Rheumatoid arthritis (RA) is a disorder comprising a

Transcription

1 Antiyli Citrullinated Protein Antibodies Are Impliated in the Development of Cardiovasular Disease in Rheumatoid Arthritis Nuria Barbarroja, Carlos Pérez-Sanhez, Patriia Ruiz-Limon, Carmen Castro-Villegas, Maria Angeles Aguirre, Rosario Carretero, Pedro Segui, Yolanda Jimenez-Gomez, Manuela Sanna, Antonio Rodriguez-Ariza, Eduardo Collantes-Estevez, Alejandro Esudero, Chary López-Pedrera Downloaded from by on Deember 27, 218 Objetive Previous studies have suggested a relationship between antiyli itrullinated protein (CCP) levels and development of ardiovasular disease in rheumatoid arthritis (RA). However, a limited number of studies have demonstrated an involvement of s in those proesses. This study was aimed to define the speifi role of these auto-antibodies in the pro-oxidative, inflammatory, and proatherogeni profile observed in leukoytes from. Approah and Results Seventy-five and 31 healthy donors were enrolled. Carotid intima media thikness was evaluated as atheroslerosis marker. Several prooagulant and inflammatory fators, leukoyte ativation, and oxidative stress markers were analyzed in plasma and leukoyte subsets. were purified from plasma of, and in vitro treatment of healthy leukoytes was onduted. High titers of s were assoiated to altered expression of prothromboti and inflammatory markers, high oxidative stress, and pathologial arotid intima media thikness in. Notably, gene expression analysis showed that lymphoytes were major players in altered inflammatory profile, monoytes were responsible for the protromboti and atherogeni status, and neutrophils mainly displayed a prooxidative feature. In vitro treatment with purified s fully reapitulated that pathogeni profile, promoting the ativation of leukoytes. Conlusions are key players in the inflammatory and proatherogeni status of. The effets are speifi of the immune ell targeted, promoting overexpression of thromboti, inflammatory, and pro-oxidative markers in monoytes; pro-oxidative status in neutrophils; and proinflammatory profile in lymphoytes. Targeting these autoantibodies would be an exellent strategy to prevent the development of ardiovasular disease in RA. (Arteriosler Thromb Vas Biol. 214;34: ) Key Words: ardiovasular disease inflammation leukoytes rheumatoid arthritis Rheumatoid arthritis (RA) is a disorder omprising a omplex onset mehanism and many assoiated ompliations. In partiular, ardiovasular disease (CVD) signifiantly ontributes to morbidity and mortality in these patients, ausing the 39% to % of deaths, 1 and atheroslerosis at early stage of the disease is onsidered a possible prelinial manifestation. In fat, the risk of CVD events, suh as myoardial infartion, is inreased in the 2 years preeding formal diagnosis of RA, 2 and one the disease is diagnosed, the risks of arotid plaques and CVD events inrease with the duration of RA. 3 Thus, the strong relationship between atheroslerosis and RA has led some authors to inlude atheroslerosis among the extra-artiular manifestations of the disease. 4 The mehanisms responsible for the premature atheroslerosis in RA are not well understood, but traditional risk fators alone are not fully responsible and a role of inflammation in this proess has been suggested. Considering that RA is a hroni inflammatory disease and the pathogeni role of inflammation in the atheroslerosis, it is likely that inflammatory mediators might be ausal in the aelerated atheroslerosis observed in RA. Reeived on: April 1, 214; final version aepted on: September 11, 214. From the Rheumatology Servie (N.B., C.P.-S., P.R.-L., C.C.-V., M.A.A., R.C., Y.J.-G., M.S., A.R.-A., E.C.-E., A.E., C.L.-P.) and Radiology Servie (P.S.), Maimonides Institute for Researh in Biomediine of Cordoba (IMIBIC)/Reina Sofia University Hospital, Cordoba, Spain; and Department of Biomedial Sienes and Centre of Exellene for Biotehnology, Development and Biodiversity Researh, University of Sassari, Italy (M.S.). These authors ontributed equally to this artile. The online-only Data Supplement is available with this artile at Correspondene to Nuria Barbarroja, PhD, Unidad de Investigaión, Hospital Reina Sofia, Avda. Menendez Pidal s/n, E-146 Córdoba, Spain. nuria.barbarroja.exts@juntadeandaluia.es 214 Amerian Heart Assoiation, In. Arteriosler Thromb Vas Biol is available at DOI: /ATVBAHA

2 Barbarroja et al Indue Cardiovasular Disease in RA 277 Downloaded from by on Deember 27, 218 Nonstandard Abbreviations and Aronyms CCP CIMT CRP CVD D MCP-1 MIP-1α RA RF TF Th TNFα Treg yli itrullinated protein arotid intima media thikness C reative protein ardiovasular disease disease ativity sore monoyte hemotati protein-1 marophage inflammatory protein-1 alpha rheumatoid arthritis rheumatoid fator tissue fator T helper tumor nerosis fator-α T regulatory Oxidative stress is another proess frequently altered in RA, whih also ontributes to atheroslerosis, being assoiated with a poorer prognosis. Neutrophils are thought to be responsible for the elevated levels of reative oxygen speies found in plasma and synovial fluid of, thus ontributing to the joint tissue damage within synovial membrane. 6 In addition, a positive orrelation has reently been shown between neutrophils reative oxygen speies levels in blood and synovial fluid with disease ativity sore (D)-28, C reative protein (CRP), and antiyli itrullinated protein (CCP) levels. 7 However, monoytes also seem to ontribute to the elevated oxidative stress in blood of through the prodution of superoxide anion. 8 Many different ell omponents an be onsidered as key elements in the long ourse of RA, inluding blood white ells, suh as lymphoytes, monoytes, and neutrophils. Ativation of these ells leads to the prodution of ytokines and mediators of inflammation, oxidative stress, and atheroslerosis. In this regard, monoytes/marophages are entral players in inflammation and have been found to be ativated in RA through the release of ytokines 9 and massive infiltration in inflammatory sites, suh as synovial membranes. 1 In addition, monoytes/marophages have been assoiated with bone erosion in RA, owing to their exessive differentiation into osteolasts. 11 Beause of its role in RA, several monoyte-targeted therapies are being developed. Lymphoytes are the most studied ells mediating the pathogenesis of RA, whih in fat is onsidered a T helper 1 (Th1) ell driven disorder with a Th1/Th2 imbalane toward higher Th1 levels ompared with healthy donors. 12 Th1 releases inflammatory moleules, suh as interferon-γ, interleukin (IL)-2, and tumor nerosis fator-α (TNFα), that prevent CD4 + T ells from differentiating into Th2 ells. Moreover, Th17 has lately been assoiated with RA pathogenesis, 13 mainly produing IL-17A, IL-6, and TNFα, ytokines that are inreased in serum of and are responsible for the ativation of other ells. On the other hand, regulatory T ells (Treg) also partiipate in the ourse of this disease. They play an important role in immunosuppression through ell ell interation and are found signifiantly dereased at advaned stage, and their funtion has been shown impaired in RA. 12 Thus, it has been proposed that the development and progression of RA is aused by the imbalane of Th1/Th2 and Th17/Treg ells. Neutrophils are the most abundant ells in the synovial fluid of affeted joints and at the pannus/artilage interfae, where tissue damage ours. They promote damage through the indution of oxidative stress, enzyme release, and expression of proinflammatory ytokines. However, the role of neutrophils in RA pathogenesis is not fully understood. 14 Among the auto-antibodies deteted in RA, the s are highly speifi for this disease beause IgM rheumatoid fator (RF) is also present in an elevated perentage of patients with several hroni infetions. Moreover, atheroslerosis status in RA is independent of RF levels. 16 In ontrast, detetion of antibodies has beome a useful diagnosti tool, partiularly at early stage. 17 Thus, the new 21 RA lassifiation riteria inluded detetion of s as a key item for diagnosing the disease. 18 High levels of s have been suggested to predit an adverse ardiovasular profile in. However, that hypothesis just arises from assoiation studies in various ohorts, and no study has still demonstrated a diret effet of s in all those proesses. Here we demonstrate the diret involvement of s, purified from, in the indution of a pro-oxidative, proinflammatory, and atherogeni status in different leukoyte subsets. In addition, speifi effets on marophage ativation and survival were also demonstrated. Materials and Methods Material and methods are available in the online-only Data Supplement. Results Clinial and Analyti Details of RA Patients Clinial details of and healthy donors inluded in this study are stated in Table 1. A.66% of was positive for RF, whereas 6.33% presented antibodies. The patients inluded in the present study had a low moderate disease ativity with a D28 of 3.49±.17, onsidering moderate level of disease D28 values from 2.4 to 3.7 and low level D28< CRP was signifiantly inreased in ompared with healthy donors. As an early atherosleroti marker, pathologial CIMT was found in 21.3% of the. Altered Expression of Prothromboti and Inflammatory Markers in Plasma of RA Patients Compared with healthy donors, displayed elevated plasma levels of prothromboti and inflammatory moleules, suh as IL-2, IL-6, IL-8, IL-1, IL-17A, IL-23, monoyte hemotati protein-1 (MCP-1), marophage inflammatory protein-1α (MIP-1α), sp-seletine, TNFα, and tissue plasminogen ativator (Table I in the online-only Data Supplement).

3 278 Arteriosler Thromb Vas Biol Deember 214 Downloaded from by on Deember 27, 218 Table 1. Clinial Details of the Rheumatoid Arthritis Patients and the Healthy Donors RA Patients Healthy Donors Clinial parameters Women/men, n/n 3/22 22/9 Age, y 7.4± ±12.82 Evolution time, y 12.33±8.93 RF positive, n 38 (.66%) antibodies, n 49 (6.33%) D ±.17 Obesity, n 7 (9.3%) 1 (3%) Diabetes mellitus, n 7 (9.3%) Hypertension, n 21 (28.%) 1 (3%) Menopause, n 2 (33.3%) 9 (29%) Smoker, n 16 (19.%) 6 (19%) CIMT, n 29 (21.3%) 1 (3%) Laboratory parameters Total holesterol, mg/dl 26.2± ±29.92 HDL-holesterol, mg/dl 3.96± ±14.88 LDL-holesterol, mg/dl ± ±24.9 Triglyerides, mg/dl 122.7± ±4.19 Apolipoprotein A, g/l ± ±29.4 Apolipoprotein B, g/l 88.76± ±19.62 ESR 24.36± ±1.68 CRP, mg/dl 14.33±2.6 1.±1.47 Creatinine kinase (U/L) 6.92± ±97.4 C ± ±49.13 C4 26.6± ±11.96 Treatments Cortiosteroids 41 (4.6%) Anti-malarials, n 12 (16.%) NSAIDS, n 9 (78.6%) Metotrexate, n 1 (68.%) Anti-TNF treatment Vitamin D 27 (36.%) Values are means±sem. indiates antiyli itrullinated proteins; CIMT, arotide intime media thikness; CRP, C reative protein; D, disease ativity sore; ESR, erythroyte sedimentation rate; HDL, high-density lipoprotein; LDL, low-density lipoprotein; NSAIDS, non-steroideal anti-inflammatory drugs; RA, rheumatoid arthritis; RF, rheumatoid fator; and TNF, tumor nerosis fator. Signifiant differenes vs ontrols (P<.1). Inreased Oxidative Status in Leukoytes and Plasma RA Intraellular peroxides and peroxynitrites were signifiantly inreased in monoytes and neutrophils from ompared with healthy donors (Figure 1A). In both ell populations, a high perentage of ells with mitohondrial membrane depolarization was further observed (Figure 1B). In ontrast, intraellular redued glutathione was found in both ell types (Figure 1C). In addition, a signifiant drop in total antioxidant apaity in plasma from was observed (P=.2). Nitri oxide levels were lower in plasma from RA patients versus healthy ontrols (P=.49), likely beause of its onsumption after reating with anion superoxide and the subsequent formation of peroxynitrite, as suggested by the elevated protein tyrosine nitration in plasma (P=.46; Table I in the online-only Data Supplement). Correlations and Assoiations Among Autoantibodies, Atheroslerosis, and Inflammatory/ Oxidative Stress Markers in RA Patients In our ohort of patients with RA, s levels positively orrelated with age and D28 index (P=.24 and P=.1, respetively). A further positive orrelation was demonstrated among the s levels and inflammatory plasma markers, suh as erythroyte sedimentation rate, CRP, IL-2, IL-8, IL-6, IL-17A, IL-23, MCP-1, MIP- 1α, TNFα, matrix metalloprotease-3, sp-seletine, tissue plasminogen ativator, and vasular endothelial growth fator-a (P<. and P<.1). In relation to oxidative stress, levels positively orrelated with tyrosine nitration levels (P=.32) and negatively with total antioxidant apaity (P=.2; Table II in the online-only Data Supplement). Assoiation studies revealed that arotid intima media thikness (CIMT) was strongly assoiated with levels (P<.1; Figure I in the online-only Data Supplement). Additionally, pathologial CIMT was assoiated with inreased levels of inflammatory moleules (CRP, MCP-1, and MIP1α) and oxidative stress markers, suh as dereased antioxidant apaity (Figure I in the online-only Data Supplement). These results indiate a diret relationship between oxidative stress and inflammation in the development of atheroslerosis in RA. No signifiant orrelation was observed between the traditional risk fators (obesity, diabetes mellitus, and hypertension) and the inflammatory parameters and oxidative markers analyzed in. Differential Expression of Inflammatory Moleules in Leukoytes in RA Reverse transriptase polymerase hain reation analysis showed that monoytes were the main ontributors to the high levels of IL-1β, TNFα, MCP-1, tissue fator (TF), and IL-8 observed in (Figure 2A, 2B, and 2D 2F), whereas both monoytes and lymphoytes were the soure of IL-6 (Figure 2C). As expeted, only lymphoytes expressed IL-17 and IL-2, mrna, with a great elevation in versus healthy donors (Figure 2G and 2H). Even though IL-23 expression was deteted in all leukoyte types, only RA lymphoytes showed a signifiant inrease of this ytokine ompared with ontrols (Figure 2L). Lymphoytes from healthy donors showed higher IL-1 mrna expression ompared with RA lymphoytes, whereas RA monoytes had inreased IL-1 mrna expression, although this inrement does not seem to aount for the high levels of IL-1 found in plasma (Figure 2J), suggesting that might be a mehanism of ompensation.

4 Barbarroja et al Indue Cardiovasular Disease in RA 279 A Downloaded from by on Deember 27, 218 JC-1 (PE) B C JC1 (% ells with mitohondrial depolarization) DhRh-123 (MFI) CMF-DA (MFI) Intraellular glutathione (GSH) DhRh-123 (FITC) 19.2% 14.7% 8.8 % JC-1 (FITC) CMF-DA (FITC) Figure 1. Inreased oxidative stress in peripheral monoytes and neutrophils from rheumatoid arthritis (RA) patients. A, Peroxides and peroxynitrites prodution in neutrophils, monoytes, and lymphoytes of and healthy donors determined by flow ytometry with the fluoresent probe dihydrorhodamine-123. B, Proportion of irulating blood ells with depolarized mitohondria determined with the JC-1 (,,6,6 -tetrahloro-1,1,3,3 - tetraethylbenzimidazolylarboyanine iodide) MitoSreen assay. C, Levels of intraellular glutathione in neutrophils, monoytes, and lymphoytes from and healthy donors measured through the fluoresent probe CMF-DA (-hloromethylfluoresein diaetate). Seventy-five RA and 31 ontrol samples were examined. Representative histograms or dot plots are shown in parallel with bar graphs showing the mean±sd of median fluoresene intensity (MFI) or perentage of damaged ells of 1 measurement from 1 sample from eah patient and ontrol inluded in the study. Filled histograms represent ontrols and empty histograms represent. () indiates signifiant differenes vs healthy donors (P<.). FITC indiates fluoresein isothioyanate; and PE, phyoerythrin. Despite marophage migration inflammatory fator alpha was found largely elevated in plasma from (Table I in the online-only Data Supplement), mrna expression of marophage migration inflammatory fator alpha was signifiantly lower in RA monoytes and lymphoytes versus ontrols (Figure 2I). Neutrophils from displayed high levels of MCP-1 alongside monoytes (Figure 2D). These ells may be

5 271 Arteriosler Thromb Vas Biol Deember 214 Downloaded from by on Deember 27, 218 A B C D Relative IL1β mrna expression Relative TF mrna expression also responsible for the elevation of MIP-1α in plasma of RA patients (Figure 2K). Leukoyte Ativation in RA Patients CD11a, CD11b, CD62L, and CD66b are well-known leukoyte ativation markers needed for the adherene to the endothelium and atherogenesis proess. In our ohort of, we found a signifiantly higher expression of CD11b on monoytes and neutrophils (P<.1). RA neutrophils further showed an inrease in the ell surfae expression of CD11a and CD66b (P<.; Table III in the online-only Data Supplement). Antibodies Indue a Pro- Oxidative Status in Healthy Leukoytes To haraterize the antibodies, we investigated the sublasses of s isolated from serum of RA patients, having high titers of s and no RF. In aordane with a reent study, 2 1 followed by 4 were the predominant sublasses among antibodies (Figure II in the online-only Data Supplement). After 1 hour of treatment with U/mL of s, we found a signifiant inrease in the levels of peroxides in healthy monoytes and neutrophils when omparing to treatment with syntheti (Figure 3A). Additionally, the perentage Relative TNFα mrna expression Relative IL17 mrna expression Relative MIPα mrna expression E F G H Relative MIFα mrna expression 1 Relative IL-8 mrna expression I J K L Relative IL1 mrna expression Relative IL-6 mrna expression. Relative MCP-1 mrna expression Relative IL-2 mrna expression Relative IL23 mrna expression of ells with depolarized mitohondria signifiantly rose in all leukoyte subsets after treatment (Figure 3B). In ontrast, addition of s aused a drop in GSH (glutathione) levels (Figure 3C). Anti-CCP Antibodies Promote a Distint Proinflammatory Profile in Healthy Leukoytes signifiantly inreased the mrna expression of IL1β, TNFα, MCP-1, TF, and IL-8 in healthy monoytes, in agreement with the results found in vivo in (Figure 4A, 4B, and 4D 4F). Neutrophils also responded to the treatment with s, and an augmentation of IL1β, MCP-1, and IL-8 was further observed on these ells (Figure 4A, 4D, and 4F). In aordane with the vivo studies, s elevated the mrna expression levels of IL-6 in lymphoytes and monoytes (Figure 4C). A signifiant inrease was also notied in IL-2 and IL-23 mrna levels in lymphoytes after s addition (Figure 4H and 4L). Alongside lymphoytes, neutrophils were the ell type that expressed higher levels of IL-23 in response to treatment (Figure 4L). In agreement with the in vivo studies, in vitro treatment elevated the expression of MIP-1α mrna in healthy neutrophils, onfirming that antibodies are responsible for the Figure 2. Different mrna expression profile of inflammatory moleules in rheumatoid arthritis (RA) leukoytes. Seventy-five RA and 31 ontrol samples were examined. A-L, Quantitative real-time reverse transriptase polymerase hain reation was performed on a panel of inflammatory ytokines in lymphoytes, monoytes, and neutrophils from eah RA patient and healthy donor. () indiates signifiant differenes vs healthy donors (P<.). IL indiates interleukin; MCP, monoyte hemotati protein; MIF, marophage migration inflammatory fator; MIP, marophage inflammatory protein; TF, tissue fator; and TNF, tumor nerosis fator.

6 Barbarroja et al Indue Cardiovasular Disease in RA 2711 A Downloaded from by on Deember 27, 218 B JC-1 (PE) DhRh-123 (MFI) 2 1 DhRh-123 (FITC) JC1 (% ells with mitohondrial depolarization) 2.16% 16.64% 1.62% C CMF-DA (MFI) Intraellular glutathione (GSH) JC-1 (FITC) CMF-DA (FITC) Figure 3. Antiyli itrullinated proteins (s) antibodies indue a pro-oxidative status in healthy leukoytes. Healthy leukoytes were treated with s ( U/mL) or equivalent onentration of syntheti human for 1 hour. Four independent experiments were performed in samples from 4 different healthy donors. A, Peroxides and peroxynitrites prodution in neutrophils, monoytes, and lymphoytes of healthy donors treated with s determined by the fluoresent probe dihydrorhodamine-123. B, Proportion of irulating blood ells with depolarized mitohondria determined with the JC-1 (,,6,6 -tetrahloro-1,1,3,3 -tetraethylbenzimidazolylarboyanine iodide) MitoSreen assay. C, Levels of intraellular glutathione in neutrophils, monoytes, and lymphoytes from healthy donors treated with s measured through the fluoresent probe CMF-DA (-hloromethylfluoresein diaetate). Representative histograms or dot plots are shown in parallel with bar graphs showing the mean±sd of median fluoresene intensity (MFI) or perentage of damaged ells of 4 independent experiments. Filled histograms represent untreated ells and empty histograms represent leukoytes treated with s. () indiates signifiant differenes vs ells treated with syntheti human (P<.). FITC indiates fluoresein isothioyanate; and PE, phyoerythrin. augmented MIP-1α expression in neutrophils and its elevation in plasma from (Figure 4K). In ontrast to that observed in, no hanges in IL-17 expression were observed between and treatments in lymphoytes (data no shown). Although no differenes were seen in the expression of vasular endothelial growth fator and its reeptor, vasular endothelial growth fator-r1, omparing leukoytes from ontrols and (data not shown), we deteted a signifiant inrease of these proteins in monoytes and neutrophils after treatment

7 2712 Arteriosler Thromb Vas Biol Deember 214 Downloaded from by on Deember 27, 218 mrna TF relative expression mrna VEGF relative expression A B C D mrna IL1β relative expression (Figure 4I and 4J). Finally, the altered expression of IL-1 in RA monoytes and lymphoytes was fully reapitulated in these ell types after treatment with s (Figure 4G). To further validate the speifiity of the antibodies, healthy lymphoytes, monoytes, and neutrophils were treated with total s from 3 depleted or nondepleted in s, s from pooled sera of 4 healthy donors, s from pooled sera of 4 patients with ankylosing spondylitis (nonautoimmune inflammatory disease), and a ommerial monolonal (anti-itrullinated fibrinogen). The depletion of ativity in s- was 4% to 7%, depending on the initial ativity observed in eah RA sample, probably beause of the exess of olumn binding apaity (Figure III in the online-only Data Supplement). The results showed that the inreased expression of genes indued by isolated s (through ELISA method) was reapitulated when treating healthy leukoytes with total s from. Thus, mrna expression of MIP-1α, IL-23, and MCP-1 was inreased in neutrophils, IL-2 and IL-23 was elevated in lymphoytes, and IL-1b, MCP-1, IL-6, TF, and vasular endothelial growth fator was upregulated in monoytes after treatment with μg RA s (Figure ). To demonstrate the speifiity of s, leukoytes were treated mrna TNFα relative expression mrna IL8 relative expression mrna IL6 relative expression E F G H mrna VEGFR1 relative expression 2 1 mrna MIPα relative expression mrna IL1 relative expression I J K L mrna MCP-1 relative expression mrna IL2 relative expression mrna L23 relative expression with RA s depleted in s. A signifiant redution of the mrna expression of these genes was observed (Figure ). Moreover, the monolonal promoted a signifiant elevation of the expression of all the genes evaluated. On the ontrary, no effets in the expression of these genes were notied after treatment with s from ankylosing spondylitis patients, indiating the role of autoimmune system in the pathology of RA. Anti-CCP Antibodies Inrease the Expression of Leukoyte Ativation Markers After 6 hours treatment with s, a signifiant elevation of CD11b and CD66b was observed in neutrophils (P=.19 and P=.4, respetively) and an inrease of CD11a was deteted in monoytes (P=.1; Table IV in the online-only Data Supplement). These results indiated that s are diretly responsible for leukoyte ativation and might, therefore, be involved in the vasular ompliations and the development of atheroslerosis assoiated to RA. Anti-CCP Antibodies Indue Apoptosis and Impair Marophage Funtion Anti-CCP treatment during 24 hours indued apoptosis in THP-1 (human monoyte-like ell line) derived Figure 4. Antiyli itrullinated proteins (s) antibodies promote a distint proinflammatory profile in healthy leukoytes. Four independent experiments were performed in samples from 4 different healthy donors. A L, Quantitative real-time reverse transriptase polymerase hain reation was performed on a panel of inflammatory ytokines in lymphoytes, monoytes, and neutrophils from healthy donors treated with s or equivalent onentration of syntheti human for 6 hours. () indiates signifiant differenes vs ells treated with syntheti human (P<.). IL indiates interleukin; MCP, monoyte hemotati protein; MIP, marophage inflammatory protein; TF, tissue fator; TNF, tumor nerosis fator; and VEGF, vasular endothelial growth fator.

8 Barbarroja et al Indue Cardiovasular Disease in RA 2713 mrna MIPα relative expression MIP-1α- Neutrophils Monolonal mrna IL23 relative expression IL23 - Neutrophils Monolonal mrna MCP-1 relative expression MCP1 - Neutrophils Monolonal mrna IL2 relative expression IL2- Lymphoytes Monoloal mrna IL23 relative expression IL23- Lymphoytes Monoloal mrna IL1β relative expression IL1β-Monoytes Monolonal Downloaded from by on Deember 27, 218 mrna MCP1 relative expression MCP-1 Monoytes Monolonal marophages in a dose-dependent manner, as demonstrated by hanges in ell morphology (Figure IVA in the onlineonly Data Supplement) and the inreased perentage of Annexin V/PI positive ells (11% and 19% of apoptosis after 1 and 2 U/mL s treatment, respetively; Figure IVB in the online-only Data Supplement). These results were further supported by the inreased mrna expression of moleules involved in apoptosis, inluding Fas ell surfae death reeptor, B-ell lymphoma 2 interating killer, tumor protein P3, and B-ell CLL/lymphoma 2 (Figure IIIC in the online-only Data Supplement). Effets in the marophage inflammatory state were also notied: treatment with s inreased the mrna expression of ytokines, suh as IL1, IL23, TNFα, and IL1β in a dose-dependent manner (Figure IVC in the online-only Data Supplement). Disussion are highly speifi preditive and diagnosti markers of RA, and they have been found to be pathogeni in experimental murine models of arthritis. 24,2 In mrna IL6 relative expression IL6 Monoytes Monoloal mrna TF relative expression TF Monoytes Monoloal mrna VEGF relative expression humans, there is a lose relationship between levels and inflammation, oxidative stress, and atheroslerosis assoiated to RA. 26 Nevertheless, the diret effets of these auto-antibodies in the funtion of the human leukoytes and marophages have not been studied yet. Here we show for the first time that antibodies at as diret indutors of the pro-oxidative status, inflammatory and atherogeni profile of lymphoytes, monoytes, and neutrophils in the CVD assoiated to RA. In addition, effets on marophage ativation and alteration of its apoptoti program are also reported. Oxidative stress has been desribed in RA, and growing evidene suggests it ould be used for earlier diagnosis of RA. 27 Here, we onfirm the inreased oxidative stress and impaired antioxidant apaity in RA both at plasma and ellular levels. In our ohort, RA plasma had high levels of oxidative stress markers and diminished levels of antioxidant apaity. These results are in agreement with reent studies showing a pro-oxidative status and impairment of antioxidant apaity in RA, measured as altered levels of VEGF Monoytes Figure. Antiyli itrullinated proteins (s) speifially indue the expression of inflammatory and prothromboti moleules in white blood ells. Healthy neutrophils, lymphoytes, and monoytes were treated with μg s from 3 rheumatoid arthritis (RA) patients depleted or nondepleted in s, s from ankylosing spondylitis () patients, s from healthy donors, and 1 μg/ml monolonal (antiitrullinated fibrinogen ) for 6 hours. Before the different treatments, ells were treated with human anti-fcr (fragment, rystallizable reeptor) to blok the nonspeifi bindings for 3 minutes. Panels show the mean±sem of 2 separated experiments performed in dupliate. (a) indiates signifiant differenes vs ells treated with normal human serum s (P<.), (b) indiates signifiant differenes vs ells treated with serum s (P<.), and () indiates signifiant differenes vs ells treated with RA nondepleted in s(p<.). IL indiates interleukin; MCP, monoyte hemotati protein; MIP, marophage inflammatory protein; TF, tissue fator; and VEGF, vasular endothelial growth fator. Mooloal

9 2714 Arteriosler Thromb Vas Biol Deember 214 Downloaded from by on Deember 27, 218 malondialdehyde, GSH, GSH Px (GSH peroxidase), GSH redutase, superoxide dismutase, atalase, gluose-6-phosphate, vitamin E, and total NO. 27,28 Although studies at plasma levels are well defined, the investigations at ellular level are more limited. In our study, neutrophils and monoytes from showed mitohondrial depolarization, inreased reative oxygen speies and peroxynitrite levels, and lower levels of GSH. These results agree with a reent study showing enhaned prodution of superoxide and peroxynitrites in monoytes and neutrophils of. 29 Interestingly, in our ohort of patients, no hanges in oxidative stress were seen in lymphoytes. Yet, other authors have desribed an alteration in mitohondria polarization and ell surfae-thiols in lymphoytes from. 3,31 This ould be beause of the different disease sore or the perentage of patients with high titers. Our data were further supported by in vitro studies, demonstrating the diret effet of s in the indution of an oxidative status in monoytes and neutrophils. The assoiation between oxidative stress and s was also validated by a reent study showing that a deletion polymorphism in glutathione S-transferase Mu-1 (antioxidant enzyme) was assoiated with positivity, 32 suggesting that oxidative stress might be involved in the development of RA-speifi autoimmunity in genetially suseptible subjets. Moreover, a reent study showed an inreased synovial oxidant ativity in positive ompared with negative for s. 33 It is well established that ytokine networks play a ritial role in the pathogenesis of RA. There is an elevation of proinflammatory ytokines in RA plasma and synovium fluid (reviewed in 34 ). Aordingly, our ohort of displayed elevated plasma levels of proinflammatory ytokines, whih, in addition, strongly orrelated with elevated levels of s in plasma, suggesting a role for these autoantibodies in the inflammatory profile assoiated with RA. In support for this notion, there is evidene of a different ytokine profile assoiated with or RF antibodies. 3 Moreover, it has been shown that positive patients had signifiant higher levels of IL1β, IL-1, and IL-17 in their synovial fluid than negative patients, 36 whih is in line with our data in plasma from. Searhing the soure of these proinflammatory moleules, we observed a different profile for eah ell type: lymphoytes, monoytes, and neutrophils. RA lymphoytes partiularly expressed elevated levels of IL17, IL2, IL6, and IL23 and dereased levels of IL-1. Interestingly, in vitro treatment with s fairly reapitulated the proinflammatory profile seen in RA lymphoytes, inreasing the expression of IL2, IL6, and IL23 and lowering the expression of IL-1. The role of these ytokines has extensively been desribed in the development of different T ell subsets in RA, suh as Th1, Th17, and Treg ells. 34 Our study suggests that s might promote the differentiation of Th1 and the impairment of Treg ells, main harateristis of the physiopathology of RA. IL-17 is a newly identified ytokine assoiated with the pathogenesis of RA and desribing the speifi ell subtype: Th17. Number of Th17 ells is inreased in peripheral blood of, and IL-17, whih is greatly released in plasma and synovial fluid, has relevant roles mediating inflammation, mirovasular funtion, and atheroslerosis On the ontrary, although IL-17 levels strongly orrelated with levels, in vitro treatment of normal lymphoytes with anti- CCPs did not indue its expression. In this regard, it has been demonstrated that monoytes and fibroblasts from inflamed joints are able to indue the development of Th17 ells, 4,41 suggesting that interations with other ativated ells might be further required to indue the expression of IL17. Despite RA is reognized as Th-driven disease, monoytes/ marophages are important effetors in its pathology, being entral players in inflammation and atheroslerosis. In our ohort of, a signifiant rise in IL1β, TNFα, IL6, MCP1, IL-8, and TF mrna was observed in peripheral monoytes. Of note, in vitro treatment of healthy monoytes and THP-1 derived marophages fully reapitulated this prooagulant and inflammatory profile. Additionally, this aspet was further supported by the diret relationship found in vivo between titers and levels of inflammatory moleules. Marophages plastiity plays a entral role during the development of RA. 42 Anti-CCP treatment promoted the expression of inflammatory ytokines typially assoiated to the lassial inflammatory marophage M1 state. This data points to antibodies as indutors of the M1 polarization in RA. Interestingly, inreasing doses promoted apoptosis in marophages. Aumulating evidene indiates that impaired apoptosis represents a mehanism underlying the pathogenesis of RA. Although apoptoti ells are rarely found in RA tissues in vivo, diverse RA ells have often been observed to express high levels of Fas ell surfae death reeptor and are highly suseptible to apoptosis in vitro, likely refleting multiple antiapoptoti proesses or alteration in the normal apoptosis program in RA. 43 In that way, elevated levels of Fas ell surfae death reeptor mrna and other pro/ antiapoptoti moleules (p3, B-ell lymphoma 2 interating killer, and Bl-2) was shown after treatment with s, suggesting that the alteration in the apoptosis proess in RA ould be beause of the ation of antibodies. Neutrophils play a relevant role in RA, being the predominant ells in synovial fluid. MIP-1α is highly expressed by RA neutrophils from synovial fluid, ontributing to the reruitment of mononulear ells from the bloodstream into synovium. 44 Aording to previous evidene, we found high MIP-1α levels in plasma from. 4 Adding new data about the soure of this moleule in plasma, we show that RA neutrophils were the ells expressing highest MIP- 1α mrna levels ompared with healthy donors, indiating that they are the main responsible for elevated blood MIP-1α levels in RA. Moreover, a strong orrelation was observed between MIP-1α and levels, indiating that s are diretly involved in the prodution of this hemokine. We further onfirm this hypothesis showing that s signifiantly inreased levels of MIP-1α in healthy neutrophils, whereas no hanges were seen in monoytes and lymphoytes. Neutrophils have been shown to produe a large variety of proinflammatory ytokines and proresolving mediators. In addition, the diret interation of neutrophils with

10 Barbarroja et al Indue Cardiovasular Disease in RA 27 Downloaded from by on Deember 27, 218 marophages, dendriti ells, natural killer ells, and lymphoytes modulates the immune response. 46 In that way, we ould demonstrate that treatment of neutrophils with s indued in the expression of higher number of inflammatory moleules than those observed in vivo in RA neutrophils. These data further suggest the relevant role of s as indutors of an inflammatory profile in neutrophils. The prevalene of CVD in is well doumented, and antibodies have been linked to an adverse ardiovasular profile. 26 In agreement with the existing body of literature, we found altered diverse parameters strongly assoiated with the development of atheroslerosis in RA patients. Thus, an inreased presene of ativation markers in neutrophils and monoytes and elevated expression of TF in monoytes were seen in. Additionally, 21.3% RA patients had pathologial CIMT, and there was a strong assoiation between pathologial CIMT and D28, CRP levels, MCP-1 plasma, MIP-1α levels and low TAC and levels. These results are in agreement with a study, indiating that positive had higher CIMT than patients without evidene of these antibodies. 47 Moreover, it has been shown that itrullinated proteins are prevalent within atherosleroti plaques, and ertain antibodies are orrelated with the CIMT 48 and assoiated with the atherosleroti burden. 49 In this sense, treatment of healthy leukoytes with s resulted in an elevation of ell ativation markers in monoytes and neutrophils, an inrease of TF expression, inflammation, and oxidative stress, indiating the partiipation of these autoantibodies in the development of atheroslerosis and CVD. Although positive and negative patients display similar linial features, they have a different disease ourse, being more severe in positive. Here we demonstrate that antibodies may diretly inrease the risk of CVD in RA. This was onfirmed through both the assoiations among levels and parameters related to inflammation, oxidative stress, and atheroslerosis in vivo and the in vitro treatment of healthy leukoytes with s. Thus, s ativate leukoytes and hange their phenotype to an atherogeni profile through the (a) alteration of the oxidative status in monoytes and neutrophils, (b) promotion of the Th-1 differentiation and the impairment of Treg ells, () modulation of the MIP-1α expression in neutrophils, inreasing the migratory properties of mononulear ells, (d) elevation of the inflammatory and prooagulant monoyte ativity, and (e) the polarization of marophages to a lassial inflammatory M1 state and alteration of the apoptosis program in marophages. Targeting these autoantibodies would be an exellent strategy to prevent the development of CVD assoiated with RA. Aknowledgments We thank all the patients for their kind partiipation in this study. We thank J.M. Martinez-Moreno for his exellent tehnial support. Soures of Funding This work was supported by grants from the Junta de Andaluia (CTS-794, PI ), the Spanish Rheumatology Foundation (FER), the Ministry of Health (PI12/11). C. López-Pedrera was supported by a ontrat from the Spanish Junta de Andaluia. None. Dislosures Referenes 1. del Rinón ID, Williams K, Stern MP, Freeman GL, Esalante A. High inidene of ardiovasular events in a rheumatoid arthritis ohort not explained by traditional ardia risk fators. Arthritis Rheum. 21;44: Maradit-Kremers H, Crowson CS, Niola PJ, Ballman KV, Roger VL, Jaobsen SJ, Gabriel SE. Inreased unreognized oronary heart disease and sudden deaths in rheumatoid arthritis: a population-based ohort study. Arthritis Rheum. 2;2: Gonzalez-Juanatey C, Llora J, Testa A, Revuelta J, Garia-Porrua C, Gonzalez-Gay MA. Inreased prevalene of severe sublinial atherosleroti findings in long-term treated rheumatoid arthritis patients without linially evident atherosleroti disease. Mediine (Baltimore). 23;82: Ku IA, Imboden JB, Hsue PY, Ganz P. Rheumatoid arthritis: model of systemi inflammation driving atheroslerosis. Cir J. 29;73: Libby P, Ridker PM, Maseri A. Inflammation and atheroslerosis. Cirulation. 22;: Hithon CA, El-Gabalawy HS. Oxidation in rheumatoid arthritis. Arthritis Res Ther. 24;6: Kundu S, Ghosh P, Datta S, Ghosh A, Chattopadhyay S, Chatterjee M. Oxidative stress as a potential biomarker for determining disease ativity in patients with rheumatoid arthritis. Free Radi Res. 212;46: Ostrakhovith EA, Afanas ev IB. Oxidative stress in rheumatoid arthritis leukoytes: suppression by rutin and other antioxidants and helators. Biohem Pharmaol. 21;62: Lioté F, Boval-Boizard B, Weill D, Kuntz D, Wautier JL. Blood monoyte ativation in rheumatoid arthritis: inreased monoyte adhesiveness, integrin expression, and ytokine release. Clin Exp Immunol. 1996;16: Kinne RW, Stuhlmuller B, Burmester GR. Cells of the synovium in rheumatoid arthritis. Marophages [Review]. Arthritis Res Ther. 27; 9: Takayanagi H. Osteoimmunology: shared mehanisms and rosstalk between the immune and bone systems. Nat Rev Immunol. 27;7: Chen J, Li J, Gao H, Wang C, Luo J, Lv Z, Li X. Comprehensive evaluation of different T-helper ell subsets differentiation and funtion in rheumatoid arthritis. J Biomed Biotehnol. 212;212: Lubberts E. IL-17/Th17 targeting: on the road to prevent hroni destrutive arthritis? Cytokine. 28;41: Németh T, Mósai A. The role of neutrophils in autoimmune diseases. Immunol Lett. 212;143: van Venrooij WJ, Hazes JM, Visser H. Antiitrullinated protein/peptide antibody and its role in the diagnosis and prognosis of early rheumatoid arthritis. Neth J Med. 22;6: Kerola AM, Kerola T, Kauppi MJ, Kautiainen H, Virta LJ, Puolakka K, Nieminen TV. Cardiovasular omorbidities antedating the diagnosis of rheumatoid arthritis. Ann Rheum Dis. 213;72: Nielen MM, van Shaardenburg D, Reesink HW, van de Stadt RJ, van der Horst-Bruinsma IE, de Koning MH, Habibuw MR, van der Horst- Bruinsma IE, Dijkmans BA. Speifi autoantibodies preede the symptoms of rheumatoid arthritis: a study of serial measurements in blood donors. Arthritis Rheum. 24;: Aletaha D, Neogi T, Silman AJ, et al. 21 Rheumatoid arthritis lassifiation riteria: an Amerian College of Rheumatology/European League Against Rheumatism ollaborative initiative. Arthritis Rheum. 21;62: Fransen J, Stuki G, van Riel LCM. Rheumatoid Arthritis Measures. Arthritis Rheum (Arthritis Care & Researh). 23;49:S214 S Engelmann R, Brandt J, Eggert M, Karberg K, Krause A, Neek G, Mueller-Hilke B. 1 and 4 are the predominant sublasses among auto-antibodies against two itrullinated antigens in RA. Rheumatology (Oxford). 28;47: Nishimura K, Sugiyama D, Kogata Y, Tsuji G, Nakazawa T, Kawano S, Saigo K, Morinobu A, Koshiba M, Kuntz KM, Kamae I, Kumagai S. Meta-analysis: diagnosti auray of anti-yli itrullinated peptide antibody and rheumatoid fator for rheumatoid arthritis. Ann Intern Med. 27;146:

11 2716 Arteriosler Thromb Vas Biol Deember 214 Downloaded from by on Deember 27, Avoua J, Gosse L, Dougados M. Diagnosti and preditive value of antiyli itrullinated protein antibodies in rheumatoid arthritis: a systemati literature review. Ann Rheum Dis. 26;6: Bos WH, Wolbink GJ, Boers M, Tijhuis GJ, de Vries N, van der Horst- Bruinsma IE, Tak PP, van de Stadt RJ, van der Laken CJ, Dijkmans BA, van Shaardenburg D. Arthritis development in patients with arthralgia is strongly assoiated with anti-itrullinated protein antibody status: a prospetive ohort study. Ann Rheum Dis. 21;69: Lundberg K, Nijenhuis S, Vossenaar ER, Palmblad K, van Venrooij WJ, Klareskog L, Zendman AJ, Harris HE. Citrullinated proteins have inreased immunogeniity and arthritogeniity and their presene in arthriti joints orrelates with disease severity. Arthritis Res Ther. 2;7:R48 R Kuhn KA, Kulik L, Tomooka B, Brashler KJ, Arend WP, Robinson WH, Holers VM. Antibodies against itrullinated proteins enhane tissue injury in experimental autoimmune arthritis. J Clin Invest. 26;116: Arnab B, Biswadip G, Arindam P, Shyamash M, Anirban G, Rajan P. Anti-CCP antibody in patients with established rheumatoid arthritis: does it predit adverse ardiovasular profile? J Cardiovas Dis Res. 213;4: Kalpakioglu B, Senel K. The interrelation of glutathione redutase, atalase, glutathione peroxidase, superoxide dismutase, and gluose- 6-phosphate in the pathogenesis of rheumatoid arthritis. Clin Rheumatol. 28;27: Hassan SZ, Gheita TA, Kenawy SA, Fahim AT, El-Sorougy IM, Abdou MS. Oxidative stress in systemi lupus erythematosus and rheumatoid arthritis patients: relationship to disease manifestations and ativity. Int J Rheum Dis. 211;14: Ostrakhovith EA, Afanas ev IB. Oxidative stress in rheumatoid arthritis leukoytes: suppression by rutin and other antioxidants and helators. Biohem Pharmaol. 21;62: Moodley D, Mody G, Patel N, Chuturgoon AA. Mitohondrial depolarisation and oxidative stress in rheumatoid arthritis patients. Clin Biohem. 28;41: Pedersen-Lane JH, Zurier RB, Lawrene DA. Analysis of the thiol status of peripheral blood leukoytes in rheumatoid arthritis patients. J Leuko Biol. 27;81: Mikuls TR, Gould KA, Bynoté KK, et al. Antiitrullinated protein antibody (ACPA) in rheumatoid arthritis: influene of an interation between HLA-DRB1 shared epitope and a deletion polymorphism in glutathione S-transferase in a ross-setional study. Arthritis Res Ther. 21;12:R Ediz L, Hiz O, Ozkol H, Gulu E, Toprak M, Ceylan MF. Relationship between antibodies and oxidant and anti-oxidant ativity in patients with rheumatoid arthritis. Int J Med Si. 211;8: MInnes IB, Shett G. Cytokines in the pathogenesis of rheumatoid arthritis. Nat Rev Immunol. 27;7: Hithon CA, Alex P, Erdile LB, Frank MB, Dozmorov I, Tang Y, Wong K, Centola M, El-Gabalawy HS. A distint multiytokine profile is assoiated with anti-ylial itrullinated peptide antibodies in patients with early untreated inflammatory arthritis. J Rheumatol. 24;31: Signifiane 36. Gómez-Puerta JA, Celis R, Hernández MV, Ruiz-Esquide V, Ramírez J, Haro I, Cañete JD, Sanmartí R. Differenes in synovial fluid ytokine levels but not in synovial tissue ell infiltrate between anti-itrullinated peptide/protein antibody-positive and -negative rheumatoid arthritis patients. Arthritis Res Ther. 213;:R van Hamburg JP, Corneth OB, Paulissen SM, Davelaar N, Asmawidjaja PS, Mus AM, Lubberts E. IL-17/Th17 mediated synovial inflammation is IL-22 independent. Ann Rheum Dis. 213;72: Marder W, Khalatbari S, Myles JD, Henh R, Yalavarthi S, Lustig S, Brook R, Kaplan MJ. Interleukin 17 as a novel preditor of vasular funtion in rheumatoid arthritis. Ann Rheum Dis. 211;7:. 39. van Es T, van Puijvelde GHM, Ramos OH Segers FM, Joosten LA, van den Berg WB, Mihon IM, de Vos P, van Berkel TJ, Kuiper J. Attenuated atheroslerosis upon IL-17R signaling disruption in LDLr defiient mie. Biohem Biophys Res Commun 29;388: Evans HG, Gullik NJ, Kelly S, Pitzalis C, Lord GM, Kirkham BW, Taams LS. In vivo ativated monoytes from the site of inflammation in humans speifially promote Th17 responses. Pro Natl Aad Si U S A. 29;16: Noak M, Thiam N, Miosse P. Interation between fibroblast and T lymphoytes is needed for IL-17 seretion and Th17 differentiation requires T ell ativation. Ann Rheum Dis. 214;73(Suppl 1):A Li J, Hsu HC, Mountz JD. Managing marophages in rheumatoid arthritis by reform or removal. Curr Rheumatol Rep. 212;14: Peng SL. Fas (CD9)-related apoptosis and rheumatoid arthritis. Rheumatology. 26;4: Hatano Y, Kasama T, Iwabuhi H, Hanaoka R, Takeuhi HT, Jing L, Mori Y, Kobayashi K, Negishi M, Ide H, Adahi M. Marophage inflammatory protein 1 alpha expression by synovial fluid neutrophils in rheumatoid arthritis. Ann Rheum Dis. 1999;8: Koh AE, Kunkel SL, Harlow LA, Mazarakis DD, Haines GK, Burdik MD, Pope RM, Strieter RM. Marophage inflammatory protein-1 alpha. A novel hemotati ytokine for marophages in rheumatoid arthritis. J Clin Invest. 1994;93: Mantovani A, Cassatella MA, Costantini C, Jaillon S. Neutrophils in the ativation and regulation of innate and adaptive immunity. Nat Rev Immunol. 211;11: Gerli R, Bartoloni Boi E, Sherer Y, Vaudo G, Mosatelli S, Shoenfeld Y. Assoiation of anti-yli itrullinated peptide antibodies with sublinial atheroslerosis in patients with rheumatoid arthritis. Ann Rheum Dis. 28;67: El-Barbary AM, Kassem EM, El-Sergany MA, Essa SA, Eltomey MA. Assoiation of anti-modified itrullinated vimentin with sublinial atheroslerosis in early rheumatoid arthritis ompared with anti-yli itrullinated peptide. J Rheumatol. 211;38: Sokolove J, Brennan MJ, Sharpe O, Lahey LJ, Kao AH, Krishnan E, Edmundowiz D, Lepus CM, Wasko MC, Robinson WH. Brief report: itrullination within the atherosleroti plaque: a potential target for the anti-itrullinated protein antibody response in rheumatoid arthritis. Arthritis Rheum. 213;6: This is a novel detailed study whih identifies the speifi role of antiyli itrullinated proteins in the indution of the proinflammatory and proatherogeni profile in rheumatoid arthritis (RA) patients. Speifially, the key novel findings are the identifiation of lymphoytes as major players in altered inflammatory profile, monoytes in the prothromboti and atherogeni status, and neutrophils in a pro-oxidative feature highlights the relevane of eah ell type in the RA physiopathology. In vitro treatment of healthy monoytes, lymphoytes, and neutrophils with antiyli itrullinated proteins purified from linial samples fully reapitulated the pathogeni profile observed in leukoytes from, promoting their ativation and hanging their phenotype to an atherogeni profile. Thus, our study stresses the importane of antiyli itrullinated protein antibodies in the pathogenesis of RA and speifially in the onset of atheroslerosis. Targeting the mehanisms involved would be an exellent strategy to prevent the development of ardiovasular disease related to RA.