skin reactions were observed at the in- horse was healthy and not sensitized to jection sites of IgE obtained from the con- Culicoides. The sera were passed over a trol horse (0/2). IgG fractions of the same Protein G column to deplete IgG anti- donor horses were injected as controls in bodies (Fig. 1A). The IgG-depleted serum this experiment. Only one weak reaction fraction 2 (Fig. 1A) was then applied to an (score 1) was observed at the IgG injection anti-equine IgE Sepharose column to pu- sites (1/8). As mentioned above, the puri- rify IgE. IgG and IgE fractions 3 and 4 fied IgG used in the transfer experiment 818 B. Wagner et al.

Figure 1. SDS-PAGE and western blotting of equine immunoglobulin fractions. IgG and IgE antibodies were purified from horse serum using two affinity columns, a Protein G column and an anti-equine IgE column. The fractions obtained by this procedure were separated in 7.5% SDS-gel under non-reducing conditions. (A) The Coomassie Blue stained gel shows the serum and purified fractions of one horse: (1) serum before purification; (2) flowthrough after Protein G and before anti-IgE affinity purification; (3) pH 3.0 eluate from Protein G column; (4) concentrated pH 3.0 eluate from anti-IgE column. (B) The IgG fractions (lane 3 in A) obtained from four horses were separated on 7.5% SDS-gels under non-reducing conditions and transferred to PVDF membranes by western blotting. Each lane on the individual blots represents the purified IgG fraction of one horse. The membranes were incubated with monoclonal antibodies to equine IgGa, IgGb, IgGc or IgG(T).

Figure 2. Passive transfer of an immediate skin reaction by intradermal (i.d.) injection of purified serum IgE from allergic horses. Purified IgG and IgE were obtained from three horses affected with summer eczema (SE+ #1, #2, #3) and one healthy animal (SE-). On day 1, IgE and IgG (composed of IgGa, IgGb and IgGc) were injected into the skin of two healthy recipient horses (A and B). The next day (day 2), Culicoides extract was injected i.d. at the same locations of the lateral side of the neck. Saline, histamine, or Culicoides extract without previous immunoglobulin injection served as controls. The intensity of the immediate skin reaction was determined 15–20 min after Culicoides allergen injection and in relation to the histamine positive (score 4) and the saline negative controls (score 0) of the same horse. Skin allergy in horses 819

Figure 3. Immediate skin reactions induced by intradermal injection of different anti-equine isotype antibodies. Saline served as a negative, histamine as a positive control. Two allergen preparations (Quack gras and Culicoides), and a polyclonal goat anti-horse IgG(H+L) antibody which induced no skin reaction in this horse, were injected as additional controls for skin reactivity. All anti-equine IgE and IgG reagents were monoclonal mouse IgG1 antibodies. The reaction was read 15 min after i.d. injection and graded from 0 (saline) to 4 (histamine). was composed of IgGa, IgGb and IgGc, obtained after i.d. injection of the two anti- but not IgG(T). Except for the histamine IgE antibodies. Although anti-IgE134 in- control, the other two recipient horses did duced a strong immediate skin reaction in not show any skin reaction regardless of all animals, anti-IgE176 did not. which antibodies were injected the day before. All skin reactions at responsive sites were reduced after 4 h and had completely 3.4. Specificity of the anti-IgE and anti- disappeared by 24 h. IgG(T) antibodies The specificity of the anti-IgE and anti- 3.3. Intradermal injection of anti- IgG(T) antibodies is crucial for the inter- equine IgE and anti-equine IgG(T) pretation of skin reactivity induced after antibodies provoked an immediate i.d. injection of these reagents. To con- skin reaction firm that no cross-reactivity occurred between the anti-IgE antibodies and IgG(T) To find more evidence that IgE is indeed or the anti-IgG(T) antibody and equine the major antibody causing allergic reac- IgE, various combinations of ELISA test- tions, antibodies to equine IgE and various ing were performed to detect purified IgE IgG were injected into the skin of six clin- or IgG(T) from horse serum. The dif- ically healthy horses. The immediate skin ferent combinations of coating and de- reaction of one horse in this experiment is tection antibodies in this test are shown shown in Figure 3. All horses showed an in Figure 4. The assay using anti-IgE176 immediate skin reaction after injection of for coating and anti-IgE134 for detection anti-IgE134, and 5 out of 6 animals reacted specifically detected equine IgE, but not to anti-IgG(T) application (Tab. I). In con- IgG(T). When anti-IgE and anti-IgG(T) trast, after anti-IgGa or anti-IgGb injection antibodies were used in one test, neither only very weak or negative reactions were IgE nor IgG(T) were recognized. The poly- observed. These results suggested that both clonal anti-horse IgG(H+L) antibody can IgE and IgG(T) are bound to skin mast react with both IgG(T) and IgE because cells and can cause an immediate allergic of its specificity for immunoglobulin light reaction in horses. Different results were chains. However, only IgG(T) could be 820 B. Wagner et al.

Table I. Individual skin reactivity scores of six horses 15 min after injection of monoclonal antibodies to horse IgE and IgG.

Horse Controls Monoclonal antibodies saline histamine anti-IgE134 anti-IgE176 anti-IgGa anti-IgGb anti-IgG(T) 10440003 20430002 30420020 40430003 50430003 60430002 No. of reactions 0/6 6/66/6 0/60/61/6 5/6 Median score 0 4 3 0 0 0 2.5

Figure 4. The specificity of monoclonal anti-equine IgE and anti-equine IgG(T) antibodies as determined by ELISA. Anti-IgE176, anti-IgG(T) or polyclonal goat anti-horse IgG(H+L) antibodies were used for coating. Purified equine IgE or IgG(T) was then applied as the test sample. For detection, biotinylated anti-IgE134 or anti-IgG(T) were used and streptavidin conjugated peroxidase was added in an additional step. The bars represent means and standard deviations of three mea- surements. detected by anti-IgG(T) in this test. The but anti-IgE176 did not provoke any re- ELISA data confirmed the specificity of action. An additional experiment was per- the anti-IgE and anti-IgG(T) antibodies for formed to explore the differences in the IgE and IgG(T), respectively. The assay induction of IgE-mediated degranulation also confirmed that both anti-IgE antibod- by these two reagents. The only cells in ies detect soluble IgE. peripheral blood containing histamine are basophils. Similar to mast cells, basophils 3.5. Histamine release from peripheral can bind IgE to their cell membranes via blood basophils using anti-IgE134 FcεRI. Crosslinking of the receptor bound and anti-IgE176 IgE subsequently induces histamine release. Peripheral blood was obtained from After i.d. injection, anti-IgE134 induced four horses with summer eczema and a his- an immediate skin reaction in all horses, tamine release assay was performed. The Skin allergy in horses 821

Table II. Relative histamine release (%) from peripheral blood basophils of four horses affected with summer eczema after stimulation with Culicoides extract, anti-IgE134 or anti-IgE176 antibodies.

Histamine releasea Horse #1 Horse #2 Horse #3 Horse #4 Median (min-max) Spontaneous 7355 5(3–7) Culicoides 82 68 83 50 75 (50–83) Anti-IgE134 57 43 54 51 52.5 (43–57) Anti-IgE176 2 4 3 11 3.5 (2–11)b a The maximal release of each sample was set as 100%. All other samples were described relative to the corresponding maximal release sample. b The one-sided p-value comparing the spontaneous release and the release induced by anti-IgE176 was 0.56. amount of free histamine was determined cause of typical, seasonal recurrent clinical in the supernatants using a competitive signs and immediate skin reactions in- radioimmuno assay. Maximal release of duced by intradermal testing with allergen histamine was obtained by boiling a sam- extracts [4, 13, 27]. By assuming analogy ple aliquot. The evaluation was performed to molecular mechanisms of human allergy setting the maximal release of each sam- and atopic disease, it has been concluded ple as 100% (Tab. II). The spontaneous that IgE-mediated mast cell degranulation release, the histamine release induced by plays a major role in the pathogenesis of Culicoides extract and that induced by equine summer eczema [3, 19, 33]. How- anti-IgE antibodies were calculated rela- ever, the direct impact of IgE-mediated tive to the corresponding maximal release mechanisms in the disease had not previ- samples for each horse. Histamine release ously been shown. We performed two in was induced after incubation of the sam- vivo experiments (1) to confirm the causal ples with Culicoides extract and also with role of allergen-specific IgE antibodies in anti-IgE134. In contrast, the release of his- the pathogenesis of equine summer eczema tamine induced by anti-IgE176 was not and (2) to address the question of whether different from the spontaneous release (p = other immunoglobulin isotypes, such as 0.56). The finding that anti-IgE134, but IgG, are also able to mediate skin mast cell not anti-IgE176, mediated degranulation of degranulation in horses. blood peripheral basophils was in accor- For many decades the classical test to dance with the results after i.d. injection of induce allergic reactions in non-allergic in- these two antibodies. dividuals has been the Prausnitz-Küstner reaction, first performed in 1921. Heinz Küstner was allergic to cooked fish and 4. DISCUSSION developed an immediate systemic reaction after exposure to it. In the original experi- The goal of this study was to investi- ment, the allergic reaction was transferred gate antibody mediated mechanisms in the to Carl Prausnitz by intradermal injection skin that contribute to mast cell degranu- of serum from Küstner, followed by injec- lation and clinical allergy in horses. Skin tion of a small quantity of boiled fish ex- diseases such as equine summer eczema tract. The fish extract provoked an immedi- are believed to be of allergic origin be- ate wheal-and-flare reaction at the injection 822 B. Wagner et al. site of Prausnitz’s skin [6]. We performed In addition to IgE, we tested whether a modified P-K experiment in horses by IgG antibodies might be involved in the transferring purified serum IgE or IgG pathogenesis of equine summer eczema. fractions, followed by challenge with al- Equine serum IgG antibodies recognized lergen extract. The experiment showed that different proteins from Culicoides extracts the allergic response can be transferred to by western blotting or proteins in the healthy non-allergic horses using IgE from saliva glands of fixed sections of Culi- affected horses and confirmed the impor- coides midges by immunohistochemistry tant role of IgE in the pathogenesis of [33]. Here, we used purified serum IgG equine summer eczema. from horses with summer eczema for pas- However, the allergic reaction could sive transfer. After Culicoides allergen only be transferred to two out of four challenge, immediate skin reactions were horses by IgE and not at each injection not observed in the recipients although the site of IgE from allergic horses. This could IgG concentration we used (300 µg/mL) have been caused by a dose-dependent ef- was 6-fold higher than that of purified IgE. fect and that we were working with the The only exception was a single weak skin minimal concentration of allergen-specific reaction (score 1) which was observed in IgE. Although some attempts have recently one recipient animal out of 16 total IgG been made, the inducing allergen of equine injection sites in four horses. The IgG frac- summer eczema is not yet defined [1] tions used in the P-K experiment were ob- and assays to quantify Culicoides-specific tained by Protein G purification. This pro- IgE antibodies in horse serum are cur- cedure selectively enriched certain serum rently not available. Out of the total volume IgG with high-affinity to Protein G, such of 100 µL that we used per i.d. injec- as IgGa, IgGb and IgGc [25]. Others like tion (50 µg/mL IgE), only a small frac- IgG(T) were not bound to the affinity col- tion of the IgE antibodies was expected to umn and were consequently not present in be allergen-specific. Compared to humans, the i.d. injected IgG fraction. The transfer horses have relatively high total serum IgE experiment using purified IgG suggested concentrations (in the µg/mL range) and that neither IgGa, IgGb nor IgGc can me- no differences in total serum IgE levels diate skin mast cell degranulation in equine were found between healthy horses and an- summer eczema. imals affected with summer eczema [30]. Injecting monoclonal antibodies to Most likely, the majority of IgE antibod- equine IgE and various IgG directly into ies obtained from serum were directed the skin tested whether the corresponding against other antigens, such as parasite Ig were already bound to skin mast cells antigens, and the concentration of Culi- via Fc-receptors. In other species, IgE coides-specific IgE in the injected fractions antibodies are bound to the high-affinity was rather small. Variations in parameters FcεRI [11]. In humans and rodents, IgG such as the amount of Culicoides-specific antibodies can also bind to various IgG IgE in the purified IgE preparations, the receptors (FcγRs) on mast cells and number of mast cells in the injection area activate or inhibit degranulation depending of the recipient horses, and the depth of on the receptor [7, 21, 22]. IgG binding to the individual injection sites might have canine mast cells via Fcγ-receptors has influenced the results. Thus, the amount also been reported [23]. In our experiments of Culicoides-specific IgE might just have using six healthy horses, an immediate been sufficient to induce an immediate skin skin reaction was induced in all horses by reaction in some but not all recipients after anti-IgE134 and in five out of six horses allergen challenge. by the anti-IgG(T) antibody CVS40. At Skin allergy in horses 823

15 min after i.d. injection, the immediate body, which is consistent with the intrader- skin reaction induced by anti-IgG(T) was mal skin testing result indicating degranu- also slightly lower (median score 2.5) lation from mast cells. than the reactivity induced by anti-IgE134 Previous studies demonstrated that (median score 3). This was similar to these two antibodies detect cell bound human mast cells that showed decreased IgE on peripheral blood cells equally IgG-dependent degranulation compared to by flow cytometric analysis. In contrast, degranulation mediated by IgE [32]. they recognized different epitopes on the Equine IgG(T) must have already been equine IgE heavy chain constant region bound to Fc-receptors on mast cells to by SDS-PAGE and western blotting. induce an immediate skin reaction after Anti-IgE134 detected a conformational anti-IgG(T) injection. The observation that epitope under non-reducing conditions skin mast cells were already sensitized by which was destroyed after treatment of IgE IgG(T) suggests that free IgG(T) antibod- with reducing substances. The IgE heavy ies can bind to high-affinity Fcγ-receptors chain epitope recognized by anti-IgE176 on equine skin mast cells and that these was still detectable under reducing con- Fcγ-receptors can activate cellular degran- ditions. This suggested that anti-IgE176 ulation. This is similar to the interaction recognizes a linear epitope [30]. of IgG and the high-affinity FcγRI on A key event in mast cell and basophil human mast cells [21] and differs from degranulation is the cross-linking of the binding of IgG to low affinity FcγRII and IgE/FcεRI complex. Cross-linking of the FcγRIII that requires prior IgG/allergen receptor via IgE antibodies initiates sig- complex formation [22]. Equine IgG(T) naling through the FcεRI and subsequently is not an isotype; rather, it represents a results in cellular degranulation [11]. Al- fraction of equine IgG composed of IgG3 though both antibodies can bind to FcεRI and IgG5 [31]. The anti-IgG(T) antibody bound IgE, we conclude that anti-IgE176 CVS40 recognizes both isotypes, although did not cross-link the IgE/FcεRI complex. it seems to have a higher affinity to IgG52. Consequently, degranulation was not in- We do not yet know whether skin mast cell duced by the anti-IgE176 antibody. Both sensitization and immediate skin reactivity anti-IgE antibodies are useful tools for was mediated by IgG3 or IgG5 or by both continuing investigations on the molecu- isotypes. lar mechanisms of IgE-mediated allergic After i.d. injection of the two anti-IgE diseases in the horse. The monoclonal anti- mabs, only anti-IgE134 induced an imme- IgE176 might also be a potential candidate diate skin reaction, but not anti-IgE176. for in vivo applications and treatment of The potential of both anti-IgE mabs to allergic conditions, because of its lack in induce degranulation via IgE bound to mediating degranulation in sensitized mast Fcε-receptors was tested using a histamine cells and basophils. release assay on a different cell popula- In summary, our experiments using tion, namely peripheral blood cells. This transfer of IgE from allergic animals and assay targeted basophils which also ex- intradermal injection of anti-IgE antibod- press the high-affinity FcεRI and are the ies provide strong evidence that IgE me- only cells in the peripheral blood contain- diates mast cell degranulation and allergy ing histamine [18]. Histamine release from in horses. The modified Prausnitz-Küstner peripheral blood basophils was induced by experiment also resulted in the first di- anti-IgE134, but not by anti-IgE176 anti- rect evidence that Culicoides-specific IgE antibodies contribute to the pathogenesis 2 Wagner B., unpublished observation. of equine summer eczema. Intradermal 824 B. Wagner et al.

injection of monoclonal antibodies indi- [10] Halldordsottir S., Larsen H.J., An epidemio- cate that IgG(T) can also activate skin mast logical study of summer eczema in Icelandic horses in Norway, Equine Vet. J. 23 (1991) cells and might have a potential role in al- 296–299. lergic conditions of the horse. [11] Kinet J.P., The high-affinity IgE receptor (FcεRI): from physiology to pathology, Annu. Rev. Immunol. 17 (1999) 931–972. ACKNOWLEDGEMENTS [12] Lange S., Hamann H., Deegen E., Ohnesorge B., Distl O., Investigation of the prevalence This work was supported by the Harry M. of summer eczema in Icelandic horses in Zweig Memorial Fund for Equine Research. northern Germany, Berl. Munch. Tierarztl. Wochenschr. 118 (2005) 481–489. [13] Larsen H.J., Bakke S.H., Mehl R., REFERENCES Intradermal challenge of Icelandic horses in Norway and Iceland with extracts of [1] Althaus H., Muller N., Busato A., Mellor Culicoides spp., Acta Vet. Scand. 29 (1988) P.S., Torsteinsdottir S., Marti E., Cloning and 311–314. sequencing of a cDNA expressing a riboso- [14] Lebis C., Bourdeau P., Marzin-Keller F., Culicoides nubeculosus mal P0 peptide from Intradermal skin tests in equine dermatol- (Diptera), Vet. Immunol. Immunopathol. 99 ogy: a study of 83 horses, Equine Vet. J. 34 (2004) 99–111. (2002) 666–671. [2] Anderson G.S., Belton P., Kleider N., [15] Littlewood J.D., Incidence of recurrent sea- Hypersensitivity of horses in British Columbia to extracts of native and ex- sonal pruritus (“sweet itch”) in British and otic species of Culicoides (Diptera: German shire horses, Vet. Res. 142 (1998) Ceratopogonidae), J. Med. Entomol. 30 66–67. (1993) 657–663. [16] Lunn D.P., Holmes M.A., Antczak D.F., [3] Anderson G.S., Belton P., Jahren E., Lange Agerwal N., Baker J., Bendali-Ahcene H., Kleider N., Immunotherapy trial for S., Blanchard-Channell M., Byrne K.M., horses in British Columbia with Culicoides Cannizzo K., Davis W., Hamilton M.J., (Diptera: Ceratopogonidae) hypersensitivity, Hannant D., Kondo T., Kydd J.H., Monier J. Med. Entomol. 33 (1996) 458–466. M.C., Moore P.F., O’Neil T., Schram B.R., Sheoran A., Stott J.L., Sugiura T., [4] Braverman Y., Preferred landing Vagnoni K.E., Report of the Second Equine sites of Culicoides species (Diptera: Leukocyte Antigen Workshop, Squaw val- Ceratopogonidae) on a horse in Israel and ley, California, July 1995, Vet. Immunol. its relevance to summer seasonal recurrent Immunopathol. 62 (1998) 101–143. dermatitis (sweet itch), Equine Vet. J. 20 (1988) 426–429. [17] Mansfield J.R., Burrell M.H., Curl V.J., Valler B.J., Treatment of equine allergic dis- [5] Brostrom H., Larsson A., Troedsson M., eases with allergy neutralization. A field Allergic dermatitis (sweet itch) of Icelandic study, J. Nutr. Environ. Med. 8 (1998) 329– horses in Sweden: an epidemiological study, 334. Equine Vet. J. 19 (1987) 229-236. [18] Marone G., Casolaro V., Patella V., Florio [6] Cohen S.G., Zelaya-Quesada M., Prausnitz G., Triggiani M., Molecular and cellular bi- and Küstner phenomenon: the P-K reaction, ology of mast cells and basophils, Int. Arch. J. Allergy Clin. Immunol. 114 (2004) 705– Allergy Immunol. 114 (1997) 207–217. 710. [19] Matthews A.G., Imlah P., McPherson E.A., [7] Daeron M., Fc receptor biology, Annu. Rev. A reagin-like antibody in horse serum: 1. Immunol. 15 (1997) 203–234. Occurrence and some biological properties, [8] Greiner E.C., Fadok V.A., Rabin E.B., Vet. Res. Commun. 6 (1983) 13–23. Equine Culicoides hypersensitivity in [20] McGorum B.C., Dixon P.M., Halliwell Florida: biting midges aspirated from R.E.W., Quantification of histamine in horses, Med. Vet. Entomol. 4 (1990) plasma and pulmonary fluids from horses 375–381. with chronic obstructive pulmonary disease, [9] Greiner E.C., Entomologic evaluation of in- before and after “natural (hay and straw) sect hypersensitivity in horses, Vet. Clin. challenges”, Vet. Immunol. Immunopathol. North Am. Equine Pract. 11 (1995) 29–41. 36 (1993) 223–237. Skin allergy in horses 825

[21] Okayama Y., Kirshenbaum A.S., Metcalfe [27] Ungar-Waron H., Braverman Y., Gluckman D.D., Expression of a functional high- A., Trainin Z., Immunogenicity and al- affinity IgG receptor, FcγRI, on human mast lergenicity of Culicoides imicola (Diptera: cells: up-regulation by IFN-γ, J. Immunol. Ceratopogonidae) extracts, J. Vet. Med. B 37 164 (2000) 4332–4339. (1990) 64–72. [22] Ravetch J.V., Bolland S., IgG Fc receptors, [28] Wagner B., Radbruch A., Richards C., Annu. Rev. Immunol. 19 (2001) 275–290. Leibold W., Monoclonal equine IgM and IgG immunoglobulins, Vet. Immunol. [23] Sato Y., Teshima R., Nakamura R., Takagi Immunopathol. 47 (1995) 1–12. K., Sasaki N., Sawada J., Kitani S., Canine [29] Wagner B., Siebenkotten G., Leibold W., mast cell activation via human IgG1 and Radbruch A., Expression of a 4-(Hydroxy- IgG4, Int. Arch. Allergy Immunol. 135 3-nitro-phenyl) acetyl (NP) specific equi- (2004) 154–160. murine IgE antibody that mediates histamine release in vitro and a Type-I skin reaction in [24] Schmallenbach K.H., Rahman I., Sasse vivo, Equine Vet. J. 34 (2002) 657–665. H.H.L., Dixon P.M., Halliwell R.E.W., McGorum B.C., Crameri R., Miller [30] Wagner B., Radbruch A., Rohwer J., Leibold W., Monoclonal anti-equine IgE antibodies H.R.P., Studies on pulmonary and sys- ff temic Aspergillus fumigatus-specific IgE with specificity for di erent epitopes on the and IgG antibodies in horses affected with immunoglobulin heavy chain of native IgE, chronic obstructive pulmonary disease Vet. Immunol. Immunopathol. 92 (2003) (COPD), Vet. Immunol. Immunopathol. 66 45–60. (1998) 245–256. [31] Wagner B., Immunoglobulins and immunoglobulin genes of the horse, Dev. [25] Sheoran A.S., Holmes M.A., Separation of Comp. Immunol. 30 (2006) 155–164. equine IgG subclasses (IgGa, IgGb and [32] Woolhiser M.R., Okayama Y., Gilfillan IgG(T)) using their different binding char- A.M., Metcalfe D.D., IgG-dependent acti- acteristics for staphylococcal protein A vation of human mast cells following up- and streptococcal protein G, Vet. Immunol. regulation of FcgammaRI by IFN-gamma, Immunopathol. 55 (1996) 33–43. Eur. J. Immunol. 31 (2001) 3298–3307. [26] Sheoran A.S., Lunn D.P., Holmes M.A., [33] Wilson A.D., Harwood L.J., Bjornsdottir S., Monoclonal antibodies to subclass- Marti E., Day M.J., Detection of IgG and specific antigenic determinants on IgE serum antibodies to Culicoides salivary equine immunoglobulin gamma chains gland antigens in horses with insect dermal and their characterization, Vet. Immunol. hypersensitivity (sweet itch), Equine Vet. J. Immunopathol. 62 (1998) 153–165. 33 (2001) 707–713.

To access this journal online: www.edpsciences.org/vetres