The results of this study show an increased expression of CCR9 on circulating monocytes in the blood of patients with RA. Not only the number of CCR9 receptors per monocytic cell increases but also the percentage of monocytic cells bearing this receptor increases. Furthermore, in the synovium there is an increase in the number and percentage of macrophages that express CCR9 in RA. The percentage of CCR9+ monocytes in RA synovium (81%) is greater than that in blood (40%), similarly in non-RA the percentage of CCR9+ monocytes in the synovium (66%) is increased compared to that in blood (16%). This suggests there may be a preferential recruitment of monocytes bearing this receptor into the synovium or an up-regulation of the receptor, possibly by cytokines (for example, TNF), once the cells have migrated into the tissue.
Monocytes migrate from the blood into tissues and differentiate into resident macrophages. This occurs via sublining blood vessels in the normal and RA synovium. However, in the latter the number of monocytes recruited and the accumulation of macrophages is markedly increased. When monocytes differentiate into macrophages they up-regulate pattern recognition receptors, such as scavenger receptors and toll-like receptors [31
]. Like other scavenger receptors CD36 acts as a phagocytic receptor [32
] and was used in the current study as a marker of monocyte to macrophage differentiation [33
]. In addition CD36 has been detected on macrophages in the RA joint [34
A major finding of the current study was the stimulatory effect of CCL25 on monocyte differentiation. The results show that in RA the basal CD36 levels on monocytes are similar to those of normal monocytes. However, upon CCL25 stimulation RA monocytes are significantly more responsive in terms of CD36 expression. This increased responsiveness may be related to the increased abundance of CCR9 found on RA blood monocytes, as shown by the MFI flow cytometry data. In addition, the increased responsiveness was not only CCL25-related since PMA, a known inducer of monocyte differentiation [35
], also elicited a greater response in RA monocytes compared to normal. CCL25 appears to function like CXCL4 which is another chemokine reported to stimulate the differentiation of monocytes to macrophages [36
]. It is unknown, as yet, whether CCR9 is modulated during monocyte differentiation and differentially expressed by polarized (M1 and M2) macrophages, together with the effect of CCL25 on markers of macrophage polarization.
CCL25 was originally defined as being chemotactic for activated, but not resting, human monocytic THP-1 cells, and activated mouse peritoneal macrophages [23
]. This implies the existence of CCR9 on activated monocytes and macrophages, although this was not formally shown experimentally. In order to further this study of Vicari et al.
], we showed that activation of THP-1 cells with TNFα results in increased CCR9 expression. In RA, monocytes in the circulation show evidence of increased activation prior to their entry into the synovium, as well as in the synovium itself [10
]. Therefore the increase in the expression of CCR9 may be related to the enhanced activation state of monocytes/macrophages in the RA circulation and synovial tissue. The mechanism behind the up-regulation of CCR9 on monocytes in the blood is unknown but may relate to elevated cytokines, such as TNFα, in the circulation which occurs in RA [37
]. Alternatively the increased CCR9 expression on monocytes may be due to changes occurring in the bone marrow which are known to occur in the disease [38
The level of CCR9 expression on blood monocytes in RA and healthy patients was compared with other chemokine receptors using flow cytometry. In terms of MFI and the percentage positive cells, CCR9 expression was similar to CCR5 and CXCR4, although was not as highly expressed as CCR2 and CCR1.
CCR9 is described as an important chemokine receptor in the gut environment where it is constitutively expressed in the majority of CD4+ and CD8+ T lymphocytes of the small intestine [15
]. In the circulation it is also expressed by discrete subsets of memory CD4+ and CD8+ lymphocytes expressing the intestinal homing receptor α4β7 [15
]. Adoptive transfer models and in vivo
neutralisation experiments have revealed that CCR9 and CCL25 play important roles in the localisation of effector T cells to the small intestinal mucosa [43
]. In the present study, CCR9 was not detected in T lymphocytes in the RA and non-RA synovium. This suggests that the CCR9+/α4β7+ subset of T cells are not recruited to the synovium and other chemokine receptors and adhesion molecules are important for T cell migration into this tissue [17
]. Furthermore, our data support the notion of a tissue-specific address code for T cell recruitment to the synovium that is different from the gut. In inflammatory gut disease, CCR9+ T cells are more abundant in the PB of patients with celiac disease or Crohn's disease [42
]. We observed a similar pattern for monocytes in RA patients. However, the percentage of CCR9+ T lymphocytes was shown to be reduced in the small intestine with Crohn's disease [42
] whereas in the present study the percentage of CCR9+ macrophages in RA ST did not reduce compared with normal, but was increased.
CCL25 has been shown to be produced by dendritic and epithelial cells in the normal thymus [23
]. In addition, in the small intestine CCL25 is constitutively expressed by epithelial and endothelial cells [17
]. In inflamed small intestine in Crohn's disease CCL25 expression is in epithelial cells in proximity to lymphocytic infiltrates and is not detectable on endothelial cells [42
]. In the present study CCL25 localised to CD14+ and CD68+ cells in the synovial sublining indicating that macrophages are a source of this chemokine. Furthermore the production of CCL25 by synovium is supported by the presence of CCL25 mRNA in RA and non-RA synovial tissue, although levels were low. Our results suggest a positive autocrine loop. Macrophages present in the synovial sublining produce CCL25 leading to the local stimulation of monocyte differentiation to macrophages. Interestingly CD14+ macrophages are likely to represent a recently recruited immature subpopulation of macrophages in the synovial sublining, losing CD14 expression upon maturity [3
]. Therefore CCR9 on these CD14+ cells may play a role in the differentiation of recently immigrated monocytes. CCL25 was also present in non-inflamed control synovia and this chemokine stimulated CD36 up-regulation by monocytes from normal healthy donors, although less so compared to RA. Therefore it is possible that CCL25 may enhance the differentiation of monocytes to macrophages under constitutive conditions in the non-inflamed synovium. The effect of CCL25 on monocyte chemotaxis was not consistent; therefore this chemokine may be more effective at influencing differentiation rather than monocyte recruitment.
A very recent in vivo
paper by Jacobs et al.
] reported the use the K/BxN serum-transfer model of arthritis in mice. CXCR2 was found to be critical for the development of autoantibody-mediated arthritis, and neutrophil recruitment to the joints. This model reflects the effector phase of arthritis rather than the initial adaptive immune response. Other chemokine receptors such as CCR9, CCR1-7, CXCR3, CXCR5 and CX3CR1 were not critical in the model. Many of these receptors have been shown to play a role in arthritis using alternative models that involve an adaptive immune response, probably functioning at that stage. There have been no studies targeting CCR9 in these adaptive immune response models and it would be interesting to perform such a study. An alternative approach to assess the role of CCR9 in vivo
in RA would be to perform a human study. There is currently a CCR9 antagonist that shows promise in clinical trials for Crohn's disease [49
], in fact this is one of the few clinical trials targeting chemokine receptors that show hope in the treatment of inflammatory diseases. So a clinical study using this antagonist in human RA would be of interest, also overcoming problems of lack of exact correlation between animal models of RA and human disease.