Imprinting of T cells with a small intestinal homing address code (i.e. CCR9) is mediated by retinoic acid (RA)18
. RA in the presence of TGF-β additionally determines the induction of FoxP319
, while TGF-β also promotes the expression of CD10320
. Yet although the induction of these regulatory molecules shares common pathways, little is known about the role of the CCL25/CCR9 axis for Treg trafficking and regulation of immune responses in the small intestine. Here we show high concordant expression of CCR9, FoxP3, and CD103 in Tregs, suggesting that their induction might be coupled within inductive compartments. As CCL25 expression is restricted to small intestine, we evaluated its role in the TNFΔARE model that develops inflammation in the distal small intestine. We show that CCR9+
T cells undergo late inflammation-driven expansion. Furthermore, during late disease, the amount of CCR9 mRNA transcripts and CCL25 protein levels increased in TNFΔARE mice ilea, yet CCL25 protein expression remained localized to the ileal epithelium. Evaluation of disease severity in CCR9-deficient mice revealed more severe disease than that of CCR9-sufficient counterparts. Thus, we investigated whether a preferential dependence by Tregs on the CCR9/CCL25 axis for homing into the chronically-inflamed ileum may account for the accelerated disease. We noticed a deficiency of CD4+
Tregs within the LP and MLN of CCR9-deficient mice. In addition, Tregs preferentially expressed CCR9 on their surface than their non-regulatory counterparts and the CD8+
subset decreased the proliferation of CD4+
T cells in vitro
. Moreover, anti-CCR9 antibodies exacerbated disease and altered the pro- to anti-inflammatory CD8+
T cell ratio, confirming a predominant role for CCR9/CCL25 for the regulation of ileitis.
CCR9-expressing cells are increased in peripheral blood of patients with small bowel CD5
, thus inflammatory signals might modulate their frequency. The expression of CCR9 is known to be induced by RA18
, while a high antigen dose is suppressive21
. It is not known whether inflammatory signals (e.g. cytokines) might alter CCR9 expression. Our data suggests that during the late disease of TNFΔARE mice either CCR9 is induced (predominantly in CD8+
T cells) or survival/retention signals result in increases in the number of CCR9-expressing cells within the MLN and LP. Furthermore it appears that CCR9 might be upregulated as the CD8+
T cells migrate out of the LP into the MLN and the periphery. Thus, CCR9 expression appears to be a dynamic process, with cells being able to upregulate and downregulate surface expression perhaps after ligand binding, resulting in internalization, as shown with other chemokine receptors (e.g. CCR7). This may account for the lower mean fluorescence intensities observed in the LP22
. As we utilized CCR9-deficient lymphocytes as controls and two different mAb clones, our studies convincingly report the true expression of this molecule within the different subsets and lymphoid compartments.
As it has been shown for other chemokines, CCL25 is likely transcytosed across endothelium to be presented at the apical surface23
where it induces adhesion of CCR9-expressing cells24
. Although CCL25 has been considered a homeostatic chemokine25
more recent data have led to the suggestion that its expression may be inducible in intestinal inflammation26, 27
. CCL25 is additionally expressed aberrantly in the liver of patients with primary sclerosing cholangitis6
. Yet the specific inflammatory signals that induce its expression in intestine or liver remain to be defined. Ericsson and colleagues exposed small intestinal cell lines to TNF and IFN-γ and mice to LPS in vivo
and although they observed increased expression of other chemokines, CCL25 remained unchanged28
. Although our data might suggest that TNF (overproduced systemically in TNFΔARE mice), might be responsible, we observed localized induction within the affected terminal ileum rather than generalized small bowel expression, suggesting that perhaps a combination of mediators within the effector organ triggers expression.
Few studies have looked at the role of CCL25/CCR9 pathway in models of IBD and the limited available literature appears discordant. We have previously evaluated its role in the SAMP1/YitFc model, which spontaneously develops Crohn's-like ileitis and found that the therapeutic efficacy of CCR9 immunoblockade was limited to early disease27
. However, other work has shown that CCR9 deficiency had no effect on the severity of ileitis in TNFΔARE mice9
, whereas a small molecule inhibitor of CCR9 prevented disease onset29
. By contrast our studies in TNFΔARE mice showed that CCR9 deficiency exacerbates ileitis. Several differences may account for the discrepancy. First, we evaluated the entire course of the disease from 4- to 20-weeks-of-age, whereas the prior study evaluated disease between 8- and 14-weeks-of-age. Second, while the prior study evaluated the histological severity of disease in cohorts of 3-6 mice, our experimental cohorts included a minimum of 17 mice9
. Finally, in our studies all mouse strains were on C57BL6/J background and the genetic background affects the severity of IBD in animal models30
Tregs, defined by their expression of the transcription factor forkhead box P3 (FoxP3)31
and production of IL-10 accumulate in intestine of patients with IBD32
. Expansion of CD4+
T cells has also been reported in animal models33
. The molecular basis of their recruitment to the inflamed small intestine has yet to be elucidated. There is evidence to support that CD4+
cells from the small intestinal LP may express more CCR9 on their surface than similar cells from the spleen34
and a subset of CD4+
cells possesses a Treg cytokine profile7
. We hypothesized that Treg deficiency in CCR9 deficient mice, due to an increased Treg dependence on CCR9 to traffic to the ileum resulted in exacerbated inflammation. In fact Treg frequency in CCR9-deficient mice was similar to that of controls and lower than in TNFΔARE mice, despite their worse inflammation, supporting a reliance on CCR9/CCL25 for Treg traffic to the inflamed small intestine.
T cell subsets in TNFΔARE mice have either an effector or a regulatory function14, 15,16
. The latter are characterized by expression of CD103 16
which is present murine and human CD8+
T cells that home to mucosal surfaces. CD4+
T cells that express CD103 exert a regulatory role in colitis35
. We have previously described that the CD8+
subset inhibited the proliferation of CD4+
cells in vitro
and exerted a regulatory role during TNF-mediated ileitis16
. Here we observe an imbalance between these subsets in CCR9-deficient TNFΔARE mice, which exhibited increased proinflammatory CD8+
T cells while displaying decreased pro-regulatory CD8+
. This imbalance has implications on the course and severity of disease due to the large disparity between effector and regulatory subsets in CCR9-deficient TNFΔARE mice.
T cells induce colitis in RAG-deficient or SCID mice30
, whereas co-transfer of Tregs prevented and cured IBD33
, highlighting the critical role of Tregs for maintenance of homeostasis. Lymphocyte recruitment to the small intestine requires local expression of CCL25 that interacts with CCR9 on the T cell surface. However, little is known about the specific homing determinants of Tregs. We show that in TNFΔARE mice, CD4+
Tregs show preferential expression of CCR9 compared with their non-regulatory counterparts. The potential dependency on this pathway by Tregs for recirculation to terminal ileum may contribute to the observed exacerbation of disease resulting from genetic ablation of CCR9.
We confirmed that CCR9-expressing CD8+ T cells inhibited the proliferation of CD4+ T cells in a dose-dependent manner. To further define whether this was also true in vivo we tested the effect on an anti-CCR9 mAb on disease severity. Antibody treatment significantly worsened all histological parameters and significantly reduced the percentage of the CD8+/CD103+ Tregs in ilea and MLN and increased the effector CD8+/CD44+ subset, confirming a critical role for CCR9-expressing cells for the regulation of ileitis.
Recent trials examined the efficacy of Traficet EN, CCX282-B, which targets the CCL25/CCR9 axis in patients with CD8
. The preliminary results appear promising both for induction and maintenance of remission (http://www.drugs.com/clinical_trials/chemocentryx-reports-traficet-maintains-remission-crohn-s-digestive-week-ddw-2010-conference-9318.html
). In addition, preclinical data on the preventive effect of Traficet EN has been published36
. As the chemokine is not expressed in normal or inflamed colon, it is interesting that the above studies did not focused on patient with ileitis but also included colonic and ileocolonic CD. Our observations of exacerbated disease and decreased Tregs after treatment with an anti-CCR9 mAb or genetic ablation of CCR9 supports the hypothesis that CCR9 plays a major role in the trafficking of Tregs to the chronically inflamed small intestine. Thus, the CCR9/CCL25 axis appears to be more important for the regulation of immune responses than for its induction in this TNF-mediated ileitic model.