Contribution of a Single DQ/3 Chain Residue to Multiple Sclerosis in French Canadians David G. Haegert and Gordon S. Francis
A B S T R A C T : Putative disease susceptibility and resistance HLA class II alleles were studied in 78 French Canadian multiple sclerosis (MS) patients and 79 controls by using sequence-specific oligonucleotide probes to analyze in vitro amplified DNA (PCR-SSOP) typing). In this relatively homogeneous ethnic group, MS was positively associated with DRB5*0101, DQBI*0602, and DQAI*0102 and negatively associated with DQB 1"0301. The strongest disease association was with DQBI*0602. Complete DQB1 typing of these individuals, plus RFLP DQfl typing of an additional five patients showed that 98% of patients compared with 73% of controis carry DQB1 alleles encoding leucine at residue 26.
In contrast, 16% of patients compared with 38% of controls carry DQB1 alleles encoding tyrosine at the same residue, and 22% of patients versus 44% of controls carry DQB1 alleles encoding glycine at residue 26. The positive disease correlation was confirmed with SSO probes designed to hybridize to codons for amino acids 22-27 of DQBI*0602, 0603, 0604, 0302, 0303 or to codons for amino acids 25-31 of DQBl*0201; all of these alleles encode Leu 26. These findings suggest that DQfl chain polymorphisms at a single residue contribute to the development of MS in the French Canadian population. Human Immunology 34, 85-90 (1992)
ABBREVIATIONS
PCR SSOP RFLP
polymerase chain reaction sequence-specific oligonucleotide probe restriction fragment length polymorphism
PCR-SSOP FC MS
polymerase chain reaction amplification and sequence-specific oligonucleotide probe French Canadian multiple sclerosis
INTRODUCTION Genetic factors play a role in susceptibility to multiple sclerosis (MS) [1]. T h e H L A gene complex contributes to this genetic susceptibility and in various population studies an association between certain H L A class II specificities and MS has been identified. Interestingly, the MS-class II associations vary to some extent with the population studied and MS has been reported to be associated with DR2 and D Q w l specificities in most Caucasian groups [2, 3], with DR4 in Italians [4, 5] and Jordanian Arabs [6] and with D R w 6 in Mexicans [7] From the Disctpline of Pathology (D.G.H.), Memortal University of Newfoundland, St. John's; and the Montreal Neurological Hospital and McGill University (G.S.F.), Montreal, Canada. Address reprint requests to Dr. D. G. Haegert, Discipline of Pathology, Health Sciences Centre, Memorial University, St. John's, Newfoundland, Canada A1B 3V6. Received September 3, 1991; acceptedMarch 16, 1992. Human Immunology 34, 85-90 (1992) © American Society for Histocompatibility and Immunogenetics, 1992
and Japanese [8]. In view of the different ethnic associations between class II molecules and MS it was hypothesized that polymorphic amino acid sequences shared by different class II molecules might be important for disease d e v e l o p m e n t [9]. Subsequent investigation showed that the majority of Norwegian MS patients carry DQfl chains that share polymorphic amino acid sequences in the first domain [9, 10]. Conversely, however, no evidence was obtained for similar sharing of DQfl chain polymorphic sequences by MS patients of mixed ethnic origin [11]. The present study was therefore instigated to determine whether shared class II nucleotides and amino acids are critical to the development of MS in French Canadians, who form a relatively homogeneous ethnic population [12, 13]. Class II typing of polymerase chain reaction (PCR)-amplified genomic D N A using sequence-specific oligonucleotide 85 0198-8859/92/$5.00
86
probes (PCR-SSOP typing) indicated in French Canadians that MS is positively associated with DRB5*0101, DQBI*0602, and DQAI*0102 and negatively associated with DQB1*0301. The strongest disease association was with DQBI*0602. Sequence comparisons of DQBI*0602,* 0301, and other DQB1 alleles [14-16] with subsequent application of appropriate SSO probes demonstrated that a single DQB1 codon is at least partly responsible for HLA class II associations with MS in French Canadians. MATERIALS A N D M E T H O D S
Patients and controls. Seventy-eight French Canadian (FC) MS patients followed at the Montreal Neurological Hospital MS clinic were selected at random. All patients had definite MS by standard criteria and 90% were classifiable into one of two clinical subgroups; relapsingremitting (n -- 48) and progressive MS (n -- 22). Controls consisted of 79 FCs who were not age matched to the MS patients, but had no evidence of MS or other autoimmune disease. All patients and controls were unrelated and 25% of patients gave a family history of MS. Oligonucleotide primers and controls. The generic DR primers and probes and the DR allele-specific probes have been previously described by Vaughan et al. [17]; the DR probes were a gift from M. Howell (Tenovus Laboratory, Southampton, England). Sequences of primers and probes for complete DQB1 and DQA1 allelic typing and for hybridization with codons 31-37 of DQAI*0102, 0103, and 0501 (probe QA-05) and codons 31-37 of DQAI*0401 and 0601 (probe QA06) were derived from Fernandez-Vina et al. [18, 19]. The DR primers and DQ primers and probes were synthesized by the Sheldon Biotechnology Center (Montreal, Canada). An additional probe termed L26, 5'GAGCGCGTGCGTCTTGTG was designed to hybridize with codons 22-27 shared by DQBI*0602, 0603, 0604, 0302, 0303 and 0201 and probe DQw2, 5'-TCTTGTGAGCAGAAGCATC was designed to hybridize with codons 25-31 of DQBI*0201; the sequences of L26 and DQw2 probes were derived from published sequences [14-16]. In vitro DNA amplification and PCR-SSOP HLA class II typing. DNA was isolated from peripheral blood samples by phenol-chloroform extraction [20]. DNA aliquots (1/*g) were denatured at 94°C for 7 minutes and then subjected to 35 cycles of PCR using a Perkin Elmer-Cetus thermal cycler (2 minutes at 94°C, 2 minutes at 55°C, and 2 minutes at 72°C) and Taq DNA polymerase (Perkin Elmer-Cetus or Promega). Onefiftieth of the amplified DNA was dot blotted onto Hy-
D.G. Haegert and G. S. Francis
bond N nylon membranes (Amersham), prehybridized for 1.5 hours in 3 M tetramethylammoniumchloride, 50 mM Tris-HC1, pH 8.0, 2 mM EDTA, 0.1% SDS, 100 /,g/ml salmon sperm DNA. Hybridization was carried out for 1.5 hours with 3'-end-labeled probes ([~2P]dCTP) at 55°C followed by washing as described by Fernandez-Vina et al. [19]. Autoradiography was then performed for 1-24 hours.
Data analysis. The X2 (chi-squared) test was used to analyze differences between patients and controls in 2 x 2 tables. Fisher's exact test was used when any expected cell value in a 2 x 2 table was less than 5. P values were corrected by the Bonferoni inequality method, i.e., by multiplying by the number of comparisons made, e.g., 15 in Table 1. Relative risk (RR) is an odds ratio (number of patients with allele per number of patients without allele)/(number of controls with allele per number of controls without allele) [21]; to correct for cells of less than 5, Haldane's modification of Woolf's method was used to estimate relative risk [22]. RESULTS
Associations of DR, DQB1, and DQAI alleles with MS. Using the generic DR probes (Table 1), DR2 was found to be MS associated (X2 = 18.4; corrected P value [PC] = 0.0003) and of the DR2 alleles tested DRB5*0101 was the major MS-associated DR2 allele (X2 = 17.14; PC = 0.0004) with RR = 4.33 (2.00
A B S T R A C T : Putative disease susceptibility and resistance HLA class II alleles were studied in 78 French Canadian multiple sclerosis (MS) patients and 79 controls by using sequence-specific oligonucleotide probes to analyze in vitro amplified DNA (PCR-SSOP) typing). In this relatively homogeneous ethnic group, MS was positively associated with DRB5*0101, DQBI*0602, and DQAI*0102 and negatively associated with DQB 1"0301. The strongest disease association was with DQBI*0602. Complete DQB1 typing of these individuals, plus RFLP DQfl typing of an additional five patients showed that 98% of patients compared with 73% of controis carry DQB1 alleles encoding leucine at residue 26.
In contrast, 16% of patients compared with 38% of controls carry DQB1 alleles encoding tyrosine at the same residue, and 22% of patients versus 44% of controls carry DQB1 alleles encoding glycine at residue 26. The positive disease correlation was confirmed with SSO probes designed to hybridize to codons for amino acids 22-27 of DQBI*0602, 0603, 0604, 0302, 0303 or to codons for amino acids 25-31 of DQBl*0201; all of these alleles encode Leu 26. These findings suggest that DQfl chain polymorphisms at a single residue contribute to the development of MS in the French Canadian population. Human Immunology 34, 85-90 (1992)
ABBREVIATIONS
PCR SSOP RFLP
polymerase chain reaction sequence-specific oligonucleotide probe restriction fragment length polymorphism
PCR-SSOP FC MS
polymerase chain reaction amplification and sequence-specific oligonucleotide probe French Canadian multiple sclerosis
INTRODUCTION Genetic factors play a role in susceptibility to multiple sclerosis (MS) [1]. T h e H L A gene complex contributes to this genetic susceptibility and in various population studies an association between certain H L A class II specificities and MS has been identified. Interestingly, the MS-class II associations vary to some extent with the population studied and MS has been reported to be associated with DR2 and D Q w l specificities in most Caucasian groups [2, 3], with DR4 in Italians [4, 5] and Jordanian Arabs [6] and with D R w 6 in Mexicans [7] From the Disctpline of Pathology (D.G.H.), Memortal University of Newfoundland, St. John's; and the Montreal Neurological Hospital and McGill University (G.S.F.), Montreal, Canada. Address reprint requests to Dr. D. G. Haegert, Discipline of Pathology, Health Sciences Centre, Memorial University, St. John's, Newfoundland, Canada A1B 3V6. Received September 3, 1991; acceptedMarch 16, 1992. Human Immunology 34, 85-90 (1992) © American Society for Histocompatibility and Immunogenetics, 1992
and Japanese [8]. In view of the different ethnic associations between class II molecules and MS it was hypothesized that polymorphic amino acid sequences shared by different class II molecules might be important for disease d e v e l o p m e n t [9]. Subsequent investigation showed that the majority of Norwegian MS patients carry DQfl chains that share polymorphic amino acid sequences in the first domain [9, 10]. Conversely, however, no evidence was obtained for similar sharing of DQfl chain polymorphic sequences by MS patients of mixed ethnic origin [11]. The present study was therefore instigated to determine whether shared class II nucleotides and amino acids are critical to the development of MS in French Canadians, who form a relatively homogeneous ethnic population [12, 13]. Class II typing of polymerase chain reaction (PCR)-amplified genomic D N A using sequence-specific oligonucleotide 85 0198-8859/92/$5.00
86
probes (PCR-SSOP typing) indicated in French Canadians that MS is positively associated with DRB5*0101, DQBI*0602, and DQAI*0102 and negatively associated with DQB1*0301. The strongest disease association was with DQBI*0602. Sequence comparisons of DQBI*0602,* 0301, and other DQB1 alleles [14-16] with subsequent application of appropriate SSO probes demonstrated that a single DQB1 codon is at least partly responsible for HLA class II associations with MS in French Canadians. MATERIALS A N D M E T H O D S
Patients and controls. Seventy-eight French Canadian (FC) MS patients followed at the Montreal Neurological Hospital MS clinic were selected at random. All patients had definite MS by standard criteria and 90% were classifiable into one of two clinical subgroups; relapsingremitting (n -- 48) and progressive MS (n -- 22). Controls consisted of 79 FCs who were not age matched to the MS patients, but had no evidence of MS or other autoimmune disease. All patients and controls were unrelated and 25% of patients gave a family history of MS. Oligonucleotide primers and controls. The generic DR primers and probes and the DR allele-specific probes have been previously described by Vaughan et al. [17]; the DR probes were a gift from M. Howell (Tenovus Laboratory, Southampton, England). Sequences of primers and probes for complete DQB1 and DQA1 allelic typing and for hybridization with codons 31-37 of DQAI*0102, 0103, and 0501 (probe QA-05) and codons 31-37 of DQAI*0401 and 0601 (probe QA06) were derived from Fernandez-Vina et al. [18, 19]. The DR primers and DQ primers and probes were synthesized by the Sheldon Biotechnology Center (Montreal, Canada). An additional probe termed L26, 5'GAGCGCGTGCGTCTTGTG was designed to hybridize with codons 22-27 shared by DQBI*0602, 0603, 0604, 0302, 0303 and 0201 and probe DQw2, 5'-TCTTGTGAGCAGAAGCATC was designed to hybridize with codons 25-31 of DQBI*0201; the sequences of L26 and DQw2 probes were derived from published sequences [14-16]. In vitro DNA amplification and PCR-SSOP HLA class II typing. DNA was isolated from peripheral blood samples by phenol-chloroform extraction [20]. DNA aliquots (1/*g) were denatured at 94°C for 7 minutes and then subjected to 35 cycles of PCR using a Perkin Elmer-Cetus thermal cycler (2 minutes at 94°C, 2 minutes at 55°C, and 2 minutes at 72°C) and Taq DNA polymerase (Perkin Elmer-Cetus or Promega). Onefiftieth of the amplified DNA was dot blotted onto Hy-
D.G. Haegert and G. S. Francis
bond N nylon membranes (Amersham), prehybridized for 1.5 hours in 3 M tetramethylammoniumchloride, 50 mM Tris-HC1, pH 8.0, 2 mM EDTA, 0.1% SDS, 100 /,g/ml salmon sperm DNA. Hybridization was carried out for 1.5 hours with 3'-end-labeled probes ([~2P]dCTP) at 55°C followed by washing as described by Fernandez-Vina et al. [19]. Autoradiography was then performed for 1-24 hours.
Data analysis. The X2 (chi-squared) test was used to analyze differences between patients and controls in 2 x 2 tables. Fisher's exact test was used when any expected cell value in a 2 x 2 table was less than 5. P values were corrected by the Bonferoni inequality method, i.e., by multiplying by the number of comparisons made, e.g., 15 in Table 1. Relative risk (RR) is an odds ratio (number of patients with allele per number of patients without allele)/(number of controls with allele per number of controls without allele) [21]; to correct for cells of less than 5, Haldane's modification of Woolf's method was used to estimate relative risk [22]. RESULTS
Associations of DR, DQB1, and DQAI alleles with MS. Using the generic DR probes (Table 1), DR2 was found to be MS associated (X2 = 18.4; corrected P value [PC] = 0.0003) and of the DR2 alleles tested DRB5*0101 was the major MS-associated DR2 allele (X2 = 17.14; PC = 0.0004) with RR = 4.33 (2.00
