Molecular mechanisms of HLA associations with myasthenia gravis

Abstract: Myasthenia gravis (MG) is a neuromuscular disease characterised by muscle weakness due to an autoimmune attack against the nicotinic acetylcholine receptor (nAChR). MG has a complex pattem of inheritance and several genetic and environmental factors are important in conferring susceptibility to the disease. An increased risk for development of MG is conferred by genetic factors in the human leukocyte antigen (HLA) region on the short arm of chromosome 6. The first aim of this study was to identify the primary HLA polymorphisms which confer MG susceptibility and protection in a Swedish population. Seventy-nine patients and 155 healthy, unrelated, population-based controls were typed for polymorphisms in the TAP, HLA-DQ, HLA-DR and TNF loci. The DQBI'02, DRB1'03 and TNFa2 alleles on the extended DR3-haplotype conferred an increased risk for development of MG. Profound linkage disequilibrium exists between these alleles and it was difficult to determine which allele that was responsible for the association. Stratification analysis of the associations indicates that the susceptibility factor(s) for MG are located telomeric of the TAP loci and that there might be two factors conferring susceptibility to MG on the extended DR3-haplotype, one located close to the TNF loci and another in the HLA-DQ region. TNFa2 was the strongest associated allele with MG. The DQBI'0603 (DQ6) allele was negatively associated with MG. MG is a heterogeneous disease and it is possible to divide the patients into different subgroups. The second aim of this study was to analyse the heterogeneity in HLA associations with different disease subgroups. The extended DR3-haplotype was found to be increased only in patients with thymic hyperplasia and an early onset of disease (below 30 years). Clusters of polymorphic amino acid residues in the HLA-DQ molecules were found in different frequencies in the subgroups. A cluster common to DQA1'01 and another cluster common to DQB1'05 and 06 alleles were negatively associated in patients with thymic hyperplasia and in patients with an early onset of disease. A cluster unique for DQB1'02 (DQ2) was positively associated with MG patients without thymomas and in patients with an early disease onset. Overall, the patients with thymic hyperplasia or an early disease onset had different associations with the clusters than patients with thymoma. The third aim of this study was to characterise structural properties of the HLA-DQ molecules which are positively and negatively associated with MG by molecular modelling. Three dimensional models of the positively associated DQ2 and negatively associated DQ6 (DQB1'0603) molecules were constructed using homology modelling techniques. Four pockets were identified in the peptide- binding grooves and the differences in the peptide binding groove was primarily located to the fourth pocket (P9). The pocket is larger in DQ2 than in DQ6 and the differences in electrostatic surface potentials between the molecules were most pronounced around this pocket. The fourth aim of this study was to analyse the correlation between HLA genotype and TCR V-gene usage by peripheral blood CD4+ and CD8+ T Iymphocytes in MG patients. TCR V gene expressions were deterrnined in 29 of the HLA-typed patients by FACS analysis with 12 monoclonal antibodies that detected 30-40 % of the TCRs. No correlation between HLA genotypes and TCR V gene expansions and no restricted usage of TCR V gene segments was seen. Half of the patients had abnormal T-cell expansions in peripheral blood and there was an increased expression of TCR V gene segments in CD8+, compared with CD4+, T-cells. TCR V gene expression was increased in unthymectomized patients in especially the CD8+ population, compared with thymectomized patients.