Dual Functions of Fractalkine/CX3C Ligand 1 in Trafficking of Perforin+/Granzyme B+ Cytotoxic Effector Lymphocytes That Are Defined by CX3CR1 Expression

Miyuki Nishimura, Hisanori Umehara, Takashi Nakayama, Osamu Yoneda, Kunio Hieshima, Mayumi Kakizaki, Naochika Dohmae, Osamu Yoshie and Toshio Imai
J Immunol June 15, 2002, 168 (12) 6173-6180; DOI: https://doi.org/10.4049/jimmunol.168.12.6173

Abstract

Fractalkine/CX3C ligand 1 and its receptor CX3CR1 are known to mediate both cell adhesion and cell migration. Here we show that CX3CR1 defines peripheral blood cytotoxic effector lymphocytes commonly armed with intracellular perforin and granzyme B, which include NK cells, γδ T cells, and terminally differentiated CD8+ T cells. In addition, soluble fractalkine preferentially induced migration of cytotoxic effector lymphocytes. Furthermore, interaction of cytotoxic effector lymphocytes with membrane-bound fractalkine promoted subsequent migration to the secondary chemokines, such as macrophage inflammatory protein-1β/CC ligand 4 or IL-8/CXC ligand 8. Thus, fractalkine expressed on inflamed endothelium may function as a vascular regulator for cytotoxic effector lymphocytes, regardless of their lineage and mode of target cell recognition, through its ability to capture them from blood flow and to promote their emigration in response to other chemokines.

Migration and microenvironmental localization of specific lymphocyte populations are finely regulated at multiple steps by chemokines and adhesion molecules (1). Recent studies have shown that various lymphocyte subsets with differential tissue tropism in accordance with their particular developmental stages and/or functional properties express specific chemokine receptors (2). For example, CCR5 and CXCR3 are preferentially associated with Th1 cells, whereas CCR3, CCR4, and CCR8 are mainly expressed on Th2 cells (3, 4, 5, 6).

Fractalkine/CX3C ligand 1 is the sole member of the CX3C subfamily and is an exceptional molecule as a chemokine for being a transmembrane protein consisting of a chemokine domain with a unique CXXXC motif atop an extended mucin-like stalk (7). In the classic pathway of leukocyte emigration, leukocytes tether and roll on the surface of endothelium by weak interactions between selectins and selectin ligands. This allows leukocytes to be exposed to locally produced chemokines being presented on the surface of endothelium through binding to glycosaminoglycans. Signals transduced via chemokine receptors then activate integrins and induce firm adhesion of rolling leukocytes on the endothelium. Finally, leukocytes start diapedesis through the endothelium and into tissues along the gradient of chemokines (8, 9, 10). Previously, we and others have shown that fractalkine and its specific receptor CX3CR1 represent a novel type of leukocyte trafficking regulator, performing both adhesive and chemotactic functions (11, 12, 13). The membrane-bound fractalkine rapidly induces firm adhesion of CX3CR1-expressing cells under both static and flow conditions without requiring selectin-mediated rolling or activation of integrins (11, 12, 13). In addition, soluble fractalkine released from the cell surface by proteolytic cleavage (14) induces calcium mobilization, integrin activation, and migration of CX3CR1-expressing cells similar to other soluble chemokines (11, 12, 13). Thus, through both membrane-bound and soluble forms, fractalkine is likely to play important roles in trafficking of cells expressing CX3CR1 (11).

The expression of fractalkine was originally demonstrated on human umbilical endothelial cells upon treatment with TNF-α or IL-1 in vitro (7). Enhanced expression of fractalkine and infiltration of CD16+ NK cells were also demonstrated in crescent glomerulonephritis of human patients (15). Recently, a polymorphism in CX3CR1, which reduces its binding activity to fractalkine, was reported to increase the risk of HIV diseases, but to reduce the risk of atherosclerosis (16, 17, 18). Collectively, fractalkine and its receptor, CX3CR1, are likely to involved in tissue accumulation of leukocytes through inflamed endothelium.

Cytotoxic lymphocytes, which include a diverse variety of cells such as NK cells, CD8+ T cells, and γδ T cells, are the major effector cells in both innate and acquired immunity against intracellular pathogens and tumor cells (19). Even though the roles of chemokines and chemokine receptors in the trafficking regulations of CD4+ Th cells have been fairly well understood recently, similar trafficking regulations remain mostly unknown for cytotoxic lymphocytes (2). In the present study we have demonstrated that surface expression of CX3CR1 defines PBL commonly possessing high levels of intracellular perforin and granzyme B, and that soluble fractalkine preferentially attracts these perforin- and granzyme B-positive lymphocytes. Thus, fractalkine and CX3CR1 are likely to play important roles in trafficking of cytotoxic effector lymphocytes regardless of their lineage and mode of target cell recognition. Furthermore, we have demonstrated that the membrane-bound fractalkine enhances their migration to other chemokines. These results suggest that fractalkine regulates tissue emigration of cytotoxic effector lymphocytes through inflamed endothelium.

Materials and Methods

FACS analysis

mAbs to human CX3CR1 were generated from WKY/Ncrj rats immunized with CX3CR1 transfectants by a standard method. Two mAbs, 2A9-1 (IgG2b) and 1F2-2 (IgG2a), were obtained. There specificity was examined by flow cytometry using a panel of murine L1.2 cells expressing all known human chemokine receptors (CCR1∼10, CXCR1∼6, XCR1, and CX3CR1). Both mAbs reacted only with murine L1.2 cells expressing CX3CR1. FITC-, PE-, PC5-, or allophycocyanin-labeled anti-CD3 (UCHT1, IgG1), anti-CD4 (13B8.2, IgG1), anti-CD8 (B9.11, IgG1), anti-CD11a (S25.3.1, IgG1), anti-CD14 (RMO52, IgG2a), anti-CD16 (3G8, IgG1), anti-CD19 (J4.119, IgG1), anti-CD27 (1A4-CD27, IgG1), anti-CD28 (CD28.2, IgG1), anti-CD45RA (HI100, IgG2b), anti-CD57 (NC1, IgM), anti-CD62 ligand (Dreg56, IgG1), anti-TCR αβ (BMA031, IgG2b), and anti-TCR γδ (IMMU510, IgG1) were purchased from Coulter (Hialeah, FL). FITC- or R-PE-labeled anti-CD3 (UCHT1, IgG1), anti-CD4 (MT310, IgG1), anti-CD8 (DK25, IgG1), anti-CD11b (2LPM19c, IgG1), anti-CD14 (TUK14, IgG2a), anti-CD25 (ACT-1, IgG1), anti-CD45RA (4KB5, IgG1), anti-CD45RO (UCHL1, IgG2a), anti-CD95 (DX2, IgG1), and anti-CCR2 (no. 48607.211, IgG2b) were purchased from DAKO (Kyoto, Japan). PerCP-labeled anti-CD4 (SK3, IgG1) and anti-CD8 (SK1, IgG1) were purchased from BD Biosciences (Mountain View, CA). R-PE-labeled anti-CCR5 (2D7, IgG2a), anti-CXCR1 (5A12, IgG2b), anti-CXCR2 (6C6, IgG1), and anti-CXCR3 (1C6, IgG1) as well as CyChrome-labeled anti-CD45RO (UCHL1, IgG2a) were purchased from BD PharMingen (San Diego, CA). mAbs to human CCR7 (6B3, IgG1) described previously (20) were provided by Dr. Hasegawa. PBMC prepared from healthy adult donors by the standard method using Ficoll-Paque (Pharmacia, Uppsala, Sweden) were stained first with anti-CX3CR1 (2A9-1) and then with FITC-conjugated goat anti-rat IgG(H + L) F(ab′)2 (mouse serum absorbed, no. CLCC40101; Cedarlane Laboratories, Hornsby, Canada). In some experiments another anti-CX3CR1 mAb IF2-2 was used to confirm the results. After incubation with 1% normal rat serum for 20 min, the cells were further stained with labeled mAbs to various surface Ags. After washing, cells were analyzed on FACSCalibur (BD Biosciences).

Intracellular staining

After staining for surface CX3CR1 (see above), cells were resuspended and fixed with IntraPrep Reagent 1 (Coulter) at room temperature for 15 min. After washing once with PBS, cells were permeabilized with IntraPrep Reagent 2 at room temperature for 5 min. Cells were then incubated with PE-labeled anti-perforin (δG9, IgG2b; BD PharMingen) or PE-labeled anti-granzyme B (CLB-GB11, IgG1; Research Diagnostics, Flanders, NJ) at room temperature for 15 min in the dark. For negative controls, PE-labeled isotype controls purchased from BD PharMingen were used. After washing once with PBS, cells were resuspended in PBS/0.5% formaldehyde and analyzed on FACSCalibur (BD Biosciences).

Chemotaxis assay

Migration of PBL through ECV304 cells was conducted as described previously (11). In brief, ECV304 cells were seeded into Transwell culture inserts (Costar, Cambridge, MA) with pore size of 5 μm (2 × 105 cells/insert) and cultured for 48–72 h in medium 199 supplemented with 10% FCS. Chemokines diluted to 10 nM in the migration medium (RPMI 1640/medium 199 (1:1), 0.5% BSA, and 20 mM HEPES, pH 7.4) were added to 24-well tissue culture plates (600 μl/well), and inserts coated with ECV304 cells were placed. PBMC were then added to the upper chambers (106 cells in 100 μl). After incubation at 37°C for 4 h, cells migrated into the lower chambers were stained for various markers as stated above and analyzed by flow cytometry.

Isolation of CX3CR1+ and CX3CR1− CD8+ T cell subsets

CD8+ T cells were prepared from PBMC by two rounds of positive selection using the MACS system (Miltenyi Biotec). In brief, PBMC at 4 × 107 cells/70 μl MACS buffer (PBS containing 1% FCS and 5 mM EDTA) were incubated with the FcR blocking reagent (20 μl for 4 × 107 cells) and anti-CD8 microbeads (10 μl for 4 × 107 cells) at 4°C for 20 min. After washing, cells were resuspended in MACS buffer (107 cells/400 μl), and selection was conducted using LS column and Midi MACS magnet according to the manufacturer’s instruction. The resulting CD8+ T cells were >97% CD8+CD3+CD16− as determined by flow cytometry. Purified CD8+ T cells were next stained with anti-CX3CR1 mAb 2A9-1 and then with FITC-conjugated goat anti-rat IgG(H + L) F(ab′)2. CX3CR1+ and CX3CR1− populations with >95% purity were obtained through sorting on FACSVantage (BD Biosciences).

Cytotoxicity assay

Cytotoxicity was determined using the anti-CD3 mAb-mediated redirected Eu3+ release assay as described previously (21). In brief, FcR-bearing P815 target cells (2 × 106) were suspended in 1 ml labeling buffer: the mixture of 880 μl HEPES buffer (50 mM HEPES (pH 7.4), 93 mM NaCl, 5 mM KCl, and 2 mM MgCl2), 140 μl dextran sulfate stock solution (HEPES buffer containing 0.5% dextran; m.w., 500,000), and 80 μl europium (Eu) stock solution (mixture of 1.52 ml Eu atomic absorption standard solution from Aldrich, 0.5 ml 100 mM diethylene-triamine-penta-acetic acid (DTPA) solution and 7.98 ml HEPES buffer). After incubation at 37°C for 20 min, cells were washed twice with repairing buffer (HEPES buffer containing 2 mM CaCl2 and 10 mM glucose) and three times with RPMI 1640 containing 10% FCS. Subsets of purified CD8+ T cells (see above) were mixed with 5 × 103 labeled P815 cells at various E:T cell ratios in the presence or the absence of anti-CD3 mAb (UCHT1, Genzyme, Cambridge, MA). After incubation at 37°C for 3 h, supernatants (60 μl) of triplicate cultures were collected and mixed with 140 μl enhancement solution (Wallac/Berthord, Gaithersburg, MD). After shaking for 5 min, fluorescence was measured with the time-resolved fluorometer (ARVO-1240sx; Wallac/Berthold). Specific cytotoxicity was determined according to the formula: % specific lysis = 100 × ((experimental Eu release − spontaneous Eu release)/(maximal Eu release − spontaneous Eu release)).

Results

Expression of CX3CR1 on lymphocyte subsets

We have generated two novel rat mAbs, 2A9-1 and 1F2-2, against human CX3CR1. We confirmed their specificity to human CX3CR1 by using a panel of transfectants stably expressing all known chemokine receptors, including CCR1 ∼10, CXCR1 ∼6, XCR1, and CX3CR1 (data not shown). Using these mAbs, we examined expression of CX3CR1 on PBL from healthy adult donors by flow cytometry (Table I⇓). Essentially identical results were obtained by two different anti-CX3CR1 Abs (2A9-1 and 1F2-2). As shown in Fig. 1⇓a, most CD16+ NK cells (>90%) and a substantial fraction of CD3+ T cells (10–30%) expressed CX3CR1 with similar intensities. In contrast, CD19+ B cells were essentially negative. Among T cells, CX3CR1 were expressed on ∼40% of αβ CD8+ T cells, ∼5% of αβ CD4+ T cells, and ∼70% of γδ T cells (Fig. 1⇓b). CX3CR1 was also expressed on 40–85% of CD56+CD16−CD3− NK cells and 50–80% of CD3+CD56+CD16− NKT cells (data not shown).

FIGURE 1.
FIGURE 1.

Flow cytometric analysis of surface expression of CX3CR1 on circulating lymphocytes. Freshly isolated PBMC were stained indirectly with 2A9-1 anti-CX3CR1 mAb and directly with labeled mAbs to various surface markers. a, Analysis of CX3CR1 expression on CD16+ NK cells, CD3+ T cells, and CD19+ B cells. Comparable results were obtained with another anti-CX3CR1 mAb, 1F2-2. b, Analysis of CX3CR1 expression on αβ+CD8+ T cells, αβ+CD4+ T cells and γδ T cells. c, Analysis of CX3CR1 expression on naive and memory/effector T cell subsets defined by expression of CD45RA and CD27.

View this table:
Table I.

Phenotypic characterization of CX3CR1-expressing lymphocytesa

We next examined the expression of CX3CR1 in naive and memory/effector T cell subsets. Differential expression of CD45RA and the costimulatory molecule CD27 can divide CD4+ T cells into three distinct subpopulations: CD45RA+CD27+ naive subset, and CD45RA−CD27+ and CD45RA−CD27− memory subsets (22, 23). CX3CR1 was hardly expressed on CD45RA+CD27+ naive cells and CD45RA−CD27+ memory cells (Fig. 1⇑c). However, ∼40% of CD45RA−CD27− terminally differentiated memory/effector CD4+ T cells expressed CX3CR1. Similarly, CD8+ T cells can be divided into four subpopulations according to differential expression of CD45RA and CD27 (23, 24): CD45RA+CD27+ naive subset, CD45RA−CD27+ and CD45RA−CD27− memory subsets, and CD45RA+CD27− fully differentiated effector subset (Fig. 1⇑c). CX3CR1 was hardly expressed on CD45RA+CD27+ naive cells. In contrast, ∼30% CD45RA−CD27+ and ∼50% CD45RA−CD27− memory cells expressed CX3CR1. Furthermore, most (∼90%) CD45RA+CD27− CD8+ T cells, which correspond to terminally differentiated effector CD8+ T cells (24), expressed CX3CR1. These results strongly suggest that CX3CR1 expression is closely associated with the development of effector functions in CD4+ and CD8+ T cells.

Expression of various surface markers on CX3CR1-expressing lymphocytes

To further characterize the phenotypes of cells expressing CX3CR1, CD8+ T cells, CD4+ T cells, γδ T cells, and CD16+ NK cells were analyzed for coexpression of CX3CR1 and various functional and/or activation makers. As shown in Fig. 2⇓, CD57 (HNK-1) and CD11b, which are good markers for cytotoxic lymphocytes (23), were mostly coexpressed with CX3CR1 on all three T cell subsets and CD16+ NK cells. In contrast, costimulatory molecules such as CD27 and CD28 were mostly negative or only intermediate on CX3CR1-expressing subsets. CD95 (Fas), which is intermediate in terminally differentiated effector CD8+ T cells (22), was mostly expressed on CX3CR1-expressing cells at intermediate levels. CD25 (an activation maker) was mostly negative on CX3CR1-expressing subsets. CD62 ligand, which is essential for lymphocyte homing to lymph nodes (25), was mostly negative on CX3CR1-expressing CD8+ T cells, CD4+ T cells, and CD16+ NK cells. Notably, however, the majority of CX3CR1-expressing γδ T cells were positive for CD62 ligand. CD11a, which is involved in trafficking to inflamed tissues and thus correlates well with terminally differentiated effector T cells (24), was consistently expressed at high levels in CX3CR1-expressing CD8+ T cells, CD4+ T cells, γδ T cells, and CD16+ NK cells. Collectively, the surface phenotypes of CX3CR1-expressing cells, regardless of their major lineage and mode of target cell recognition, are very similar to those previously reported for terminally differentiated effector CD8+ T cells (24). Thus, CX3CR1-expressing lymphocytes are likely to represent a group of fully differentiated effector lymphocytes ready to migrate into inflamed tissues.

FIGURE 2.
FIGURE 2.

Flow cytometric analysis of surface expression of CX3CR1 and various functional molecules on CD8+ T cells, CD4+ T cells, γδ T cells, and CD16+ NK cells. Freshly isolated PBMC were stained indirectly with 2A9-1 anti-CX3CR1 mAb and directly with labeled mAbs to various surface molecules.

Selective expression of CX3CR1 on perforin- and granzyme B-positive lymphocytes

Cytotoxic effector lymphocytes such as NK cells, cytotoxic CD8+ T cells (CTLs) and the majority of γδ T cells contain specialized cytoplasmic granules storing the pore-forming protein (perforin) and serine proteases (granzymes), which are released from these cells upon activation to mediate target cell lysis (26). We therefore examined CD16+ NK cells, CD4+ T cells, CD8+ T cells, and γδ T cells for surface expression of CX3CR1 and intracellular staining of perforin and granzyme B. As shown in Fig. 3⇓, surface expression of CX3CR1 highly correlated with intracellular staining of perforin and granzyme B in all four classes of lymphocytes. Thus, CX3CR1 is commonly expressed on circulating cytotoxic effector lymphocytes containing intracellular cytotoxic granules regardless of their lineage and mode of target cell recognition.

FIGURE 3.
FIGURE 3.

Flow cytometric analysis of surface expression of CX3CR1 and intracellular content of perforin or granzyme B in CD16+ NK cells, CD8+ T cells, CD4+ T cells, and γδ T cells. Freshly isolated PBMC were stained indirectly with 2A9-1 anti-CX3CR1 mAb and directly with labeled mAbs to various surface molecules. After surface staining, cells were fixed, permeabilized, and stained with anti-perforin mAb or anti-granzyme B mAb. Regardless of the surface markers, CX3CR1-positive cells almost completely overlapped with those containing high amounts of perforin (upper panels) and granzyme B (lower panel).

Correlation of cytotoxic activity with CX3CR1 expression

To test the immediate cytotoxic activity of CX3CR1-expressing T cells, we purified CD8+ T cells by positive selection and sorted them into CX3CR1+ and CX3CR1− fractions. We measured the cytotoxic activities of the original and sorted populations by anti-CD3 mAb-mediated redirected cytotoxicity assay (27). As shown in Fig. 4⇓, CX3CR1+CD8+ T cells indeed showed much greater cytotoxic activity than presorted CD8+ T cells. In contrast, CX3CR1−CD8+ T cells only showed marginal cytotoxic activity. We confirmed that the cytotoxic activity was dependent on anti-CD3 mAb and was not affected by the anti-CX3CR1 mAb used for cell sorting (data not shown). Thus, circulating CD8+ T lymphocytes with immediate cytotoxic activity selectively express CX3CR1.

FIGURE 4.
FIGURE 4.

Cytotoxic activity of CX3CR1+ and CX3CR1− CD8+ T cell subsets. CD8+ T cells were purified from freshly isolated PBMC. Using anti-CX3CR1 mAb 2A9-1, purified CD8+ T cells were further sorted into CX3CR1+ and CX3CR1− fractions. The cytotoxic activity of unfractionated (presorted) and CX3CR1-fractionated CD8+ T cells was immediately analyzed against P815 target cells at various E:T cell ratios in the presence of anti-CD3 mAb. The observed cytotoxic activities were mostly dependent on the presence of anti-CD3 mAb. Thus, contribution from NK-like activity was minimum in the present system. Representative results from three separate experiments are shown. Data are the mean ± SE of triplicate cultures.

Selective migration of lymphocytes with intracellular perforin and granzyme B to soluble fractalkine

The selective expression of CX3CR1 on circulating lymphocytes containing cytotoxic granules suggests that fractalkine plays an important role in the recruitment of cytotoxic lymphocytes from the blood. To support this idea, we conducted Transwell chemotaxis assays using soluble fractalkine and stained input and migrated T cells for intracellular perforin and granzyme B. Soluble fractalkine attracted CD16+ NK cells, CD8+ T cells, and CD4+ T cells with efficiencies consistent to their percent positivity of CX3CR1 (data not shown) (11). Although original CD8+ T cells were heterogeneous in terms of intracellular contents of perforin and granzyme B, the cells attracted to soluble fractalkine displayed high strong staining for intracellular perforin and granzyme B (Fig. 5⇓a). Similar results were obtained for CD4+ T cells (data not shown). If spontaneous migration was subtracted, soluble fractalkine almost exclusively attracted cells possessing intracellular perforin and granzyme B (Fig. 5⇓b). We also confirmed that neutralizing mAbs to fractalkine (3A5 and 3H7), but not control mouse IgG, effectively blocked fractalkine-induced migration of CD16+ NK cells and granzyme B-positive CD8+ and CD4+ T cells (Fig. 5⇓c).

FIGURE 5.
FIGURE 5.

Preferential migration of T cells containing intracellular perforin and granzyme B to soluble fractalkine. Cells migrated through a monolayer of ECV304 cells into lower chambers containing soluble fractalkine at 10 nM were stained for surface CD8 or CD4, fixed, permeabilized, and stained for intracellular perforin or granzyme B. a, Flow cytometric profiles of intracellular perforin and granzyme B from input CD8+ T cells and CD8+ T cells migrated into lower chambers containing soluble fractalkine. b, Preferential migration of perforin- and granzyme B-containing CD8+ T cells and CD4+ T cells to soluble fractalkine. c, Inhibition of migration of granzyme B-containing CD8+ and CD4+ T cells and CD16+ NK cells to soluble fractalkine by neutralizing anti-fractalkine mAbs. Representative results from three separate experiments are shown as the mean ± SEM.

Differential expression of CX3CR1 and other chemokine receptors

We next examined coexpression of CX3CR1 and other chemokine receptors, namely, CCR5, CXCR3, CXCR1, CXCR2, CCR2, and CCR7. As shown in Fig. 6⇓, even though most CD16+ NK cells expressed CX3CR1 and CXCR1, only small fractions were found to coexpress CCR5 and CXCR3. In contrast, most CX3CR1-expressing γδ T cells coexpressed CCR5, while coexpression of CXCR3 or CXCR1 was variable from donor to donor. In the case of CD8+ and CD4+ T cells, substantial fractions of CX3CR1-expressing cells coexpressed CCR5 and CXCR1 at high to intermediate levels, while only minor fractions were found to coexpress CXCR3 or CCR2. In contrast, CCR7, the chemokine receptor essential for lymphocyte homing to secondary lymphoid tissues (2) was clearly negative on CX3CR1-expressing cells regardless of their lineages. CXCR2 was negative on any lymphocyte subsets examined, although it was clearly positive on both monocytes and granulocytes. These results indicate that expression of CX3CR1 is regulated quite differently from other chemokine receptors.

FIGURE 6.
FIGURE 6.

Differential expression of CX3CR1 and other chemokine receptors in T cell subsets and NK cells. Freshly isolated PBMC were indirectly stained with anti-CX3CR1 mAb 2A9-1 and directly labeled with mAbs to CD16, γδ TCR, CD8, or CD4 as well as those to CCR5, CXCR3, CXCR1, CCR2, and CCR7.

Promoting effect of membrane-bound fractalkine on migration of CX3CR1-expressing lymphocytes to secondary chemokines

The membrane-bound fractalkine has been shown to mediate firm adhesion and activation of CX3CR1-expressing cells (11, 12, 28, 29). However, it is not clear how membrane-bound fractalkine affects their migration. We therefore performed transmigration assay using ECV304 cells and fractalkine-transfected ECV304 (ECV-FKN) cells to examine its effects on migration of CX3CR1-expressing lymphocytes to soluble fractalkine and the second chemokine. We have previously shown that TNF-activated HUVECs and ECV-FKN cells express similar levels of fractalkine, indicating that ECV-FKN cells express physiologically relevant levels of fractalkine (12). Since a substantial fraction of CX3CR1-expressing CD8+ T cells coexpress CCR5, but not CCR2 (Fig. 6⇑), we chose macrophage inflammatory protein-1β (MIP-1β)3/CC ligand 4 (the ligand for CCR5) as a test chemokine and macrophage chemotactic protein-1 (MCP-1)/CC ligand 2 (the ligand for CCR2) as a negative control.

As shown in Fig. 7⇓a, transmigration of CD8+ T cells to soluble fractalkine was reduced by the expression of membrane-bound fractalkine (ECV-FKN). This suggests that adhesion by the membrane-bound form of fractalkine is dominant over cell migration to the soluble form of fractalkine. In contrast, transmigration of CD8+ T cells to MIP-1β, but not to MCP-1, was significantly enhanced in the presence of membrane-bound fractalkine (Fig. 7⇓a). Furthermore, only CD8+ T cells carrying granzyme B (the CX3CR1+ fraction) showed enhanced migration to MIP-1β in the presence of membrane-bound fractalkine (Fig. 7⇓b). The enhancing effects were completely blocked by a neutralizing mAb to fractalkine (3A5), but not by control IgG. The enhancing effects were not observed when PBMCs were added to the upper chamber together with 10 nM soluble fractalkine or preactivated with 300 nM soluble fractalkine (data not shown). In contrast, migration of CD8+ T cells without granzyme B (the CX3CR1− fraction) to MIP-1β or migration induced by MCP-1, the latter being selective for the granzyme B− fraction, was not affected by the presence of membrane-bound fractalkine (Fig. 7⇓b). The enhanced migration of granzyme B-positive CD8+ T cells was mainly due to augmentation of maximal response to MIP-1β by membrane-bound fractalkine (Fig. 7⇓c).

FIGURE 7.
FIGURE 7.

Effects of membrane-bound fractalkine on migration of CD8+ T cells to soluble fractalkine and other chemokines. Transmigration assays were conducted using Transwell plates with a preformed monolayer of ECV304 or ECV304 stably expressing membrane-bound fractalkine (ECV-FKN). a, Reduced migration of CD8+ T cells to soluble fractalkine (10 nM) and enhanced migration of CD8+ T cells to MIP-1β (10 nM) by membrane-bound fractalkine. b, Selective enhancement of migration of granzyme B-positive CD8+ T cells to MIP-1β by membrane-bound fractalkine. The enhancing effects were effectively abrogated with neutralizing mAbs to fractalkine. No such enhancement by membrane-bound fractalkine was seen in migration of granzyme B-negative CD8+ T cells to MIP-1β or MCP-1 (10 nM each). c, Migration of granzyme B-positive CD8+ T cells to various concentrations of MIP-1β in the presence or the absence of membrane-bound fractalkine. Migration efficiency to MIP-1β was mainly enhanced by membrane-bound fractalkine.

Since the majority of NK cells coexpress CX3CR1 and CXCR1 (Fig. 6⇑) (30), we further examined the effect of membrane-bound fractalkine on migration of CD16+ NK cells to IL-8/CXC ligand 8 (the ligand of CXCR1). As shown in Fig. 8⇓a, migration of CD16+ NK cells to IL-8 was also strongly enhanced by membrane-bound fractalkine. The enhancing effects were again effectively blocked by a neutralizing mAb to fractalkine (3A5). The enhanced migration of CD16+ NK cells was mainly due to augmentation of the maximal response to IL-8 by membrane-bound fractalkine (Fig. 8⇓c). Thus, membrane-bound fractalkine not only mediates the adhesion of CX3CR1-expressing cells, but also enhances their migration to the secondary chemokines.

FIGURE 8.
FIGURE 8.

Effects of membrane-bound fractalkine on migration of CD16+ NK cells to IL-8. Transmigration assays were conducted using Transwell plates with a preformed monolayer of ECV304 or ECV304 stably expressing membrane-bound fractalkine (ECV-FKN). a, Enhanced migration of CD16+ NK cells to IL-8 (10 nM) by membrane-bound fractalkine. The enhancing effects were effectively abrogated with neutralizing mAbs to fractalkine. b, Migration of CD16+ NK cells to various concentrations of IL-8 in the presence or the absence of membrane-bound fractalkine. Migration efficiency to IL-8 was mainly enhanced by membrane-bound fractalkine.

Discussion

Cytotoxic lymphocytes, which include NK cells, γδ T cells, CD8+ T cells, and a minor fraction of CD4+ T cells, are essential for host defense against intracellular pathogens and altered self cells (19). Their excessive activities, however, lead to organ-specific autoimmune diseases (31, 32, 33). They are commonly endowed with cytotoxic mechanisms such as cytoplasmic granules containing perforin and granzymes and surface expression of death-signaling Fas ligand (19, 26, 34). In the present study we have clearly demonstrated that the fractalkine receptor CX3CR1 is selectively expressed on various lineages of lymphocytes with high contents of intracellular perforin and granzyme B (Fig. 3⇑). Most CD16+ NK cells and the majority of γδ T cells, which play important roles in innate immunity as immediate effector killer cells (19), express CX3CR1 along with cytoplasmic perforin and granzyme B (Figs. 1⇑ and 3⇑). Similarly, CX3CR1-expressing CD8+ T cells, but not those without surface CX3CR1, were of terminally differentiated phenotypes, armed with perforin and granzyme B, and displayed strong cytotoxic activity (Figs. 3⇑ and 4⇑). Furthermore, CD4+ T cells expressing CX3CR1 were also terminally differentiated and contained intracellular perforin and granzyme B (Figs. 1⇑ and 3⇑), demonstrating that CX3CR1 is a good surface marker for CD4+ T cells with cytotoxic activity. Predominant expression of Fas ligand on terminally differentiated CD8+ T cells with remarkable lytic activity has also been demonstrated (24). Thus, we propose to call CX3CR1-expressing T cells, which are terminally differentiated subsets with cytotoxic activity, cytotoxic effector T cells (TCE).

Previously, we and others have shown that soluble fractalkine induces migration of NK cells and CD8+ T cells, and that the membrane-bound fractalkine mediates efficient capture, firm adhesion, and activation of CX3CR1-expressing cells in both static and flow conditions (11, 12, 13). In the present study we have further demonstrated that soluble fractalkine is a selective chemoattractant for lymphocytes displaying cytotoxic effector phenotypes (Fig. 5⇑). In addition, the membrane-bound fractalkine enhances migration of CX3CR1-expressing cells to other chemokines, such as MIP-1β and IL-8, acting on the same cells (Fig. 7⇑). Even though it remains to be seen whether the membrane-bound fractalkine indeed enhances chemotaxis of cells expressing CX3CR1 to other chemokines in vivo, we propose that fractalkine provides an initial clue for circulating NK cells and TCE to emigrate through inflamed endothelium as P- and E-selectin mediate selective recruitment of Th1 cells by inducing their rolling under flow (35). At present we do not know the exact molecular mechanisms of this enhancement, but integrin activation and cell polarization induced by membrane-bound fractalkine may have promoting effects on migratory responses to secondary chemokines (12, 28). In this context it is interesting to examine whether P- and E-selectin also promote the migration of Th1 cells toward Th1-directed chemokines even in the static condition. Thus, fractalkine expressed on inflamed endothelium plays dual functions: 1) as a selective capturing molecule for rapidly circulating killer effector lymphocytes in the blood, and 2) as a subsequent cosignaling molecule promoting their migration to secondary chemokines.

Sallusto et al. (36) have shown that surface expression of CCR7, the chemokine receptor that promotes lymphocyte homing to secondary lymphoid tissues, divides human memory T cells into two functionally distinct subsets. CCR7− memory T cells, termed effector memory T cells, preferentially migrate to inflamed tissues and display immediate effector functions. In contrast, CCR7+ T memory cells, termed central memory T cells, lack immediate effector functions and preferentially home to secondary lymphoid tissues, where they are efficiently stimulated by dendritic cells and differentiate into CCR7− effector cells upon secondary stimulation (36). It has been also reported that CD8+ cytotoxic T cells start to express perforin and granzymes during differentiation to memory/effector stages after antigenic stimulation (19). Consistently, we have observed that most CX3CR1+ T cells express CD11a, and cytoplasmic perforin and granzyme B, but lack L-selectin and/or CCR7 (Figs. 2⇑, 3⇑, and 6⇑), indicating that expression of CX3CR1 defines the terminally differentiated TCE ready to infiltrate into inflamed tissues.

Recent studies have shown that the expression of some chemokine receptors correlates with Th1 or Th2 polarization and/or with tissue-selective migration. We found that the majorities of CX3CR1-expressing CD8+ T cells and CD4+ T cells were also found to coexpress CCR5 and, less frequently, CXCR3 (Fig. 6⇑), the chemokine receptors known to be highly selective for Th1 cells (2), suggesting that CX3CR1-expressing CD8+ and CD4+ T cells partly overlap with Tc1 and Th1, respectively. In this regard, Fraticelli et al. (35) reported that fully polarized Th1 cells selectively expressed CX3CR1 and migrated to fractalkine. Most γδ T cells were also found to coexpress CCR5 and CXCR3 (Fig. 6⇑). In contrast, except for CX3CR1 and CXCR1, CD16+ NK cells were found to be negative for most other chemokine receptors (Fig. 6⇑). Therefore, CX3CR1 and CXCR1 (Fig. 7⇑) seem to be the major chemokine receptors involved in trafficking of CD16+ NK cells. Consistently, Campbell et al. (30) reported that CD16+ NK cells were the predominant population in peripheral blood migrating to IL-8 and fractalkine.

Recently, the second membrane-type chemokine, CXC ligand 16, has been described (37, 38). In humans its receptor CXCR6/Bonzo is expressed on fractions of CD45RO+ CD8+ and CD4+ T cells as well as small subsets of CD16+ NK cells and γδ T cells and is coordinately regulated with CCR5 (39). A recent study has further shown that CXCR6 is expressed by subsets of Th1 and Tc1, but not by Th2 or Tc2 cells (40). In contrast to the almost complete match of surface expression of CX3CR1 and possession of preformed perforin and granzyme B in various lymphocyte lineages, including CD8+ T cells (Fig. 3⇑), however, only an average 34% of granzyme A+ CD8+ T cells were shown to be positive for CXCR6 (40). Thus, we assume that, like CCR5, CXCR6 is expressed only on subsets of CX3CR1-expressing cytotoxic effector lymphocytes and on noncytotoxic effector cells. It remains to be seen whether there are any differences between the CXCR6+CX3CR1+ and CXCR6−CX3CR1+ subsets of cytotoxic lymphocytes in terms of their functional property and tissue migration (40, 41).

In conclusion, we have demonstrated that fractalkine and CX3CR1 are likely to play dual roles in the recruitment of effector lymphocytes with full cytotoxic activity, suggesting important pathophysiological roles of fractalkine in vascular and tissue injuries (42). CX3CR1 is also valuable as a highly selective surface marker for terminally differentiated TCE, whose selective dysfunction is likely to be associated with various chronic diseases, such as HIV-1 infection (30).

Acknowledgments

We thank Dr. Yoshimi Takai for constant support and encouragement; Drs. Hiroshi Kawamoto, Tatsuo Kina, and Yoshimoto Katsura for support with cell sorting; and Dr. Hitoshi Hasegawa for an anti-CCR7 mAb.

Footnotes

  • ↵1 Current address: Department of Rheumatology and Clinical Immunology, Kyoto University Graduate School of Medicine, Kyoto 606-8507, Japan.

  • ↵2 Address correspondence and reprint requests to Dr. Toshio Imai, Kan Research Institute, Science Center Building 3, Kyoto Research Park, 1 Chydoji Awata-cho, Simogyo-ku, Kyoto 600-8815, Japan. E-mail address: t-imai{at}kan.gr.jp

  • ↵3 Abbreviations used in this paper: MIP, macrophage inflammatory protein; DTPA, diethylene-triamine-penta-acetic acid; MCP, macrophage chemotactic protein; TCE, cytotoxic effector T cell.

  • Received December 19, 2001.
  • Accepted April 16, 2002.

References

  1. ↵
    Butcher, E. C., M. Williams, K. Youngman, L. Rott, M. Briskin. 1999. Lymphocyte trafficking and regional immunity. Adv. Immunol. 72: 209
    OpenUrlCrossRefPubMed
  2. ↵
    Yoshie, O., T. Imai, H. Nomiyama. 2001. Chemokines in immunity. Adv. Immunol. 78: 57
    OpenUrlCrossRefPubMed
  3. ↵
    Sallusto, F., A. Lanzavecchia, C. R. Mackay. 1998. Chemokines and chemokine receptors in T-cell priming and Th1/Th2-mediated responses. Immunol. Today 19: 568
    OpenUrlCrossRefPubMed
  4. ↵
    Imai, T., M. Nagira, S. Takagi, M. Kakizaki, M. Nishimura, J. Wang, P. W. Gray, K. Matsushima, O. Yoshie. 1999. Selective recruitment of CCR4-bearing Th2 cells toward antigen-presenting cells by the CC chemokines thymus and activation-regulated chemokine and macrophage-derived chemokine. Int. Immunol. 11: 81
    OpenUrlAbstract/FREE Full Text
  5. ↵
    Yamamoto, J., Y. Adachi, Y. Onoue, Y. S. Adachi, Y. Okabe, T. Itazawa, M. Toyoda, T. Seki, M. Morohashi, K. Matsushima, et al 2000. Differential expression of the chemokine receptors by the Th1- and Th2-type effector populations within circulating CD4+ T cells. J. Leukocyte Biol. 68: 568
    OpenUrlAbstract/FREE Full Text
  6. ↵
    Andrew, D. P., N. Ruffing, C. H. Kim, W. Miao, H. Heath, Y. Li, K. Murphy, J. J. Campbell, E. C. Butcher, L. Wu. 2001. C-C chemokine receptor 4 expression defines a major subset of circulating nonintestinal memory T cells of both Th1 and Th2 potential. J. Immunol. 166: 103
    OpenUrlAbstract/FREE Full Text
  7. ↵
    Bazan, J. F., K. B. Bacon, G. Hardiman, W. Wang, K. Soo, D. Rossi, D. R. Greaves, A. Zlotnik, T. J. Schall. 1997. A new class of membrane-bound chemokine with a CX3C motif. Nature 385: 640
    OpenUrlCrossRefPubMed
  8. ↵
    Butcher, E. C.. 1991. Leukocyte-endothelial cell recognition: three (or more) steps to specificity and diversity. Cell 67: 1033
    OpenUrlCrossRefPubMed
  9. ↵
    Springer, T. A.. 1994. Traffic signals for lymphocyte recirculation and leukocyte emigration: the multistep paradigm. Cell 76: 301
    OpenUrlCrossRefPubMed
  10. ↵
    Butcher, E. C., L. J. Picker. 1996. Lymphocyte homing and homeostasis. Science 272: 60
    OpenUrlAbstract
  11. ↵
    Imai, T., K. Hieshima, C. Haskell, M. Baba, M. Nagira, M. Nishimura, M. Kakizaki, S. Takagi, H. Nomiyama, T. J. Schall, et al 1997. Identification and molecular characterization of fractalkine receptor CX3CR1, which mediates both leukocyte migration and adhesion. Cell 91: 521
    OpenUrlCrossRefPubMed
  12. ↵
    Fong, A. M., L. A. Robinson, D. A. Steeber, T. F. Tedder, O. Yoshie, T. Imai, D. D. Patel. 1998. Fractalkine and CX3CR1 mediate a novel mechanism of leukocyte capture, firm adhesion, and activation under physiologic flow. J. Exp. Med. 188: 1413
    OpenUrlAbstract/FREE Full Text
  13. ↵
    Haskell, C. A., M. D. Cleary, I. F. Charo. 1999. Molecular uncoupling of fractalkine-mediated cell adhesion and signal transduction: rapid flow arrest of CX3CR1-expressing cells is independent of G-protein activation. J. Biol. Chem. 274: 10053
    OpenUrlAbstract/FREE Full Text
  14. ↵
    Garton, K. J., P. J. Gough, C. P. Blobel, G. Murphy, D. R. Greaves, P. J. Dempsey, E. W. Raines. 2001. TACE (ADAM17) mediates the cleavage and shedding of fractalkine (CX3CL1). J. Biol. Chem. 276: 37993
    OpenUrlAbstract/FREE Full Text
  15. ↵
    Furuichi, K., T. Wada, Y. Iwata, N. Sakai, K. Yoshimoto, M. Shimizu, K. Kobayashi, K. Takasawa, H. Kida, S. Takeda, et al 2001. Upregulation of fractalkine in human crescentic glomerulonephritis. Nephron 87: 314
    OpenUrlCrossRefPubMed
  16. ↵
    McDermott, D. H., J. S. Colla, C. A. Kleeberger, M. Plankey, P. S. Rosenberg, E. D. Smith, P. A. Zimmerman, C. Combadiere, S. F. Leitman, R. A. Kaslow, et al 2000. Genetic polymorphism in CX3CR1 and risk of HIV disease. Science 290: 2031
    OpenUrlCrossRefPubMed
  17. ↵
    Faure, S., L. Meyer, D. Costagliola, C. Vaneensberghe, E. Genin, B. Autran, J. F. Delfraissy, D. H. McDermott, P. M. Murphy, P. Debre, et al 2000. Rapid progression to AIDS in HIV+ individuals with a structural variant of the chemokine receptor CX3CR1. Science 287: 2274
    OpenUrlAbstract/FREE Full Text
  18. ↵
    Moatti, D., S. Faure, F. Fumeron, M. Amara, P. Seknadji, D. H. McDermott, P. Debre, M. C. Aumont, P. M. Murphy, D. de Prost, et al 2001. Polymorphism in the fractalkine receptor CX3CR1 as a genetic risk factor for coronary artery disease. Blood 97: 1925
    OpenUrlAbstract/FREE Full Text
  19. ↵
    Shresta, S., C. T. Pham, D. A. Thomas, T. A. Graubert, T. J. Ley. 1998. How do cytotoxic lymphocytes kill their targets?. Curr. Opin. Immunol. 10: 581
    OpenUrlCrossRefPubMed
  20. ↵
    Hasegawa, H., T. Nomura, M. Kohno, N. Tateishi, Y. Suzuki, N. Maeda, R. Fujisawa, O. Yoshie, S. Fujita. 2000. Increased chemokine receptor CCR7/EBI1 expression enhances the infiltration of lymphoid organs by adult T-cell leukemia cells. Blood 95: 30
    OpenUrlAbstract/FREE Full Text
  21. ↵
    Granberg, C., K. Blomberg, I. Hemmila, T. Lovgren. 1988. Determination of cytotoxic T lymphocyte activity by time-resolved fluorometry using europium-labelled concanavalin A-stimulated cells as targets. J. Immunol. Methods 114: 191
    OpenUrlCrossRefPubMed
  22. ↵
    Trowbridge, I. S., M. L. Thomas. 1994. CD45: an emerging role as a protein tyrosine phosphatase required for lymphocyte activation and development. Annu. Rev. Immunol. 12: 85
    OpenUrlCrossRefPubMed
  23. ↵
    Hamann, D., M. T. Roos, R. A. van Lier. 1999. Faces and phases of human CD8 T-cell development. Immunol. Today 20: 177
    OpenUrlCrossRefPubMed
  24. ↵
    Hamann, D., P. A. Baars, M. H. Rep, B. Hooibrink, S. R. Kerkhof-Garde, M. R. Klein, R. A. van Lier. 1997. Phenotypic and functional separation of memory and effector human CD8+ T cells. J. Exp. Med. 186: 1407
    OpenUrlAbstract/FREE Full Text
  25. ↵
    Arbones, M. L., D. C. Ord, K. Ley, H. Ratech, C. Maynard-Curry, G. Otten, D. J. Capon, T. F. Tedder. 1994. Lymphocyte homing and leukocyte rolling and migration are impaired in L-selectin-deficient mice. Immunity 1: 247
    OpenUrlCrossRefPubMed
  26. ↵
    Kagi, D., B. Ledermann, K. Burki, R. M. Zinkernagel, H. Hengartner. 1996. Molecular mechanisms of lymphocyte-mediated cytotoxicity and their role in immunological protection and pathogenesis in vivo. Annu. Rev. Immunol. 14: 207
    OpenUrlCrossRefPubMed
  27. ↵
    Mentzer, S. J., J. A. Barbosa, S. J. Burakoff. 1985. T3 monoclonal antibody activation of nonspecific cytolysis: a mechanism of CTL inhibition. J. Immunol. 135: 34
    OpenUrlAbstract/FREE Full Text
  28. ↵
    Goda, S., T. Imai, O. Yoshie, O. Yoneda, H. Inoue, Y. Nagano, T. Okazaki, H. Imai, E. T. Bloom, N. Domae, et al 2000. CX3C-chemokine, fractalkine-enhanced adhesion of THP-1 cells to endothelial cells through integrin-dependent and -independent mechanisms. J. Immunol. 164: 4313
    OpenUrlAbstract/FREE Full Text
  29. ↵
    Yoneda, O., T. Imai, S. Goda, H. Inoue, A. Yamauchi, T. Okazaki, H. Imai, O. Yoshie, E. T. Bloom, N. Domae, et al 2000. Fractalkine-mediated endothelial cell injury by NK cells. J. Immunol. 164: 4055
    OpenUrlAbstract/FREE Full Text
  30. ↵
    Campbell, J. J., S. Qin, D. Unutmaz, D. Soler, K. E. Murphy, M. R. Hodge, L. Wu, E. C. Butcher. 2001. Unique subpopulations of CD56+ NK and NK-T peripheral blood lymphocytes identified by chemokine receptor expression repertoire. J. Immunol. 166: 6477
    OpenUrlAbstract/FREE Full Text
  31. ↵
    Wu, Z., E. R. Podack, J. M. McKenzie, K. J. Olsen, M. Zakarija. 1994. Perforin expression by thyroid-infiltrating T cells in autoimmune thyroid disease. Clin. Exp. Immunol. 98: 470
    OpenUrlPubMed
  32. ↵
    Muller, S., J. Lory, N. Corazza, G. M. Griffiths, K. Z’Graggen, L. Mazzucchelli, A. Kappeler, C. Mueller. 1998. Activated CD4+ and CD8+ cytotoxic cells are present in increased numbers in the intestinal mucosa from patients with active inflammatory bowel disease. Am. J. Pathol. 152: 261
    OpenUrlPubMed
  33. ↵
    Rubesa, G., E. R. Podack, J. Sepcic, D. Rukavina. 1997. Increased perforin expression in multiple sclerosis patients during exacerbation of disease in peripheral blood lymphocytes. J. Neuroimmunol. 74: 198
    OpenUrlCrossRefPubMed
  34. ↵
    Yasukawa, M., H. Ohminami, J. Arai, Y. Kasahara, Y. Ishida, S. Fujita. 2000. Granule exocytosis, and not the fas/fas ligand system, is the main pathway of cytotoxicity mediated by alloantigen-specific CD4+ as well as CD8+ cytotoxic T lymphocytes in humans. Blood 95: 2352
    OpenUrlAbstract/FREE Full Text
  35. ↵
    Austrup, F., D. Vestweber, E. Borges, M. Lohning, R. Brauer, U. Herz, H. Renz, R. Hallmann, A. Scheffold, A. Radbruch, et al 1997. P- and E-selectin mediate recruitment of T-helper-1 but not T-helper-2 cells into inflamed tissues. Nature 385: 81
    OpenUrlCrossRefPubMed
  36. ↵
    Sallusto, F., D. Lenig, R. Forster, M. Lipp, A. Lanzavecchia. 1999. Two subsets of memory T lymphocytes with distinct homing potentials and effector functions. Nature 401: 708
    OpenUrlCrossRefPubMed
  37. ↵
    Matloubian, M., A. David, S. Engel, J. E. Ryan, J. G. Cyster. 2000. A transmembrane CXC chemokine is a ligand for HIV-coreceptor Bonzo. Nat. Immunol. 1: 298
    OpenUrlCrossRefPubMed
  38. ↵
    Wilbanks, A., S. C. Zondlo, K. Murphy, S. Mak, D. Soler, P. Langdon, D. P. Andrew, L. Wu, M. Briskin. 2001. Expression cloning of the STRL33/BONZO/TYMSTR ligand reveals elements of CC, CXC, and CX3C chemokines. J. Immunol. 166: 5145
    OpenUrlAbstract/FREE Full Text
  39. ↵
    Unutmaz, D., W. Xiang, M. J. Sunshine, J. Campbell, E. Butcher, D. R. Littman. 2000. The primate lentiviral receptor Bonzo/STRL33 is coordinately regulated with CCR5 and its expression pattern is conserved between human and mouse. J. Immunol. 165: 3284
    OpenUrlAbstract/FREE Full Text
  40. ↵
    Kim, C. H., E. J. Kunkel, J. Boisvert, B. Johnston, J. J. Campbell, M. C. Genovese, H. B. Greenberg, E. C. Butcher. 2001. Bonzo/CXCR6 expression defines type 1-polarized T-cell subsets with extralymphoid tissue homing potential. J. Clin. Invest. 107: 595
    OpenUrlCrossRefPubMed
  41. ↵
    Masopust, D., V. Vezys, A. L. Marzo, L. Lefrancois. 2001. Preferential localization of effector memory cells in nonlymphoid tissue. Science 291: 2413
    OpenUrlAbstract/FREE Full Text
  42. ↵
    Umehara, H., E. Bloom, T. Okazaki, N. Domae, T. Imai. 2001. Fractalkine and vascular injury. Trends Immunol. 22: 602
    OpenUrlCrossRefPubMed
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section