B cells play a significant role in established Rheumatoid Arthritis (RA). However, it is unclear to what extent differentiated B cells are present in joint tissue already at the onset of disease. Here, we studied synovial biopsies (n = 8) captured from untreated patients at time of diagnosis. 3414 index-sorted B cells underwent RNA sequencing and paired tissue pieces were subjected to spatial transcriptomics (n = 4). We performed extensive bioinformatics analyses to dissect the local B cell composition. Select plasma cell immunoglobulin sequences were expressed as monoclonal antibodies and tested by ELISA. Memory and plasma cells were found irrespective of autoantibody status of the patients. Double negative memory B cells were prominent, but did not display a distinct transcriptional profile. The tissue architecture implicate both local B cell maturation via T cell help and plasma cell survival niches with a strong CXCL12–CXCR4 axis. The immunoglobulin sequence analyses revealed clonality between the memory B and plasma cell pools further supporting local maturation. One of the plasma cell-derived antibodies displayed citrulline autoreactivity, demonstrating local autoreactive plasma cell differentiation in joint biopsies captured from untreated early RA. Hence, plasma cell niches are not a consequence of chronic inflammation, but are already present at the time of diagnosis.

Based on the epidemiological link between periodontitis and rheumatoid arthritis (RA), and the unique feature of the periodontal bacterium Porphyromonas gingivalis to citrullinate proteins, it has been suggested that production of anti-citrullinated protein antibodies (ACPA), which are present in a majority of RA patients, may be triggered in the gum mucosa. To address this hypothesis, we investigated the antibody response to a citrullinated P. gingivalis peptide in relation to the autoimmune ACPA response in early RA, and examined citrulline-reactivity in monoclonal antibodies derived from human gingival B cells. Antibodies to a citrullinated peptide derived from P. gingivalis (denoted CPP3) and human citrullinated peptides were analyzed by multiplex array in 2,807 RA patients and 372 controls; associations with RA risk factors and clinical features were examined. B cells from inflamed gingival tissue were single-cell sorted, and immunoglobulin (Ig) genes were amplified, sequenced, cloned and expressed (n=63) as recombinant monoclonal antibodies, and assayed for citrulline-reactivities by enzyme-linked immunosorbent assay. Additionally, affinity-purified polyclonal anti-cyclic-citrullinated peptide (CCP2) IgG, and monoclonal antibodies derived from RA blood and synovial fluid B cells (n=175), were screened for CPP3-reactivity. Elevated anti-CPP3 antibody levels were detected in RA (11%), mainly CCP2+ RA, compared to controls (2%), p<0.0001, with a significant association to HLA-DRB1 shared epitope alleles, smoking and baseline pain, but with low correlation to autoimmune ACPA fine-specificities. Monoclonal antibodies derived from gingival B cells showed cross-reactivity between P. gingivalis CPP3 and human citrullinated peptides, and a CPP3+/CCP2+ clone, derived from an RA blood memory B cell, was identified. Our data support the possibility that immunity to P. gingivalis derived citrullinated antigens, triggered in the inflamed gum mucosa, may contribute to the presence of ACPA in RA patients, through mechanisms of molecular mimicry.

Background: Anti-citrullinated protein antibodies (ACPAs) play an important role in rheumatoid arthritis (RA) pathogenesis. We hypothesized that the effect of these antibodies is mediated by their binding to synovial fibroblasts and inducing an increased mobility of fibroblasts 1 . Objectives: In our study, we analyzed and compared fibroblast modulation by ACPA pools obtained from different patients or by a set of monoclonal ACPAs with different fine specificity that were obtained from different tissue sites. Methods: Synovial fibroblasts were isolated from RA patients synovial tissue biopsies. Individual polyclonal ACPA and control IgGs were purified from sera of four ACPA-positive RA patients by affinity purification on protein G and CCP-2 columns. Monoclonal antibodies were derived from memory B cell isolated from blood2, synovial fluid or bronchoalveolar lavage of RA patients. Whole antibodies and F(ab’)2 fragments were tested in fibroblast migration using IncuCyte live-cell analysis. Blocking experiments were performed with soluble citrullinated proteins in SF migration. Cross-reactivity of the antibodies to citrullinated and acetylated epitopes was tested using PAD inhibitors (Cl-amidine and GSK199), histone acetyltransferases (anacardic acid) and deacetylases (trichostatin A). Binding patterns of monoclonal ACPAs, both whole and F(ab’)2 fragments were analyzed in synovial biopsies obtained from both healthy donors and RA patients. Results: Three out of four tested individual ACPA were able to promote fibroblast migration. Five out of nine tested monoclonal ACPAs stimulated fibroblast migration. One of these antibodies, clone 1325:01B09 is characterized by cross-reactivity to citrullinated, homocitrullinated and acetylated targets. The effect of 1325:01B09 on fibroblast migration was completely abolished by Cl-amidine or by pre-incubating the antibody with citrullinated fibrinogen or histone but not citrullinated enolase or vimentin. Despite the cross-reactivity to acetylated epitopes, neither anacardic acid nor trichostatin A could modulate the 1325:01B09 effect on fibroblast migration. F(ab’)2 fragments of this antibody stimulated fibroblast migration and labelled podoplanin-positive fibroblasts in inflamed RA synovium similarly to the intact antibody, indicating an Fc-independent effect. Conclusion: The effect on fibroblast mobility was likely to be mediated by binding to citrullinated epitopes but not through Fc receptors. Detection of fibroblast modulating ACPAs in majority of RA patients indicated that fibroblasts might be key cellular targets in disease pathogenesis, although individual variability might exist in the composition of ACPA cellular targets. References: [1]Sun M, Rethi B, Krishnamurthy A, et al. Anticitrullinated protein antibodies facilitate migration of synovial tissue-derived fibroblasts. Ann Rheum Dis 2019;78(12):1621-31. doi: 10.1136/annrheumdis-2018-214967 [published Online First: 2019/09/05] [2]Amara K, Lena Israelsson, Ragnhild Stålesen, et al. A Refined Protocol for Identifying Citrulline-specific Monoclonal Antibodies from Single Human B Cells from Rheumatoid Arthritis Patient Material. Bio-protocol 2019;9(16) Disclosure of Interests: Meng Sun: None declared, Bence Réthi: None declared, Akilan Krishnamurthy: None declared, Vijay Joshua: None declared, Alexandra Circiumaru: None declared, Marianne Engström: None declared, Caroline Grönwall: None declared, Vivianne Malmström Grant/research support from: VM has had research grants from Janssen Pharmaceutica, Khaled Amara: None declared, Lars Klareskog: None declared, Heidi Wähämaa: None declared, Anca Catrina: None declared

Background We have previously shown that structural changes, increased tissue citrullination, signs of local inflammation and ACPA are present in the pulmonary compartment of early seropositive RA. These findings suggest a potential role for the lungs in generation of RA-associated autoimmunity. Objectives To identify citrulline reactive B cells in the lung compartment of early untreated RA patients and to generate and characterize the corresponding monoclonal antibodies. Methods Bronchoalveolar lavage (BAL) fluid cells (13 and 22.5 million respectively) were obtained from two early untreated non-smoking ACPA positive RA patients and single CD19+ B cells were sorted by flow cytometry. Immunoglobulin variable region genes were sequenced and expressed to generate recombinant monoclonal antibodies (mAbs). The citrulline reactivity was determined by in-house ELISA against different citrullinated peptides and controls. Results Single sorted CD19+ B cells (n=768) from each patient (RA.1 and RA.2) were processed and the variable region amplification and sequencing yielded 192 (RA.1) and 213 (RA.2) paired heavy and light chain sequences. Based on previously published data showing SHM-introduced N-linked glycosylation in the Fab regions of ACPA [1], 22 sequences from each patient were pre-selected with such features, and monoclonal antibodies were expressed. Two antibodies from each patient were ACPAs as determined by their reactivity against CCP2. The 4 ACPAs have varying ACPA fine specificity against citrullinated enolase, filaggrin, vimentin and fibrinogen peptides (figure 1). Sequence analysis of the heavy chain variable region revealed unique V gene usage of the ACPAs arising from different patients, V4-39 for the 2 mAbs from RA.1 and V3-49 for RA.2. Conclusion We demonstrate for the first time that citrulline-reactive B cells are present in the lung compartment of early untreated RA. Further analysis and functional characterization are needed and ongoing to understand their role in RA. Reference [1] Lloyd, K.A., et al., Variable domain N-linked glycosylation and negative surface charge are key features of monoclonal ACPA: Implications for B-cell selection. Eur J Immunol, 2018. 48(6): p.1030-1045. Disclosure of Interests Vijay Joshua: None declared, Malena Loberg-Haarhaus: None declared, Heidi Wähämaa: None declared, Aase Hensvold: None declared, Magnus Sköld: None declared, Johan Grunewald: None declared, Lars Klareskog Grant/research support from: Yes, but not for the presented study., Vivianne Malmström: None declared, Anca Catrina Grant/research support from: Yes, but not for the presented study.

Background:Malondialdehyde (MDA) is a highly reactive compound produced by lipid-peroxidation in situations associated with oxidative stress. MDA can irreversibly modify proteins residues such as lysine, arginine and histidine. In addition, MDA adducts can further react with acetaldehyde to generate malondialdehyde-acetaldehyde (MAA) modifications. Such modifications can give rise to immunogenic neo-epitopes that are recognized by autoantibodies. In fact, anti-MDA/MAA IgG antibodies are significantly increased in the serum of patients with autoimmune diseases, such as rheumatoid arthritis (RA) (1) and systemic lupus erythematosus (2). Recently, we have shown that anti-MDA/MAA IgG antibodies are able to promote osteoclast (OC) differentiationin vitro(1).Objectives:To investigate the molecular mechanisms triggered by anti-MDA/MAA autoantibodies during osteoclastogenesis.Methods:OCs were generated from monocyte-derived macrophages in the presence of the cytokines RANK-L and M-CSF. The development of OCs was monitored by light microscopy following tartrate-resistant acid phosphatase (TRAP) staining and erosion area on synthetic calcium phosphate-coated plates. Three different recombinant human monoclonal anti-MDA/MAA antibodies, cloned from single synovial B cells of RA patients, control antibodies and Fab fragments of the antibodies were added to OC cultures. Glycolysis was inhibited by 2-deoxyglucose, and Fc-gamma receptor I or II by anti-CD64 or anti-CD16 neutralizing antibodies. IL-8 levels were measured by enzyme linked immunosorbent assay. Cellular metabolism was monitored using Seahorse XF Analyzer (extracellular acidification rate and oxygen consumption) and a colorimetric L-Lactate assay.Results:Lactic acid production correlated with the osteoclastogenetic effect of some but not all anti-MDA/MAA antibodies on OCs (R=0.4758, p=0.0252) suggesting an antibody-mediated regulation of glycolysis. Further, extracellular acidification (ECAR) and oxygen consumption (OCR) rate of the developing OCs were increased by the osteoclastogenic anti-MDA/MAA clones (maximum increase of 54% for the ECAR and 78% for the OCR by clone 146+:01G07, and maximum increase of 28% for the ECAR and 39% for the OCR by clone 1103:01H05), but not by the non-osteoclastogenetic anti-MDA/MAA clones or control antibodies. The glycolysis inhibitor 2-deoxyglucose completely abolished the osteoclastogenetic effect of the anti-MDA/MAA clones at drug concentrations that did not influenced baseline OC development. Fab2 fragments of the osteocalstogenetic anti-MDA/MAA clones had no effect on OC development and metabolism. In accordance with this, Fc-gamma receptor I neutralizing antibodies completely abolished the osteocalstogenetic effect of the anti-MDA/MAA clones. The osteoclastogenetic effect of the anti-MDA/MAA antibodies was independent of IL-8 production. In contrast to anti MDA/MAA antibodies, ACPA-mediated osteoclastogenesis was independent of glycolysis and Fc-gamma receptors but dependent on IL-8.Conclusion:Our results describe a novel glycolysis-dependent mechanism by which anti-MDA/MAA antibodies promote osteoclast development that is different from the one previously described for ACPA.References:[1] C. Grönwall et al. Journal of Autoimmunity 84 (2017) 29-45.[2] C. Wang et al. Arthritis and Rheumatism 62 (2010) 2064-2072Disclosure of Interests:Koji Sakuraba: None declared, Akilan Krishnamurthy: None declared, Alexandra Circiumaru: None declared, Meng Sun: None declared, Vijay Joshua: None declared, Marianne Engström: None declared, Xiaowei Zheng: None declared, Cheng Xu: None declared, Khaled Amara: None declared, Vivianne Malmström Grant/research support from: VM has had research grants from Janssen Pharmaceutica, Sergiu-Bogdan Catrina: None declared, Caroline Grönwall: None declared, Bence Réthi: None declared, Anca Catrina: None declared

OBJECTIVE:Anti-citrullinated protein antibodies (ACPAs) are a hallmark of seropositive rheumatoid arthritis (RA). Yet, the precise disease-relevant autoantigens that are targeted by ACPAs remains a matter of debate. This study utilized patient-derived monoclonal ACPAs, rather than serum autoantibody analysis, to characterize the multireactivity to different protein modifications and to reveal autoantibody subsets in patients with RA.METHODS:Twelve human monoclonal ACPAs (positive by the second-generation cyclic citrullinated peptide test) were generated from 6 RA patients, and a head-to-head comparison of their reactivities was performed. For profiling, we used a complementary DNA-based protein array (Engine GmbH) and 3 peptide-screening platforms with RA autoantigens (Thermo Fisher Scientific), citrullinated and carbamylated peptides (NimbleGen/Roche), or histone-derived peptides with different posttranslational modifications (JPT Histone Code), covering >207,000 peptides (>7,800 gene products).RESULTS:The fine-specificity profiles of the investigated ACPAs varied, but all of the monoclonal ACPAs displayed multireactivity to a large number of citrullinated peptides/proteins, each characterized by specific binding properties. ACPA subsets could be defined by clone-distinct consensus binding motifs (e.g., Cit-Gly, Gly-Cit, or Arg-Cit-Asp), with the most common ACPA recognition being that of a Gly in the +1 flanking position, but with additional amino acid preferences. For ACPA protein recognition, we observed a preference for citrullinated RNA-binding proteins with high Arg/Gly content. Six of the 12 ACPA clones also bound acetylated lysine (KAc) or homocitrulline peptide motifs, displaying a similar affinity or higher apparent affinity than that for citrullinated peptides.CONCLUSION:ACPAs and anti-modified protein autoantibodies represent overlapping facets of RA autoimmunity and bind to a wide variety of modified proteins, extending well beyond the historically recognized set of RA autoantigens. So far, KAc reactivity has been detected only in the context of anti-carbamylated and anti-citrullinated peptide autoantibody responses, postulating the existence of hierarchies of autoreactivity in RA. Future investigations of ACPA fine specificities and functionality should take into consideration the presence of consensus Cit/Carb/KAc motifs and the multireactivity of these autoantibodies in patients with RA.

We describe here a detailed, refined protocol for the generation of citrulline-specific monoclonal antibodies from single human B cells from rheumatoid arthritis (RA) patients. This protocol provides a detailed guide of the procedure starting from single B cells of your choice and followed by amplification of the variable region of immunoglobulin genes by RT-PCR, subsequent immunoglobulin gene cloning, recombinant IgG1 monoclonal antibody (mAb) production and quality controls. The produced mAbs can be used for further studies including reactivity towards candidate antigens and functionality both in vitro and in vivo. This protocol can be used to generate antigen-specific mAbs from B cells derived from different tissues and compartments, including peripheral blood, synovial fluid, digested biopsies, bone marrow aspirations, and bronchoalveolar lavage fluid. Notably, although examples are given on how to identify citrulline-specific autoantibodies the general methods can also be applied to other reactivities.

Background: Antibodies targeting citrullinated proteins (ACPA) are highly specific for rheumatoid arthritis (RA). However, the etiology and molecular basis for ACPA production is still unclear. Based on an epidemiological association between RA and periodontitis (PD), and the unique ability of the oral pathogen P.gingivalis (Pg) to express a PAD enzyme that can citrullinate both bacterial and human proteins, it has been hypothesised that the ACPA response may be triggered in the gum mucosa in response to Pg. Objectives: The main purpose of this study was to investigate if citrulline-reactive B cells reside in inflamed gingival tissue, and to examine ACPA cross-reactivity between citrullinated bacterial and human epitopes on a monoclonal level. Methods: Using a single-cell antibody cloning approach, 55 recombinant monoclonal antibodies (mAbs) were generated from gingival tissue (GT) CD19+ B cells (n=1 ACPA+ RA/PD patient). Citrulline reactivity was determined using the anti-CCP2 kit (EuroDiagnostica AB), and in-house peptide ELISAs (including a citrullinated peptide derived from Pg PAD (CPP3) and citrullinated peptides derived from human α-enolase, fibrinogen, vimentin, fillaggrin and histone 4). Reactivity against CCP2 and CPP3 was also investigated in: 19 mAbs from bronchoalveolar lavage (BAL) CD19+ B cells (n=2 ACPA+ RA patients); 29 mAbs from bone marrow (BM) plasma cells (n=4 ACPA+ RA patients); 142 mAbs from synovial fluid (SF) plasma cells (n=5 ACPA+ RA patients); and 36 mAbs from peripheral blood memory/plasma cells (n=4 ACPA+ RA patients). Predicted germline versions were produced for two of the mAbs. Results: Among 55 GT mAbs, 14 unique clones (25%) were reactive to the bacterial CPP3-peptide. We also detected CPP3-reactive mAbs from BAL (n=9/19), BM (n=3/29), SF (n=1/142) and peripheral blood (n=1/36). Interestingly, 11 out of 28 (39%) CPP3-reactive clones also bound citrullinated peptides derived from human proteins. Notably, three of these clones were positive in the clinical anti-CCP2 test, and when converted back to the predicted germline sequence, these clones became CCP2 negative, while maintaining reactivity against the bacterial CPP3 peptide. Conclusion: Our data show that B cells reactive with a citrullinated peptide derived from Pg PAD are present in gingival tissue, lungs, bone marrow, blood and the inflamed joint of ACPA+ RA patients. Moreover, the finding that a number of these clones are cross-reactive with citrullinated peptides derived from human proteins as well as the gold standard CCP2 test suggests mechanisms of molecular mimicry in the generation of ACPA. Importantly, the germline versions of these ACPA were Pg-reactive but not autoreactive, supporting the hypothesis of a bacterial origin for this ACPA response. Disclosure of Interests: Natalia Sherina: None declared, Vijay Joshua: None declared, Radha Thyagarajan: None declared, Natalie Sippl: None declared, Lena Israelsson: None declared, Heidi Wähämaa: None declared, Ragnhild Stålesen: None declared, Kaja Eriksson: None declared, Tülay Yucel-Lindberg: None declared, Aase Hensvold: None declared, Caroline Grönwall: None declared, Anca Catrina Grant/research support from: Yes, but not for the presented study., Vivianne Malmström: None declared, Antonio Lanzavecchia: None declared, Luca Piccoli: None declared, Khaled Amara: None declared, Karin Lundberg: None declared

Background: Antiphospholipid syndrome (APS) is defined by arterial, venous, or microvascular thrombosis or obstetric morbidity together with confirmed positive test results, at least 12 weeks apart, for antiphospholipid antibodies (aPLs), including at least 1 of the following tests: anticardiolipin (anti-CL), anti2-glycoprotein I (anti-2GPI), or the functional lupus anticoagulant test (1, 2). Long-term anticoagulation prevents new thrombotic events in most patients with APS. Information on prothrombotic aPLs in patients with myocardial infarction (MI) is conflicting, due to limited sample sizes, selected populations, and nonstandardized methods in previous studies (3). Objective: To assess the frequency of anti-2GPI and anti-CL of IgG/IgA/IgM isotypes and antinuclear antibodies among patients with first-time MI and matched control participants in a large, multicenter, casecontrol study. Methods and Findings: The PAROKRANK (Periodontitis and Its Relation to Coronary Artery Disease) study included 805 consecutive patients younger than 75 years who were hospitalized for a first MI at 17 Swedish hospitals between 2010 and 2014 (4). Population control participants (n= 805) were individually matched for age, sex, and region to the patients. Persons who had prior MI or heart valve replacement or were unable to follow the protocol were excluded. Patients were recruited during hospitalization and scheduled for an inclusion visit 6 to 10 weeks later at their local cardiology clinic. The national quality registry SWEDEHEART (www.swedeheart.se) was used to collect medical information at the hospitalization and standardized follow-up data at the inclusion visit. More detailed clinical protocol and extensive questionnaires were also completed at the inclusion visit (5). The same information was collected for the matched control participants. All participants fasted and abstained from smoking for 12 hours before blood sampling at the inclusion visit. Autoantibodies targeting CL and 2GPI of IgG/IgM/IgA isotypes and specific nuclear antigens, as specified in the Table, were analyzed using multiplexed beads (BioPlex 2200 Multiplex Testing, Bio-Rad). Table. Clinical and Serologic Characteristics in Patients with Myocardial Infarction and Matched Control Participants* The study was approved by the Regional Ethics Committee in Stockholm, Sweden (Dnr:2008/152-31/2), and all participants provided written informed consent. The Table shows participant characteristics and laboratory results. We found that IgG anti-2GPI and/or IgG anti-CL (referred to as aPL IgG positive) were more common in patients with MI (aPL IgG positive: n= 88 [11.1%]) than in control participants (n= 10 [1.3%]) (P< 0.001), whereas IgA/IgM isotypes did not differ. Many patients with MI had high IgG levels (Figure). Anti-CL and anti-2GPI of the same isotype correlated strongly (IgM: rs = 0.92; IgG: rs = 0.85; IgA: rs = 0.86). Figure. Units of antiphospholipid antibodies in patients with MI and matched controls. Each circle represents 1 participant. Anti-2GPI = anti2glycoprotein I antibodies; Anti-CL = anticardiolipin antibodies; APL = IgA antiphospholipid units; GPL = IgG antiphospholipid units; Ig = immunoglobulin; MI = myocardial infarction; MPL = IgM antiphospholipid units; NS = nonsignificant. In multivariable conditional logistic regression analysis (SAS software, version 9.4; SAS Institute) adjusted for traditional risk factors (Table), aPL IgG positivity (odds ratio, 7.8 [95% CI, 4.0 to 15.3]; P< 0.001) and current smoking (odds ratio, 2.6 [CI, 1.9 to 3.5]; P< 0.001) remained associated with MI, whereas diabetes, hypertension, and body mass index did not. Using t tests and 2 tests as appropriate, we found that age did not differ among aPL IgGpositive (n= 88) and negative (n= 704) patients with MI (63.1 vs. 61.8 years; P= 0.146), but aPL IgG positivity was nonsignificantly more common among women (26.1% vs. 17.9%; P= 0.062) and current smokers (33.7% vs. 25.1%; P= 0.085). Discussion: We report a high prevalence (11%) and strong positive associations between aPL IgG positivity and first-time MI. These antibodies were 10 times more common in patients with MI than control participants, and they occurred independently of traditional cardiovascular risk factors. Notably, only aPLs of the IgG isotype were elevated. No differences between patients with MI and control participants regarding IgM, included in the Sydney criteria (1), or IgA were observed. Blood samples were drawn 6 to 10 weeks after MI, so whether the antibodies were present before the MI or whether they were persistent or transient are unknown. Persistent aPLs together with MI would indicate definite APS (1), for which current guidelines recommend indefinite warfarin treatment (2). Transient aPLs might represent an immunologic reaction to myocardial damage during MI or an infection, and whether these aPLs are associated with enhanced risk for thrombosis is not known. This study includes a large and representative study population with first-time MI, both sexes, and well-matched control participants. It used standardized methods for aPL measurements and collection of participant characteristics. Study limitations are that aPL determinations were investigated only once and that the lupus anticoagulant test could not be performed due to lack of citrated plasma. A strong and independent association between aPL IgG positivity and first-time MI may represent an important, but until now neglected, risk factor for MI in the general population. If confirmed in long-term cohort studies, these findings may lead to improved management, treatment, and outcomes for many patients who have had or are at high risk for MI.

Career situation of first and presenting author Post-doctoral fellow Introduction Anti-citrullinated antibodies (ACPA) in RA have been postulated to contribute to disease pathogenesis. This has been supported by studies of monoclonal APCA generated by cloning from single cell sorting of different B cell subsets from peripheral blood and synovial fluid. It has been debated if the bone marrow (BM) compartment of RA patients hosts long lived plasma cells producing ACPA or if ACPA are primarily coming from plasmablasts at the sites of inflammation, e.g. the synovium. Objectives The main objectives were to study primary lymphoid tissue from RA patients, the BM plasma cell IgG repertoire and further characterize recombinantly expressed antibodies from selected sequences to explore if ACPA producing long lived plasma cells reside in this niche. Methods For this study we collected bone marrow from proximal or distal femur of RA patients undergoing hip replacement surgery due to secondary osteoarthritis. BM samples were processed from four ACPA+RF+ and one ACPA-RF+ RA patients. Mononuclear cells were obtained by Ficoll separation and CD138+ plasma cells were single cell sorted by flow cytometry. Paired heavy and light chains were PCR amplified, sequenced and analyzed by V-Quest and IgBLAST towards the IMGT database to annotate variable gene usage. Selected sequences were cloned and expressed as IgG in Expi293 cells. Results Overall 465 paired IgG BM plasma cell sequences were obtained, 97 from the ACPA- patient and 368 (range 20–194) from ACPA+ patients. We observed statistically significant changes in heavy chain variable gene usage with lower VH-1 and higher VH-3 frequency in the ACPA+ sequences compared to ACPA- seqences. We also found statistically significant increase in VH N-glycosylation in ACPA+sequences (22.6% ACPA+ vs 12.4% ACPA-; p=0.03). There was however no difference in mutation numbers. From these, 34 clones were selected, based on mutation numbers and presence of Fab N-glycosylation sites, for subsequent mAb-expression. Among the 34 clones we found two CCP2+ and cit-peptide positive clones and one malondialdehyde-acetaldehyde (MAA) adduct reactive clone, originating from different APCA+ patients. Conclusions We could identify BM plasma cells producing autoantibodies to malondialdehyde-modified proteins and citrullinated peptides, isolated from RA patients. Hence, we show for the first time, that not just at the site of inflammation but also long-lived plasma cells in bone marrow do produce RA specific autoantibodies. Acknowledgements Orthopedic staff at Karolinska University Hospital. Danika Shepis, Louise Berg and Christina Gerstner for sorting help. Disclosure of Interest None declared

Background: Anti-citrullinated protein antibodies (ACPA) are specific markers with pathological effects in rheumatoid arthritis (RA). ACPA-specific B-cells have been identified in synovial joint fluid and peripheral blood. An increased concentration of ACPA is found in synovial fluid and lung tissue compared to blood suggest local production, partly confirmed by detection ACPA synovial plasma blasts, but primary lymphoid tissue have not previously been examined. Objectives: In this study we investigated the plasma cell repertoire in the bone marrow and occurrence of autoantibody producing plasma cells. Methods: For this study we developed a method to collected and process bone marrow samples from proximal femur in RA patients undergoing hip joint replacement. Bone marrow samples were processed and mononuclear cells were obtained by Ficoll separation. CD138+ plasma cells were single cell sorted by flow cytometry. Paired heavy and light chains were PCR amplified, sequenced, and analyzed by V-Quest and IgBLAST towards the IMGT database to annotate variable gene usage. To enrich for ACPA, sequences were selected based on high somatic hypermutations number and Fab N-glycosylation sites. Selected sequences were cloned and expressed as IgG in Expi293 cells. Monoclonal antibody reactivity to citrullinated and arginine form of vimentin, enolase, fibrinogen, histone, tenascin C peptides; carbamylated and malondialdehyde-acetaldehyde bovine serum albumin, and tetanus toxoid antigens were examined by ELISA. Included patients were clinical examined and donating peripheral blood samples. Results: After method development and quality analysis we were able to include five bone marrow samples in this study. The five included RA patients, median age 74 (range 66-82) years, had in median 20 (range 4-44) years of destructive disease, all were RF positive, four were females and anti-CCP2-positive, and all except one were treated with anti-rheumatic drugs. Overall, from the processed bone marrow, 465 paired heavy-light IgG plasma cell sequences were obtained, in total 368 (range 20-194) from ACPA positive patients and 97 from the ACPA negative patient. We observed statistically significant changes in heavy chain variable gene usage with lower VH-1 and higher VH-3 frequency in sequences from ACPA positive RA patients compared to sequences from the ACPA negative RA patient. We also found statistically significant increase in VH N-glycosylation in sequences from the ACPA positive RA patients (22.6% ACPA positive vs 12.4% ACPA negative; p=0.03) but no difference in mutation numbers. From obtained sequences, 34 interesting clones were selected for monoclonal antibody-expression. Among the 34 clones we found two citrulline specific reactive clones (reactivity towards CCP2, citrullinated vimentin and fibrinogen peptides) and one malondialdehyde-acetaldehyde (MAA) reactive clone, originating from three different APCA positive patients with the most longstanding disease. None of these clones had reactivity towards control antigens such as arginine versions of peptides. Conclusion: Plasma cells producing ACPA are present in primary lymphoid tissue from RA patients. Further studies of difference in characteristics, IgG gene usage and N-glycosylation, between ACPA-positive and ACPA-negative patients are needed. Disclosure of Interests: Radha Thyagarajan: None declared, Aase Hensvold: None declared, Lena Israelsson: None declared, Johanna Steen: None declared, Heidi Wähämaa: None declared, Annika van Vollenhoven: None declared, Khaled Amara: None declared, Anca Catrina Grant/research support from: Yes, but not for the presented study., Vivianne Malmström: None declared, Caroline Grönwall: None declared

Introduction Increasing evidence points to the autoimmune process of rheumatoid arthritis (RA) originating at mucosal surfaces. Previous work from our group described a subset of B cells in the synovium and synovial fluid (SF) of RA patients distinguishable by their expression of Fc-like receptor 4 (FcRL4) and elevated expression of RANKL, indicating a unique pathogenic function.1,2 B cells expressing FcRL4 were originally described as a distinct memory B cell subset in human tonsils where they are localised in the epithelium.3 We have recently shown in RA that they are enriched in clones recognising citrullinated autoantigens.3 Recent in vitro work suggested that FcRL4 is a low affinity receptor for aggregated IgA.4 Objectives Investigate the interactions between IgA and FcRL4 and FcRL4+ B cells using tonsil and SF B cells, and transfected cell lines. Examine the distribution of Ig classes by flow cytometry and PCR. Methods Mononuclear cells were isolated from SF (n=10) and tonsil, labelled for CD19, FcRL4, IgA, and IgG and analysed by flow cytometry. In experiments identifying BCRs and relative loads of surface bound IgA; SFMCs were washed in an acidic buffer to remove receptor-bound proteins before staining. Heat-aggregated purified human IgA was added to assess IgA binding to FcRL4 +B cells. Single B cells were sorted from SF of RA patients and their constant region genes probed for identification of their Ig subclass by PCR. Results We show that ex vivo SF FcRL4+ B cells have a significantly higher load of IgA bound to their surface (p=0.0001) than FcRL4- B cells. Following in vitro removal of surface bound IgA, FcRL4 +B cells bind heat-aggregated IgA (p=0.03 vs control). We also demonstrate that a significantly higher proportion of FcRL4+ B cells use IgA BCRs (p=0.0061) by flow cytometry, and by probing constant region genes by PCR an enrichment for Ig genes for coding the IgA1 isotype (p=0.009) was found. Furthermore, three out of eight of the antibodies recognising citrullinated peptides had been cloned from FcRL4+ IgA+ B cells. Conclusions In conclusion, their specificity for citrullinated antigens, ability to capture IgA immune complexes through via FcRL4 and enrichment in IgA1 subclass expression, suggest a role for FcRL4 +B cells in the mucosal origin of joint inflammation. References . Yeo L, et al. Ann Rheum Dis2011;70:2022–2028. doi:10.1136/ard.2011.153312 . Yeo L, et al. Ann Rheum Dis2015;74:928–935. doi:10.1136/annrheumdis-2013-204116 . Amara K, et al. J. Autoimmun2017. doi:10.1016/j.jaut.2017.03.004 . Wilson TJ, et al. J Immunol2012;188;4741–4745. doi:10.4049/jimmunol.1102651 Acknowledgements This work was supported by the Medical Research Council UK, Rheumatoid Arthritis Pathogenesis Centre of Excellence, and Karolinska Institutet Foundations for Rheumatology Research. Disclosure of interest None declared

Autoreactive B cells have a central role in the pathogenesis of rheumatoid arthritis (RA), and recent findings have proposed that anti-citrullinated protein autoantibodies (ACPA) may be directly pathogenic. Herein, we demonstrate the frequency of variable-region glycosylation in single-cell cloned mAbs. A total of 14 ACPA mAbs were evaluated for predicted N-linked glycosylation motifs in silico, and compared to 452 highly-mutated mAbs from RA patients and controls. Variable region N-linked motifs (N-X-S/T) were strikingly prevalent within ACPA (100%) compared to somatically hypermutated (SHM) RA bone marrow plasma cells (21%), and synovial plasma cells from seropositive (39%) and seronegative RA (7%). When normalized for SHM, ACPA still had significantly higher frequency of N-linked motifs compared to all studied mAbs including highly mutated HIV broadly-neutralizing and malaria-associated mAbs. The Fab glycans of ACPA-mAbs were highly sialylated, contributed to altered charge, but did not influence antigen binding. The analysis revealed evidence of unusual B-cell selection pressure and SHM-mediated decrease in surface charge and isoelectric point in ACPA. It is still unknown how these distinct features of anti-citrulline immunity may have an impact on pathogenesis. However, it is evident that they offer selective advantages for ACPA+ B cells, possibly through non-antigen driven mechanisms.

BACKGROUND:Systemic lupus erythematosus (SLE) is a systemic autoimmune disease, which exhibits multiple B cell abnormalities including expanded populations of memory B cells and elevated levels of autoantibodies. Belimumab is a monoclonal antibody targeting the B cell cytokine BAFF (a.k.a. BLyS), approved for the treatment of SLE.METHODS:In this prospective cohort study, B cells from peripheral blood of 23 SLE patients initiating belimumab treatment and followed longitudinally for up to three years, were assessed using mass cytometry.FINDINGS:B cells decreased during the study period, with a rapid decrease of both naïve and CD11c+CD21- B cells at the first follow-up visit, followed by a continuous reduction at subsequent follow-ups. In contrast, plasma cells and switched memory B cells remained stable throughout the study. The observed immunological changes correlated with early, but not late, clinical improvements. Moreover, high baseline B cell counts were predictive of failure to attain low disease activity. In summary, our data unveiled both rapid and gradual later therapy-associated alterations of both known and unforeseen B cell phenotypes.INTERPRETATION:Our results suggest that evaluation of B cell counts might prove useful prior to initiation of belimumab treatment and that early treatment evaluation and discontinuation might underestimate delayed clinical improvements resultant of late B cell changes.