Measurement of canine IgE using the alpha chain of the human high affinity IgE receptor

Veterinary Immunology and Immunopathology 78 (2001) 349±355

Short communication

Measurement of canine IgE using the alpha chain of the human high af®nity IgE receptor K. Stedman, K. Lee, S. Hunter, B. Rivoire, C. McCall*, D. Wassom Heska Corporation, 1613 Prospect Parkway, Fort Collins, CO 80525, USA Received 29 December 2000; received in revised form 27 January 2001; accepted 28 January 2001

Abstract In vitro assays for allergen speci®c immunoglobulin E (IgE) are a convenient and reproducible alternative to intradermal skin testing in dogs. Such tests may be used to support a diagnosis of atopic dermatitis and to de®ne appropriate allergens for immunotherapy. Current in vitro assays rely upon monoclonal or polyclonal antibodies as IgE detection reagents. However, in sera where allergen-speci®c IgG occurs in great excess, any IgE:IgG cross-reactivity of the detection reagent may result in lowered assay speci®city. Therefore, we have developed an assay for canine IgE which uses a recombinant form of the extracellular part of the alpha chain of the human high af®nity IgE receptor (FceRIa). Biotinylated FceRIa shows no signi®cant binding to puri®ed canine IgG, and recognizes a heat labile antibody in serum, with a detection limit of 73±146 pg/ml. Comparison of assay signals using the labeled FceRIa and a highly speci®c anti-canine IgE monoclonal antibody (MAb) shows good agreement. The FceRIa is therefore a sensitive and speci®c alternative to polyclonal or monoclonal antibodies for canine serum IgE measurement. # 2001 Elsevier Science B.V. All rights reserved. Keywords: IgE receptor; Canine; Immunoglobulin E; MAb; ELISA

IgE is present in very low quantities in normal human serum (0±0.003 mg/ml), and its detection by in vitro methods is a technical challenge. Dogs are reported to have much higher levels of IgE than humans, presumably as a result of their exposure to parasite infestation, but the quantity of IgE in serum is still far exceeded by that of IgG (Hill et al., 1995). The reagents most commonly used to detect IgE in ELISA are polyclonal or monoclonal antibody preparations. While the best of these are very sensitive, they may demonstrate some, albeit low, cross-reactivity with IgG. In sera containing a *

Corresponding author. Tel.: ‡1-970-493-7272; fax: ‡1-970-472-1644. E-mail address: [email protected] (C. McCall). 0165-2427/01/$ ± see front matter # 2001 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 5 - 2 4 2 7 ( 0 1 ) 0 0 2 4 2 - 2

350

K. Stedman et al. / Veterinary Immunology and Immunopathology 78 (2001) 349±355

100±1000-fold excess of antigen-speci®c IgG, this cross-reactivity could theoretically result in lowered assay speci®city. In vivo, IgE binds to the FceRI on the surface of mast cells and basophils with very high af®nity (1010 Mÿ1), and the isotype speci®city of the receptor, as measured by competition experiments with excess quantities of IgG, is absolute. The cDNAs coding for FceRI alpha chains for human (Kinet et al., 1987), rat (Shimuzu et al., 1988), mouse (Ra et al., 1989), dog (Goitsuka et al., 1999), horse and sheep (McAleese et al., 2000) have been cloned and sequenced. The human receptor consists of four polypeptide chains Ð one alpha, one beta and two gamma, and the ability to bind IgE resides wholly in the alpha chain (Blank et al., 1991). The human receptor can bind murine IgE, although the reciprocal binding does not occur (Conrad et al., 1983). Human skin can be sensitized by passive transfer of canine serum containing antigen-speci®c IgE, suggesting that the human FceRI can also bind canine IgE, but again, the reciprocal does not hold true, i.e. human IgE cannot sensitize canine skin (Lowenthal et al., 1993). The in vitro correlation of this observation has also been demonstrated, i.e. human basophils may be sensitized with canine IgE and induced to degranulate by the addition of anti-IgE antibodies or appropriate allergen (Sainte-Laudy and Prost, 1996). We reasoned, therefore, that the isolated human FceRI alpha chain itself might be used as a reagent for the in vitro detection of canine IgE. We set out to develop an ELISA for canine allergen-speci®c IgE using a recombinant form of the extracellular domain of the alpha chain of the human high af®nity IgE receptor (FceR1a). The truncated form of the FceRIa chain, (extracellular amino acids 1±172), was produced in the baculovirus expression system using standard techniques. FceRIa was puri®ed from baculovirus-infected insect cell supernatant by Q-sepharose anion exchange chromatography, followed by gel ®ltration. The puri®ed FceRIa was biotinylated by periodate oxidation followed by coupling to biotin-hydrazide. The labeled receptor was then tested in a variety of ELISA formats, using either horseradish peroxidase or alkaline phosphatase as the readout enzyme system. Unbiotinylated FceRIa was coupled to sepharose beads to produce an af®nity column which was used as the ®nal step to purify IgE from a pool of sera obtained from ¯easensitized healthy laboratory beagles. Balb/c mice were immunized with this puri®ed IgE and a panel of monoclonal antibodies was made using standard techniques. MAbs were screened against puri®ed canine IgE and IgG by ELISA and western blot, and one MAb (H207) with high sensitivity and speci®city for IgE was selected for assay development. H207 was puri®ed by protein A af®nity chromatography and was labeled by coupling to biotin-NHS ester. Table 1 shows the binding of the biotinylated FceRIa to highly puri®ed canine IgE and canine IgG in ELISA. The assay was performed as follows: immulon II microtiter plates (Dynatech, Chantilly, VA) were coated with 100 ng/well of puri®ed IgE or 1 mg/well of puri®ed canine IgG (Jackson Immunoresearch Laboratories, West Grove, PA), covered and incubated overnight at 48C. The plates were then decanted and blocked with 200 ml/ well Assay Buffer (``Assay Buffer'': PBS ‡ 4% fetal calf serum ‡ 0:05% Tween-20) for 1 h at room temperature (RT). A four-cycle wash on an automated plate washer (Dynatech, Chantilly, VA) was performed before the addition of serial two-fold duplicate dilutions of biotinylated FceRIa (starting concentration of 0.13 ug/ml or 0.65 nmol) in

K. Stedman et al. / Veterinary Immunology and Immunopathology 78 (2001) 349±355

351

Table 1 Binding of biotinylated FceR1a to puri®ed canine IgE and IgG in ELISAa FceRIa (pmol/well)

IgE (100 ng/well)

IgG (1000 ng/well)

650 325 162 81 40 20 10 5

1376 1481 1811 1445 1139 947 506 276

110 64 50 46 31 22 19 14

a

Results are expressed as optical density of test sample minus optical density of buffer only background  1000. A value of 150 is considered positive.

100 ml/well of Assay Buffer, and incubation for 1 h at RT. After washing, 10 ng/well streptavidin-HRP (KPL Labs, Gaithersburg, MD) in 100 ml Assay Buffer was added and incubated for a ®nal 1 h at RT. Plates were washed and 100 ml/well of TMB peroxidase substrate (KPL #50-76-00) was added for 30 min. The reaction was stopped with 100 ml/ well of 1 M H3PO4, and optical density (OD) was determined at A450 on an automated ELISA plate reader. ODs were corrected for background and shown as integers  1000. FceRIa showed a dose dependent binding to the puri®ed canine IgE, but did not show signi®cant binding to puri®ed canine IgG, even at a 10-fold higher coating concentration. Canine IgE antibodies, like those of humans, suffer loss of skin sensitizing capacity if denatured by heating to 568C (Schwartzman et al., 1971). Table 2 shows the effects of heat inactivation of serum on the binding of the biotinylated FceRIa. The sera were obtained from six beagle dogs of a strain that had been selected to mount persistent high IgE responses if sensitized in the perinatal period by subcutaneous injection of allergens. Dogs 100, 200 and 300 were sensitized to Meadow fescue grass pollen extract, and mounted IgE responses to this and other closely related grass pollens. These dogs had no measurable IgE antibodies directed towards the dust mites D. farinae or D. pteronyssinus. Table 2 The effect of heat inactivation on the binding of FceR1a in ELISAa Allergen

Dog 100

Dog 200

Dog 300

Dog 500

Dog 600

Dog 700

ÿ

‡

ÿ

‡

ÿ

‡

ÿ

‡

ÿ

‡

ÿ

‡

0 0 0 0

1884 1461 0 0

7 0 0 0

1664 990 0 0

11 0 0 0

0 0 4042 3718

0 0 350 87

0 0 3881 1392

0 0 100 131

0 0 4042 2972

0 0 162 60

Fescue 669 Ryegrass 482 D. farinae 0 D. pteronyssinus 0

a Genetically high IgE responder beagles were immunized by repeated s.c. injection with either Meadow fescue extract (dogs 100, 200, 300) or D. farinae extract (dogs 500, 600, 700) (Center Laboratories, 10 mg protein per 1 mg alum) starting within 10 days of birth. Sera were assayed intact (ÿ) or heat inactivated (‡) in the FceRIa-based ELISA, using alkaline phosphatase as enzymatic readout. Results are expressed as EA units. A value of 150 is considered positive.

352

K. Stedman et al. / Veterinary Immunology and Immunopathology 78 (2001) 349±355

Dogs 500, 600 and 700 were sensitized to D. farinae extract, and mounted IgE responses to this and the related D. pteronyssinus mite, but not to the grasses. The assay conditions used for this experiment, and subsequent experiments reported herein, were essentially similar to those of the commercial test (ALLERCEPTTM Detection System) performed in Heska's veterinary diagnostic laboratories. This assay has thus been completely optimized and validated. Nunc microtiter plates were coated with D. farinae or Meadow fescue extract (Center Laboratories, Port Washington, NY) at 30 protein nitrogen units (PNU)/well and dried. Serum, either intact or heat inactivated for 4 h at 568C, was diluted to 1:30 in Tris-buffered saline containing 0.05% Tween-20 (TBST), added to the plate and incubated overnight at 4±88C. The plates were then washed once with TBST and 100 ml of appropriately diluted biotinylated FceR1a added for 2 h at 228C. The plates were washed four times with TBST and 100 ml of appropriately diluted streptavidin-alkaline phosphatase (Jackson Immunoresearch Laboratories, West Grove, PA), added for 1 h at 228C. After four washes in TBST, 100 ml of pNPP substrate (p-nitrophenyl phosphate Ð 0.35 nmol/L, Moss, Pasadena, MD) was added for 1 h at 228C. The assay was stopped by the addition of 50 ml of 20 mM L-cysteine and read at A405 on an automated ELISA plate reader. Background OD (wells with no serum added) was subtracted from experimental OD and the resulting values were normalized as follows. A four-point standard curve (expected curve) was established from average results of repeated experiments with a pooled positive reference serum standard. Results obtained with reference standards in any subsequent ELISA were used to construct an actual curve. The normalization factor was then calculated as the mean of ratios between expected and actual values for each of the four points on the standard curve. Normalized OD values were multiplied by 1000 and expressed as EA units. In all cases, heat treatment of the sera abolished or greatly diminished the binding of labeled FceRIa. Residual binding observed for sera from dogs 500, 600 and 700 might occur for a number of reasons. These animals had very high levels of IgE to house dust mite and some IgE might remain even after heat inactivation. The existence of two forms of canine IgE has been described (Peng et al., 1997), but both forms were reported to be heat labile. A short-term skin-sensitizing IgG (IgGd) has also been described in atopic dogs (Willemse et al., 1985) which is heat stable. Although the FceRIa does not bind to the puri®ed canine IgG preparation in the experiment described herein, we cannot formally rule out the possibility, that this IgG subclass is present in small quantities in the atopic beagles, and bound by FceRIa, but is absent from the commercial IgG preparation used. In order to measure the sensitivity of the FceRIa-based ELISA in a truly quantitative manner, we ®rst made puri®ed allergen-speci®c canine IgE from pooled ¯ea allergic beagle serum. This was achieved by ammonium sulfate precipitation followed by double af®nity puri®cation Ð ®rst on a monoclonal anti-IgE column (using H207) and then on a column coated with recombinant ¯ea allergen Cte f 1(McDermott et al., 2000). The puri®ed anti-Cte f 1 IgE was titrated into serum from a speci®c pathogen free (SPF) beagle, each `spiked' serum sample diluted to 1:30, and the FceRIa-based ELISA was performed, using alkaline phosphatase readout and wells coated with 100 ng rCte f 1. The results are shown in Fig. 1. A clear dose response curve was obtained, with OD signal

K. Stedman et al. / Veterinary Immunology and Immunopathology 78 (2001) 349±355

353

Fig. 1. Puri®ed ¯ea-allergen-speci®c IgE was titrated into serum from a speci®c pathogen-free dog, the samples were diluted to 1:30 and tested by FceRIa-based ELISA, using plates coated with 100 ng/well recombinant Cte f 1.

maximizing at an anti-Cte f 1 IgE concentration of 18,000 pg/ml. The limit of detection of allergen-speci®c IgE in this assay using a cut-off of 150 EA units, was between 73 and 146 pg/ml. The biotinylated FceRIa and the biotinylated MAb H207 were compared for their ability to detect anti-D. farinae IgE in a large number (1086) of sera from putatively atopic dogs, submitted to Heska's veterinary diagnostic laboratories during the period

Fig. 2. Anti-D. farinae IgE measured in 1086 canine sera using monoclonal antibody H207 or FceRIa-based ELISA.

354

K. Stedman et al. / Veterinary Immunology and Immunopathology 78 (2001) 349±355

March 1999±2000. The assay signal obtained with the FceRIa versus signal obtained with the monoclonal antibody is plotted in Fig. 2. For both reagents, the sera gave a wide range of ELISA values ranging from 0 to 4400 (the maximum observed in the assay). The signals obtained with these two IgE detection reagents correlated very well over the entire signal range of the assay (R2 value of 0.972). These data show that the FceRIa is a novel and exquisitely speci®c reagent for the measurement of canine IgE in ELISA. When used in conjunction with high quality allergen preparations, FceRIa forms the basis of a sensitive and speci®c test for IgE in canine serum (Bevier et al., 1997; McCall et al., 1997). FceRIa is also a physiologically relevant reagent in that its primary function in vivo is to bind IgE with very high speci®city to the key cells mediating allergic responses, namely mast cells and basophils. The biotinylated FceRIa does not bind to canine IgG. This means that allergen-speci®c IgE may be measured accurately in serum, even in individuals with great excesses of allergen-speci®c IgG. Acknowledgements The authors would like to acknowledge John Arguello and Sean Clarke for expert technical assistance, Carol Talkington Verser for critical reading of the manuscript and Nicole Hoem for administrative assistance. References Bevier, D., Mondesire, R., Rose, B., Wassom, D., 1997. FceRIa-based ELISA technology for in vitro determination of allergen-speci®c IgE in a population of intradermal skin-tested normal and atopic dogs (Supplement to compendium on continuing education for the practicing veterinarian). Clin. Adv. 3, 10±16. Blank, U., Ra, C., Kinet, J.-P., 1991. Characterization of truncated a chain products from human, rat and mouse immunoglobulin E. J. Biol. Chem. 266, 2639±2646. Conrad, D., Wingard, J., Ishizaka, T., 1983. The interaction of human IgE with the human basophil IgE receptor. J. Immunol. 130, 327±333. Goitsuka, R., Hayashi, N., Nagase, M., Sasaki, N., Ra, C., Tsujimoto, H., Hasegawa, A., 1999. Molecular cloning of cDNAs encoding dog high af®nity IgE receptor alpha, beta and gamma chains. Immunogenetics 49, 580±582. Hill, P., Moriello, K., DeBoer, D., 1995. Concentrations of total serum IgE, IgA and IgG in atopic and parasitized dogs. Vet. Immunol. Immunopathol. 44, 105±113. Kinet, J.-P., Metzger, H., Hakimi, J., Kochan, J., 1987. A cDNA presumptively coding for the alpha subunit of the receptor with high af®nity for immunoglobulin E. Biochemistry 26, 4605±4610. Lowenthal, M., Patterson, R., Harris, K., 1993. Passive transfer of IgE-mediated cutaneous reactivity in heterologous species. Ann. Allergy 71, 481±484. McAleese, S., Halliwell, R., Miller, H., 2000. Cloning and sequencing of the horse and sheep high-af®nity IgE receptor alpha chains. Immunogenetics 51, 878±881. McCall, C., Stedman, K., Penner, S., Hunter, S., Bevier, D., Rose, B., Mondesire, R., Wassom, D., 1997. FceRIabased measurement of anti-¯ea saliva IgE in dogs (Supplement to compendium of continuing education for the practicing veterinarian). Clin. Adv. 3, 24±28. McDermott, M., Weber, E., Hunter, S., Stedman, K., Best, E., Frank, G., Wang, R., Escudero, J., Kuner, J., McCall, C., 2000. Identi®cation, cloning and characterization of a major cat ¯ea salivary allergen. Mol. Immunol. 37, 361±375.

K. Stedman et al. / Veterinary Immunology and Immunopathology 78 (2001) 349±355

355

Peng, Z., Arthur, G., Rector, E., Kierek-Jaszczuk, D., Simons, F., Becker, A.B., 1997. Heterogeneity of polyclonal IgE characterized by differential charge, af®nity to protein A and antigenicity. J. Allergy Clin. Immunol. 100, 87±95. Ra, C., Jouvain, M.-H.E., Kinet, J.-P., 1989. Complete structure of the mouse mast cell receptor for IgE (FceR1) and surface expression of chimeric receptors (rat±mouse±human) on transfected cells. J. Biol. Chem. 264, 15323±15327. Sainte-Laudy, J., Prost, C., 1996. Binding of canine anaphylactic antibodies on human basophils: application to canine allergy diagnosis. Vet. Dermatol. 7, 185±191. Schwartzman, R., Rockey, J., Halliwell, R., 1971. Canine reaginic antibody. Characterization of the spontaneous anti-ragweed and anti-dinitrophenol reaginic antibodies of the atopic dog. Clin. Exp. Immunol. 9, 549±569. Shimuzu, A., Tepler, I., Benfy, P., Berenstein, E., Siraganian, R., Leder, P., 1988. Human and rat mast cell highaf®nity immunoglobulin E receptors: characterization of putative alpha chain gene products. In: Proceedings of the National Academy of Sciences of the USA, Vol. 85, pp. 1907±1911. Willemse, A., Noordzij, A., Rutten, V.P.M.G., Bernadina, W.E., 1985. Induction of non-IgE anaphylactic antibodies in dogs. Clin. Exp. Immunol. 59, 351±358.