You are here
HLA and anti-citrullinated protein antibodies: Building blocks in RA
Best Practice & Research Clinical Rheumatology, Volume 29, Issue 6, December 2015, Pages 692 - 705
Antibodies against citrullinated proteins (ACPAs) are specific for rheumatoid arthritis (RA). ACPA-positive RA is a chronic inflammatory disease resulting from the complex interaction between genetic (mainly HLA class II genes) and environmental factors (mainly smoking). Recent findings have offered new insights into where, when and how anti-citrulline immunity develops. Some studies have found that a mucosal site, such as the lungs, may function as the initiating site for the immune response against citrullinated proteins, in line with the known association between smoking and ACPA. Other studies, focusing rather on the HLA associations, have suggested that cross-reactivity between microbial sequences and citrullinated self-proteins may lead to ACPA formation. Once ACPAs have developed, they can circulate throughout the body and upon reaching the joints exert direct pathogenic effects themselves. ACPAs can target first the bone compartment of the joints to activate osteoclasts and release interleukin (IL)-8 that in turn will promote bone loss and pain-like behaviour. In the current review, we will present the current understanding of the genetic associations in RA contributing to ACPA occurrence and offer insight in the latest findings explaining how and why autoimmunity generated in the lungs of genetically susceptible hosts might lead to chronic inflammation in the joints.
Keywords: Anti citrullinated proteins antibodies (ACPA), HLA, Citrullination, Osteoclast activation, Interleukin-8.
Rheumatoid arthritis (RA) is a chronic disease mainly characterized by inflammation of small joints of hands and feet. The underlying pathogenic abnormality is systemic autoimmunity with main clinical manifestations on the level of joints, tendons and bursae. Several markers of the underlying immunological aberrancies can be detected in the blood of RA patients, including autoantibodies, in particular. Among the autoantibodies, anti-citrullinated protein antibodies (ACPAs) have been found to be more specific for RA than the traditionally used rheumatoid factor (RF) . Based on several findings, which will be discussed in more detail in this review, ACPA appears to have a closer link to the underlying immunopathogenesis compared to, for example, RF.
ACPAs can be found up to 10 years before RA onset and can predict the development of RA. Prior to disease onset, the ACPA response matures with a rise in ACPA levels, increased isotype usage and epitope spreading to a broader range of citrullinated antigens  and . This results in a broad, mature autoantibody response at the time of disease onset, which does not expand any further later. The discovery of ACPA has led to a paradigm shift in the pathophysiological hypotheses concerning the development of RA. Nearly all genetic and environmental risk factors known for RA have been found to be exclusively associated with ACPA-positive disease. This also applies to the most potent genetic risk factor encoded in the human leukocyte antigen (HLA) class II locus.
In this review, we will first provide an in-depth description of the HLA alleles and their relationship to RA, followed by a presentation of the most recent current pathogenic views linking HLA genes to the development of ACPA-positive RA.
HLA molecules: structure and function
Molecules belonging to the major histocompatibility complex (MHC), known in humans as HLA, are best known for their antigen-presenting function. For a complete understanding of the current hypotheses about the role of HLA in the development of RA, we need to take a closer look at the exact structure and function of these molecules and the genomic organization of their encoding locus .
HLA molecules are divided into two classes differing in the site of expression, structure and function (Table 1). HLA class I molecules are present on all nucleated cells, whereas HLA class II molecules are only expressed on immune cells, and especially those cells involved in antigen presentation such as dendritic cells, B cells and macrophages. The overall protein structure of HLA class I and class II molecules is similar in that one side of these molecules is attached to the cell membrane and the other side has a long cleft or groove to bind peptides. On a lower structural level of the subunit, however, the class I and II molecules do differ. Class I molecules consist of two polypeptide chains, of which only the longer alpha chain encoded in the HLA genetic locus is polymorphic. It is bound to the non-polymorphic molecule beta2-microglobulin. The peptide-binding cleft is formed entirely by the alpha chain. Class II molecules, however, consist of two equally large chains: the alpha and beta chain, both of which are polymorphic and encoded in the HLA genomic region. Both contribute to the peptide-binding cleft. The composition of class II molecules consisting of two chains adds a higher degree of variability, but the alpha chains of these molecules are generally less polymorphic than the beta chains. Furthermore, the combination of alpha and beta chains encoded on different chromosomes can increase the number of different HLA molecules expressed by a cell, with two alpha and two beta chains leading to four different molecules. Murine studies have shown, however, that not all combinations lead to stable dimers.
HLA Class I and II characteristics and associations with RA.
|Class I||Class II|
|Expression||All nucleated cells||Mainly antigen-presenting cells|
|Chains encoded within HLA locus||1 alpha-chain/molecule||1 alpha and 1 beta-chain/molecule|
|Length of presented peptides||8–10 amino acids||≥13 amino acids|
|Origin of presented peptides||Cytosol and endoplasmic reticulum||Lysosomes|
|Recognized by||CD8-positive T cells||CD4-positive T cells|
|Genomic locus (with number of alleles)||HLA A (3285)
HLA B (4077)
HLA C (2801)
|HLA DP A (42), HLA DP B (587)
HLA DQ A (54), HLA DQ B (876)
HLA DR A (7), HLA DR B (1932)
|Alleles associated with ACPA-positive RA in Caucasians||Predisposing:
DRB1*01:01 and 01:02
|Amino acids associated with ACPA-positive RA in Caucasians||HLA-B 9||HLA-DRβ1 11, 71, 74
The way in which peptides are bound by the different molecules is a key aspect to understanding how some HLA molecules may predispose to RA. Crystallography has revealed that the major difference between class I and II molecules is the ‘openness’ at the ends of the peptide-binding cleft, that is, the ends of a peptide bound to class I are buried within the class I molecule, while the ends of a peptide bound to class II can dangle out. Peptides bound to class I are generally of length 8–10 amino acids because contacts between their free amino-terminus and carboxy-terminus on one side and the peptide-binding cleft on the other side are very important for stable binding. Longer peptides can therefore not be readily accommodated. Besides the termini, amino acid residues at two or three particular positions (called anchor residues) are essential for binding to a given HLA class I molecule. Class II molecules, however, can bind to much longer peptides of at least 13 amino acids held in the peptide groove by side chains that protrude into shallow and deep pockets of the class II molecules. These binding pockets of the class II molecules are more liberal than class I molecules in their accommodation of different peptide residues. It has therefore proven considerably more difficult to define anchor residues for class II molecules and predict the peptides can bind to them. This has impeded the identification of a common peptide sequence which might be accommodated by the various class II molecules associated with RA, a so-called ‘arthritogenic peptide.’
The structural differences between HLA class I and II molecules also control their functional differences. The T cell surface molecule CD8 binds to invariant sites on HLA class I, while CD4 binds to HLA class II molecules, thereby determining the peptide–HLA complex and its corresponding T cell response. As class I molecules are loaded in the endoplasmic reticulum with peptides derived from the cytosol, these peptides which can originate from, for example, viruses lead to a cytotoxic CD8-positive T cell response. By contrast, class II molecules are loaded in acidified lysosomes with peptides derived from, for example, endocytosed pathogens. They activate CD4 T cells specialized to activate other immune cells such as B cells or macrophages.
Besides their role in the activation of T cells at the start of an immune response, it is important to realize that HLA molecules also play a key role in the selection of immune cells, which are allowed to develop and mature to finally patrol the body. T cells originate from the bone marrow, but undergo most of their development in the thymus where they are subjected to the process of positive and negative selection. During positive selection, only T cells capable of recognizing self-HLA molecules are selected for further development, implying that most thymocytes die at this stage. As a result, the T cell repertoire of an individual is restricted to the HLA molecules that this individual harbours and expresses in the thymus. This concept of HLA-restricted T cell responses is critical for understanding some of the current hypotheses about the role of HLA polymorphisms during RA pathogenesis.
HLA genes: structure and nomenclature
The genetic region encoding for HLA molecules is found on chromosome 6. The class I region contains the genes for the various alpha chains of three different class I molecules: HLA A, B, and C. HLA B is the most diverse of these three with over 4000 different possible alleles leading to over 3000 nonsynonymous proteins. The class II region also harbours the genes for three different cell surface molecules including HLA DP, DQ and DR. In this case, the alleles encoding the beta chain of the HLA–DR molecule exist in variable forms with >1500 different allelic variants (Table 1) .
Throughout the years, the nomenclature of HLA alleles has changed substantially. This is mainly due to the evolution in methods used to characterize or ‘type’ the alleles. Originally, testing for the presence of a certain HLA allele was performed by adding serum-containing antibodies to the HLA allele of interest and to cells with an unknown HLA type. With serotyping, the alleles were called, for example. DR1, which we now know are different HLA DRB1 molecules. A similar principle of alloreactivity was also employed by mixed lymphocyte cultures, in which T cells from for example, a potential organ recipient would react to the presence of minor allele differences with cells from the potential donor. This typing led to the Dw types in which DR1 was typed as either Dw1 or Dw20. Finally, these typing methods were replaced by genetic typing which makes use of differences in the genetic sequences of the various alleles. Based on the current HLA nomenclature, alleles are called: HLA locus name*xx:yy in which the first two digits (xx) stand for a group of alleles and the third and fourth digit (yy) specify a nonsynonymous allele. For example, the alleles previously known as DR1 are now designated as DRB1*01:01 through DRB1*01:72.
Within the HLA region, genes coding for the separate molecules are often inherited together due to linkage disequilibrium. This leads to typical haplotypes consisting of a certain combination of alleles, for example, A*01:01; C*07:01; B*08:01; DRB1*03:01; DQA1*05:01; DQB1*02:01, which is a well-known ancestral haplotype. This haplotype structure can make it difficult to distinguish between the separate alleles causing the association with a disease such as RA.
HLA associations with RA
The link between HLA alleles and RA was originally discovered by mixed lymphocyte reactions as described earlier, in which cells from RA patients were found to not be reactive against those from other RA patients, thus implying they must have similar/the same HLA alleles  and . Using serotyping, the common allele was originally identified as the DRw4 alloantigen , although it was later discovered that several of the (then called) HLA–DR alleles were associated with RA. This led to the formulation of the shared epitope (SE) hypothesis which was based on the observation that all HLA–DR alleles predisposing to RA have a similar amino acid sequence (the so-called shared epitope) at position 70–74 of the HLA–DRB1 molecule . The fact that this sequence is located in the peptide-binding groove suggests that it may lead to the presentation of specific peptides involved in the activation of T cells directed against joint-specific antigens. It has, however, proven to be very challenging to distinguish distinct peptide-binding motifs associated with disease, as illustrated by the fact that this has also not been successful for the much less accommodating HLA–B27 allele . As mentioned earlier, the discovery of anti-citrullinated protein antibodies (ACPA) has led to a significant change of perspective concerning the association between risk factors and RA. Most genetic risk factors including the HLA shared epitope (SE) alleles have now been found to be specifically associated with ACPA-positive disease . Fig. 1 provides an example of an SE allele binding to a citrullinated peptide which will be discussed in more detail later.
Structure of the HLA-DRB1*04:01 α–chain (yellow), β-chain (blue) and citrullinated vimentin (vimentin-71Cit66-78) (purple). Specific residues of the β-chain, indicated by amino acid code and position number, interact with citrulline via hydrogen bonds and van der Waals interactions. Figure reproduced with permission from .
Over the years, many studies have reported associations between numerous HLA alleles or single nucleotide polymorphisms (SNPs) and autoantibody-positive/autoantibody-negative RA in different populations  and . Various classifications incorporating combinations of different alleles based on amino acid similarities such as the de Montcel , DERAA , de Vries , and D70 classification have been put forward. Overall, the finding in Caucasians is in accordance with a large meta-analysis performed in four European populations . In ACPA-positive RA, there are clearly alleles conferring an increased risk which largely overlap with the well-known SE alleles belonging to DRB1*01, *04 and *10 allelic groups. There is a hierarchy of risk among these alleles with *04 alleles being associated with the strongest effect . In addition to the SE alleles, the DRB1*09 alleles have also repeatedly been reported to confer risk to (ACPA-positive) RA  and . In the past, it has been questioned whether there are also alleles associated with a true protective effect, or whether the lower prevalence of presumed protective alleles in RA patients might be due to the high prevalence of risk-conferring alleles in this population . In recent years, however, several publications have shown that while taking into account the effects of predisposing alleles, a protective effect can be shown for other alleles, such as most prominently DRB1*13 alleles  and . Protective effects of certain rare alleles such as HLA DRB1*04:02 cannot be excluded due to the lack of power. In conclusion, there is a large body of literature on the association between HLA DR (serotype)/DRB1 alleles (genotype) and RA; the consistent observations have been summarized in Table 1. Reports about associations with ACPA-positive RA outside the HLA–DRB1 region exist  and , but are more limited in number, as is also the case for publications on associations between HLA and autoantibody-negative RA , , and .
Modern-day statistical advances have provided a new viewpoint on the association between HLA and RA by focusing not on entire alleles, but rather on separate amino acids. Using large reference datasets containing information on HLA SNPs as well as classically genotyped HLA alleles, it has become possible to impute HLA genotypes in SNP-typed individuals paving the way for re-analysis of datasets generated in genome-wide association studies . In this manner, three HLA–DRB1 amino acids located in the peptide-binding groove were found to be particularly relevant for the association between this locus and autoantibody-positive RA . Perhaps not surprisingly, haplotypes constructed of these three alleles and recoded using the classical four-digit nomenclature revealed a hierarchy of risk and protection similar to previous studies using HLA genotyping. The imputation method furthermore identified independent risk alleles in HLA–B and HLA–DPB1. The same method has been used for ACPA-negative RA resulting in the identification of the same amino acid positions as for ACPA-positive RA, but with different effects of the individual amino acid residues  and .
Although these findings concerning amino acids have provided an interesting new perspective, the suggestion that these amino acids explain the entire genetic risk for RA conferred by this region should be interpreted with some caution for several reasons. The amino acid positions DRβ1 11 and 13 identified as being the most important contributing amino acids are also the most polymorphic residues and thus provide the best differentiation between different classical HLA alleles . Most importantly, HLA crystallization and peptide elution studies have shown that peptide-binding pockets are shaped by multiple different amino acids  and that the combination of all these amino acids will eventually determine the ligands that can be presented by the HLA molecule. From a structural biological point of view, it therefore appears questionable that the contribution of HLA molecules to the pathogenesis of RA can be explained by a very limited number of specific amino acids.
Hypotheses explaining the HLA–RA association
Based on the antigen-presenting properties of the HLA molecules, the most obvious explanation for the association between certain HLA alleles and RA would be that these predisposing alleles are particularly avid presenters of a joint-derived peptide that would activate T cells and thus initiate the immune response leading to the joint inflammation typical for RA. However as mentioned earlier, it has thus far not been possible to identify an arthritogenic peptide or peptide sequence which would function in this manner. Several other hypotheses, some incorporating modifications of the antigen-presenting function of these alleles, have therefore been proposed to explain the HLA–RA association.
Based on the specific association between HLA SE alleles and ACPA-positive RA, several studies have investigated the capacity of HLA SE alleles to present citrullinated antigens. In a study using HLA DRB1*04:01-transgenic mice, conversion of arginine to citrulline was found to significantly increase the peptide–HLA affinity especially at the P4 pocket . This led to the following overarching hypothesis to explain the differential effects of the various HLA DRB1 alleles on RA susceptibility. Concerning predisposing alleles, the very high affinity between citrullinated (and not arginine-containing) peptides and HLA SE molecules could lead to a high density of peptide–HLA complexes on the cell surface of antigen-presenting cells (APCs) enabling T cell activation. Furthermore, non-RA-associated/neutral HLA molecules are hypothesized to have P4 pockets that can accommodate neither citrulline nor arginine, thereby basically leaving activation of T cells directed against these residues unaffected. By contrast, protective HLA molecules may be able to bind both arginine and citrulline. This would lead to a low density of peptide–MHC complexes on the cell surface, which may polarize cells toward a regulatory phenotype or lead to negative selection of the recognizing T cells.
These results were corroborated by a study investigating the crystal structure of citrullinated peptides (vimentin and aggrecan) bound to both predisposing and protective HLA molecules . Fig. 1 shows the binding of a citrullinated vimentin peptide to HLA DRB1*04:01. Regarding the risk-conferring HLA DRB1*04:01 allele and resistance allele HLA DRB1*04:02, the P4 pocket in 04:01 was found to preferentially accommodate citrulline, while the P4 pocket of 04:02 enabled binding of both citrulline and arginine. In this study, the corresponding SE-restricted vimentin- and aggrecan-specific CD4-positive T cells were also identified in the peripheral blood of RA patients, thus establishing the association between HLA-antigen presentation and T cell reactivity.
Another study has also highlighted this significant step from HLA-molecules to T cells, but proposed a different hypothesis using protective HLA alleles as its starting point . Similar to the shared epitope amino acid sequence present in predisposing alleles, this hypothesis originates from a sequence shared by protective HLA alleles: the DERAA-sequence, which has in the meantime been refined to mainly HLA-DRB1*13  and . Contrary to the previously described assumptions about a direct antigen-presenting role of the RA-associated HLA molecules, this hypothesis suggests that the DERAA sequence of the protective HLA molecules may in fact function as an antigen. Peptides derived from endogenous HLA molecules are known to be presented by other HLA molecules; the DERAA sequence of HLA–DRB1*13, for example, can be presented by HLA–DQ molecules. In DERAA-positive individuals (i.e. individuals with RA-protective HLA alleles), presentation of DERAA-containing peptides in the thymus would lead to deletion of the corresponding T cells, amounting to central tolerance to this sequence. In individuals without protective HLA alleles, however, T cells can be found which respond to the DERAA sequence, which is also found in many microbes. By molecular mimicry, these T cells could subsequently react to one of the few human proteins carrying the DERAA sequence vinculin. Vinculin is expressed in the synovium and can be detected in citrullinated form in RA synovial fluid. This hypothesis therefore provides an explanation for the emergence of anti-citrulline immune responses in individuals lacking protective HLA alleles using the following scenario: T cells directed against DERAA-containing proteins could be primed by microbes and via cross-reactivity to vinculin aid B cells producing antibodies to citrullinated vinculin, thus resulting in the production of ACPA. Fig. 2 illustrates the shared epitope and the protective allele theories.
Schematic representation of the functions of T and B cells in RA. CD4-positive T cells from RA patients can recognize citrullinated self-proteins presented by predisposing HLA molecules as depicted on the left. These T cells might also (cross) react to noncitrullinated epitopes as illustrated on the right. T cells provide efficient help to B cells leading to the production of broadly crossreactive ACPAs. Figure reproduced with permission from .
Besides these theories, which are more or less based on the antigen-presenting/antigen-containing properties of the HLA molecules, a completely different mechanism for the association between the HLA SE alleles and RA has been proposed. HLA SE molecules have been described to function as signal transduction ligands in an allele-specific manner . By binding to (potentially citrullinated) calreticulin on the surface of dendritic cells, the SE molecules may activate an NO-mediated pathway which would then lead to aberrant T cell polarization by the affected dendritic cell. This has been described to lead to an increase in Th17 cells and a decrease in T regulatory cells. This hypothesis thereby provides a strikingly different view on the potential role of the HLA SE molecules and can rightfully be called the (signalling) road less taken.
Timing of the HLA effect
The strong association between ACPA and the predisposing and protective HLA alleles has changed the views on the role of these alleles in disease development. The fact that ACPA can be found years prior to disease onset raised the question whether the HLA alleles known to be associated with RA would also be associated with the presence of ACPA in individuals without RA. A study making use of the Swedish twin registry in which 226 ACPA-positive persons without RA were identified provided an intriguing answer to this question . Compared to the well-known association between the HLA SE alleles and ACPA-positive RA (with an odds ratio (OR) in this study of 7.24, 95% confidence interval (CI): 3.96–13.22), the association between these alleles and ACPA without RA was much smaller, OR: 1.42 (1.06–1.90). Thus, these major genetic risk factors, while important for development of ACPA appear to be even more important for determining the ACPA-positive individuals who will eventually develop disease.
From genes and environment to antibodies
As discussed in detail earlier, genetic factors (either risk or protective factors) are important but not sufficient to explain the development of RA, with environmental and stochastic factors being significant in determining the healthy individuals who develop ACPAs . The strongest known environmental risk factor for antibody-positive RA is smoking, directly and primary affecting the lungs , , , , , , , and . This led to the idea of a potential triggering in the lung of the disease-relevant immunity against citrullinated proteins. Original observations have shown increased expression of citrullinated proteins in the lungs of healthy smokers as compared with non-smokers, also associated with an increased expression of PAD enzymes, suggesting that citrullinated targets might be primarily presented to the immune system in the lungs . Citrullination, however, is a widespread post-translational modification of proteins that can occur either in physiological conditions such as in brain development  and  and terminal differentiation of the epidermis , , , , and  or in any inflammatory context at different anatomic locations . Therefore, citrullination itself cannot explain initiation of the specific citrulline autoimmunity. During recent years, we gained more insights in possible additional mechanisms governing the interaction between genes and environment and leading to pathogenic immunity. First, several studies have demonstrated that distinct MHC class II alleles interact with smoking, silica and textile dust in providing high risk for ACPA-positive, but not ACPA-negative RA , , , , and , suggesting that environmental exposures to the lung might drive the production of antibodies in carriers of certain MHC class II genes. Second, restriction of the antibody response against citrullination of proteins is partially related to the HLA type of the host and specifically the presence of the shared epitope of HLA–DR , with citrullinated peptides binding stronger to the HLA–DR molecules as compared to non-citrullinated ones, increasing therefore the possibility to be presented to the immune system. In addition, distinct shared epitope (SE) alleles of the HLA–DR class II molecule appear to have restrictive relationships with different patterns of reactivities to epitopes of citrullinated autoantigens (such as vimentin, fibrinogen, enolase and collagen II) . Third, it is highly conceivable (but still only superficially investigated in the context of ACPA-positive RA) that exogenous irritants to the lung are able to activate the innate immune responses that in turn will enhance antigen presentation and activate the adaptive immune system. It has been, for example, shown that cigarette smoke can activate innate immune cells by triggering pattern recognition receptors to release danger signals , altering the local microbiome through increased bacterial colonization  and increasing danger-associated molecular patterns produced by the cells . A yet unsolved issue, however, is why citrullinated proteins are recognized as antigens, when citrullination is such a widespread phenomenon both during embryonic development and after birth. One possibility would be the failure of central thymic selection due to low/scarce expression of citrullinated as compared to non-citrullinated proteins in the thymus and its epithelial cells, allowing low-affinity auto-reactive T cells to exist in the circulation of healthy individuals  and . In addition, tissue-specific conditions (such as in the lungs affected by smoking and potential others air pollutants) with extensive local post-translational modifications and high expression of target antigens may generate a repertoire with significant differences from the one in the thymus to be presented as autoantigens and to activate otherwise low-affinity auto-reactive T cells. In addition to auto-reactive T cells, auto-reactive B cells able to produce ACPAs following cigarette smoke exposure have been identified in both RA patients and healthy individuals carrying the SE alleles, suggesting that environment and genes also influence the pool of auto-reactive B cells in healthy subjects .
If lung changes would be an initiating event in RA as suggested by the strong epidemiological association, these changes should be present in susceptible individuals and identifiable in patients already at diagnosis. Older studies have reported the presence of lung disease before diagnosis of RA  and  and more recently the presence of unexplained dyspnoea and immune activation in the lungs years before development of arthritis . In addition, individuals at risk of developing RA being positive for any RFs and/or ACPAs isotypes have a higher incidence of lung changes (as detected by high-resolution computer tomography) than healthy controls . Besides, in some of these individuals, certain ACPA isotypes could be detected in the sputum but not in the blood compartment suggesting that these antibodies might be locally produced in the lungs . This is in line with the observation of increased HRCT changes in the lungs of ACPA-positive early-untreated RA patients as compared to ACPA-negative RA patients and healthy controls, independent of age, gender and smoking status . More detailed pathology studies have recently shown signs of inflammation and immune activation in the lungs of ACPA-positive early untreated RA patients with occasional formation of germinal centre-like structures positively staining for citrullinated peptides . This is accompanied by enrichment of both IgA and IgG ACPAs in the bronchoalveolar fluid lavage (BAL) as compared to blood . Interestingly, increased inflammation and expression of citrullinated proteins were observed in ACPA-positive RA individuals, both smokers and non-smokers, suggesting that factors other than smoking might also be active in these early stages of disease development. In this respect, the mucosal microbiome partially determined by the host genome has recently been highlighted to play an important role in the early stages of RA development. Studies in early-untreated RA revealed dysbiosis in the gut and mouth microbiome that could be quantified and used as a diagnostic and treatment response predictor in a fairly good manner . Preliminary data addressing the role of the lung microbiome showed subtle changes in asymptomatic subjects with an elevated risk of future RA when compared to healthy controls  by examining induced sputum samples, but further detailed investigations in true lung compartments such as the bronchoalveolar space are needed.
From antibodies to joint disease
While the well-studied interaction between genes and environment gives important clues on how ACPAs develop in a susceptible host, less has been known until recently on the self-pathogenic contribution of this autoimmunity and on the longitudinal events responsible for propagation of this autoimmunity from extra-articular sites specifically to the joints.
Several studies have now addressed the pathogenic effects mediated by ACPAs. While it has been difficult to prove a clear-cut arthritis-inducing capacity for ACPAs in mice, passive transfer of ACPAs in mice with the existing minimal joint disease exacerbates disease . ACPAs have been shown to activate the complement system , stimulate macrophage, mast cells and platelets , , and , facilitate release of neutrophil-derived extra-cellular traps (NETs) from neutrophils , recognize citrullinated peptides in the NETs , mediate osteoblast apoptosis  and promote bone destruction  and . Taken together, these data demonstrate that ACPAs have direct pathogenic effects on a large array of cells all relevant for RA pathogenesis and that these effects are mediated through either Fc-mediated mechanisms or direct binding of the ACPAs on cell surface targets.
In contrast to the rapid accumulation of pathogenic proofs for ACPAs, little information on the mechanisms responsible for the ACPA-induced initiation of joint inflammation has been gained until recently. A first piece of evidence came from the identification of shared citrullination patterns in biopsies obtained from both lungs and joints of RA patients . Several such targets were described, with one specific citrullinated vimentin peptide being present in a majority of the studied biopsies. Furthermore, antibodies against this particular peptide were identified in the peripheral blood of a subgroup of RA patients, suggesting a compelling hypothesis, where existence of same citrullinated targets in the lungs and the joints of RA patients might explain the localization of the immune events generated in the lungs to the joints. However, currently, we lack studies on the citrullination patterns in other organs and in different conditions (both inflammatory and on inflammatory) that would help understanding if site-specific expression of citrullinated targets might be responsible for the secondary localization in the joints. Recently, investigation of a tissue collection of large lung samples and some few control tissues revealed the presence of citrullinated peptides in the lungs, in close relationship to the degree of tissue inflammation and low levels of expression in healthy tissues (such as lymph nodes, skeletal muscles and kidney) . This is in accordance with previous studies showing low amounts of citrullinated proteins, if any, in healthy non-inflamed tissues (such as synovial membrane and skeletal muscles) with upregulation during inflammation , but does not yet solve the question if site-specific shared immune targets could explain the joint preference of the extra-articular originating antibodies. A largely accepted hypothesis suggests the need for a second hit that will allow development of joint inflammation. According to this theory, some yet unknown stimuli primarily affecting the joints, at a time when ACPAs are already present in the blood stream, will promote local inflammation with exposure of citrullinated targets in a joint-specific pattern and/or in sufficient high amounts to overcome the threshold for the pathogenic capacity of the antibodies. Speculations on this second hit suggest minor joint trauma and/or subclinical infections as the favourite candidates, but with no evidence. By contrast, two recent publications raised a novel theory to explain propagation of the immune events from the lungs to the joints.
Prior to arthritis onset, a gradual increase in the levels and the number of different ACPA specificities occurs  and , associated with measureable signs of bone loss  but also diffuse symptoms such as pain and fatigue. The new reports demonstrate that ACPAs can specifically bind to the surface of developing osteoclasts (OCs) and promote bone loss both in vivo and in vitro. Interestingly, binding of ACPAs to the OCs occurs in conditions mimicking the physiological development of these cells in the absence of any other inflammatory stimuli. This intriguing observation was explained by identification of a novel essential mechanism for OC development, namely citrullination by PAD enzymes. In this way, OCs specifically residing in calcium-enriched environment are dependent on the PAD enzyme activity and express citrullinated targets on their surface in the absence of any other secondary stimuli. This might explain ACPAs binding to OCs but not other cells, with a consecutive release of gradually increasing interleukin (IL)-8, acting as an autocrine stimuli to further increase OC activation and bone destruction as demonstrated both in vitro in macrophage-derived OC cultures and in vivo in mice injected with ACPAs . Interestingly, the OC-dependent IL-8 secretion is responsible not only for bone loss but also for inducing pain-like behaviour in the absence of joint inflammation . Mechanisms responsible for the next transition, from the bone to the synovial membrane, remain to be investigated, but IL-8 might lead to a gradual accumulation of neutrophils with local release of NETs that will further stimulate synovial fibroblasts and generate chronic joint inflammation. Taken together, these novel data offer novel explanations for how and why antibodies generated at extra-articular sites might secondarily specifically localize into the joints and offer new ideas on how these new described pathogenic mechanisms might be used for specific targeting in ACPA-positive individuals presenting with pain and showing signs of bone loss but not yet having RA.
Summary and future agenda
The interaction between genes and environmental exposures is essential for development of ACPA and ACPA-positive RA. Large epidemiological investigations unrevealed the complexity of this interaction with both risk and protective genetic traits being essential in allowing an environmental insult (such as smoking) to lead to autoimmunity against posttranslational modified endogenous proteins. More recent findings offer novel insights on how an extra-articular-generated immune response might in genetic susceptible hosts localize in the joints to first induce bone loss and pain and only later on synovial inflammation. This new pathogenic scenario is exemplified in Fig. 3. Future research agenda should on one hand focus on a more refined investigation of the genetic–environmental interaction (to explore the role of microbiome and the relation of this interaction to more thorough defined autoimmunity) and on the other hand on the possibilities to target the novel pathogenic mechanisms recently described (such as IL-8 and OC targeting) in ACPA-positive individuals presenting with pain and bone loss but not yet having arthritis.
- The presence of both RA protective and risk alleles of the HLA class II occurring with different frequencies in the general population and interacting with a large array of environmental exposures (including smoking) requires a more thorough characterization of these interactions in well-defined subsets of disease and individuals at risk of developing RA.
- The demonstration of a novel pathogenic chain events responsible for bone loss and pain-like behaviour indicate the importance of further studies to address how this events leads to synovial inflammation.
- Identification of the pivotal role of osteoclasts, citrullination and IL-8 in bone loss and pain induction warrants future studies to address the potential therapeutic value of blocking strategies.
Schematic representation of the longitudinal development of ACPA-positive RA starting with the interaction between genes and environment and leading to chronic joint inflammation. In susceptible hosts, exposure to environmental factors (such as cigarette smoke) induces post-translational modifications of arginine (arg) containing peptides in to citrulline (cit)-containing peptides at mucosal surfaces (such as in the lung). Citrullinated peptides are presented by antigen-presenting cells (APCs) expressing certain MCH class II molecules to the immune system with consecutive activation of auto-reactive T and B cells and production of anti-citrullinated proteins antibodies (ACPAs) by plasma cells. These events lead to presence of ACPAs in the blood stream in the absence of any clinical symptom. Antibodies will bind citrullinated peptides expressed on the surface of developing osteoclasts (OCs) in the vicinity of the joints leading to gradually increase in interleukin (IL)-8 production and resulting in a clinical state characterized by ACPA positivity, joint pain and bone loss. Later on when enough IL-8 is produced locally this will first result in local recruitment of neturophils with consecutive chemoattraction of immune cells to the synovial membrane and activation of synovial fibroblast to result in chronic inflammation and ACPA-positive RA.
Conflict of interest statement
No conflicts of interest to declare.
-  K. Nishimura, D. Sugiyama, Y. Kogata, et al. Meta-analysis: diagnostic accuracy of anti-cyclic citrullinated peptide antibody and rheumatoid factor for rheumatoid arthritis. Annals of Internal Medicine. 2007;146:797-808
-  S. Rantapaa-Dahlqvist, B.A. de Jong, E. Berglin, et al. Antibodies against cyclic citrullinated peptide and IgA rheumatoid factor predict the development of rheumatoid arthritis. Arthritis & Rheumatology. 2003;48:2741-2749
-  D. van der Woude, S.R. Dahlqvist, A. Ioan-Facsinay, et al. Epitope spreading of the anti-citrullinated protein antibody response occurs before disease onset and is associated with the disease course of early arthritis. Annals of the Rheumatic Diseases. 2010;69:1554-1561
-  C.A. Janeway, M. Walport, P. Travers, et al. in: Immunobiology: the immune system in health and disease. vol. 6th (Garland Science Publishing, New York, 2005)
-  The IMGT/HLA database was accessed on December 11th 2015. http://www.ebi.ac.uk/ipd/imgt/hla/.
-  G.P. Astorga, R.C. Williams Jr. Altered reactivity in mixed lymphocyte culture of lymphocytes from patients with rheumatoid arthritis. Arthritis & Rheumatology. 1969;12:547-554
-  P. Stastny. Mixed lymphocyte cultures in rheumatoid arthritis. Journal of Clinical Investigation. 1976;57:1148-1157
-  P. Stastny. Association of the B-cell alloantigen DRw4 with rheumatoid arthritis. New England Journal of Medicine. 1978;298:869-871
- * P.K. Gregersen, J. Silver, R.J. Winchester. The shared epitope hypothesis. An approach to understanding the molecular genetics of susceptibility to rheumatoid arthritis. Arthritis & Rheumatology. 1987;30:1205-1213
-  R.B. Schittenhelm, T. Sian, P.G. Wilmann, et al. Revisiting the arthritogenic peptide theory quantitative not qualitative changes in the peptide repertoire of HLA-B27 allotypes. Arthritis & Rheumatology. 2015;67:702-713
- * T.W. Huizinga, C.I. Amos, A.H. van der Helm-van Mil, et al. Refining the complex rheumatoid arthritis phenotype based on specificity of the HLA-DRB1 shared epitope for antibodies to citrullinated proteins. Arthritis & Rheumatology. 2005;52:3433-3438
-  S.L. Mackie, J.C. Taylor, S.G. Martin, et al. A spectrum of susceptibility to rheumatoid arthritis within HLA-DRB1: stratification by autoantibody status in a large UK population. Genes and Immunity. 2012;13:120-128
-  E. Lundstrom, H. Kallberg, M. Smolnikova, et al. Opposing effects of HLA-DRB1*13 alleles on the risk of developing anti-citrullinated protein antibody-positive and anti-citrullinated protein antibody-negative rheumatoid arthritis. Arthritis & Rheumatology. 2009;60:924-930
-  S.T. du Montcel, L. Michou, E. Petit-Teixeira, et al. New classification of HLA-DRB1 alleles supports the shared epitope hypothesis of rheumatoid arthritis susceptibility. Arthritis & Rheumatology. 2005;52:1063-1068
-  I.E. van der Horst-Bruinsma, H. Visser, J.M. Hazes, et al. HLA-DQ-associated predisposition to and dominant HLA-DR-associated protection against rheumatoid arthritis. Human Immunology. 1999;60:152-158
-  N. de Vries, H. Tijssen, P.L. van Riel, et al. Reshaping the shared epitope hypothesis: HLA-associated risk for rheumatoid arthritis is encoded by amino acid substitutions at positions 67-74 of the HLA-DRB1 molecule. Arthritis & Rheumatology. 2002;46:921-928
-  D.L. Mattey, P.T. Dawes, M.A. Gonzalez-Gay, et al. HLA-DRB1 alleles encoding an aspartic acid at position 70 protect against development of rheumatoid arthritis. Journal of Rheumatology. 2001;28:232-239
- * D. van der Woude, B.A. Lie, E. Lundstrom, et al. Protection against anti-citrullinated protein antibody-positive rheumatoid arthritis is predominantly associated with HLA-DRB1*1301: a meta-analysis of HLA-DRB1 associations with anti-citrullinated protein antibody-positive and anti-citrullinated protein antibody-negative rheumatoid arthritis in four European populations. Arthritis & Rheumatology. 2010;62:1236-1245
-  H.S. Lee, K.W. Lee, G.G. Song, et al. Increased susceptibility to rheumatoid arthritis in Koreans heterozygous for HLA-DRB1*0405 and *0901. Arthritis & Rheumatology. 2004;50:3468-3475
-  S.L. Bridges Jr., J.M. Kelley, L.B. Hughes. The HLA-DRB1 shared epitope in Caucasians with rheumatoid arthritis: a lesson learned from tic-tac-toe. Arthritis & Rheumatology. 2008;58:1211-1215
-  S. Oka, H. Furukawa, A. Kawasaki, et al. Protective effect of the HLA-DRB1*13:02 allele in Japanese rheumatoid arthritis patients. PLoS One. 2014;9:e99453
-  G.B. Nordang, S.T. Flam, M.T. Maehlen, et al. HLA-C alleles confer risk for anti-citrullinated peptide antibody-positive rheumatoid arthritis independent of HLA-DRB1 alleles. Rheumatology (Oxford). 2013;52:1973-1982
-  B. Ding, L. Padyukov, E. Lundstrom, et al. Different patterns of associations with anti-citrullinated protein antibody-positive and anti-citrullinated protein antibody-negative rheumatoid arthritis in the extended major histocompatibility complex region. Arthritis & Rheumatology. 2009;60:30-38
-  C. Terao, K. Ohmura, Y. Kochi, et al. A large-scale association study identified multiple HLA-DRB1 alleles associated with ACPA-negative rheumatoid arthritis in Japanese subjects. Ann Rheum Dis. 2011;70:2134-2139
-  K.N. Verpoort, F.A. van Gaalen, A.H. van der Helm-van Mil, et al. Association of HLA-DR3 with anti-cyclic citrullinated peptide antibody-negative rheumatoid arthritis. Arthritis & Rheumatology. 2005;52:3058-3062
-  L. Bossini-Castillo, C. de Kovel, H. Kallberg, et al. A genome-wide association study of rheumatoid arthritis without antibodies against citrullinated peptides. Annals of the Rheumatic Diseases. 2015;74:10
-  S. Leslie, P. Donnelly, G. McVean. A statistical method for predicting classical HLA alleles from SNP data. American Journal of Human Genetics. 2008;82:48-56
-  S. Raychaudhuri, C. Sandor, E.A. Stahl, et al. Five amino acids in three HLA proteins explain most of the association between MHC and seropositive rheumatoid arthritis. Nature Genetics. 2012;44:291-296
-  B. Han, D. Diogo, S. Eyre, et al. Fine mapping seronegative and seropositive rheumatoid arthritis to shared and distinct HLA alleles by adjusting for the effects of heterogeneity. American Journal of Human Genetics. 2014;94:522-532
-  J. van Heemst, T.J.W. Huizinga, D. van der Woude, et al. Fine-mapping the human leukocyte antigen locus in rheumatoid arthritis and other rheumatic diseases: identifying causal amino acid variants?. Current Opinion in Rheumatology. 2015;27:256-261
-  L.J. Stern, J.H. Brown, T.S. Jardetzky, et al. Crystal-structure of the human class-II MHC protein HLA-DR1 complexed with an influenza-virus peptide. Nature. 1994;368:215-221
-  J.A. Hill, S. Southwood, A. Sette, et al. Cutting edge: the conversion of arginine to citrulline allows for a high-affinity peptide interaction with the rheumatoid arthritis-associated HLA-DRB1*0401 MHC class II molecule. Journal of Immunology. 2003;171:538-541
-  S.W. Scally, J. Petersen, S.C. Law, et al. A molecular basis for the association of the HLA-DRB1 locus, citrullination, and rheumatoid arthritis. Journal of Experimental Medicine. 2013;210:2569-2582
- * J. van Heemst, D. Jansen, S. Polydorides, et al. Crossreactivity to vinculin and microbes provides a molecular basis for HLA-based protection against rheumatoid arthritis. Nature Communications. 2015;6
-  S. Ling, E.N. Cline, T.S. Haug, et al. Citrullinated calreticulin potentiates rheumatoid arthritis shared epitope signaling. Arthritis & Rheumatology. 2013;65:618-626
- * A.H. Hensvold, P.K.E. Magnusson, V. Joshua, et al. Environmental and genetic factors in the development of anticitrullinated protein antibodies (ACPAs) and ACPA-positive rheumatoid arthritis: an epidemiological investigation in twins. Annals of the Rheumatic Diseases. 2015;74:375-380
-  L. Padyukov, C. Silva, P. Stolt, et al. A gene-environment interaction between smoking and shared epitope genes in HLA-DR provides a high risk of seropositive rheumatoid arthritis. Annals of Rheumatic Diseases. 2004;50:3085-3092
- * L. Klareskog, P. Stolt, K. Lundberg, et al. A new model for an etiology of rheumatoid arthritis: smoking may trigger HLA-DR (shared epitope)-restricted immune reactions to autoantigens modified by citrullination. Annals of Rheumatic Diseases. 2006;54:38-46
-  M. Pedersen, S. Jacobsen, M. Klarlund, et al. Environmental risk factors differ between rheumatoid arthritis with and without auto-antibodies against cyclic citrullinated peptides. Arthritis Research and Therapy. 2006;8:R133
-  E.W. Karlson, S.C. Chang, J. Cui, et al. Gene-environment interaction between HLA-DRB1 shared epitope and heavy cigarette smoking in predicting incident rheumatoid arthritis. Annals of Rheumatic Diseases. 2010;69:54-60
-  C.L. Too, A. Yahya, S. Murad, et al. Smoking interacts with HLA-DRB1 shared epitope in the development of anti-citrullinated protein antibody-positive rheumatoid arthritis: results from the Malaysian Epidemiological Investigation of Rheumatoid Arthritis (MyEIRA). Arthritis Research and Therapy. 2012;14:R89
-  P. Stolt, A. Yahya, C. Bengtsson, et al. Silica exposure among male current smokers is associated with a high risk of developing ACPA-positive rheumatoid arthritis. Annals of Rheumatic Diseases. 2010;69:1072-1076
-  Too CL, Muhamad NA, Padyukov L, et al: Gene-environment interaction between HLA-DRB1 shared epitope and occupational textile dust exposure in the risk of ACPA-positive rheumatoid arthritis in Female patients: evidence from the Malaysian epidemiological investigation of rheumatoid arthritis case-control study. Arthritis and Rheumatism 65:S457–S458.
-  D. Makrygiannakis, M. Hermansson, A.K. Ulfgren, et al. Smoking increases peptidylarginine deiminase 2 enzyme expression in human lungs and increases citrullination in BAL cells. Annals of the Rheumatic Diseases. 2008;67:1488-1492
-  M.A. Moscarello, D.D. Wood, C. Ackerley, et al. Myelin in multiple sclerosis is developmentally immature. The Journal of Clinical Investigation. 1994;94:146-154
-  D.R. Beniac, D.D. Wood, N. Palaniyar, et al. Cryoelectron microscopy of protein-lipid complexes of human myelin basic protein charge isomers differing in degree of citrullination. Journal of Structural Biology. 2000;129:80-95
-  T. Senshu, S. Kan, H. Ogawa, et al. Preferential deimination of keratin K1 and filaggrin during the terminal differentiation of human epidermis. Biochemical and Biophysical Research Communications. 1996;225:712-719
-  T. Senshu, K. Akiyama, K. Nomura. Identification of citrulline residues in the V subdomains of keratin K1 derived from the cornified layer of newborn mouse epidermis. Experimental Dermatology. 1999;8:392-401
-  C.R. Harding, I.R. Scott. Histidine-rich proteins (filaggrins): structural and functional heterogeneity during epidermal differentiation. Journal of Molecular Biology. 1983;170:651-673
-  D.J. Pearton, B.A. Dale, R.B. Presland. Functional analysis of the profilaggrin N-terminal peptide: identification of domains that regulate nuclear and cytoplasmic distribution. The Journal of Investigative Dermatology. 2002;119:661-669
-  I.R. Scott, C.R. Harding, J.G. Barrett. Histidine-rich protein of the keratohyalin granules. Source of the free amino acids, urocanic acid and pyrrolidone carboxylic acid in the stratum corneum. Biochimica et Biophysica Acta. 1982;719:110-117
-  D. Makrygiannakis, E. af Klint, I.E. Lundberg, et al. Citrullination is an inflammation-dependent process. Annals of the Rheumatic Diseases. 2006;65:1219-1222
-  K. Lundberg, C. Bengtsson, N. Kharlamova, et al. Genetic and environmental determinants for disease risk in subsets of rheumatoid arthritis defined by the anticitrullinated protein/peptide antibody fine specificity profile. Annals of Rheumatic Diseases. 2013;72:652-658
-  Y. Arnson, Y. Shoenfeld, H. Amital. Effects of tobacco smoke on immunity, inflammation and autoimmunity. Journal of Autoimmunity. 2010;34:J258-J265
-  J.M. Larsen, D.B. Steen-Jensen, J.M. Laursen, et al. Divergent pro-inflammatory profile of human dendritic cells in response to commensal and pathogenic bacteria associated with the airway microbiota. PLoS One. 2012;7:e31976
-  J. Sokolove, X. Zhao, P.E. Chandra, et al. Immune complexes containing citrullinated fibrinogen costimulate macrophages via Toll-like receptor 4 and Fcgamma receptor. Arthritis & Rheumatology. 2011;63:53-62
-  O. Snir, M. Rieck, J.A. Gebe, et al. Identification and functional characterization of T cells reactive to citrullinated vimentin in HLA-DRB1*0401-positive humanized mice and rheumatoid arthritis patients. Arthritis & Rheumatology. 2011;63:2873-2883
-  E.A. James, M. Rieck, J. Pieper, et al. Citrulline-specific Th1 cells are increased in rheumatoid arthritis and their frequency is influenced by disease duration and therapy. Arthritis & Rheumatology. 2014;66:1712-1722
-  M.F. Bellatin, M. Han, M. Fallena, et al. Production of autoantibodies against citrullinated antigens/peptides by human B cells. Journal of Immunology. 2012;188:3542-3550
-  T. Bongartz, C. Nannini, Y.F. Medina-Velasquez, et al. Incidence and mortality of interstitial lung disease in rheumatoid arthritis: a population-based study. Arthritis & Rheumatology. 2010;62:1583-1591
-  J. Chen, Y. Shi, X. Wang, et al. Asymptomatic preclinical rheumatoid arthritis-associated interstitial lung disease. Clinical and Developmental Immunology. 2013;2013:406927
-  A. Fischer, J.J. Solomon, R.M. du Bois, et al. Lung disease with anti-CCP antibodies but not rheumatoid arthritis or connective tissue disease. Respiratory Medicine. 2012;106:1040-1047
-  M.K. Demoruelle, M.H. Weisman, P.L. Simonian, et al. Brief report: airways abnormalities and rheumatoid arthritis-related autoantibodies in subjects without arthritis: early injury or initiating site of autoimmunity?. Arthritis & Rheumatology. 2012;64:1756-1761
-  V.C. Willis, M.K. Demoruelle, L.A. Derber, et al. Sputum autoantibodies in patients with established rheumatoid arthritis and subjects at risk of future clinically apparent disease. Arthritis & Rheumatism. 2013;65:2545-2554
-  G. Reynisdottir, R. Karimi, V. Joshua, et al. Structural changes and antibody enrichment in the lungs are early features of anti-citrullinated protein antibody-positive rheumatoid arthritis. Arthritis & Rheumatology. 2014;66:31-39
-  G. Reynisdottir, H. Olsen, V. Joshua, et al. Signs of immune activation and local inflammation are present in the bronchial tissue of patients with untreated early rheumatoid arthritis. Annals of Rheumatic Diseases. 2015;
-  X. Zhang, D. Zhang, H. Jia, et al. The oral and gut microbiomes are perturbed in rheumatoid arthritis and partly normalized after treatment. Nature Medicine. 2015;21:895-905
-  M.K. Demoruelle, J.M. Norris, V.M. Holers, et al. The lung microbiome differs in asymptomatic subjects at elevated risk of future rheumatoid arthritis compared with healthy control subjects. Annals of the American Thoracic Society. 2014;11(Suppl 1):S74
-  K.A. Kuhn, L. Kulik, B. Tomooka, et al. Antibodies against citrullinated proteins enhance tissue injury in experimental autoimmune arthritis. Journal of Clinical Investigation. 2006;116:961-973
-  L.A. Trouw, E.M. Haisma, E.W. Levarht, et al. Anti-cyclic citrullinated peptide antibodies from rheumatoid arthritis patients activate complement via both the classical and alternative pathways. Arthritis & Rheumatology. 2009;60:1923-1931
-  J. Suurmond, F. Rivellese, A.L. Dorjee, et al. Toll-like receptor triggering augments activation of human mast cells by anti-citrullinated protein antibodies. Annals of Rheumatic Diseases. 2015;74:1915-1923
-  K.L. Habets, L.A. Trouw, E.W. Levarht, et al. Anti-citrullinated protein antibodies contribute to platelet activation in rheumatoid arthritis. Arthritis Research and Therapy. 2015;17:209
-  A. Kleyer, S. Finzel, J. Rech, et al. Bone loss before the clinical onset of rheumatoid arthritis in subjects with anticitrullinated protein antibodies. Annals of Rheumatic Diseases. 2013;
-  E. Corsiero, M. Bombardieri, E. Carlotti, et al. Single cell cloning and recombinant monoclonal antibodies generation from RA synovial B cells reveal frequent targeting of citrullinated histones of NETs. Annals of Rheumatic Diseases. 2015;
-  M.C. Lu, C.L. Yu, H.C. Yu, et al. Anti-citrullinated protein antibodies promote apoptosis of mature human Saos-2 osteoblasts via cell-surface binding to citrullinated heat shock protein 60. Immunobiology. 2016;221(1):76-83
- * U. Harre, D. Georgess, H. Bang, et al. Induction of osteoclastogenesis and bone loss by human autoantibodies against citrullinated vimentin. The Journal of Clinical Investigation. 2012;122(5):1791-1802
-  U. Harre, S.C. Lang, R. Pfeifle, et al. Glycosylation of immunoglobulin G determines osteoclast differentiation and bone loss. Nature Communications. 2015;6:6651
-  A.J. Ytterberg, V. Joshua, G. Reynisdottir, et al. Shared immunological targets in the lungs and joints of patients with rheumatoid arthritis: identification and validation. Annals of Rheumatic Diseases. 2014;
-  E.B. Lugli, R. Correia, R. Fischer, et al. Expression of citrulline and homocitrulline residues in the lungs of non-smokers and smokers: implications for autoimmunity in rheumatoid arthritis. Arthritis Research and Therapy. 2015;17:9
-  M. Brink, M. Hansson, L. Mathsson, et al. Multiplex analyses of antibodies against citrullinated peptides in individuals prior to development of rheumatoid arthritis. Arthritis & Rheumatology. 2013 Apr;65(4):899-910
-  J. Sokolove, R. Bromberg, K.D. Deane, et al. Autoantibody epitope spreading in the pre-clinical phase predicts progression to rheumatoid arthritis. PloS One. 2012;7:e35296
- * A. Krishnamurthy, V. Joshua, A. Haj Hensvold, et al. Identification of a novel chemokine-dependent molecular mechanism underlying rheumatoid arthritis-associated autoantibody-mediated bone loss. Annals of Rheumatic Diseases. 2015;
- * G. Wigerblad, D.B. Bas, C. Fernades-Cerqueira, et al. Autoantibodies to citrullinated proteins induce joint pain independent of inflammation via a chemokine-dependent mechanism. Annals of Rheumatic Diseases. 2015;
- * F. Koning, R. Thomas, J. Rossjohn, et al. Coeliac disease and rheumatoid arthritis: similar mechanisms, different antigens. Nature Reviews Rheumatology. 2015 Aug;11(8):450-461
© 2016 Elsevier Ltd, All rights reserved.