Genetics of rheumatoid arthritis.

In the past four decades, a plethora of genetic association studies have been carried out in cohorts of patients with rheumatoid arthritis. These studies have highlighted key aspects of disease pathogenesis and suggested causal mechanisms. In this review, we discuss major advances in our understanding of the genetic architecture of rheumatoid arthritis susceptibility, severity and treatment response and explain how genetics supports current models of disease pathogenesis and outcome. We outline future research directions, like Mendelian randomisation, and present a number of potential avenues for clinical translation, including risk and outcome prediction, patient stratification into treatment response groups and pharmacological applications.

Development and application of EpitopeScan, a Python3 toolset for mutation tracking in SARS-CoV-2 immunogenic epitopes.

Introduction: Outbreaks of coronaviruses and especially the recent COVID-19 pandemic emphasize the importance of immunological research in this area to mitigate the effect of future incidents. Bioinformatics approaches are capable of providing multisided insights from virus sequencing data, although currently available software options are not entirely suitable for a specific task of mutation surveillance within immunogenic epitopes of SARS-CoV-2. Method: Here, we describe the development of a mutation tracker, EpitopeScan, a Python3 package with command line and graphical user interface tools facilitating the investigation of the mutation dynamics in SARS-CoV-2 epitopes via analysis of multiple-sequence alignments of genomes over time. We provide an application case by examining three Spike protein-derived immunodominant CD4+ T-cell epitopes restricted by HLA-DRB1*04:01, an allele strongly associated with susceptibility to rheumatoid arthritis (RA). Mutations in these peptides are relevant for immune monitoring of CD4+ T-cell responses against SARS-CoV-2 spike protein in patients with RA. The analysis focused on 2.3 million SARS-CoV-2 genomes sampled in England. Results: We detail cases of epitope conservation over time, partial loss of conservation, and complete divergence from the wild type following the emergence of the N969K Omicron-specific mutation in November 2021. The wild type and the mutated peptide represent potential candidates to monitor variant-specific CD4+ T-cell responses. EpitopeScan is available via GitHub repository https://github.com/Aleksandr-biochem/EpitopeScan.

Using Polygenic Risk Scores to Aid Diagnosis of Patients With Early Inflammatory Arthritis: Results From the Norfolk Arthritis Register.

OBJECTIVE: There is growing evidence that genetic data are of benefit in the rheumatology outpatient setting by aiding early diagnosis. A genetic probability tool (G-PROB) has been developed to aid diagnosis has not yet been tested in a real-world setting. Our aim was to assess whether G-PROB could aid diagnosis in the rheumatology outpatient setting using data from the Norfolk Arthritis Register (NOAR), a prospective observational cohort of patients presenting with early inflammatory arthritis. METHODS: Genotypes and clinician diagnoses were obtained from patients from NOAR. Six G-probabilities (0%-100%) were created for each patient based on known disease-associated odds ratios of published genetic risk variants, each corresponding to one disease of rheumatoid arthritis, systemic lupus erythematosus, psoriatic arthritis, spondyloarthropathy, gout, or "other diseases." Performance of the G-probabilities compared with clinician diagnosis was assessed. RESULTS: We tested G-PROB on 1,047 patients. Calibration of G-probabilities with clinician diagnosis was high, with regression coefficients of 1.047, where 1.00 is ideal. G-probabilities discriminated clinician diagnosis with pooled areas under the curve (95% confidence interval) of 0.85 (0.84-0.86). G-probabilities <5% corresponded to a negative predictive value of 96.0%, for which it was possible to suggest >2 unlikely diseases for 94% of patients and >3 for 53.7% of patients. G-probabilities >50% corresponded to a positive predictive value of 70.4%. In 55.7% of patients, the disease with the highest G-probability corresponded to clinician diagnosis. CONCLUSION: G-PROB converts complex genetic information into meaningful and interpretable conditional probabilities, which may be especially helpful at eliminating unlikely diagnoses in the rheumatology outpatient setting.

Automated clustering reveals CD4+ T cell subset imbalances in rheumatoid arthritis.

Background: Despite the report of an imbalance between CD4+ T helper (Th) cell subsets in rheumatoid arthritis (RA), patient stratification for precision medicine has been hindered by the discovery of ever more Th cell subsets, as well as contradictory association results. Objectives: To capture previously reported Th imbalance in RA with deep immunophenotyping techniques; to compare hypothesis-free unsupervised automated clustering with hypothesis-driven conventional biaxial gating and explore if Th cell heterogeneity accounts for conflicting association results. Methods: Unstimulated and stimulated peripheral blood mononuclear cells from 10 patients with RA and 10 controls were immunophenotyped with a 37-marker panel by mass cytometry (chemokine receptors, intra-cellular cytokines, intra-nuclear transcription factors). First, conventional biaxial gating and standard definitions of Th cell subsets were applied to compare subset frequencies between cases and controls. Second, unsupervised clustering was performed with FlowSOM and analysed using mixed-effects modelling of Associations of Single Cells (MASC). Results: Conventional analytical techniques fail to identify classical Th subset imbalance, while unsupervised automated clustering, by allowing for unusual marker combinations, identified an imbalance between pro- and anti-inflammatory subsets. For example, a pro-inflammatory Th1-like (IL-2+ T-bet+) subset and an unconventional but pro-inflammatory IL-17+ T-bet+ subset were significantly enriched in RA (odds ratio=5.7, p=2.2 x 10-3; odds ratio=9.7, p=1.5x10-3, respectively). In contrast, a FoxP3+ IL-2+ HLA-DR+ Treg-like subset was reduced in RA (odds ratio=0.1, p=7.7x10-7). Conclusion: Taking an unbiased approach to large dataset analysis using automated clustering algorithms captures non-canonical CD4+ T cell subset imbalances in RA blood.

Pre-treatment calprotectin (MRP8/14) provides no added value to testing CRP alone in terms of predicting response to TNF inhibitors in rheumatoid arthritis in a post hoc analysis.

OBJECTIVES: The inflammatory protein calprotectin (MRP8/14) has been identified as a promising biomarker of treatment response in rheumatoid arthritis (RA). Our aim was to test MRP8/14 as a biomarker of response to tumour necrosis factor (TNF)-inhibitors in the largest RA cohort to date and to compare with C-reactive protein (CRP). METHODS: Serum MRP8/14 was measured in 470 patients with RA about to commence treatment with adalimumab (n=196) or etanercept (n=274). Additionally, MRP8/14 was measured in the 3-month sera of 179 adalimumab-treated patients. Response was determined using European League against Rheumatism (EULAR) response criteria calculated using the traditional 4-component (4C) DAS28-CRP and alternate validated versions using 3-component (3C) and 2-component (2C), clinical disease activity index (CDAI) improvement criteria and change in individual outcome measures. Logistic/linear regression models were fitted for response outcome. RESULTS: In the 3C and 2C models, patients with RA were 1.92 (CI: 1.04 to 3.54) and 2.03 (CI: 1.09 to 3.78) times more likely to be classified as EULAR responders if they had high (75th quartile) pre-treatment levels of MRP8/14 compared with low (25th quartile). No significant associations were observed for the 4C model. When only using CRP as a predictor, in the 3C and 2C analyses, patients above the 75th quartile were 3.79 (CI: 1.81 to 7.93) and 3.58 (CI: 1.74 to 7.35) times more likely to be EULAR responders and addition of MRP8/14 did not significantly improve model fit (p values=0.62 and 0.80, respectively). No significant associations were observed in the 4C analysis. Exclusion of CRP from the outcome measure (CDAI) did not result in any significant associations with MRP8/14 (OR 1.00 (CI: 0.99 to 1.01), suggesting that the associations were due to the correlation with CRP and that there is no additional utility of MRP8/14 beyond use of CRP in patients with RA starting TNFi therapy. CONCLUSION: Beyond correlation with CRP, we found no evidence to suggest that MRP8/14 explains additional variability in response to TNFi in patients with RA over and above CRP alone.

Multi-ancestry genome-wide association analyses identify novel genetic mechanisms in rheumatoid arthritis.

Rheumatoid arthritis (RA) is a highly heritable complex disease with unknown etiology. Multi-ancestry genetic research of RA promises to improve power to detect genetic signals, fine-mapping resolution and performances of polygenic risk scores (PRS). Here, we present a large-scale genome-wide association study (GWAS) of RA, which includes 276,020 samples from five ancestral groups. We conducted a multi-ancestry meta-analysis and identified 124 loci (P < 5 × 10-8), of which 34 are novel. Candidate genes at the novel loci suggest essential roles of the immune system (for example, TNIP2 and TNFRSF11A) and joint tissues (for example, WISP1) in RA etiology. Multi-ancestry fine-mapping identified putatively causal variants with biological insights (for example, LEF1). Moreover, PRS based on multi-ancestry GWAS outperformed PRS based on single-ancestry GWAS and had comparable performance between populations of European and East Asian ancestries. Our study provides several insights into the etiology of RA and improves the genetic predictability of RA.

HLA-DRB1 haplotypes predict cardiovascular mortality in inflammatory polyarthritis independent of CRP and anti-CCP status.

BACKGROUND: Haplotypes defined by amino acids at HLA-DRB1 positions 11, 71 and 74 associated with susceptibility to rheumatoid arthritis (RA) are associated with radiological outcome, anti-TNF response and all cause-mortality in RA. RA is associated with cardiovascular (CV) morbidity and mortality, but the increased prevalence of risk factors of CV disease in RA only partially explains this association. The aim of this study was to investigate whether amino acids at positions 11, 71 and 74 of HLA-DRB1 are associated with cardiovascular (CV) mortality in inflammatory polyarthritis (IP). METHODS: The Norfolk Arthritis Register (NOAR) is an incidence register of IP: recruitment 1990-2007, final follow-up 2011. Two thousand five hundred fourteen patients had available genetic and mortality data. Amino acids at positions 11, 71 and 74 of HLA-DRB1 were determined. Univariate Cox proportional hazard models were applied to assess the association of genetic markers and both all-cause mortality and cardiovascular mortality. RESULTS: Among 2514 participants, 643 (25.6%) died during the study, and 343 (53.3%) of these deaths were attributed to CV causes. One thousand six hundred fifty (65.6%) participants were female, 709 (32.3%) were anti-CCP-positive and the median age of participants was 54. HLA-DRB1 haplotypes associated with susceptibility to rheumatoid arthritis (RA) consistently show the same magnitude and direction of association for overall and CV mortality in IP. For example, the SEA-haplotype, associated with the lowest susceptibility to RA, and the best radiographic outcome, was found to be associated with decreased CV mortality (HR 0.67, 95% CI 0.47, 0.91, p=0.023). Mediation analysis revealed associations were independent of anti-CCP status. CONCLUSIONS: HLA-DRB1 haplotypes associated with susceptibility to RA also predispose to increased risk of CV mortality in IP, independent of known CV risk factors. Associations were independent of anti-CCP status, which suggests in the future, genetic factors will add to the prediction of risk of cardiovascular mortality beyond serological markers.

Pre-defined gene co-expression modules in rheumatoid arthritis transition towards molecular health following anti-TNF therapy.

BACKGROUND: No reliable biomarkers to predict response to TNF inhibitors (TNFi) in RA patients currently exist. The aims of this study were to replicate changes in gene co-expression modules that were previously reported in response to TNFi therapy in RA; to test if changes in module expression are specific to TNFi therapy; and to determine whether module expression transitions towards a disease-free state in responding patients. METHOD: Published transcriptomic data from the whole blood of disease-free controls (n = 10) and RA patients, treated with the TNFi adalimumab (n = 70) or methotrexate (n = 85), were studied. Treatment response was assessed using the EULAR response criteria following 3 or 6 months of treatment. Change in transcript expression between pre- and post-treatment was recorded for previously defined modules. Linear mixed models tested whether modular expression after treatment transitioned towards a disease-free state. RESULTS: For 25 of the 27 modules, change in expression between pre- and post-treatment in the adalimumab cohort replicated published findings. Of these 25 modules, six transitioned towards a disease-free state by 3 months (P < 0.05), irrespective of clinical response. One module (M3.2), related to inflammation and TNF biology, significantly correlated with response to adalimumab. Similar patterns of modular expression, with reduced magnitude, were observed in the methotrexate cohort. CONCLUSION: This study provides independent validation of changes in module expression in response to therapy in RA. However, these effects are not specific to TNFi. Further studies are required to determine whether specific modules could assist molecular classification of therapeutic response.

Using the Immunophenotype to Predict Response to Biologic Drugs in Rheumatoid Arthritis.

Tumour necrosis factor (TNF)-α is a key mediator of inflammation in rheumatoid arthritis, and its discovery led to the development of highly successful anti-TNF therapy. Subsequently, other biologic drugs targeting immune pathways, namely interleukin-6 blockade, B cell depletion, and T cell co-stimulation blockade, have been developed. Not all patients respond to a biologic drug, leading to a knowledge gap between biologic therapies available and the confident prediction of response. So far, genetic studies have failed to uncover clinically informative biomarkers to predict response. Given that the targets of biologics are immune pathways, immunological study has become all the more pertinent. Furthermore, advances in single-cell technology have enabled the characterization of many leucocyte subsets. Studying the blood immunophenotype may therefore, define biomarker profiles relevant to each individual patient's disease and treatment outcome. This review summarises our current understanding of how immune biomarkers might be able to predict treatment response to biologic drugs.

Validity of a two-component imaging-derived disease activity score for improved assessment of synovitis in early rheumatoid arthritis.

OBJECTIVES: Imaging of joint inflammation provides a standard against which to derive an updated DAS for RA. Our objectives were to develop and validate a DAS based on reweighting the DAS28 components to maximize association with US-assessed synovitis. METHODS: Early RA patients from two observational cohorts (n = 434 and n = 117) and a clinical trial (n = 59) were assessed at intervals up to 104 weeks from baseline; all US scans were within 1 week of clinical exam. There were 899, 163 and 183 visits in each cohort. Associations of combined US grey scale and power Doppler scores (GSPD) with 28 tender joint count and 28 swollen joint count (SJC28), CRP, ESR and general health visual analogue scale were examined in linear mixed model regressions. Cross-validation evaluated model predictive ability. Coefficients learned from training data defined a re-weighted DAS28 that was validated against radiographic progression in independent data (3037 observations; 717 patients). RESULTS: Of the conventional DAS28 components only SJC28 and CRP were associated with GSPD in all three development cohorts. A two-component model including SJC28 and CRP outperformed a four-component model (R2 = 0.235, 0.392, 0.380 vs 0.232, 0.380, 0.375, respectively). The re-weighted two-component DAS28CRP outperformed conventional DAS28 definitions in predicting GSPD (Δtest log-likelihood <-2.6, P < 0.01), Larsen score and presence of erosions. CONCLUSION: A score based on SJC28 and CRP alone demonstrated stronger associations with synovitis and radiographic progression than the original DAS28 and should be considered in research on pathophysiological manifestations of early RA. Implications for clinical management of RA remain to be established.

Analysis of SNP-SNP interactions and bone quantitative ultrasound parameter in early adulthood.

BACKGROUND: Osteoporosis individual susceptibility is determined by the interaction of multiple genetic variants and environmental factors. The aim of this study was to conduct SNP-SNP interaction analyses in candidate genes influencing heel quantitative ultrasound (QUS) parameter in early adulthood to identify novel insights into the mechanism of disease. METHODS: The study population included 575 healthy subjects (mean age 20.41; SD 2.36). To assess bone mass QUS was performed to determine Broadband ultrasound attenuation (BUA, dB/MHz). A total of 32 SNPs mapping to loci that have been characterized as genetic markers for QUS and/or BMD parameters were selected as genetic markers in this study. The association of all possible SNP pairs with QUS was assessed by linear regression and a SNP-SNP interaction was defined as a significant departure from additive effects. RESULTS: The pairwise SNP-SNP analysis showed multiple interactions. The interaction comprising SNPs rs9340799 and rs3736228 that map in the ESR1 and LRP5 genes respectively, revealed the lowest p value after adjusting for confounding factors (p-value = 0.001, ß (95% CI) = 14.289 (5.548, 23.029). In addition, our model reported others such as TMEM135-WNT16 (p = 0.007, ß(95%CI) = 9.101 (2.498, 15.704), ESR1-DKK1 (p = 0.012, ß(95%CI) = 13.641 (2.959, 24.322) or OPG-LRP5 (p = 0.012, ß(95%CI) = 8.724 (1.936, 15.512). However, none of the detected interactions remain significant considering the Bonferroni significance threshold for multiple testing (p<0.0001). CONCLUSION: Our analysis of SNP-SNP interaction in candidate genes of QUS in Caucasian young adults reveal several interactions, especially between ESR1 and LRP5 genes, that did not reach statistical significance. Although our results do not support a relevant genetic contribution of SNP-SNP epistatic interactions to QUS in young adults, further studies in larger independent populations would be necessary to support these preliminary findings.

Genotypic variability based association identifies novel non-additive loci DHCR7 and IRF4 in sero-negative rheumatoid arthritis.

Sero-negative rheumatoid arthritis (RA) is a highly heterogeneous disorder with only a few additive loci identified to date. We report a genotypic variability-based genome-wide association study (vGWAS) of six cohorts of sero-negative RA recruited in Europe and the US that were genotyped with the Immunochip. A two-stage approach was used: (1) a mixed model to partition dichotomous phenotypes into an additive component and non-additive residuals on the liability scale and (2) the Levene's test to assess equality of the residual variances across genotype groups. The vGWAS identified rs2852853 (P = 1.3e-08, DHCR7) and rs62389423 (P = 1.8e-05, near IRF4) in addition to two previously identified loci (HLA-DQB1 and ANKRD55), which were all statistically validated using cross validation. DHCR7 encodes an enzyme important in cutaneous synthesis of vitamin D and DHCR7 mutations are believed to be important for early humans to adapt to Northern Europe where residents have reduced ultraviolet-B exposure and tend to have light skin color. IRF4 is a key locus responsible for skin color, with a vitamin D receptor-binding interval. These vGWAS results together suggest that vitamin D deficiency is potentially causal of sero-negative RA and provide new insights into the pathogenesis of the disorder.

Genetics of rheumatoid arthritis susceptibility, severity, and treatment response.

A decade after the first genome-wide association study in rheumatoid arthritis (RA), a plethora of genetic association studies have been published on RA and its clinical or serological subtypes. We review the major milestones in the study of the genetic architecture of RA susceptibility, severity, and response to treatment. We set the scientific context necessary for non-geneticists to understand the potential clinical applications of human genetics and its significance for a stratified approach to the management of RA in the future.

A genetic risk score composed of rheumatoid arthritis risk alleles, HLA-DRB1 haplotypes, and response to TNFi therapy - results from a Swedish cohort study.

BACKGROUND: To prevent debilitating and irreversible joint damage, rheumatoid arthritis (RA) is often treated with tumor necrosis factor inhibitor (TNFi), but many patients do not respond to this costly therapy. Few predictors for response are known, and it has been proposed that genetic factors which influence the development of RA may also influence disease severity and response to therapy. Several previous studies have attempted to confirm this but results remain inconclusive. We expand on previous studies by including more RA risk alleles, and maximize power by combining them into a genetic risk score. METHOD: We linked genotyped RA patients from the Epidemiological Investigation of Rheumatoid Arthritis study to the Swedish Rheumatology Quality Register, identifying patients who started a TNFi as their first biological disease-modifying anti-rheumatic drug, with a return visit within 2-8 months after treatment start (N = 867). We calculated risk scores from 76 established RA risk SNPs, and four HLA-DRB1 amino acid positions, and tested whether risk scores or individual genetic risk factors could predict the European League Against Rheumatism (EULAR) response. RESULTS: We found no association between any of the risk scores or HLA-DRB1 haplotypes and EULAR response, neither overall nor stratified by anti-citrullinated protein/peptide antibody (ACPA) status. When evaluating each of the 76 SNPs, we found that the number of SNPs presenting significant associations was not higher than expected by chance (5/76 SNPs had p < 0.05 in ACPA-positive RA, 4/76 in ACPA-negative RA). CONCLUSION: Overall, known RA risk SNPs do not predict response to TNFi, either individually or when combined into a risk score. This does not support the hypothesis that genes influencing RA onset would also influence its prognosis and treatment response.

HLA-DRB1 Amino Acid Positions 11/13, 71, and 74 Are Associated With Inflammation Level, Disease Activity, and the Health Assessment Questionnaire Score in Patients With Inflammatory Polyarthritis.

OBJECTIVE: Rheumatoid arthritis (RA) susceptibility HLA-DRB1 haplotypes based on amino acid positions 11/13, 71, and 74 predict radiographic damage. The mechanism of action is unknown, but it may be mediated by inflammation. We undertook this study to systematically investigate the effect of these amino acids on nonradiographic measures of disease activity/outcomes. METHODS: We tested the association of RA susceptibility HLA-DRB1 amino acids with the C-reactive protein (CRP) level, the tender joint count (TJC), the swollen joint count (SJC), the Disease Activity Score in 28 joints (DAS28), and the Health Assessment Questionnaire (HAQ) score in the Norfolk Arthritis Register (NOAR) and Early Rheumatoid Arthritis Study (ERAS) cohorts. Longitudinal modeling of disease activity/outcomes was performed using generalized linear latent and mixed models. Mediation analysis was performed using directed acyclic graphs to investigate the paths from genetic factors to outcome. RESULTS: A total of 2,158 patients were available for analysis in the NOAR cohort. Valine at position 11 showed the strongest association with the CRP level (P = 2.21 × 10-6 ), the SJC (P = 7.51 × 10-6 ), and the DAS28 (P = 0.002); it was marginally associated with the HAQ score (P = 0.044) but not with the TJC. The same amino acid and haplotype risk hierarchy observed for susceptibility and radiographic severity was observed for the CRP level and nonradiographic measures of disease activity/outcome, apart from the TJC. The results were replicated in the ERAS cohort. The effect of valine at position 11 on the SJC was mainly mediated by anti-citrullinated protein antibody status, the effect of which was mainly mediated by inflammation; however, the effect of valine at position 11 was also independent of the CRP level (P = 1.6 × 10-4 ). CONCLUSION: Genetic markers of RA susceptibility located within HLA-DRB1 determine the levels of clinical and systemic inflammation independently, and also determine all objective measures of disease activity and outcome.

Association Between Genetic Variation in FOXO3 and Reductions in Inflammation and Disease Activity in Inflammatory Polyarthritis.

OBJECTIVE: Genetic variation in FOXO3 (tagged by rs12212067) has been associated with a milder course of rheumatoid arthritis (RA) and shown to limit monocyte-driven inflammation through a transforming growth factor ß1-dependent pathway. This genetic association, however, has not been consistently observed in other RA cohorts. We sought to clarify the contribution of FOXO3 to prognosis in RA by combining detailed analysis of nonradiographic disease severity measures with an in vivo model of arthritis. METHODS: Collagen-induced arthritis, the most commonly used mouse model of RA, was used to assess how Foxo3 contributes to arthritis severity. Using clinical, serologic, and biochemical methods, the arthritis that developed in mice carrying a loss-of-function mutation in Foxo3 was compared with that which occurred in littermate controls. The association of rs12212067 with nonradiographic measures of RA severity, including the C-reactive protein level, the swollen joint count, the tender joint count, the Disease Activity Score in 28 joints, and the Health Assessment Questionnaire score, were modeled longitudinally in a large prospective cohort of patients with early RA. RESULTS: Loss of Foxo3 function resulted in more severe arthritis in vivo (both clinically and histologically) and was associated with higher titers of anticollagen antibodies and interleukin-6 in the blood. Similarly, rs12212067 (a single-nucleotide polymorphism that increases FOXO3 transcription) was associated with reduced inflammation, both biochemically and clinically, and with lower RA activity scores. CONCLUSION: Consistent with its known role in restraining inflammatory responses, FOXO3 limits the severity of in vivo arthritis and, through genetic variation that increases its transcription, is associated with reduced inflammation and disease activity in RA patients, effects that result in less radiographic damage.

Major histocompatibility complex harbors widespread genotypic variability of non-additive risk of rheumatoid arthritis including epistasis.

Genotypic variability based genome-wide association studies (vGWASs) can identify potentially interacting loci without prior knowledge of the interacting factors. We report a two-stage approach to make vGWAS applicable to diseases: firstly using a mixed model approach to partition dichotomous phenotypes into additive risk and non-additive environmental residuals on the liability scale and secondly using the Levene's (Brown-Forsythe) test to assess equality of the residual variances across genotype groups per marker. We found widespread significant (P < 2.5e-05) vGWAS signals within the major histocompatibility complex (MHC) across all three study cohorts of rheumatoid arthritis. We further identified 10 epistatic interactions between the vGWAS signals independent of the MHC additive effects, each with a weak effect but jointly explained 1.9% of phenotypic variance. PTPN22 was also identified in the discovery cohort but replicated in only one independent cohort. Combining the three cohorts boosted power of vGWAS and additionally identified TYK2 and ANKRD55. Both PTPN22 and TYK2 had evidence of interactions reported elsewhere. We conclude that vGWAS can help discover interacting loci for complex diseases but require large samples to find additional signals.

Replication of Associations of Genetic Loci Outside the HLA Region With Susceptibility to Anti-Cyclic Citrullinated Peptide-Negative Rheumatoid Arthritis.

OBJECTIVE: Genetic polymorphisms within the HLA region explain only a modest proportion of anti-cyclic citrullinated peptide (anti-CCP)-negative rheumatoid arthritis (RA) heritability. However, few non-HLA markers have been identified so far. This study was undertaken to replicate the associations of anti-CCP-negative RA with non-HLA genetic polymorphisms demonstrated in a previous study. METHODS: The Rheumatoid Arthritis Consortium International densely genotyped 186 autoimmune-related regions in 3,339 anti-CCP-negative RA patients and 15,870 controls across 6 different populations using the Illumina ImmunoChip array. We performed a case-control replication study of the anti-CCP-negative markers with the strongest associations in that discovery study, in an independent cohort of anti-CCP-negative UK RA patients. Individuals from the arcOGEN Consortium and Wellcome Trust Case Control Consortium were used as controls. Genotyping in cases was performed using Sequenom MassArray technology. Genome-wide data from controls were imputed using the 1000 Genomes Phase I integrated variant call set release version 3 as a reference panel. RESULTS: After genotyping and imputation quality control procedures, data were available for 15 non-HLA single-nucleotide polymorphisms in 1,024 cases and 6,348 controls. We confirmed the known markers ANKRD55 (meta-analysis odds ratio [OR] 0.80; P = 2.8 × 10(-13) ) and BLK (OR 1.13; P = 7.0 × 10(-6) ) and identified new and specific markers of anti-CCP-negative RA (prolactin [PRL] [OR 1.13; P = 2.1 × 10(-6) ] and NFIA [OR 0.85; P = 2.5 × 10(-6) ]). Neither of these loci is associated with other common, complex autoimmune diseases. CONCLUSION: Anti-CCP-negative RA and anti-CCP-positive RA are genetically different disease subsets that only partially share susceptibility factors. Genetic polymorphisms located near the PRL and NFIA genes represent examples of genetic susceptibility factors specific for anti-CCP-negative RA.

Loci associated with N-glycosylation of human IgG are not associated with rheumatoid arthritis: a Mendelian randomisation study.

OBJECTIVES: A recent study identified 16 genetic variants associated with N-glycosylation of human IgG. Several of the genomic regions where these single nucleotide polymorphisms (SNPs) reside have also been associated with autoimmune disease (AID) susceptibility, suggesting there may be pleiotropy (genetic sharing) between loci controlling both N-glycosylation and AIDs. We investigated this by testing variants associated with levels of IgG N-glycosylation for association with rheumatoid arthritis (RA) susceptibility using a Mendelian randomisation study, and testing a subset of these variants in a less well-powered study of treatment response and severity. METHODS: SNPs showing association with IgG N-glycosylation were analysed for association with RA susceptibility in 14 361 RA cases and 43 923 controls. Five SNPs were tested for association with response to anti-tumour necrosis factor (TNF) therapy in 1081 RA patient samples and for association with radiological disease severity in 342 patients. RESULTS: Only one SNP (rs9296009) associated with N-glycosylation showed an association (p=6.92×10(-266)) with RA susceptibility, although this was due to linkage disequilibrium with causal human leukocyte antigen (HLA) variants. Four regions of the genome harboured SNPs associated with both traits (shared loci); although statistical analysis indicated that the associations observed for the two traits are independent. No SNPs showed association with response to anti-TNF therapy. One SNP rs12342831 was modestly associated with Larsen score (p=0.05). CONCLUSIONS: In a large, well-powered cohort of RA patients, we show SNPs driving levels of N-glycosylation have no association with RA susceptibility, indicating colocalisation of associated SNPs are not necessarily indicative of a shared genetic background or a role for glycosylation in disease susceptibility.