THE ROLE OF CITRULLINATION IN THE DEVELOPMENT
OF MOUSE AND HUMAN INFLAMMATORY ARTHRITIS
by
VAN CRANSTON WILLIS
B.S., Brigham Young University, 2007
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Doctor of Philosophy
Molecular Biology Program
2012
This thesis for the Doctor of Philosophy degree by
Van Cranston Willis
has been approved for the
Molecular Biology Program
by
James DeGregori, Chair
V. Michael Holers, Advisor
Kevin Deane
Kathryn Haskins
Linda van Dyk
Date 11/20/12
ii
Willis, Van, Cranston (Ph.D., Molecular Biology)
The Role of Citrullination in the Development of Mouse and Human Inflammatory
Arthritis
Thesis directed by Professor V. Michael Holers
ABSTRACT
Rheumatoid arthritis (RA) is associated with the development of autoantibodies to citrullinated protein antigens (ACPAs). Citrullination is the post-translational modification of arginine residues into citrulline by protein arginine deiminases (PADs).
ACPAs can be present years before the onset of clinical disease and represent an increased risk for future development of RA. While high ACPA titers are associated with increased disease severity, the fact that ACPAs are present for years prior to the onset of clinically apparent disease questions their contribution, and the contribution of citrullination, to the pathogenesis of RA. In order to understand the contribution of
ACPAs and citrullinated epitopes to the pathogenesis of disease, mice with collagen- induced arthritis (CIA) were treated with the PAD inhibitors Cl-amidine, a pan-PAD inhibitor, or GSK283, a PAD4 inhibitor. Mice treated with either PAD inhibitor showed reduced clinical and histological disease severity, eptiope spreading, and in the case of
Cl-amidine, total citrulline levels in the joint and serum. Interestingly epitope spreading in mice treated with PAD inhibitors was reduced in a citrulline-independent manner, suggesting that the ameliorative effects of PAD inhibitor treatment are not necessarily derived from the elimination of citrullinated epitopes and/or ACPA. Cl-amidine treatment failed to ameliorate collagen antibody-induced arthritis (CAIA), suggesting that the effector phase of the arthritis response is unaffected by PAD inhibition. As part of a
iii
CCTSI-funded clinical research project, sputum samples were obtained from subjects at- risk for the future development of RA. These samples were evaluated for the presence of
RA-related autoantibodies (ACPAs and rheumatoid factors [RF]) to test the hypothesis that autoimmunity may begin in the lung in RA. RA-related autoantibodies were detected in the sputa of at-risk subjects as well as patients with early RA (≤ 1 year of diagnosis) as compared to healthy controls. Furthermore, the detection of RA-related autoantibodies in the absence of serological detection of these same autoantibodies suggests that in in some subjects, RA may begin in the lung. Together these data suggest that PAD inhibition ameliorates disease by altering the immune response more broadly than eliminating citrullinated epitope availability and that in some cases RA may begin in the lung.
The form and content of this abstract are approved. I recommend its publication.
Approved: V. Michael Holers
iv
TABLE OF CONTENTS
CHAPTER
I. INTRODUCTION...... 1
General overview of rheumatoid arthritis...... 1
General disease background...... 1
Genetic contributions...... 1
Environmental contributions and citrullination...... 3
Extra articular manifestations of disease...... 5
Protein Arginine Deiminases (PADs)...... 7
Antibodies to Citrullinated Protein Antigens (ACPAs)...... 8
Summary of pathogenesis...... 10
PAD inhibitors as a tool to investigate the pathogenesis of disease...... 11
Animal models of disease...... 13
Dissertation aims ...... 14
II. MATERIALS AND METHODS...... 15
Mice...... 15
Collagen-induced arthritis ...... 15
Collagen antibody-induced arthritis ...... 16
PAD inhibitor treatments...... 17
Measurement of total citrulline...... 17
Synovial proteome autoantibody microarrays...... 18
Autoantigen reactivity by ELISA...... 19
Flow cytometric analysis of immune cell populations ...... 19
Measurement of antibodies...... 20
Measurement of antibody-producing cells ...... 21 v
mRNA isolation and determination of transcript expression ...... 22
Evaluation of antigen presentation ...... 23
Determination of macrophage polarization ...... 24
Human subjects recruitment and consent ...... 25
Sputa collection and processing ...... 25
Human sample autoantibody testing...... 26
Statistical analysis...... 26
III. CL-AMIDINE TREATMENT AMELIORATES CIA ...... 28
Background and rationale...... 28
Results ...... 30
Cl-amidine treatment reduces clinical and histological disease severity in CIA...... 30
Cl-amidine treatment reduces serum and synovial citrulline content in CIA...... 31
Cl-amidine treatment does not alter immune cell populations...... 34
Table I. Frequency of immune cells in Cl-amidine-treated and control mice with CIA...... 36
Cl-amidine treatment does not ameliorate CAIA...... 36
Treatment with Cl-amidine decreases IgG2a and IgG1 anti-mouse CII antibodies...... 37
Treatment with Cl-amidine decreases antibody responses to both native and citrullinated synovial epitopes...... 39
Summary and conclusions...... 43
IV. PAD4 INHIBITION IS SUFFICIENT FOR THE AMELIORATION OF CIA...... 48
Background and rationale...... 48
Results ...... 49
Treatment with GSK283 decreases clinical disease severity in CIA. .49
vi
Treatment with GSK283 decreases histological disease severity in CIA...... 50
Treatment with GSK283 does not significantly alter joint total citrulline levels...... 52
Treatment with GSK283 does not significantly alter CII autoantibody production...... 56
Treatment with GSK283 reduces epitope spreading in CIA...... 60
Summary and conclusions...... 63
V. INVESTIGATION OF THE MECHANISM OF ACTION FOR ARTHRITIS AMELIORATION BY PAD INHIBITORS ...... 66
Background and rationale...... 66
Results ...... 69
Cl-amidine treatment does not affect autoantibody-producing cell population numbers...... 69
Cl-amidine treatment does not affect CD4+CD25+FoxP3+ cell numbers...... 70
Cl-amidine treatment does not alter antigen presentation...... 73
Cl-amidine treatment does not increase p53 expression or activity. ...73
Cl-amidine treatment alters complement protein gene expression...... 76
Cl-amidine treatment inhibits M1 macrophage polarization...... 80
Summary and conclusions...... 80
VI. DETECTION OF AUTOANTIBODIES IN THE SPUTA OF SUBJECTS AT-RISK FOR FUTURE RA ...... 85
Background and rationale...... 85
Results ...... 86
Subject demographics...... 86
Sputa autoantibody levels and positivity...... 86
Sputa versus sera autoantibody positivity...... 89
vii
Summary and conclusions...... 89
VII. DISCUSSION ...... 93
The contribution of citrullination to disease...... 93
The contribution of PAD4 to disease ...... 97
The effects of PAD inhibition on macrophage polarization...... 102
The detection of autoantibodies in the sputa of subjects at-risk for future RA ...103
VIII. FUTURE DIRECTIONS...... 105
REFERENCES ...... 110
viii
LIST OF TABLES
Table
I. Frequency of immune cells in Cl-amidine-treated and control mice with CIA...... 36
II. Absolute number (x106) of immune cells in Cl-amidine-treated and control mice with CIA...... 36
III. Subject characteristics and autoantibody positivity in sera and sputa...... 87
ix
LIST OF FIGURES
Figure
1. Cl-amidine treatment reduces clinical disease activity and joint destruction in CIA....33
2. Cl-amidine treatment reduces serum and synovial citrulline content in vivo...... 35
3. Cl-amidine treatment does not affect passively transferred arthritis...... 38
4. Cl-amidine treatment reduces IgG1 and IgG2a autoantibody response to mouse but not bovine CII...... 41
5. Cl-amidine treatment reduces autoantibody responses in CIA...... 42
6. PAD4 inhibition decreases clinical disease severity in CIA...... 51
7. GSK283 treatment reduces histological disease severity...... 54
8. GSK283 treatment does not significantly alter joint total citrulline levels...... 55
9. Plots of GSK283 serum levels over time...... 58
10. GSK283 treatment does not reduce autoantibody titers in CIA...... 59
11. GSK283 treatment reduces epitope spreading in CIA...... 62
12. Cl-amidine treatment does not decrease the number of autoantibody producing cells in CIA...... 71
13. Cl-amidine treatment does not decrease Treg cell numbers in the spleen or lymph node...... 72
14. Cl-amidine treatment does not affect antigen presentation...... 74
15. Cl-amidine treatment does not increase p53 or p21 mRNA levels...... 75
16. Cl-amidine treatment does not affect spleen or lymph node cell apoptosis...... 77
17. Cl-amidine treatment reduces C3 mRNA levels in the joint and spleen...... 78
18. Cl-amidine treatment increases splenic expression of factor H but does not affect complement protein expression in the knee at day 35...... 79
19. Cl-amidine treatment inhibits M1 polarization in vitro...... 81
x
20. Autoantibody levels in the sputa...... 88
21. Numbers of subjects positive for autoantibodies in the sputum only, serum only, neither, or both...... 90
xi
CHAPTER I
INTRODUCTION
General overview of rheumatoid arthritis
General disease background.
Rheumatoid arthritis (RA) is an autoimmune arthritis that affects 0.5-1.0% of the adult population (1) with a gender bias of a 2-3:1 female:male ratio. Symptoms include pain and swelling of the small peripheral joints. Debilitating joint destruction resulting from bone erosion and pannus formation is characteristic of RA. Pannus is a hyperplastic growth of the synovial lining associated with immune cell accumulation that invades the surrounding joint promoting bone erosion and joint destruction (reviewed in (2)). While the primary manifestation of disease is inflammation of the joints, RA is a systemic autoimmune disease and complications from RA affecting the cardiovascular system, lungs, skin, bones and eyes are proposed components of the increased mortality observed in RA patients (3). About 60% of patients with RA test positive for antibodies to citrullinated protein antigens (ACPAs) (4) which are highly specific for the disease and can be present for years before clinical onset of disease (5-7).
Genetic contributions.
The most significant genetic risk allele for RA resides in the human leukocyte antigen (HLA) genes. HLA molecules are expressed on antigen-presenting cells (e.g. macrophages, dendritic cells and B cells) and function to bind and present peptide antigens to T cells as part of the immunologic response. The HLA-DRB1 allele has been characterized in detail as the major genetic component of disease in RA (8, 9). The DRB1 1 alleles that impart risk for developing RA have a region of highly similar sequence that has been identified as the “shared epitope” (SE). The SE hypothesis proposes that these shared amino acids, Q(R)/KRAA, in positions 67-74 of the protein encoded by the HLA
DRB1 gene are responsible for the genetic susceptibility for RA (10). The amino acids in the SE motif coordinate the binding of peptides to be presented on HLA molecules and are thus thought to contribute to autoimmunity by preferentially presenting self peptides, especially citrullinated peptides (11). Supporting this hypothesis are genetic studies showing that SE alleles are specifically associated with the production of ACPAs (12-
14), and there is a complex hierarchy of risk conferred by the various DRB1 risk genotypes (15-17).
HLA alleles are estimated to account for one third of the genetic susceptibility to
RA (18). PTPN22, a protein tyrosine phosphatase associated with T and B cell signaling, is the next strongest genetic risk factor for RA after HLA-DRB1 (19). It is associated with Caucasian, but not Asian populations and is a risk allele for various autoimmune diseases with a humoral component. Knockout mouse experiments suggest that the
PTPN22 risk allele may enhance T-cell activation and increase antibody production (20).
Following PTPN22 in proportionate contribution to RA risk is TRAF1-C5 (21). This risk allele lies between the TRAF1 and C5 genes, both important in immunological responses.
C5 encodes complement factor 5, a crucial innate response factor in collagen-induced arthritis (CIA) (22, 23). Its impact on human disease is unclear since clinical trials targeting the C5a receptor in humans have failed to reduce synovial inflammation (24).
TNF receptor-associated factor 1 (TRAF1) is an adaptor protein involved in linking TNF family members to downstream signaling (25). TRAF1 is known to be antiproliferative
2
(26) and influence CD40 signaling (27), an interesting finding given that a CD40 allele has also been linked to increased risk for RA (28). There are a series of other signaling and immunologic genes that have been identified as contributing modest risk increases for RA such as STAT4 (29), the 6q23 region (30, 31), and the 4q27 region (32, 33).
Perhaps the most notable non-HLA risk allele in RA is the association between
PADI4 and RA that has been observed in Asian (34), but not Caucasian (35) populations.
This PADI4 association provides another link to citrullination in RA in addition to HLA
DRB1 alleles. Follow-up studies on the PADI4 polymorphism showed increased expression of anti-CCP antibodies (34) and increased enzyme activity (36) in the presence of the susceptible haplotype.
Environmental contributions and citrullination.
Critical to the genetic observations about susceptibility to RA is the low concordance rate (12-15%) between identical twins (37). This suggests that other factors, such as environmental influences, in addition to genetic factors make major contributions to the development of disease. A number of environmental influences are known to contribute to the risk of developing RA. By far the most influential environmental contributor to RA is smoking. Smoking contributes 18-25% of the population burden of
RA (38, 39), is dose related and stronger in males and in the ACPA+ RA subset. There is an apparent latency of 10-20 years from cessation of smoking to reduction in disease risk
(40). Studies examining the interactions between smoking and the SE HLA alleles suggest that their interaction increases the risk of RA (41), but follow-up studies indicate that this interaction may be ethnically or geographically restricted (42-44). Limited analyses indicate that passive smoking and smokeless tobacco are not been associated 3 with increased RA risk (38, 45). Emerging investigations are evaluating the interaction of smoking with PTPN22 and PADI4 risk alleles (46-48).
In addition to smoking, other environmental risk factors for RA are related to lung exposure. For example, air pollution (49) has been linked to increased risk for developing
RA, but its importance is still unclear (50). Additionally, the inhalation of silica dust or mineral oil have been suggested as triggers for lung inflammation that may result in immune dysregulation in RA (51).
Various lines of evidence suggest that the microbiome (reviewed in (52)) and/or infection of mucosal surfaces with specific microorganisms, such as Proteus mirabilis and Porphrymonas gingivalis, may be associated with RA (53-55). This agrees with findings in RA patients of alterations in commensal bacteria in the gut (56), suggesting that microorganisms may play a role in disease progression as well as initiation.
Finally, oral contraceptive use (57) and dietary factors such as omega-3 fatty acids (58) have demonstrated a protective effect in RA, while low levels of serum anti- oxidants beta-carotene and alpha-tocopherol are proposed to increase risk (59).
Many of the identified environmental risk factors for RA have a connection to citrullination and ACPA+ disease. Smoking is known to upregulate the expression of the citrullinating enzyme protein arginine deiminase 2 (PAD) in the lung and increase the citrullination of bronchoalveolar lavage (BAL) cells (60). Porphrymonas gingivalis, one of the primary agents of periodontal disease, expresses a deiminase that has the potential to citrullinate various human and bacterial proteins (61, 62) whose immune recognition may cross-react with joint antigens to initiate disease (63). Interestingly periodontal
4 disease is closely associated with RA (64-66), but it is unclear if periodontitis is a causal factor for RA (67, 68) or if RA is a causal factor for periodontitis (69).
Extra articular manifestations of disease.
As a systemic autoimmune disease RA manifests in extra-articular sites in addition to the joints. Cardiovascular complications (70) such as heart failure (71, 72) and myocardial infarction are well documented in RA, have been linked to seropositivity
(73), and can be reduced when RA is treated (74). A systemic extra-articular manifestation of RA is increased infection that correlates with disease activity (75).
Lymphoma incidence is not only increased but seems related to the severity of RA (76).
Significant increases in osteoporosis (77) are also seen in RA, which have been confirmed as independent of corticosteroids (78-80) used for the treatment of RA.
The importance of the extra-articular manifestations of RA is highlighted by recent studies that have extended findings of increased interstitial lung disease (ILD) (81) and lung abnormalities (82) in RA. Emerging studies of at-risk and early RA subjects demonstrate growing evidence of lung abnormalities and ILD in early RA (83) and even seropositive individuals that have yet to develop clinical RA (84, 85). Other studies focusing on the gut (86) and periodontia (87-89) suggest that the pulmonary-oral tract may be an extra-articular site of RA-related immune dysregulation.
A growing body of evidence supports the hypothesis that the pulmonary-oral tract is an extra-articular site of RA-related immune dysregulation: 1) The strong association of inhaled substances, such as tobacco pollution and dust with increased risk for RA (41,
49, 90), 2) the high prevalence of lung disease in patients with RA, even those with early
RA (82, 83, 91, 92), 3) case series detailing findings of lung disease in several patients 5 accompanied by elevated RA autoantibody levels, which later developed symptomatic
RA (84, 85, 93). In addition to these RA-specific findings, other findings relevant to lung biology further support this argument. Data from myositis, emphysema and cystic fibrosis, while not of RA origin, demonstrate that autoimmune responses can be mounted in the lung (94-96). With regard to environmental exposures in the lung initiating a systemic immune response, antigens encountered in the lung can induce a systemic immune response through presentation by antigen presenting cells to lymphocytes in regional lymph nodes (97). Worth noting is the formation of ‘inducible bronchus- associated lymphatic tissue’ (iBALT) in response to infection or environmental factors
(98). iBALT consists of mature lymphoid tissue containing B cells, T cells and antigen presenting cells. Once formed iBALT are capable of initiating localized immune responses and antibody production. iBALT related antibodies are mostly IgA (although
IgG is generated as well) and secreted onto the mucosal surface (99). Of note, iBALT observed in RA patients produce plasma cells generating ACPAs and RF and are thought to contribute to the pathogenesis of lung disease in RA. These same iBALT were associated with increased levels of ACPAs in BAL, strongly suggesting the production of
ACPA in the lungs of these patients (100). Finally, while porphrymonas gingivalis or other oral bacteria that are associated with RA may contribute to RA-related autoimmunity, they may do so in the lung in addition to or instead of at the oral site. It is possible that they influence autoimmunity in the lung after micro-aspiration to the lungs.
Supporting this, Porphrymonas gingivalis is involved in airways inflammation in acute pneumonia in patients with cystic fibrosis (101, 102).
6
Protein Arginine Deiminases (PADs).
Citrulline is a non-coded amino acid formed by post-translational modification of arginine to citrulline. PADs are the only enzymes known to catalyze protein citrullination. There are 5 PAD isozymes in mammals, all encoded on a single gene cluster on chromosome 1 (1p36) (reviewed in (103)). PADs have varied tissue-expression patterns in humans and mice: PADs 1, 3 and 6 (named so for historical reasons) are expressed in the skin, hair follicles and early embryo respectively (104-108). PAD4 is mainly expressed in white blood cells such as neutrophils and macrophages, but has been detected in tumors and cell lines (109-112). PAD2 is nearly ubiquitously expressed at high levels (113-115). With regard to RA, PAD2 and PAD4 are of most interest since they are expressed in mast cells, neutrophils and macrophages--cell types with the potential to affect inflammation and immunologic responses.
All PAD enzymes are located in the cytoplasm (116), with the exception of
PAD4, which can translocate to the nucleus (117) and citrullinate various nuclear proteins including histones (110, 112, 118, 119). The subcellular localization of these enzymes is especially pertinent to rheumatoid arthritis given that sites of inflammation show elevated levels of both intracellular and extracellular proteins (120), and this distinction may be important in autoimmune response to citrullinated proteins.
The structure and function of PAD enzymes themselves is well defined through experiments using PAD4 as a model enzyme (121, 122), which have revealed that PADs require Ca+ (> 10-5 mol/l) to attain a conformation that exposes the active site of the protein (121, 123). The catalytic activity of PADs is analogous to that of cysteine proteases (124-126) where the active site cysteine initiates an attack on the iminium
7 carbon of an arginine residue to form a tetrahedral intermediate. The loss of ammonia from this reaction forms a covalent intermediate that is hydrolyzed releasing both enzyme and a now citrullinated protein target.
The physiological roles of PADs are poorly understood, but PAD activity and citrullination have been linked to apoptosis, cellular differentiation, neutrophil extracellular trap (NET) formation (127), antigen processing in autophagy (128, 129), and oogenesis (111, 130, 131). Citrullination is also observed during inflammation (132), the cornification of skin (133), chemokine regulation (134-136), during spinal cord injury repair (137), and inflammatory bowel disease (138). The citrullination of proteins has been shown to affect protein conformation (133), protein-protein and ligand-receptor interactions (135, 139, 140) as well as gene expression (141, 142). Of note, there is evidence of autocitrullination of PAD4 (143, 144), suggesting that citrullination may be regulated by PAD activity itself. Studies utilizing knockout mice for PAD2 (145) and
PAD4 (146) have not reported any overt phenotypes suggesting physiological roles for
PADs.
Antibodies to Citrullinated Protein Antigens (ACPAs).
Although citrullination is a normal (if poorly understood) physiological process, humoral autoimmunity to citrullinated proteins (in the form of ACPA) is a phenomenon specific to RA (147-149). The presence of ACPA up to years before the onset of disease
(5-7) and their correlation with disease severity (150, 151) suggest they may play a role in the pathogenesis of arthritis. The impact that ACPA may have on disease development is further highlighted by the fact that the development of RA in ACPA- patients appears to be genetically distinct from that of ACPA+ patients (37, 152). 8
While there is a significant relationship between disease severity and ACPAs, the question of whether reactivity to citrullinated antigens is pathogenic remains. Evidence supporting the pathogenicity of ACPAs comes from several in vivo studies. The first showed that mice tolerized to citrullinated peptides developed arthritis that was ~60% less severe than controls and that mice treated with a monoclonal antibody recognizing citrullinated peptides amplified disease severity (153). Another study observed that transgenic mice containing the human HLA DRB1-0401 allele immunized with citrullinated human fibrinogen, but not native fibrinogen or mouse fibrinogen, develop arthritis (154). Third, Uysal et. al demonstrated the pathogenic capability of antibodies to citrullinated CII in vivo, including a detailed analysis of the interaction of citrullinated peptides with Fab complexes from these same antibodies (155). Immune complexes to citrullinated fibrinogen induced TNF α production in macrophages from an in vitro model of human disease (156) and ACPAs from RA patients have been shown to activate complement (157).
If ACPAs seem to be pathogenic then why is it that ACPAs can be present so long before disease onset? Various studies have investigated in some detail the ACPA response in early and pre-clinical arthritis. These studies show that epitope spreading
(158, 159) and avidity maturation (160) occur before disease onset and that the specific reactivities observed varies from individual to individual. In fact, a variety of citrullinated autoantigens have been detected in RA including vimentin (161), type II collagen (CII)
(162), α enolase (163), filaggrin (164) and fibrin (165), further suggesting that the initial ACPAs formed may not be pathogenic or able to react with epitopes in the joint
9 and that a maturation and spreading of the immune response is needed for disease initiation.
Summary of pathogenesis.
Although the details of RA pathogenesis are unclear, what is known reveals a complex disease process that involves the interaction of genes, environmental influences and various components of the immune system. In the pre-clinical phase of arthritis individuals with genetic risk alleles (principally the shared epitope) are exposed to an environmental trigger (e.g. injury, infection or smoking). This exposure begins an immune response that results in the production of autoantibodies (e.g. ACPA and RF) and is marked by an increase in certain cytokine and chemokine levels (166). As mentioned previously, the pre-clinical phase of disease may last years before clinical onset of RA. Little is known about this pre-clinical stage of arthritis due to the difficulty in identifying disease onset and obtaining biopsies from these ostensibly healthy individuals. Much of what we do know about the early stages of disease development comes from research performed in animal models and observations from human subjects with RA.
It appears that complement factors (reviewed in (167)), antibodies, B cells
(reviewed in (168)), T cells (169), mast cells (170-172), macrophages (reviewed in
(173)), and many other cell types can each play a role in disease initiation and progression. These various cell types and factors could be acting concomitantly, separately or in sequence to produce the inflammation necessary to generate citrullinated self proteins that are presented to T cells which activate B cells to induce the production of ACPAs. These ACPAs may originate at a site outside of the joint and are circulating 10 for up to years before another stimulus (e.g. injury, inflammation, epitope spreading) causes an influx of inflammatory cells into the joint. This can result in the citrullination of joint proteins (120, 174, 175) or the deposition of circulating citrullinated proteins in the joint (165) and the opportunity for circulating ACPAs to bind to citrullinated joint antigens (176). Thus begins the establishment of chronic joint inflammation characterized by immune complex formation and localized production of pro-inflammatory cytokines and ACPAs that recruit further rounds of inflammation, citrullination and joint destruction.
PAD inhibitors as a tool to investigate the pathogenesis of disease
Due to the close association between citrullination and ACPA+ RA, it is essential to understand if citrullination is pathogenic or simply a byproduct of the disease state in developing RA. Studies utilizing PAD knockout mice (145, 146), while informative, are limited by close proximity of the PAD genes making the knockout of multiple PADs simultaneously extremely difficult. The characterization of the effects of N-α-Benzoyl-
N5-(2-Chloro-1-Iminoethyl)-L-Ornithine Amide, a pan-PAD inhibitor better known as
Cl-amidine, and a PAD4 specific inhibitor named GSK283, on the development of inflammatory arthritis (IA) in mice are the focus of this dissertation.
The structure of Cl-amidine is similar to benzoyl arginine amide, a small molecule PAD substrate, except that it incorporates a reactive haloacetamidine warhead in place of the substrate guanidinium. Cl-amidine’s mechanism of action and function have been studied extensively (126, 177, 178). These studies have revealed that Cl- amidine has an IC50 of 5.9±0.3 µM and irreversibly inhibits PAD activity in a time and 11 concentration dependent manner. Structural studies suggest that Cl-amidine functions by forming a covalent bond with cysteine 645 in lieu of the amidino-Cys intermediate formed during normal catalysis (121, 124, 126, 179). This is only possible in the presence of Ca+ as the active site is inaccessible unless PADs are calcium bound. Due to the covalent nature of the bond, Cl-amidine inhibition appears to be irreversible. Cl-amidine can be taken up in cells (126), though the mechanism of transport into cells is unknown.
Prior to or concomitant with the work of this dissertation various studies were performed utilizing Cl-amidine to identify if PADs and citrullination play a role in the decondensation of chromatin for NET formation (180), spinal cord injury response (137), regulation of p53-target gene expression through histone modification and protein interactions (141, 142, 181), autophagy (unpublished results), colitis (138) and breast cancer (unpublished results). These studies establish the safety and utility of Cl-amidine in in vitro and in vivo systems.
GSK283 is a novel small molecule inhibitor identified from a screen at
GlaxoSmithKline (GSK), which selectively targets PAD4 over the other PAD family members. It can inhibit NH3 release from benzoyl-arginine ethyl ester (BAEE) substrate with a potency of ca 100 nM and inhibits histone H3 citrullination in cells in the low uM range. In contrast to Cl-amidine, inhibition is reversible and the inhibitor shows a binding for the low Calcium form of PAD4 and so may affect the equilibrium to stabilize the inactive form of the enzyme. The pharmacokinetics of GSK283 have been extensively studied and a dose of 30 mg/kg given subcutaneously was selected on the basis of modelled data which suggests that this dose level would achieve unbound Cmax levels above the biochemical potencies for almost 6 hours.
12
Animal models of disease
CIA is an MHC-resticted B and T cell dependent inducible mouse model of arthritis (reviewed in (182)) that is widely used because of the similarity of CIA immunologic and histopathologic lesions to human disease (183). Disease is induced by intradermal immunizations with bovine CII with CFA. In genetically susceptible mice, cross-reactivity to mouse CII develops and severe inflammation of the joints follows a booster injection of bovine CII. This is an excellent model for determining if citrullination is pathogenic in RA due to its widespread use, induction of citrullination in the joints (176) and serum (our unpublished observations), the ability of ACPAs to increase disease (153) and reliance on epitope spreading to self-antigens for disease initiation. CIA pathogenesis is dependent on B (184, 185) and T cells (186, 187) and chronic inflammation seems to be primarily mediated by anti-CII antibody deposition in the joints and ensuing complement activation and Fc receptor engagement.
Collagen antibody-induced arthritis (CAIA) is a passive transfer model of arthritis that bypasses the adaptive immune response in mice to focus on innate immune modulators. Arthritis in CAIA is MHC independent and induced by injecting a cocktail of antibodies to mouse CII into mice (188). These antibodies bind in the joint and activate complement, initiating an influx of neutrophils. This model is useful as a control for determining the role of citrullination in inflammatory arthritis in that it can determine the effects of PAD inhibition on the antibody-mediated effector phase of the immune response and non-citrulline mediated immunity.
13
Dissertation aims
The role of citrullination and ACPAs in the pathogenesis of RA remains to be clearly demonstrated and the limited capability of generating PAD knockout mice make them an incomplete method for addressing this question. The use of PAD inhibitors can help to elucidate the intriguing role of citrullination and ACPAs in the disease process of
RA. The aims of this dissertation are: 1) to characterize the effect of PAD inhibition on the generation of citrullinated epitopes, autoimmunity and epitope spreading in experimental inflammatory arthritis in mice 2) identify the potential mechanism(s) of disease amelioration by PAD inhibition and 3) as part of a CCTSI-funded clinical research project, to determine if autoantibodies to citrullinated proteins are present in the airways of subjects at risk for developing RA.
14
CHAPTER II
MATERIALS AND METHODS
Mice
DBA1/J and C57BL/6 mice were purchased from the Jackson laboratory (Bar
Harbour, ME). All studies were conducted in accordance with the GSK Policy on the
Care, Welfare and Treatment of Laboratory Animals and were reviewed by the
Institutional Animal Care and Use Committee either at GSK or by the ethical review process at the University of Colorado School of Medicine.
Collagen-induced arthritis
For induction of CIA, 6–8-wk-old DBA/1J mice were injected intradermally following our lab’s protocol on days 0 and 21 with 100 µl incomplete Freund’s adjuvant
(IFA) containing 200 mg bovine CII (Elastin Products) and 200 mg inactivated
Mycobacterium tuberculosis (H37Ra; Difco) to make complete Freund’s adjuvant (CFA)
(189). The emulsion for immunization is made by mixing CII and CFA between 2 glass syringes until the mixture no longer separates when dropped into water.
Disease incidence in mice was determined by an individual blinded to treatment.
Disease severity was scored 3 times a week by examining paws for signs of clinical inflammation. Clinical scores were assigned by giving each paw a number 0-3 based on visible swelling, with 0 representing no inflammation and 3 representing maximal inflammation, and a total possible score of 12 per mouse. Each point of the clinical score
15 is derived from inflammation in the digits, paw or wrist/ankle. Serum from each mouse was also collected at several time points for ELISA.
Unless noted otherwise, mice were euthanized at day 35 and both forepaws and one hind limb (including the paw, ankle, and knee) were surgically removed and fixed in
10% buffered formalin (Biochemical Sciences). The remaining hind limb was flash frozen in liquid nitrogen for total citrulline or mRNA studies. Formalin-fixed tissue samples were embedded in paraffin, sectioned, and stained with H&E for histopathology and with toluidine blue (T-blue) for specific evaluation of cartilage changes. Joint sections were scored by an individual blinded to treatment status for levels of synovial inflammation, pannus, cartilage damage, and bone damage, each category scored on a scale of 0–5. Quantitative immunohistological analysis for mouse C3 deposition in the synovium and cartilage was scored on a scale of 0–3.
Collagen antibody-induced arthritis
6-8-wk-old DBA/1J mice were immunized IP at day 0 with 1 mg of a cocktail of
4 monoclonal antibodies to type II collagen (Arthrogen) according to the manufacturer’s
(Chondrex) instructions. On day 3 of the protocol, mice were injected IP with 50 µg of lipopolysaccharide (LPS) to synchronize the onset of arthritis (190). Arthritis was scored daily after LPS injection until animal sacrifice at day 10. Limbs were taken for histological analysis of disease severity as described above. All clinical and histological scoring was performed on these joints as described above.
16
PAD inhibitor treatments
Pan-PAD inhibition was achieved by treating CIA or CAIA mice with 1 mg/kg,
10 mg/kg or 50 mg/kg Cl-amidine in PBS daily via IP injection. Vehicle-treated control
CIA or CAIA mice were treated with 200 µl PBS IP daily. All mice received treatments beginning on day 0 of CIA or day -7 of CAIA and continued to receive daily treatments until the endpoint of each experiment.
PAD4 inhibition was achieved by treating CIA mice with 10 mg/kg, 30 mg/kg or
30 mg/kg bis in die (bid) GSK283 in 0.9% saline (Hospira) daily via subcutaneous injection. Vehicle-treated control CIA mice were treated with 0.9% saline subcutaneously
(SQ) daily. All mice received treatments beginning on day 0 of CIA and continued to receive daily treatments until the endpoint of the experiment.
Measurement of total citrulline
For total citrulline determination in Cl-amidine or control treated mice, animals were administered one intra-dermal injection of CII in CFA (as described above) and were then treated daily for 21 days with PBS or 1 mg/kg, 10 mg/kg or 50 mg/kg Cl- amidine intraperitoneally (IP). On day 21, the entire knee joint was surgically removed from control and Cl-amidine–treated DBA/1J CIA mice and flash-frozen in liquid nitrogen. Synovial tissue was enriched from the joints by further dissection. Tissue homogenates were prepared according to the method of Moscarello and colleagues (114,
191). Briefly, individual tissue samples were resuspended in 50 mM HEPES (pH 7.6), 1.0 mM EDTA, 0.5 mM DTT, and 0.43 mM PMSF at a final concentration of 20 mg tissue/0.1 ml buffer. Cell extracts were prepared with a Dounce homogenizer and 17 subsequently clarified by centrifugation. Total citrulline content was analyzed by adding
60 µl synovial lysate or 20 µl serum lysate to a reaction buffer containing 50 mM NaCl,
10 mM CaCl2, 2 mM DTT, and 100 mM Tris (pH 7.6). A color development solution that detects citrullinated proteins (200 ml) was added (COLDER) (192). The sample was vortexed and incubated at 95°C for 30 min. The absorbance at 540 nm was measured and compared to a standard curve of known citrulline concentrations. Measurements were made in duplicate, and the data were normalized using protein concentrations for each sample determined using the BCA total protein assay (Pierce).
Total citrulline in knee joints from GSK283 treated mice was assayed by removing knees from CIA mice at day 35 that had received 0.9% saline, 10 mg/kg, 30 mg/kg or 30 mg/kg bid GSK283 daily SQ from day 0 through day 35 of CIA. Knees were flash frozen in liquid nitrogen, processed and assayed for total protein and total citrulline content as described above.
Synovial proteome autoantibody microarrays
For Cl-amidine treated mice, synovial proteome arrays containing 191 proteins and peptides representing candidate autoantigens in RA were used to perform multiplex characterization of autoantibody responses in sera derived from Cl-amidine–treated mice and control mice (193, 194). The synovial proteome arrays were previously validated with a panel of mAbs and reference sera specific for many of the spotted proteins (194).
Significance analysis of microarrays (SAM; version 3.08) was used to identify autoantibody reactivities that exhibited significant differences between Cl-amidine– and
PBS-treated CIA mice. The mice and their Ag array reactivities were arranged using 18 hierarchical cluster analysis (Cluster 3.0 software) and displayed as a heatmap (Java
TreeView software version 1.1.3 created by Alok) (195, 196). An accession number
(GSE23731) for complete microarray data has been assigned and approved in the Gene
Expression Omnibus database
(http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE23731).
For GSK283 treated mice synovial proteome arrays were performed as above but an accession number has not yet been assigned in any database.
Autoantigen reactivity by ELISA
Murine IgG reactive with putative autoantigens identified by synovial proteome autoantibody microarrays [HCgp39(130-145), HCgp39(160-175), DEAE pool 1 and
Biglycan(227-246)] were measured in duplicate by coating 96-well plates with 10µg/ml antigen in PBS overnight at 4°C. After washing with PBS + 0.05% tween 20 (PBST), plates were blocked with 1% BSA in PBS for 4 hours at 4°C. Following blocking,
50µl/well of a 1:150 dilution of D0 or D35 serum in PBS was added to plates and incubated overnight at 4°C. Plates were washed in PBST and antibody binding was detected using HRP conjugated anti-mouse IgG visualized with TMB substrate.
Flow cytometric analysis of immune cell populations
Spleens obtained from mice with CIA that had been treated with either Cl- amidine or GSK283 were surgically removed at day 35 and placed in complete media
(RPMI 1640, 10% heat-inactivated FBS, penicillin-streptomyocin solution [Cellgro]) on 19 ice. Splenocytes were isolated through a cell strainer (BD Falcon 352350) in a 60mm culture dish containing 4 ml complete medium. Using the flat end of a syringe plunger spleens were pressed through cell strainers to yield a single-cell suspension. Cell strainers were then washed with 6ml complete medium and cells were centrifuged at 1000 rpm at
4°C for 3 minutes. Following centrifugation red blood cells were lysed with 1 ml red cell lysing buffer (Sigma) and incubated for 5 minutes at 37°C. Following 2 washes in complete medium cells were counted on a hemocytometer with a viability dye (trypan blue 1:20). 1x106 cells per condition were then stained for surface or intracellular markers as indicated using BD Pharmingen antibodies and staining protocols.
Measurement of antibodies
Antibody titers to bovine and mouse CII were measured using serum collected from all mice on days 0, 21 and 35 (diluted 1:2000 in PBS). IgG2a and IgG1 antibody titers to murine and bovine CII were measured in triplicate by specific ELISA by coating
96-well plates with 5 mg/ml murine or bovine CII (Chondrex). A standard pool of anti-
CII Abs was obtained by combining sera from several mice with severe CIA; the levels of IgG1 and IgG2a Abs in this pool were considered 100 U/ml. A standard curve was obtained using serum dilutions, and the Michaelis–Menton equation was used to convert
OD values into units.
20
Measurement of antibody-producing cells
The number of cells producing antibodies specific for bovine or mouse CII was determined by specific ELISPOT (MabTech). Briefly, ELISPOT plates with 0.45 µm hydrophobic polyvinylfluoride (PVDF) membranes (Millipore) were pre-wet with 50 µl
70% ethanol per well for 2 minutes. After 5 washes with sterile water pre-wet wells were then coated with 100 µl of 15 µg/ml anti-IgG capture antibody in PBS overnight at 4°C.
After coating, plates were washed 5 times with sterile PBS and blocked with 200 µl complete medium for 30 minutes at room temperature. Following blocking 150,000 cells isolated from splenocytes or lymph nodes (as detailed above) were added to each well and incubated at 37°C with 5% carbon dioxide for 18 hours. To detect spots formed by antigen-specific or total IgG producing cells, plates were washed 5 times with PBS and
100 µl 0.1 µg/ml biotinylated bovine CII or mouse CII (for antigen-specific spot detection) or 100 µl 1 µg/ml anti-IgG biotin (for total IgG spot detection) was added per well and incubated for 2 hours at room temperature. Following this incubation plates were washed 5 times with PBS and 100 µl of streptavidin-ALP (1:1000) in PBS-0.5%
FBS per well was added with plates incubating for 1 hour at room temperature. To visualize spots, plates were washed 5 times with PBS and 100 µl of BCIP/NBT plus substrate solution was added to each well and allowed to develop for 30 minutes at room temperature. This reaction was stopped by rinsing plates with tap water. Plates were allowed to dry overnight and spots were counted using an ImmunoSpot ELISPOT reader.
21 mRNA isolation and determination of transcript expression
Joint mRNA (the hind limb from the knee down) was isolated in 1 ml TriReagent by homogenizing with a Polytron PT 10-35 homogenizer. Joint homogenates were then centrifuged at 2000 rpm for 3 minutes at 4°C. The supernatant from this centrifugation was transferred to a screw cap microtube containing 200 µl chloroform and vortexed for
15 seconds. This mixture was centrifuged at 14,000 rpm for 15 minutes at 4°C. Following centrifugation the aqueous (top) phase was transferred to an RNase-free microtube and mixed with 500 µl RLT buffer (RNeasy, Qiagen). RNA isolation was completed using the RNeasy protocol beginning from the addition of RLT buffer to tissue samples. Lymph node and spleen mRNA was isolated from pelleted single-cell suspensions (1 x 107 cells/pellet isolated as detailed above) using the RNeasy protocol.
The mRNA for p53 (ABI- Assay ID: Mm01731287_m1), p21 (ABI-Assay ID:
Mm 00432448_m1), c3 [FWD (5’-3’):-GAT TTT GAT GAG TAC ACC ATG ACC A;
Rev (5’-3’): GCT GCC CTG CCT GCA C; Probe (5-3’): 6FAM-CCA GCA GGT CAT
CAA GTC AGG CTC AGA], fB (FWD (5’-3’):- GGC AAC AGC TGG TAC CCT CTT;
Rev (5’-3’) : CCT TTA GCC AGG GCA GCA C; Probe (5-3’): 6FAM-CGG GAC TTC
CAC ATC AAC CTC TTC CA) and fH [FWD (5’-3’):- TCA CCA CCT TCT GGG TAT
TCC T; Rev (5’-3’): TCC TGA CGC ATG GGA CTT C; Probe (5-3’): 6FAM- CTT
CGT TGC ACA GCA CAA GGG TGG] was measured by real-time quantitative RT-
PCR using an ABI Prism 7900 Sequence detector. PCR primers and probes were purchased from Applied Biosystems. The TaqMan probes were 5’labeled with 6- carboxyfluorescein (FAM). 1ug total RNA was used to synthesize cDNA using the High
Capacity c-DNA Reverse Transcription kit (ABI-P/N 4368814). cDNA was diluted 1:2
22 before PCR amplification. Real time PCR reactions were carried out in MicroAmp optical tubes (PE ABI) in a 25 ul mix containing 8% glycerol, 1X TaqMan buffer A (500 mM KCl, 100 mM Tris-HCl, 0.1 M EDTA, 600 nM passive reference dye ROX, pH 8.3 at room temperature), 300 uM each of dATP, dGTP, dCTP and 600uM dUTP, 5.5 mM
MgCl2, 1X primer-probe mix, 1.25 U AmpliTaq Gold DNA and 5 ul template cDNA.
Thermal cycling conditions were as follows: Initiation was performed at 50°C for 2 min followed by activation of TaqGold at 95°C for 10 min. Subsequently 40 cycles of amplification were performed at 95°C for 15 secs and 60°C for 1 min. Experiments were performed with duplicates for each data point. Each PCR run included the standard curve
(10 fold serially diluted pooled cDNA from control and experimental samples), test samples, no-template and NORT controls. The standard curve is then used to calculate the relative amounts of targets in test samples. Quantities of targets in test samples were normalized to the corresponding 18s rRNA (PE ABI, P/N 4308310).
Evaluation of antigen presentation
Peritoneal exudate cells (PECs) were obtained by intraperitoneal injection with 1 ml thioglycollate or 100 mg concanavalin A. After 4 d, PECs were harvested and used as
APCs in antigen presentation experiments. PECs were incubated overnight in the presence or absence of Cl-amidine with 10 µM or 1 µM hen egg lysozyme (HEL) protein or 10 µM HEL peptide. T cell hybridomas reactive to citrullinated HEL or both citrullinated and uncitrullinated HEL were then added and hybridoma activation was measured by IL-2 secretion as assayed by proliferation of the IL-2–dependent line CTLL.
23
Determination of macrophage polarization
Bone marrow (BM) macrophages were derived from C57BL/6 mice by removing one femur and bone marrow cells harvested by flushing 1 ml of sterile PBS through the bone marrow cavities with a 25 5/8 gauge syringe. Cells were counted on a hemacytometer and 20,000 cells were cytospun onto a slide, with differential cell counts performed. On average, 2 to 4% of the extruded bone marrow cells are monocytes.
BM derived macrophages from C57BL/6 mice were cultured and stimulated for
24 hr with X-Vivo media alone, IL-4 + IL-13 (20 ng/ml), IFN-γ (50U), LPS (100 ng/mL) or LPS + IFN-γ to determine macrophage programming status. In additional wells, 200
µM Cl-amidine in PBS was added to wells containing IL-4 (20 ng/ml), IL-13 (20 ng/ml),
IL-4 + IL-13 (20 ng/ml), IFN-γ (50U), LPS (100 ng/mL) or LPS + IFN-γ.
Media from treated BM macrophages was analyzed for NaNO2 levels by incubation for 10 min in Greiss Reagent. Samples were read at 550 nm OD and compared against a NaNO2 standard curve to determine concentration.
Protein lysates from treated BM macrophages was analyzed for arginase activity in a Mn-Tris solution (10 mM, 50 mM, pH 7.5) for 10 min at 55°C followed by the addition of 0.5 M Arginine (pH 9.7) and incubated for 1 hr at 37°C. A sulfuric acid:phosphoric acid (1 H2SO4:3 H3PO4:7 H2O) solution was added to stop the reaction. Nine percent isonitrosopropiophenone (ISPF) in EtOH was added and samples were incubated for 45 min at 100°C. Samples were read at 540 nm OD and compared against a urea standard curve to determine concentration as previously described (197).
24
Human subjects recruitment and consent
Subjects were recruited from the Studies of the Etiology of RA (SERA) project, a prospective study established to investigate the natural history of RA(198). Enrollment was between February 2011 and February 2012, and included members from 4 subgroups followed in the SERA project: patients with early (<12 months since diagnosis) seropositive RA (Early RA) fulfilling 1987 ACR criteria(199), subjects without inflammatory arthritis (IA), by history and by clinical examination by a rheumatologist or trained study nurse, who were either first-degree relatives (FDRs) of probands with RA, or subjects identified through community health-fair screening (Health-Fair), and finally healthy seronegative controls recruited through local advertising. Institutional Review
Boards at participating institutions approved all study procedures.
Sputa collection and processing
Sputa was obtained using an induced sputa protocol of inhalation of nebulized 5% saline over 15 minutes, with expectorated sputum collected at defined intervals and when subjects sensed the need for coughing/sputum release. Once collected, sputa samples were diluted with 3 mLs of PBS per 1 gram of sputa and mechanically disrupted through an 18 g needle (4 cycles per gram or a minimum of 12 cycles). Processed sputa were then centrifuged at 1200 rpm for 10 minutes. The supernatant was further centrifuged at 3500 rpm for 20 minutes. The final supernatant was used for all sputa analyses. To ensure that adequate bronchial samples were analyzed, only sputa samples with a squamous epithelial cell count of <10 cells per high-powered field were utilized for these experiments (excluding 8% of subjects) (200). 25
Human sample autoantibody testing
Sera samples from all subjects were tested for the following autoantibodies: RF isotypes IgM, IgA and IgG by ELISA (QUANTA Lite™ kits, INOVA Diagnostics, Inc.), anti-CCP2 (IgG ELISA; Diastat, Axis-Shield Diagnostics, Ltd.), and anti-CCP3.1
(IgA/IgG ELISA; INOVA Diagnostics, Inc.). Serum positivity for each RF isotype was established based on levels positive in <5% of 491 blood donor controls; standard kit cut- offs were used for serum anti-CCP positivity (anti-CCP2 >5 units; anti-CCP3.1 ≥20 units).
All sputa samples were collected simultaneously with sera, and tested using the autoantibody assays described above, substituting 100 uL of processed sputa for diluted sera. A ‘final’ value for each autoantibody was established using an average concentration from two ELISA wells (there was no significant difference in mean levels between wells). In addition, the total protein concentration of sera and sputa was determined using the BCA protein assay (Pierce).
Statistical analysis
ANOVA, with tests for multiple comparisons, was used to examine the data for
CIA and CAIA clinical disease activity scores, histology and anti-CII Ab levels and synovial and serum citrulline content. Flow cytometry, grouped synovial citrulline, antigen presentation, tissue mRNA level analysis and autoantigen ELISAs data were analyzed by Student’s t test. Pearson’s correlation coefficient was calculated for comparisons between histology and clinical disease scoring. The Shapiro-Wilks' "W" test was used to test all histological and disease activity data for normal distribution 26 before applying subsequent parametric tests. Data were expressed as the mean ± SEM with p < 0.05 considered significant.
For human sample studies autoantibody levels were compared across groups using Kruskal-Wallis testing. Autoantibody positivity in sputa and sera was evaluated using matched-pair analyses.
27
CHAPTER III
CL-AMIDINE TREATMENT AMELIORATES CIA1
Background and rationale
The pathophysiologic significance of protein citrullination in RA is unclear, as it is unknown whether citrullination reflects ongoing inflammation or plays a causal role in the pathogenesis of disease. The presence of ACPA in the sera of individuals who ultimately develop seropositive RA for an average of 4-5 years prior to the onset of clinically apparent disease (5-7) and their association with disease severity (150, 151), suggest that the development of these autoantibodies may be an early event in the onset and progression of RA. Other evidence from animal and in vitro studies strengthens the likelihood of a pathogenic role for ACPAs in arthritis: 1) mice tolerized to citrullinated peptides develop less severe arthritis than controls and mice treated with a monoclonal antibody recognizing citrullinated peptides exhibit increased disease severity (153); 2) transgenic mice containing the human HLA DRB1-0401 allele develop arthritis when immunized with citrullinated human fibrinogen, but not native fibrinogen or mouse fibrinogen (154); 3) antibodies to citrullinated CII have demonstrated pathogenic capabilities in vivo in mice (155); 4) Immune complexes to citrullinated fibrinogen can induce TNF α production in macrophages (156); 5) ACPAs from RA patients have been shown to activate complement (157).
1 This chapter is adapted from N-{alpha}-Benzoyl-N5-(2-Chloro-1-Iminoethyl)-L- Ornithine Amide, a Protein Arginine Deiminase Inhibitor, Reduces the Severity of Murine Collagen-Induced Arthritis. J Immunol 186:4396-4404. 28
By preventing citrullination, the role of immunity to citrullinated proteins in RA can be evaluated. The only enzymes known to catalyze citrullination (or deimination) are
PADs, a small family of five calcium-dependent enzymes (PADs 1, 2, 3, 4, and 6; there is no PAD5). Inhibiting PAD activity can be achieved by three primary methods: first through the genetic knockout of PAD genes, second through pharmacological inhibition of PAD enzymes or third by utilizing silencing RNA (siRNA). The utility of the genetic knockout approach is limited by the clustering of the PAD genes on the same locus. This makes the creation of knockouts for multiple PADs extremely difficult and time consuming. siRNA inhibition of PADs is an excellent tool for in vitro studies, but presents many challenges for in vivo approaches. In contrast the pharmacological inhibition of PADs presents a methodology that has the potential to target one or multiple
PADs at any time during the development of disease and, while containing the potential for off-target effects and toxicity, is an excellent tool for the initial determination of the role of PADs in RA.
Through our collaboration with Paul Thompson (The Scripps Research Institute,
Florida), we gained access to a pan-PAD inhibitor denoted Cl-amidine. Cl-amidine is a mechanism-based PAD inactivator that irreversibly inhibits all of the known active PAD isozymes, i.e. PADs 1-4, with low µM potency. In addition to Cl-amidine’s ability to inhibit PADs in vitro, its has been evaluated in a cell-based assay of PAD4 activity demonstrating that Cl-amidine is cell permeable, can transit membranes and inhibit
PAD4 activity in the nucleus (126).
We chose to evaluate Cl-amidine’s potential to alter the development of arthritis in the CIA model of RA because it is a widely used model that recapitulates various
29 aspects of human disease and importantly these mice are known to develop serum antibody reactivity to citrullinated epitopes as determined by synovial proteome array analysis (153, 201) and have shown alteration of disease course in previous experiments targeting citrulline reactivity.
The hypothesis tested in this chapter is that the development of citrullinated epitopes and ACPA is pathogenic in RA. To test this hypothesis we examined whether the pan-PAD inhibitor Cl-amidine could ameliorate the signs and symptoms of murine
CIA by blocking the development of citrullinated epitopes and ACPA generation and/or affecting other PAD-dependent cellular processes.
Results
Cl-amidine treatment reduces clinical and histological disease severity in CIA.
To test if inhibition of PAD activity would result in decreased disease severity as a result of reducing the evolution of ACPA epitopes and autoantibody production, mice were immunized on days 0 and 21 with bovine CII and clinical disease activity was monitored during days 21-35. Although all of the mice developed disease, mice receiving daily IP injections of Cl-amidine in PBS, beginning on day 0 and continuing until the endpoint of the experiment, exhibited reduced clinical disease activity ~50% on days 24 through 36 (p < 0.05 for all 3 doses of Cl-amidine at day 35) (Figure 1). These results indicate that Cl-amidine treatment reduces clinical disease activity in CIA.
To examine the effects of treatment with Cl-amidine on joint destruction, histological analyses of joints were performed at the endpoint of the experiment, day 35.
30
Histopathology scores for synovial inflammation, pannus invasion, cartilage damage, and bone damage were significantly lower in the 10 mg/kg and 50 mg/kg per day treatment groups compared to the PBS vehicle group (p < 0.03) (Figure 1B). Cl-amidine treatment at 50 mg/kg also reduced C3 deposition in the synovium and at 10 mg/kg in the cartilage
(Figure 1C). The clinical disease activity and histology scores showed a significant correlation with each other (R squared value of 0.8, p < 0.001) (Figure 1D). No differences in neutrophil (Figure 1E) or macrophage (Figure 1F) infiltration of CIA joints was observed in mice treated with Cl-amidine, suggesting that recruitment of these cells was not blocked by Cl-amidine. These results suggest that Cl-amidine treatment prevents joint destruction and reduces C3 deposition in CIA.
Cl-amidine treatment reduces serum and synovial citrulline content in CIA.
To help confirm that Cl-amidine derives its effects from its ability to inhibit PAD activity, we measured changes in synovial citrulline content, i.e. the product of the PAD reaction, as a proxy of PAD inhibition. As expected, synovial citrulline content trended downward in response to increasing Cl-amidine dose (Figure 2A, left panel). We then compared the synovial citrulline content in all Cl-amidine treated mice (all doses combined) with all control mice (mice receiving no treatment or PBS). Mice treated daily for 21 days with Cl-amidine, beginning with one CII immunization, exhibited significantly lower synovial citrulline content compared to PBS treated or untreated control mice with CIA (p < 0.05), and showed similar synovial citrulline content in comparison to unimmunized mice (Figure 2A, right panel). These results indicated that treatment with daily injections of Cl-amidine for 21 days decreased citrulline content in the joints. 31
Figure 1 i' B w A (/) 4· +I 12 1: m 11 Qj 10 ...Q) 3· g 0 9 (.) ** ...Qj 8 (/) 0 7 Q) 2 (.) Illm ~l (/) 6 Q) 5 Ill ro i ·;:;~ 4 i5 1 :;; (.) 3 <( 2 0 Qj 1 Ill 0 m ;.,0<:- .1'.... Qj ~~. {f Ill 21 24 27 30 33 36 ~ ~· ·~'If c Days After First Clllnjection ~~ 32 Figure 1. Cl-amidine treatment reduces clinical disease activity and joint destruction in CIA. Arthritis was induced by two injections, at day 0 and 21, of bovine CII in CFA. Mice were treated as described in Materials and Methods and assessed three times weekly for clinical disease activity. A) All treatment groups showed significantly reduced clinical disease activity scores compared to the PBS vehicle control beginning on day 27 and continuing through day 35. For 10 mg group * p < 0.05 at day 24 and p < 0.03 at day 27. #p < 0.01 for 50 mg group. ***p < 0.03 for 10 and 50 mg groups and p < 0.05 for 1 mg group. (n=8 except for 1 mg group where n=7). B) Histologic sections were scored for changes in synovial inflammation, pannus, cartilage damage and bone damage, all on a scale of 0–5. C) Sections were scored for C3 deposition in both the synovium and cartilage. D) Plot depicting the correlation between clinical disease activity and histological disease score. Data are expressed as the mean ± SEM based upon a set of five joints per animal. Treatment with Cl-amidine resulted in decreased histopathology scores in all parameters at the 10 mg/kg and 50 mg/kg per day treatment groups compared to the PBS vehicle group. (n=8 except for 1 mg group where n=7); *p ≤ 0.03. **p < 0.02, ***p < 0.01. 33 Additionally, we measured the serum citrulline content in the mice with CIA (Figure 1) at day 35 to verify the activity of Cl-amidine. As expected, serum citrulline content trended downward in response to increasing Cl-amidine dose (Figure 2B, left panel). We then compared serum citrulline content in all Cl-amidine treated mice (all doses combined) with all control mice (mice receiving no treatment or PBS). Mice treated daily for 35 days with Cl-amidine, beginning with their first CII immunization, exhibited significantly lower serum citrulline content compared to PBS treated control mice with CIA (Figure 2B, right panel) (p = 0.0018). These data indicated that Cl- amidine treatment reduces PAD-mediated citrullination. Cl-amidine treatment does not alter immune cell populations. To examine whether Cl-amidine treatment has effects on the number and distribution of immune cell populations, flow cytometry was performed on splenocytes from mice with CIA treated with Cl-amidine daily for 21 days beginning with CII immunization. No differences were seen in the frequency (Table I) or absolute number (Table II) of T, B, NK cells or monocytes in comparison to cells from control mice treated with PBS. Therefore, we concluded that the beneficial effects of Cl-amidine on CIA are not accompanied by a depletion of major subsets of immune cells. 34 Figure 2. Cl-amidine treatment reduces serum and synovial citrulline content in vivo. A) Synovial citrulline content. Arthritis was induced in control and Cl-amidine treated mice by a single injection of bovine CII in CFA at day 0. Mice were treated daily, beginning on day 0, for 21 days with Cl-amidine (1, 10 or 50 mg/kg) in PBS, PBS, or were untreated. Results are displayed by dose (left panel) and grouped (right panel) as unimmunized (WT), control (PBS and untreated) and Cl-amidine (1, 10 and 50 mg/kg) treated mice. B) Serum citrulline content was measured in mice at day 35. Results are displayed by dose (left panel) and grouped (right panel) as control (untreated and PBS) and Cl-amidine (1, 10, and 50 mg/kg) treated mice. 35 Table I. Frequency of immune cells in Cl-amidine-treated and control mice with CIA. CD4 CD8 CD19+B220+ DX5 CD11b CD11c PBS 10.1 ± 0.9 4.0 ± 0.4 59.8 ± 1.6 2.7 ± 0.3 1.6 ± 0.2 1.6 ± 0.2 CL50 10.4 ± 0.8 3.8 ± 0.3 56.3 ± 1.7 2.8 ± 0.1 1.6 ± 0.2 1.9 ± 0.1 Splenocytes from CIA mice treated with Cl-amidine or PBS for 21 days were stained with the listed cell surface markers, and the frequency of each population was determined. Data are mean ± SEM. CL50 = 50 mg/kg/day Cl-amidine. Table II. Absolute number (x106) of immune cells in Cl-amidine-treated and control mice with CIA. Spl CD4 CD8 CD19+B220+ DX5 CD11b CD11c PBS 74 ± 14 6.8 ± 0.78 2.7 ± 0.34 44 ± 7.9 2.0 ± 0.42 1.2 ± 0.29 1.2 ± 0.29 CL50 96 ± 24 9.0 ± 1.7 3.3 ± 0.67 52 ± 15 2.6 ± 0.63 1.5 ± 0.43 1.8 ± 0.46 Splenocytes (Spl) from CIA mice treated with Cl-amidine or PBS for 21 days were stained with the listed cell surface markers, and the absolute number of cells in each population was determined. Data are mean ± SEM. CL50 = 50 mg/kg/day Cl-amidine. Cl-amidine treatment does not ameliorate CAIA. To examine the possibility that the beneficial effects of Cl-amidine in CIA are mediated at the effector phase, we studied the CAIA mouse model of inflammatory arthritis. Unlike CIA, the CAIA mouse model bypasses the initial antigen recognition phases of the immune response through IP administration of a cocktail of mAb to CII (Arthrogen). This is followed by an IP injection of LPS on day 3, leading to the rapid and consistent development of inflammatory arthritis. 36 DBA/1J mice were administered 1, 10 or 50 mg/kg Cl-amidine in PBS or PBS daily IP for 7 days prior to receiving 4 mg/mouse of Arthrogen IP to induce disease. Mice continued receiving daily treatment with PBS or Cl-amidine until sacrifice 10 days after disease induction. Cl-amidine pre-treatment in mice with CAIA had no effect on disease activity (Figure 3A) or histologically evaluated joint damage (Figure 3B). Clinical disease activity and histology scores showed a significant correlation with each other (R squared value of 0.68, p=0.0063, not shown). These results in CAIA contrast with the decrease in both disease activity and joint destruction in CIA induced by treatment with Cl-amidine (Figure 1). We concluded that Cl-amidine treatment does not ameliorate inflammation and joint tissue destruction in CAIA by altering the effector phase of disease. Treatment with Cl-amidine decreases IgG2a and IgG1 anti-mouse CII antibodies. Both xeno (bovine) and self (mouse) anti-CII Ab evolve over time in CIA, reflecting an expanding loss in tolerance to self-epitopes. To gain a better understanding of how Cl-amidine treatment affects self-tolerance, sera from each treatment group were incubated on 96-well plates coated with bovine or mouse CII. Since anti-mouse CII IgG2a autoantibodies are known to be pathogenic in CIA (202), sera were analyzed by ELISA specific for mouse IgG1 and IgG2a Ab to bovine and murine CII . 37 Figure 3. Cl-amidine treatment does not affect passively transferred arthritis. Mice were given daily IP injections of PBS, 1, 10, or 50 mg/kg Cl-amidine for 7 days prior to receiving 4 mg Arthrogen IP 3 days later mice were injected IP with 50 µg LPS. Mice continued to receive PBS or Cl-amidine daily on days 0-10. Mice were scored for disease severity daily. A) Disease severity scores. B) Histologic sections were scored for changes in synovial inflammation, pannus, cartilage damage and bone damage, all on a scale of 0–5. (PBS n=8, Cl 1&10 n=4, Cl50 n=9). 38 Cl-amidine treatment did not alter the anti-bovine CII IgG1 or IgG2a antibody levels at any dose (Figure 4A). However, there were significantly lower (p ≤ 0.05) IgG1 and IgG2a anti-mouse CII antibody levels in the sera from mice treated with 50 and 10 mg/kg/day Cl-amidine (respectively) compared to treatment with PBS alone at day 35 (Figure 4B). Notably, Cl-amidine treatment at 10 mg/kg reduced the titer of IgG2a autoantibodies, which are known to be pathogenic in CIA (Figure 4C) (202, 203). These results demonstrated that Cl-amidine treatment decreases the humoral response to native but not foreign CII. Treatment with Cl-amidine decreases antibody responses to both native and citrullinated synovial epitopes. To examine the effects of Cl-amidine treatment on Ab responses, a synovial antigen array analysis was performed using day 35 sera derived from mice with CIA treated with 50 mg/kg/day Cl-amidine or PBS. These synovial arrays contain 191 peptides and proteins representing candidate antigens in RA, including overlapping peptides spanning the full length of the and β chains of fibrinogen in citrullinated and unmodified forms, as well as peptides from other candidate citrullinated autoantigens, notably vimentin (154). Significance Analysis of Microarrays (SAM) identified 8 autoantibody reactivities that were statistically decreased in CIA mice treated with Cl- amidine compared to PBS treatment (Figure 5). 39 Figure 4 Bovine Cll .-. lgG1 'E...... lgG2a :;) ~ Ill Ill .... 0.25 .... 0.9 Cll Cll 0.8 i=- i=- ca 0.20 ... 0.7 N 0.6 ".El ".El 0.1 5 1: 0.5 1: Cll Cll tn 0.4 tn .!!! .!!! 0.10 0 0.3 (.) 0 I 0.2 <.( 0.05 :.;:::: :.;:::: 1: 0.1 1: c:t c:t O.O...L..-...c:;;...----r----...--- Cll 0.00 ~ 0 21 :.;:::: 0 21 35 :.;::::> 35 ca Qj Day .!!! Day a: a:Cll Mouse Cll .-...... 'E :;) ~ Ill.... 0.25 Cll -i= -+-No Treatment -e- PBS CE' :5 ~ Ill.... Cll i=- ca 0.20 N ".El 0.15 1: Cll ~0 .1 0 0 <.( 0.05 :.;:::: 1: Cll O.O..LL---L....L-....&.....&.-...L....L.-....L... ~ 0.00 ..LL-.....L.....L-...1....1.-...L....L.-..L... > :.;:::: > ca :.;:::: Qj ca Qj a: a: c:::::J No Treatment c:::::J PBS c:::::J 1mg Cl-amidine c::::J 1Omg Cl-amidine - 50mg Cl-am idine 40 Figure 4. Cl-amidine treatment reduces IgG1 and IgG2a autoantibody response to mouse but not bovine CII. Sera from individual mice were incubated on a 96-well plate coated with mouse (B) or bovine (A) CII. The presence of anti-mouse IgG1 or IgG2a antibody was detected by ELISA. The 50 mg/kg/day group showed significantly lower IgG1 anti-mouse CII antibody titers, while the 10 mg/kg/day group had significantly lower IgG2a anti-mouse CII antibody titers when compared to the PBS vehicle control group at day 35. C) Histograms of IgG1 and IgG2a antibody titers at day 35. *p ≤ 0.05. (n=8 except for 1 mg group where n=7). 41 Figure 5. Cl-amidine treatment reduces autoantibody responses in CIA. Sera were collected from Cl-amidine and control treated mice, and autoantibodies in these samples were profiled using (A) Synovial Antigen Arrays and an anti-IgG/M goat- anti-mouse secondary antibody or (B) specific ELISA. (A) Significance Analysis of Microarrays (SAM) was used to analyze the antigen array datasets, and identified 8 antigens with a significant difference in autoantibody reactivity (false discovery rate (q) < 5%). Hierarchical clustering was then performed to elucidate the relationships between the autoantibody profiles. The results are displayed as a heatmap (blue being negative, yellow intermediate, red strong positive). The Cl-amidine treated mice clustered together (on the left side of the heatmap) and exhibited lower autoantibody titers against all of the 8 antigens identified by SAM, including autoantibody reactivity to native epitopes derived from cartilage and gp39 (native antigens in black font) as well as citrulline- modified filaggrin peptides (red font). (B) Serum IgG reactivity to the indicated putative autoantigens from the array was performed by specific ELISA using serum from PBS and Cl-amidine (10 & 50 mg/kg/day treated) mice. D0 serum reactivity was treated as baseline and subtracted from D35 serum reactivity. Cl-amidine treated mice showed a statistically significant decrease to Biglycan (227-246) (p=0.019). 42 Over 150 native and citrullinated epitopes on the arrays did not exhibit significant differences in reactivity. Sera from mice treated with 50 mg/kg/day Cl-amidine demonstrated decreased IgG reactivity to native epitopes including human cartilage gp39, biglycan, and hnRNP-A, as well as decreased autoantibody reactivity to citrullinated filaggrin peptides cfc2 and cfc4, as compared to sera from controls treated with PBS. There were no differences between treatment with Cl-amidine or PBS alone in the levels of autoantibodies to 38 other citrullinated proteins in the array. Reactivity to a subset of antigens in Figure 5 was evaluated by ELISA (Figure 5B). Mice treated with 50 mg/kg/day Cl-amidine showed a trend of lower reactivity to all antigens tested [HCgp39 (130-145), HCgp39 (160-175) and DEAE pool 1] and a statistically significant decrease to biglycan (227-246) compared to controls. These data demonstrate that Cl-amidine treatment decreases the development of autoantibodies in CIA and also suggest that Cl- amidine treatment modestly impairs epitope spreading. Both xeno (bovine) and self (mouse) anti-CII Ab evolved over time in CIA, reflecting an expanding loss in tolerance to self-epitopes. Summary and conclusions The hypothesis examined in these studies was that the development of citrullinated epitopes and ACPA is pathogenic in RA. To test this hypothesis we examined whether the pan-PAD inhibitor Cl-amidine ameliorated the signs and symptoms of murine CIA by blocking the development of citrullinated epitopes and ACPA generation and/or affecting other PAD-dependent cellular processes. 43 We found that mice treated with Cl-amidine exhibit a decrease of ~50% in clinical disease activity as well as significant reductions in joint inflammation and destruction in CIA. In contrast, Cl-amidine treatment had no effect on CAIA, suggesting that Cl-amidine does not alter the antibody-mediated effector phase of disease. Interestingly the main source of PAD2 and PAD4 in the joint is thought to be inflammatory cells, primarily macrophages and neutrophils. It is therefore notable that Cl-amidine did not ameliorate CAIA, a disease model dependent on the function of these cells. These results indicate that pan-PAD inhibition is able to blunt the severity of arthrtitis when given during the development of disease, but that this therapy alone is unable to affect ongoing inflammation under the conditions tested. Future studies focusing on different time points of therapeutic intervention with PAD inhibitors could shed light onto the role of citrullination in the development of disease. We found that Cl-amidine treatment returned CIA synovial citrulline content to WT levels. This suggests that Cl-amidine treatment, at the doses utilized in this study, does not inhibit all citrullination. It is not likely that insufficient Cl-amidine was bioavailable because the 10 mg/kg/day and 50 mg/kg/day doses both showed similar levels of beneficial clinical effect. However, since Cl-amidine preferentially binds activated PAD enzymes (126), it is possible that Cl-amidine is only able to inhibit “active” PADs, i.e. PADs that are present in a sufficiently calcium-rich environment (e.g., extracellularly). Since all PADs are normally intracellular and increased levels of citrullination are observed on extracellular proteins during disease (120, 175), it is possible that despite its cell-permeability, Cl-amidine mainly inhibits extracellular or aberrant PAD activity. 44 We found that the number and proportion of major immune cell populations appear unaltered with Cl-amidine treatment. This is an important observation arguing that drug toxicity of immune cells is not likely the cause of the phenotype that we observe, nor is the elimination of a relevant immune cell population by the inhibition of citrullination. The populations we evaluated were by design broad and encompass various subsets of cell types. Further studies with additional cell markers may reveal changes in more specific immune cell populations that will help elucidate the mechanism of disease reduction by PAD inhibition. In our study, Cl-amidine treatment impaired the IgG1 and IgG2a response to mouse CII without affecting normal humoral responses to bovine CII. The mechanism by which Cl-amidine affects the humoral autoimmune response to other than citrulline epitopes remains to be determined. However, our results may reflect the effects of Cl- amidine on the expression of co-stimulatory or other molecules. While Cl-amidine had no apparent effect on immune cell frequency or numbers, we cannot rule out the possibility of Cl-amidine altering a regulatory cell population (e.g. Treg or Breg) to specifically affect the humoral autoimmune response. The balance of regulatory and pathogenic cells is clearly altered in autoimmune arthritis in mice and in RA (204-206). This balance is in part controlled by soluble cytokines and chemokines. For example, CXCL12 has been reported to alter the balance between Tregs and TH17 cells in EAE (207). Interestingly, citrullination of inflammatory chemokines including CXCL12 (134-136) has been shown to alter immune responses in vitro. Thus, if Cl-amidine is affecting regulatory, effector, or other lymphocytes, it could do so either directly or indirectly. Additionally, Cl-amidine may have an effect on any number of targets involved in maintaining tolerance, like 45 indoleamine 2,3 dioxygenase (IDO), which then results in enhanced autoreactive cell responses (208). In accordance with the mouse IgG1 and IgG2a data discussed above, synovial proteome array analysis of sera from Cl-amidine treated CIA mice compared to controls revealed not only decreased reactivity to many native peptides but also a decreased response to a subset of citrullinated peptides. The arrays utilized have been particularly useful in studies of CIA, as they have demonstrated citrulline-specific reactivity with peptides as well as epitope spreading during the evolution of disease. Our array data also agrees with previous studies that have suggested that citrulline-related epitope spreading occurs in RA due to neoantigen formation as a consequence of inflammatory citrullination (201). Limiting of the spread of autoreactivity may be another explanation of how Cl-amidine treatment reduces disease severity in CIA. Contrary to our hypothesis, these data indicate that the ACPA humoral response is not uniquely affected by Cl- amidine treatment as both citrullinated and uncitrullinated self-antigen reactivity was affected by PAD inhibition. These results suggest that Cl-amidine treatment may exert broader effects on autoantibody responses than those resulting from decreased citrullinated epitope generation or that the effects observed are secondary effects resulting from the amelioration of disease. In conclusion, Cl-amidine treatment ameliorates the clinical and histological disease severity of CIA, reduces serum and joint citrulline levels and decreases epitope spreading in a citrulline-independent manner. Our results suggest that PAD inhibition ameliorates disease, at least in part, by reducing development of autoantigens. Furthermore, these results suggest that Cl-amidine treatment may exert broader effects on 46 autoantibody responses than those resulting from decreased citrullinated epitope generation. 47 CHAPTER IV PAD4 INHIBITION IS SUFFICIENT FOR THE AMELIORATION OF CIA Background and rationale While PAD activity is important in the development of CIA (as discussed in the previous chapter), the roles of specific PAD enzymes in the pathogenesis of disease is unclear. Of particular relevance to RA are PAD2 and PAD4, the only PADs expressed in RA synovium (174), likely due to their expression in immune cells such as neutrophils and macrophages. PAD4 specifically is associated with RA in a number of ways: 1) as the only PAD expressed exclusively in immune cell populations (PAD2 is expressed nearly ubiquitously); 2) the PADI4 gene is associated with increased risk for developing RA in Asian populations (34-36); 3) PAD4 has the ability to translocate to the nucleus and citrullinate histones (117) and has been demonstrated to affect gene transcription (141, 142, 181, 209); 4) autoantibodies to PAD4 are associated with more severe disease and are present in ~45% of patients with RA (210-214). Through the inhibition of individual PADs their contribution to the disease process can be evaluated. Of note, previous studies have shown that PAD2 knockout mice do not citrullinate myelin sheaths in a mouse model of multiple sclerosis (145), but the prevention of citrullination had no effect on disease severity and that PAD4 knockout mice developed the same severity of arthritis as controls in the KBxN mouse model of RA (146). These studies suggest that PADs may not play a significant role in altering the course of autoimmune diseases individually. However, KBxN arthritis is a model that relies on passive transfer of antibodies for the induction of arthritis, not unlike CAIA. Our 48 experiments demonstrated that PAD inhibition had no effect on the course of CAIA (see chapter III), unlike CIA. These data indicate that PAD activity may not be essential for disease development in passive transfer models of arthritis, but that PAD activity is important for the development of autoimmunity in full autoimmune models of disease. As a consequence of our previous work on the effects of PAD inhibition in CIA, GlaxoSmithKline asked us to test their PAD4 inhibitor, GSK283, in CIA. We were excited to undertake this experiment in order to replicate the effects of PAD inhibition on arthritis severity in CIA with a distinct PAD inhibitor (suggesting that PAD inhibition, and not nonspecific drug effects is responsible for disease amelioration) and to test if citrullination mediated specifically by PAD4 is important in the development of disease. The hypothesis tested in this chapter is that PAD4 inhibition is sufficient for disease amelioration in CIA by blocking the development of citrullinated epitopes and ACPA generation and/or affecting other PAD-dependent cellular processes and that PAD inhibition (not off-target drug effects) is responsible for disease amelioration in CIA. Results Treatment with GSK283 decreases clinical disease severity in CIA. To test if inhibition of PAD4 activity would result in decreased disease severity as as observed with pan-PAD inhibition, mice were immunized on days 0 and 21 with bovine CII and clinical disease activity was monitored during days 21-35. From day 0 mice began receiving either 10 mg/kg, 30 mg/kg or 30 mg/kg bid GSK283 SQ, 0.25 mg/kg Dexamethasone IP, 1 mg anti-C5 monoclonal antibody (BB5.1) IP, 0.9% saline 49 SQ or no treatment. Mice continued to receive these treatments daily throughout the course of the experiment with the exception of BB5.1, which was given twice per week. As expected, mice receiving 0.25 mg/kg Dexamethasone developed very little clinical disease, but mice receiving the vehicle control (0.9% saline) showed a 37% decrease in disease severity (measured by area under the curve [AUC]) compared to control mice receiving no treatment, an observation that is sometimes observed in CIA but not fully understood. Mice receiving BB5.1 showed a reduction of disease activity (37%) similar to that of mice receiving saline. However, mice receiving daily SQ injections of GSK283 in 0.9% saline exhibited a reduced clinical disease activity of ~40% compared to saline- treated control mice and of ~61% when compared to control mice receiving no treatment (Figure 6). Interestingly, mice receiving 30 mg/kg GSK283 bid showed less benefit from treatment than mice receiving the 30 mg/kg dose. These results suggested that treatment with GSK283 reduces clinical disease activity in CIA. Treatment with GSK283 decreases histological disease severity in CIA. To evaluate the effects of treatment with GSK283 on joint destruction in CIA, histological analyses of joints were performed at the endpoint of the experiment, day 35. As anticipated, histopathology scores for synovial inflammation, pannus invasion, cartilage damage, and bone damage were markedly reduced (p < 0.05) for the 0.25 mg/kg Dexamethasone and BB5.1 treatment groups compared to saline controls (Figure 7A-D). Scores for all GSK283 treatment groups trended downward in each category, with the 30 mg/kg and 30 mg/kg bid treatment groups demonstrating significantly reduced inflammation and bone damage scores (p < 0.05). The clinical disease activity and histology scores showed a significant correlation with each other (R squared value of 50 Figure 6. PAD4 inhibition decreases clinical disease severity in CIA. Arthritis was induced by two injections, at day 0 and 21, of bovine CII in CFA. Mice were treated as described in Materials and Methods and assessed three times weekly for clinical disease activity. A) All treatment groups (including the saline vehicle control) showed reduced clinical disease activity scores compared to mice receiving no treatment. B) Plot depicting AUC for clinical disease activity for each mouse by treatment group. AUC was decreased by 37% in the saline group compared to no treatment while 24%, 54% and 36% AUC reductions were observed in 10 mg/kg, 30 mg/kg, and 30 mg/kg bid groups respectively when compared to saline treated control mice. Statistically significant changes in AUC compared to saline treated mice were observed in 0.25 mg/kg/day dexamethasone treated mice and mice receiving no treatment *p < 0.05 (n=7 except for the no tx group where n=8). 51 0.92, p < 0.0001) (data not shown). C3 deposition was reduced in the synovium (Figure 7E) and cartilage (Figure 7F) of mice treated with GSK283. Remarkably, treatment with 30 mg/kg bid GSK283 resulted in the elimination of C3 deposition in both compartments (p < 0.01) while treatment with 10 or 30 mg/kg GSK283 also significantly reduced C3 deposition (p < 0.01), marking C3 deposition as the most responsive endpoint of GSK283 treatment. Similar to the clinical disease activity, treatment with saline alone resulted in a decrease in C3 deposition in the synovium and cartilage. These results indicated that GSK283 treatment prevents joint destruction and reduces C3 deposition in CIA. Treatment with GSK283 does not significantly alter joint total citrulline levels. To confirm that the effects of GSK283 are derived from its ability to inhibit PAD4 activity, we measured changes in synovial citrulline content, as a proxy of PAD inhibition. Synovial citrulline content trended downward in response to all GSK283 doses (Figure 8A). Of note, the total citrulline content of joints from saline treated animals was similar to that of untreated control mice. This suggested that while saline treatment decreased clinical disease activity and joint destruction compared to untreated control mice, it did not do so by inhibiting PAD activity. We also measured the serum citrulline content in mice with CIA at day 35 to verify the activity of GSK283. Similar to joint total citrulline levels serum citrulline content trended downward with GSK283 treatment (Figure 8B) but only WT (untreated mice without CIA) and BB5.1 mice demonstrated a statistically significant decrease in total citrulline levels in the serum (p < 0.05). 52 Figure 7 a> A f B 0 3.5 0 3 (J (J . .. tn ~3.0 ~ • C) C) ..2 2.5 0 0 * :s 2 ]i 2.0 .. -~ J: * J: • ; 1.5 .. ** s:::: Q) ** m 1 E 1.0 .. E ** :5 0.5 ••• ...... 0 • ~ 0.0 f c 0 3 (J tn • ~ C) 0 0 2 tn -:E •• s:::: ** co • Q) .. E • • ** ** .. s:::: ...... _ ** ** -·c;...... • ...... E F ** 3.5 • 2.5 f 3.0 •• 0 • (J ..!... en 2.5 - s:::: • 0 2.0 ;; ·u; 0 1.5 a. ** ~ 1.0 M (..) 0.5 ** ** ** 53 Figure 7. GSK283 treatment reduces histological disease severity. Histologic sections were scored for changes in A) synovial inflammation, B) pannus formation, C) cartilage damage and D) bone damage, all on a scale of 0–5. Sections were scored for C3 deposition in both E) the synovium and F) cartilage on a scale of 0-3. Data are expressed as the mean ± SEM based upon a set of five joints per animal. Treatment with 30 mg/kg or 30 mg/kg bid GSK283 resulted in decreased histopathology scores in inflammation and bone damage compared to the saline vehicle group. (n=7 except for no tx where n=8); *p ≤ 0.05. **p < 0.01. 54 Figure 8. GSK283 treatment does not significantly alter joint total citrulline levels. A) Synovial citrulline content. Arthritis was induced in mice by injections of bovine CII in CFA at day 0 and 21. Mice were treated daily, beginning on day 0, with GSK283 (10 mg/kg, 30 mg/kg,or 30 mg/kg bid) in saline, saline, 0.25 mg/kg Dexamethasone, 1 mg BB5.1 twice weekly or were untreated. WT mice received no CII immunizations and are used as a baseline reference. B) Serum citrulline content was measured in mice at day 35. GSK283 treated mice showed slightly decreased levels of total citrulline in the knee and serum, but only WT and BB5.1 mice had statistically significant decreases in citrulline content compared to saline-treated controls. 55 We anticipated that since GSK283 was not a pan-PAD inhibitor total citrulline levels might not be affected as dramatically as we observed with Cl-amidine treatment. To verify that serum GSK283 levels were sufficient to inhibit PAD4 during our study GSK283 treated mice were bled at various timepoints throughout the study for PK determination. As depicted in Figure 9, modeled serum levels of GSK283 suggested that PAD inhibition was achieved (above the IC50) every day for 2, 8 or 15 hours in mice treated with 10 mg/kg, 30 mg/kg or 30 mg/kg bid GSK283 respectively. These data modeled from sporadic blood testing indicated that in vivo levels of GSK283 were sufficiently high throughout the study to inhibit PAD4 and that the lack of decrease in total citrulline levels might be attributed to citrullination by uninhibited PADs. Treatment with GSK283 does not significantly alter CII autoantibody production. As detailed in chapter 2 xeno (bovine) and self (mouse) anti-CII antibodies evolve over time in CIA, reflecting an expanding loss in tolerance to self-epitopes. To see if GSK283 treatment recapitulated the effects of Cl-amidine treatment on autoimmune antibody production, sera from mice in each treatment group were incubated on 96-well plates coated with bovine or mouse CII. The sera were analyzed by ELISA specific for mouse IgG1 and IgG2a Ab to bovine and murine CII. Unlike Cl-amidine treatment, GSK283 treatment did not alter anti-bovine or anti- mouse CII IgG1 or IgG2a antibody levels at any dose (Figure 10). These results demonstrated that GSK283 treatment does not decrease the humoral response to native or foreign CII. 56 Figure 9 A 10 <:: .,0 lo"' c: 0.1 "'u <:: 0 (..) "0 0 0 05 0.01 -10 mg/kg Mean D1 profile • • CIA 10 mglkg D1 • PAD4 SCP IC50 0.001 0 4 8 16 20 24 B Tirritf(h) 100 10 :E 2. <:: ~ ~ <:: u "'<:: 0 (..) 0.1 "0 0 0 • 05 -30 mg/kg Mean D1 QD profile 0.01 I • CIA 30 mgl kg QD D1 PAD4 SCP IC50 0.001 0 4 8 16 20 24 c Tirritf(h) 100 :E 10 2. <:: 0 :;::: c:~ "'u <:: 0 (..) "0 0 0 05 0.1 -30 mgl kg Mean D1 BID profile • CIA 30 mg/kg BID D1 0.01 0 4 8 16 20 24 Tirritf(h) 57 Figure 9. Plots of GSK283 serum levels over time. Plots of modeled PK measurements in CIA mice treated with A) 10 mg/kg, B) 30 mg/kg or C) 30 mg/kg bid GSK283. Red dashed line represents GSK283’s in vitro determined IC50 and solid blue lines depict serum levels of GSK283 modeled at each time point from sample measurements (green squares). Models suggest that serum levels of GSK283 remained above the IC50 for 2, 8 or 15 hours in the 10 mg/kg, 30 mg/kg, and 30 mg/kg bid groups respectively. 58 Figure 10. GSK283 treatment does not reduce autoantibody titers in CIA. Sera from individual mice were incubated on a 96-well plate coated with bovine (A & B) or mouse (C & D) CII. The presence of anti-mouse IgG1 (A & C) or IgG2a antibody (B & D) was detected by ELISA. The 0.25 mg/kg Dexamethasone group showed the lowest autoantibody titers, but no titers from any groups showed statistically significant decreases (n=7 except for no tx where n=8). 59 Treatment with GSK283 reduces epitope spreading in CIA. To examine the effects of GSK283 treatment on antibody responses, a synovial antigen array analysis was performed using day 35 sera derived from mice with CIA treated with 30 mg/kg, 30 mg/kg bid GSK283 or saline. These synovial arrays were described in detail in the previous chapter. Briefly, these arrays contain nearly 200 peptides and proteins, both citrullinated and uncitrullinated, representing candidate antigens in RA (154). SAM identified 7 and 14 autoantibody reactivities that were statistically decreased in CIA mice treated with 30 mg/kg or 30 mg/kg bid GSK283 respectively, compared to PBS treatment (Figure 11). Over 150 native and citrullinated epitopes on the arrays did not exhibit significant differences in reactivity. Sera from mice treated with 30 mg/kg GSK283 exhibited decreased IgG reactivity to native epitopes such as peptides from human fibrinogen alpha and beta chains and keratin while sera from mice treated with 30 mg/kg bid GSK283 demonstrated decreased IgG reactivity to native epitopes including collagen IV, heat shock protein 65, peptides from human fibrinogen alpha and beta chains – two of which match those that were decreased in the 30 mg/kg group, and vimentin as compared to saline controls (q < 0.1%). Decreased autoantibody reactivity to citrullinated peptides from vimentin and fibromodulin were also observed in sera from mice treated with 30 mg/kg bid GSK283 compared to controls treated with saline (q < 0.1%). There were no differences between treatment with GSK283 or saline alone in the levels of autoantibodies to 38 other citrullinated proteins in the array. These data demonstrated that GSK283 treatment decreases the development of autoantibodies in CIA and impairs epitope spreading to a degree similar to that observed with Cl-amidine treatment. 60 Figure II A q < 0.1% I ,...... L:O ....L1 I Scale 0 0 0 0 0 0 0 CD CD CD CD CD CD CD 0> 0> 0> 0> 0> 0> 0> ..!o<: ..!o<: ..!o<: ..!o<: ..!o<: ..!o<: ..!o<: 0, 0, 0, 0, 0, 0, 0, Q) Q) Q) Q) Q) Q) Q) E E E E E E E -~ -~ -~ -~ -~ -~ -~ ro ro ro ro ro ro ro C"') C"') C"') C"') C"') C"') C"') C/) C/) C/) C/) C/) C/) C/) =I =I =I =I =I =I =I I I I I I I I C"') ...... r-. c-.J CD L.C') CD c-.J C"') r-. 00 L.C') ""'" c-.J ""'"c-.J c-.J c-.J c-.J c-.J c-.J =r-. =r-. =r-. =r-. =r-. =r-. =r-. r-. r-. r-. r-. r-. r-. r-. A4285 Collagen type IV A4100 H2B/f 1-20 A4250 Vim 376-395 cit2 * A4307 HSP65 A4190 hFibB 301-320 A4187 hFibB 256-275 A4090 H2B/a 1-20 A4132 hFibA 181-200 A4064 Fibromodulin 345-364 cit2 * A4159 hFibA 481-500 A4246 Vim 181-200 A4042 COMP 45-64 A4229 PG41196-1215 A4133 hFibA 196-215 B I q < 0.1% Scale I ,L1 r::L, >3000 0) 0) 0) 0) 0) 0) If ~ ~ ~ ~ ~ ~ 1000 0) 0) 0) 0) 0) 0) Q) 0) Q) Q) Q) Q) Q) Q) ------E E E E E E c -E c c c c c c 300 0 0 0 0 0 0 co 0 co co co co co co C"') C"') C"') C"') C"') C"') (f) C"') (f) (f) (f) (f) (f) (f) 100 I I I I I I I I I I I I I I <10 co N 0 C"') ...-- <.0 0'> I'- L!) co C"') N ...-- 0 ...-- ...-- ...-- ...-- N 0 N N N N N N ""'"I'- I'- I'- I'- I'- I'- I'- I'- I'- I'- I'- I'- ""'"I'- I'- A4190 hFibB 301-320 A4095 H2B/b 1-20 • A4127 hFibA 151-170 A4175 hFibB 106-125 A4159 hFibA481-500 A4293 Keratin, native A4128 hFibA 16-35 61 Figure 11. GSK283 treatment reduces epitope spreading in CIA. Sera were collected from A) 30 mg/kg bid GSK283 or B) 30 mg/kg GSK283 and control treated mice, and autoantibodies in these samples were profiled using Synovial Antigen Arrays and an anti-IgG/M goat-anti-mouse secondary antibody. Significance Analysis of Microarrays (SAM) was used to analyze the antigen array datasets, and identified 14 and 7 antigens respective to each treatment group with a significant difference in autoantibody reactivity (false discovery rate (q) < 0.1). Hierarchical clustering was then performed to elucidate the relationships between the autoantibody profiles. The results are displayed as a heatmap (blue being negative, yellow intermediate, red strong positive). GSK283 treated mice clustered together (on the left side of the heatmap) and exhibited lower autoantibody titers against all of the 14 and 7 antigens (respectively) identified by SAM, including autoantibody reactivity to native epitopes derived from collagen IV and human fibrinogen alpha and beta chains as well as citrulline-modified vimentin and fibromodulin peptides (red asterisks). 62 Summary and conclusions The hypotheses tested in this chapter were 1) that PAD4 inhibition is sufficient for disease amelioration in CIA by blocking the development of citrullinated epitopes and ACPA generation and/or affecting other PAD-dependent cellular processes and 2) that PAD inhibition (not off-target drug effects) is responsible for disease amelioration in CIA. We found that treatment with GSK283, a PAD4 inhibitor, resulted in the amelioration of clinical and histological disease severity to levels comparable to those achieved with pan-PAD inhibition. Epitope spreading, as measured by peptide array analysis, was reduced in a similar fashion to that seen with Cl-amidine treatment, but no overlap in the targeting of epitope spreading reduction was observed. This observation must be interpreted with the caveat that only a small fraction of antigen reactivities were altered for either compound and that the sensitivity of this assay makes it prone to variation between studies and compounds. Notably, PAD4 inhibition did not significantly alter joint total citrulline levels or antibody titers to mouse CII. Finally, we observed significant decreases in C3 deposition in the joints of mice treated with the highest doses of GSK283. These findings suggest that 1) PAD inhibition, not off-target drug effects is responsible for disease amelioration in CIA, 2) PAD4 inhibition is sufficient for disease amelioration in CIA and 3) that PAD inhibition does not appear to ameliorate CIA by blocking the development of citrullinated epitopes or ACPA generation. Our observation that a distinct PAD inhibitor that utilizes a different mechanism of action, composition and delivery route than Cl-amidine ameliorates CIA argues that PAD inhibition is responsible for disease amelioration. It would be unlikely that two unrelated PAD inhibitors could bring about such similar disease amelioration due to non- 63 specific effects. That Cl-amidine, a pan-PAD inhibitor, and GSK283, a PAD4 inhibitor, ameliorate CIA similarly in the disease parameters evaluated further suggests that PAD4 inhibition is sufficient for disease amelioration in CIA. It is true that total citrulline content in the joints and serum of GSK283 treated mice were not significantly changed compared to vehicle-treated and possibly, Cl-amidine treated mice. However, both PAD2 and PAD4 are expressed in the inflamed joint (174) so it is reasonable to assume that in the presence of GSK283 uninhibited PAD2 continues to citrullinate joint proteins and could mask any decrease in total citrulline levels by PAD4 inhibition. In contrast to our experiments with Cl-amidine we did not observe any significant differences in anti-mouse CII autoantibody titers in GSK283 treated mice. This suggests that PAD4 activity does not regulate autoantibody production directly. However, GSK283 treated mice did display lower autoantibody reactivity to RA peptide arrays. Importantly, mice treated with Cl-amidine or GSK283 did not show reduced reactivity to citrullinated peptides specifically, but demonstrated reduced reactivity to both citrullinated and native peptides. These observations are consistent with a hypothesis of PAD inhibition’s effects on autoantibody production being a secondary effect of the amelioration of disease, rather than a direct result of PAD inhibition. The significant reduction of C3 deposition in the joints of animals treated with the highest doses of GSK283 suggests that PAD4 inhibition may ameliorate disease by altering the complement response. Future experiments must determine the mechanism of the reduction of C3 deposition in the joint but as PAD4 is known to affect gene transcription (141, 142, 181) it is plausible that PAD4 may be regulating complement 64 proteins at the transcript level. Addressing this possibility is explored in chapter V and the Discussion and Future Directions sections. In conclusion, GSK283 treatment decreases epitope spreading in a citrulline- independent manner and ameliorates the clinical and histological disease severity of CIA. However, GSK283 treatment does not appear to reduce total citrulline levels or anti- mouse CII autoantibody titers. Considered with our results using the pan-PAD inhibitor Cl-amidine, these results suggest that PAD inhibition ameliorates disease by exerting broader effects on the immune system than altering the generation of citrullinated epitopes. Further, these results suggest that inhibition of PAD4 is sufficient to ameliorate CIA and that PAD inhibition may play a role in the regulation of complement protein deposition in the joint. 65 CHAPTER V INVESTIGATION OF THE MECHANISM OF ACTION FOR ARTHRITIS AMELIORATION BY PAD INHIBITORS Background and rationale The body of data from the previous two chapters clearly indicates that PAD inhibition ameliorates CIA. However, the mechanism by which PAD inhibition ameliorates CIA is unclear. Since citrullination affects gene transcription, protein activity and conformation, and has been linked to facets of the autoimmune response there are many possible explanations for how PAD inhibition ameliorates CIA. This chapter details the exploration of five different hypotheses to test the mechanism of action of PAD inhibition on the development of inflammatory arthritis: 1) PAD inhibition affects autoantibody production in CIA, 2) PAD inhibition ameliorates CIA by altering antigen processing and/or presentation, 3) PAD inhibition ameliorates CIA by increasing p53 expression and/or activity, 4) PAD inhibition ameliorates CIA by reducing complement factor transcription, and 5) PAD inhibition alters macrophage polarization. The first hypothesis stems from data presented in chapters III and IV: PAD inhibition affects autoantibody production in CIA. Our studies utilizing Cl-amidine, a pan-PAD inhibitor, in CIA initially suggested that PAD inhibition affects the production of autoantibodies but not antibodies to foreign antigens (215). Studies with the PAD4 inhibitor GSK283 confirmed the reduction of epitope spreading observed with Cl- amidine treatment in CIA but failed to replicate the CII autoantibody-specific effects observed previously. Is autantibody production affected by PAD inhibition directly or 66 does the reduction in autoantibody reactivity reflect the amelioration of disease? If PAD inhibition is affecting autoantibody production, it could be doing so by affecting the number of autantibody producing cells or alternately a regulatory cell populations such as Tregs, which can dampen the autoimmune response. The second hypothesis has origins in the reduction of epitope spreading observed in Cl-amidine and GSK283 treated mice as well as work by Ireland et al. that highlights a role for citrullination in antigen processing and presentation during autophagy (128, 129): PAD inhibition ameliorates CIA by altering antigen processing and/or presentation. While we did not observe a citrulline-specific reduction in antigen reactivity in CIA with pan-PAD or PAD4 inhibition, it is possible that the reduction in epitope spreading we observe is due to an alteration of antigen processing and/or presentation regulated by citrullination. Third is a hypothesis based on data suggesting that p53 may play a regulatory role in RA: PAD inhibition ameliorates CIA by increasing p53 expression and/or activity. Studies by Gary Firestein’s lab have demonstrated that p53 knockout mice develop more severe CIA than mice with normal p53 expression, that p53 expression increases in the joints of CIA mice with the onset and progression of disease, and that the absence of p53 does not affect CAIA (216, 217). The clinical data presented in these investigations matches our results with PAD inhibition. Furthermore, p53 is known to affect T cell function in RA (218), can interact directly with PAD enzymes or other transcription factors regulated by citrullination (139, 141, 181) and is known to be expressed at increased levels when PAD activity is inhibited (138, 139, 141, 181). These data show that PAD activity and p53 are linked in several immunologically relevant physiological 67 roles and that PAD inhibition may ameliorate CIA by increasing p53 expression and/or activity. A fourth hypothesis investigating the mechanism of action of PAD inhibition is that PAD inhibition ameliorates CIA by reducing complement factor transcription. The classical and alternative activation pathways of complement may be involved in RA (219) and as mentioned previously ACPA from patients with RA can activate complement in vitro (157). In mice the alternative pathway is necessary and sufficient for CAIA and KBxN arthritis (190, 220). Treatment with Cl-amidine or GSK283 resulted in significantly reduced C3 deposition in the joint in CIA. Strikingly, mice treated with the highest doses of GSK283 showed a near complete elimination of C3 deposition in the joint. These data suggest that PAD inhibition may regulate the complement system. However, Cl-amidine treatment in CAIA had no effect on disease (Figure 3), highlighting the need for further investigation of the effects of PAD inhibition on complement. Since PADs can affect transcription it is possible that PAD inhibition affects complement by altering complement protein transcription, or alternately by affecting an immune cell population that produces complement proteins (e.g. macrophages). The fifth hypothesis investigated to elucidate the mechanism of action of PAD inhibition on the amelioration of inflammatory arthritis in mice is that PAD inhibition alters macrophage polarization. PAD2 and PAD4 are expressed in the joint in RA patients in close association with inflammation (174). Immune cells expressing these PADs include neutrophils, mast cells and macrophages. Studies focusing on citrullination in the formation of NETs have reported no phenotypic abnormalities in neutrophils from PAD4 knockout mice (180) and in a study of KBxN arthrtitis, PAD4 knockout mice did 68 not demonstrate an altered disease course (146). A recent study showed that mast cells predominantly express PAD2 and that P2X7 activation of mast cells induces PAD2 activity and intracellular protein citrullination (221). The physiological roles of citrullination in macrophages are not well understood, however. PADs 2 and 4 are expressed over the course of differentation from monocyte to macrophage (222, 223) and PAD activity has been demonstrated to affect macrophage immunological signaling (224). Given that disease amelioration in CIA seems to be PAD4 dependent and that neutrophil and mast cell PAD activity does not appear to be directly related to PAD4, macrophages remain as an immune cell population likely to be mediating the effects of PAD inhibition on disease. The broad pro-inflammatory, destructive, and remodeling potential of macrophages (reviewed in (225)) combined with their expression of PAD4 make macrophages a likely mediator of amelioration of arthritis by PAD inhibition. In an effort to elucidate the mechanism of action of the amelioration of disease in CIA five hypotheses were tested. The hypotheses tested in this chapter are that PAD inhibition ameliorates CIA by: 1) affecting autoantibody production in CIA, 2) altering antigen processing and/or presentation, 3) increasing p53 expression and/or activity, 4) reducing complement factor transcription and 5) altering macrophage polarization. Results Cl-amidine treatment does not affect autoantibody-producing cell population numbers. To test if inhibition of PAD activity altered autoantibody-producing cell numbers DBA/1J mice were given a single intradermal immunization of bovine CII and CFA or 69 CFA alone. Mice received daily treatment of PBS or 50 mg/kg Cl-amidine in PBS IP daily including the day of the CII immunization. At day 21 mice were sacrificed and the spleen and draining lymph node cells were used in ELISPOT assays to determine the number of anti-bovine CII and anti-mouse CII antibody producing cells in mice from each treatment group. The number of cells producing anti-mouse CII or anti-bovine CII antibodies was quantified and normalized to the number of cells producing IgG. No difference was observed between treatment groups in the numbers of anti-mouse CII, anti-bovine CII, or IgG producing cells (Figure 12). These data suggest that Cl-amidine treatment does not decrease the number of autoantibody producing cells in the spleen or lymph node. Cl-amidine treatment does not affect CD4+CD25+FoxP3+ cell numbers. In order to test if the reductions in autoantibody reactivity associated with Cl- amidine treatment reflect an alteration of Treg numbers, spleen and lymph node cells from CIA mice at day 14 treated daily beginning at day 0 with PBS or 50 mg/kg Cl- amidine in PBS IP were stained for Treg markers (CD4, CD25 and Foxp3) and analyzed by flow cytometry (Figure 13). No difference in CD4+CD25+Foxp3+ cell numbers was observed in either the spleen or lymph nodes of mice treated with 50 mg/kg Cl-amidine or PBS. These data suggest that Cl-amidine treatment does not decrease autoantibody production by decreasing the number of Tregs in the spleen or lymph node. 70 Figure 12. Cl-amidine treatment does not decrease the number of autoantibody producing cells in CIA. Spleen (A, C, E) and lymph node (B, D, F) cells from mice 21 days after immunization with CFA or bovine CII plus CFA and treated with PBS IP or 50 mg/kg Cl-amidine IP daily were assayed by ELISPOT for their ability to secrete anti-bovine CII (A, B), anti- mouse CII (C, D) or total IgG (E, F). No difference in the number of cells secreting IgG or antibodies specific to mouse or bovine CII were observed in Cl-amidine treated mice compared to PBS treated mice. 71 Figure 13. Cl-amidine treatment does not decrease Treg cell numbers in the spleen or lymph node. Splenocytes and draining lymph node cells were harvested from CIA mice 14 days after immunization with bovine CII. Mice received daily treatment IP of either PBS or 50 mg/kg Cl-amidine in PBS starting at day 0. Flow cytometry analysis of the number of CD4+CD25+Foxp3+ Tregs in each tissue showed no differences between treatment groups. 72 Cl-amidine treatment does not alter antigen presentation. To test if Cl-amidine treatment reduces epitope spreading by modifying antigen presentation, PECs were incubated overnight with 10 µM or 1 µM HEL protein or 10 µM HEL peptide and 0, 100 or 200 µM Cl-amidine. Following this incubation T cell hybridomas recognizing citrullinated HEL (Granny) or both citrullinated and uncitrullinated HEL (3A9) were added and hybridoma activation was determined by IL-2 secretion as determine by CTLL proliferation (Figure 14). Cl-amidine treatment of PECs did not affect the activation of Granny or 3A9 hybridomas. These data indicated that PAD inhibition does not alter antigen presentation. Cl-amidine treatment does not increase p53 expression or activity. To test if Cl-amidine treatment increases p53 expression or activity CIA mice were treated daily, beginning on day 0, with PBS or 50 mg/kg Cl-amidine IP. On day 14 or 35 mice were sacrificed and limbs spleens and draining lymph nodes were harvested for mRNA and flow cytometry analysis. qRTPCR using mRNA from mouse hind limbs (from the knee down), spleen or draining lymph nodes showed no difference in p53 transcript levels between Cl-amidine treated mice and controls at day 14 of CIA (Figure 15A-C). Attempts at determining p53 protein levels in the knee were unsuccessful. To determine if Cl-amidine treatment affected p53 activation, transcript levels of p21, a p53 target gene, were determined using qRTPCR. No increase in p21 transcript levels in the knee was observed in Cl-amidine treated mice compared to controls at day 14 (Figure 15D). 73 Figure 14. Cl-amidine treatment does not affect antigen presentation. IL-2 production from T cell hybridomas recognizing citrullinated HEL (Granny) or both citrullinated and uncitrullinated HEL (3A9) were added to PECs incubated overnight with 0, 100 or 200 µM Cl-amidine and A) 10 µM or B) 1 µM HEL protein or C) 10 µM HEL peptide. Data from each panel are representative of at least two independent experiments and each column represents triplicate wells. 74 Figure 15. Cl-amidine treatment does not increase p53 or p21 mRNA levels. mRNA transcript levels from mice at day 14 of CIA treated daily with PBS or 50 mg/kg Cl-amidine (CL50) IP. Levels of p53 transcripts in the A) limb, B) spleen or C) lymph node were unchanged with Cl-amidine treatment. D) Transcript levels of the p53 target gene p21 were also unchanged by Cl-amidine treatment. All transcript levels are normalized to 18-s mRNA transcript expression. 75 To further confirm that Cl-amidine treatment did not affect p53 activation, annexin V staining was performed on splenocytes and draining lymph node cells from these same mice. No significant difference in annexin V, a marker of apoptosis, was observed any of the analyzed conditions (Figure 16). Cl-amidine treatment alters complement protein gene expression. In order to test if Cl-amidine treatment decreases complement gene expression mRNA from the hind limbs, spleen or draining lymph nodes of CIA mice treated daily, beginning on day 0, with PBS or 50 mg/kg Cl-amidine IP was assayed by qRTPCR. Transcript levels for C3 were significantly decreased in the limb and spleen at day 14 of CIA in mice treated with Cl-amidine compared to controls (Figure 17 A-B) while transcript levels of factor B, an activator of the alternative pathway, trended lower in the limb and spleen but did not reach statistical significance (Figure 17 C-D). Transcript levels of factor H, a regulator of complement activity were unchanged in the limb and draining lymph nodes but significantly increased in the spleen (Figure 18 A-C). Of note, Cl-amidine treatment did not change mRNA levels of C3, factor B or factor H in the knee at day 35 of CIA, after the onset of clinical disease (Figure 18 D-F). These results suggested that Cl-amidine treatment may decrease the expression of the complement activating proteins C3 and factor B while increasing expression of the complement regulator protein factor H. 76 Figure 16. Cl-amidine treatment does not affect spleen or lymph node cell apoptosis. Percent apoptotic cells (Annexin V+PI-) from mice at day A) 14 or C) 35 of CIA treated daily with PBS or 50 mg/kg Cl-amidine (CL50) IP. The percent of apoptotic cells in the spleen (A&C) and draining lymph nodes (B) of mice treated with Cl-amidine was not significantly different from vehicle treated controls. 77 Figure 17. Cl-amidine treatment reduces C3 mRNA levels in the joint and spleen. mRNA transcript levels of C3 (A-C) and factor B (D-F) in the knee (A, D), spleen (B, E) and draining lymph nodes (C, F) of CIA mice at day 14 treated with 50 mg/kg Cl- amidine (CL50) or PBS daily from day 0 as measured by qRTPCR. mRNA levels were normalized to 18-s mRNA expression. Treatment with Cl-amidine significantly reduced expression of C3 in the knee and spleen while levels of factor B trended downward with Cl-amidine treatment. *p<0.05, **p<0.01. 78 Figure 18. Cl-amidine treatment increases splenic expression of factor H but does not affect complement protein expression in the knee at day 35. (A-C) mRNA transcript levels of factor H in the A) knee, B) spleen, and C) draining lymph nodes of CIA mice at day 14 treated with 50 mg/kg Cl-amidine (CL50) or PBS daily from day 0 as measured by qRTPCR. mRNA transcript levels of D) C3, E) factor B and F) factor H in the knee of CIA mice at day 35 treated with 50 mg/kg Cl-amidine (CL50) or PBS daily from day 0 as measured by qRTPCR. Treatment with Cl-amidine significantly increased expression of factor H in the spleen compared to controls while trasncript levels for all complement factors measured in the knee were unchanged at day 35. *p<0.05. 79 Cl-amidine treatment inhibits M1 macrophage polarization. To test if Cl-amidine treatment affects macrophage polarization bone marrow macrophages were isolated from C57BL/6 mice and cultured for 24 hours with 20 ng/ml IL-4 and 20 ng/ml IL-13 to induce M2 polarization or 50U IFN γ plus 100 ng/ml LPS to induce M1 polarization, in the presence or absence of 200 µM Cl-amidine. Macrophage polarization was assessed by measuring nitric oxide levels in the supernatant of cultures (for M1) and assaying arginase activity of cell lysates (for M2). Cl-amidine treatment of cells in vitro prevented M1 polarization but did not alter M2 polarization (Figure 19). These data indicated that Cl-amidine treatment prevents M1 polarization in vitro. Summary and conclusions The hypotheses tested in this chapter were that PAD inhibition ameliorates CIA by: 1) affecting autoantibody production in CIA, 2) altering antigen processing and/or presentation, 3) increasing p53 expression and/or activity, 4) reducing complement factor transcription and 5) altering macrophage polarization. These hypotheses were pursued in an effort to identify the mechanism by which PAD inhibition ameliorates CIA. Data from chapter III (Figure 4) suggested that Cl-amidine treatment decreased autoantibody responses specifically. However, inhibition of PAD4 by GSK283 failed to replicate the decrease in autoantibody response, but did confirm the observation of epitope spreading seen with Cl-amidine. Together these data suggested that PAD inhibition might affect autoantibody production in CIA. 80 Figure 19. Cl-amidine treatment inhibits M1 polarization in vitro. BM macrophages were incubated for 24 hours with cytokines to induce polarization in the presence or absence of Cl-amidine. A) M1 polarization was assayed via nitric oxide production and B) M2 polarization was determined by arginase activity. Cl-amidine treatment inhibited M1 polarization but showed no change in M2 polarization capability. ***p<0.001. 81 Two approaches were used to test this hypothesis. First, the number of cells producing antibodies specific for mouse (self) and bovine (foreign) CII in mice with CIA were quantified using ELISPOT. No difference was observed in the number of cells producing antibodies to mouse or bovine CII. These data suggest that the original finding of decreased autoantibody titers while statistically significant, may be an anomaly resulting from biological variability or a secondary effect accompanying disease amelioration. The second approach used to test the hypothesis that Cl-amidine treatment affects autoantibody production in CIA was to measure the number of Treg cells in mice with CIA that were treated with Cl-amidine or PBS. No difference was seen in the number of Tregs under either condition in any of the tissues analysed further indicating that Cl- amidine treatment does not affect autoantibody production specifically. While the data in this chapter argue against a role for PAD inhibition in altering auoantibody responses specifically, experiments using both PAD inhibitors showed a significant decrease in epitope spreading. Two explanations could readily explain this observation: reductions in epitope spreading might be a secondary effect of disease amelioration or it could be connected to the role of citrullination in antigen processing and/or presentation. To test if Cl-amidine treatment affects antigen processing and presentation PECs incubated with HEL protein or peptide were incubated with Cl- amidine. No difference in the activation of hybridomas specific for both citrullinated and uncitrullinated HEL were observed in any condition. These results indicate that Cl- amidine treatment does not inhibit antigen processing and/or presentation and that the reductions in epitope spreading are likely a secondary effect of disease amelioration. 82 PAD inhibition has been demonstrated in multiple in vitro and in vivo studies to increase p53 expression. Experiments by the Firestein lab suggest that p53 may play a protective role in CIA. To test if Cl-amidine treatment increases p53 expression or activity levels of p53 and p21, a target gene of p53 activity, were assayed in limbs of mice treated with Cl-amidine or PBS. This strategy was decided upon after numerous failed attempts at detecting p53 protein in CIA joints. No induction of p53 or p21 gene expression was observed in knees of mice with CIA treated with Cl-amidine compared to controls. Since p53 is mostly regulated at the protein level the lack of change in p53 mRNA levels is unsurprising, however, if p53 gene expression and/or activity were induced by Cl-amidine treatment it is very likely that p21 mRNA levels would be increased in these samples. In an attempt to assay p53 activation, apoptosis of spleen and draining lymph node cells from mice with CIA treated with Cl-amdine was determined by flow cytometry. No difference in the number of annexin+PI- cells was observed in either group indicating that Cl-amidine treatment does not increase p53 activity in CIA. To follow up the significant decrease of C3 deposition in the joints of mice with CIA treated with GSK283, mRNA analyses were performed for complement proteins C3, factor B and factor H. We observed that mice with CIA before the onset of arthritis (day 14) expressed lower levels of C3 in the limb and spleen, with factor B levels trending downward in those same tissues. Furthermore, the regulator of complement activity factor H mRNA levels were increased in the spleens of these same mice. Intriguingly we found that there was no difference in the mRNA expression of these same complement factors after the onset of arthritis (day 35). It is possible that PAD inhibition regulates the expression of complement protein genes, but that this regulation can be overwhelmed 83 during the development of disease by the positive feedback loops of the complement system. This will be explored in greater detail in the Discussion and Future Directions sections of this dissertation. In order to test if macrophages were affected by PAD inhibition, macrophage polarization experiments were performed in vitro to assess Cl-amidine’s ability to alter M1 and M2 polarization. Cl-amidine treatment of polarizing BM macrophages inhibited M1 polarization while having no effect on M2 polarization. The mechanism of this effect remains to be determined, but it is likely that PAD inhibition alters the expression of genes essential for M1 polarization, either directly or indirectly through interactions with transcription factors or the citrullination of histone tails. The little that is known about citrullination in macrophages suggests that it may play a role in regulating NF-κB activity (224). The possible roles of citrullination in macrophage polarization and activity will be expounded upon in the Discussion and Future Directions. 84 CHAPTER VI DETECTION OF AUTOANTIBODIES IN THE SPUTA OF SUBJECTS AT-RISK FOR FUTURE RA Background and rationale The finding of elevations of rheumatoid factor (RF) and antibodies to citrullinated protein antigens (ACPAs) prior to the symptomatic onset of inflammatory arthritis (IA) suggests that rheumatoid arthritis (RA)-related autoimmunity is initiated outside of the joints (226). Although the anatomic site of initiation of RA-related autoantibody production is unknown, emerging data, including the identification of elevations of IgA autoantibodies in subjects prior to the onset of symptomatic RA, suggest that it may be a mucosal site (227). Furthermore, the association of inhaled factors including smoking with increased risk for RA (51), and findings of inflammatory lung abnormalities associated with RA-related autoantibody positivity preceding the onset of articular symptoms in RA (228), suggest that the lung may be a mucosal site of RA initiation. Evaluating the lung to determine that it is a site of initial generation of RA-related autoimmunity presents many challenges; however, prior work demonstrating the generation of autoantibodies within the lung through comparative studies of sputa and sera suggest that such an approach may be used to identify the lung as a site of generation of RA-related autoantibodies in the early development of RA (229). In order to test the hypothesis that RA-related autoantibodies are generated in the lung, we evaluated ACPAs and RF in simultaneously-collected sputa and sera in healthy 85 subjects, subjects at elevated risk for developing RA due to family history of RA or seropositivity for ACPAs, and subjects with early, classified RA. Results Subject demographics. Subjects’ characteristics are presented in Table III. The control group was younger than the other groups; however, there were no significant differences in age, gender or smoking status across FDR, Health-Fair and Early RA subjects. Aside from mild, self-limited post-procedural coughing in 20% of subjects, there were no significant complications from sputa induction. Sputa autoantibody levels and positivity. There was a trend for higher median sputa levels of all autoantibodies in FDRs, Health-Fair and Early RA subjects when compared to healthy controls (Figure 20); however, only levels of CCP2, CCP3.1 and RF-IgM were statistically significantly higher in Early RA subjects when compared to controls (p<0.05 for each autoantibody). There are no established levels for RA-related autoantibody positivity in sputa. Therefore, we established a positive level for each autoantibody by applying a 2 standard deviation increase to mean levels in processed sputa from healthy controls. With this approach, Early RA subjects had the highest prevalence of autoantibody positivity in sera and sputa, although only prevalence of sputa anti-CCP2 and RF-IgM was significantly higher in Early RA compared to FDR and Health-Fair subjects (p<0.05) (Table III and 86 Table III. Subject characteristics and autoantibody positivity in sera and sputa. Healthy FDRs Health-Fair Early RA Controls (N=29) (N=18) (N=12) (N=17) Age, Median (range) 32 54 (23-79) 57 (30-70) 50 (29-75)2 (26-58) Female, N 13 (77%) 21 (72%) 9 (50%) 8 (67%) (% female) Ever smoker, 5 (29%) 11 (38%) 8 (44%) 4 (33%) N (%)* History of lung 1 (6%) 7 (24%) 4 (22%) 2 (17%) disease1 Autoantibody Serum Serum Sputum Serum Sputum Serum Sputum positivity, N (%) Anti-CCP2 0 (0%) 0 (0%) 3 (10%) 3 (17%) 4 (22%) 11 (92%) 6 (50%)4 Anti-CCP3.1 0 (0%) 2 (7%)3 9 (31%)3 14(78%)3 4 (22%)3 11 (92%)3 5 (42%)3 RF-IgA 0 (0%) 0 (0%)3 4 (14%)3 0 (0%) 3 (17%) 8 (67%) 5 (42%) RF-IgG 0 (0%) 6 (21%) 5 (17%) 4 (22%) 1 (6%) 8 (67%)3 2 (17%)3 RF-IgM 0 (0%) 4 (14%) 5 (17%) 3 (17%) 4 (22%) 11 (92%) 8 (67%)4 Any autoantibody 0 (0%) 8 (28%) 12 (41%) 15 (83%) 8 (44%) 12 (100%) 9 (75%) 0 autoantibodies 17 (100%) 21 (72%) 17 (59%) 3 (17%) 10(56%) 0 (0%) 3 (25%) 1 autoantibody 0 (0%) 4 (14%) 5 (17%) 7 (39%) 3 (17%) 1 (8%) 1 (8%) 2 autoantibodies 0 (0%) 4 (14%) 3 (10%) 7 (39%) 2 (11%) 1 (8%) 1 (8%) 3 autoantibodies 0 (0%) 0 (0%) 1 (3%) 1 (6%) 2 (11%) 2 (17%) 4 (33%) 4 autoantibodies 0 (0%) 0 (0%) 3 (10%) 0 (0%) 0 (0%) 0 (0%) 2 (17%) 5 autoantibodies 0 (0%) 0 (0%) 0 (0%) 0 (0%) 1 (6%) 8 (67%) 1 (8%) 1) Self-reported history of lung disease (asthma, emphysema, chronic bronchitis or bronchiectasis) diagnosed by a health-care provider at some point prior to the sputa collection visit. 2) P-value <0.01 for comparison of median age across all groups (Healthy Controls are younger); however, there were no significant differences in median age across the FDR, Health-Fair and Early RA groups. 3) P-values <0.05 in paired analyses comparing within each group the proportion of subjects with positivity for each autoantibody at each site (sera vs. sputa). 4) P-values <0.05 comparing the proportions of Early RA subjects with sputa positivity for anti-CCP2 and RF-IgM to FDRs and Health-Fair subjects. *There was no significant association between smoking status and autoantibody positivity in sera or sputa. Figure 20). Of note, only 2 subjects were positive for all 5 autoantibodies in their sputa; 1 subject had Early RA, and the other was a Health-Fair subject positive only for CCP3.1 in serum who developed classifiable articular RA approximately 6 months after the sputa collection visit. Of note, to evaluate for potential non-specific inflammation and increases in total sputa proteins affecting our results, we normalized the results of sputa autoantibodies to total protein levels. This resulted in no significant changes in sputa autoantibody positivity compared to the unadjusted values (data not shown). 87 Figure 20. Autoantibody levels in the sputa. Plots depicting autoantibody levels in the sputa of healthy controls, first-degree relatives (FDR), Health-Fair and Early Rheumatoid Arthritis (RA) subjects. Solid lines represent the median for each group. Dashed lines indicate cutoff values generated by adding 2 standard deviations to the healthy control group mean. Open circles denote ever-smokers. *For anti-CCP2, the median level was significantly elevated in Early RA subjects compared to healthy controls, FDR and Health-Fair subjects (p<0.01); for anti-CCP3.1, the median level was significantly elevated in Early RA compared to healthy controls (p<0.05); for RF-IgM, the median level was significantly elevated in Early RA subjects compared to healthy controls and FDRs (p<0.05). 88 Sputa versus sera autoantibody positivity. Comparisons of positivity for each autoantibody in sera and sputa are presented in Table III and Figure 21. Overall, for each of the autoantibodies, a higher proportion of Early RA and Health-Fair subjects were positive in their sera compared to sputa, although in each of these groups some individuals were positive for specific autoantibodies only in their sputa. In the FDRs, 21/29 (72%) were seronegative for all autoantibodies, and of these seronegative FDRs, 8 (38%) were positive for at least 1 autoantibody in their sputa. Specifically, 3 seronegative FDRs were positive in their sputa for 1 autoantibody, 2 for 2, 1 for 3 and 2 for 4, with the autoantibody types including anti-CCP2 (2 subjects), anti- CCP3.1 (7 subjects), RF-IgA (3 subjects), RF-IgG (4 subjects), and RF-IgM (2 subjects). In analyses of all FDRs (N=29), CCP3.1 and RF-IgA were positive in a higher proportion of FDRs’ sputa compared to sera (CCP3.1: 31% vs. 7%, p=0.04; RF-IgA: 14% vs 0%; p=0.04). Summary and conclusions We demonstrate herein that elevations of RA-related autoantibodies are detectable in the sputa of subjects that have established RA, as well as in subjects who are at elevated risk for future RA. The finding in a subset of subjects of autoantibody positivity in sputa but not sera suggests that the lung may be a site of generation of some of these autoantibodies. Furthermore, in FDRs, the finding of significantly more subjects with sputa compared to sera positivity for CCP3.1 and RF-IgA, assays that both detect IgA, 89 Figure 21. Numbers of subjects positive for autoantibodies in the sputum only, serum only, neither, or both. Counts of (A) First-degree relatives (FDR), (B) Health-Fair, and (C) Early RA subjects testing positive for autoantibodies in the serum, sputum, neither or both, are represented by stacked bars. Counts for each classification are listed in their respective bars. In paired analyses of the proportions of subjects positive at each site (sputa vs. sera), in FDRs, CCP3.1 and RF-IgA were elevated in significantly more subjects’ sputa than sera (p=0.04 for each autoantibody); in Health-Fair subjects, anti-CCP3.1 was elevated in significantly more subjects’ sera than sputa (p<0.01); in Early RA subjects, anti-CCP3.1 and RF-IgG were elevated in significantly more subjects’ sera than sputa (p=0.01 and 0.03, respectively). 90 suggests that the lung may be a site of initial generation of autoantibodies, and related to IgA responses. There are several lines of evidence to support that the mucosa of the lung is a site of generation of RA-related autoimmunity. In particular, within the airways, inducible bronchus-associated lymphatic tissue (iBALT) can develop in response to infections or other stimuli (99, 100). Once present, iBALT produces immune factors, including antibodies (IgA, IgG and IgM isotypes), that can be detected at the mucosal surface of the lung (99, 100). Importantly, Rangel-Moreno and colleagues have linked iBALT to RA by demonstrating that a high prevalence of patients with RA-related lung disease have iBALT (100). Additionally, they demonstrated that plasma cells within RA- associated iBALT generate both RF and ACPAs (100). Therefore, it is possible that our findings of elevated RA-related autoantibodies in induced sputa, which preferentially samples the bronchial tree, reflects underlying generation of RA-related autoimmunity within the airways. While we have not demonstrated inflammatory airways disease or iBALT in the subjects studied herein, the potential that they have such inflammation associated with sputa autoantibodies is in accord with our published work demonstrating that a high proportion of autoantibody positive subjects without IA, and patients with Early RA, have airways abnormalities consistent with inflammation on high-resolution computed tomographic imaging (228). Caveats to this include the possibility that autoantibodies are present in sputa due to translocation from the circulation; however, we believe this is unlikely given the differences between sputa and blood positivity across subject groups. Elevations of autoantibodies in sputa of subjects with established RA, and those at-risk for developing disease, suggest that sputa testing may be a safe and informative 91 method to investigate the lung’s role in the pathogenesis of RA, and especially the lung as a site of initiation of RA-related autoimmunity. 92 CHAPTER VI DISCUSSION The contribution of citrullination to disease The original hypothesis proposed in this dissertation was that the development of citrullinated epitopes and ACPA is pathogenic in RA. We tested this hypothesis by treating mice with Cl-amidine, a pan-PAD inhibitor, from day 0 of CIA until the experiment’s endpoint. We demonstrated that Cl-amidine treatment decreases clinical and histological disease severity, lowers serum and joint total citrulline levels, and decreases epitope spreading as measured by peptide array. This evidence was confirmed by similar observations of decreased disease severity and epitope spreading utilizing GSK283, a specific inhibitor of PAD4, in the same model. These results suggest that the development of citrullinated epitopes and ACPA is pathogenic in RA. However, other results from our work, and the work of others, indicate that it is unclear if the development of citrullinated epitopes and ACPA is pathogenic in RA. If the development of citrullinated epitopes and ACPA are pathogenic in RA then why doesn’t Cl-amidine or GSK283 treatment prevent disease initiation? It is possible that we were unable to prevent the development of disease because at the doses utilized in our studies PAD inhibition does not eliminate all citrullination. However, it is questionable that increased doses of Cl-amidine and GSK283 would result in increased beneficial clinical since a plateu of clinical effect was observed at the highest doses of each compound. Cl-amidine preferentially binds activated PAD enzymes (126) and it is possible that Cl-amidine is only able to inhibit “active” PADs, i.e. PADs that are present 93 in a sufficiently calcium-rich environment (e.g., extracellularly). Since all PADs are normally intracellular and increased levels of citrullination are observed on extracellular proteins during disease (120, 175), it is possible that despite its cell-permeability, Cl- amidine mainly inhibits extracellular or aberrant PAD activity. If such is the case citrulline-reactive T and B cells could still be activated in a bystander manner as part of a nonself response (230). Autoreactive citrulline-specific clones activated in the inflammatory environment following immunization with CFA-containing antigen would be primed to react with intracellularly citrullinated molecules (e.g. histones) that are released during cell turnover and normal immune processes (221). These clones could then initiate a break in self-tolerance if Cl-amidine or GSK283 are unable to inhibit intracellular citrullination. However, PAD2 knockout mice develop experimental autoimmune encephalomyelitis (EAE) in the complete absence of myelin sheath citrullination (145), suggesting that even if all citrullination is inhibited it may not alter the development of disease. An alternate explanation for the inability of Cl-amidine or GSK283 treatment to prevent CIA onset completely is that anti-citrulline reactivity is not responsible for the initial break in tolerance in CIA, but amplifies disease severity after disease onset. This is consistent with our previously published data, where infusion of ACPA did not initiate disease in mice but augmented the severity of CAIA induced by a sub-maximal dose of anti-CII Abs (153). This has also been demonstrated in another citrulline-related autoimmune disease model (231), EAE. In EAE it has been shown that citrullinated- antigen-reactive T cells were unable to induce disease, but were able to augment disease in the presence of disease-initiating non-citrulline autoreactive clones. Also, mice 94 immunized with native fibrinogen, but not citrullinated fibrinogen, develop mild arthritis (232). Of relevance are findings showing that immunization with citrullinated antigens can give rise to mild arthritis in some cases (154, 233). Nonetheless, it is possible that PAD inhibition prevents the activation of citrulline-reactive but not disease initiating cells resulting in decreased disease activity but not complete prevention. Relevant to the discussion of citrullination’s ability to initiate disease is the post- translational modification homocitrullination (carbamylation). Homocitrullination is the nonenzymatic reaction of urea-derived cyanate with free NH2- groups on lysine residues. Recent studies have demonstrated that homocitrullinated proteins can give rise to antibodies to both citrullinated and homocitrullinated proteins and that citrulline detection reagents can cross-react with homocitrulline (234). This may explain in part why our assays did not detect an elimination of citrulline under conditions of PAD inhibition. However, a study of subjects with RA demonstrated that while ~40% of subjects with RA test positive for anti-carbamylated protein (CarP) antibodies, ACPA and anti-CarP antibodies are different antibody families that have little cross-reactivity (235). Furthermore one study demonstrated that mice immunized with homocitrullinated peptides followed by an intra-articular injection of citrullinated peptide developed significantly enhanced arthritis compared to mice immunized with citrullinated peptides followed by an intra-articular injection of either homocitrullinated or citrullinated peptides (236). This is another demonstration of citrullination’s ability to enhance disease severity, and argues that disease initiation may occur in response to an uncitrullinated antigen. In fact, myeloperoxidase-catalyzed protein homocitrullination is linked to inflammation and smoking (237), both relevant factors in the initiation of RA. 95 ACPA epitope spreading (158, 238) and avidity maturation (160) expand prior to RA onset, arguing that they may play a role in clinical disease onset. Our studies show no evidence of reduced ACPA development by peptide array, but they do demonstrate reduced epitope spreading to native and some citrullinated proteins in the joint. This is consistent with these effects being a secondary effect of decreased joint inflammation and destruction by some other mechanism. However, it could also represent the effects of PAD inhibition on an immune cell population involved in regulating autoimmunity rather than the alteration of citrullinated epitope generation. We were unable to detect any differences in major immune cell populations in the spleen and draining lymph nodes or pro-inflammatory and anti-inflammatory cytokine mRNA levels in the joint (data not shown). Furthermore, in follow-up experiments we detected no difference in the number of CD4+CD25+FoxP3+ regulatory T cells. These data indicate that PAD inhibition does not affect the autoimmune response by altering Tregs or cyokines influencing other immune players but cannot rule out the possibility of effects that were not assayed (e.g. Treg functional studies and more specific flow cytometry staining of immune cell populations). Another explanation for the amelioration of disease in CIA with PAD inhibtitors is the post-translational regulation of chemokines at the protein level, which would be undetectable by our mRNA assays. Citrullination of chemokines is known to affect immunological signaling and chemoattractivity (134-136) and so it is possible that mediators of cell recruitment and inflammation are regulated in this way. Two lines of evidence argue contrary to this hypothesis. First is our data showing no difference in neutrophil and macrophage recruitment to the inflamed joint in PAD inhibitor treated 96 mice; second is the lack of effect of Cl-amidine in CAIA, a mouse model of arthritis highly dependent on neutrophil recruitment and activation. Instead, these results argue that PAD inhibition affects the adaptive immune response rather than the effector phase of immunity. The amelioration of disease in CIA by the inhibition of PADs suggests that ACPA and the development of citrullinated epitopes are pathogenic in CIA. However, our experiments with two distinct PAD inhibitors resulting in no change in pathogenic autoantibody production and the reduction of epitope spreading in a citrulline- independent manner indicate that PAD inhibition may be exerting its ameliorative effects in CIA by affecting broader immunological processes than ACPA generation and citrullinated epitope generation. Indeed, PAD inhibition or knockdown has shown efficacy in reducing clinical severity in models of colitits (138) and breast cancer (239), conditions unassociated with citrulline-specific autoimmunity. This further suggests that PAD inhibition is altering disease in a manner independent of citrullinated epitope generation. Reactivity to citrullinated proteins is pathogenic in that it is likely responsible for the amplification of disease, but the question of whether citrullinated proteins have the capability of initating disease directly remains to be answered. The contribution of PAD4 to disease PAD2 and PAD4, as the only PADs expressed in the RA synovium are of particular relevance to RA (174). We tested the hypothesis that PAD4 inhibition is sufficient for disease amelioration in CIA by treating mice with CIA with GSK283, an inhibitor of PAD4. Mice treated with GSK283 developed less severe clinical and 97 histological disease severity than mice treated with vehicle alone, had unchanged levels of total citrulline in the joints and serum, and demonstrated reduced epitope spreading in the absence of altered anti-mouse or anti-bovine CII titers. These results indicate that PAD4 inhibition is sufficient for disease amelioration in CIA. No decrease in total citrulline levels in the joint or serum was observed after treatment with GSK283. This is perhaps unsurprising given that other PADs were presumably active and citrullinating protein targets. These results are consistent with a recent report indicating that the localization of citrullinated peptides and not necessarily their presence is RA specific (240) in that they suggest that not all citrullination is necessarily involved in the development of RA. Our data are consistent with an effect mediated not by citrullinated epitope or ACPA reduction, but by the role of PAD4 activity in some other immunologic process. While we cannot rule out the contribution of nonspecific effects on the course of disease, the replication of disease amelioration with a distinct PAD inhibitor suggests that such is not likely the case. GSK283 treatment, like Cl-amidine treatment, reduced epitope spreading in a citrulline-independent manner as measured by peptide array. However, none of the antigens to which antibody reactivity was reduced were the same as those identified with Cl-amidine treatment. Some of this variability could be attributed to biological variation between experiments but the distinct antigens recognized even between arrays from mice treated with 30 mg/kg or 30 mg/kg bid GSK283 argue against such a hypothesis. ACPA autoimmunity is known to evolve from a single reactivity to many as disease onset approaches, but the primal citrullinated antigen varies between individuals suggesting that perhaps no shared original autoantigen is present in RA (238). Furthermore very 98 limited overlap between antigens from our studies and those performed to originally validate the peptide arrays we utilized exists (201). These data suggest that once disease is initiated the evolution of autoimmunity and epitope spreading may proceed with some variability. Unlike Cl-amidine treatment, GSK283 treatment did not reduce anti-mouse CII antibody titers. This data in conjunction with our observation of unchanged numbers of anti-mouse CII producing cells as measured by ELIPSOT in Cl-amidine treated mice suggests that PAD inhibition may not be directly related to a reduction in autoantibody titers. Since anti-mouse CII antibody titers do not necessarily track tightly with disease severity in CIA, as evidenced by our own dexamethasone treated mice who developed very little clinical or histological disease but mounted anti-mouse CII antibody titers similar to mice developing severe arthritis, it is likely that the decrease in anti-mouse CII antibody titers with Cl-amidine treatment is not a direct effect of PAD inhibition, or represents a PAD2 specific effect that would not be replicated with GSK283 treatment. Perhaps the most striking observation of PAD4 inhibition is the decrease in C3 deposition in the joints of mice with CIA treated with GSK283. GSK283 treatment reduced C3 deposition in a dose-dependent manner that did not necessarily predict decreased joint destruction or inflammation. This is evidenced through the lack of increased clinical or histological disease amelioration between the 30 mg/kg and 30 mg/kg bid treatment groups, but a significant dose dependent decrease in C3 deposition. These results could arise from an effect on complement activation (e.g. C3 convertase activity) or the generation of complement factors themselves. 99 The amelioration of disease that we observe with GSK283 treatment is consistent with an effect on one or more complement activation pathways. The alternative pathway of complement is known to be important in both human (157) and mouse arthritis (190, 241) and acts as an amplifier of antibody-mediated classical pathway activation (220). Consequently it is possible that PAD4 inhibition affects the ability of the alternative pathway of complement to amplify the generation and deposition of C3 by antibody- mediated classical pathway initiation. This would reduce the generation of C5a, a major proinflammatory complement factor involved in arthritic joint destruction and inflammation (242). Our data are consistent with such a scenario in that we demonstrated a reduction in clinical and histological disease severity, but not C3 deposition, in the joints of mice treated with BB5.1, an inhibitor of C5. However, in CAIA arthritis induction has been demonstrated to be dependent upon the alternative pathway of complement (190). This contrasts with our study of Cl-amidine treatment in CAIA where no effect on disease severity or C3 deposition was apparent with PAD inhibition. It is possible that PAD4 but not PAD2 is responsible for regulating these changes in complement activation, especially considering the superior potency and selectivity of GSK283 to inhibit PAD4. An alternate explanation for this discrepancy could be that PAD inhibition can only control a threshold amount of complement activation that was overwhelmed by our strongly activating CAIA conditions. Experiments testing GSK283 treatment in CAIA and treatment with PAD inhibitors using sub-maximal doses of arthritogenic antibodies in the induction of CAIA are needed to clarify these questions. Another explanation for altered C3 deposition with GSK283 treatment is that PAD inhibition alters the levels of complement proteins. Since PAD activity can regulate 100 gene expression (110, 118, 139, 181) it would be expected that PAD inhibtion might be affecting complement activity at the mRNA level by modifying the ability of transcription factors like hepatocyte nuclear factor 4 to promote complement factor transcription (243). Our preliminary studies of complement factor mRNA levels in the joint and lymphoid organs confirm this hypothesis. mRNA levels of C3 and factor B were significantly lower in the joints of mice immunized with bovine CII and CFA treated with Cl-amidine than in controls. Of note is the observation that these transcript levels were decreased in the absence of arthritis. No difference in the levels of these same transcripts was apparent when measured after the onset of arthritis. However, this may be a reflection of the use of different inhibitors with different capabilities to prevent PAD4 activity or could also be explained by the fact that complement protein activation is amplified at various steps, which may mask the effects of PAD inhibition on constraining complement activation. This may also be another explanation for why disease is ameliorated, but not prevented, by PAD inhibition in CIA. While complement protein levels may be regulated by PAD inhibition at the transcript level, the location of their regulation is of some importance. Complement proteins activated in the inflamed joint can originate from the serum or the synovial fluid (219, 244). Osteoclasts and macrophages in the joint can produce complement proteins (245, 246) locally in the joint. Decreases in joint C3 deposition may be the result of the effects of PAD inhibition on these complement-generating cell types. Our preliminary experiments on the ability of PAD inhibitors to alter macrophage polarization support the concept of this hypothesis, but experiments designed to compare complement protein levels in the serum and joint are needed to adequately address this question. 101 The effects of PAD inhibition on macrophage polarization We found that Cl-amidine treatment inhibited the ability of BM macrophages to polarize to M1 activated macrophages but had no effect on their ability to polarize to M2 activated macrophages. Since these results were generated using exogenous cytokines they suggest that PAD inhibition’s effects on M1 polarization are due to an effect of citrullination on gene transcription or signaling proteins rather than a decrease in the activity of polarizing cytokines themselves from citrullination. PAD2 protein levels have been demonstrated to increase during the differentiation of monocytes to macrophages (222, 223), suggesting that PAD2 is involved in macrophage differentiation. The mechanism by which PAD2 affects macrophage differentiation is unclear but it is likely that it exerts effects on gene transcription by interacting with transcription factors. Although little is understood about the role of PADs in signal regulation, Lee et al. demonstrated that PAD2 coimmunoprecipitates with IKKγ, a stimulator of NF-κB activity, and that this interaction inhibits IKKγ’s stimulatory potential (224). This report indicates that the inhibition of PAD activity would be pro-inflammatory for this signaling pathway. However, the role for PADs in macrophage inflammatory signaling may be more nuanced than shown by this single study and from the data we have presented it is plausible that citrullination could regulate the activity of other signaling molecules like STAT1, which is involved in IFNγ and LPS M1 polarization, but not STAT6, which is involved in M2 polarization (247). Additionally, it is possible that PAD2 and PAD4 regulate distinct signaling pathways. This could be evaluated by treating BM macrophages under polarizing conditions with GSK283 to see if M1 polarization is inhibited. Furthermore, it is possible that PADs are involved in regulating both pro and 102 anti-inflammatory signaling pathways by interacting with different scaffolding or signaling proteins. All of these possibilities need to be tested and are discussed further in Chapter VII, Future Directions. The detection of autoantibodies in the sputa of subjects at-risk for future RA Our studies of sputa from subjects at-risk for future RA demonstrate that RA- related autoantibodies are elevated in the sputa of subjects that have established RA, as well as in subjects who are at elevated risk for future RA. That a subset of subjects were autoantibody positive in sputa but not sera suggests that the lung may be a site of generation of some of these autoantibodies. Furthermore, in FDRs, the finding of significantly more subjects with sputa compared to sera positivity for CCP3.1 and RF- IgA, assays that both detect IgA, suggests that the lung may be a site of initial generation of autoantibodies, and related to IgA responses. However, the presence in several FDRs and the majority of Health-Fair and Early RA subjects of greater positivity for RA-related autoantibodies in sera versus sputa indicates that autoantibodies may also be generated at sites other than the airways, and there are several potential explanations for these findings. First, it is unlikely the lung is the site of generation of autoantibodies in all subjects during the natural history of RA. Other mucosal sites including the periodontal region and gut are also candidate sites for generation of autoantibodies. Additionally, the joints are known to be a source of autoantibody generation in subjects with established RA, and that may be an important factor when assessing the lung generation of autoantibodies in subjects with active synovitis (248). Second, changes within the lung in autoantibody production may occur 103 over time, resulting in greater positivity for autoantibodies in sera versus sputa. Such changes include the development of autoantibody isotypes and subclasses that preferentially move to the circulation rather than sputa, and the movement of autoantibody production from a mucosal source such as iBALT that produces autoantibodies readily detected in sputa, to production in regional lymph nodes leading to greater levels of systemic rather than mucosal autoantibodies (99). Notably, the Health- Fair subjects were seropositive for RA-related autoantibodies ≥1 year prior to their participation in this study, and we assume that the Early RA subjects similarly had a prolonged period of autoimmunity at the time of their sputa evaluation. As such, if these subjects had initial generation of autoimmunity in their lungs, over time these types of changes may have led to greater autoantibody positivity in sera versus sputa. Finally, in subjects with greater positivity of autoantibodies in sera versus sputa, the lung may still be a site of generation of autoantibodies, although not detected in sputa because of sampling issues, levels that fluctuate over time, or generation of very low levels of autoantibodies. (These latter arguments also apply to evaluations of autoantibodies in sera.) 104 CHAPTER VII FUTURE DIRECTIONS Future experiments to address the questions raised by this work regarding PAD inhibition in arthritis revolve around three major themes: 1) clarifying the role of PAD4 versus pan-PAD inhibition in the amelioration of arthritis; 2) exploring the effects of PAD inhibition on complement protein activation and transcription; 3) understanding the molecular mechanism of inhibiting M1 polarization with PAD inhibition. To evaluate if there are any mechanistic differences between disease amelioration by Cl-amdine or GSK283 treatment, the use of GSK283 in CAIA is necessary. Furthermore, utilizing both inhibitors in experiments with sub-maximal doses of arthritogenic antibodies in CAIA may reveal ameliorative effects that are masked by the strong disease induction procedures that were used in this work. These experiments also contribute to understanding the role of PAD inhibition on complement protein transcription and/or activation as described below. In addition, the exploratory experiments identifying complement transcription and macrophage polarization (described below) as potential mechanisms of disease amelioration using Cl-amidine need to be performed using GSK283. The preliminary results of this dissertation demonstrating an effect on complement protein transcription by PAD inhibition need to be repeated and expanded to 1) confirm that PAD inhibtion does affect complement protein transcription and/or activity and 2) to evaluate the mechanism by which this is brought about. In order to confirm that PAD inhibition affects complement transcription and/or protein activity complement transcript levels need to be measured in additional tissues (e.g. the liver, a 105 major tissue source of complement protein production), at additional time points (preceding CII immunization as well as during the interval following the booster CII immunization but prior to clinical disease onset) and in in vitro cell production systems. Cytokine stimulation of macrophages in culture is known to enhance complement protein production (249) and could be assayed in the presence or absence of PAD inhibitors under cytokine stimulating or resting conditions. Additionally, experiments determining serum complement protein levels and activating potential in both arthrtitic and unmanipulated mice treated with PAD inhibitors are necessary to understand if any effects of PAD inhibition are systemic, as well as to confirm that the reductions observed in mRNA transcript levels affect protein levels. If an effect on complement is confirmed there are numerous in vitro and in vivo models and assays that can be utilized to elucidate the mechanism by which PAD inhibition alters complement protein production and/or activity. These assays include the use of knockout mice for various complement factors, the CAIA model of arthritis and in vitro and ex vivo evaluations of complement protein activity. If reductions in complement protein transcript levels are confirmed an mRNA array could help guide investigations of the effects of PAD inhibtion on transcription factors known to regulate complement protein gene expression, such as hepatocyte nuclear factor 4. In this dissertation Cl-amidine treatment was shown to inhibit M1 macrophage polarization in vitro. To expand the understanding of how this is brought about in vitro macrophage experiments should be repeated using GSK283 to elucidate PAD4’s role in macrophage polarization, and experiments utilizing both Cl-amidine and GSK283 under additional M1 polarizing cytokine stimulus (e.g. TNFα, alone and in combination) to 106 dissect if the effects of PAD inhibition are upstream of all signaling pathways involved in M1 polarization or if PAD inhibition only affects some signaling pathways. Gene expression array experiments in BM macrophages treated with M1 polarizing cytokines in the presence and absence of PAD inhibitors will be very informative for targeting further studies investigating the signaling pathways involved in M1 polarization inhibition. If signaling molecule gene expression is unchanged with PAD inhibition this may imply that post-translational modification of signaling proteins such as STAT1, JAK1 or PIAS is responsible for M1 polarization inhibition under these conditions. Further in vivo macrophage polarization experiments are needed to confirm the relevance of our in vitro findings. Ideally the evaluation of macrophages isolated from inflamed synovium in mice treated with PAD inhibitors or vehicle control will be performed, but due to technical restrictions this will prove to be a difficult undertaking. Preceding these experiments macrophage polarization in the presence of PAD inhibitors could be performed in a tissue with greater macrophage abundance, like the lung. The induction of M1 macrophages by tracheal delivery of LPS in the lungs of mice treated with or without PAD inhibitors would reveal the in vivo effects of PAD inhibition on macrophage polarization and help to inform future studies of macrophage polarization in the joint. We observed RA-related autoantibodies in the sputa of subjects at-risk for developing future RA. Going forward, our findings need to be replicated in larger studies that include longitudinal comparisons of simultaneously-collected biospecimens from multiple sites (e.g. the oral cavity, lung, gut, genitourinary tract, blood, and joints) to investigate how autoimmunity may develop in the lung or elsewhere, as well as to investigate how autoimmunity may be generated at a mucosal site and then progress to arthritis. Such 107 studies should include investigations of the roles of cells, immunoglobulin class- switching, subtype development, epitope spreading, and autoantigens and inflammation in the development of mucosal autoimmunity. Studies should also evaluate the potential mechanisms, including isotype and specific antigen detection, which explain the differences seen herein in rates of positivity between CCP2 and CCP3.1 assays in order to understand the biology of autoimmunity as well as to determine the optimal assays to identify biomarkers of early RA in the lung. Many of these investigations could utilize sputa samples, which are relatively inexpensive and safe to obtain, although due to the paucity of T and B cells in sputa, and potential of sputa assessments to miss factors that may be present in underlying mucosal tissue, other methods such as bronchoscopy with lavage or lung biopsy may be necessary to explore fully the role of the lung in early RA pathogenesis (99, 200). Future studies should also investigate the potential impact of production of autoimmunity in the lung on lung injury, and investigate genetic and environmental factors (such as the shared epitope, tobacco smoke or microorganisms) that may drive the generation of autoimmunity in the lung (of note, in our study, smoking was not associated with sera or sputa positive for autoantibodies). Importantly, given our finding of RA-related autoantibodies in the sputa of seronegative FDRs, it is likely that subjects who are at-risk for future RA but who have not yet developed circulating autoimmunity are ideal for investigating the earliest aspects of the initiation of RA. Such subjects could be identified through projects such as SERA, or the similar North American Native project that has demonstrated elevations of circulating cytokines/chemokines in FDRs in the absence of serum positivity for ACPAs or RF 108 (250); investigating such subjects for the generation of mucosal autoimmunity should be highly informative. 109 REFERENCES 1. Gabriel, S. E., and K. Michaud. 2009. Epidemiological studies in incidence, prevalence, mortality, and comorbidity of the rheumatic diseases. Arthritis Res Ther 11:229. 2. McInnes, I. B., and J. R. O'Dell. 2010. State-of-the-art: rheumatoid arthritis. Ann Rheum Dis 69:1898-1906. 3. Gabriel, S. E., C. S. Crowson, and W. M. O'Fallon. 1999. Mortality in rheumatoid arthritis: have we made an impact in 4 decades? J Rheumatol 26:2529-2533. 4. Whiting, P. F., N. Smidt, J. A. Sterne, R. Harbord, A. Burton, M. Burke, R. Beynon, Y. Ben-Shlomo, J. Axford, and P. Dieppe. 2010. Systematic review: accuracy of anti-citrullinated Peptide antibodies for diagnosing rheumatoid arthritis. Ann Intern Med 152:456-464; W155-466. 5. Majka, D. S., and V. M. Holers. 2003. Can we accurately predict the development of rheumatoid arthritis in the preclinical phase? Arthritis Rheum 48:2701-2705. 6. Nielen, M. M., D. van Schaardenburg, H. W. Reesink, J. W. Twisk, R. J. van de Stadt, I. E. van der Horst-Bruinsma, M. H. de Koning, M. R. Habibuw, and B. A. Dijkmans. 2006. Simultaneous development of acute phase response and autoantibodies in preclinical rheumatoid arthritis. Ann Rheum Dis 65:535-537. 7. Rantapaa-Dahlqvist, S., K. Boman, A. Tarkowski, and G. Hallmans. 2007. Up regulation of monocyte chemoattractant protein-1 expression in anti-citrulline antibody and immunoglobulin M rheumatoid factor positive subjects precedes onset of inflammatory response and development of overt rheumatoid arthritis. Ann Rheum Dis 66:121-123. 8. Hall, F. C., D. E. Weeks, J. P. Camilleri, L. A. Williams, N. Amos, C. Darke, K. Gibson, K. Pile, B. P. Wordsworth, and J. D. Jessop. 1996. Influence of the HLA- DRB1 locus on susceptibility and severity in rheumatoid arthritis. QJM 89:821- 829. 9. Nepom, G. T. 1998. Major histocompatibility complex-directed susceptibility to rheumatoid arthritis. Adv Immunol 68:315-332. 10. Gregersen, P. K., J. Silver, and R. J. Winchester. 1987. The shared epitope hypothesis. An approach to understanding the molecular genetics of susceptibility to rheumatoid arthritis. Arthritis Rheum 30:1205-1213. 110 11. Hill, J. A., S. Southwood, A. Sette, A. M. Jevnikar, D. A. Bell, and E. Cairns. 2003. 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. J Immunol 171:538-541. 12. Ding, B., L. Padyukov, E. Lundstrom, M. Seielstad, R. M. Plenge, J. R. Oksenberg, P. K. Gregersen, L. Alfredsson, and L. Klareskog. 2009. 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 Rheum 60:30-38. 13. van Gaalen, F. A., J. van Aken, T. W. Huizinga, G. M. Schreuder, F. C. Breedveld, E. Zanelli, W. J. van Venrooij, C. L. Verweij, R. E. Toes, and R. R. de Vries. 2004. Association between HLA class II genes and autoantibodies to cyclic citrullinated peptides (CCPs) influences the severity of rheumatoid arthritis. Arthritis Rheum 50:2113-2121. 14. Irigoyen, P., A. T. Lee, M. H. Wener, W. Li, M. Kern, F. Batliwalla, R. F. Lum, E. Massarotti, M. Weisman, C. Bombardier, E. F. Remmers, D. L. Kastner, M. F. Seldin, L. A. Criswell, and P. K. Gregersen. 2005. Regulation of anti-cyclic citrullinated peptide antibodies in rheumatoid arthritis: contrasting effects of HLA-DR3 and the shared epitope alleles. Arthritis Rheum 52:3813-3818. 15. Barnetche, T., A. Constantin, A. Cantagrel, A. Cambon-Thomsen, and P. A. Gourraud. 2008. New classification of HLA-DRB1 alleles in rheumatoid arthritis susceptibility: a combined analysis of worldwide samples. Arthritis Res Ther 10:R26. 16. Gourraud, P. A., P. Dieude, J. F. Boyer, L. Nogueira, A. Cambon-Thomsen, B. Mazieres, F. Cornelis, G. Serre, A. Cantagrel, and A. Constantin. 2007. A new classification of HLA-DRB1 alleles differentiates predisposing and protective alleles for autoantibody production in rheumatoid arthritis. Arthritis Res Ther 9:R27. 17. Gourraud, P. A., J. F. Boyer, T. Barnetche, M. Abbal, A. Cambon-Thomsen, A. Cantagrel, and A. Constantin. 2006. A new classification of HLA-DRB1 alleles differentiates predisposing and protective alleles for rheumatoid arthritis structural severity. Arthritis Rheum 54:593-599. 18. Deighton, C. M., D. J. Walker, I. D. Griffiths, and D. F. Roberts. 1989. The contribution of HLA to rheumatoid arthritis. Clin Genet 36:178-182. 19. Gregersen, P. K., H. S. Lee, F. Batliwalla, and A. B. Begovich. 2006. PTPN22: setting thresholds for autoimmunity. Semin Immunol 18:214-223. 111 20. Hasegawa, K., F. Martin, G. Huang, D. Tumas, L. Diehl, and A. C. Chan. 2004. PEST domain-enriched tyrosine phosphatase (PEP) regulation of effector/memory T cells. Science 303:685-689. 21. Barton, A., W. Thomson, X. Ke, S. Eyre, A. Hinks, J. Bowes, L. Gibbons, D. Plant, A. G. Wilson, I. Marinou, A. Morgan, P. Emery, S. Steer, L. Hocking, D. M. Reid, P. Wordsworth, P. Harrison, and J. Worthington. 2008. Re-evaluation of putative rheumatoid arthritis susceptibility genes in the post-genome wide association study era and hypothesis of a key pathway underlying susceptibility. Hum Mol Genet 17:2274-2279. 22. Wang, Y., S. A. Rollins, J. A. Madri, and L. A. Matis. 1995. Anti-C5 monoclonal antibody therapy prevents collagen-induced arthritis and ameliorates established disease. Proc Natl Acad Sci U S A 92:8955-8959. 23. Wang, Y., J. Kristan, L. Hao, C. S. Lenkoski, Y. Shen, and L. A. Matis. 2000. A role for complement in antibody-mediated inflammation: C5-deficient DBA/1 mice are resistant to collagen-induced arthritis. J Immunol 164:4340-4347. 24. Vergunst, C. E., D. M. Gerlag, H. Dinant, L. Schulz, M. Vinkenoog, T. J. Smeets, M. E. Sanders, K. A. Reedquist, and P. P. Tak. 2007. Blocking the receptor for C5a in patients with rheumatoid arthritis does not reduce synovial inflammation. Rheumatology (Oxford) 46:1773-1778. 25. Arch, R. H., R. W. Gedrich, and C. B. Thompson. 1998. Tumor necrosis factor receptor-associated factors (TRAFs)--a family of adapter proteins that regulates life and death. Genes Dev 12:2821-2830. 26. Tsitsikov, E. N., D. Laouini, I. F. Dunn, T. Y. Sannikova, L. Davidson, F. W. Alt, and R. S. Geha. 2001. TRAF1 is a negative regulator of TNF signaling. enhanced TNF signaling in TRAF1-deficient mice. Immunity 15:647-657. 27. Bishop, G. A. 2004. The multifaceted roles of TRAFs in the regulation of B-cell function. Nat Rev Immunol 4:775-786. 28. Raychaudhuri, S., E. F. Remmers, A. T. Lee, R. Hackett, C. Guiducci, N. P. Burtt, L. Gianniny, B. D. Korman, L. Padyukov, F. A. Kurreeman, M. Chang, J. J. Catanese, B. Ding, S. Wong, A. H. van der Helm-van Mil, B. M. Neale, J. Coblyn, J. Cui, P. P. Tak, G. J. Wolbink, J. B. Crusius, I. E. van der Horst- Bruinsma, L. A. Criswell, C. I. Amos, M. F. Seldin, D. L. Kastner, K. G. Ardlie, L. Alfredsson, K. H. Costenbader, D. Altshuler, T. W. Huizinga, N. A. Shadick, M. E. Weinblatt, N. de Vries, J. Worthington, M. Seielstad, R. E. Toes, E. W. Karlson, A. B. Begovich, L. Klareskog, P. K. Gregersen, M. J. Daly, and R. M. Plenge. 2008. Common variants at CD40 and other loci confer risk of rheumatoid arthritis. Nat Genet 40:1216-1223. 112 29. Remmers, E. F., R. M. Plenge, A. T. Lee, R. R. Graham, G. Hom, T. W. Behrens, P. I. de Bakker, J. M. Le, H. S. Lee, F. Batliwalla, W. Li, S. L. Masters, M. G. Booty, J. P. Carulli, L. Padyukov, L. Alfredsson, L. Klareskog, W. V. Chen, C. I. Amos, L. A. Criswell, M. F. Seldin, D. L. Kastner, and P. K. Gregersen. 2007. STAT4 and the risk of rheumatoid arthritis and systemic lupus erythematosus. N Engl J Med 357:977-986. 30. Consortium, W. T. C. C. 2007. Genome-wide association study of 14,000 cases of seven common diseases and 3,000 shared controls. Nature 447:661-678. 31. Plenge, R. M., C. Cotsapas, L. Davies, A. L. Price, P. I. de Bakker, J. Maller, I. Pe'er, N. P. Burtt, B. Blumenstiel, M. DeFelice, M. Parkin, R. Barry, W. Winslow, C. Healy, R. R. Graham, B. M. Neale, E. Izmailova, R. Roubenoff, A. N. Parker, R. Glass, E. W. Karlson, N. Maher, D. A. Hafler, D. M. Lee, M. F. Seldin, E. F. Remmers, A. T. Lee, L. Padyukov, L. Alfredsson, J. Coblyn, M. E. Weinblatt, S. B. Gabriel, S. Purcell, L. Klareskog, P. K. Gregersen, N. A. Shadick, M. J. Daly, and D. Altshuler. 2007. Two independent alleles at 6q23 associated with risk of rheumatoid arthritis. Nat Genet 39:1477-1482. 32. Todd, J. A., N. M. Walker, J. D. Cooper, D. J. Smyth, K. Downes, V. Plagnol, R. Bailey, S. Nejentsev, S. F. Field, F. Payne, C. E. Lowe, J. S. Szeszko, J. P. Hafler, L. Zeitels, J. H. Yang, A. Vella, S. Nutland, H. E. Stevens, H. Schuilenburg, G. Coleman, M. Maisuria, W. Meadows, L. J. Smink, B. Healy, O. S. Burren, A. A. Lam, N. R. Ovington, J. Allen, E. Adlem, H. T. Leung, C. Wallace, J. M. Howson, C. Guja, C. Ionescu-Tirgoviste, M. J. Simmonds, J. M. Heward, S. C. Gough, D. B. Dunger, L. S. Wicker, and D. G. Clayton. 2007. Robust associations of four new chromosome regions from genome-wide analyses of type 1 diabetes. Nat Genet 39:857-864. 33. van Heel, D. A., L. Franke, K. A. Hunt, R. Gwilliam, A. Zhernakova, M. Inouye, M. C. Wapenaar, M. C. Barnardo, G. Bethel, G. K. Holmes, C. Feighery, D. Jewell, D. Kelleher, P. Kumar, S. Travis, J. R. Walters, D. S. Sanders, P. Howdle, J. Swift, R. J. Playford, W. M. McLaren, M. L. Mearin, C. J. Mulder, R. McManus, R. McGinnis, L. R. Cardon, P. Deloukas, and C. Wijmenga. 2007. A genome-wide association study for celiac disease identifies risk variants in the region harboring IL2 and IL21. Nat Genet 39:827-829. 34. Suzuki, A., R. Yamada, X. Chang, S. Tokuhiro, T. Sawada, M. Suzuki, M. Nagasaki, M. Nakayama-Hamada, R. Kawaida, M. Ono, M. Ohtsuki, H. Furukawa, S. Yoshino, M. Yukioka, S. Tohma, T. Matsubara, S. Wakitani, R. Teshima, Y. Nishioka, A. Sekine, A. Iida, A. Takahashi, T. Tsunoda, Y. Nakamura, and K. Yamamoto. 2003. Functional haplotypes of PADI4, encoding citrullinating enzyme peptidylarginine deiminase 4, are associated with rheumatoid arthritis. Nat Genet 34:395-402. 113 35. Plenge, R. M., L. Padyukov, E. F. Remmers, S. Purcell, A. T. Lee, E. W. Karlson, F. Wolfe, D. L. Kastner, L. Alfredsson, D. Altshuler, P. K. Gregersen, L. Klareskog, and J. D. Rioux. 2005. Replication of putative candidate-gene associations with rheumatoid arthritis in >4,000 samples from North America and Sweden: association of susceptibility with PTPN22, CTLA4, and PADI4. Am J Hum Genet 77:1044-1060. 36. Hung, H. C., C. Y. Lin, Y. F. Liao, P. C. Hsu, G. J. Tsay, and G. Y. Liu. 2007. The functional haplotype of peptidylarginine deiminase IV (S55G, A82V and A112G) associated with susceptibility to rheumatoid arthritis dominates apoptosis of acute T leukemia Jurkat cells. Apoptosis 12:475-487. 37. van der Woude, D., J. J. Houwing-Duistermaat, R. E. Toes, T. W. Huizinga, W. Thomson, J. Worthington, A. H. van der Helm-van Mil, and R. R. de Vries. 2009. Quantitative heritability of anti-citrullinated protein antibody-positive and anti- citrullinated protein antibody-negative rheumatoid arthritis. Arthritis Rheum 60:916-923. 38. Costenbader, K. H., D. Feskanich, L. A. Mandl, and E. W. Karlson. 2006. Smoking intensity, duration, and cessation, and the risk of rheumatoid arthritis in women. Am J Med 119:503 e501-509. 39. Criswell, L. A., L. A. Merlino, J. R. Cerhan, T. R. Mikuls, A. S. Mudano, M. Burma, A. R. Folsom, and K. G. Saag. 2002. Cigarette smoking and the risk of rheumatoid arthritis among postmenopausal women: results from the Iowa Women's Health Study. Am J Med 112:465-471. 40. Lahiri, M., C. Morgan, D. P. Symmons, and I. N. Bruce. 2011. Modifiable risk factors for RA: prevention, better than cure? Rheumatology (Oxford) 51:499-512. 41. Klareskog, L., P. Stolt, K. Lundberg, H. Kallberg, C. Bengtsson, J. Grunewald, J. Ronnelid, H. E. Harris, A. K. Ulfgren, S. Rantapaa-Dahlqvist, A. Eklund, L. Padyukov, and L. Alfredsson. 2006. A new model for an etiology of rheumatoid arthritis: smoking may trigger HLA-DR (shared epitope)-restricted immune reactions to autoantigens modified by citrullination. Arthritis Rheum 54:38-46. 42. Lee, H. S., P. Irigoyen, M. Kern, A. Lee, F. Batliwalla, H. Khalili, F. Wolfe, R. F. Lum, E. Massarotti, M. Weisman, C. Bombardier, E. W. Karlson, L. A. Criswell, R. Vlietinck, and P. K. Gregersen. 2007. Interaction between smoking, the shared epitope, and anti-cyclic citrullinated peptide: a mixed picture in three large North American rheumatoid arthritis cohorts. Arthritis Rheum 56:1745-1753. 43. Bang, S. Y., K. H. Lee, S. K. Cho, H. S. Lee, K. W. Lee, and S. C. Bae. Smoking increases rheumatoid arthritis susceptibility in individuals carrying the HLA- DRB1 shared epitope, regardless of rheumatoid factor or anti-cyclic citrullinated peptide antibody status. Arthritis Rheum 62:369-377. 114 44. Mikuls, T. R., H. Sayles, F. Yu, T. Levan, K. A. Gould, G. M. Thiele, D. L. Conn, B. L. Jonas, L. F. Callahan, E. Smith, R. Brasington, L. W. Moreland, R. J. Reynolds, and S. L. Bridges, Jr. 2010. Associations of cigarette smoking with rheumatoid arthritis in African Americans. Arthritis Rheum 62:3560-3568. 45. Carlens, C., M. P. Hergens, J. Grunewald, A. Ekbom, A. Eklund, C. O. Hoglund, and J. Askling. 2010. Smoking, use of moist snuff, and risk of chronic inflammatory diseases. Am J Respir Crit Care Med 181:1217-1222. 46. Kallberg, H., L. Padyukov, R. M. Plenge, J. Ronnelid, P. K. Gregersen, A. H. van der Helm-van Mil, R. E. Toes, T. W. Huizinga, L. Klareskog, and L. Alfredsson. 2007. Gene-gene and gene-environment interactions involving HLA-DRB1, PTPN22, and smoking in two subsets of rheumatoid arthritis. Am J Hum Genet 80:867-875. 47. Costenbader, K. H., S. C. Chang, I. De Vivo, R. Plenge, and E. W. Karlson. 2008. Genetic polymorphisms in PTPN22, PADI-4, and CTLA-4 and risk for rheumatoid arthritis in two longitudinal cohort studies: evidence of gene- environment interactions with heavy cigarette smoking. Arthritis Res Ther 10:R52. 48. Keenan, B. T., L. B. Chibnik, J. Cui, B. Ding, L. Padyukov, H. Kallberg, C. Bengtsson, L. Klareskog, L. Alfredsson, and E. W. Karlson. 2010. Effect of interactions of glutathione S-transferase T1, M1, and P1 and HMOX1 gene promoter polymorphisms with heavy smoking on the risk of rheumatoid arthritis. Arthritis Rheum 62:3196-3210. 49. Hart, J. E., F. Laden, R. C. Puett, K. H. Costenbader, and E. W. Karlson. 2009. Exposure to traffic pollution and increased risk of rheumatoid arthritis. Environ Health Perspect 117:1065-1069. 50. Hart, J. E., H. Kallberg, F. Laden, T. Bellander, K. H. Costenbader, M. Holmqvist, L. Klareskog, L. Alfredsson, and E. W. Karlson. 2012. Ambient air pollution exposures and risk of rheumatoid arthritis: results from the Swedish EIRA case-control study. Ann Rheum Dis. 51. Klareskog, L., J. Ronnelid, K. Lundberg, L. Padyukov, and L. Alfredsson. 2008. Immunity to citrullinated proteins in rheumatoid arthritis. Annu Rev Immunol 26:651-675. 52. Scher, J. U., and S. B. Abramson. 2011. The microbiome and rheumatoid arthritis. Nat Rev Rheumatol 7:569-578. 53. Chou, C. T., J. Uksila, and P. Toivanen. 1998. Enterobacterial antibodies in Chinese patients with rheumatoid arthritis and ankylosing spondylitis. Clin Exp Rheumatol 16:161-164. 115 54. Senior, B. W., G. A. Anderson, K. D. Morley, and M. A. Kerr. 1999. Evidence that patients with rheumatoid arthritis have asymptomatic 'non-significant' Proteus mirabilis bacteriuria more frequently than healthy controls. J Infect 38:99-106. 55. de Pablo, P., T. Dietrich, and T. E. McAlindon. 2008. Association of periodontal disease and tooth loss with rheumatoid arthritis in the US population. J Rheumatol 35:70-76. 56. Vaahtovuo, J., E. Munukka, M. Korkeamaki, R. Luukkainen, and P. Toivanen. 2008. Fecal microbiota in early rheumatoid arthritis. J Rheumatol 35:1500-1505. 57. Bhatia, S. S., D. S. Majka, J. M. Kittelson, L. A. Parrish, E. D. Ferucci, K. D. Deane, W. P. Arend, M. Rewers, V. Michael Holers, and J. M. Norris. 2007. Rheumatoid factor seropositivity is inversely associated with oral contraceptive use in women without rheumatoid arthritis. Ann Rheum Dis 66:267-269. 58. Shapiro, J. A., T. D. Koepsell, L. F. Voigt, C. E. Dugowson, M. Kestin, and J. L. Nelson. 1996. Diet and rheumatoid arthritis in women: a possible protective effect of fish consumption. Epidemiology 7:256-263. 59. Heliovaara, M., P. Knekt, K. Aho, R. K. Aaran, G. Alfthan, and A. Aromaa. 1994. Serum antioxidants and risk of rheumatoid arthritis. Ann Rheum Dis 53:51-53. 60. Makrygiannakis, D., M. Hermansson, A. K. Ulfgren, A. P. Nicholas, A. J. Zendman, A. Eklund, J. Grunewald, C. M. Skold, L. Klareskog, and A. I. Catrina. 2008. Smoking increases peptidylarginine deiminase 2 enzyme expression in human lungs and increases citrullination in BAL cells. Ann Rheum Dis 67:1488- 1492. 61. McGraw, W. T., J. Potempa, D. Farley, and J. Travis. 1999. Purification, characterization, and sequence analysis of a potential virulence factor from Porphyromonas gingivalis, peptidylarginine deiminase. Infect Immun 67:3248- 3256. 62. Travis, J., R. Pike, T. Imamura, and J. Potempa. 1997. Porphyromonas gingivalis proteinases as virulence factors in the development of periodontitis. J Periodontal Res 32:120-125. 63. Kinloch, A. J., S. Alzabin, W. Brintnell, E. Wilson, L. Barra, N. Wegner, D. A. Bell, E. Cairns, and P. J. Venables. 2011. Immunization with porphyromonas gingivalis enolase induces autoimmunity to mammalian alpha-enolase and arthritis in DR4-IE-transgenic mice. Arthritis Rheum 63:3818-3823. 64. Marotte, H., P. Farge, P. Gaudin, C. Alexandre, B. Mougin, and P. Miossec. 2006. The association between periodontal disease and joint destruction in rheumatoid arthritis extends the link between the HLA-DR shared epitope and severity of bone destruction. Ann Rheum Dis 65:905-909. 116 65. Abou-Raya, S., A. Abou-Raya, A. Naim, and H. Abuelkheir. 2008. Rheumatoid arthritis, periodontal disease and coronary artery disease. Clin Rheumatol 27:421- 427. 66. Pischon, N., T. Pischon, J. Kroger, E. Gulmez, B. M. Kleber, J. P. Bernimoulin, H. Landau, P. G. Brinkmann, P. Schlattmann, J. Zernicke, F. Buttgereit, and J. Detert. 2008. Association among rheumatoid arthritis, oral hygiene, and periodontitis. J Periodontol 79:979-986. 67. Moen, K., J. G. Brun, T. M. Madland, T. Tynning, and R. Jonsson. 2003. Immunoglobulin G and A antibody responses to Bacteroides forsythus and Prevotella intermedia in sera and synovial fluids of arthritis patients. Clin Diagn Lab Immunol 10:1043-1050. 68. Moen, K., J. G. Brun, M. Valen, L. Skartveit, E. K. Eribe, I. Olsen, and R. Jonsson. 2006. Synovial inflammation in active rheumatoid arthritis and psoriatic arthritis facilitates trapping of a variety of oral bacterial DNAs. Clin Exp Rheumatol 24:656-663. 69. Queiroz-Junior, C. M., M. F. Madeira, F. M. Coelho, V. V. Costa, R. L. Bessoni, L. F. Sousa, G. P. Garlet, D. D. Souza, M. M. Teixeira, and T. A. Silva. 2011. Experimental Arthritis Triggers Periodontal Disease in Mice: Involvement of TNF-{alpha} and the Oral Microbiota. J Immunol. 70. Maradit-Kremers, H., P. J. Nicola, C. S. Crowson, K. V. Ballman, and S. E. Gabriel. 2005. Cardiovascular death in rheumatoid arthritis: a population-based study. Arthritis Rheum 52:722-732. 71. Crowson, C. S., P. J. Nicola, H. M. Kremers, W. M. O'Fallon, T. M. Therneau, S. J. Jacobsen, V. L. Roger, K. V. Ballman, and S. E. Gabriel. 2005. How much of the increased incidence of heart failure in rheumatoid arthritis is attributable to traditional cardiovascular risk factors and ischemic heart disease? Arthritis Rheum 52:3039-3044. 72. Nicola, P. J., H. Maradit-Kremers, V. L. Roger, S. J. Jacobsen, C. S. Crowson, K. V. Ballman, and S. E. Gabriel. 2005. The risk of congestive heart failure in rheumatoid arthritis: a population-based study over 46 years. Arthritis Rheum 52:412-420. 73. Goodson, N. J., N. J. Wiles, M. Lunt, E. M. Barrett, A. J. Silman, and D. P. Symmons. 2002. Mortality in early inflammatory polyarthritis: cardiovascular mortality is increased in seropositive patients. Arthritis Rheum 46:2010-2019. 74. Myasoedova, E., C. S. Crowson, P. J. Nicola, H. Maradit-Kremers, J. M. Davis, 3rd, V. L. Roger, T. M. Therneau, and S. E. Gabriel. 2011. The influence of rheumatoid arthritis disease characteristics on heart failure. J Rheumatol 38:1601- 1606. 117 75. Doran, M. F., C. S. Crowson, G. R. Pond, W. M. O'Fallon, and S. E. Gabriel. 2002. Frequency of infection in patients with rheumatoid arthritis compared with controls: a population-based study. Arthritis Rheum 46:2287-2293. 76. Baecklund, E., A. Iliadou, J. Askling, A. Ekbom, C. Backlin, F. Granath, A. I. Catrina, R. Rosenquist, N. Feltelius, C. Sundstrom, and L. Klareskog. 2006. Association of chronic inflammation, not its treatment, with increased lymphoma risk in rheumatoid arthritis. Arthritis Rheum 54:692-701. 77. Haugeberg, G., T. Uhlig, J. A. Falch, J. I. Halse, and T. K. Kvien. 2000. Bone mineral density and frequency of osteoporosis in female patients with rheumatoid arthritis: results from 394 patients in the Oslo County Rheumatoid Arthritis register. Arthritis Rheum 43:522-530. 78. van Staa, T. P., P. Geusens, J. W. Bijlsma, H. G. Leufkens, and C. Cooper. 2006. Clinical assessment of the long-term risk of fracture in patients with rheumatoid arthritis. Arthritis Rheum 54:3104-3112. 79. Lane, N. E., A. R. Pressman, V. L. Star, S. R. Cummings, and M. C. Nevitt. 1995. Rheumatoid arthritis and bone mineral density in elderly women. The Study of Osteoporotic Fractures Research Group. J Bone Miner Res 10:257-263. 80. Kroger, H., R. Honkanen, S. Saarikoski, and E. Alhava. 1994. Decreased axial bone mineral density in perimenopausal women with rheumatoid arthritis--a population based study. Ann Rheum Dis 53:18-23. 81. Dawson, J. K., H. E. Fewins, J. Desmond, M. P. Lynch, and D. R. Graham. 2001. Fibrosing alveolitis in patients with rheumatoid arthritis as assessed by high resolution computed tomography, chest radiography, and pulmonary function tests. Thorax 56:622-627. 82. Metafratzi, Z. M., A. N. Georgiadis, C. V. Ioannidou, Y. Alamanos, M. P. Vassiliou, A. K. Zikou, G. Raptis, A. A. Drosos, and S. C. Efremidis. 2007. Pulmonary involvement in patients with early rheumatoid arthritis. Scand J Rheumatol 36:338-344. 83. Gabbay, E., R. Tarala, R. Will, G. Carroll, B. Adler, D. Cameron, and F. R. Lake. 1997. Interstitial lung disease in recent onset rheumatoid arthritis. Am J Respir Crit Care Med 156:528-535. 84. Fischer, A., J. J. Solomon, R. M. du Bois, K. D. Deane, A. L. Olson, E. R. Fernandez-Perez, T. J. Huie, A. D. Stevens, M. B. Gill, A. M. Rabinovitch, D. A. Lynch, D. A. Burns, I. S. Pineiro, S. D. Groshong, R. D. Duarte Achcar, K. K. Brown, R. J. Martin, and J. J. Swigris. 2012. Lung disease with anti-CCP antibodies but not rheumatoid arthritis or connective tissue disease. Respir Med 106:1040-1047. 118 85. Demoruelle, M. K., M. H. Weisman, P. L. Simonian, D. A. Lynch, P. B. Sachs, I. F. Pedraza, A. R. Harrington, J. R. Kolfenbach, C. C. Striebich, Q. N. Pham, C. D. Strickland, B. D. Petersen, M. C. Parish, L. A. Derber, J. M. Norris, V. M. Holers, and K. D. Deane. 2011. Airways abnormalities and rheumatoid arthritis- related autoantibodies in subjects without arthritis: Early injury or initiating site of autoimmunity? Arthritis Rheum. 86. Gomez, A., D. Luckey, C. J. Yeoman, E. V. Marietta, M. E. Berg Miller, J. A. Murray, B. A. White, and V. Taneja. 2012. Loss of sex and age driven differences in the gut microbiome characterize arthritis-susceptible 0401 mice but not arthritis-resistant 0402 mice. PLoS ONE 7:e36095. 87. Kasser, U. R., C. Gleissner, F. Dehne, A. Michel, B. Willershausen-Zonnchen, and W. W. Bolten. 1997. Risk for periodontal disease in patients with longstanding rheumatoid arthritis. Arthritis Rheum 40:2248-2251. 88. Mercado, F., R. I. Marshall, A. C. Klestov, and P. M. Bartold. 2000. Is there a relationship between rheumatoid arthritis and periodontal disease? J Clin Periodontol 27:267-272. 89. Scher, J. U., C. Ubeda, M. Equinda, R. Khanin, Y. Buischi, A. Viale, L. Lipuma, M. Attur, M. H. Pillinger, G. Weissmann, D. R. Littman, E. G. Pamer, W. A. Bretz, and S. B. Abramson. 2012. Periodontal disease and the oral microbiota in new-onset rheumatoid arthritis. Arthritis Rheum. 90. Oliver, J. E., and A. J. Silman. 2006. Risk factors for the development of rheumatoid arthritis. Scand J Rheumatol 35:169-174. 91. Brown, K. K. 2007. Rheumatoid lung disease. Proc Am Thorac Soc 4:443-448. 92. Gochuico, B. R., N. A. Avila, C. K. Chow, L. J. Novero, H. P. Wu, P. Ren, S. D. MacDonald, W. D. Travis, M. P. Stylianou, and I. O. Rosas. 2008. Progressive preclinical interstitial lung disease in rheumatoid arthritis. Arch Intern Med 168:159-166. 93. Gizinski, A. M., M. Mascolo, J. L. Loucks, A. Kervitsky, R. T. Meehan, K. K. Brown, V. M. Holers, and K. D. Deane. 2009. Rheumatoid arthritis (RA)-specific autoantibodies in patients with interstitial lung disease and absence of clinically apparent articular RA. Clin Rheumatol 28:611-613. 94. Verstappen, S. M., M. J. McCoy, C. Roberts, N. E. Dale, A. B. Hassell, and D. P. Symmons. 2009. Beneficial effects of a 3-week course of intramuscular glucocorticoid injections in patients with very early inflammatory polyarthritis: results of the STIVEA trial. Ann Rheum Dis 69:503-509. 119 95. Birnbaum, H., C. Pike, R. Kaufman, and M. Cifaldi. 2009. Employer model of workplace impacts of anti-TNF therapy for rheumatoid arthritis. J Occup Environ Med 51:1167-1176. 96. Levine, S. M., N. Raben, D. Xie, F. B. Askin, R. Tuder, M. Mullins, A. Rosen, and L. A. Casciola-Rosen. 2007. Novel conformation of histidyl-transfer RNA synthetase in the lung: the target tissue in Jo-1 autoantibody-associated myositis. Arthritis Rheum 56:2729-2739. 97. Reynolds, H. 2004. Respiratory structure and function. 22nd ed. Philadelphia: Saunders. 98. Moyron-Quiroz, J. E., J. Rangel-Moreno, K. Kusser, L. Hartson, F. Sprague, S. Goodrich, D. L. Woodland, F. E. Lund, and T. D. Randall. 2004. Role of inducible bronchus associated lymphoid tissue (iBALT) in respiratory immunity. Nat Med 10:927-934. 99. Twigg, H. L., 3rd. 2005. Humoral immune defense (antibodies): recent advances. Proc Am Thorac Soc 2:417-421. 100. Rangel-Moreno, J., L. Hartson, C. Navarro, M. Gaxiola, M. Selman, and T. D. Randall. 2006. Inducible bronchus-associated lymphoid tissue (iBALT) in patients with pulmonary complications of rheumatoid arthritis. J Clin Invest 116:3183-3194. 101. Okuda, K., R. Kimizuka, S. Abe, T. Kato, and K. Ishihara. 2005. Involvement of periodontopathic anaerobes in aspiration pneumonia. J Periodontol 76:2154- 2160. 102. van der Gast, C. J., A. W. Walker, F. A. Stressmann, G. B. Rogers, P. Scott, T. W. Daniels, M. P. Carroll, J. Parkhill, and K. D. Bruce. 2010. Partitioning core and satellite taxa from within cystic fibrosis lung bacterial communities. ISME J 5:780-791. 103. Shirai, H., T. L. Blundell, and K. Mizuguchi. 2001. A novel superfamily of enzymes that catalyze the modification of guanidino groups. Trends Biochem Sci 26:465-468. 104. Guerrin, M., A. Ishigami, M. C. Mechin, R. Nachat, S. Valmary, M. Sebbag, M. Simon, T. Senshu, and G. Serre. 2003. cDNA cloning, gene organization and expression analysis of human peptidylarginine deiminase type I. Biochem J 370:167-174. 120 105. Kanno, T., A. Kawada, J. Yamanouchi, C. Yosida-Noro, A. Yoshiki, M. Shiraiwa, M. Kusakabe, M. Manabe, T. Tezuka, and H. Takahara. 2000. Human peptidylarginine deiminase type III: molecular cloning and nucleotide sequence of the cDNA, properties of the recombinant enzyme, and immunohistochemical localization in human skin. J Invest Dermatol 115:813-823. 106. Kubilus, J., R. F. Waitkus, and H. P. Baden. 1980. Partial purification and specificity of an arginine-converting enzyme from bovine epidermis. Biochim Biophys Acta 615:246-251. 107. Rogers, G., B. Winter, C. McLaughlan, B. Powell, and T. Nesci. 1997. Peptidylarginine deiminase of the hair follicle: characterization, localization, and function in keratinizing tissues. J Invest Dermatol 108:700-707. 108. Wright, P. W., L. C. Bolling, M. E. Calvert, O. F. Sarmento, E. V. Berkeley, M. C. Shea, Z. Hao, F. C. Jayes, L. A. Bush, J. Shetty, A. N. Shore, P. P. Reddi, K. S. Tung, E. Samy, M. M. Allietta, N. E. Sherman, J. C. Herr, and S. A. Coonrod. 2003. ePAD, an oocyte and early embryo-abundant peptidylarginine deiminase- like protein that localizes to egg cytoplasmic sheets. Dev Biol 256:73-88. 109. Chang, X., and J. Han. 2006. Expression of peptidylarginine deiminase type 4 (PAD4) in various tumors. Mol Carcinog 45:183-196. 110. Cuthbert, G. L., S. Daujat, A. W. Snowden, H. Erdjument-Bromage, T. Hagiwara, M. Yamada, R. Schneider, P. D. Gregory, P. Tempst, A. J. Bannister, and T. Kouzarides. 2004. Histone deimination antagonizes arginine methylation. Cell 118:545-553. 111. Liu, G. Y., Y. F. Liao, W. H. Chang, C. C. Liu, M. C. Hsieh, P. C. Hsu, G. J. Tsay, and H. C. Hung. 2006. Overexpression of peptidylarginine deiminase IV features in apoptosis of haematopoietic cells. Apoptosis 11:183-196. 112. Wang, Y., J. Wysocka, J. Sayegh, Y. H. Lee, J. R. Perlin, L. Leonelli, L. S. Sonbuchner, C. H. McDonald, R. G. Cook, Y. Dou, R. G. Roeder, S. Clarke, M. R. Stallcup, C. D. Allis, and S. A. Coonrod. 2004. Human PAD4 regulates histone arginine methylation levels via demethylimination. Science 306:279-283. 113. Asaga, H., M. Yamada, and T. Senshu. 1998. Selective deimination of vimentin in calcium ionophore-induced apoptosis of mouse peritoneal macrophages. Biochem Biophys Res Commun 243:641-646. 114. Moscarello, M. A., L. Pritzker, F. G. Mastronardi, and D. D. Wood. 2002. Peptidylarginine deiminase: a candidate factor in demyelinating disease. J Neurochem 81:335-343. 121 115. Watanabe, K., and T. Senshu. 1989. Isolation and characterization of cDNA clones encoding rat skeletal muscle peptidylarginine deiminase. J Biol Chem 264:15255-15260. 116. Vossenaar, E. R., A. J. Zendman, W. J. van Venrooij, and G. J. Pruijn. 2003. PAD, a growing family of citrullinating enzymes: genes, features and involvement in disease. Bioessays 25:1106-1118. 117. Mastronardi, F. G., D. D. Wood, J. Mei, R. Raijmakers, V. Tseveleki, H. M. Dosch, L. Probert, P. Casaccia-Bonnefil, and M. A. Moscarello. 2006. Increased citrullination of histone H3 in multiple sclerosis brain and animal models of demyelination: a role for tumor necrosis factor-induced peptidylarginine deiminase 4 translocation. J Neurosci 26:11387-11396. 118. Hagiwara, T., Y. Hidaka, and M. Yamada. 2005. Deimination of histone H2A and H4 at arginine 3 in HL-60 granulocytes. Biochemistry 44:5827-5834. 119. Nakashima, K., T. Hagiwara, and M. Yamada. 2002. Nuclear localization of peptidylarginine deiminase V and histone deimination in granulocytes. J Biol Chem 277:49562-49568. 120. De Rycke, L., A. P. Nicholas, T. Cantaert, E. Kruithof, J. D. Echols, B. Vandekerckhove, E. M. Veys, F. De Keyser, and D. Baeten. 2005. Synovial intracellular citrullinated proteins colocalizing with peptidyl arginine deiminase as pathophysiologically relevant antigenic determinants of rheumatoid arthritis- specific humoral autoimmunity. Arthritis Rheum 52:2323-2330. 121. Arita, K., H. Hashimoto, T. Shimizu, K. Nakashima, M. Yamada, and M. Sato. 2004. Structural basis for Ca(2+)-induced activation of human PAD4. Nat Struct Mol Biol 11:777-783. 122. Arita, K., T. Shimizu, H. Hashimoto, Y. Hidaka, M. Yamada, and M. Sato. 2006. Structural basis for histone N-terminal recognition by human peptidylarginine deiminase 4. Proc Natl Acad Sci U S A 103:5291-5296. 123. Hidenari Takahara, H. O. a. K. S. 1986. Calcium-dependent Properties of Peptidylarginine Deiminase from Rabbit Skeletal Muscle. Agric. Biol. Chem 50:2899-2904. 124. Kearney, P. L., M. Bhatia, N. G. Jones, L. Yuan, M. C. Glascock, K. L. Catchings, M. Yamada, and P. R. Thompson. 2005. Kinetic characterization of protein arginine deiminase 4: a transcriptional corepressor implicated in the onset and progression of rheumatoid arthritis. Biochemistry 44:10570-10582. 125. Knuckley, B., M. Bhatia, and P. R. Thompson. 2007. Protein arginine deiminase 4: evidence for a reverse protonation mechanism. Biochemistry 46:6578-6587. 122 126. Luo, Y., K. Arita, M. Bhatia, B. Knuckley, Y. H. Lee, M. R. Stallcup, M. Sato, and P. R. Thompson. 2006. Inhibitors and inactivators of protein arginine deiminase 4: functional and structural characterization. Biochemistry 45:11727- 11736. 127. Wang, Y., M. Li, S. Stadler, S. Correll, P. Li, D. Wang, R. Hayama, L. Leonelli, H. Han, S. A. Grigoryev, C. D. Allis, and S. A. Coonrod. 2009. Histone hypercitrullination mediates chromatin decondensation and neutrophil extracellular trap formation. J Cell Biol 184:205-213. 128. Ireland, J., J. Herzog, and E. R. Unanue. 2006. Cutting edge: unique T cells that recognize citrullinated peptides are a feature of protein immunization. J Immunol 177:1421-1425. 129. Ireland, J. M., and E. R. Unanue. 2011. Autophagy in antigen-presenting cells results in presentation of citrullinated peptides to CD4 T cells. J Exp Med 208:2625-2632. 130. Esposito, G., A. M. Vitale, F. P. Leijten, A. M. Strik, A. M. Koonen-Reemst, P. Yurttas, T. J. Robben, S. Coonrod, and J. A. Gossen. 2007. Peptidylarginine deiminase (PAD) 6 is essential for oocyte cytoskeletal sheet formation and female fertility. Mol Cell Endocrinol 273:25-31. 131. Nakashima, K., T. Hagiwara, A. Ishigami, S. Nagata, H. Asaga, M. Kuramoto, T. Senshu, and M. Yamada. 1999. Molecular characterization of peptidylarginine deiminase in HL-60 cells induced by retinoic acid and 1alpha,25- dihydroxyvitamin D(3). J Biol Chem 274:27786-27792. 132. Makrygiannakis, D., E. af Klint, I. E. Lundberg, R. Lofberg, A. K. Ulfgren, L. Klareskog, and A. I. Catrina. 2006. Citrullination is an inflammation-dependent process. Ann Rheum Dis 65:1219-1222. 133. Candi, E., R. Schmidt, and G. Melino. 2005. The cornified envelope: a model of cell death in the skin. Nat Rev Mol Cell Biol 6:328-340. 134. Loos, T., A. Mortier, M. Gouwy, I. Ronsse, W. Put, J. P. Lenaerts, J. Van Damme, and P. Proost. 2008. Citrullination of CXCL10 and CXCL11 by peptidylarginine deiminase: a naturally occurring posttranslational modification of chemokines and new dimension of immunoregulation. Blood 112:2648-2656. 135. Proost, P., T. Loos, A. Mortier, E. Schutyser, M. Gouwy, S. Noppen, C. Dillen, I. Ronsse, R. Conings, S. Struyf, G. Opdenakker, P. C. Maudgal, and J. Van Damme. 2008. Citrullination of CXCL8 by peptidylarginine deiminase alters receptor usage, prevents proteolysis, and dampens tissue inflammation. J Exp Med 205:2085-2097. 123 136. Struyf, S., S. Noppen, T. Loos, A. Mortier, M. Gouwy, H. Verbeke, D. Huskens, S. Luangsay, M. Parmentier, K. Geboes, D. Schols, J. Van Damme, and P. Proost. 2009. Citrullination of CXCL12 differentially reduces CXCR4 and CXCR7 binding with loss of inflammatory and anti-HIV-1 activity via CXCR4. J Immunol 182:666-674. 137. Lange, S., S. Gogel, K. Y. Leung, B. Vernay, A. P. Nicholas, C. P. Causey, P. R. Thompson, N. D. Greene, and P. Ferretti. 2011. Protein deiminases: new players in the developmentally regulated loss of neural regenerative ability. Dev Biol 355:205-214. 138. Chumanevich, A. A., C. P. Causey, B. A. Knuckley, J. E. Jones, D. Poudyal, A. P. Chumanevich, T. Davis, L. E. Matesic, P. R. Thompson, and L. J. Hofseth. 2011. Suppression of colitis in mice by Cl-amidine: a novel peptidylarginine deiminase inhibitor. Am J Physiol Gastrointest Liver Physiol 300:G929-938. 139. Guo, Q., and W. Fast. 2011. Citrullination of inhibitor of growth 4 (ING4) by peptidylarginine deminase 4 (PAD4) disrupts the interaction between ING4 and p53. J Biol Chem 286:17069-17078. 140. Nakayama-Hamada, M., A. Suzuki, H. Furukawa, R. Yamada, and K. Yamamoto. 2008. Citrullinated fibrinogen inhibits thrombin-catalysed fibrin polymerization. J Biochem 144:393-398. 141. Li, P., H. Yao, Z. Zhang, M. Li, Y. Luo, P. R. Thompson, D. S. Gilmour, and Y. Wang. 2008. Regulation of p53 target gene expression by peptidylarginine deiminase 4. Mol Cell Biol 28:4745-4758. 142. Yao, H., P. Li, B. J. Venters, S. Zheng, P. R. Thompson, B. F. Pugh, and Y. Wang. 2008. Histone Arg modifications and p53 regulate the expression of OKL38, a mediator of apoptosis. J Biol Chem 283:20060-20068. 143. Andrade, F., E. Darrah, M. Gucek, R. N. Cole, A. Rosen, and X. Zhu. 2010. Autocitrullination of human peptidylarginine deiminase 4 regulates protein citrullination during cell activation. Arthritis Rheum. 144. Slack, J. L., L. E. Jones, M. M. Bhatia, and P. R. Thompson. 2011. Autodeimination of Protein Arginine Deiminase 4 Alters Protein-Protein Interactions but Not Activity. Biochemistry. 145. Raijmakers, R., J. Vogelzangs, J. Raats, M. Panzenbeck, M. Corby, H. Jiang, M. Thibodeau, N. Haynes, W. J. van Venrooij, G. J. Pruijn, and B. Werneburg. 2006. Experimental autoimmune encephalomyelitis induction in peptidylarginine deiminase 2 knockout mice. J Comp Neurol 498:217-226. 124 146. Rohrbach, A. S., S. Hemmers, S. Arandjelovic, M. Corr, and K. A. Mowen. 2012. PAD4 is not essential for disease in the K/BxN murine autoantibody-mediated model of arthritis. Arthritis Res Ther 14:R104. 147. Avouac, J., L. Gossec, and M. Dougados. 2006. Diagnostic and predictive value of anti-cyclic citrullinated protein antibodies in rheumatoid arthritis: a systematic literature review. Ann Rheum Dis 65:845-851. 148. Bizzaro, N., G. Mazzanti, E. Tonutti, D. Villalta, and R. Tozzoli. 2001. Diagnostic accuracy of the anti-citrulline antibody assay for rheumatoid arthritis. Clin Chem 47:1089-1093. 149. Goldbach-Mansky, R., J. Lee, A. McCoy, J. Hoxworth, C. Yarboro, J. S. Smolen, G. Steiner, A. Rosen, C. Zhang, H. A. Menard, Z. J. Zhou, T. Palosuo, W. J. Van Venrooij, R. L. Wilder, J. H. Klippel, H. R. Schumacher, Jr., and H. S. El- Gabalawy. 2000. Rheumatoid arthritis associated autoantibodies in patients with synovitis of recent onset. Arthritis Res 2:236-243. 150. Plant, D., W. Thomson, M. Lunt, E. Flynn, P. Martin, S. Eyre, T. Farragher, D. Bunn, J. Worthington, D. Symmons, and A. Barton. 2010. The role of rheumatoid arthritis genetic susceptibility markers in the prediction of erosive disease in patients with early inflammatory polyarthritis: results from the Norfolk Arthritis Register. Rheumatology (Oxford) 50:78-84. 151. Bukhari, M., W. Thomson, H. Naseem, D. Bunn, A. Silman, D. Symmons, and A. Barton. 2007. The performance of anti-cyclic citrullinated peptide antibodies in predicting the severity of radiologic damage in inflammatory polyarthritis: results from the Norfolk Arthritis Register. Arthritis Rheum 56:2929-2935. 152. Ohmura, K., C. Terao, E. Maruya, M. Katayama, K. Matoba, K. Shimada, A. Murasawa, S. Honjo, K. Takasugi, S. Tohma, K. Matsuo, K. Tajima, N. Yukawa, D. Kawabata, T. Nojima, T. Fujii, R. Yamada, H. Saji, F. Matsuda, and T. Mimori. 2010. Anti-citrullinated peptide antibody-negative RA is a genetically distinct subset: a definitive study using only bone-erosive ACPA-negative rheumatoid arthritis. Rheumatology (Oxford) 49:2298-2304. 153. Kuhn, K. A., L. Kulik, B. Tomooka, K. J. Braschler, W. P. Arend, W. H. Robinson, and V. M. Holers. 2006. Antibodies against citrullinated proteins enhance tissue injury in experimental autoimmune arthritis. J Clin Invest 116:961- 973. 154. Hill, J. A., D. A. Bell, W. Brintnell, D. Yue, B. Wehrli, A. M. Jevnikar, D. M. Lee, W. Hueber, W. H. Robinson, and E. Cairns. 2008. Arthritis induced by posttranslationally modified (citrullinated) fibrinogen in DR4-IE transgenic mice. J Exp Med 205:967-979. 125 155. Uysal, H., R. Bockermann, K. S. Nandakumar, B. Sehnert, E. Bajtner, A. Engstrom, G. Serre, H. Burkhardt, M. M. Thunnissen, and R. Holmdahl. 2009. Structure and pathogenicity of antibodies specific for citrullinated collagen type II in experimental arthritis. J Exp Med. 156. Clavel, C., L. Nogueira, L. Laurent, C. Iobagiu, C. Vincent, M. Sebbag, and G. Serre. 2008. Induction of macrophage secretion of tumor necrosis factor alpha through Fcgamma receptor IIa engagement by rheumatoid arthritis-specific autoantibodies to citrullinated proteins complexed with fibrinogen. Arthritis Rheum 58:678-688. 157. Trouw, L. A., E. M. Haisma, E. W. Levarht, D. van der Woude, A. Ioan-Facsinay, M. R. Daha, T. W. Huizinga, and R. E. Toes. 2009. Anti-cyclic citrullinated peptide antibodies from rheumatoid arthritis patients activate complement via both the classical and alternative pathways. Arthritis Rheum 60:1923-1931. 158. van der Woude, D., S. Rantapaa-Dahlqvist, A. Ioan-Facsinay, C. Onnekink, C. M. Schwarte, K. N. Verpoort, J. W. Drijfhout, T. W. Huizinga, R. E. Toes, and G. J. Pruijn. 2010. Epitope spreading of the anti-citrullinated protein antibody response occurs before disease onset and is associated with the disease course of early arthritis. Ann Rheum Dis 69:1554-1561. 159. van de Stadt, L. A., M. H. de Koning, R. J. van de Stadt, G. Wolbink, B. A. Dijkmans, D. Hamann, and D. van Schaardenburg. Development of the anti- citrullinated protein antibody repertoire prior to the onset of rheumatoid arthritis. Arthritis Rheum 63:3226-3233. 160. Suwannalai, P., L. A. van de Stadt, H. Radner, G. Steiner, H. S. El-Gabalawy, C. M. Zijde, M. J. van Tol, D. van Schaardenburg, T. W. Huizinga, R. E. Toes, and L. A. Trouw. 2011. Avidity maturation of anti-citrullinated protein antibodies in rheumatoid arthritis. Arthritis Rheum 64:1323-1328. 161. Van Steendam, K., K. Tilleman, and D. Deforce. The relevance of citrullinated vimentin in the production of antibodies against citrullinated proteins and the pathogenesis of rheumatoid arthritis. Rheumatology (Oxford) 50:830-837. 162. Burkhardt, H., T. Koller, A. Engstrom, K. S. Nandakumar, J. Turnay, H. G. Kraetsch, J. R. Kalden, and R. Holmdahl. 2002. Epitope-specific recognition of type II collagen by rheumatoid arthritis antibodies is shared with recognition by antibodies that are arthritogenic in collagen-induced arthritis in the mouse. Arthritis Rheum 46:2339-2348. 163. Lundberg, K., N. Wegner, T. Yucel-Lindberg, and P. J. Venables. 2010. Periodontitis in RA-the citrullinated enolase connection. Nat Rev Rheumatol 6:727-730. 126 164. Girbal-Neuhauser, E., J. J. Durieux, M. Arnaud, P. Dalbon, M. Sebbag, C. Vincent, M. Simon, T. Senshu, C. Masson-Bessiere, C. Jolivet-Reynaud, M. Jolivet, and G. Serre. 1999. The epitopes targeted by the rheumatoid arthritis- associated antifilaggrin autoantibodies are posttranslationally generated on various sites of (pro)filaggrin by deimination of arginine residues. J Immunol 162:585-594. 165. Masson-Bessiere, C., M. Sebbag, E. Girbal-Neuhauser, L. Nogueira, C. Vincent, T. Senshu, and G. Serre. 2001. The major synovial targets of the rheumatoid arthritis-specific antifilaggrin autoantibodies are deiminated forms of the alpha- and beta-chains of fibrin. J Immunol 166:4177-4184. 166. Deane, K. D., C. I. O'Donnell, W. Hueber, D. S. Majka, A. A. Lazar, L. A. Derber, W. R. Gilliland, J. D. Edison, J. M. Norris, W. H. Robinson, and V. M. Holers. 2010. The number of elevated cytokines and chemokines in preclinical seropositive rheumatoid arthritis predicts time to diagnosis in an age-dependent manner. Arthritis Rheum 62:3161-3172. 167. Holers, V. M. 2008. The spectrum of complement alternative pathway-mediated diseases. Immunol Rev 223:300-316. 168. Silverman, G. J., and D. A. Carson. 2003. Roles of B cells in rheumatoid arthritis. Arthritis Res Ther 5 Suppl 4:S1-6. 169. Kurosaka, M., and M. Ziff. 1983. Immunoelectron microscopic study of the distribution of T cell subsets in rheumatoid synovium. J Exp Med 158:1191-1210. 170. Hueber, A. J., D. L. Asquith, A. M. Miller, J. Reilly, S. Kerr, J. Leipe, A. J. Melendez, and I. B. McInnes. 2010. Mast cells express IL-17A in rheumatoid arthritis synovium. J Immunol 184:3336-3340. 171. Nigrovic, P. A., B. A. Binstadt, P. A. Monach, A. Johnsen, M. Gurish, Y. Iwakura, C. Benoist, D. Mathis, and D. M. Lee. 2007. Mast cells contribute to initiation of autoantibody-mediated arthritis via IL-1. Proc Natl Acad Sci U S A 104:2325-2330. 172. Shin, K., P. A. Nigrovic, J. Crish, E. Boilard, H. P. McNeil, K. S. Larabee, R. Adachi, M. F. Gurish, R. Gobezie, R. L. Stevens, and D. M. Lee. 2009. Mast cells contribute to autoimmune inflammatory arthritis via their tryptase/heparin complexes. J Immunol 182:647-656. 173. Hamilton, J. A., and P. P. Tak. 2009. The dynamics of macrophage lineage populations in inflammatory and autoimmune diseases. Arthritis Rheum 60:1210- 1221. 127 174. Foulquier, C., M. Sebbag, C. Clavel, S. Chapuy-Regaud, R. Al Badine, M. C. Mechin, C. Vincent, R. Nachat, M. Yamada, H. Takahara, M. Simon, M. Guerrin, and G. Serre. 2007. Peptidyl arginine deiminase type 2 (PAD-2) and PAD-4 but not PAD-1, PAD-3, and PAD-6 are expressed in rheumatoid arthritis synovium in close association with tissue inflammation. Arthritis Rheum 56:3541-3553. 175. Kinloch, A., K. Lundberg, R. Wait, N. Wegner, N. H. Lim, A. J. Zendman, T. Saxne, V. Malmstrom, and P. J. Venables. 2008. Synovial fluid is a site of citrullination of autoantigens in inflammatory arthritis. Arthritis Rheum 58:2287- 2295. 176. Vossenaar, E. R., S. Nijenhuis, M. M. Helsen, A. van der Heijden, T. Senshu, W. B. van den Berg, W. J. van Venrooij, and L. A. Joosten. 2003. Citrullination of synovial proteins in murine models of rheumatoid arthritis. Arthritis Rheum 48:2489-2500. 177. Luo, Y., B. Knuckley, M. Bhatia, P. J. Pellechia, and P. R. Thompson. 2006. Activity-based protein profiling reagents for protein arginine deiminase 4 (PAD4): synthesis and in vitro evaluation of a fluorescently labeled probe. J Am Chem Soc 128:14468-14469. 178. Luo, Y., B. Knuckley, Y. H. Lee, M. R. Stallcup, and P. R. Thompson. 2006. A fluoroacetamidine-based inactivator of protein arginine deiminase 4: design, synthesis, and in vitro and in vivo evaluation. J Am Chem Soc 128:1092-1093. 179. Stone, E. M., T. H. Schaller, H. Bianchi, M. D. Person, and W. Fast. 2005. Inactivation of two diverse enzymes in the amidinotransferase superfamily by 2- chloroacetamidine: dimethylargininase and peptidylarginine deiminase. Biochemistry 44:13744-13752. 180. Li, P., M. Li, M. R. Lindberg, M. J. Kennett, N. Xiong, and Y. Wang. 2010. PAD4 is essential for antibacterial innate immunity mediated by neutrophil extracellular traps. J Exp Med. 181. Li, P., D. Wang, H. Yao, P. Doret, G. Hao, Q. Shen, H. Qiu, X. Zhang, Y. Wang, and G. Chen. 2010. Coordination of PAD4 and HDAC2 in the regulation of p53- target gene expression. Oncogene 29:3153-3162. 182. Kannan, K., R. A. Ortmann, and D. Kimpel. 2005. Animal models of rheumatoid arthritis and their relevance to human disease. Pathophysiology 12:167-181. 183. Courtenay, J. S., M. J. Dallman, A. D. Dayan, A. Martin, and B. Mosedale. 1980. Immunisation against heterologous type II collagen induces arthritis in mice. Nature 283:666-668. 184. Joe, B., M. M. Griffiths, E. F. Remmers, and R. L. Wilder. 1999. Animal models of rheumatoid arthritis and related inflammation. Curr Rheumatol Rep 1:139-148. 128 185. Watson, W. C., P. S. Brown, J. A. Pitcock, and A. S. Townes. 1987. Passive transfer studies with type II collagen antibody in B10.D2/old and new line and C57Bl/6 normal and beige (Chediak-Higashi) strains: evidence of important roles for C5 and multiple inflammatory cell types in the development of erosive arthritis. Arthritis Rheum 30:460-465. 186. Corthay, A., A. Johansson, M. Vestberg, and R. Holmdahl. 1999. Collagen- induced arthritis development requires alpha beta T cells but not gamma delta T cells: studies with T cell-deficient (TCR mutant) mice. Int Immunol 11:1065- 1073. 187. Ranges, G. E., S. Sriram, and S. M. Cooper. 1985. Prevention of type II collagen- induced arthritis by in vivo treatment with anti-L3T4. J Exp Med 162:1105-1110. 188. Terato, K., D. S. Harper, M. M. Griffiths, D. L. Hasty, X. J. Ye, M. A. Cremer, and J. M. Seyer. 1995. Collagen-induced arthritis in mice: synergistic effect of E. coli lipopolysaccharide bypasses epitope specificity in the induction of arthritis with monoclonal antibodies to type II collagen. Autoimmunity 22:137-147. 189. Banda, N. K., D. Kraus, A. Vondracek, L. H. Huynh, A. Bendele, V. M. Holers, and W. P. Arend. 2002. Mechanisms of effects of complement inhibition in murine collagen-induced arthritis. Arthritis Rheum 46:3065-3075. 190. Banda, N. K., J. M. Thurman, D. Kraus, A. Wood, M. C. Carroll, W. P. Arend, and V. M. Holers. 2006. Alternative complement pathway activation is essential for inflammation and joint destruction in the passive transfer model of collagen- induced arthritis. J Immunol 177:1904-1912. 191. Pritzker, L. B., T. A. Nguyen, and M. A. Moscarello. 1999. The developmental expression and activity of peptidylarginine deiminase in the mouse. Neurosci Lett 266:161-164. 192. Knipp, M., and M. Vasak. 2000. A colorimetric 96-well microtiter plate assay for the determination of enzymatically formed citrulline. Anal Biochem 286:257-264. 193. Hueber, W., B. A. Kidd, B. H. Tomooka, B. J. Lee, B. Bruce, J. F. Fries, G. Sonderstrup, P. Monach, J. W. Drijfhout, W. J. van Venrooij, P. J. Utz, M. C. Genovese, and W. H. Robinson. 2005. Antigen microarray profiling of autoantibodies in rheumatoid arthritis. Arthritis Rheum 52:2645-2655. 194. Robinson, W. H., C. DiGennaro, W. Hueber, B. B. Haab, M. Kamachi, E. J. Dean, S. Fournel, D. Fong, M. C. Genovese, H. E. de Vegvar, K. Skriner, D. L. Hirschberg, R. I. Morris, S. Muller, G. J. Pruijn, W. J. van Venrooij, J. S. Smolen, P. O. Brown, L. Steinman, and P. J. Utz. 2002. Autoantigen microarrays for multiplex characterization of autoantibody responses. Nat Med 8:295-301. 129 195. Tusher, V. G., R. Tibshirani, and G. Chu. 2001. Significance analysis of microarrays applied to the ionizing radiation response. Proc Natl Acad Sci U S A 98:5116-5121. 196. Eisen, M. B., P. T. Spellman, P. O. Brown, and D. Botstein. 1998. Cluster analysis and display of genome-wide expression patterns. Proc Natl Acad Sci U S A 95:14863-14868. 197. Takemoto, K., K. Ogino, M. Shibamori, T. Gondo, Y. Hitomi, T. Takigawa, D. H. Wang, J. Takaki, H. Ichimura, Y. Fujikura, and H. Ishiyama. 2007. Transiently, paralleled upregulation of arginase and nitric oxide synthase and the effect of both enzymes on the pathology of asthma. Am J Physiol Lung Cell Mol Physiol 293:L1419-1426. 198. Kolfenbach, J. R., K. D. Deane, L. A. Derber, C. O'Donnell, M. H. Weisman, J. H. Buckner, V. H. Gersuk, S. Wei, T. R. Mikuls, J. O'Dell, P. K. Gregersen, R. M. Keating, J. M. Norris, and V. M. Holers. 2009. A prospective approach to investigating the natural history of preclinical rheumatoid arthritis (RA) using first-degree relatives of probands with RA. Arthritis Rheum 61:1735-1742. 199. Arnett, F. C., S. M. Edworthy, D. A. Bloch, D. J. McShane, J. F. Fries, N. S. Cooper, L. A. Healey, S. R. Kaplan, M. H. Liang, H. S. Luthra, and et al. 1988. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 31:315-324. 200. Pavord, I. D., M. M. Pizzichini, E. Pizzichini, and F. E. Hargreave. 1997. The use of induced sputum to investigate airway inflammation. Thorax 52:498-501. 201. Kidd, B. A., P. P. Ho, O. Sharpe, X. Zhao, B. H. Tomooka, J. L. Kanter, L. Steinman, and W. H. Robinson. 2008. Epitope spreading to citrullinated antigens in mouse models of autoimmune arthritis and demyelination. Arthritis Res Ther 10:R119. 202. Watson, W. C., and A. S. Townes. 1985. Genetic susceptibility to murine collagen II autoimmune arthritis. Proposed relationship to the IgG2 autoantibody subclass response, complement C5, major histocompatibility complex (MHC) and non- MHC loci. J Exp Med 162:1878-1891. 203. Wooley, P. H., H. S. Luthra, J. M. Stuart, and C. S. David. 1981. Type II collagen-induced arthritis in mice. I. Major histocompatibility complex (I region) linkage and antibody correlates. J Exp Med 154:688-700. 204. Benito-Miguel, M., Y. Garcia-Carmona, A. Balsa, C. Perez de Ayala, T. Cobo- Ibanez, E. Martin-Mola, and M. E. Miranda-Carus. 2009. A dual action of rheumatoid arthritis synovial fibroblast IL-15 expression on the equilibrium between CD4+CD25+ regulatory T cells and CD4+CD25- responder T cells. J Immunol 183:8268-8279. 130 205. Nistala, K., and L. R. Wedderburn. 2009. Th17 and regulatory T cells: rebalancing pro- and anti-inflammatory forces in autoimmune arthritis. Rheumatology (Oxford) 48:602-606. 206. Ehrenstein, M. R., J. G. Evans, A. Singh, S. Moore, G. Warnes, D. A. Isenberg, and C. Mauri. 2004. Compromised function of regulatory T cells in rheumatoid arthritis and reversal by anti-TNFalpha therapy. J Exp Med 200:277-285. 207. Meiron, M., Y. Zohar, R. Anunu, G. Wildbaum, and N. Karin. 2008. CXCL12 (SDF-1alpha) suppresses ongoing experimental autoimmune encephalomyelitis by selecting antigen-specific regulatory T cells. J Exp Med 205:2643-2655. 208. Scott, G. N., J. DuHadaway, E. Pigott, N. Ridge, G. C. Prendergast, A. J. Muller, and L. Mandik-Nayak. 2009. The immunoregulatory enzyme IDO paradoxically drives B cell-mediated autoimmunity. J Immunol 182:7509-7517. 209. Tanikawa, C., M. Espinosa, A. Suzuki, K. Masuda, K. Yamamoto, E. Tsuchiya, K. Ueda, Y. Daigo, Y. Nakamura, and K. Matsuda. 2012. Regulation of histone modification and chromatin structure by the p53-PADI4 pathway. Nat Commun 3:676. 210. Auger, I., M. Martin, N. Balandraud, and J. Roudier. 2009. Rheumatoid arthritis- specific autoantibodies to peptidyl arginine deiminase type 4 inhibit citrullination of fibrinogen. Arthritis Rheum 62:126-131. 211. Halvorsen, E. H., E. A. Haavardsholm, S. Pollmann, A. Boonen, D. van der Heijde, T. K. Kvien, and O. Molberg. 2009. Serum IgG antibodies to peptidylarginine deiminase 4 predict radiographic progression in patients with rheumatoid arthritis treated with tumour necrosis factor-alpha blocking agents. Ann Rheum Dis 68:249-252. 212. Halvorsen, E. H., S. Pollmann, I. M. Gilboe, D. van der Heijde, R. Landewe, S. Odegard, T. K. Kvien, and O. Molberg. 2008. Serum IgG antibodies to peptidylarginine deiminase 4 in rheumatoid arthritis and associations with disease severity. Ann Rheum Dis 67:414-417. 213. Harris, M. L., E. Darrah, G. K. Lam, S. J. Bartlett, J. T. Giles, A. V. Grant, P. Gao, W. W. Scott, Jr., H. El-Gabalawy, L. Casciola-Rosen, K. C. Barnes, J. M. Bathon, and A. Rosen. 2008. Association of autoimmunity to peptidyl arginine deiminase type 4 with genotype and disease severity in rheumatoid arthritis. Arthritis Rheum 58:1958-1967. 214. Zhao, J., Y. Zhao, J. He, R. Jia, and Z. Li. 2008. Prevalence and significance of anti-peptidylarginine deiminase 4 antibodies in rheumatoid arthritis. J Rheumatol 35:969-974. 131 215. Willis, V. C., A. M. Gizinski, N. K. Banda, C. P. Causey, B. Knuckley, K. N. Cordova, Y. Luo, B. Levitt, M. Glogowska, P. Chandra, L. Kulik, W. H. Robinson, W. P. Arend, P. R. Thompson, and V. M. Holers. 2011. N-{alpha}- Benzoyl-N5-(2-Chloro-1-Iminoethyl)-L-Ornithine Amide, a Protein Arginine Deiminase Inhibitor, Reduces the Severity of Murine Collagen-Induced Arthritis. J Immunol 186:4396-4404. 216. Simelyte, E., S. Rosengren, D. L. Boyle, M. Corr, D. R. Green, and G. S. Firestein. 2005. Regulation of arthritis by p53: critical role of adaptive immunity. Arthritis Rheum 52:1876-1884. 217. Yamanishi, Y., D. L. Boyle, M. J. Pinkoski, A. Mahboubi, T. Lin, Z. Han, N. J. Zvaifler, D. R. Green, and G. S. Firestein. 2002. Regulation of joint destruction and inflammation by p53 in collagen-induced arthritis. Am J Pathol 160:123-130. 218. Tang, B. X., X. You, L. D. Zhao, Y. Li, X. Zhang, F. L. Tang, D. N. Ba, and W. He. 2011. p53 in fibroblast-like synoviocytes can regulate T helper cell functions in patients with active rheumatoid arthritis. Chin Med J (Engl) 124:364-368. 219. Okroj, M., D. Heinegard, R. Holmdahl, and A. M. Blom. 2007. Rheumatoid arthritis and the complement system. Ann Med 39:517-530. 220. Banda, N. K., K. Takahashi, A. K. Wood, V. M. Holers, and W. P. Arend. 2007. Pathogenic complement activation in collagen antibody-induced arthritis in mice requires amplification by the alternative pathway. J Immunol 179:4101-4109. 221. Arandjelovic, S., K. R. McKenney, S. S. Leming, and K. A. Mowen. 2012. ATP Induces Protein Arginine Deiminase 2-Dependent Citrullination in Mast Cells through the P2X7 Purinergic Receptor. J Immunol. 222. Vossenaar, E. R., T. R. Radstake, A. van der Heijden, M. A. van Mansum, C. Dieteren, D. J. de Rooij, P. Barrera, A. J. Zendman, and W. J. van Venrooij. 2004. Expression and activity of citrullinating peptidylarginine deiminase enzymes in monocytes and macrophages. Ann Rheum Dis 63:373-381. 223. Hojo-Nakashima, I., R. Sato, K. Nakashima, T. Hagiwara, and M. Yamada. 2009. Dynamic expression of peptidylarginine deiminase 2 in human monocytic leukaemia THP-1 cells during macrophage differentiation. J Biochem 146:471- 479. 224. Lee, H. J., M. Joo, R. Abdolrasulnia, D. G. Young, I. Choi, L. B. Ware, T. S. Blackwell, and B. W. Christman. 2010. Peptidylarginine deiminase 2 suppresses inhibitory {kappa}B kinase activity in lipopolysaccharide-stimulated RAW 264.7 macrophages. J Biol Chem 285:39655-39662. 225. Kennedy, A., U. Fearon, D. J. Veale, and C. Godson. 2011. Macrophages in synovial inflammation. Front Immunol 2:52. 132 226. van de Sande, M. G., M. J. de Hair, C. van der Leij, P. L. Klarenbeek, W. H. Bos, M. D. Smith, M. Maas, N. de Vries, D. van Schaardenburg, B. A. Dijkmans, D. M. Gerlag, and P. P. Tak. 2011. Different stages of rheumatoid arthritis: features of the synovium in the preclinical phase. Ann Rheum Dis 70:772-777. 227. Kokkonen, H., I. Soderstrom, J. Rocklov, G. Hallmans, K. Lejon, and S. Rantapaa Dahlqvist. 2010. Up-regulation of cytokines and chemokines predates the onset of rheumatoid arthritis. Arthritis Rheum 62:383-391. 228. Demoruelle, M. K., M. H. Weisman, P. L. Simonian, D. A. Lynch, P. B. Sachs, I. F. Pedraza, A. R. Harrington, J. R. Kolfenbach, C. C. Striebich, Q. N. Pham, C. D. Strickland, B. D. Petersen, M. C. Parish, L. A. Derber, J. M. Norris, V. M. Holers, and K. D. Deane. 2012. Brief Report: Airways abnormalities and rheumatoid arthritis-related autoantibodies in subjects without arthritis: Early injury or initiating site of autoimmunity? Arthritis Rheum 64:1756-1761. 229. Schiotz, P. O., E. M. Egeskjold, N. Hoiby, and H. Permin. 1979. Autoantibodies in serum and sputum from patients with cystic fibrosis. Acta Pathol Microbiol Scand C 87:319-324. 230. Nickdel, M. B., P. Conigliaro, G. Valesini, S. Hutchison, R. Benson, R. V. Bundick, A. J. Leishman, I. B. McInnes, J. M. Brewer, and P. Garside. 2009. Dissecting the contribution of innate and antigen-specific pathways to the breach of self-tolerance observed in a murine model of arthritis. Ann Rheum Dis 68:1059-1066. 231. Carrillo-Vico, A., M. D. Leech, and S. M. Anderton. 2010. Contribution of myelin autoantigen citrullination to T cell autoaggression in the central nervous system. J Immunol 184:2839-2846. 232. Ho, P. P., L. Y. Lee, X. Zhao, B. H. Tomooka, R. T. Paniagua, O. Sharpe, M. J. Benbarak, P. E. Chandra, W. Hueber, L. Steinman, and W. H. Robinson. 2009. Autoimmunity against Fibrinogen Mediates Inflammatory Arthritis in Mice. J Immunol. 233. Thiele, G. M., M. J. Duryee, A. Dusad, C. D. Hunter, J. P. Lacy, D. R. Anderson, D. Wang, J. R. O'Dell, T. R. Mikuls, and L. W. Klassen. 2012. Citrullinated mouse collagen administered to DBA/1J mice in the absence of adjuvant initiates arthritis. Int Immunopharmacol 13:424-431. 234. Turunen, S., M. K. Koivula, L. Risteli, and J. Risteli. 2010. Anti-citrulline antibodies can be caused by homocitrulline-containing proteins in rabbits. Arthritis Rheum. 133 235. Shi, J., R. Knevel, P. Suwannalai, M. P. van der Linden, G. M. Janssen, P. A. van Veelen, N. E. Levarht, A. H. van der Helm-van Mil, A. Cerami, T. W. Huizinga, R. E. Toes, and L. A. Trouw. 2011. Autoantibodies recognizing carbamylated proteins are present in sera of patients with rheumatoid arthritis and predict joint damage. Proc Natl Acad Sci U S A 108:17372-17377. 236. Mydel, P., Z. Wang, M. Brisslert, A. Hellvard, L. E. Dahlberg, S. L. Hazen, and M. Bokarewa. 2010. Carbamylation-Dependent Activation of T Cells: A Novel Mechanism in the Pathogenesis of Autoimmune Arthritis. J Immunol. 237. Wang, Z., S. J. Nicholls, E. R. Rodriguez, O. Kummu, S. Horkko, J. Barnard, W. F. Reynolds, E. J. Topol, J. A. DiDonato, and S. L. Hazen. 2007. Protein carbamylation links inflammation, smoking, uremia and atherogenesis. Nat Med 13:1176-1184. 238. van de Stadt, L. A., M. H. de Koning, R. J. van de Stadt, G. Wolbink, B. A. Dijkmans, D. Hamann, and D. van Schaardenburg. 2011. Development of the anti-citrullinated protein antibody repertoire prior to the onset of rheumatoid arthritis. Arthritis Rheum 63:3226-3233. 239. Cherrington, B. D., X. Zhang, J. L. McElwee, E. Morency, L. J. Anguish, and S. A. Coonrod. 2012. Potential Role for PAD2 in Gene Regulation in Breast Cancer Cells. PLoS ONE 7:e41242. 240. Makrygiannakis, D., S. Revu, M. Engstrom, E. af Klint, A. P. Nicholas, G. J. Pruijn, and A. I. Catrina. 2012. Local administration of glucocorticoids decreases synovial citrullination in rheumatoid arthritis. Arthritis Res Ther 14:R20. 241. Katschke, K. J., Jr., K. Y. Helmy, M. Steffek, H. Xi, J. Yin, W. P. Lee, P. Gribling, K. H. Barck, R. A. Carano, R. E. Taylor, L. Rangell, L. Diehl, P. E. Hass, C. Wiesmann, and M. van Lookeren Campagne. 2007. A novel inhibitor of the alternative pathway of complement reverses inflammation and bone destruction in experimental arthritis. J Exp Med 204:1319-1325. 242. Macor, P., P. Durigutto, L. De Maso, C. Garrovo, S. Biffi, A. Cortini, F. Fischetti, D. Sblattero, C. Pitzalis, R. Marzari, and F. Tedesco. 2012. Treatment of experimental arthritis by targeting synovial endothelium with a neutralizing recombinant antibody to C5. Arthritis Rheum 64:2559-2567. 243. Garnier, G., A. Circolo, and H. R. Colten. 1996. Constitutive expression of murine complement factor B gene is regulated by the interaction of its upstream promoter with hepatocyte nuclear factor 4. J Biol Chem 271:30205-30211. 244. Sjoholm, A. G., K. Berglund, U. Johnson, A. B. Laurell, and G. Sturfelt. 1986. C1 activation, with C1q in excess of functional C1 in synovial fluid from patients with rheumatoid arthritis. Int Arch Allergy Appl Immunol 79:113-119. 134 245. Ignatius, A., P. Schoengraf, L. Kreja, A. Liedert, S. Recknagel, S. Kandert, R. E. Brenner, M. Schneider, J. D. Lambris, and M. Huber-Lang. 2011. Complement C3a and C5a modulate osteoclast formation and inflammatory response of osteoblasts in synergism with IL-1beta. J Cell Biochem 112:2594-2605. 246. Perlmutter, D. H., and H. R. Colten. 1986. Molecular immunobiology of complement biosynthesis: a model of single-cell control of effector-inhibitor balance. Annu Rev Immunol 4:231-251. 247. Kotowicz, K., R. E. Callard, K. Friedrich, D. J. Matthews, and N. Klein. 1996. Biological activity of IL-4 and IL-13 on human endothelial cells: functional evidence that both cytokines act through the same receptor. Int Immunol 8:1915- 1925. 248. Humby, F., M. Bombardieri, A. Manzo, S. Kelly, M. C. Blades, B. Kirkham, J. Spencer, and C. Pitzalis. 2009. Ectopic lymphoid structures support ongoing production of class-switched autoantibodies in rheumatoid synovium. PLoS medicine 6:e1. 249. Andoh, A., Y. Fujiyama, K. Kitoh, M. Niwakawa, K. Hodohara, T. Bamba, and S. Hosoda. 1993. Macrophage colony-stimulating factor (M-CSF) enhances complement component C3 production by human monocytes/macrophages. Int J Hematol 57:53-59. 250. El-Gabalawy, H. S., D. B. Robinson, I. Smolik, D. Hart, B. Elias, K. Wong, C. A. Peschken, C. A. Hitchon, X. Li, C. N. Bernstein, M. M. Newkirk, and M. J. Fritzler. 2012. Familial clustering of the serum cytokine profile in the relatives of rheumatoid arthritis patients. Arthritis Rheum. 135