During the current reporting period we have focused on 3 major projects: 1) Genetics of Scleroderma in the African-American Population Through a collaborative group denoted GRASP (Genome Research in African American Scleroderma Patients), we have secured the largest collection of African-American SSc patients ever assembled. GRASP consists of 23 centers outside of the NIH. With help from the Scleroderma Research Foundation, we have collected DNA samples from 1170 African-American SSc patients, and we have identified 1039 antinuclear antibody (ANA) negative African-American controls provided by Charles Rotimi. Phenotypic characteristics of the GRASP cohort were summarized in the report from the previous reporting period and were published during the current reporting period. Consistent with the underlying assumptions of this study, GRASP patients exhibited more severe phenotypes than a control European ancestry cohort. 1008 SSc patients and 1008 SSc controls have been genotyped on 2 Illumina GWAS arrays. We have completed whole-exome sequencing (WES) on 400 patients and 482 controls. Variants identified in the discovery phase by GWAS and WES are now undergoing sequencing in a replication study of 600 SSc patients and 360 controls. The Illumina MEGA array was used to genotype 934 patients and 946 controls to conduct the first GWAS of SSc in African Americans. A total of 1.7 million variants were genotyped and after quality control filtering 1.4 million variants remained and were imputed into the 1000 Genomes Phase 3 v5 reference panel. Admixture and principal component analysis confirmed that the patients and controls were well-matched. GWAS identified class II genes as the strongest risk factor in SSc susceptibility and discovered two non-HLA, African ancestry-specific loci, IFT43/TGFB3 and FSD2/HOMER2. rs35915063 was the top scoring variant in the HLA-DQB1 gene (P=2.2 x 10e-17, OR=1.96). Stepwise conditional regression analysis using an additive model revealed the DQA1 variant rs9271620 and the DPB1 variant rs2071354 as independent risk variants. Upon performing eQTL analysis in whole blood from the GTEx RNA sequencing data, the African-American SSc-associated alleles of these 3 risk variants were associated with increased expression of HLA-DQB2/DQA2 and decreased expression of HLA-DQB1 and HLA-DQA1 genes. rs59063398 showed the strongest non-MHC association (P=2.0 x 10e-8, OR=0.47). This variant is part of an evolutionarily conserved haploblock in the IFT43 gene region that is African ancestry-specific. On performing eQTL analysis of the GTEx RNA sequencing data of sun-exposed skin of the lower leg, we observed decreased expression of TGFB3 associated with the minor variant in the IFT43/TGFB3 region (P=0.02). We used MHC region genotypes from the MEGA array to impute classical HLA types and analyzed their associations with scleroderma in the GRASP cohort. We extracted the genotypes of 25,256 markers from 20 Mb encompassing the greater MHC region to impute classical HLA types using the HLA*IMP:03 web server. The most significantly SSc-associated HLA type was a predominantly African allele, HLA-DRB1*08:04, OR=2.95. Regression analysis conditioning on the disease-associated alleles identified another African DRB1 allele, *11:02, as well as HLA-DPB1*13:01 and HLA-DRB4*01:01 as independent contributors to disease risk, but no association was found for MHC class I alleles. 34.6 percent of GRASP cases carry either DRB1*08:04 or *11:02 compared with 16.3 percent of controls. On stratifying the SSc samples by autoantibodies, very strong and specific HLA allele associations were identified, with HLA-DRB1*0804 increasing risk by 7.2-fold in the anti-fibrillarin antibody subset of African-American SSc and HLA-DPB1*1301 increasing risk by 4.1-fold in the anti-topoisomerase I antibody subset. In the anti-topoisomerase I antibody subset of SSc patients of European ancestry, HLA-DPB1*1301 increased SSc risk by 12.5-fold. 2) Genetic and Immunologic Studies of PFAPA PFAPA is the most common periodic fever syndrome in children. Patients present with recurrent episodes of high fever with pharyngitis, cervical adenitis, and aphthous ulcers with regular timing. Tonsillectomy is curative in most patients. In order to untangle the mechanism of PFAPA, we are studying the immunologic features of tonsils and genetic variants linked to the disease. We studied the cellular profile of tonsils removed from children with PFAPA during an asymptomatic interval and compared them with tonsils from children with obstructive sleep apnea and with structural oropharyngeal disorders without tonsil pathology. By flow-cytometry, we found fewer T follicular helper (Tfh) cells, increased T follicular regulatory cells, fewer plasma cells and IgG class switching, and increased numbers of NK cells. Gene expression analysis of sorted tonsillar CD4+ T cells by Nanostring revealed increased expression of genes associated with Th1 cell activation in patients with PFAPA. Sorted tonsillar myeloid cells, on the other hand, showed depressed expression of genes involved in innate immune pathway signaling. We also screened 141 Caucasian patients with PFAPA for single nucleotide polymorphisms previously associated with another ulcerative disorder, Behcet's disease. We found that a variant upstream of IL12A was strongly associated with PFAPA (OR=2.06, p=6x10e-7), compared with Caucasian controls from the gnoMAD database. IL12A encodes the p35 subunit of both the inflammatory cytokine, IL-12, and the inhibitory cytokine, IL-35. We hypothesize that patients with PFAPA and this variant have heightened production of IL-12 and IL-35 during the activation and resolution phases of the disease, respectively, leading to excess Th1 polarization and germinal center activation during fever flares, and germinal center and myeloid cell suppression during asymptomatic periods between flares as we find in the tonsils. We believe that this cycling between states of heightened immune activation and suppression lead to the regular timing of episodes in this disease. Functional studies of monocytes from patients with and without the IL12A variant and comparison of cells and cytokines during flare and non-flare periods are underway to test these hypotheses. 3) Trisomy 8-Associated Autoinflammatory Disease We have characterized inflammatory symptoms in a cohort of approximately 50 patients with constitutional trisomy 8 mosaicism or chromosome 8 duplication; nearly half have mucosal ulcerations and 40% have recurrent stereotypical fevers. Mucosal ulcerations are often large and persist for over one week and may be present in the genital or gastrointestinal tract in addition to the mouth. Affected patients have elevated expression of TLR, NF-kappaB, and IL-1-related genes in their whole blood, suggesting activation of myeloid cells. We have also found that CD14+ monocytes in affected patients have a significantly higher percentage of chromosome 8 mosaicism in comparison to B and T cells (copy number of chromosome 8 genes is 2.69 in CD14+ cells vs. 2.48 in CD19+ cells vs. 2.27 in CD3+, p = 0.01 for CD14+ vs. CD19+ and p=0.001 for CD14+ vs. CD3+), suggesting that myeloid cells with trisomy 8 are afforded a survival advantage in addition to exhibiting excess activation. We have treated two patients with anakinra and one with etanercept with significant improvement in symptoms. We are continuing to study the functional effects of trisomy 8 on survival and cytokine production in monocytes. We also plan to assess the frequency of trisomy 8 mosaicism and chromosome 8 duplications among a cohort of patients with Behcet's disease.
Showing the most recent 10 out of 13 publications
