γδ T lymphocytes play an important role in the immune defense against infection, based on the unique reactivity of human Vδ2Vγ9 γδ T cells toward bacterial phosphoantigens. Chemokines and their corresponding receptors orchestrate numerous cellular reactions, including leukocyte migration, activation, and degranulation. In this study we investigated the expression of various receptors for inflammatory and homeostatic chemokines on peripheral blood γδ T cells and compared their expression patterns with those on αβ T cells. Although several of the analyzed receptors (including CCR6, CCR7, CXCR4, and CXCR5) were not differentially expressed on γδ vs αβ T cells, γδ T cells expressed strongly increased levels of the RANTES/macrophage inflammatory protein-1α/-1β receptor CCR5 and also enhanced levels of CCR1–3 and CXCR1–3. CCR5 expression was restricted to Vδ2 γδ T cells, while the minor subset of Vδ1 γδ T cells preferentially expressed CXCR1. Stimulation with heat-killed extracts of Mycobacterium tuberculosis down-modulated cell surface expression of CCR5 on γδ T cells in a macrophage-dependent manner, while synthetic phosphoantigen isopentenyl pyrophosphate and CCR5 ligands directly triggered CCR5 down-modulation on γδ T cells. The functionality of chemokine receptors CCR5 and CXCR3 on γδ T cells was demonstrated by Ca2+ mobilization and chemotactic response to the respective chemokines. Our results identify high level expression of CCR5 as a characteristic and selective feature of circulating Vδ2 γδ T cells, which is in line with their suspected function as Th1 effector T cells.
Chemokines are a large family of low m.w. proteins that play important roles in leukocyte migration, activation, and degranulation. They are classified on the basis of structural features into major subclasses of CXC chemokines, where two of four conserved cysteines are separated by an amino acid X, and CC chemokines, where these two cysteines are located side by side. Minor subgroups of chemokines are characterized by the absence of two cysteines (C chemokine lymphotactin) or the presence of three amino acids between two cysteines (CX3C chemokines) (see Ref. 1 for review). Based on such structural features, a new systematic nomenclature for human chemokines has been proposed (2). The corresponding chemokine receptors are seven-transmembrane G protein-coupled receptors that share structural features and can be grouped according to the corresponding ligand into CXCR, CCR, CR, and CX3CR families. An alternative classification of chemokines and receptors is based on functional and physiological features and distinguishes between inflammatory (or inducible) and homeostatic (or constitutive) chemokines (3). Inflammatory chemokines are frequently up-regulated in nonlymphoid tissue under inflammatory conditions and are instrumental in the recruitment of effector T lymphocytes.
With regard to T lymphocyte biology, chemokines and their corresponding cellular receptors are involved in intrathymic T cell development (4, 5), in the orchestration of T-B cell interactions, as well as in the differentiation of effector T cells and the development of memory T cells (6, 7). Although naive T lymphocytes express CXCR4 and CCR7, various CCR and CXCR are differentially expressed on effector and memory T cells (6, 8). Coordinated expression of chemokine receptors has been associated with functionally distinct T lymphocyte subsets. In this respect, CCR5, CXCR3, and CCR1 have been found preferentially on Th1 cells producing IFN-γ, while polarized Th2 cells producing IL-4 frequently express the eotaxin (CCL11) receptor CCR3 (9, 10, 11, 12). Furthermore, subsets of memory T lymphocytes can be distinguished on the basis of their CCR7 expression. It appears that CCR7 is gradually lost as T cells differentiate from CCR7+ naive cells via CCR7+ lymph node-homing noneffector memory cells toward CCR7− tissue-homing effector memory cells (7, 13). Importantly, the cell surface expression of some chemokine receptors is modulated by cytokines and Ag recognition via the TCR, suggesting that T cells might change their migration pattern after antigenic stimulation (14, 15). Moreover, it is quite clear that the correlation of a committed functional phenotype with a particular pattern of cell surface chemokine receptors is not absolute. CCR3 can thus be induced on polarized Th1 cells, and CCR5 can be induced on Th2 cells in the presence of an appropriate cytokine milieu (16).
Although the vast majority of mature T lymphocytes expresses a heterodimeric αβ TCR, a small subset (1–5%) of peripheral blood T cells carries the alternative γδ TCR (17). Major differences between αβ and γδ T cells concern the diversity of the TCR germline repertoire and the Ags recognized by the respective TCR molecules (see Ref. 18 for review). The majority (50–95%) of human peripheral blood γδ T cells express a Vδ2Vγ9-encoded TCR, whereas γδ T cells using other Vδ/Vγ elements are usually rare in peripheral blood, but constitute major T cell populations in other anatomical localizations such as the small intestine (17, 19, 20). Vδ2Vγ9 T cells recognize small phosphorylated molecules derived from bacterial metabolic pathways (phosphoantigens) that cannot be seen by αβ T cells (17, 21, 22, 23, 24). These features together with results from in vivo studies in animal models suggest that γδ T cells play an important and nonredundant role in the immune defense against infectious micro-organisms (17, 25). It is likely that γδ T cells have additional functions, such as the immune surveillance of stressed cells and of certain tumor cells (17, 26, 27).
Although the expression and significance of chemokine receptors on αβ T lymphocytes has been the subject of extensive studies, little information is available on the chemokine receptor expression of γδ T cells. Functional studies with purified γδ T cells and γδ T cell clones indicate that human γδ T cells migrate in response to CC chemokines such as monocyte chemoattractant protein 1 (MCP-1)4 (6) (or CCL2), RANTES (CCL5), macrophage inflammatory protein 1α (MIP-1α or CCL3), and MIP-1β (CCL4), but not in response to CXC chemokines IL-8 (CXCL8) or IFN-inducible protein 10 (IP-10 or CXCL10) (28). The expression of the corresponding chemokine receptors was not investigated in this study (28). More recently, it was shown by RNase protection assay that phosphoantigen-activated Vδ2 γδ T cell lines rapidly down-regulated their expression of CC chemokine receptors, most notably CCR5, following re-exposure to the synthetic phosphoantigen isopentenyl pyrophosphate (IPP) (29). In addition, CXCR3 expression was found on TCRγδ-expressing thymocytes that migrated in response to the corresponding ligands IP-10 (CXCL10), monokine-induced by IFN-γ (or CXCL9), and IFN-inducible T cell α-chemoattractant (or CXCL11) (30).
In this study, we present the first comparative analysis of chemokine receptor expression on peripheral blood γδ and αβ T lymphocytes. Our results reveal significant differences between circulating αβ and γδ T cells in their surface expression of certain chemokine receptors, most notably CCR5. We discuss these findings with respect to the migration properties and effector phenotype of peripheral blood γδ T cells.
Materials and Methods
PBMC were isolated from buffy coats or from heparinized peripheral blood obtained from healthy adult donors by Ficoll-Hypaque density gradient centrifugation. αβ and γδ T cell clones were established from E-rosette-purified T cells (for αβ clones) or from MACS-purified γδ T cells by limiting dilution as previously described (19, 31). Briefly, T cells were seeded at 0.3 cells/well in 96-well microtiter plates in the presence of 2 × 105 irradiated PBMC feeder cells, PHA (0.5 μg/ml), and IL-2 (10 U/ml). T cell clones were expanded in RPMI 1640 medium (Life Technologies, Karlsruhe, Germany) supplemented with l-glutamine (2 mM), antibiotics, 10% heat-inactivated FCS, and IL-2 and were restimulated every 2 wk with irradiated feeder cells and PHA as previously described (31). To investigate activation-induced modulation of chemokine receptor expression, PBMC or E-rosette-purified T cells (1 × 106/ml) were stimulated for 24 h in 24-well culture plates with 1 μg/ml PHA, a 1/10.000 dilution of heat-killed Mycobacterium tuberculosis (M. tb.) H37Ra extract (32), 1 μg/ml LPS (L 2654 from Sigma-Aldrich, Deisenhofen, Germany), or 2 μg/ml IPP (Sigma-Aldrich) (33).
PBMC were stained with FITC-conjugated pan-TCRαβ mAb (Endogen, Woburn, MA) or pan-TCRγδ mAb (BD PharMingen, Heidelberg, Germany) and anti-human chemokine receptor mAb. All chemokine receptor Ab were used as PE conjugates with the exception of anti-CCR7 and anti-CXCR4, which were not directly fluorochrome labeled. PE-conjugated goat F(ab′)2 anti-mouse Ab (Caltag Laboratories, Burlingame, CA) was used as a second-step reagent to detect these primary Abs. We used the following anti-chemokine receptor mAb: CCR1, CCR2, CCR3, CCR6, CCR7, CXCR1, CXCR2, CXCR3, CXCR4, and CXCR5 (all from R&D Systems, Wiesbaden, Germany) and CCR5 (BD PharMingen). In addition, we used Tricolor-conjugated mAb against CD45RO and CD45RA (Caltag Laboratories). Appropriate fluorochrome-labeled isotype controls were included. The TCR Vγ and Vδ usage of established γδ T cell clones was analyzed with appropriate mAb as previously described (19). All analyses were measured on a FACSCalibur flow cytometer using CellQuest software (BD Biosciences). Results are presented as histograms or dot plots of mean fluorescence intensity. In the latter case the background of isotype controls has been subtracted. Statistical analysis was performed using paired Student’s t test.
Changes in the cytosolic free Ca2+ concentration in response to chemokine receptor ligands was visualized in Fluo-3/AM-loaded γδ T cell clones by flow cytometry. Briefly, T cell clones (5 × 106/ml) were incubated (25 min, 37°C) in HBSS containing 4 μm Fluo-3/AM (Molecular Probes, Leiden, The Netherlands). Afterward, T cell suspensions were diluted 1/5 in RPMI 1640 and incubated for an additional 30 min at 37°C. Cells were washed three times and resuspended (5 × 106 cells/ml) in assay buffer (137 mM NaCl, 5 mM KCl, 1 mM Na2HPO4, 5 mM glucose, 1 mM CaCl, 0.5 mM MgCl2, 10 mM HEPES, and 1 g/L BSA (pH 7.4)). Before each assay, 100 μl of the cell suspension was incubated for 3 min at 37°C in a thermoblock and subsequently stimulated with the following recombinant chemokines (500 ng/ml): RANTES, MIP-1α, MIP-1β, MIP-3α, IL-8, and IP-10 (all from R&D Systems). For comparison, a mixture of human defensins hBD2 (PeproTech, Rocky Hill, NJ) and HNP1 and HNP2 (Sigma-Aldrich, Steinheim, Germany) was used each at a final concentration of 500 ng/ml.
Lymphocyte chemotaxis was measured using a 48-well Boyden’s chamber (NeuroProbe, Cabin John, MD). Chemokines were serially diluted in RPMI 1640 (without phenol red) containing 0.1% BSA, 0.9 mM CaCl2, and 0.5 mM MgCl2, and 30 μl of the respective solutions were added to the bottom wells of the chamber. These were covered with a polycarbonate membrane (pore size, 5 μm; Costar Nucleopore, Tubingen, Germany), and the top wells received 1 × 105 T cell clone cells suspended in 50 μl RPMI 1640 supplemented with 0.1% BSA, 0.9 mM CaCl2, and 0.5 mM MgCl2. After incubation for 2.5 h at 37°C in an atmosphere containing 5% CO2, the assay was stopped by replacing the cell suspension in the upper well with ice-cold medium for 10 min. Thereafter, fresh cold medium was added for another 10 min to completely detach migrated cells from the bottom of the filters. Then filters were removed, and the migrated cells were transferred from the bottom wells to a microtiter plate. Residual cells in the bottom wells received 20 μl medium, were lysed by adding 5 μl 1% Triton X-100 (v/v) for 10 min, and were combined with the cells transferred to the microtiter plate, and cell lysis was continued for 10 min. Fifty microliters of 0.01 M p-nitrophenyl-β-glucuronide (Sigma-Aldrich) in 0.1 sodium acetate buffer (pH 4) was added for 40 h at 37°C, and the enzymatic reaction was stopped by adding 100 μl 0.4 M glycine buffer (pH 10). The OD was determined at 405 nm in a microplate reader. The number of migrated cells was calculated from a standard of lysed cells run in parallel.
Differential expression of chemokine receptors on circulating γδ and αβ T lymphocytes
We analyzed the expression of a panel of CC and CXC chemokine receptors on circulating peripheral blood γδ and αβ T lymphocytes by two-color flow cytometry. The results of a representative experiment are shown in Fig. 1. Compared with αβ T cells, the expression of several CC receptors (CCR1, CCR2, CCR3, CCR6) and CXC receptors (CXCR1, CXCR2, CXCR5) was higher on γδ T cells of this healthy adult blood donor, while other receptors, such as CCR7 and CXCR4, were equally expressed on both T cell subsets. The most striking differences were observed for CCR5 and CXCR3, which were both strongly expressed on γδ T cells, but only at low levels on αβ T lymphocytes. Because the expression of chemokine receptors varies among individuals, we analyzed the expression on αβ and γδ T cells in nine additional healthy donors. The results of the 10 separate experiments are summarized for CCR in Fig. 2. As can be seen, the strongly increased expression of CCR5 compared with that on αβ T cells is a general feature of peripheral blood γδ T cells (p < 0.001). γδ T cells also expressed increased levels of the MCP receptor CCR2 (p < 0.01) and the MCP/eotaxin-receptor CCR3 (p < 0.01; but note the different scales of mean fluorescence intensity in Fig. 2), while differences in CCR6 and CCR7 expression between γδ and αβ T lymphocytes were not statistically significant. αβ T cells are composed of subpopulations with regard to CD4/CD8 expression and naive vs memory (CD45RA vs CD45RO) phenotype, while the vast majority of the dominant γδ T cell population (Vδ2Vγ9) is double negative and CD45RO+ (34, 35). Therefore, we compared CCR5 expression on CD45RA and CD45RO subsets of CD4+, CD8+, and γδ T cells. The results of a representative experiment are shown in Fig. 3,A. As can be seen, CCR5 expression was primarily confined to the CD45RO+ and CD45RA− subsets. The relative percentages of CCR5-positive cells within the CD45RA+ and CD45RO+ subsets of CD4+, CD8+, and γδ T cells measured in eight healthy individuals are shown Fig. 3,B. In all three T cell populations, greater percentages of CD45RO+ cells expressed CCR5 compared with CD45RA+ subsets. Most γδ T cells were CD45RO+ and not CD45RA+ (see Fig. 3 A); CCR5, however, was also expressed on the few CD45RA+ γδ T cells.
Fig. 4 summarizes the results of CXC receptor expression on γδ vs αβ T cells and illustrates a substantial heterogeneity among the 10 analyzed healthy individuals. Nevertheless, significantly increased expression on γδ T cells was observed for CXCR1 (p < 0.05), CXCR2 (p < 0.005), and CXCR3 (p < 0.05), while differences in CXCR4 and CXCR5 expression did not reach statistical significance (note again the different scales of mean fluorescence intensity in Fig. 4). CXCR3 expression has been described for both central memory and effector memory T cells (13). Therefore, we again investigated the expression of CXCR3 on CD45RA and CD45RO subsets of CD4+, CD8+, and γδ T cells. A representative experiment is shown in Fig. 5,A. CXCR3 was expressed to varying degrees on both CD45RA+ and CD45RO+ subsets of CD4+ and CD8+ T cells as well as on the vast majority of γδ T cells. As for CCR5, CXCR3 was expressed on almost all CD45RO+ γδ T cells as well as on most of the few CD45RA+ or CD45RO− γδ T cells. The results obtained with the CXCR3 subset analysis of eight donors are summarized in Fig. 5,B. Taken together, the results presented in Figs. 4 and 5 indicate that the expression of CCR5 and CXCR3 is higher on CD45RO+ than on CD45RA+ cells. Both chemokine receptors, however, are also expressed on most of the few CD45RA+ naive γδ T cells.
Next we investigated chemokine receptor expression on subsets of peripheral blood γδ T cells with different TCR usage. The majority (50–95%) of circulating γδ T cells express Vδ2 paired with Vγ9, while 10–20% of peripheral blood γδ T cells use Vδ1, usually paired with a TCR Vγ element other than Vγ9 (19). In five separate experiments, we observed strongly enhanced expression of CCR5 on Vδ2, but not on Vδ1, γδ T cells compared with αβ T cells (p < 0.05), while Vδ1 γδ T cells expressed higher levels of CXCR1 than Vδ2 (and αβ) T cells (p < 0.005; Fig. 6). The moderately increased expression of CXCR3 on γδ T cells (Fig. 4) is primarily due to the dominant Vδ2 population and not to Vδ1 T cells (Fig. 6; p < 0.1; note again the different mean fluorescence intensity scales).
CCR7 was weakly expressed on γδ T cells (Fig. 2). Subsets of central memory (CCR7+) and effector memory (CCR7−) CD4+ αβ T cells can be differentiated on the basis of CCR7 expression (13). To investigate whether similar subsets exist within the γδ T cells, we analyzed CCR5, CXCR3, and CCR7 expression on gated CD45RO+ γδ T cells. As illustrated in Fig. 7, on the average only 40% of the gated cells were CCR7+, thus clearly indicating that a significant fraction of ex vivo analyzed CD45RO+ γδ T cells lacks CCR7 expression and thus could be classified as effector memory T cells (13). The simultaneous expression of CCR5 and CXCR3 on most of the CCR7− effector memory γδ T cells, as shown in Fig. 7, might help these cells to migrate into the inflamed tissue.
Modulation of chemokine receptor expression by cellular activation
Chemokine receptor expression is known to be modulated (up- or down-regulated) in response to cellular activation or ligand binding. To investigate the possible modulation of chemokine receptor expression on γδ T cells in response to cellular activation, PBMC were stimulated with PHA, M. tb., or phosphoantigen IPP, and chemokine receptors were analyzed after 24 h on gated γδ T cells. As illustrated in Fig. 8 (top), ex vivo expressed CXCR4 and CCR5 were strongly down-regulated by PHA, while other chemokine receptors with little expression on freshly isolated γδ T cells, including CXCR1 and CXCR2, were up-regulated. We also analyzed chemokine receptor modulation on γδ T cells in response to M. tb. and IPP, two well-known ligands for TCR-dependent recognition by Vδ2Vγ9 γδ T cells (22, 32). Interestingly, there was a dramatic down-regulation of CCR5 on γδ T cells in response to M. tb., whereas much less down-modulation was observed in the presence of IPP (Fig. 8, bottom). As also shown in Fig. 8, there was moderate modulation of other chemokine receptors on γδ T cells in response to M. tb. and/or IPP, which, however, was clearly less pronounced than the effect of PHA and also was more variable when analyzed in different donors (marked with an asterisk in Fig. 8). The finding that M. tb. induced a much stronger CCR5 down-modulation than IPP on γδ T cells when PBMC were used as responder cells suggested that the modulation triggered by M. tb. might be an indirect effect, in part due to the M. tb.-induced macrophage activation and subsequent cytokine and/or chemokine production by macrophages (36). To address this issue, we compared the CCR5 modulation on gated γδ T cells when PBMC (containing ∼30% monocytes) or E-rosette-purified T cells were used as responder cells. As illustrated in Fig. 9, the Vδ2Vγ9 ligand IPP triggered comparable CCR5 down-modulation on γδ T cells in both cases, as did a mixture of CCR5 ligands RANTES, MIP-1α, and MIP-1β. In contrast, M. tb. induced strong CCR5 down-modulation on γδ T cells only when PBMC were used as responder cells and not with purified responder T cells, suggesting a significant contribution of M. tb.-activated macrophages. This was further supported by the finding that stimulation of PBMC with LPS triggered CCR5 down-modulation on γδ T cells as efficiently as did M.tb., whereas LPS did not have any effect when purified responder T cells were used (Fig. 9). Moreover, the down-modulation of CCR5 on γδ T cells in response to M. tb. or IPP stimulation of PBMC was completely inhibited by a mixture of neutralizing Ab against CCR5 ligands (anti-RANTES, anti-MIP-1α, and anti-MIP-1β), suggesting that CC chemokine production by macrophages (in response to M. tb.) and/or γδ T cells (in response to IPP) was critically involved (Fig. 10).
Functional chemokine receptor expression on γδ T cell clones
In contrast to the uniformly strong ex vivo expression of CCR5 on Vδ2Vγ9 γδ T cells, chemokine receptor expression was more variable on established γδ T cell clones and varied with the activation status (not shown). To demonstrate that the cell surface chemokine receptors on γδ T cells are functional, we measured the Ca2+ influx in Fluo-3/AM-loaded γδ T cell clones in response to the corresponding ligands. The results of a representative experiment with a Vδ1 clone are illustrated in Fig. 11. This clone expressed CCR5, CCR6, and CXCR3 (Fig. 11,B) and responded to the respective ligands MIP-1α, MIP-1β, RANTES (for CCR5), MIP-3α (for CCR6), and IP-10 (for CXCR3) with rapid Ca2+ influx (Fig. 11,A), whereas no Ca2+ influx was elicited by IL-8, in line with the absent expression of IL-8R CXCR1 and CXCR2. Comparable results were obtained with other γδ T cell clones displaying different Vδ/Vγ TCR (not shown). In addition, γδ T cells expressing the relevant chemokine receptors migrated in response to the corresponding ligands in a chemotactic assay, as illustrated for a representative CCR5+ and CXCR3+ Vδ2Vγ9 clone in response to RANTES in Fig. 12. A clear chemotactic response was also seen with IP-10, although the intensity of the response was more variable than that obtained for RANTES (not shown).
Based on the expressed TCR Vδ/Vγ repertoire, subpopulations of murine and human γδ T cells can be identified that preferentially localize to different anatomical compartments. In humans, there are six expressed Vγ genes and a similar number of expressed Vδ genes (17, 37). The usage of this small germline repertoire is strikingly skewed. Although Vδ1 cells dominate in the peripheral blood of newborns, a gradual expansion of Vδ2 cells takes place during childhood, thereby leading to the characteristic predominance of Vδ2 cells, which usually account for 50–95% of all γδ T cells in the peripheral blood of healthy adults (38). γδ T cells expressing other Vδ elements such as Vδ1 are rare in the peripheral blood of adults (usually <20%), but constitute the dominant population within intraepithelial γδ T cells in the small intestine (20). Although Vδ1 is usually combined with any of the expressed Vγ elements (Vγ2, -3, -4, -5, -8, -9), the Vδ2 chain of peripheral blood γδ T cells is almost exclusively paired with Vγ9 (19, 39). Vδ2Vγ9 γδ T cells recognize in a TCR-dependent manner microbial ligands, most notably intermediates of the microbial nonmevalonate isoprenoid biosynthesis pathway (phosphoantigens), as well as some lymphoma cells (18, 21, 22, 23, 24). It has been postulated that the increase in peripheral blood Vδ2Vγ9 cells during childhood results from continuous exposure to such bacterial ligands (38). In fact, the Vδ2Vγ9 γδ T cells present in the peripheral blood of healthy adult donors are not naive T cells, but, rather, express markers characteristic of memory cells, such as CD45RO (34, 35). In addition, the constitutive expression of serine esterase also suggests an activated state of Vδ2Vγ9 T cells in vivo (40). Moreover, the results of short-term culture of these cells supports the idea that they are primed toward a Th1 phenotype, as evidenced by their rapid production of IFN-γ in response to phosphoantigens (41). However, ex vivo-isolated Vδ2Vγ9 cells can also be polarized toward a Th2 phenotype when cultured under appropriate Th2 priming conditions (42).
We have investigated the expression of chemokine receptors on human γδ T cells with two goals in mind. First, we aimed at a comparative analysis of chemokine receptor expression on peripheral blood γδ vs αβ T cells to determine whether major differences in the expression patterns exist. Secondly, we asked whether the expression of cell surface chemokine receptors on the dominant Vδ2Vγ9 γδ T cell population would correlate with their previous classification as memory and Th1 polarized cells (based on CD45RO expression and cytokine pattern) (34, 41). Our results reveal striking differences in the expressed chemokine receptor repertoire between peripheral blood γδ and αβ T cells and their CD4/CD8 subpopulations. Despite substantial interindividual heterogeneity, γδ T cells expressed increased levels of some of the analyzed CCR (CCR1, CCR2, CCR3), but most significantly of CCR5, while there was no significant difference in the expression of other CCR such as CCR6 and the CCL19/CCL21 receptor CCR7. Further analysis revealed that CCR5 was expressed on the vast majority of the CD45RO+ γδ T cells as well as on most of the few CD45RA+ γδ T cells. Although CCR5 was also preferentially expressed on the CD45RO+ compared with CD45RA+ subsets when CD4+ and CD8+ cells were investigated (Fig. 3), our results clearly identify γδ T cells as the subset within ex vivo-analyzed CD45RO+ peripheral blood T cells with the highest fraction of CCR5-expressing cells. Most CD45RO+ γδ T cells expressed CCR5 and CXCR3, while, on the average, <50% also expressed CCR7 (Fig. 7). On the basis of their CCR7 expression, memory T cells have been subdivided into effector memory (CCR7−) and central memory (CCR7+) T cells (13), and CCR5 has been reported to be preferentially expressed on CCR7− effector memory T cells. Our present results indicate that CCR5 is also expressed on at least subsets of CCR7− CD45RO+ γδ T cells. Functional studies with cell sorter-purified CCR5+ CD45RO+ γδ T cells coexpressing, or not, CCR7 are required to define their functional phenotype in terms of the proposed effector memory and central memory classification (7, 13). In accordance with CCR5, a more detailed analysis of CXCR3 expression also revealed the strongest expression (in terms of the percentage of positive cells) within CD45RO+ (and the few CD45RA+) γδ T cells compared with the respective subsets of CD4+ and CD8+ cells (Fig. 5). The expression of the RANTES/MIP-1α/MIP-1β receptor CCR5 together with CCR1 and CXCR3 has been associated with polarized Th1 cells (9, 10, 11). Although the studies of deletion mutants indicate that CCR5 is not absolutely essential for Th1 function in vivo (43), the high level expression of CCR5 on ex vivo-analyzed γδ T cells and the increased expression of CXCR3 clearly support the idea that circulating γδ T cells are primed Th1-type effector cells. Importantly, our results show that it is the dominant population of Vδ2Vγ9 cells that strongly expresses CCR5 (and less so CXCR3), while the minor population of circulating Vδ1 γδ T cells does not significantly differ from αβ T cells in this respect; both display rather low levels of CCR5 on their surface. Taken together, the strong expression of CCR5 reported in this study together with the known expression of CD45RO (34, 35) and the constitutive expression of serine esterase (40) all support the assumption that circulating Vδ2Vγ9 (but not Vδ1) γδ T cells in the peripheral blood of healthy adults are experienced cells, perhaps due to chronic exposure to microbial ligands, and are ready to rapidly respond to TCR-dependent ligand recognition by Th1-like cytokine production (41) and cytotoxic effector activity (44).
Although CCR5 was expressed on almost all Vδ2 γδ T cells, we observed a preferential expression of CXCR1 on the minor γδ T cell subset of Vδ1 cells, suggesting that the CXCR1 ligand IL-8 might preferentially act on this subpopulation. Although no chemotactic migration of γδ T cells in response to IL-8 was observed in a previous study, the TCR Vδ usage of the analyzed γδ T cell populations (total peripheral blood γδ T cells or established clones) was not reported (28).
The expression of chemokine receptors is modulated by cellular activation, which can result in up-regulation (45, 46, 47, 48, 49) or down-modulation (50). Not unexpectedly (50), CCR5 expression on ex vivo-analyzed γδ T cells was dramatically down-regulated by activation of PBMC with PHA or M. tb. and less strikingly by exposure to IPP. We speculated that the difference between the effects of the γδ T cell ligands M. tb. and IPP (Fig. 8) might be due to the fact that IPP is recognized exclusively by Vδ2Vγ9 γδ T cells within the PBMC, while the M. tb. lysate used in this study also activates monocytes/macrophages (36). Indeed, our further analysis with E-rosette-purified T cells confirmed this assumption, because CCR5 down-modulation on γδ T cells in response to M. tb. was much less dramatic when purified T cells were analyzed, while down-modulation induced by IPP was unchanged (which, in fact, was comparable to chemokine-induced down-modulation; Fig. 9). Moreover, LPS stimulation of PBMC, but not of purified T cells, also triggered CCR5 down-modulation on γδ T cells, again suggesting that macrophages activated by M. tb. (or LPS) contribute to CCR5 down-modulation on γδ T cells when unseparated PBMC are used as responder cells. Down-modulation of CCR5 by IPP (and M. tb.) involved the production of CC chemokines by γδ T cells (and monocytes in the case of M. tb.) as it could be completely prevented by a cocktail of neutralizing anti-RANTES/MIP-1α/MIP-1β mAb, well in line with the reported CC chemokine production of phosphoantigen-stimulated Vδ2Vγ9 γδ T cells (29, 51). In addition to CCR5, other chemokine receptors were also up-regulated or down-modulated by cellular activation, but with considerable variation among individual donors, thereby excluding definitive conclusions about the significance of these observations.
To address the functionality of chemokine receptors on γδ T cells, we measured Ca2+ influx and chemotactic migration in response to chemokines and observed a clear correlation between chemokine receptor expression and responsiveness to the respective ligands. Specifically, we observed Ca2+ mobilization with RANTES, MIP-1α, and MIP-1β in established CCR5+ clones, with MIP-3α in CCR6+ clones, and with IP-10 in CXCR3+ clones, while no response was obtained with IL-8, in line with the absence of CXCR1 and CXCR2 on the analyzed γδ T cell clones. In this regard our results extend a recent report showing that human thymic γδ T cells expressing CXCR3 migrate in response to IP-10, monokine-induced by IFN-γ, or IFN-inducible T cell α-chemoattractant, which all are produced by thymic epithelial cells (30), while no transendothelial chemotaxis of freshly isolated peripheral blood γδ T cells in response to IP-10 was observed in a previous study (28) where chemokine receptor expression was not analyzed. The functional significance of CCR5 expression was further confirmed in our experiments by the demonstration that CCR5+ γδ T cell clones showed a chemotactic response to RANTES. Similarly, CXCR3+ γδ T cell clones migrated in response to IP-10, even though there was more variability than with the chemotactic response of CCR5+ cells to RANTES.
In conclusion, our studies have identified strong CCR5 expression as a selective feature of ex vivo-analyzed peripheral blood Vδ2Vγ9 γδ T cells, which distinguishes these cells from NK cells and most other circulating T lymphocytes (52), including the majority of αβ T cells and other subsets (Vδ1) of γδ T cells. Future studies will address the functional significance of high level CCR5 expression on ex vivo-isolated Vδ2Vγ9 cells by analyzing the consequences of MIP-1α/MIP-1β/RANTES binding at the level of cellular activation and signal transduction of Vδ2Vγ9 T cells.
This work was supported by the Deutsche Forschungsgemeinschaft (DFG Ka 502/7-2).
Abbreviations used in this paper: MCP, monocyte chemoattractant protein; IP-10, IFN-inducible protein 10; IPP, isopentenyl pyrophosphate; MIP, macrophage inflammatory protein; M. tb., Mycobacterium tuberculosis.