Presented in part at the 69th Annual Scientific Meeting of the American College of Rheumatology, San Diego, CA, November 2005.
Increased serum levels of soluble fractalkine (CX3CL1) correlate with disease activity in rheumatoid vasculitis†
Version of Record online: 30 OCT 2006
Copyright © 2006 by the American College of Rheumatology
Arthritis & Rheumatism
Volume 54, Issue 11, pages 3408–3416, November 2006
How to Cite
Matsunawa, M., Isozaki, T., Odai, T., Yajima, N., Takeuchi, H. T., Negishi, M., Ide, H., Adachi, M. and Kasama, T. (2006), Increased serum levels of soluble fractalkine (CX3CL1) correlate with disease activity in rheumatoid vasculitis. Arthritis & Rheumatism, 54: 3408–3416. doi: 10.1002/art.22208
- Issue online: 30 OCT 2006
- Version of Record online: 30 OCT 2006
- Manuscript Accepted: 9 AUG 2006
- Manuscript Received: 13 APR 2006
- Ministry of Health and Welfare of Japan
To determine levels of soluble fractalkine (sFkn) in rheumatoid arthritis (RA) patients with and without rheumatoid vasculitis (RV), and to assess the relationship of sFkn levels to disease activity.
Serum was obtained from 98 RA patients (54 without vasculitis, 36 with extraarticular manifestations but without histologically proven vasculitis, and 8 with histologically proven vasculitis) and from 38 healthy individuals. Levels of sFkn were measured by enzyme-linked immunosorbent assay. Expression of Fkn and CX3CR1 was quantified by real-time polymerase chain reaction. Vasculitis disease activity was assessed using the Birmingham Vasculitis Activity Score and the Vasculitis Activity Index.
Serum sFkn levels were significantly higher in patients with RA than in controls and were significantly higher in RA patients with RV than in those without vasculitic complications. Statistically significant correlations were observed between serum sFkn levels in RA patients and levels of C-reactive protein, rheumatoid factor, immune complex, and complement. In the RV group, sFkn levels also correlated with disease activity. Immunohistochemical analysis indicated that Fkn levels were associated mainly with endothelial cells in vasculitic arteries. In addition, expression of CX3CR1 messenger RNA was significantly greater in peripheral blood mononuclear cells from patients with active RV than in those from other RA patients or controls. Notably, serum sFkn levels were significantly diminished following successful treatment and clinical improvement.
These findings suggest that Fkn and CX3CR1 play crucial roles in the pathogenesis of RV and that sFkn may serve as a serologic inflammatory marker of disease activity in RA patients with vasculitis.
Rheumatoid vasculitis (RV) is an uncommon but severe complication of rheumatoid arthritis (RA) which can cause skin disorders such as rash, cutaneous ulcerations, and gangrene, as well as neuropathy, eye symptoms, and systemic inflammation (1). Although little is known about the molecular mechanism underlying RV, it is well known that a chronic imbalance in the expression of chemokines and proinflammatory cytokines is likely important in the orchestration of the inflammatory responses observed in patients with RA (2, 3), and similar dysregulation of cytokine and chemokine networks has been suggested to occur in patients with RV.
The chemokine fractalkine (Fkn; CX3CL1) is synthesized as a type I transmembrane protein by endothelial cells (ECs) (4). Its unique CX3C chemokine domain is attached to a 241–amino acid mucin stalk, a 19–amino acid transmembrane domain, and a 37–amino acid intracellular domain of unknown function (4, 5). The soluble form of Fkn (sFkn) reportedly exerts a chemotactic effect on monocytes, natural killer (NK) cells, and T lymphocytes, and acts via its receptor, CX3CR1, as an adhesion molecule that is able to promote the firm adhesion of a subset of leukocytes to ECs under conditions of physiologic flow (6, 7). Thus, Fkn appears to possess immunoregulatory properties that affect inflammatory/immune cell–EC interactions and inflammatory responses at sites of inflammation. Indeed, numerous studies have implicated Fkn in a variety of inflammatory disorders, including glomerulonephritis, RA, systemic sclerosis, and systemic lupus erythematosus (8–12). The aim of the present study was to determine the levels of sFkn and CX3CR1 in patients with RV and to assess the relationship between sFkn levels and RV disease activity.
PATIENTS AND METHODS
We classified 98 patients with RA that fulfilled the 1987 criteria of the American College of Rheumatology (formerly, the American Rheumatism Association) (13) into 3 groups: 1) patients with no clinical signs of vasculitis (RA group), 2) patients with recent-onset (<6 months) extraarticular manifestations, including peripheral neuropathy, skin ulcers, necrotizing glomerulonephritis, fibrosing alveolitis, ischemic colitis, nailfold lesions, pericarditis, pleuritis, (epi-)scleritis, weight loss (≥10% in 6 months), fever, and multiple rheumatoid nodules, but without histologically proven vasculitis (EAM-RA group), and 3) patients with recent-onset extraarticular manifestations and histologically proven vasculitis (RV group). Vasculitis was defined as fibrinoid necrosis in a vessel wall seen on muscle or sural nerve biopsy, or as leukocytoclasis seen on skin biopsy. Serum samples were collected once or twice from all patients during periods of active disease, and from RV patients during both periods of active disease and periods of inactive disease (RA group 71 samples from 54 patients, EAM-RA group 58 samples from 36 patients, and RV group 26 samples from 8 patients). In addition, serum samples were collected from 38 healthy individuals (1 sample from each healthy subject).
Serum CRP, RF, and C4 levels were determined in the clinical laboratory at our hospital, using a latex photometric immunoassay. Levels of C1q immune complex were measured using an enzyme-linked immunosorbent assay (ELISA). Erythrocyte sedimentation rate was measured by the Westergren method. Vasculitis disease activity was assessed using both the Birmingham Vasculitis Activity Scores (BVAS) (14) and Vasculitis Activity Index (VAI) (15). All experiments were carried out in accordance with protocols approved by the Human Subjects Research Committee at our institution, and informed consent was obtained from all patients and healthy volunteers.
Patient characteristics are summarized in Table 1. There were significant differences in serum C-reactive protein (CRP) levels and rheumatoid factor (RF) titers between the RV group and the other 2 groups.
|RA (n = 54)||EAM-RA (n = 36)||RV (n = 8)|
|No. female/no. male||41/13||28/8||5/3|
|Age, years||64.3 ± 1.1||66.7 ± 1.3||69.5 ± 1.0|
|RA duration, years||6.8 ± 1.0||8.9 ± 1.3||8.3 ± 1.6|
|CRP, mg/dl||2.6 ± 0.4||2.9 ± 0.4||4.4 ± 0.8†|
|ESR, mm/hour||58.0 ± 4.4||70.5 ± 4.1||66.0 ± 5.5|
|RF, IU/ml||139.8 ± 31.8||259.1 ± 47.8||551.7 ± 122.3†|
Monoclonal and biotinylated polyclonal antibodies against human Fkn were purchased from Genzyme/Techne (Cambridge, MA).
Soluble Fkn in serum was quantified using a double-ligand ELISA according to a previously described procedure (12), with modifications. Monoclonal murine anti-human Fkn (4 μg/ml) was used as the primary antibody, and biotinylated polyclonal goat anti-Fkn (0.25 μg/ml) as the secondary antibody. This ELISA detects the chemokine domain of human Fkn, and its limit of sensitivity is ∼150 pg/ml. Concentrations of soluble intercellular adhesion molecule 1 (ICAM-1) were quantitated in serum from patients with RV, using a commercial ELISA kit according to the instructions of the manufacturer (Endogen, Rockford, IL).
Isolation of total RNA and real-time polymerase chain reaction (PCR).
Total RNA was extracted from peripheral blood mononuclear cells (PBMCs) and reverse transcribed. Using the complementary DNA (cDNA) obtained, real-time quantitative PCR was carried out as described previously (12), using the LightCycler PCR Detection System with a FastStart DNA Master SYBR Green I kit (Roche Diagnostics, Mannheim, Germany). To compare quantitative results between different samples, a dilution series of cDNA from unstimulated human umbilical vein endothelial cells (HUVECs) and normal human PBMCs, which served as internal standards for Fkn and CX3CR1, respectively, was loaded each time and assigned a value of 100 units. The primers used for real-time PCR were as follows: for human CX3CR1 5′-AGC-AGG-CAT-GGA-AGT-GTT-CT (sense), 5′-GTT-GTT-TTG-TGT-GCA-TTG-GG (antisense); for human Fkn 5′-GCT-GAG-GAA-CCC-ATC-CAT (sense), 5′-GAG-GCT-CTG-GTA-GGT-GAA-CA (antisense); for β-actin (internal control) 5′-CCC-AAG-GCC-AAC-CGC-GAG-AAG-AT (sense), 5′-GTC-CCG-GCC-AGC-CAG-GTC-CAG (antisense).
Fkn antigens were visualized immunohistochemically as previously described (16). Biopsied specimens were embedded in OCT compound (TissueTek II; Miles, Naperville, IL) and snap frozen. Before staining, 5-μm–thick frozen sections were fixed for 30 minutes in ice-cold acetone, after which endogenous peroxidase activity was quenched by incubating the slides for an additional 30 minutes in a mixture of absolute methanol and 3% hydrogen peroxide. The slides were then incubated with polyclonal goat anti-Fkn antibody (1:500 dilution; Genzyme/Techne) or preimmune goat IgG. Biotinylated rabbit anti-goat IgG (Nichirei, Tokyo, Japan) and peroxidase-conjugated streptavidin were used as the second and third reagents, respectively, after which the optimal color was developed for 5–10 minutes using a 3,3′-diaminobenzidine tetrahydrochloride detection kit (Nichirei). After rinsing with distilled water, the slides were counterstained with Mayer's hematoxylin. Tissue sections were also examined by immunostaining with anti-human von Willebrand factor (vWF) polyclonal antibody (Enzyme Research Laboratories, South Bend, IN) to identify vascular ECs.
Data were analyzed with a Macintosh computer using a statistical software package (StatView 4.5; Abacus Concept, Berkeley, CA) and expressed as the mean ± SEM. The significance of the differences between groups was evaluated using the Mann-Whitney U test. Followup data were evaluated by Wilcoxon's test. The relationships between sFkn levels and other parameters were evaluated using Spearman's rank correlation. P values less than 0.05 were considered significant.
Serum sFkn levels.
We initially used ELISAs to compare the sFkn levels in serum samples from all RA patients, with or without extraarticular manifestations (n = 155), with those in samples from healthy individuals (n = 38). We found that sFkn levels were markedly higher in RA patients than in healthy controls (P < 0.0001) (Figure 1A), which is consistent with the results of previous studies (9, 17). Moreover, as shown in Figure 1B, serum sFkn levels correlated positively with CRP levels (r = 0.501, P < 0.0001), RF titers (r = 0.693, P < 0.0001), and C1q immune complex levels (r = 0.743, P < 0.0001) and negatively with C4 levels (r = −0.368, P = 0.0004). Because it is possible that the anti-IgG activity of the RF in the samples may have augmented the reactivity of the Fkn ELISA, we also assayed sFkn in 5 randomly chosen serum samples before and after depleting RF using RF Stripper (The Binding Site, Birmingham, UK). No significant differences in serum sFkn levels were found between the RF-depleted and untreated sera (data not shown).
Effect of serum sFkn levels on extraarticular manifestations and vasculitis in RA patients.
The results summarized above suggest that serum sFkn levels correlate with the disease and inflammatory activity of RA. Fkn is expressed as a membrane-bound chemokine on vascular ECs, where it plays an important role in mediating vascular inflammation (7, 18). Therefore, to investigate the role played by Fkn in extraarticular manifestations of RA, we compared serum sFkn levels among the RA, EAM-RA, and RV groups. As shown in Figure 2, serum sFkn levels were higher in the RV group (mean ± SEM 4,980.1 ± 719.3 pg/ml) and in the EAM-RA group (3,946.9 ± 575.3 pg/ml) than in the RA group (1,846.7 ± 410.9 pg/ml; P = 0.005, RV versus RA and P = 0.012, EAM-RA versus RA). Although serum sFKN levels were higher in the RV group than in the EAM-RA group, the difference was not significant (P = 0.376).
Correlation between serum sFkn levels and features of RV.
To determine whether elevated sFkn levels contribute to the development of vasculitis or vascular damage in RA, we next examined the relationship between serum sFkn levels and serologic parameters thought to parallel the degree of RV activity. We found that serum C4 levels in RV patients showed a significant negative correlation with sFkn levels (r = −0.558, P = 0.007), while C1q immune complex levels showed a significant positive correlation (r = 0.836, P < 0.0001) (Figure 3A). Importantly, sFkn levels in RV patients were also significantly correlated with RV disease activity, as indicated by both the BVAS (r = 0.450, P = 0.0006) and the VAI (r = 0.372, P = 0.0028) (Figure 3B). In addition, we found a significant positive correlation between sFkn and ICAM-1 levels in patients with RV (r = 0.474, P = 0.017) (Figure 3C), which is noteworthy given that circulating ICAM-1 levels may also have diagnostic value for the assessment of RV and provide a clue to the pathogenesis of systemic vasculitis (19).
Immunohistochemical localization of Fkn and expression of Fkn and CX3 CR1 messenger RNA (mRNA).
When we examined the in situ expression of Fkn in skin biopsy samples from patients with active RV, we found that Fkn was associated mainly with impaired arterial ECs within arteries exhibiting vasculitis (Figure 4A). This was also indicated by the colocalization of anti-vWF (results not shown). There was little or no nonspecific staining in tissue sections incubated with control IgG (Figure 4A). Furthermore, quantitative real-time PCR revealed that CX3CR1 mRNA was more strongly expressed in PBMCs from RV patients than in those from either RA patients (P = 0.035) or healthy controls (P = 0.008) (Figure 4B). In contrast, Fkn expression in PBMCs from all 3 groups was markedly weaker than in HUVECs (standard), and there were no significant differences among the 3 groups, although levels in the controls tended to be lower (data not shown).
Followup studies of the effects of treatment on serum sFkn levels.
Figure 5 summarizes the results of followup studies of serum sFkn levels in RV patients after treatment with glucocorticoids and other immunosuppressive drugs. Notably, serum sFkn levels in 7 of 8 patients with active RV were significantly diminished following clinical improvement with treatment (glucocorticoids alone in 3 patients, glucocorticoids plus methotrexate in 3 patients, glucocorticoids plus methotrexate and infliximab in 1 patient, and glucocorticoids plus intravenous cyclophosphamide and intravenous immunoglobulin in 1 patient). No significant differences in the course of the changes in serum sFkn levels were found among the different treatment regimens (data not shown).
In the present study, we found that serum sFkn levels are significantly higher in patients with RV than in RA patients without vasculitis or in healthy controls. In addition, we observed that sFkn levels in RV patients are positively correlated with BVAS and VAI scores and serologic features including RF titers and C1q immune complex levels and negatively correlated with C4 levels, and that sFkn levels decrease in RV patients after successful treatment with glucocorticoids and other immunosuppressive drugs.
Generally, RV is defined histologically as the presence of an inflammatory infiltrate with destruction of the vessel wall (1, 20), and is caused by circulating C1q immune complexes containing RF and autoantibodies that form deposits in vessel walls, where they trigger an inflammatory reaction, which leads to EC injury and activation (21, 22). ECs exert significant proinflammatory activities, amplifying and perpetuating inflammatory processes, and are thus critically involved in the pathogenesis of systemic vasculitis (23–25). The major cellular source of Fkn in the periphery is the endothelium, and EC-derived Fkn likely plays a pivotal role in pathologic conditions such as vascular inflammation, glomerulonephritis, and pulmonary arterial hypertension (7, 26–29). More recently, we and others have demonstrated the involvement of Fkn in systemic lupus erythematosus and in lupus nephritis in MRL mice (12, 30).
In accordance with the results of earlier studies (9, 17), we found that serum sFkn levels were elevated in patients with RA, and that the elevation was especially pronounced in patients with RV, who also exhibited augmented Fkn expression on impaired ECs. Moreover, Ruth et al have shown that CX3CR1 is expressed on peripheral T cells and monocytes of patients with RA (9). In addition, enhanced expression of CX3CR1 in PBMCs from RA patients was observed in both CD4 and CD8 T cells, with Th1 and Tc1 phenotype, respectively (31), which is consistent with the observations that CX3CR1 is preferentially expressed in Th1 cells and that Th1 cells, but not Th2 cells, respond to Fkn (32).
In the present study, we demonstrated increased expression of CX3CR1 mRNA in PBMCs from RV patients. Fkn also acts, via CX3CR1, as an adhesion molecule and as a chemoattractant, recruiting monocytes, NK cells, and T lymphocytes to ECs (33). We suggest that the up-regulated expression of Fkn on ECs and CX3CR1 on PBMCs, along with the accumulation of activated inflammatory cells, reflects the pathophysiologic events leading to vasculitis in RA patients. In this regard, we did not find significantly increased expression of CX3CR1 mRNA in PBMCs from RA patients without vasculitis as compared with those from healthy controls, a finding that was inconsistent with previous reports (9, 31). Such differences in patterns of CX3CR1 expression may reflect the methodology used for assessment of receptor expression. CX3CR1 expression in the earlier studies was assessed, using flow cytometry, as protein expression, while we used reverse transcriptase–PCR analysis to assess mRNA expression.
We found that serum sFkn concentrations in RV patients were significantly correlated with levels of ICAM-1 (Figure 3C), which are indicative of endothelial damage or vascular inflammation (34, 35). Additionally, we demonstrated a significant positive correlation between serum sFkn levels and circulating C1q immune complex levels, and a significant negative correlation between serum sFkn levels and serum C4 levels (Figure 3A) in RV patients. It has been previously demonstrated that C1q immune complexes stimulate ECs to express adhesion molecules such as ICAM-1 and E-selectin (36). Although it remains to be determined whether increased levels of serum immune complexes, RF, or complement consumption have some direct effects on Fkn induction in ECs, the present results, taken together, indicate that the up-regulation of Fkn expression in ECs may be promoted through a mechanism similar to that seen in ICAM-1 induction.
The finding that sFkn levels were higher in RA patients with extraarticular manifestations than in RA patients without extraarticular manifestations might mean that extraarticular manifestations are an indicator of undiagnosed vasculitis in RA patients or might support the frequently proposed hypothesis that vasculopathy underlies the extraarticular manifestations in RA. On the other hand, serum sFkn was undetectable in 60.6% of the patients in the RA group, 28.6% of those in the EAM-RA group, and 38.5% of those in the RV group (Figure 2), and the mean serum CRP levels in these groups were 1.7 mg/dl, 1.8 mg/dl, and 2.6 mg/dl, respectively, whereas patients with detectable serum sFkn had comparatively higher levels of CRP (mean 3.6 mg/dl, 3.3 mg/dl, and 6.9 mg/dl in the RA, EAM-RA, and RV groups, respectively, with detectable sFkn). This indicates that serum sFkn could be a significant serologic inflammatory marker not only for RV, but also for RA and EAM-RA with higher disease activity. Further investigations on a larger sample of patients are needed to understand the clinical significance and roles of Fkn during the progression of RV.
It has become apparent that cytokine abnormalities are central to the pathogenesis of rheumatic disease, but the molecular mechanisms underlying the vasculitis seen in RA remain unclear. There is increasing evidence to suggest that, by mediating vascular endothelial activation, tumor necrosis factor α (TNFα) plays a key role in the pathophysiology of vasculitis in RA (37, 38). Indeed, patients with clinical signs of systemic vasculitis have shown significantly higher levels of TNFα compared with those with no evidence of systemic involvement (39), and TNFα is known to be a potent inducer of Fkn in ECs (18, 40, 41). Taken together, these findings suggest that dysregulation of cytokine cascades affecting Fkn and other inflammatory cytokines, especially TNFα, contributes significantly to the development of RV. This makes sFkn a potentially useful indicator of disease activity in RA with vasculitis, especially among patients with active disease, as well as of the effects of treatment.
We would like to thank Dr. Jun Shimizu (Department of Neurology, University of Tokyo) for expert assistance in the histologic analysis of biopsy samples.
- 36Identification of C1q as the heat-labile serum cofactor required for immune complexes to stimulate endothelial expression of the adhesion molecules E-selectin and intercellular and vascular cell adhesion molecules 1. Proc Natl Acad Sci U S A 1995; 92: 8378–82., , , , , , et al.