Association Between IgE Response Against B e t v I, the Major Allergen of Birch Pollen, and HLA-DRB Alleles Gottfried F. Fischer, Winfried F. Pickl, Ingrid Fa6, Christof Ebner, Fatima Ferreira, Heimo Breiteneder, Elisabeth Vikoukal, Otto Scheiner, and Dietrich Kraft A B S T R A C T : The association of the human IgE response against Bet v I, the major allergen of birch pollen, and the H L A - D R and D Q phenotype was studied. Birch pollen allergic patients showed a typical case history, positive skin-prick test, and positive RAST with birch pollen extracts. They were divided into two groups. Group 1 (n = 37) consisted of in&viduals generating IgE antibodies that selectively reacted with Bet v L Their serum IgE did not react with minor allergens from birch pollen as tested by lmmunoblot analysis, nor did they show a response against allergens from a panel of grass and other tree pollen or perennial allergens from animals and fungi as determined by skin-prick test. Patients belonging to group II (n - 34) possessed IgE reacting with Bet v I plus one or more additional allergens. The control group consisted of 637 healthy blood donors. Comparison of the frequencies of RFLP-defined HLA-DR and DQ alleles in patients and the control group revealed that the distribution of DRB3 alleles in group I patients differed signifi-
cantly from that in the control group: A higher frequency of the DRw52a/c alleles in comparison to the control group (p ..... < 0.02) was observed. In addition, alleles defined by nucleotide sequences coding for the amino acid sequence tyrosine-phenylalanine-histidine at positions 30-32 of the/3 chain of DR molecules were found with a higher frequency in patient group I(pcorf < 0.02), too. These alleles comprise DRw52a/c and some DRB1 alleles. In patients of group II, no significant differences in the distribution of class II alleles in comparison with the control group were observed. The fact that the assooation with DRB alleles is observed only in patients with a selective allergen-specific IgE response could mean that, among the genes that contribute to disease susceptibility, the class lI genes play a more dominant role in these patients. In conclusion, certain alleles encoded in DRB3 and DRB 1 genes of the HLA complex might play a functional role in birch pollen allergy. Human Immunology 33, 2 5 9 - 2 6 5 (1992)
ABBREVIATIONS EF HPLC PAGE PCR RAST
etiologic fraction high performance liquid chromatography polyacrylamide gel electrophoresis polymerase chain reaction radioallergosorbent test
RFLP RIST RR SDS TMAC
restriction fragment length polymorphism radioimmunosorbent test relative risk sodium dodecyl sulfate tetra methyl ammonium chloride
INTRODUCTION M o s t p a t i e n t s ( 9 5 % ) suffering f r o m t y p e I allergy to birch p o l l e n elicit an IgE r e s p o n s e against the 1 7 - k D
From the Institute for Blood Group Serology (National Blood Group Reference Laboratory, WHO," and National Tzssue Typing Laboratory, Coum'd of Europe (G.F.F.: W.F.P.; I.F.); the Institute of General and Experimental Pathology (C.E.; F.F.," H.B.; O.S.; D.K.), University of Vienna; and the Allergy Clinic (E.V.), Vienna, Austria. Address reprint requests to Dr. Gottfried F. Fischer, who is now at the Laboratory of Molecular Biology, MRC, Hills Road, Cambridge CB2 2QH, UK. Received October 22, 1991; accepted December 27, 199•.
Human Immunology33, 259-265 (1992) © AmericanSocietyfor Histocompatibilityand lmmunogenetics, 1992
p r o t e i n Bet v I [1, 2]. Bet v I b e l o n g s to a family o f 10 I g E - b i n d i n g isoallergens o f Betula verrucosa [3] and was r e c e n t l y c l o n e d and s e q u e n c e d [4]. In s o m e birch p o l l e n allergic patients, the IgE response is selective for Bet v I, with no d e t e c t a b l e IgE a n t i b o d i e s specific for o t h e r allergens, for e x a m p l e , min o r birch p o l l e n allergens or o t h e r seasonal or p e r e n n i a l allergens. In the sera o f o t h e r patients, h o w e v e r , IgE a n t i b o d i e s with specificities to variable n u m b e r s o f allergens o t h e r than t r e e o r grass p o l l e n o r p e r e n n i a l allergens in a d d i t i o n to Bet v I can be d e t e c t e d [2]. 259 0198-8859/92/S5.00
260
For allergens from ragweed pollen (Amb a V and Ra5G ), rye-grass pollen (Lolp III), and house dust mite, it was demonstrated that certain class II alleles can serve as genetic markers for the human immune response [58]. In that respect only little is known for the Bet v I allergy [9]. Therefore we characterized polymorphic HLA-DR and DQ alleles by restriction fragment length polymorphism (RFLP) analysis in individuals with an IgE response specific for Bet v I and compared their frequencies with those of a control population.
PATIENTS A total of 71 allergic patients were chosen for this study. Determination of IgE serum levels was performed by the radioimmunosorbent test (RIST) (Pharmacia, Uppsala, Sweden). A typical case history indicated birch pollen allergy. The patients were tested by skin-prick test against a panel of allergens from tree and grass pollen (wheat, maize, Juncus conglomeratus, rye, Artemisia vulgaris, chrysanthemum, ragweed, Urtica vulgaris, Plantago lanceolata, sorrel, Alnus glutinosa, hazel, birch, Fagus, oak, ash, popolar, Salix alba, ulmus, and plane-tree) and perennial allergens (Dermatophagoides pteronyssinus, Dermatophagoides farinae, Penicillium notatum, Alternaria alt., Aspergillus fumigatus, Botyris cinerea, Candida albicans, Cladosporium herbarum, Mucor racemosus, Pullularia pull., dog, cat, rabbit, guinea pig, horse, cattle, pig, hamster, feathers from duck, goose, and budgerigar) (Alk, Copenhagen, Denmark). Skin-prick reactivity was confirmed by radioallergosorbent test (RAST) (classes 2.6-5; Pharmacia). The patients were subdivided into two groups: Group I consisted of 37 patients with a monovalent sensitization to birch pollen. Their serum IgE antibodies reacted selectively with Bet v I, the major allergen of birch pollen when tested in immunoblot analysis, and did not recognize any minor allergens. From the patients' history and results of skin-prick test, no evidence for IgE response against other allergens was detected. Group II consisted of 34 patients. Their serum IgE antibodies reacted with Bet v I and some of other allergens tested. The control group consisted of 637 healthy Austrian blood donors typed for their class II alleles by RFLP analysis [10].
MATERIALS A N D M E T H O D S RFLP T y p i n g DNA analysis followed conventional methods. Briefly, DNA was prepared from peripheral blood by the salting
G.F. Fischer et al.
out procedure [11]; 6/*g of DNA were digested with the restriction enzyme Taq I (Boehringer, Mannheim, Germany) according to the manufacturer's recommendations. Electrophoresis was performed in a 0.7~ agarose gel (Gibco-BRL). DR/3 probe pRTV1 [12], DQ8 probe pII-/3-1 [13], and DQ ~ probe pDCH1 [14] were kind gifts from Drs. J. Bidwell, L. Rask, and C. Auffray, respectively. Probes were labeled with 32pdCTP (Amersham International, Amersham, UK) by the random priming technique [15]. Hybridization was performed at 62°C in the presence of 0.75 M NaCI, 75 mM sodium citrate, 10 mM Tris-HC1 pH 8, 2 mM EDTA, 2% SDS, 10% Dextran sulfate (Pharmacia), and 10 tzg/ml tRNA (Boehringer-Mannheim). Washes at high stringency were performed in the presence of 30 mM NaCI and 3 mM sodium citrate at 62°C. Autoradiography was performed for time periods ranging from overnight to 1 week. Band patterns were assigned to DR and DQ types was performed according to the method described by Bidwell et al. [16]. For easy comparisons, we use also their terminology [16]. T y p i n g of Polymerase Chain Reaction (PCR)-Amplified Products with Sequence-Specific Oligonucleotides Genomic DNA (0.5/zg) was amplified by PCR by standard procedures [17] except that the pH of the PCR reaction buffer was 9.0. With the primer pair DRB3'-I and DRB5'-I, exons 2 of the DRB1, DRB3, DRB4, and DRB5 loci were amplified. The sequences of the primers were taken from Shinomiya et al. [18]. The sequences of oligonucleotide probes identifying DRB3*0101, DRB3*02, and DRB3*0301 have been described by Smrzka et al. [19]. The sequence of the oligonucleotide DRB2901 spanning the part of the DRB1 and DRB3 loci, which codes for amino acid positions 29-34 and includes the amino acid sequence tyrosine-phenylalanine-histidine (YFH) at positions 3032 is 5'-AGA TAC TTC CAT AAC CAG-3'. This sequence was deduced from published sequences [20]. Five percent of the amplification product was dot blotted and fixed on a nylon membrane (Gelman Sciences) by baking (80°C for 2 hours); 5 pM of probes were labeled with 32p-ATP (Amersham International) with 10 U T4-polynucleotide kinase (BoehringerMannheim). Hybridization was carried out at 56°C in the presence of 3 M tetramethyl-ammonium chloride (TMAC), 50 mM Tris pH 8.0, 2 mM EDTA, 0,5~: SDS, and 10 /xg/ml tRNA. Washes at high stringency were performed at 57°C in a buffer containing TMAC, Tris, SDS, and EDTA in the same concentrations as used for the hybridization buffer. Autoradiography was performed for 2 hours and overnight.
Birch Pollen Allergy Is Associated with HLA-DR
Statistical Analysis Disease association was calculated by comparing the distribution of all RFLP defined DRB, DQB, and D Q A alleles in the patient groups with that in the control population. The chi-square test was performed on 2 × 2 tables consisting of numbers of each allele present or absent in the patient or control group. The p value obtained by the chi-square test was multiplied by the number of comparisons n, where n = 33, to calculate the corrected probability p ...... [21].
Preparation of Birch Pollen Extracts This was exactly performed as described elsewhere [2].
Purification of Recombinant B e t v I Expression of recombinant nonfusion Bet v I in Escherichia coll. The recombinant nonfusion protein was expressed by using the p K K 223.3 expression vector (Pharmacia LKB Biotechnology) whose promotor is regulated by the lac repressor and can be induced by addition of isopropyl-3-D-thiogalactoside (IPTG) to the medium. Competent E. coliJM 105 cells transformed with pKK 223.3 containing the complete Bet v I c D N A [4] were streaked on LB-ampicillin plates. Individual colonies were picked and grown overnight; 1 ml of an overnight culture was added to 500 ml of LB medium containing 100 /*g/ml ampicillin. Cultures were grown until the OD600 reached 0.4. T o induce the synthesis of the recombinant nonfusion protein, I P T G was then added to a final concentration of 0.5 mM and the cultures were grown for 3.5 hours at 37°C. Crude cell extract. The cells of a 500-ml culture of E. toll were collected by centrifugation and resuspended in 1 5 - 2 0 ml of 50 mM T r i s - H C l buffer (pH 7.5) containing 220 mM NaCl. The cells were broken by freezing in liquid nitrogen and thawing at 37°C. This procedure was repeated twice. Following centrifugation at 30,000 g for 25 min, the supernatant was dialyzed with four changes of 25 mM i m i d a z o l e - H C l buffer, p H 7.4. Chromatofocusing. The crude cell extract was applied to a PBE-94 exchanger column (10-ml bed volume; Pharmacia AB) at room temperature. The column was washed with 25 mM imidazole-HC1 buffer (pH 7.4) and the protein eluted with 12.5 % (vol/vol) Polybuffer 74 titrated to p H 4.0 with HC1. High-performance liquid chromatography (HPLC). The pooled Bet v I fractions from the chromatofocusing step were subjected to HPLC employing a linear gradient of 2-propanol created within 50 min at room temperature
261
(C8, RP column; solvent A: 0.1% trifluoroacetic acid in water; solvent B: 90% 2-propanol/trifluoroacetic acid; gradient of 0 - 6 0 % solvent B; flow rate 1.0 ml/minute). Protein assay. Protein concentration of purified sample was determined by a micro-Kjeldahl method [22] using glycine as standard.
S D S - P o l y a c r y l a m i d e Gel E l e c t r o p h o r e s i s and
Protein Transfer SDS PAGE and protein transfer onto nitrocellulose were performed as previously described [2]. Briefly, 1 mg/ml gel of crude birch pollen extract or 800/,g/ml gel of purified recombinant Bet v I preparation were applied to 15% polyacrylamide gels. The separated proteins were subsequently transferred onto nitrocellulose by using a blotting apparatus (Hoefer, San Francisco, CA) at 150 mA for 6 hours at room temperature. IgE Blots IgE blots were performed essentially as described elsewhere [2]. Briefly, after the protein-blotting procedure, nitrocellulose membranes were cut into strips and blocked with a buffer containing 0.5% bovine serum albumin and 0.5% Tween 20. Thereafter, the strips were incubated with a fourfold dilution of patients' sera. The specific IgE was detected after incubation with 125 I-labeled rabbit anti-human IgE (Pharmacia) and autoradiography.
RESULTS Reaction of patients' IgE with pollen extracts from Betulla verrucosa. The serum IgE level was tested in 34 of the 37 group I patients and amounted to 82 + 66 IU/ml (mean -+ standard deviation). Normal levels range between 10 and 150 IU IgE/ml. Only three of the patients tested had levels above the normal range (160, 190, and 320 IU IgE/ml). Having shown a positive skin-prick test with Betula veruccosa extracts, which was confirmed by RAST, patients' sera were tested for their IgE-binding pattern to birch pollen proteins by immunoblot analysis. Figure 1 shows representative reaction patterns of the two patient groups. The serum IgE of an allergic individual representative for group I patients bound selectively to Bet v I (molecular mass, 17 kD) in birch pollen extract (lane C). N o minor allergens from the birch pollen extract are recognized. It showed identical IgE-binding characteristics to purified recombinant Bet v I (lane D). The case history and skin-prick testing showed no evidence for reactivity against any other allergens.
262
G.F. Fischer et al.
A
kDa 97.4
-
46.0
-
30.0
-
21.5
-
B
C
D
E
F
G
H
14.3-
IgE-binding pattern from two representative birch-pollen-allergic patients. Birch pollen extracts (lanes A, C, E, and G) or purified recombinant Bet v I (lanes B, D, F, and H) were separated on a polyacrylamide gel and blotted onto nitrocellulose. Specific serum IgE of a patient from group II (lanes A and B) and from group I (lanes C and D) is shown. Lanes E - H represent the negative control reactions where normal human serum pool (lanes E and F) or buffer (lanes G and H) were used as first step reagent. FIGURE 1
Beside reactivity with birch pollen allergens, patients of group II showed IgE reactivity against further allergens as revealed by case history, skin-prick test, RAST, and immunoblot analysis. In the case of a group II patient assayed in lane A, the serum IgE recognized in addition to Bet v I several o t h e r - - m i n o r - - p r o t e i n s in birch pollen extract. The serum of this patient also showed IgE antibody reactivity with recombinant Bet v I (lane B) proving its Bet v I specificity. Lanes E - H show the negative control reactions for the birch pollen extract (lanes E and G) and recombinant Bet v I preparation (lanes F and H) using a normal human serum pool and buffer instead of patients' sera.
other allergens were recognized by the serum IgE (=group II) (Table 1). Results obtained by RFLP indicating a DRw52 a/c phenotype were confirmed by typing with allele-specific oligonucleotides: All of the RFLP defined DRw52a/c subjects belonged to the DRB3*0101 type (data not shown). A n amino acid sequence shared by D R B 1 and D R B 3 alleles might contribute to disease susceptibility. The amino acid
sequence tyrosine-phenylalanine-histidine (YFH) is located in the second hypervariable region of DR molecules at position 3 0 - 3 2 . It is shared by certain alleles from the DRB3 (DRw52a/c) and the DRB1 locus (DR3, some DRw13, -w14) [20]. All of these alleles can be identified by RFLP analysis. Comparison of the number of patients; possessing alleles with the Y F H sequence in their D R B I or DRB3 loci, with the control group revealed that the distribution differences of these alleles remained significant. This association comprised 81% of patients, however (Table 1). To strengthen this theoretical consideration, patients were studied for their YFH positiveness by probing their DRB genes with a sequence specific oligonucleotide. Figure 2 shows the hybridization pattern of the 37 patients belonging to patients' group I, whose exons 2 of their functional DRB loci were amplified and blotted to nylon membranes. All individuals expected from their RFLP H L A - D R type to be YFH positive indeed showed a positive reaction with the oligonucleotide
TABLE 1
Association between an H L A D R w 5 2 a / c phenotype and IgE response against Bet v I. The D R and D Q alleles in birch
pollen allergic patients were characterized by RFLP analyses [16]. The distribution of DQA-, DQB-, and D R B l - e n c o d e d HLA class II molecules did not differ in birch pollen allergic patients and the control population (data not shown). The frequency of the DRB3 allele DRw52a/c in 37 patients showing a selective IgE response against Bet v I (=group I) differed significantly from that in the control population: This allele was found in 62% of patients, whereas it was present in only 33% of the control population (Table 1). The p .... was
cantly from that in the control group: A higher frequency of the DRw52a/c alleles in comparison to the control group (p ..... < 0.02) was observed. In addition, alleles defined by nucleotide sequences coding for the amino acid sequence tyrosine-phenylalanine-histidine at positions 30-32 of the/3 chain of DR molecules were found with a higher frequency in patient group I(pcorf < 0.02), too. These alleles comprise DRw52a/c and some DRB1 alleles. In patients of group II, no significant differences in the distribution of class II alleles in comparison with the control group were observed. The fact that the assooation with DRB alleles is observed only in patients with a selective allergen-specific IgE response could mean that, among the genes that contribute to disease susceptibility, the class lI genes play a more dominant role in these patients. In conclusion, certain alleles encoded in DRB3 and DRB 1 genes of the HLA complex might play a functional role in birch pollen allergy. Human Immunology 33, 2 5 9 - 2 6 5 (1992)
ABBREVIATIONS EF HPLC PAGE PCR RAST
etiologic fraction high performance liquid chromatography polyacrylamide gel electrophoresis polymerase chain reaction radioallergosorbent test
RFLP RIST RR SDS TMAC
restriction fragment length polymorphism radioimmunosorbent test relative risk sodium dodecyl sulfate tetra methyl ammonium chloride
INTRODUCTION M o s t p a t i e n t s ( 9 5 % ) suffering f r o m t y p e I allergy to birch p o l l e n elicit an IgE r e s p o n s e against the 1 7 - k D
From the Institute for Blood Group Serology (National Blood Group Reference Laboratory, WHO," and National Tzssue Typing Laboratory, Coum'd of Europe (G.F.F.: W.F.P.; I.F.); the Institute of General and Experimental Pathology (C.E.; F.F.," H.B.; O.S.; D.K.), University of Vienna; and the Allergy Clinic (E.V.), Vienna, Austria. Address reprint requests to Dr. Gottfried F. Fischer, who is now at the Laboratory of Molecular Biology, MRC, Hills Road, Cambridge CB2 2QH, UK. Received October 22, 1991; accepted December 27, 199•.
Human Immunology33, 259-265 (1992) © AmericanSocietyfor Histocompatibilityand lmmunogenetics, 1992
p r o t e i n Bet v I [1, 2]. Bet v I b e l o n g s to a family o f 10 I g E - b i n d i n g isoallergens o f Betula verrucosa [3] and was r e c e n t l y c l o n e d and s e q u e n c e d [4]. In s o m e birch p o l l e n allergic patients, the IgE response is selective for Bet v I, with no d e t e c t a b l e IgE a n t i b o d i e s specific for o t h e r allergens, for e x a m p l e , min o r birch p o l l e n allergens or o t h e r seasonal or p e r e n n i a l allergens. In the sera o f o t h e r patients, h o w e v e r , IgE a n t i b o d i e s with specificities to variable n u m b e r s o f allergens o t h e r than t r e e o r grass p o l l e n o r p e r e n n i a l allergens in a d d i t i o n to Bet v I can be d e t e c t e d [2]. 259 0198-8859/92/S5.00
260
For allergens from ragweed pollen (Amb a V and Ra5G ), rye-grass pollen (Lolp III), and house dust mite, it was demonstrated that certain class II alleles can serve as genetic markers for the human immune response [58]. In that respect only little is known for the Bet v I allergy [9]. Therefore we characterized polymorphic HLA-DR and DQ alleles by restriction fragment length polymorphism (RFLP) analysis in individuals with an IgE response specific for Bet v I and compared their frequencies with those of a control population.
PATIENTS A total of 71 allergic patients were chosen for this study. Determination of IgE serum levels was performed by the radioimmunosorbent test (RIST) (Pharmacia, Uppsala, Sweden). A typical case history indicated birch pollen allergy. The patients were tested by skin-prick test against a panel of allergens from tree and grass pollen (wheat, maize, Juncus conglomeratus, rye, Artemisia vulgaris, chrysanthemum, ragweed, Urtica vulgaris, Plantago lanceolata, sorrel, Alnus glutinosa, hazel, birch, Fagus, oak, ash, popolar, Salix alba, ulmus, and plane-tree) and perennial allergens (Dermatophagoides pteronyssinus, Dermatophagoides farinae, Penicillium notatum, Alternaria alt., Aspergillus fumigatus, Botyris cinerea, Candida albicans, Cladosporium herbarum, Mucor racemosus, Pullularia pull., dog, cat, rabbit, guinea pig, horse, cattle, pig, hamster, feathers from duck, goose, and budgerigar) (Alk, Copenhagen, Denmark). Skin-prick reactivity was confirmed by radioallergosorbent test (RAST) (classes 2.6-5; Pharmacia). The patients were subdivided into two groups: Group I consisted of 37 patients with a monovalent sensitization to birch pollen. Their serum IgE antibodies reacted selectively with Bet v I, the major allergen of birch pollen when tested in immunoblot analysis, and did not recognize any minor allergens. From the patients' history and results of skin-prick test, no evidence for IgE response against other allergens was detected. Group II consisted of 34 patients. Their serum IgE antibodies reacted with Bet v I and some of other allergens tested. The control group consisted of 637 healthy Austrian blood donors typed for their class II alleles by RFLP analysis [10].
MATERIALS A N D M E T H O D S RFLP T y p i n g DNA analysis followed conventional methods. Briefly, DNA was prepared from peripheral blood by the salting
G.F. Fischer et al.
out procedure [11]; 6/*g of DNA were digested with the restriction enzyme Taq I (Boehringer, Mannheim, Germany) according to the manufacturer's recommendations. Electrophoresis was performed in a 0.7~ agarose gel (Gibco-BRL). DR/3 probe pRTV1 [12], DQ8 probe pII-/3-1 [13], and DQ ~ probe pDCH1 [14] were kind gifts from Drs. J. Bidwell, L. Rask, and C. Auffray, respectively. Probes were labeled with 32pdCTP (Amersham International, Amersham, UK) by the random priming technique [15]. Hybridization was performed at 62°C in the presence of 0.75 M NaCI, 75 mM sodium citrate, 10 mM Tris-HC1 pH 8, 2 mM EDTA, 2% SDS, 10% Dextran sulfate (Pharmacia), and 10 tzg/ml tRNA (Boehringer-Mannheim). Washes at high stringency were performed in the presence of 30 mM NaCI and 3 mM sodium citrate at 62°C. Autoradiography was performed for time periods ranging from overnight to 1 week. Band patterns were assigned to DR and DQ types was performed according to the method described by Bidwell et al. [16]. For easy comparisons, we use also their terminology [16]. T y p i n g of Polymerase Chain Reaction (PCR)-Amplified Products with Sequence-Specific Oligonucleotides Genomic DNA (0.5/zg) was amplified by PCR by standard procedures [17] except that the pH of the PCR reaction buffer was 9.0. With the primer pair DRB3'-I and DRB5'-I, exons 2 of the DRB1, DRB3, DRB4, and DRB5 loci were amplified. The sequences of the primers were taken from Shinomiya et al. [18]. The sequences of oligonucleotide probes identifying DRB3*0101, DRB3*02, and DRB3*0301 have been described by Smrzka et al. [19]. The sequence of the oligonucleotide DRB2901 spanning the part of the DRB1 and DRB3 loci, which codes for amino acid positions 29-34 and includes the amino acid sequence tyrosine-phenylalanine-histidine (YFH) at positions 3032 is 5'-AGA TAC TTC CAT AAC CAG-3'. This sequence was deduced from published sequences [20]. Five percent of the amplification product was dot blotted and fixed on a nylon membrane (Gelman Sciences) by baking (80°C for 2 hours); 5 pM of probes were labeled with 32p-ATP (Amersham International) with 10 U T4-polynucleotide kinase (BoehringerMannheim). Hybridization was carried out at 56°C in the presence of 3 M tetramethyl-ammonium chloride (TMAC), 50 mM Tris pH 8.0, 2 mM EDTA, 0,5~: SDS, and 10 /xg/ml tRNA. Washes at high stringency were performed at 57°C in a buffer containing TMAC, Tris, SDS, and EDTA in the same concentrations as used for the hybridization buffer. Autoradiography was performed for 2 hours and overnight.
Birch Pollen Allergy Is Associated with HLA-DR
Statistical Analysis Disease association was calculated by comparing the distribution of all RFLP defined DRB, DQB, and D Q A alleles in the patient groups with that in the control population. The chi-square test was performed on 2 × 2 tables consisting of numbers of each allele present or absent in the patient or control group. The p value obtained by the chi-square test was multiplied by the number of comparisons n, where n = 33, to calculate the corrected probability p ...... [21].
Preparation of Birch Pollen Extracts This was exactly performed as described elsewhere [2].
Purification of Recombinant B e t v I Expression of recombinant nonfusion Bet v I in Escherichia coll. The recombinant nonfusion protein was expressed by using the p K K 223.3 expression vector (Pharmacia LKB Biotechnology) whose promotor is regulated by the lac repressor and can be induced by addition of isopropyl-3-D-thiogalactoside (IPTG) to the medium. Competent E. coliJM 105 cells transformed with pKK 223.3 containing the complete Bet v I c D N A [4] were streaked on LB-ampicillin plates. Individual colonies were picked and grown overnight; 1 ml of an overnight culture was added to 500 ml of LB medium containing 100 /*g/ml ampicillin. Cultures were grown until the OD600 reached 0.4. T o induce the synthesis of the recombinant nonfusion protein, I P T G was then added to a final concentration of 0.5 mM and the cultures were grown for 3.5 hours at 37°C. Crude cell extract. The cells of a 500-ml culture of E. toll were collected by centrifugation and resuspended in 1 5 - 2 0 ml of 50 mM T r i s - H C l buffer (pH 7.5) containing 220 mM NaCl. The cells were broken by freezing in liquid nitrogen and thawing at 37°C. This procedure was repeated twice. Following centrifugation at 30,000 g for 25 min, the supernatant was dialyzed with four changes of 25 mM i m i d a z o l e - H C l buffer, p H 7.4. Chromatofocusing. The crude cell extract was applied to a PBE-94 exchanger column (10-ml bed volume; Pharmacia AB) at room temperature. The column was washed with 25 mM imidazole-HC1 buffer (pH 7.4) and the protein eluted with 12.5 % (vol/vol) Polybuffer 74 titrated to p H 4.0 with HC1. High-performance liquid chromatography (HPLC). The pooled Bet v I fractions from the chromatofocusing step were subjected to HPLC employing a linear gradient of 2-propanol created within 50 min at room temperature
261
(C8, RP column; solvent A: 0.1% trifluoroacetic acid in water; solvent B: 90% 2-propanol/trifluoroacetic acid; gradient of 0 - 6 0 % solvent B; flow rate 1.0 ml/minute). Protein assay. Protein concentration of purified sample was determined by a micro-Kjeldahl method [22] using glycine as standard.
S D S - P o l y a c r y l a m i d e Gel E l e c t r o p h o r e s i s and
Protein Transfer SDS PAGE and protein transfer onto nitrocellulose were performed as previously described [2]. Briefly, 1 mg/ml gel of crude birch pollen extract or 800/,g/ml gel of purified recombinant Bet v I preparation were applied to 15% polyacrylamide gels. The separated proteins were subsequently transferred onto nitrocellulose by using a blotting apparatus (Hoefer, San Francisco, CA) at 150 mA for 6 hours at room temperature. IgE Blots IgE blots were performed essentially as described elsewhere [2]. Briefly, after the protein-blotting procedure, nitrocellulose membranes were cut into strips and blocked with a buffer containing 0.5% bovine serum albumin and 0.5% Tween 20. Thereafter, the strips were incubated with a fourfold dilution of patients' sera. The specific IgE was detected after incubation with 125 I-labeled rabbit anti-human IgE (Pharmacia) and autoradiography.
RESULTS Reaction of patients' IgE with pollen extracts from Betulla verrucosa. The serum IgE level was tested in 34 of the 37 group I patients and amounted to 82 + 66 IU/ml (mean -+ standard deviation). Normal levels range between 10 and 150 IU IgE/ml. Only three of the patients tested had levels above the normal range (160, 190, and 320 IU IgE/ml). Having shown a positive skin-prick test with Betula veruccosa extracts, which was confirmed by RAST, patients' sera were tested for their IgE-binding pattern to birch pollen proteins by immunoblot analysis. Figure 1 shows representative reaction patterns of the two patient groups. The serum IgE of an allergic individual representative for group I patients bound selectively to Bet v I (molecular mass, 17 kD) in birch pollen extract (lane C). N o minor allergens from the birch pollen extract are recognized. It showed identical IgE-binding characteristics to purified recombinant Bet v I (lane D). The case history and skin-prick testing showed no evidence for reactivity against any other allergens.
262
G.F. Fischer et al.
A
kDa 97.4
-
46.0
-
30.0
-
21.5
-
B
C
D
E
F
G
H
14.3-
IgE-binding pattern from two representative birch-pollen-allergic patients. Birch pollen extracts (lanes A, C, E, and G) or purified recombinant Bet v I (lanes B, D, F, and H) were separated on a polyacrylamide gel and blotted onto nitrocellulose. Specific serum IgE of a patient from group II (lanes A and B) and from group I (lanes C and D) is shown. Lanes E - H represent the negative control reactions where normal human serum pool (lanes E and F) or buffer (lanes G and H) were used as first step reagent. FIGURE 1
Beside reactivity with birch pollen allergens, patients of group II showed IgE reactivity against further allergens as revealed by case history, skin-prick test, RAST, and immunoblot analysis. In the case of a group II patient assayed in lane A, the serum IgE recognized in addition to Bet v I several o t h e r - - m i n o r - - p r o t e i n s in birch pollen extract. The serum of this patient also showed IgE antibody reactivity with recombinant Bet v I (lane B) proving its Bet v I specificity. Lanes E - H show the negative control reactions for the birch pollen extract (lanes E and G) and recombinant Bet v I preparation (lanes F and H) using a normal human serum pool and buffer instead of patients' sera.
other allergens were recognized by the serum IgE (=group II) (Table 1). Results obtained by RFLP indicating a DRw52 a/c phenotype were confirmed by typing with allele-specific oligonucleotides: All of the RFLP defined DRw52a/c subjects belonged to the DRB3*0101 type (data not shown). A n amino acid sequence shared by D R B 1 and D R B 3 alleles might contribute to disease susceptibility. The amino acid
sequence tyrosine-phenylalanine-histidine (YFH) is located in the second hypervariable region of DR molecules at position 3 0 - 3 2 . It is shared by certain alleles from the DRB3 (DRw52a/c) and the DRB1 locus (DR3, some DRw13, -w14) [20]. All of these alleles can be identified by RFLP analysis. Comparison of the number of patients; possessing alleles with the Y F H sequence in their D R B I or DRB3 loci, with the control group revealed that the distribution differences of these alleles remained significant. This association comprised 81% of patients, however (Table 1). To strengthen this theoretical consideration, patients were studied for their YFH positiveness by probing their DRB genes with a sequence specific oligonucleotide. Figure 2 shows the hybridization pattern of the 37 patients belonging to patients' group I, whose exons 2 of their functional DRB loci were amplified and blotted to nylon membranes. All individuals expected from their RFLP H L A - D R type to be YFH positive indeed showed a positive reaction with the oligonucleotide
TABLE 1
Association between an H L A D R w 5 2 a / c phenotype and IgE response against Bet v I. The D R and D Q alleles in birch
pollen allergic patients were characterized by RFLP analyses [16]. The distribution of DQA-, DQB-, and D R B l - e n c o d e d HLA class II molecules did not differ in birch pollen allergic patients and the control population (data not shown). The frequency of the DRB3 allele DRw52a/c in 37 patients showing a selective IgE response against Bet v I (=group I) differed significantly from that in the control population: This allele was found in 62% of patients, whereas it was present in only 33% of the control population (Table 1). The p .... was
