Key points

Background

Rheumatic diseases are autoimmune diseases that are characterized by inflammation and destruction of joints, muscles, blood vessels, and organs. Both genetic and environmental factors typically contribute to the onset of rheumatic diseases. For example, familial and epidemiologic studies have demonstrated that approximately 50% of the disease risk of rheumatoid arthritis (RA), one of the most common rheumatic diseases that affect synovial joints, is explained by genetic factors. Recent developments in human genome sequencing technologies, such as high-density single nucleotide polymorphism (SNP) microarrays and next-generation sequencing (NGS), have substantially contributed to the elucidation of the genetic architecture of human complex traits. In particular, genome-wide association studies (GWAS), a method of statistical genetics that massively evaluates the disease risk of genome-wide SNPs, has successfully identified several human disease risk genes. Specifically for RA, a large-scale transethnic GWAS identified more than 100 risk genetic loci, with implications for novel drug discovery. In this review, the authors highlight recent findings on genomics of rheumatic diseases and their application to translational research.

Background

Rheumatic diseases are autoimmune diseases that are characterized by inflammation and destruction of joints, muscles, blood vessels, and organs. Both genetic and environmental factors typically contribute to the onset of rheumatic diseases. For example, familial and epidemiologic studies have demonstrated that approximately 50% of the disease risk of rheumatoid arthritis (RA), one of the most common rheumatic diseases that affect synovial joints, is explained by genetic factors. Recent developments in human genome sequencing technologies, such as high-density single nucleotide polymorphism (SNP) microarrays and next-generation sequencing (NGS), have substantially contributed to the elucidation of the genetic architecture of human complex traits. In particular, genome-wide association studies (GWAS), a method of statistical genetics that massively evaluates the disease risk of genome-wide SNPs, has successfully identified several human disease risk genes. Specifically for RA, a large-scale transethnic GWAS identified more than 100 risk genetic loci, with implications for novel drug discovery. In this review, the authors highlight recent findings on genomics of rheumatic diseases and their application to translational research.

Roles of the major histocompatibility region to risk of rheumatic diseases

The major histocompatibility complex (MHC) region, a genetic locus located at chromosome 6p23, is known to have a strong impact on the genetic risk of rheumatic diseases. Although this region is only 0.1% of the length of the human genome, the MHC region confers most of the risk for most rheumatic diseases. The initial identification of the genetic risk loci was reported to be associated with the HLA genes located in the MHC region, such as HLA-DRB1 for RA, HLA-C for psoriasis, HLA-B for ankylosing spondylitis, and HLA-DPB1 for Graves disease. However, delineation of the detailed disease risk of HLA alleles has been challenging and controversial owing to the complex structures of the polymorphisms in the MHC region. For RA, the HLA-DRB1 alleles, which share a conserved amino acid sequences at positions 70 to 74 and are called shared epitope (SE) alleles, confer strong risk in multiple populations, but non-SE HLA-DRB1 alleles contribute risk as well.

Recently, the method of statistical genetics called HLA imputation was developed. This approach computationally imputes (ie, estimates) HLA alleles of the individuals using SNP genotyping data and, therefore, allows comprehensive HLA allele risk assessment using existing large-scale GWAS data without additional genotyping costs. Application of the HLA imputation method to GWAS data had facilitated successful fine-mapping of the risk HLA variants of multiple diseases. HLA imputation-based analysis of GWAS data in autoantibody-positive RA revealed that most of the MHC risk was explained by amino acid sequence polymorphisms at positions 11 and 13 of HLA-DRβ1 molecule in multiple populations, including Europeans, East Asians, Japanese, and African Americans ( Fig. 1 ). The HLA imputation method contributed to HLA allele fine-mapping of other rheumatic or immune-related diseases, including psoriasis, Graves diseases, type 1 diabetes, systemic lupus erythematosus, the subset of patients with lung adenocarcinoma with mutations in the epidermal growth factor receptor, and natural killer T-cell lymphoma.

Amino acid positions of RA risk on 3-dimensional structure of the HLA-DRβ1 molecule. HLA-DRβ1 amino acid sequence alterations at positions 11 and 13 confer strong risk of RA in multiple populations.

Most of these HLA fine-mapping studies indicated that the amino acid sequence polymorphisms of the classic HLA genes were disease risk variants. However, the recent HLA imputation analysis of RA GWAS data in the Japanese populations reported that the synonymous variant of HLA-DOA , one of the nonclassic HLA genes, had risk independently from other classic HLA genes, including HLA-DRB1 . The HLA-DOA risk variant demonstrated an expression quantitative trait locus effect on HLA-DOA mRNA expressions, suggesting a dosage contribution of the nonclassic HLA genes to the disease biology.

Genome-wide association studies identified many non–major histocompatibility complex risk genes

Genetic variants outside of the MHC region confer relatively weaker disease risk compared with those in the MHC region, so many subjects and controls are required to have sufficient statistical power to identify these variants. Implementation of GWAS helped to address this issue, and this approach has successfully identified many disease risk genes of rheumatic diseases. Initially, GWAS was performed for single populations; but currently, large-scale transethnic meta-analyses are routinely conducted.

Although the GWAS strategy successfully identified disease risk genes, only a small fraction of the disease risk was explained by GWAS-identified risk variants. This problem is called missing heritability. Multiple hypotheses have been proposed to explain missing heritability (also see Vincent A. Laufer and colleagues article, “ Integrative Approaches to Understanding the Pathogenic Role of Genetic Variation in Rheumatic Diseases ,” in this issue), such as rare risk variants not represented on SNP genotyping panels. With the development of NGS technologies, rare risk variants could be identified ; it was discovered that these rare variants had small contributions to the genetic risk of common diseases in the populations, so they did not contribute substantially to solving the problem of missing heritability. Assessment of the role of rare variant risks by whole-exome sequencing or whole-genome sequencing will require relatively larger sample sizes compared with evaluation of common variants by GWAS, and further synthesis of the studies through international collaborative partnerships will be necessary. Although further analytical assessments may be warranted, polygenic effects from genome-wide common SNPs with very weak risk have also been considered to potentially explain missing heritability. Other genetic factors not typically assessed in the GWAS approaches, such as a nonadditive effect of the genetic risk variants and gene-gene interactions, have also been suggested for filling in the unexplained missing heritability.

Another future goal of genetic studies is to identify the variants associated with disease prognosis or severity. Previous studies suggested that the genetic factors for diseases onset and prognosis may not be identical, and genetic studies specifically focusing on clinical outcomes are warranted.