10.5.1. CCR5, Genetics, and RA
As mentioned earlier, a 32-base pair (bp) deletion in the CCR5 gene (A32) results in a frameshift and premature termination in the region encoding the second extracellular loop of the CCR5 receptor, producing a nonfunctional receptor that remains intracellular (67). About 1% of the Caucasian population is likely to be homozygous for A32 (68). Notably, Gomez-Reino et al. (69) reported that in their cohort of 673 Spanish patients with RA, there were no patients homozygous for A32, but the frequency of this A32 mutation was normal in lupus patients and healthy controls. This striking observation was not replicated later on in four other studies performed in different populations in which RA patients homozygous for A32 were found, suggesting that CCR5 deficiency does not prevent rheumatoid arthritis (70-73). A recent study of 516 RA patients and 985 healthy controls (74) also failed to find A32 homozygous RA patients, while 1.2% of the healthy controls were homozygous.
Although having a A32 allele may not totally prevent the development of RA, it may influence disease severity and progression. There are several contradictory reports that have addressed this question. In 163 Danish patients with RA, those carrying the A32 allele were more likely to be seronegative for RF (suggesting a more benign disease) and had less swollen joints (71). This conclusion has been substantiated in another study (72). In contrast with these results, a more detailed study using 5 years of follow-up data (including hand and foot radiographs, C-reactive protein (CRP) levels, and RF) from 438 patients found that the A32 allele does not predict disease progression (73). This conclusion is also supported by two additional studies (74,75).
Collectively, the data from genetic-association studies regarding the effect of A32 state on the risk of developing the disease, RA progression, severity, as well as the data from experimental models of RA, seem to suggest that CCR5 may play a minor role in RA pathogenesis and that according to the circumstances, its effects may or may not be noticeable (i.e., large enough cohort of specific ethnic origin or the use of B6X129 mice for CIA instead of the more susceptible DBA/1j mice).
CCR2 is expressed by a large proportion of mononuclear cells present in inflamed joints and in synovial effusions of patients with different forms of arthritis (76,77) and in adjuvant-induced murine arthritis (78). The cognate ligands of CCR2 (CCL2) and CCR5 (CCL3, CCL4, and CCL5) have been detected in the synovial fluid of patients with RA (79-82), at sites of bone remodeling (83), in articular cartilage (84), and in inflamed joints in CIA (85). Also, analogues of CCL2, neutralizing antibodies against CCL2, and DNA vaccines that encode these chemokines have all been shown to inhibit disease initiation or progression in rodent models of arthritis (Tables 4 and 5).
CCR2-null mice exhibit a profound defect in induced leukocyte migration, particularly monocyte/macrophage (86,87), and DC migration associated with skewing toward T helper 2 (Th2) cytokine production (88). Likewise, blockade of CCR2 with monoclonal antibody MC-21 markedly impairs monocytes/macrophage recruitment in vivo (89). Because of these defective inflammatory responses and the fact that CIA can be ameliorated by administration of Th2 cytokines (90), it was hypothesized that CCR2-null mice would be resistant to the development of CIA or its manifestations and that the blockade of CCR2 with monoclonal antibodies would result in a similar outcome. However, CCR2-null mice immunized with type II bovine collagen (CII) not only were fully susceptible to CIA but also showed accelerated and more severe disease than wild-type (WT) control mice (43). CII-immunized CCR2-null mice had significantly higher serum levels of anti-collagen-specific IgG1 and IgG2a, increased levels of total IgE, and even positive titers of RF. Diseased joints in CCR2-null mice showed severe bone destruction and increased numbers of neutrophils and osteoclasts. The lymph nodes in these mice contained higher numbers of activated T and B cells, including an increase in the number of RANKL-bearing T cells. The severe disease phenotype of CCR2-null mice was transferred using bone marrow cells to lethally irradiated WT mice suggesting that absence of CCR2 in the hematopoietic compartment was responsible for the increase susceptibility of CCR2-null mice to CIA. Furthermore, in a complementary model of arthritis based on the intravenous administration of anti-CII antibodies, it was shown that WT and CCR2-null mice had comparable disease during the acute phases of the disease (first 2 weeks after anti-CII administration); the resolution phase of the disease that commonly begins 3 weeks after Ab injection was delayed in CCR2-null mice (43).
Additional support for a negative regulatory role of CCR2 in arthritis was partially provided by blockade of CCR2 with specific monoclonal antibody (91). Blockade during disease initiation (days 0 to 15) significantly reduced the severity of arthritis, whereas blockade of CCR2 in late phase of the disease (e.g., days 21 to 36) markedly aggravated clinical and histologic signs of arthritis. Blockade of CCR2 in early disease phases resulted in lower serum levels of anti-collagen-specific IgG, whereas the late blockade of CCR2 markedly increased the humoral immune response against collagen, induced a higher influx of monocytes and neutrophils into the joints, and increased bone and cartilage erosion.
Despite the bulk of evidence pointing to a negative regulatory role for CCR2 in experimental arthritis, results obtained using a novel rodent CCR2 antagonist, INCB3344 (Incyte Corporation Wilmington, Delaware), seem to challenge that notion (92). INCB3344 was reported to be 100-fold more selective for CCR2 than for other closely related chemokine receptors. Administration of this small-molecule antagonist starting at day 9 after adjuvant treatment significantly reduced arthritis severity in rat adjuvant-induced arthritis. INCB3344 treatment was associated with 82% inhibition of joint inflammation and 64% reduction in bone reabsorption.
Interestingly, in a potentially new murine model, germ-line inactivation of CCR2 was also associated with increased susceptibility to arthritis development. Intradermal Mycobacterium avium challenge of WT and CCR2-null mice on DBA/1j and BALB/c background was associated with the development of clinically evident chronic persistent polyarthritis only in the CCR2-null DBA/1j mice but not in the WT mice nor the CCR2-null mice in the BALB/c background (93). Increased susceptibility to infection was unlikely to be the only cause of arthritis as BALB/c CCR2-null mice had a high degree of susceptibility to the pathogen but did not develop arthritis. Histologic analysis revealed that the arthritic phenotype present in the DBA/1j CCR2-null mice was characterized by synovial hyperplasia, synovitis, synovial cyst formation, fibroblast proliferation, bone remodeling, osteoid deposition, and osteoclast activity. Interestingly, this disease phenotype was highly reminiscent of the autoimmune arthritis present in CCR2-null mice immunized with collagen (43). Indeed, immunologic studies showed that compared with DBA/1j WT, BALB/c WT, or CCR2-null, DBA/1j CCR2-null mice had significantly higher titers of anticollagen type II antibodies of IgG1 subtype.
How does one interpret the data available so far regarding the role of CCR2 in RA? Arguably, the more contrasting data relate to that between the severe CIA phenotype in CCR2-null mice and the antiarthritic effects of INCB3344 in the adjuvant-induced arthritis model. A working hypothesis is that the phenotype of the KO mouse is driven by immune abnormalities existing prior to immunization and that cannot be fully recapitulated after acute treatment with an antagonist that is initiated after the disease has begun. For instance, compared with the WT mice, CCR2-null mice have substantially lower numbers of specific monocyte subtypes in the blood (94) and DCs in the spleen (88). These types of cellular changes are likely to be adaptations occurring as a consequence of the chronic absence of a key molecule in the system. If this hypothesis is correct, long-term administration of the CCR2 antagonist or neutralizing Ab may lead to a phenotype that may be concordant with that of CCR2-null mice.
Intriguingly, results from a recent clinical trial in RA patients suggest that the role of chemokines in RA may be proinflammatory. CCL2 is the major ligand for CCR2, and administration of a CCL2 blocking Ab was not associated with any clinical improvement in these subjects. On the contrary, these trials found an increase in the levels of C-reactive protein (a disease activity marker) and enhanced accumulation of macrophages in the joint (95).
Was this article helpful?