Ankylosing spondylitis (AS) is a polygenic autoimmune disease of the axial skeleton affecting 0.3-0.5 % of the human population and causing a debilitating condition. AS starts with inflammation around ligament and tendon attachments (enthesitis) and is frequently associated with inflammation of the peripheral joints. Also, AS is often associated with extra-skeletal manifestations such as anterior uveitis and inflammatory bowel disease. Although the etiology of AS is unknown, genetic and environmental components have been prominently implicated as predisposing factors. The major genetic contributor is the major histocompatibility complex (MHC)-encoded (human leukocyte antigen) HLA-B27 first described in 1973. However, the presence of HLA-B27 alone is not sufficient for disease development. A few non-MHC candidate genes have also been implicated in AS. However, progress in AS-related genomic research has been hampered by the extreme genetic heterogeneity of the human population, and the fact that human individuals cannot be genetically manipulated. Animal models are invaluable tools for understanding the mechanisms of human (autoimmune) disorders. Proteoglycan [PG]-induced spondyloarthritis (PGISpA), which develops in genetically homogeneous inbred BALB/c mice upon immunization with cartilage PG, is permissive to genetic manipulation. PGISpA is the only inducible model of AS in which spine involvement is associated with sacroiliitis, uveitis and frequently but not always peripheral arthritis. Using genome-wide screening, we identified Pgis2 (on chromosome [Chr] 2, syntenic with human Chr9:SPA locus) as one of the most prominent quantitative trait loci (QTLs) that affects both the onset and severity of spondylitis. We generated congenic lines in which relevant Chr2 intervals from PGISpA-resistant DBA/2 were inserted into the PGISpA-susceptible BALB/c genome. A special value of this combination is that both mouse strains carry the same MHC (H2d), thus, the major genetic predisposition genetic factor (~50%) is silenced. These congenic strains were tested for association of PGISpA susceptibility. The protective/suppressive regions of DBA/2 origin were stepwise reduced to ~3.0 mega basepair (Mbp) size, and 4 parents (DBA/2 and BALB/c) and 6 PGISpA-resistant and PGISpA-susceptible congenic mice were sequenced using a next-generation high-throughput sequencing method. The 10 genomic sequences were aligned to each other and to the reference (database) C57Bl/6 (B6) genomic sequence. Over 93% of indels (mutations: polymorphisms, deletions, insertions found in DBA/2 alleles) were localized in 3 relatively small genomic regions affecting a cluster of 3 (Gpr107-Nsc1-Hmcn2) genes and 2 other genes (St6galnac6 and Lmx1b) and their intergenic regions. Although non-sense mutations were not found in the coding sequences of these 5 genes, an unusually high number of mutations occurred in their intronic and intergenic sequences. We hypothesize that these genes contribute to PGISpA in BALB/c, and protect disease development in DBA/2 mice or in congenic strains carrying DBA/2 alleles of these regions. In vivo results with IVSC strains also suggest that genes in the three mutated Chr2 subregions have gene-gene (epistatic) interactions, i.e., they may act synergistically to completely prevent the development of PGISpA. Moreover, St6galnac6 was highly expressed in inflamed soft tissues of the spine (in areas of peridiscitis and enthesitis) in BALB/c mice, and the St6galnac6 promoter was heavily methylated (silenced) in PGISpA- resistant DBA/2 mice, indicating that epigenetic alterations have profound effects on the regulation of expression of this gene. We further hypothesize that both genetic mutations and epigenetic alterations are involved in the pathological mechanisms of PGISpA.
In Aim 1, we will narrow, and (if possible) separate genes of the Gpr107-Nsc1-Hmcn2 cluster via new recombinations by generating new IVSC mice (with even smaller Chr intervals or single genes), which will be intercrossed with St6galnac6- and/or Lmx1b-specific IVSC mice to test epistatic effect among the mutated genes.
In Aim 1 B, we will generate transgenic mice carrying (non-mutated) Hmcn2 alone or in combination with Gpr107 and Nsc1 of non-DBA/2 origin (B6) to reverse the protective effect (i.e., restore susceptibility) of any of this region on PGISpA. In Ai 2, we will use St6galnac6-deficient and viable Lmx1b+/- mice to test the independent or combined effects of these genes on PGISpA, and Lmx1b-LacZ transgenic mice to detect expression of the Lmx1b transcription factor during the progression of SpA. Because epigenetic alterations may be important etiologic components of autoimmune diseases, and St6galnac6 expression appears to play a key role in early inflammatory reactions, in Aim 3 we will focus on how the altered methylation of the St6galnac6 gene affects the development of inflammation of in PGISpA.
The successful accomplishment of this project may result in significant information. We combine expertise and knowledge in two areas of research in genetics of arthritis/spondylitis and epigenomics. During the last 15 years, we developed and characterized spondyloarthritis in mice, which resembles human ankylosing spondylitis (AS). The PI and co-investigators have extensive experience in the appropriate areas, and they hypothesized that epigenomic alteration may control AS, a disease which affects approximately 0.2-0.5% of the human population. The modification of the regulatory regions of a gene or gene clusters in the genome (by the presence of certain chemical groups, e.g., methylation) may affect the expression of a gene. These alterations of the genome can control the function of immune cells, thus these modifications may be responsible for autoimmune responses. This form of the modified genome (epigenome) cannot be determined or identified by standard genome-screening or sequencing methods. This may be the reason that the etiology of AS is still completely unknown more than 35 years after the first genetic association was found between AS and the major histocompatibility gene (HLA-B27) and single nucleotide polymorphism (SNP) studies achieved the limit: localized disease-associated genomic regions, but not genes. Because this concept has never been tested in AS, and the research team has a unique model of spondylitis which mimics human AS, the central hypothesis first will be tested in a genetically homogenous animal model system, and then in well-controlled (registered) human patients with AS. Recently, this team used second-generation high-throughput sequencing method, and found three genomic regions (5 genes of the 62 within 3 million basepair region) which have high density of mutations.
