Rheumatoid arthritis (RA) is an inflammatory disorder characterized by infiltration of monocytes, T cells, and polymorphonuclear cells into the synovial joints. The CC chemokines monocyte chemoattractant protein 1 (MCP-1)/CCL2, macrophage inflammatory protein 1α (MIP-1α)/CCL3, MIP-1β/CCL4, and RANTES/CCL5 in part account for the presence of monocytes and T cells in RA synovial fluid (SF) and synovial tissue (ST) (1–3). RANTES, MIP-1α, MCP-1, and MIP-1β are chemotactic for monocytes and T cells. RA synovial fibroblasts produce RANTES, MIP-1α, MCP-1, and MIP-1β upon stimulation by tumor necrosis factor α (TNFα), interleukin-1α (IL-1α), or IL-1β (4). Chemokines are chemotactic cytokines that induce their effects by binding to an array of G protein–coupled receptors (GPCRs) (5, 6). It is known that CC chemokine receptors (CCRs) differ mainly in their NH2-terminal portions, which might explain the different specificity of the CC chemokines for particular receptor interactions. An additional factor which may enhance the specificity of CC chemokines for target cells is the cellular distribution of the receptors. Moreover, most CCRs bind to multiple CC chemokines with different affinities (3, 7–9).
Findings from our laboratory indicate that CCR1, the receptor for RANTES/MIP-1α, is expressed on monocytes in normal and RA peripheral blood (PB) and on a minority of SF monocytes. This may imply that CCR1 is important in the initial recruitment of monocytes from the circulation to sites of inflammation (10). CCR2, the receptor for MCP-1, is mainly expressed on monocytes in normal and RA PB. Moreover, the percentage of T cells expressing CCR2 is elevated in RA PB compared with normal PB. Memory T cells have higher CCR2 expression than naive T cells in RA PB. Investigators in our group and others have reported that CCR5, the receptor for RANTES, MIP-1α, and MIP-1β, is expressed on a greater percentage of monocytes and memory T cells in RA SF compared with RA or normal PB (3, 8, 10). Researchers in our laboratory have also demonstrated CCR5 immunostaining on a majority of RA ST macrophages, fibroblasts, vascular smooth muscle cells, and perivascular lymphocytes. Results of studies using RA synovial fluid, peripheral blood, and synovial tissues, as well as the results of the present study, strongly suggest that specific CCRs play a critical role in RA and its animal models.
Rat adjuvant-induced arthritis (AIA) is a commonly used model of RA, autoimmune disease, and inflammation. An advantage of using this model is the ability to assess the role of several pathogenic factors, including chemokines and their receptors, at various phases of the disease, which is generally not feasible in humans. In this study, we have quantified the β-chemokines and CCRs in rat AIA. Since CCR1, CCR2, and CCR5 messenger RNA (mRNA) expression levels were up-regulated with peak inflammation in the AIA model, we examined whether these CCRs were activated by phosphorylation on tyrosine residues. Identifying the key intermediate proteins in the signal transduction cascade associated with inflammatory cytokines (IL-1, TNF, and IL-6) or CCRs is important and may have therapeutic potential in inflammatory diseases such as RA. It is known that IL-6 signals through the JAK/signal transducer and activator of transcription (STAT) pathway (11). CCRs dimerize upon chemokine binding and associate with the downstream pathways. In this report, we present evidence for the association of CCRs with JAK/STAT-1/STAT-3 pathways in AIA rat joints; furthermore, we demonstrate cell type expression of these proteins.
- Top of page
- MATERIALS AND METHODS
In this study, we induced AIA and harvested rat joints at different time points after injection of adjuvant. Joint mRNA was analyzed by TaqMan real-time PCR. The data presented in Figure 1A demonstrate that the CCR1 mRNA concentration in rat AIA increases progressively from day 14, peaks on day 18, and decreases thereafter. Interestingly, the ingress of macrophages and lymphocytes into the joints reported in AIA (16) follows the same pattern of expression. CCR1- and CCR2-immunoreactive cells were found in RA ST and colocalized with CD68+ macrophages (10). In the present study, CCR1 mRNA levels increased with peak inflammation in the rat AIA joint. This contrasts with the surface expression of CCR1, which was found to be decreased in human RA SF and PB compared with human normal SF and PB (10). CCR1+ monocytes may be depleted from the PB as they are recruited into the joint.
As shown in Figure 1B, CCR2 showed the same pattern of mRNA expression as CCR1 in AIA. CCR2 and CCR1 have also been shown to have the same pattern of protein expression in human RA PB and SF (10). Results from our laboratory also indicate that CD3+ lymphocytes expressing CCR2 are significantly elevated in RA PB compared with normal PB (23). Since CCR1 expression and CCR2 expression increase with peak inflammation in the AIA model, they may be important for initial recruitment of monocytes and T cells from the circulation to the site of inflammation. Additional evidence supporting the importance of CCR2 in the initial phases of inflammation is that MCP-1 (one of the ligands for CCR2) has 3 times more immunoreactivity on day 18 than on day 25 in rat AIA (following the same pattern of expression as that of CCR2 in rat AIA) (24).
CCR4 mRNA expression differs from that of CCR1 and CCR2 (Figure 1C). CCR4 expression increases successively until day 18 and then is rapidly up-regulated until day 24, dropping significantly on day 29. Investigators in our group also showed that CCR4 was expressed on significantly more monocytes in RA PB than in RA SF or normal PB (10). CCR4 has been associated with Th2 cells, which produce cytokines that may attenuate RA inflammation.
The CCR5 mRNA expression detected in rat AIA correlates with peak inflammation on day 18 (Figure 2A). Additional data from our laboratory indicate that RA SF monocytes exhibit the highest CCR5 expression compared with RA and normal PB monocytes. Interestingly, joint patterns of TNFα protein expression in rat AIA correlate well with CCR5 mRNA expression in the same model (25). Theoretically, CCR5 may be down-regulated after encountering the ligands in the inflamed joint, possibly by ligand binding or cytokine modulation. The mRNA expression of MIP-1β (the ligand for CCR1 and CCR5) in AIA joints follows the same bell-shaped curve as those of the expression of CCR1 and CCR5, with maximum expression on day 18 (Figure 2B). Previously, investigators in our group also showed that immunostaining for MIP-1β in RA ST was predominantly on macrophages (colocalizing with CCR1 and CCR5) (10). The greatest MIP-1α mRNA expression was detected on days 7 and 18, and lower mRNA concentrations were found on days 14, 21, 24, and 29 (Figure 2C). The overproduction of synovial MIP-1α mRNA levels on day 7 precedes clinical symptoms and thus may contribute to recruiting monocytes prior to day 18. Furthermore, MIP-1α mRNA levels follow the same pattern as that of total leukocyte counts in AIA. In fact, the pattern of expression of MIP-1α mRNA is similar to that of MIP-1α protein in rat AIA reported by us previously (25).
The binding of CC chemokines to GPCRs follows a series of events including dimerization, phosphorylation, and association with downstream signaling pathways (26). In this study, we observed phosphotyrosine activation of CCR1 (Figures 3A and B), CCR2 (Figures 3C and D), and CCR5 (Figures 4A and B). Interestingly, the duration of CCR1 tyrosine phosphorylation (days 14, 18, 21, and 24) is longer than that of CCR2 (days 14 and 18) or CCR5 (days 14, 18, and 21). Nevertheless, the baseline CCR1 mRNA levels are lower than those of the other two receptors. Furthermore, our data demonstrate higher mRNA levels and shorter tyrosine phosphorylation duration for CCR2 compared with CCR5. All 3 CCRs are tyrosine phosphorylated on days 14 and 18, which is the period between initial and peak inflammation.
The data suggest that tyrosine phosphorylation is triggered in the early phases of inflammation and that it coincides with up-regulation of CCR mRNA (11). Recently, it has been reported that CCRs, including CCR2 and CCR5, signal through the JAK/STAT pathway. In the CCR5-transfected HEK 293 cell response to RANTES, CCR5 is rapidly tyrosine phosphorylated, and JAK-1 associates with CCR5. JAK-1 association in response to RANTES promotes STAT-5b transcription factor association to the receptor, as well as its activation (11, 27). A CCR2 dominant-negative mutant that retains its capacity to form homodimers in response to MCP-1 but that cannot be tyrosine phosphorylated is unable to recruit and trigger JAK-2 phosphorylation (28).
In this study, we investigated the JAK/STAT signaling pathway associated with CCR1, CCR2, and CCR5. CCR1 is associated with JAK-1/STAT-1/STAT-3 phosphorylation on days 14 and 18 (Figures 4C, 4D, and 5), which coincides with its initial mRNA and phosphotyrosine up-regulation. CCR2 signals through the JAK-2/STAT-1/STAT-3 pathway at peak inflammation (Figures 6, 7A, and 7B). JAK-1 association with CCR5 on day 18 (Figures 7C and D) is subsequent to the initial CCR5 phosphorylation on day 14 (Figures 4A and B). Moreover, both STAT-1 and STAT-3 (Figure 8) demonstrate an increase in tyrosine phosphorylation associated with CCR5 activation on days 18 and 21.
These results suggest that CCR5 expression is up-regulated starting on day 14, and that CCR5 is simultaneously tyrosine phosphorylated. The association of STAT-1 and STAT-3 with CCR5 on days 18 and 21 correlates with JAK-1 phosphorylation and binding on day 18. It is noteworthy that CCR1 is associated with activation of the JAK/STAT-1/STAT-3 pathway during early to peak inflammation. While CCR2 signals through the same pathway only during peak inflammation, CCR5 activation is associated with the JAK/STAT pathway during peak to later stages of inflammation. Nevertheless, all 3 receptors are associated with the JAK/STAT family on day 18. This implies that membrane-localized protein tyrosine kinases are recruited to CC ligand–stimulated receptors (phosphorylated receptors), where they are activated and act in concert to initiate intracellular signaling cascades during maximal inflammation.
Shouda et al demonstrated increased STAT-3 phosphorylation in mouse collagen-induced arthritis at peak inflammation and a significant down-regulation within 10 days (29). Furthermore, the simultaneous tyrosine phosphorylation of STAT-1 and STAT-3 may be important for STAT-1/STAT-3 dimerization. In accordance with findings of these studies, we and others have shown that the phosphorylation of STAT-1 and STAT-3 is significantly higher in RA ST than in normal ST (29, 30). Consistent with our findings, it has been shown that STAT-1 decoy oligodeoxynucleotide and adenovirally delivered dominant-negative STAT-3 gene administered intraarticularly in animals with arthritis significantly suppressed the arthritis score and paw volume and clinical arthritis parameters (29, 31). TNFα induces suppressor of cytokine signaling 3 (SOCS-3), which is an endogenous JAK kinase inhibitor. It appears that SOCS-3 functions as a component of a negative feedback loop to dynamically terminate cytokine-mediated signals. It is noteworthy that the proinflammatory cytokines, such as TNFα and IL-6, signal through the same pathway as CCRs.
Immunohistochemical analysis demonstrated that CCR5, p-STAT-1, and p-STAT-3 were expressed on synovial lining cells, macrophages, and endothelial cells in arthritic rat ankles on post–adjuvant injection day 18. While the majority of the CCR5 and p-STAT-1 immunostaining was on synovial lining cells and macrophages, p-STAT-3 was predominantly expressed on endothelial cells. The expression of CCR5 and p-STAT-1 on synovial lining cells and macrophages might indicate that this pathway is involved in proinflammatory cytokine production by synovial macrophages. CCR5-deficient mice show reduced cytokine production by peritoneal macrophages following stimulation with lipopolysaccharide plus interferon-γ. Consistent with findings of the present study, investigators in our group previously showed that the CCR5-immunoreactive cells in RA ST were predominantly macrophages (10), indicating that CCR5 might have a role in monocyte migration into the ST. STAT-3 phosphoactivation on endothelial cells might be important for proliferation and chemoattraction of inflammation mediators. Tyrosine phosphorylation of STAT-1 and STAT-3 at peak inflammation in rat AIA coincides with activation in the synovial lining layer, considered an area most actively involved in chronic synovitis, and may reflect the in situ production of TNFα and IL-1.
In conclusion, CCR1, CCR2, and CCR5 mRNA levels are up-regulated and activated with peak inflammation. We demonstrate that CCR1, CCR2, and CCR5 tyrosine phosphorylation are associated with the JAK/STAT pathway and with macrophage and endothelial cell infiltration in rat AIA. Targeting this signaling pathway may provide a novel therapeutic avenue in RA.