Unique Chemotactic Response Profile and Specific Expression of Chemokine Receptors Ccr4 and Ccr8 by Cd4⁺Cd25⁺ Regulatory T Cells

The efficient operation of the immune system is critically dependent on a complex series of cellular interactions and movements to specific locations. Chemokines are small chemotactic cytokines characterized by critically positioned cysteine residues, which recruit distinct leukocyte subsets to sites of inflammation and specific microenvironments within secondary lymphoid tissues 1,2. Chemokine receptor expression is exquisitely regulated depending on the stage of activation and differentiation of T cells and coordinates tissue localization and encounters with APCs 3,4,5.

One of the fundamental features of the immune system is its ability to preserve a delicate balance between effector responses and mechanisms of immunoregulation. Given the central role of chemokines in the regulation of immunity, we postulated that some of these molecules could participate in turning off adaptive immune responses by recruiting cells with immunoregulatory functions. Various subpopulations of CD4⁺ T cells that can mediate immunosuppression both in vitro and in vivo have been described 6,7,8. Recent studies demonstrated that CD4⁺CD25⁺CD45RB^low memory T cells expressing cytotoxic T lymphocyte antigen (CTLA)-4 are regulatory T cells (Treg) that mediate tolerance to organ-specific self-antigens and can prevent autoimmunity and intestinal inflammation in mice 9,10. Recently, human peripheral blood CD4⁺CD25⁺ T cells have also been characterized as suppressor T cells 11. We now report the unique chemotactic response profile of human blood-borne CD4⁺CD25⁺ Treg cells. Treg cells specifically express the chemokine receptors CCR4 and CCR8 and respond to the chemokines macrophage-derived chemokine (MDC/CCL22), thymus and activation-regulated chemokine (TARC/CCL17), I-309/CCL1, and the virokine vMIP-I, which are agonistic ligands of these receptors. Mature DCs producing CCL17 and CCL22 were found to preferentially attract Treg cells, suggesting involvement of these chemokines in Treg cell function.

CCL1, CCL22, CCL17, vMIP-I, CCL11, CXCL11, and CXCL12 were purchased from Dictagene, whereas CXCL13, CCL19, CCL2, CCL3, CCL4, CCL20, and CCL5 were purchased from R&D Systems. Abs for various surface molecules were from BD PharMingen unless indicated otherwise. Goat anti–human CCR4 Ab was from Research Diagnostics Inc. Goat anti–human CCR8 Ab was from Alexis Corporation. Mouse anti–human CD4 Ab was from Novocastra Laboratories Ltd. All the conjugated and unconjugated secondary Abs for immunofluorescent stainings were from Jackson ImmunoResearch Laboratories.

Human CD4⁺ T cells were purified from healthy donor peripheral blood by Ficoll-Paque (Amersham Pharmacia Biotech) density gradient centrifugation followed by isolation with immunomagnetic beads (CD4⁺ T cell isolation kit; Miltenyi Biotech). CD4⁺ T cells were stained with PE-labeled anti-CD4 and FITC-labeled anti-CD25 mAbs and sorted using a FACStar™ (Becton Dickinson). Purity was routinely >95%. Human monocytes were purified from PBMCs by isolation with immunomagnetic beads (Monocyte isolation kit; Miltenyi Biotech) and cultured in RPMI 1640 with 5% FetalClone I (HyClone), 2 mM L-glutamine, 1 mM sodium pyruvate, 100 U/ml penicillin-streptomycin (Sigma-Aldrich; complete medium). Dendritic cells (DCs) were derived from human monocytes by culturing them in complete medium in the presence of 10 ng/ml of IL-4 (BD PharMingen) and 800 U/ml of GM-CSF (Myelogen; Schering-Plough) for 6 d. Maturation of DCs was induced by stimulation with 1 μg/ml of LPS in fresh complete medium. After 16 h of stimulation, DCs were washed and cultured for an additional 24 h in complete medium. Finally, the supernatants were harvested and tested in chemotaxis.

Chemotaxis assays were performed using 5-μm pore polycarbonate filters in a Transwell chamber (Corning Costar Corporation) as described 12. After migration, 1.35 × 10⁴ polystyrene beads (Polysciences Inc.) were added in the lower compartment of the Transwell. Migrated CD4⁺ T cells were recovered and stained with anti-CD4, anti-CD25, and anti-CD45RO mAbs. The number of migrated T cells was measured by flow cytometer acquisition of a fixed number of beads. For the CCL17 and CCL22 preclearing experiment, 2 ml of DC supernatants were incubated 2 h at 4°C with 4 μg of mAbs, anti-CCL17, and/or anti-CCL22 (R&D Systems) or an isotype-matched control Ab coupled to 25 μl of protein G sepharose beads (Amersham Pharmacia Biotech), and after filtration to remove beads, the supernatants were used for chemotaxis. To calculate specific migration, the number of cells in each subpopulation in the absence of chemokine was subtracted from the number of the corresponding cell subpopulation migrated in the presence of chemokines. To evaluate the percentage of specific migration, the number of specifically migrated CD4⁺CD25⁺ T cells was divided by the total number of specifically migrated cells. To calculate the chemotactic index, the number of cells migrated in response to chemokines was divided by the number of spontaneously migrated cells.

To assess the proliferative response of human CD4⁺ T cells, we purified allogeneic PBMCs and depleted T cells with anti-CD3 microbeads (Miltenyi Biotech). Then, 2.5 × 10⁴ PBMCs were irradiated (3,000 rad) and mixed to 2.5 × 10⁴ of migrated CD4⁺ T cells that were depleted or not of CD25⁺ T cells by FITC-labeled anti-CD25 mAb followed by anti-FITC microbeads (FITC multisorting kit; Miltenyi Biotech). [³H]thymidine incorporation was measured by scintillation counting on day 3 after a 6-h pulse.

Slides for double immunofluorescent stainings were postfixed with 4% paraformaldehyde, blocked with 15% FCS/PBS. After washing, the slides were incubated with the primary Ab. The slides were washed again and incubated with the fluorocrome (Rhodamine Red-X, Cy2)-conjugated secondary Abs. After washing, slides were mounted with 90% glycerol/PBS and analyzed with a confocal microscope (MRC-1024; Bio-Rad Laboratories) equipped with a 15-mW Kr/Ar laser.

Cell surface staining of purified human CD4⁺ peripheral blood T cells revealed that ∼8–10% of these cells expressed the IL-2Rα chain CD25. Human CD25⁺CD4⁺ T cells were virtually all CD45RO⁺, lacked expression of CD69, were mostly CD45RB^low and CD62L⁺, and showed constitutive intracellular expression of CTLA-4 (data not shown).

As the mechanism of immunosuppression requires contact between target and Treg cells 13, we reasoned that Treg cells could use specific chemokine receptors to localize in the vicinity of effector T cells and/or APCs. We initially analyzed the migration of purified populations of CD4⁺ T cells to a large panel of chemokines (Fig. 1). Analysis of the percentage of specific migration revealed that ∼60% of CD4⁺ T cells specifically migrating in response to CCL1 expressed CD25. Human herpesvirus 8 (HHV8)-encoded vMIP-I, CCL22, and CCL17 also showed preferential chemotactic activity on CD25⁺ Treg cells (Fig. 1 a). Given the fact that CD25⁺ T cells are CD45RO⁺ and receptors for inflammatory chemokines are expressed on effector/memory T cells, we compared the chemotactic responsiveness of CD25⁺ Treg cells with that of bulk CD25⁻CD4⁺ T cells or only with CD45RO⁺CD25⁻ CD4⁺ T cells. This analysis revealed that CCL17 and CCL22 were more potent and efficacious but less selective than CCL1 for Treg cells. The chemotactic index in response to CCL17 and CCL22 was ∼15–18 for CD25⁺ T cells, ∼5–8 for CD45RO⁺CD25⁻ T cells (Fig. 1 b), and ∼40% of specifically migrated CD4⁺ T cells expressed CD25 (Fig. 1 a). By contrast, the chemotactic index in response to CCL1 was only ∼4 for CD25⁺ T cells (Fig. 1 b), but >50% of specifically migrated CD4⁺ T cells were CD25⁺. It should be noted that Treg cells also responded to several chemokines including interferon-inducible T cell α chemoattractant (I-TAC/CXCL11) and Epstein-Barr virus–induced molecule 1 ligand chemokine (ELC/CCL19) (Fig. 1 b). CCL19 exhibited powerful chemotactic activity with a chemotactic index close to ∼100 (Fig. 1 b), but CD25⁺ Treg cells were not specifically hyperresponsive to CCL19 when compared with CD4⁺CD45RO⁺ T cells (Fig. 1 a). By contrast, the bulk CD25⁻ T cell population, which included CD45RA⁺ naive T cells, showed a greater response to the CCR7 ligand CCL19 (Fig. 1 b). Next, we investigated the ability of CD4⁺ T cells that migrated in response to CCL1 or CCL22 to proliferate. In preliminary experiments we confirmed the anergic and immunosuppressive phenotype of purified CD4⁺CD25⁺ T cells (data not shown). As expected, the allogeneic proliferative response of CD4⁺ T cells that migrated to CCL1 and CCL22 was markedly reduced in comparison to the response of nonmigrated cells or cells that migrated to CCL19 (Fig. 2 a). Depletion of CD25⁺ cells from CD4⁺ T cells that migrated to CCL1 or CCL22 resulted in enhanced T cell proliferation (Fig. 2 b).

Based on these findings, we analyzed CCR4 and CCR8 expression on purified CD4⁺CD25⁺ and CD4⁺CD25⁻ T cells. Staining with anti-CCR4 and anti-CCR8 Abs and analysis by confocal microscopy showed that the great majority of CD4⁺CD25⁺ T cells expressed chemokine CCR4 and CCR8 (Fig. 3, a and b). Staining of CD4⁺CD25⁻ T cells revealed that a relatively large fraction of these cells stained positive for CCR4 expression, whereas few cells expressed CCR8 (Fig. 3, a and b). These findings suggested coexpression of CCR4 and CCR8 on Treg cells and, consistent with the chemotaxis data, indicated that expression of CCR8 is more restricted to the Treg cell population than CCR4. In agreement with coexpression of these receptors on Treg cells, migration in response to a combination of suboptimal doses of CCL1 and CCL22 resulted in a synergistic rather than additive chemotactic effect (Fig. 3 c). Finally, to prove that these receptors were indeed responsible for the observed migration of Treg cells we took advantage of the recently described selective CCR8 antagonist, MC148, encoded by the virus of Molluscum contagiosum 14. In the presence of MC148, specific migration of Treg cells in response to CCL1 but not CCL22 was completely abolished, confirming the specific involvement of CCR8 in CCL1-mediated migration of Treg cells (Fig. 3 d). These findings document that blood-borne human Treg cells possess a unique chemotactic response profile and preferentially express the chemokine receptors CCR4 and CCR8.

Overall, our results suggest that CCR4 and CCR8 ligand chemokines CCL1, CCL17, and CCL22 might participate in downregulating inflammatory T cell–mediated responses. Previous work identified specific requirements for production of these chemokines by T cells and APCs 15,16. Mature DCs have been reported to produce large quantities of CCL17 and CCL22 17. Thus, we tested the possibility that Treg cells could be attracted by mature DCs. Maturation of monocyte-derived DCs was induced by stimulation with LPS and the supernatant from fully mature DCs was harvested and tested for the ability to attract Treg cells. In a first series of experiments, we found that the supernatants obtained from mature DCs, containing high levels of CCL17 and CCL22 and small amounts of CCL1, attracted very efficiently CCR4-expressing, but not CCR8-expressing L1.2 cells (data not shown). We then compared the migration of CD25⁻ and CD25⁺ Treg cells in response to supernatants of mature DCs and investigated the role of CCL17 and CCL22 present in the supernatants. Our data show that supernatant obtained from mature DCs preferentially attracted CD25⁺ Treg cells among blood-borne CD4⁺ T cells (Fig. 4). Neutralization of CCL22 significantly reduced the migration of Treg cells, but neutralization of both CCL17 and CCL22 was required to completely abolish the migration of Treg cells (Fig. 4).

In this report, we provide the first evidence that blood-borne human CD4⁺CD25⁺ Treg cells exhibit a distinctive chemotactic response profile and chemokine receptor expression. Treg cells exhibit chemotactic responsiveness to several inflammatory and lymphoid chemokines, but they are specifically hyperresponsive to chemokines that engage the chemokine receptors CCR4 and CCR8 (Fig. 1).

Our investigation documents a broad spectrum of responsiveness of Treg cells to inflammatory chemokines that could potentially allow access to inflamed tissues and contact responding T cells and APCs. However, the specificity of action of chemokines such as CCL17, CCL22, and CCL1 on Treg cells suggests a unique role for these chemokines and their receptors in the physiology of Treg cells. Activated T cells and professional APCs such as DCs and monocytes/macrophages can produce CCL1, CCL17, and CCL22 15,16,18. CCL17 and CCL22 secreted by activated DCs have been shown to attract activated T cells expressing CCR4 17. Recent data suggest that T cells compete for access to antigen-bearing APCs 19,20, a phenomenon that may potentially be regulated by chemokines secreted by APCs and chemokine receptors expressed on T cells. Thus, chemokine receptors, such as CCR4 and CCR8, that are upregulated upon TCR-mediated activation and are associated to critical steps of Th cell differentiation 21, may regulate the access of T cells to antigen-bearing APCs and control the dynamics of their interactions. In light of our results, we speculate that constitutive expression of CCR4 and CCR8 may endow Treg cells with a competitive advantage over other T cells for interacting with APCs that secrete chemokines acting on those receptors. Consistent with this hypothesis, we found that mature DCs preferentially attract Treg cells among circulating CD4⁺ T cells by virtue of their secretion of CCR4 agonistic chemokines. Another possible scenario, not mutually exclusive, envisions that the production of CCL1, CCL17, and CCL22 by different types of cells in inflamed tissues may recruit Treg cells to downregulate an ongoing inflammatory response. In line with this proposal, the previously reported pattern of CCL1 secretion by Fcγ receptor triggered monocytes is consistent with chemokines produced in an inflammatory environment in response to opsonizing Abs and bacterial products 16. Interestingly, FcγRI triggering has been implicated in downregulation of inflammatory responses by modulating cytokine production by monocytes 22. Furthermore, we have recently shown that IL-12 inhibits CCL1 and CCL22 production by activated T cells 18. Conceivably, production of CCL1 and CCL22 by activated monocytes and T cells 15,18,23 might promote the recruitment of Treg cells to downregulate inflammatory Th1 responses 24.

Specific expression of CCR4 and CCR8 on Treg cells may allow their migration toward APCs and activated T cells leading to inhibition of APC function or suppression of responding T cells. Although Treg cells specifically express both CCR4 and CCR8, there are important distinctions between these receptors. First, CCR8 appears more selectively expressed on Treg cells than CCR4. CCR4 is expressed on the majority of circulating CD4⁺ memory T cells, including skin-homing T cells and central memory CCR7⁺ T cells (12,25,26; Fig. 3 a). By contrast, CCR8 expression and responsiveness to CCL1 seems confined to Treg cells and to another small subset of circulating CD4⁺ T cells (unpublished data). Second, CCR4 agonistic chemokines are consistently more potent and efficacious than CCR8 agonistic chemokines in promoting migration of Treg cells. This difference may be due to different levels of chemokine receptor's expression and could imply a distinct involvement for CCR4 and CCR8 in the function of Treg cells.

Based on our findings, regulation of CCL1, CCL17, and CCL22 production during the course of inflammatory responses could dictate the extent, severity, and duration of the response by modulating recruitment of Treg cells expressing CCR4 and CCR8. Interestingly, HHV8, which is found associated with Kaposi's sarcoma and Castelman's disease, encodes for three viral chemokines (vMIP-I, II, and III), which are believed to play a role in subverting the host's immune response 27. vMIP-I and vMIP-II are selective agonists of CCR8 28,29, while vMIP-III is an agonist of CCR4 30. Based on our findings, we can speculate that production of vMIPs by HHV8 may recruit Treg cells in infected tissues and help virally infected cells to evade the host's immune response. Taken together, our data suggest that certain “regulatory” chemokines may control the termination of inflammatory responses by recruiting a specific subset of Treg cells that can limit tissue damage and prevent autoimmunity. Manipulating the recruitment of Treg cells may be useful in a variety of pathological conditions to either achieve tolerance in autoimmunity and transplantation or boost immune responses against tumors and viral infections.

We thank L. Colantonio, L. Adorini, A. Mantovani, and S. Sozzani for discussions and critical reading of the manuscript, G. Torrioni and S. Rossini for FACS^® sorting.