By Gordon Lam, MD
Fellow, Division of Rheumatology
Johns Hopkins University School of Medicine
Release Date: May 1, 2006
Expiration Date: May 1, 2008
Dr. Lam has no significant financial interest or relationships to disclose.
One of the great advances in rheumatology recently has been the identification of citrullinated proteins as a specific marker of rheumatoid arthritis. This discovery has changed our diagnostic and prognostic rubric of the disease. Hence, today I would like to talk about citrullinated proteins, peptidylarginine deiminase (PAD), which is the enzyme involved in the citrullination process, and the possible roles that they may play in the pathogenesis of rheumatoid arthritis.
Rheumatoid arthritis is a common, systemic autoimmune disease, afflicting 1/2 to 1 percent of the population. Yet despite its prevalence and decades of research devoted to it, the exact etiology of rheumatoid arthritis is still unknown. Like all autoimmune diseases, many have postulated that infectious agents, environmental triggers, and host makeup may play roles in advent of autoimmunity. Theories have been postulated implicating periodontal disease, smoking, hormones, and genetics as associative factors in rheumatoid arthritis.
Despite the various unknowns, rheumatoid arthritis is categorically characterized by chronic inflammation of the joints. This leads to massive infiltration of inflammatory cells into the joint space and synovial lining, which cause the synovium to proliferate and pannus tissue to develop. Pannus then attacks the bone and cartilage, resulting in erosions and progressive joint destruction.
As with most diseases, serologic markers have been sought to better characterize rheumatoid arthritis. To date, rheumatoid factor is the only serologic parameter that is included in the American College of Rheumatology (ACR) criteria for the classification of rheumatoid arthritis. As many of you know, rheumatoid factor is an antibody directed against the Fc portion of IgG molecules. It is found in approximately 75% to 85% of rheumatoid arthritis patients. However, rheumatoid factor is non-specific. It can be seen in other diseases such as bacterial, viral, or parasitic diseases, and it can also be present in other inflammatory conditions such as cryoglobulinemia and sarcoidosis. Hence, rheumatoid factor is a good, but not ideal, marker for rheumatoid arthritis.
Better markers are needed to more accurately diagnosis disease. In rheumatoid arthritis, better markers would allow us to differentiate it from other inflammatory polyarthropathies and to identify disease earlier in its clinical course, which would hopefully lead to more effective interventions, since they could be introduced at a stage where the disease is more modifiable. Specific markers could help to elucidate the pathophysiology of disease, allowing new therapeutic approaches. The quest for better serologic markers in rheumatoid arthritis has been ongoing for decades and has brought us to where we are today — citrullinated antigens.
History of Citrullinated Antigens
Like most great revelations in science, the history of citrullinated antigens and their association with rheumatoid arthritis has been one of discovery, oversight, and rediscovery. In 1964, Neinhuis & Mandema(1) noted a specific staining pattern on indirect immunofluorescence of human buccomucosal cells from sera of rheumatoid arthritis patients. Based on the staining pattern of this unknown protein, they called the protein the perinuclear factor. Hence, the autoantibody was called the “anti-perinuclear factor.” It was not until 15 years later that Young & colleagues(2) recapitulated these observations by performing the exact same assay with RA sera, but this time on rat esophagus tissue. Again a specific staining pattern was noticed—this time towards epithelial filaments. Thus, they called the antibody response anti-keratin antibodies. It was not until twenty years later that two independent groups led by Schellekens(3) and Girbauld- Neuhauser(4) realized that the target of both the anti-perinuclear factor and the anti-keratin antibody was the same protein: fillagrin.
What is fillagrin? Fillagrin, also known as filament aggregating protein, is a protein associated with keratin filaments. However, it undergoes post-translational modifications that create citrulline residues by deaminating arginine. Subsequent studies demonstrated that the antibody production against fillagrin was absolutely dependent on the citrullination of the protein. Since then, many other studies have demonstrated that citrullinated antigens such as vimentin and fibrin are also candidate targets in rheumatoid arthritis.
This then begs the question: What is citrullination? As the diagram below shows, citrullination is a post-translational modification whereby the amino acid arginine is modified to the non-standard residue citrulline.
This is done by the enzyme peptidylarginine-deaminase, or PAD, in the presence of calcium. The major effect of this process is that arginine, which is a positively charged molecule, is now changed to a neutral, although still polar, molecule. Intramolecular interactions are lost allowing the protein to unfold, thus making it more susceptible to proteolytic cleavage by other enzymes. In addition, new intermolecular interactions can form.
It is plausible that the process of citrullination plays an integral role in autoimmunity. Multiple post-translational modifications have been implicated in several other autoimmune diseases: glutamic acid decarboxylase has been associated with Type I diabetes mellitus; myeloperoxidase in vasculitis; modified wheat gliadin in celiac disease; and pyruvate dehydrogenase in primary biliary cirrhosis, to name a few.
Schellekens(3) demonstrated that citrullinated peptides (and not their arginine counterparts) were specifically recognized by antibodies in RA sera using a citrullinated peptide assay. In an attempt to optimize the sensitivity of the assay, Shellekans created cyclic variants of the peptides, which may have allowed better exposure and increased access to relevant binding epitopes. This second generation CCP assay was created in 2002 and increased the sensitivity of the test to approximately 82% while maintaining its high specificity of 98 – 99%.
Many studies have verified the validity of the anti-CCP assay and found that it has a similar level of sensitivity for rheumatoid arthritis as the test for rheumatoid factor, i.e. approximately 80%. However, the CCP assay has a far superior level of specificity, making it a more ideal marker than rheumatoid factor. In fact, 40% of rheumatoid factor seronegative patients with clinical signs and symptoms of rheumatoid arthritis are CCP positive. Therefore, the combination of the two tests greatly enhances the diagnostic accuracy of the disease. Both are used in clinical practice routinely today.
In addition to its utility in diagnosing rheumatoid arthritis, anti-CCP antibodies may play a role in the prognosis of disease, as well. Rantapaa-Dahlqvist(5) studied 83 patients with rheumatoid arthritis who had happened to donate blood prior to the development of clinical disease. In taking single-point serum samples from these 83 patients, they were able to detect anti-CCP antibodies in 34% of them before the clinical manifestations of disease. Interestingly, anti-CCP antibodies were detected in some sera samples up to nine years before the onset of symptoms. In a similar study by Nielen et al(6), 79 patients with rheumatoid arthritis who had donated blood serially throughout time prior to the onset of their disease were analyzed. 49% had anti-CCP antibodies and/or IgM rheumatoid factor present in their sera before the clinical onset of disease, at a median of 4.5 years before clinical manifestations. The conclusions from these studies are threefold: (1) anti-CCP antibodies can be present several years before the onset of symptoms of rheumatoid arthritis; (2) anti-CCP antibodies can predict the development of rheumatoid arthritis, and; (3) the anti-CCP assay in combination with IgM rheumatoid factor may provide better prognostic indications of disease.
To carry this last thought forward, many groups have studied the presence of anti-CCP antibodies and their correlation with erosive disease in rheumatoid arthritis. Kroot et al(7) noted that anti-CCP positive rheumatoid arthritis patients had significantly more radiologic damage after six years than their anti-CCP negative disease counterparts. Meyer et al(8) demonstrated that anti-CCP positive rheumatoid arthritis patients had a greater likelihood of increased erosive damage after five years compared to rheumatoid factor negative controls (67%; OR:2.5). And lastly, Forslind et al(9) noted that anti-CCP positive rheumatoid arthritis patients had significantly more erosions at two years compared to rheumatoid factor positive patients.
While the association of anti-CCP antibodies with RA is intriguing, equally compelling was the realization that anti-CCP antibodies were present before the manifestation of clinical symptoms. This supports a paradigm of disease that has three distinct phases. Phase 1 is the state of susceptibility, whereby an individual is predisposed to developing a disease, perhaps in the form of HLA or non-HLA gene susceptibility. In phase-2, or the pre-clinical phase of the disease, autoimmunity is triggered by some endogenous or environmental mechanism, and autoantibodies begin to form. In the case of rheumatoid arthritis, these antibodies are the anti-CCP antibodies and rheumatoid factor. The autoimmune response is magnified, perhaps via epitope spreading, leading to diversity and clinical manifestation once the response begins to cause tissue destruction. Thereafter, phase 3becomes evident, the stage where symptoms and clinical disease presents. In rheumatoid arthritis, this is detected by radiographic findings of bone erosions.
Is there a genetic association between citrullination and rheumatoid arthritis?
The shared epitope is a sequence of amino acids on HLA class 2 molecules that is associated with rheumatoid arthritis. It is present on HLA-DRB1 alleles 0101, 0401, and 0404. These molecules have a peptide-binding pocket (P4) that is positively charged. One could postulate that in patients with the shared epitope, arginine (which is positively charged) is less likely to bind to the P4 pocket, as like charges repel. Instead, citrullinated molecules, which are neutral but polar, are more likely to bind P4. This would then initiate T-cell activation, B cell production of antibodies to citrullinated proteins, and the clinical manifestations of rheumatoid arthritis.
Alternatively, in patients without the shared epitope who are less susceptible to rheumatoid arthritis, the HLA class 2 molecule has a P4 binding pocket that is negatively charged. One could postulate that arginine, having an opposite charge, is more likely to bind to this P4 pocket, which then leads to negative selection and peripheral tolerance, thus allowing for the lack of clinical development of rheumatoid arthritis.
In a landmark study, Suzuki et al(10) used complex disease gene mapping with single nucleotide polymorphisms to identify functionally relevant polymorphisms of the gene PADI4, thus implicating it as a rheumatoid arthritis gene. It was hypothesized that rheumatoid arthritis susceptibility haplotype in PADI4 produced more stable PAD transcripts, allowing increased levels of citrullinated proteins and higher levels of antibodies to them in rheumatoid arthritis patients. Conversely, in non-susceptible individuals, less stable PAD4 transcripts would theoretically abrogate clinical manifestations of rheumatoid arthritis since there would be decreased enzymatic activity, decreased citrullination of proteins, and hence, decreased autoantibody response. Replication studies have confirmed these initial findings(14), and the haplotypes have been found in other ethnic populations.
There are five different isotypes of peptidylarginine deiminase, each with a specific expression pattern and target protein. However, all are dependent upon calcium for their enzymatic activity.
|PAD1||epidermis, uterus||carotene, filagrin|
|PAD2||in cell cytoplasm of skeletal muscle, brain, spleen, and secretory glands||vimentin in skeleton and macrophages; myelin-basic protein in CNS|
|PAD3||hair follicles||thought to be trychohyalin|
|PAD4||White blood cells, specifically granulocytes and monocytes; found in the nucleus||Histones and nucleophosmin (B23)|
Of these 5 isotypes, PAD2 and PAD4 are most likely pathogenic in rheumatoid arthritis. PAD2 positive macrophages have been found in inflamed synovium, and in ionomycin-treated macrophages, one of its target proteins is vimentin, which is a known RA autoantigen. As mentioned previously, PAD4 was demonstrated to be a susceptibility locus for rheumatoid arthritis in various ethnic groups. Other studies done with immunohistochemistry noted expression of PAD4 in hematological and rheumatoid synovial tissue.
The story of citrullinated peptides and their association with rheumatoid arthritis has been one of the great discoveries in the field of rheumatology. Post-translationally modified proteins have been implicated in various autoimmune diseases, and citrullination is an example of one such post-translational process. Assays to the cyclic variants of citrullinated proteins have been found to be highly sensitive and specific in the diagnosis of rheumatoid arthritis. Equally important, antibodies to cyclic citrullinated peptides can predate the onset of disease manifestations, and hence, can play a role in disease prognosis. Peptidylarginine deiminase is an enzyme that catalyzes the citrullination process, which converts arginine residues to citrulline. Genetic associations with PAD4 have been identified with rheumatoid arthritis.
Taken together, these data strongly suggest that PAD and citrullinated proteins play vital roles in the pathogenesis of rheumatoid arthritis. However, many questions remain. Which PAD isotypes are relevant in rheumatoid arthritis? What are the pathologic stimuli that activate PAD? Is PAD itself an autoantigen? What are the relevant citrullinated antigens that drive rheumatoid arthritis? What is the role of environmental triggers, such as infections, smoking, etc., in the citrullination of autoantigens or in PAD activation? Hopefully, with the efforts of ongoing research, the answers to these questions will be found in the near future.
- Nienhuis R, Mandema E. Ann Rheum Dis 1964;23:302-5
- Young B, Mallya R, Leslie R, Clark C, Hamblin T. BMJ 1979;ii:97-9
- Schellekens G, de Jong B, van den Hoogen F, van de Putte L, van Venrooij W. J Clin Invest 1998;101:273-281.
- Girbal-Neuhauser E, Durieux J, Arnaud M, Dalbon P, Sebbag M, Vincent C, Simon M, Senshu T, Masson-Bessiere C, Jolivet-Reynaud C, Jolivet M, Serre G. J Immunol 1999;162:585-594.
- Rantapaa-Dahlqvist S, de Jong W, Berglin E, Hallmans G, Wadell G, Stenlund H, Sundin U, van Venrooij W. Arthritis Rheum 2003;48:2741-2749.
- Nielen M, van Schaardenburg D, Reesink H, van de Stadt R, van der Horst-Bruinsma I, de Koning M, Habibuw M, Vandenbroucke J, Dijkmans B. Arthritis Rheum 2004;50:380-386.
- Kroot E, de Jong B, van Leeuwen M, Swinkels, H, van den Hoogen FH, van’t Hof M, van de Putte LB, van Rijswijk MH, van Venrooij WJ, van Riel PL. Arthritis Rheum 2000;43:1831-5.
- Meyer O, Labarre C, Dougados M, Goupille P, Cantagrel A, Dubois A, Nicaise-Roland P, Sibilia J, Combe B. Ann Rheum Dis 2003;62:120-6.
- Forslind K, Ahlmen M, Eberhardt K, Hafstrom I, Svensson B; BARFOT Study Group. Ann Rheum Dis 2004;63:1090-5.
- Suzuki A, Yamada R, Chang X, et al. Nat Genet 2003;34:395-402.
- Barton A, Bowes J, Eyre S, Spreckley K, Hinks A, John S, Worthington J. Arthritis Rheum 2004;50:1117-21.
- Caponi L, Petit-Teixeira E, Sebbag M, Bongiorni F, Moscato S, Pratesi F, Pierlot C, Osorio J, Chapuy-Regaud S, Guerrin M, Cornelis F, Serre G, Migliorini P; ECRAF. Ann Rheum Dis 2005;64:587-593.
- Martinez A, Valdivia A, Pascual-Salcedo D, Lamas JR, Fernandez-Arquero M, Balsa A, Fernandez-Gutierrez B, de la Concha EG, Urcelay E. Rheumatology 2005;44:1263-1266.
- Ikari K, Kuwahara M, Nakamura T, Momohara S, Hara M, Yamanaka H, Tomatsu T, Kamatani N. Arthritis Rheum 2005;52:3054-7.