HLA Class II DNA Typing in a Large Series of European Patients with Systemic Lupus Erythematosus

Introduction The etiopathogenesis of systemic lupus erythematosus (SLE) is complex and still largely unknown. Genetic, environmental, and hormonal factors contribute to disease susceptibility (1). The importance of genetic factors is documented by 1) the pronounced difference in concordance rates between monozygotic and dizygotic twins (15); 2) the higher disease prevalence in relatives of patients with SLE (14); 3) the higher prevalence of SLE in certain ethnic groups, like African Americans and some Native Americans (18); and 4) studies of murine SLE showing that some strains invariably develop the disease (23,24). In humans, several genes contribute to lupus susceptibility. Their identification has been complicated by the fact that SLE is a heterogeneous disease, both clinically and immunologically (3); in addition, many of the susceptibility genes differ across the various populations (27), and it is well recognized that genetic predisposition is linked to autoantibody repertoires and to clinical subsets of disease (9,20,31). Many studies have focused on human leukocyte antigen (HLA) genes, mainly because they are involved in immune recognition and response, even if genetic susceptibility to SLE is in part located outside the short arm of the sixth chromosome (13,22,30,33,35,36). Studies performed by genomic scan of multiplex lupus families confirm the pathogenetic role of the HLA system in this disease (19). We decided to approach the problem of HLA-SLE association in a large sample of patients, all of Caucasian origin, to enable study of homogeneous subsets defined clinically and serologically. In this paper, we present the results of an immunogenetic study conducted in a large series of European SLE patients. Our objective was to determine the HLA class II allele association with SLE and its main clinical and autoantibody subsets. Patients and Methods Five hundred seventy-seven patients with SLE (91.2% female, 8.8% male; female:male ratio = 10:1), fulfilling 4 or more of the American Rheumatism Association (ARA) 1982 revised criteria for the classification of the disease (32), coming from 11 university hospitals of 7 European countries, were enrolled consecutively. Mean age of these patients was 37.9 ± 13.1 years (range, 13–79 yr). Mean age at the onset of the disease was 29.5 ± 12.5 years (range, 4–77 yr). Mean disease duration was 8.2 ± 3.1 years (range, 1–24 yr). To avoid regional biases, each center was asked to contribute 60 patients, with the only exception the 2 centers in Malaga, in southern Spain, that were asked to give 30 patients each. Results were available for 577 patients, whose regional distribution is depicted in Table 1.TABLE 1: Regional distribution of 577 European patients with systemic lupus erythematosusEach of the centers recorded the clinical characteristics of the patients, including the past medical history and a complete physical examination, using specific protocol forms, and collected blood and serum samples for the immunogenetic and serologic studies. All participating centers have substantial experience in the management of patients with SLE. Nevertheless, in order to minimize the possible interobserver bias, the definition of the clinical features included in the protocol forms was discussed by all participating physicians on several occasions. (For a precise definition of the clinical features, see references 5 and 29.) Blood and serum samples were taken from patients at the time clinical information was collected and were stored at −80 °C. Subsequently, they were transported in dry ice to the respective laboratories by the researchers of the clinical centers. HLA class II typing was done at the following 3 laboratories, with the same methodology: 1) Department of Medical Genetics, Institute of Clinica Medica, University of Cagliari, Italy; 2) Immunogenetics Laboratory, National Institute for Cancer Research, Genova, Italy; 3) Laboratory of Cellular Immunology and Immunogenetics, Institute of Internal Medicine, Infectious Diseases and Immunopathology, University of Milano, Italy. To guarantee quality control of the DNA typing tests and interlaboratory procedural uniformity, there was exchange of personnel as well as of biologic material (blood samples, genetic probes) between the 3 laboratories where the tests were performed. In each patient, the following HLA loci were studied: DRB1, DRB3, DRB4, DRB5, DQA1, DQB1. The frequency of HLA alleles in patients with SLE was compared with the frequency in a healthy Caucasian population, matched for ethnic origin, taken from the 12th International Histocompatibility Workshop (IHW) (11). Results of HLA class II allele typing of a healthy Polish population were from Jungerman et al (16). Various serologic tests were performed, including the detection of autoantibodies that are more commonly found with SLE. Each serologic test was performed in the same laboratory for all patients (Institute of Rheumatology, University of Siena, Italy). Centralized data handling, computing, and statistical analysis took place at the Rheumatology Division, San Camillo Hospital, Rome, Italy. The data were transferred into a computerized database program (DBASE III PLUS) and were analyzed using the Statistical Package for Social Sciences (SPSS) program. The study, approved by the Human Experimental Committee of each participating hospital, was carried out according to the principles of the Declaration of Helsinki. Informed consent was obtained from all patients. Autoantibody detection Antinuclear antibodies (ANA) were determined by indirect immunofluorescence using Hep-2 cells as substrate. Anti-dsDNA antibodies were determined by indirect immunofluorescence with Crithidia luciliae as substrate. Precipitating antibodies to extractable nuclear antigens, including Ro (SS-A), La (SS-B), U1-RNP, Sm, Jo1, SCL70, PCNA, were detected by double immunodiffusion and counterimmunoelectrophoresis using calf and rabbit thymus and human spleen extracts, as previously described (10,20). Other autoantibodies, such as anticardiolipin antibodies, anti-β2glycoprotein I antibodies, antineutrophil cytoplasmic antibodies, and antiganglioside antibodies, were also determined. Results of their correlation with clinical manifestation and HLA class II alleles have been published separately (7–10). HLA typing Molecular typing of HLA-DRB1, DRB3, DRB4, DRB5, DQA1 and DQB1 was performed according to the 11th IHW reference protocol (17) with minor variations. DNA was extracted from peripheral blood mononuclear cells using a standard phenol/chloroform technique. HLA-DRB, DQA1, DQB1 alleles were typed by the polymerase chain reaction-sequence specific oligonucleotide probes (PCR-SSOP) method, carried out using digoxigenin (DIG)-labeled probes. The primers and the oligonucleotide probes used were those validated by the 11th IHW. Briefly, the second exon of each gene is specifically PCR amplified; the amplified fragment is immobilized on a nylon positively-charged membrane (Zeta-Probe blotting membrane, Bio-Rad, Hercules, CA) and hybridized with sequence-specific probes previously 3′-end labeled with DIG-ddUTP using terminal-transferase (Boehringer Mannheim Gmbh, Germany). An anti-DIG-alkaline phosphatase conjugate (Fab fragments) (Boehringer Mannheim Gmbh) is bound to the DIGlabeled hybrids; incubation with CSPD (Tropix, Bedford, MA) substrate allows the chemiluminescent detection of the hybrids, by exposing the membranes to a X-Omat AR autoradiography film (Eastman Kodak, Rochester, NY). Statistical analysis Frequencies of HLA alleles were compared among the various groups using conventional chi-square analysis with the Yates correction or, when appropriate, the Fisher exact test (2-tailed). The p value for HLA frequency differences was corrected (pc) for the number of comparisons. A p value equal to or less than 0.05 was taken as significant. An odds ratio (OR) was calculated, when the p value was significant, by using 2×2 contingency tables. When the value of a cell in the 2×2 table was 0, relative risk (RR) was calculated by Haldane’s modification of Woolf’s method (12):EQUATION This statistical analysis was performed by means of the SPSS program using the information stored in the database program. Results Frequency of HLA class II alleles in patients and in healthy controls As expected, frequency of HLA-DRB1*03 was significantly higher in SLE patients compared with healthy controls (35.2 vs 19.4, p < 10−6, pc = 1.3 × 10−5, OR = 2.2, 95% CI [confidence intervals] = 1.8–2.9) (Table 2). In addition, DRB1*15 and DRB1*16 (DR2) were significantly associated with SLE (respectively, 27.7 vs 19.4, p = 0.003, pc = 0.04, OR = 1.6, 95% CI = 1.2–2, and 14.4 vs 10.5, p = 0.03, pc = not significant [NS], OR = 1.4, 95% CI = 1.1–2). On the contrary, DRB1*01, DRB1*04, and DRB1*11 (DR5) showed negative association with SLE. Alleles at the DRB3 locus were not associated with the disease.TABLE 2: Phenotypic frequency of HLA-DRB1 and DRB3 alleles in patients with SLE and in healthy controlsAmong the alleles at the DQA1 locus, only DQA1*0102 showed a positive association with SLE (42.6 vs 32.8, p = 2 × 10−4, pc = 0.006, OR = 1.5, 95% CI = 1.2–1.9), whereas DQA1*0101 and DQA1*03 were negatively associated (Table 3).TABLE 3: Phenotypic frequency of HLA-DQA1 and DQB1 alleles in patients with SLE and in healthy controlsSome alleles at the DQB1 locus were found with significantly increased frequency in patients with SLE compared with healthy controls. They were DQB1*0502 (17.8 vs 11.6, p = 0.001, pc = 0.03, OR = 1.6, 95% CI = 1.2–2.2), DQB1*0602 (20.7 vs 15.1, p = 0.008, pc = NS, OR = 1.5, 95% CI = 1.1–1.9), DQB1*0201 (52.7 vs 39.7, p = 2 × 10−6, pc = 5.6 × 10−5, OR = 1.7, 95% CI = 1.3–2.1), DQB1*0303 (5.8 vs 3.4, p = 0.05, pc = NS, OR = 1.7, 95% CI = 1–2.9), DQB1*0304 (1.9 vs 0.4, p = 0.02, pc = NS, OR = 4.3, 95% CI = 1.2–16.2). Conversely, DQB1*0501, DQB1*0301, and DQB1*0302 showed a negative association with SLE (see Table 3). Frequency of autoantibodies in lupus patients and correlation with HLA class II alleles Frequencies of those autoantibodies of interest for this paper are depicted in Table 4. They are in keeping with the findings in previous studies. With the exception of anti-dsDNA (pc = 0.001), the frequencies of all other autoantibodies were not statistically different among the various centers. On the other hand, the frequency of anti-dsDNA in every patient was quite variable with time, depending mainly on the disease activity and therapy.TABLE 4: Relative frequency of autoantibodies in patients with SLEAs already known, anti-Ro was overrepresented in haplotypes containing DR3 (36.6 in patients DRB1*03 positive vs 19 in their negative counterparts, p = 2 × 10−5, pc = 3 × 10−4, OR = 2.4, 95% CI = 1.6–3.7), as well as anti-La (13.1 vs 2.7, p = 1 × 10−5, pc = 1 × 10−4, OR = 5.5, 95% CI = 2.5–12.1) (Table 5). Both these autoantibodies were also strongly associated with DQB1*0201 (DQw2), which is in strong linkage disequilibrium with DRB1*03, and, obviously, showed a strong association with haplotypes containing DRB1*03 and DQB1*0201. Anti-Ro, but not anti-La, was also associated with DQB1*0502 (DQw1), and with the haplotype DR2/ DQB1*0502. It was not associated with DR2 (DRB1*15, DRB1*16). Anti-dsDNA was associated with DRB1*15 (66.2 vs 52.2, p = 0.008, pc = NS, OR = 1.8, 95% CI = 1.2–2.7), DQA1*0102 (63.6 vs 50.4, p = 0.004, pc = NS, OR = 1.7, 95% CI = 1.2–2.5), and DQB1*0602 (66 vs 53.5, p = 0.03, pc = NS, OR = 1.7, 95% CI = 1.1–2.7).TABLE 5: Relative frequency of significantly associated autoantibodies in patients positive and negative for the various HLA alleles/haplotypesAnti-Ro showed negative association with DRB1*01 (11 vs 27.6, p = 0.004, pc = NS, OR = 0.3, 95% CI = 0.1–0.7), DRB1*14 (3.6 vs 26.5, p = 0.01, pc = NS, OR = 0.1, 95% CI = 0.01–0.8), DQA1*0101 (10.2 vs 28.3, p = 6 × 10−4, pc = 0.02, OR = 0.3, 95% CI = 0.1–0.6), DQB1*0501 (10.7 vs 28, p = 0.001, pc = 0.03, OR = 0.3, 95% CI = 0.1–0.6), and DQB1*0503 (0 vs 26.5, p = 0.006, pc = NS, OR = 0.7, 95% CI = 0.6–0.8). Anti-U1-RNP was negatively associated with DRB1*03 (12 vs 21.4, p = 0.01, pc = NS, OR = 0.5, 95% CI = 0.3–0.8), and DQA1*0501 (13 vs 24, p = 0.002, pc = NS, OR = 0.5, 95% CI = 0.3– 0.7). Anti-nDNA was negatively associated with DQB1*0301 (48.2 vs 59.1, p = 0.04, pc = NS, OR = 0.6, 95% CI = 0.4–0.9). The other autoantibodies showed no association with HLA class II alleles. Correlation of clinical manifestations with HLA class II alleles and with autoantibodies Frequency of clinical manifestation was compared among the 11 centers. The following clinical manifestations showed a significantly different frequency: leukopenia (pc = 6 × 10−4, ranging from 20% in patients coming from Austria to 80% in patients from Catalonia, Spain), lymphopenia (pc = 6 × 10−4, ranging from 10% in patients from central/southern Italy to 85% in French patients), lymphadenopathy (pc = 6 × 10−4, ranging from about 37% in French and Greek patients to about 20% in patients in all other centers), central nervous system involvement (pc = 6 × 10−4, ranging from about 69% in patients from France and Poland to 15% in those from northern Italy), livedo reticularis (pc = 6 × 10−4, ranging from about 43% in patients from Poland and Greece to about 15% in those from the other centers). The association between these clinical manifestations and HLA alleles could be due to environmental factors or might be due to regional biases regarding diagnosis, and has to be considered cautiously. All other clinical manifestations showed a similar frequency across the 11 centers. Some HLA class II alleles were associated with clinical manifestations of SLE (Table 6), perhaps due to their association with autoantibodies. Indeed, DRB1*03 was associated with pleuritis, lung involvement, renal involvement, and psychosis. DRB1*16 carried a higher risk for renal involvement, discoid lupus, and lymphadenopathy. DQA1*0101 was associated with lymphadenopathy, and DQA1*03 with lung involvement, lymphopenia, and hemolytic anemia. DQB1*0201 (DQw2) was associated with leukopenia, digital skin vasculitis, and pleuritis, while DQB1*0502 (DQw1) was associated with renal involvement, discoid lupus, and livedo reticularis.TABLE 6: Relative frequency of major clinical manifestations in patients positive and negative for the various HLA alleles several HLA class II alleles were negatively associated with clinical manifestations of SLE (Table DRB1*01 showed negative with renal and DRB1*11 was negatively associated with renal disease and and DRB1*14 was negatively associated with leukopenia, lymphopenia, and digital skin the locus, DQA1*0101 showed negative with and renal disease, DQA1*0102 with involvement, DQB1*0301 with and renal involvement, DQB1*0501 with renal involvement, and DQB1*0503 with Relative frequency of major clinical manifestations in patients positive and negative for the various HLA alleles of these clinical manifestations were associated with the of autoantibodies. anti-Ro was associated with in patients anti-Ro positive vs in those anti-Ro p = 1 × 10−4, OR = 2.2, 95% CI = and with lung involvement vs p = 2 × OR = 95% CI = was associated with vs p = 0.03, OR = 95% CI = and showed negative association with renal disease vs p = 0.03, OR = 0.1, 95% CI = found in the of other clinical manifestations that are more of the are discussed in paper loci contribute to and disease is as a complex genetic that is by environmental studies are complicated because the disease is with susceptibility from a number of genes with studies in SLE are also because SLE is in several clinical and each with its genetic genes to SLE according to the ethnic of the including association and have the linkage of the disease to the in (19). It also has been that certain haplotypes association with autoantibodies in SLE than with the disease Studies of the association between HLA class II alleles and and between HLA and serologic of the disease, were in by in and we the to that paper for reference In Table and Table we have the described by other in In these studies found that was associated with Caucasian SLE patients in the Spain, and, more also in and the while in Greece the significant association was with and in northern with On the other the allele was associated with SLE in and groups including and while in patients the association was with the In SLE was found to be associated with and in both and with SLE was found to be associated with and DQB1*0201 in Caucasian patients and with in African In addition, in patients, an association has been found with and with and DQB1*0501 haplotypes In patients an association of SLE with both and has been found while in Greek SLE patients, with and have been described with SLE in different ethnic HLA class II alleles and autoantibody subsets of the and genes and autoantibody subsets of have been found between anti-dsDNA antibodies and DR2 and DR3 and their respective linked alleles DQB1*0602 and as well in those these An association of anti-dsDNA antibodies with and in French SLE patients has been and anti-La antibodies have been found to be associated with and DR3 and with the of and studies from 2 different European centers have the haplotype in susceptibility to autoantibody in In patients, the only association has been with The of anti-Ro and anti-La antibodies has been found to be associated with and DQB1*0201 alleles in patients an association has been found in Italy between and and DQB1*0201 haplotype Anti-U1-RNP antibodies have been found to be associated in both and with and serologic and using and DNA have both HLA-DRB1 and specifically between and and linked to well as other specific DQB1 alleles specific with DQB1*0302 in both and All studies have The studies not homogeneous clinical criteria for patient and they used different laboratory for the detection of autoantibodies that be In the number of patients is of the we to the characteristics of the association of SLE with HLA class II alleles in a large sample of European patients. In the large sample to patients into various defined according to clinical characteristics and autoantibody and the of the patients to the problem of ethnic on HLA-SLE On the other hand, results of study be at in part by the number of participating because we not out the of environmental factors and of biases regarding The results that the of the to the disease is less than its to certain of autoantibodies in to clinical In we have already that HLA class II alleles in the of to the of antibodies and, as a to the of clinical manifestations to the We have also found significant association between other HLA alleles and both antiganglioside antibodies and clinical manifestations In this paper, we in the same patient other HLA class II alleles susceptibility for different autoantibodies and clinical patients the DRB1*03 allele and the linked DQB1*0201 allele genetic predisposition to the of autoantibodies to and to involvement, pleuritis, and psychosis. a similar association between DQB1*0201 and antibodies has been previously described in a To we have found for the time that patients the DQB1*0502 which is in linkage disequilibrium with are to develop anti-Ro antibodies anti-La antibodies, renal disease, discoid lupus, and livedo whereas the of anti-La antibodies might be for renal The association of and described to is a of the association of this allele with antibodies DR2 (DRB1*15, was associated with anti-dsDNA antibodies, data previously described in Caucasian It is that we found for the time, as as we that DR2 haplotypes were also associated with renal disease, discoid lupus, and lymphadenopathy. DRB1*11 (DR5) to lupus, by and lymphopenia, while SLE patients showed a risk for renal and involvement, with a frequency of DRB1*14 was associated with a risk of and patients with this allele were less to develop digital skin vasculitis, leukopenia, and The negative association of anti-dsDNA antibodies with DQB1*0301 could patients with this allele less renal results confirm that the role of HLA class II genes in SLE is to the of the autoantibody and clinical whereas the onset of SLE is more the of the of many other genes, with Indeed, studies showed that the genetic predisposition to SLE is only in part located on the short arm of the sixth Other of the human to be more for SLE such as those located in chromosome and (19). In addition, studies that different susceptibility loci are involved in the of SLE in different ethnic the genetic of this In study of a large sample of Caucasian patients with lupus some HLA clinical and serologic in SLE and the that this disease is complex and that the role of genes is mainly to to serologic and clinical We conducted this study to determine the HLA class II allele in a large of patients of homogeneous ethnic with systemic lupus erythematosus The large sample to patients according to their clinical and serologic We 577 European Caucasian patients with SLE. Antinuclear antibodies anti-dsDNA antibodies and antibodies to extractable nuclear Ro (SS-A), La (SS-B), U1-RNP, Sm, Jo1, SCL70, and PCNA, were detected in all patients. Molecular typing of HLA-DRB1, DRB3, DQA1, and DQB1 loci was performed by the polymerase chain reaction-sequence specific oligonucleotide probes (PCR-SSOP) We found a significantly increased frequency of DRB1*03, and DQB1*0304 in lupus patients as compared with healthy controls. In addition, DRB1*03 was associated with anti-La, pleuritis, and involvement of and central nervous DRB1*15 and DQB1*0602 were associated with anti-dsDNA DQB1*0201 with anti-Ro and anti-La, leukopenia, digital skin vasculitis, and and DQB1*0502 was associated with renal involvement, discoid lupus, and livedo In study some HLA clinical and serologic in SLE and that the role of genes is mainly to to serologic and clinical manifestations of this