Defined by CX3CR1 Expression Cytotoxic Effector Lymphocytes

Dual Functions of Fractalkine/CX3C Ligand 1 in Trafficking of Perforin +/Granzyme B+ Cytotoxic Effector Lymphocytes That Are Defined by CX3CR1 Expression This information is current as of September 26, 2021. Miyuki Nishimura, Hisanori Umehara, Takashi Nakayama, Osamu Yoneda, Kunio Hieshima, Mayumi Kakizaki, Naochika Dohmae, Osamu Yoshie and Toshio Imai J Immunol 2002; 168:6173-6180; ;

doi: 10.4049/jimmunol.168.12.6173 Downloaded from http://www.jimmunol.org/content/168/12/6173

References This article cites 42 articles, 17 of which you can access for free at: http://www.jimmunol.org/content/168/12/6173.full#ref-list-1 http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists by guest on September 26, 2021 • Fast Publication! 4 weeks from acceptance to publication

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2002 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

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,1† Takashi Nakayama,‡ Osamu Yoneda,† Kunio Hieshima,‡ Mayumi Kakizaki,* Naochika Dohmae,† Osamu Yoshie,‡ and Toshio Imai2*‡

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, ␥␦ 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- Downloaded from 1␤/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. The Journal of Immunology, 2002, 168: 6173Ð6180.

igration and microenvironmental localization of spe- CX3CR1-expressing cells under both static and flow conditions http://www.jimmunol.org/ cific lymphocyte populations are finely regulated at without requiring selectin-mediated rolling or activation of inte- M multiple steps by chemokines and adhesion molecules grins (11Ð13). In addition, soluble fractalkine released from the (1). Recent studies have shown that various lymphocyte subsets cell surface by proteolytic cleavage (14) induces calcium mobili- with differential tissue tropism in accordance with their particular zation, integrin activation, and migration of CX3CR1-expressing developmental stages and/or functional properties express specific cells similar to other soluble chemokines (11Ð13). Thus, through both chemokine receptors (2). For example, CCR5 and CXCR3 are membrane-bound and soluble forms, fractalkine is likely to play im- preferentially associated with Th1 cells, whereas CCR3, CCR4, portant roles in trafficking of cells expressing CX3CR1 (11). and CCR8 are mainly expressed on Th2 cells (3Ð6). The expression of fractalkine was originally demonstrated on Fractalkine/CX3C ligand 1 is the sole member of the CX3C human umbilical endothelial cells upon treatment with TNF-␣ or by guest on September 26, 2021 subfamily and is an exceptional molecule as a chemokine for being IL-1 in vitro (7). Enhanced expression of fractalkine and infiltra- a transmembrane protein consisting of a chemokine domain with a ϩ tion of CD16 NK cells were also demonstrated in crescent glo- unique CXXXC motif atop an extended mucin-like stalk (7). In the merulonephritis of human patients (15). Recently, a polymorphism classic pathway of leukocyte emigration, leukocytes tether and roll in CX3CR1, which reduces its binding activity to fractalkine, was on the surface of endothelium by weak interactions between se- lectins and selectin ligands. This allows leukocytes to be exposed reported to increase the risk of HIV diseases, but to reduce the risk to locally produced chemokines being presented on the surface of of atherosclerosis (16Ð18). Collectively, fractalkine and its recep- endothelium through binding to glycosaminoglycans. Signals tor, CX3CR1, are likely to involved in tissue accumulation of leu- transduced via chemokine receptors then activate integrins and in- kocytes through inflamed endothelium. duce firm adhesion of rolling leukocytes on the endothelium. Fi- Cytotoxic lymphocytes, which include a diverse variety of cells ϩ nally, leukocytes start diapedesis through the endothelium and into such as NK cells, CD8 T cells, and ␥␦ T cells, are the major tissues along the gradient of chemokines (8Ð10). Previously, we effector cells in both innate and acquired immunity against intra- and others have shown that fractalkine and its specific receptor cellular pathogens and tumor cells (19). Even though the roles of CX3CR1 represent a novel type of leukocyte trafficking regulator, chemokines and chemokine receptors in the trafficking regulations performing both adhesive and chemotactic functions (11Ð13). The of CD4ϩ Th cells have been fairly well understood recently, sim- membrane-bound fractalkine rapidly induces firm adhesion of ilar trafficking regulations remain mostly unknown for cytotoxic lymphocytes (2). In the present study we have demonstrated that *Kan Research Institute, Kyoto, Japan; †Department of Internal Medicine, Osaka surface expression of CX3CR1 defines PBL commonly possessing Dental University, and ‡Department of Microbiology, Kinki University School of high levels of intracellular perforin and granzyme B, and that sol- Medicine, Osaka, Japan uble fractalkine preferentially attracts these perforin- and gran- Received for publication December 19, 2001. Accepted for publication April zyme B-positive lymphocytes. Thus, fractalkine and CX3CR1 are 16, 2002. likely to play important roles in trafficking of cytotoxic effector The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance lymphocytes regardless of their lineage and mode of target cell with 18 U.S.C. Section 1734 solely to indicate this fact. recognition. Furthermore, we have demonstrated that the mem- 1 Current address: Department of Rheumatology and Clinical Immunology, Kyoto brane-bound fractalkine enhances their migration to other che- University Graduate School of Medicine, Kyoto 606-8507, Japan. mokines. These results suggest that fractalkine regulates tissue 2 Address correspondence and reprint requests to Dr. Toshio Imai, Kan Research Institute, Science Center Building 3, Kyoto Research Park, 1 Chydoji Awata-cho, emigration of cytotoxic effector lymphocytes through inflamed Simogyo-ku, Kyoto 600-8815, Japan. E-mail address: [email protected] endothelium.

Materials and Methods Cytotoxicity assay FACS analysis Cytotoxicity was determined using the anti-CD3 mAb-mediated redirected 3ϩ mAbs to human CX3CR1 were generated from WKY/Ncrj rats immunized Eu release assay as described previously (21). In brief, FcR-bearing P815 target cells (2 ϫ 106 with CX3CR1 transfectants by a standard method. Two mAbs, 2A9-1 ) were suspended in 1 ml labeling buffer: the mixture of 880 ␮ (IgG2b) and 1F2-2 (IgG2a), were obtained. There specificity was examined l HEPES buffer (50 mM HEPES (pH 7.4), 93 mM NaCl, 5 mM KCl, and 2 mM MgCl ), 140 ␮l dextran sulfate stock solution by flow cytometry using a panel of murine L1.2 cells expressing all known 2 (HEPES buffer containing 0.5% dextran; m.w., 500,000), and 80 ␮ human chemokine receptors (CCR1ϳ10, CXCR1ϳ6, XCR1, and l eu- CX3CR1). Both mAbs reacted only with murine L1.2 cells expressing ropium (Eu) stock solution (mixture of 1.52 ml Eu atomic absorption stan- CX3CR1. FITC-, PE-, PC5-, or allophycocyanin-labeled anti-CD3 dard solution from Aldrich, 0.5 ml 100 mM diethylene-triamine-penta- (UCHT1, IgG1), anti-CD4 (13B8.2, IgG1), anti-CD8 (B9.11, IgG1), anti- acetic acid (DTPA) solution and 7.98 ml HEPES buffer). After incubation CD11a (S25.3.1, IgG1), anti-CD14 (RMO52, IgG2a), anti-CD16 (3G8, at 37¡C for 20 min, cells were washed twice with repairing buffer (HEPES buffer containing 2 mM CaCl and 10 mM glucose) and three times with IgG1), anti-CD19 (J4.119, IgG1), anti-CD27 (1A4-CD27, IgG1), anti- 2 RPMI 1640 containing 10% FCS. Subsets of purified CD8ϩ CD28 (CD28.2, IgG1), anti-CD45RA (HI100, IgG2b), anti-CD57 (NC1, T cells (see above) were mixed with 5 ϫ 103 IgM), anti-CD62 ligand (Dreg56, IgG1), anti-TCR ␣␤ (BMA031, IgG2b), labeled P815 cells at various E:T cell ratios in the presence or the absence of anti-CD3 mAb (UCHT1, Genzyme, and anti-TCR ␥␦ (IMMU510, IgG1) were purchased from Coulter (Hi- Cambridge, MA). After incubation at 37¡C for 3 h, supernatants (60 ␮l) of aleah, FL). FITC- or R-PE-labeled anti-CD3 (UCHT1, IgG1), anti-CD4 ␮ (MT310, IgG1), anti-CD8 (DK25, IgG1), anti-CD11b (2LPM19c, IgG1), triplicate cultures were collected and mixed with 140 l enhancement so- anti-CD14 (TUK14, IgG2a), anti-CD25 (ACT-1, IgG1), anti-CD45RA lution (Wallac/Berthord, Gaithersburg, MD). After shaking for 5 min, flu- (4KB5, IgG1), anti-CD45RO (UCHL1, IgG2a), anti-CD95 (DX2, IgG1), orescence was measured with the time-resolved fluorometer (ARVO- 1240sx; Wallac/Berthold). Specific cytotoxicity was determined according and anti-CCR2 (no. 48607.211, IgG2b) were purchased from DAKO ϭ ϫ Ϫ (Kyoto, Japan). PerCP-labeled anti-CD4 (SK3, IgG1) and anti-CD8 (SK1, to the formula: % specific lysis 100 ((experimental Eu release

Ϫ Downloaded from IgG1) were purchased from BD Biosciences (Mountain View, CA). R-PE- spontaneous Eu release)/(maximal Eu release spontaneous Eu release)). labeled anti-CCR5 (2D7, IgG2a), anti-CXCR1 (5A12, IgG2b), anti- CXCR2 (6C6, IgG1), and anti-CXCR3 (1C6, IgG1) as well as CyChrome- Results labeled anti-CD45RO (UCHL1, IgG2a) were purchased from BD Expression of CX3CR1 on lymphocyte subsets PharMingen (San Diego, CA). mAbs to human CCR7 (6B3, IgG1) described previously (20) were provided by Dr. Hasegawa. PBMC prepared We have generated two novel rat mAbs, 2A9-1 and 1F2-2, against from healthy adult donors by the standard method using Ficoll-Paque human CX3CR1. We conﬁrmed their speciﬁcity to human

(Pharmacia, Uppsala, Sweden) were stained first with anti-CX3CR1 CX3CR1 by using a panel of transfectants stably expressing all http://www.jimmunol.org/ (2A9-1) and then with FITC-conjugated goat anti-rat IgG(H ϩ L) F(abЈ) 2 known chemokine receptors, including CCR1 ϳ10, CXCR1 ϳ6, (mouse serum absorbed, no. CLCC40101; Cedarlane Laboratories, Hornsby, Canada). In some experiments another anti-CX3CR1 mAb IF2-2 XCR1, and CX3CR1 (data not shown). Using these mAbs, we was used to confirm the results. After incubation with 1% normal rat serum examined expression of CX3CR1 on PBL from healthy adult do- for 20 min, the cells were further stained with labeled mAbs to various nors by flow cytometry (Table I). Essentially identical results were surface Ags. After washing, cells were analyzed on FACSCalibur (BD obtained by two different anti-CX3CR1 Abs (2A9-1 and 1F2-2). Biosciences). ϩ As shown in Fig. 1a, most CD16 NK cells (Ͼ90%) and a sub- Intracellular staining stantial fraction of CD3ϩ T cells (10Ð30%) expressed CX3CR1 ϩ After staining for surface CX3CR1 (see above), cells were resuspended and with similar intensities. In contrast, CD19 B cells were essen- by guest on September 26, 2021 fixed with IntraPrep Reagent 1 (Coulter) at room temperature for 15 min. tially negative. Among T cells, CX3CR1 were expressed on ϳ40% After washing once with PBS, cells were permeabilized with IntraPrep of ␣␤ CD8ϩ T cells, ϳ5% of ␣␤ CD4ϩ T cells, and ϳ70% of ␥␦ Reagent 2 at room temperature for 5 min. Cells were then incubated with T cells (Fig. 1b). CX3CR1 was also expressed on 40Ð85% of PE-labeled anti-perforin (␦G9, IgG2b; BD PharMingen) or PE-labeled anti- ϩ Ϫ Ϫ ϩ ϩ CD56 CD16 CD3 NK cells and 50Ð80% of CD3 CD56 granzyme B (CLB-GB11, IgG1; Research Diagnostics, Flanders, NJ) at Ϫ room temperature for 15 min in the dark. For negative controls, PE-labeled CD16 NKT cells (data not shown). isotype controls purchased from BD PharMingen were used. After washing We next examined the expression of CX3CR1 in naive and once with PBS, cells were resuspended in PBS/0.5% formaldehyde and memory/effector T cell subsets. Differential expression of analyzed on FACSCalibur (BD Biosciences). CD45RA and the costimulatory molecule CD27 can divide CD4ϩ Chemotaxis assay T cells into three distinct subpopulations: CD45RAϩCD27ϩ naive 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 ␮m(2ϫ 105 cells/ Table I. Phenotypic characterization of CX3CR1-expressing insert) and cultured for 48Ð72 h in medium 199 supplemented with 10% a FCS. Chemokines diluted to 10 nM in the migration medium (RPMI 1640/ lymphocytes medium 199 (1:1), 0.5% BSA, and 20 mM HEPES, pH 7.4) were added to 24-well tissue culture plates (600 ␮l/well), and inserts coated with ECV304 NK ␥␦ T CD8ϩ T CD4ϩ T cells were placed. PBMC were then added to the upper chambers (106 cells in 100 ␮l). After incubation at 37¡C for 4 h, cells migrated into the lower CD57 72 Ϯ 11 72 Ϯ 15 80 Ϯ 5.6 75 Ϯ 14 chambers were stained for various markers as stated above and analyzed by CD11b 99 Ϯ 0.2 88 Ϯ 7.9 85 Ϯ 8.6 67 Ϯ 21 flow cytometry. CD28 1 Ϯ 0.1 13 Ϯ 6.6 22 Ϯ 15 20 Ϯ 12 ϩ Ϫ ϩ CD27 1 Ϯ 0.1 11 Ϯ 3.1 13 Ϯ 4.7 13 Ϯ 4.7 Isolation of CX3CR1 and CX3CR1 CD8 T cell subsets CD95 60 Ϯ 21 80 Ϯ 7.9 92 Ϯ 4.8 94 Ϯ 5.2 Ϯ Ϯ Ϯ Ϯ CD8ϩ T cells were prepared from PBMC by two rounds of positive se- CD25 1 0.1 6 4.5 4 2.6 16 20 Ϯ Ϯ Ϯ Ϯ lection using the MACS system (Miltenyi Biotec). In brief, PBMC at 4 ϫ CD62L 37 19 57 16 32 3.6 30 7.7 high Ϯ Ϯ Ϯ Ϯ 107 cells/70 ␮l MACS buffer (PBS containing 1% FCS and 5 mM EDTA) CD11a 96 0.9 92 8.9 97 1.3 97 0.4 Ϯ Ϯ Ϯ Ϯ were incubated with the FcR blocking reagent (20 ␮l for 4 ϫ 107 cells) and Perforin 96 3.0 68 18 66 23 53 19 Ϯ Ϯ Ϯ Ϯ anti-CD8 microbeads (10 ␮l for 4 ϫ 107 cells) at 4¡C for 20 min. After Granzyme B 91 7.2 85 7.3 95 5.0 90 9.3 Ϯ Ϯ Ϯ Ϯ washing, cells were resuspended in MACS buffer (107 cells/400 ␮l), and CCR5 14 7.6 91 6.4 77 14 89 6.2 Ϯ Ϯ Ϯ Ϯ selection was conducted using LS column and Midi MACS magnet ac- CXCR3 2 1.3 45 27 18 20 12 7.5 Ϯ Ϯ Ϯ Ϯ cording to the manufacturer’s instruction. The resulting CD8ϩ T cells were CXCR1 96 1.9 37 21 50 20 39 12 Ϯ Ϯ Ϯ Ϯ Ͼ97% CD8ϩCD3ϩCD16Ϫ as determined by flow cytometry. Purified CCR2 1 0.7 8 5.5 2 2.3 12 6.9 Ϯ Ϯ Ϯ Ϯ CD8ϩ T cells were next stained with anti-CX3CR1 mAb 2A9-1 and then CCR7 3 0.8 4 2.5 4 1.8 0 0.0 ϩ ϩ Ј ϩ with FITC-conjugated goat anti-rat IgG(H L) F(ab )2. CX3CR1 and a Ϫ Freshly isolated CX3CR1 lymphocytes from blood were analyzed by flow CX3CR1 populations with Ͼ95% purity were obtained through sorting cytometry. Values shown are the percentage Ϯ SD of CX3CR1ϩ lymphocytes ex- on FACSVantage (BD Biosciences). pressing the indicated cell surface or intracellular proteins. The Journal of Immunology 6175

intermediate on CX3CR1-expressing subsets. CD95 (Fas), which is intermediate in terminally differentiated effector CD8ϩ 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 CD8ϩ T cells, CD4ϩ T cells, and CD16ϩ NK cells. Notably, however, the majority of CX3CR1-expressing ␥␦ 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 CD8ϩ T cells, CD4ϩ T cells, ␥␦ T cells, and CD16ϩ 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 CD8ϩ T cells (24). Thus, CX3CR1-expressing lymphocytes are likely to represent a group of fully differentiated effector lymphocytes ready to migrate into inflamed tissues. Downloaded from

Selective expression of CX3CR1 on perforin- and granzyme B-positive lymphocytes Cytotoxic effector lymphocytes such as NK cells, cytotoxic CD8ϩ T cells (CTLs) and the majority of ␥␦ T cells contain specialized

cytoplasmic granules storing the pore-forming protein (perforin) http://www.jimmunol.org/ and serine proteases (granzymes), which are released from these FIGURE 1. Flow cytometric analysis of surface expression of CX3CR1 on circulating lymphocytes. Freshly isolated PBMC were stained indirectly cells upon activation to mediate target cell lysis (26). We therefore ϩ ϩ ϩ ␥␦ with 2A9-1 anti-CX3CR1 mAb and directly with labeled mAbs to various examined CD16 NK cells, CD4 T cells, CD8 T cells, and surface markers. a, Analysis of CX3CR1 expression on CD16ϩ NK cells, T cells for surface expression of CX3CR1 and intracellular stain- CD3ϩ T cells, and CD19ϩ B cells. Comparable results were obtained with ing of perforin and granzyme B. As shown in Fig. 3, surface ex- another anti-CX3CR1 mAb, 1F2-2. b, Analysis of CX3CR1 expression on pression of CX3CR1 highly correlated with intracellular staining ϩ ϩ ϩ ϩ ␣␤ CD8 T cells, ␣␤ CD4 T cells and ␥␦ T cells. c, Analysis of of perforin and granzyme B in all four classes of lymphocytes. CX3CR1 expression on naive and memory/effector T cell subsets deﬁned by expression of CD45RA and CD27. by guest on September 26, 2021 subset, and CD45RAϪCD27ϩ and CD45RAϪCD27Ϫ memory subsets (22, 23). CX3CR1 was hardly expressed on CD45RAϩ CD27ϩ naive cells and CD45RAϪCD27ϩ memory cells (Fig. 1c). However, ϳ40% of CD45RAϪCD27Ϫ terminally differentiated memory/effector CD4ϩ T cells expressed CX3CR1. Similarly, CD8ϩ T cells can be divided into four subpopulations according to differential expression of CD45RA and CD27 (23, 24): CD45RAϩ CD27ϩ naive subset, CD45RAϪCD27ϩ and CD45RAϪCD27Ϫ memory subsets, and CD45RAϩCD27Ϫ fully differentiated effector subset (Fig. 1c). CX3CR1 was hardly expressed on CD45RAϩ CD27ϩ naive cells. In contrast, ϳ30% CD45RAϪCD27ϩ and ϳ50% CD45RAϪCD27Ϫ memory cells expressed CX3CR1. Further- more, most (ϳ90%) CD45RAϩCD27Ϫ CD8ϩ T cells, which cor- respond to terminally differentiated effector CD8ϩ T cells (24), expressed CX3CR1. These results strongly suggest that CX3CR1 expression is closely associated with the development of effector functions in CD4ϩ and CD8ϩ T cells.

Expression of various surface markers on CX3CR1-expressing lymphocytes To further characterize the phenotypes of cells expressing CX3CR1, CD8ϩ T cells, CD4ϩ T cells, ␥␦ T cells, and CD16ϩ NK cells were analyzed for coexpression of CX3CR1 and various functional and/or activation makers. As shown in Fig. 2, CD57 FIGURE 2. Flow cytometric analysis of surface expression of CX3CR1 (HNK-1) and CD11b, which are good markers for cytotoxic lym- and various functional molecules on CD8ϩ T cells, CD4ϩ T cells, ␥␦ T phocytes (23), were mostly coexpressed with CX3CR1 on all three cells, and CD16ϩ NK cells. Freshly isolated PBMC were stained indirectly ϩ T cell subsets and CD16 NK cells. In contrast, costimulatory with 2A9-1 anti-CX3CR1 mAb and directly with labeled mAbs to various molecules such as CD27 and CD28 were mostly negative or only surface molecules. 6176 FRACTALKINE AND CX3CR1 IN MIGRATION OF CYTOTOXIC LYMPHOCYTES

cytes with immediate cytotoxic activity selectively express CX3CR1.

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 CD16ϩ NK cells, CD8ϩ T cells, and CD4ϩ T cells with efﬁciencies consistent to their percent positivity of FIGURE 3. Flow cytometric analysis of surface expression of CX3CR1 CX3CR1 (data not shown) (11). Although original CD8ϩ T cells ϩ and intracellular content of perforin or granzyme B in CD16 NK cells, were heterogeneous in terms of intracellular contents of perforin ϩ ϩ ␥␦ CD8 T cells, CD4 T cells, and T cells. Freshly isolated PBMC were and granzyme B, the cells attracted to soluble fractalkine displayed stained indirectly with 2A9-1 anti-CX3CR1 mAb and directly with labeled high strong staining for intracellular perforin and granzyme B (Fig. mAbs to various surface molecules. After surface staining, cells were ﬁxed, 5a). Similar results were obtained for CD4ϩ T cells (data not

permeabilized, and stained with anti-perforin mAb or anti-granzyme B Downloaded from mAb. Regardless of the surface markers, CX3CR1-positive cells almost shown). If spontaneous migration was subtracted, soluble frac- completely overlapped with those containing high amounts of perforin (up- talkine almost exclusively attracted cells possessing intracellular per panels) and granzyme B (lower panel). perforin and granzyme B (Fig. 5b). We also conﬁrmed that neutralizing mAbs to fractalkine (3A5 and 3H7), but not control mouse IgG, effectively blocked fractalkine-induced migration of ϩ ϩ ϩ Thus, CX3CR1 is commonly expressed on circulating cytotoxic CD16 NK cells and granzyme B-positive CD8 and CD4 T effector lymphocytes containing intracellular cytotoxic granules cells (Fig. 5c). http://www.jimmunol.org/ regardless of their lineage and mode of target cell recognition.

Correlation of cytotoxic activity with CX3CR1 expression To test the immediate cytotoxic activity of CX3CR1-expressing T cells, we puriﬁed CD8ϩ T cells by positive selection and sorted them into CX3CR1ϩ and CX3CR1Ϫ fractions. We measured the cytotoxic activities of the original and sorted populations by anti-

CD3 mAb-mediated redirected cytotoxicity assay (27). As shown by guest on September 26, 2021 in Fig. 4, CX3CR1ϩCD8ϩ T cells indeed showed much greater cytotoxic activity than presorted CD8ϩ T cells. In contrast, CX3CR1ϪCD8ϩ 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 CD8ϩ T lympho-

FIGURE 5. Preferential migration of T cells containing intracellular perforin and granzyme B to soluble fractalkine. Cells migrated through a FIGURE 4. Cytotoxic activity of CX3CR1ϩ and CX3CR1Ϫ CD8ϩ T monolayer of ECV304 cells into lower chambers containing soluble frac- cell subsets. CD8ϩ T cells were purified from freshly isolated PBMC. talkine at 10 nM were stained for surface CD8 or CD4, fixed, permeabil- Using anti-CX3CR1 mAb 2A9-1, purified CD8ϩ T cells were further ized, and stained for intracellular perforin or granzyme B. a, Flow cyto- sorted into CX3CR1ϩ and CX3CR1Ϫ fractions. The cytotoxic activity of metric profiles of intracellular perforin and granzyme B from input CD8ϩ unfractionated (presorted) and CX3CR1-fractionated CD8ϩ T cells was T cells and CD8ϩ T cells migrated into lower chambers containing soluble immediately analyzed against P815 target cells at various E:T cell ratios in fractalkine. b, Preferential migration of perforin- and granzyme B-contain- the presence of anti-CD3 mAb. The observed cytotoxic activities were ing CD8ϩ T cells and CD4ϩ T cells to soluble fractalkine. c, Inhibition of mostly dependent on the presence of anti-CD3 mAb. Thus, contribution migration of granzyme B-containing CD8ϩ and CD4ϩ T cells and CD16ϩ from NK-like activity was minimum in the present system. Representative NK cells to soluble fractalkine by neutralizing anti-fractalkine mAbs. Rep- results from three separate experiments are shown. Data are the mean Ϯ SE resentative results from three separate experiments are shown as the of triplicate cultures. mean Ϯ SEM. The Journal of Immunology 6177

Differential expression of CX3CR1 and other chemokine chose macrophage inflammatory protein-1␤ (MIP-1␤)3/CC ligand receptors 4 (the ligand for CCR5) as a test chemokine and macrophage che- We next examined coexpression of CX3CR1 and other chemokine motactic protein-1 (MCP-1)/CC ligand 2 (the ligand for CCR2) as a negative control. receptors, namely, CCR5, CXCR3, CXCR1, CXCR2, CCR2, and ϩ CCR7. As shown in Fig. 6, even though most CD16ϩ NK cells As shown in Fig. 7a, transmigration of CD8 T cells to soluble expressed CX3CR1 and CXCR1, only small fractions were found fractalkine was reduced by the expression of membrane-bound to coexpress CCR5 and CXCR3. In contrast, most CX3CR1-ex- fractalkine (ECV-FKN). This suggests that adhesion by the mem- pressing ␥␦ T cells coexpressed CCR5, while coexpression of brane-bound form of fractalkine is dominant over cell migration to CXCR3 or CXCR1 was variable from donor to donor. In the case the soluble form of fractalkine. In contrast, transmigration of ϩ ␤ of CD8ϩ and CD4ϩ T cells, substantial fractions of CX3CR1- CD8 T cells to MIP-1 , but not to MCP-1, was significantly enhanced in the presence of membrane-bound fractalkine (Fig. expressing cells coexpressed CCR5 and CXCR1 at high to inter- ϩ mediate levels, while only minor fractions were found to coexpress 7a). Furthermore, only CD8 T cells carrying granzyme B (the ϩ ␤ CXCR3 or CCR2. In contrast, CCR7, the chemokine receptor es- CX3CR1 fraction) showed enhanced migration to MIP-1 in the sential for lymphocyte homing to secondary lymphoid tissues (2) presence of membrane-bound fractalkine (Fig. 7b). The enhancing was clearly negative on CX3CR1-expressing cells regardless of effects were completely blocked by a neutralizing mAb to fracta- their lineages. CXCR2 was negative on any lymphocyte subsets lkine (3A5), but not by control IgG. The enhancing effects were examined, although it was clearly positive on both monocytes and not observed when PBMCs were added to the upper chamber to- granulocytes. These results indicate that expression of CX3CR1 is gether with 10 nM soluble fractalkine or preactivated with 300 nM Downloaded from regulated quite differently from other chemokine receptors. soluble fractalkine (data not shown). In contrast, migration of CD8ϩ T cells without granzyme B (the CX3CR1Ϫ fraction) to Promoting effect of membrane-bound fractalkine on migration of MIP-1␤ or migration induced by MCP-1, the latter being selective Ϫ CX3CR1-expressing lymphocytes to secondary chemokines for the granzyme B fraction, was not affected by the presence of membrane-bound fractalkine (Fig. 7b). The enhanced migration of The membrane-bound fractalkine has been shown to mediate firm granzyme B-positive CD8ϩ T cells was mainly due to augmenta- adhesion and activation of CX3CR1-expressing cells (11, 12, 28, tion of maximal response to MIP-1␤ by membrane-bound frac- http://www.jimmunol.org/ 29). However, it is not clear how membrane-bound fractalkine talkine (Fig. 7c). affects their migration. We therefore performed transmigration as- Since the majority of NK cells coexpress CX3CR1 and CXCR1 say using ECV304 cells and fractalkine-transfected ECV304 (Fig. 6) (30), we further examined the effect of membrane-bound (ECV-FKN) cells to examine its effects on migration of CX3CR1- fractalkine on migration of CD16ϩ NK cells to IL-8/CXC ligand expressing lymphocytes to soluble fractalkine and the second che- 8 (the ligand of CXCR1). As shown in Fig. 8a, migration of mokine. We have previously shown that TNF-activated HUVECs CD16ϩ NK cells to IL-8 was also strongly enhanced by mem- and ECV-FKN cells express similar levels of fractalkine, indicat- brane-bound fractalkine. The enhancing effects were again effec- ing that ECV-FKN cells express physiologically relevant levels of tively blocked by a neutralizing mAb to fractalkine (3A5). The fractalkine (12). Since a substantial fraction of CX3CR1-express- ϩ by guest on September 26, 2021 ϩ enhanced migration of CD16 NK cells was mainly due to aug- ing CD8 T cells coexpress CCR5, but not CCR2 (Fig. 6), we mentation of the maximal response to IL-8 by membrane-bound fractalkine (Fig. 8c). Thus, membrane-bound fractalkine not only mediates the adhesion of CX3CR1-expressing cells, but also enhances their migration to the secondary chemokines.

Discussion Cytotoxic lymphocytes, which include NK cells, ␥␦ T cells, CD8ϩ T cells, and a minor fraction of CD4ϩ 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 (31Ð33). 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 CD16ϩ NK cells and the majority of ␥␦ 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 CD8ϩ 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, CD4ϩ T cells expressing CX3CR1 were also terminally differentiated and FIGURE 6. Differential expression of CX3CR1 and other chemokine contained intracellular perforin and granzyme B (Figs. 1 and 3), receptors in T cell subsets and NK cells. Freshly isolated PBMC were indirectly stained with anti-CX3CR1 mAb 2A9-1 and directly labeled with 3 Abbreviations used in this paper: MIP, macrophage inﬂammatory protein; DTPA, mAbs to CD16, ␥␦ TCR, CD8, or CD4 as well as those to CCR5, CXCR3, diethylene-triamine-penta-acetic acid; MCP, macrophage chemotactic protein; TCE, CXCR1, CCR2, and CCR7. cytotoxic effector T cell. 6178 FRACTALKINE AND CX3CR1 IN MIGRATION OF CYTOTOXIC LYMPHOCYTES

FIGURE 8. Effects of membrane-bound fractalkine on migration of ϩ

CD16 NK cells to IL-8. Transmigration assays were conducted using Downloaded from Transwell plates with a preformed monolayer of ECV304 or ECV304 stably expressing membrane-bound fractalkine (ECV-FKN). a, Enhanced migration of CD16ϩ 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 CD16ϩ NK cells to various concentrations of IL-8 in the presence or the absence of membrane-bound frac-

talkine. Migration efﬁciency to IL-8 was mainly enhanced by membrane- http://www.jimmunol.org/ bound fractalkine.

FIGURE 7. Effects of membrane-bound fractalkine on migration of CD8ϩ T cells to soluble fractalkine and other chemokines. Transmigration assays were conducted using Transwell plates with a preformed monolayer do not know the exact molecular mechanisms of this enhancement, of ECV304 or ECV304 stably expressing membrane-bound fractalkine but integrin activation and cell polarization induced by membrane- (ECV-FKN). a, Reduced migration of CD8ϩ T cells to soluble fractalkine bound fractalkine may have promoting effects on migratory re- ϩ (10 nM) and enhanced migration of CD8 T cells to MIP-1␤ (10 nM) by sponses to secondary chemokines (12, 28). In this context it is membrane-bound fractalkine. b, Selective enhancement of migration of ϩ interesting to examine whether P- and E-selectin also promote the ␤ by guest on September 26, 2021 granzyme B-positive CD8 T cells to MIP-1 by membrane-bound frac- migration of Th1 cells toward Th1-directed chemokines even in talkine. The enhancing effects were effectively abrogated with neutralizing the static condition. Thus, fractalkine expressed on inflamed en- mAbs to fractalkine. No such enhancement by membrane-bound frac- ϩ dothelium plays dual functions: 1) as a selective capturing mole- talkine was seen in migration of granzyme B-negative CD8 T cells to MIP-1␤ or MCP-1 (10 nM each). c, Migration of granzyme B-positive cule for rapidly circulating killer effector lymphocytes in the CD8ϩ T cells to various concentrations of MIP-1␤ in the presence or the blood, and 2) as a subsequent cosignaling molecule promoting absence of membrane-bound fractalkine. Migration efficiency to MIP-1␤ their migration to secondary chemokines. was mainly enhanced by membrane-bound fractalkine. 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 demonstrating that CX3CR1 is a good surface marker for CD4ϩ T functionally distinct subsets. CCR7Ϫ memory T cells, termed ef- cells with cytotoxic activity. Predominant expression of Fas ligand fector memory T cells, preferentially migrate to inflamed tissues on terminally differentiated CD8ϩ T cells with remarkable lytic and display immediate effector functions. In contrast, CCR7ϩ T activity has also been demonstrated (24). Thus, we propose to call memory cells, termed central memory T cells, lack immediate ef- CX3CR1-expressing T cells, which are terminally differentiated fector functions and preferentially home to secondary lymphoid subsets with cytotoxic activity, cytotoxic effector T cells (TCE). tissues, where they are efficiently stimulated by dendritic cells and Previously, we and others have shown that soluble fractalkine differentiate into CCR7Ϫ effector cells upon secondary stimulation induces migration of NK cells and CD8ϩ T cells, and that the (36). It has been also reported that CD8ϩ cytotoxic T cells start to membrane-bound fractalkine mediates efficient capture, firm ad- express perforin and granzymes during differentiation to memory/ hesion, and activation of CX3CR1-expressing cells in both static effector stages after antigenic stimulation (19). Consistently, we and flow conditions (11Ð13). In the present study we have further have observed that most CX3CR1ϩ T cells express CD11a, and demonstrated that soluble fractalkine is a selective chemoattractant cytoplasmic perforin and granzyme B, but lack L-selectin and/or for lymphocytes displaying cytotoxic effector phenotypes (Fig. 5). CCR7 (Figs. 2, 3, and 6), indicating that expression of CX3CR1

In addition, the membrane-bound fractalkine enhances migration defines the terminally differentiated TCE ready to infiltrate into of CX3CR1-expressing cells to other chemokines, such as MIP-1␤ inflamed tissues. and IL-8, acting on the same cells (Fig. 7). Even though it remains Recent studies have shown that the expression of some chemo- to be seen whether the membrane-bound fractalkine indeed en- kine receptors correlates with Th1 or Th2 polarization and/or with hances chemotaxis of cells expressing CX3CR1 to other chemo- tissue-selective migration. We found that the majorities of kines in vivo, we propose that fractalkine provides an initial clue CX3CR1-expressing CD8ϩ T cells and CD4ϩ T cells were also for circulating NK cells and TCE to emigrate through inflamed found to coexpress CCR5 and, less frequently, CXCR3 (Fig. 6), endothelium as P- and E-selectin mediate selective recruitment of the chemokine receptors known to be highly selective for Th1 cells Th1 cells by inducing their rolling under flow (35). At present we (2), suggesting that CX3CR1-expressing CD8ϩ and CD4ϩ T cells The Journal of Immunology 6179 partly overlap with Tc1 and Th1, respectively. In this regard, Frati- 9. Springer, T. A. 1994. Traffic signals for lymphocyte recirculation and leukocyte celli et al. (35) reported that fully polarized Th1 cells selectively emigration: the multistep paradigm. Cell 76:301. expressed CX3CR1 and migrated to fractalkine. Most ␥␦ T cells 10. Butcher, E. C., and L. J. Picker. 1996. Lymphocyte homing and homeostasis. Science 272:60. were also found to coexpress CCR5 and CXCR3 (Fig. 6). In con- ϩ 11. Imai, T., K. Hieshima, C. Haskell, M. Baba, M. Nagira, M. Nishimura, trast, except for CX3CR1 and CXCR1, CD16 NK cells were M. Kakizaki, S. Takagi, H. Nomiyama, T. J. Schall, et al. 1997. Identification and found to be negative for most other chemokine receptors (Fig. 6). molecular characterization of fractalkine receptor CX3CR1, which mediates both Therefore, CX3CR1 and CXCR1 (Fig. 7) seem to be the major leukocyte migration and adhesion. Cell 91:521. ϩ 12. Fong, A. M., L. A. Robinson, D. A. Steeber, T. F. Tedder, O. Yoshie, T. Imai, and chemokine receptors involved in trafficking of CD16 NK cells. D. D. Patel. 1998. Fractalkine and CX3CR1 mediate a novel mechanism of leu- ϩ Consistently, Campbell et al. (30) reported that CD16 NK cells kocyte capture, firm adhesion, and activation under physiologic flow. J. Exp. were the predominant population in peripheral blood migrating to Med. 188:1413. IL-8 and fractalkine. 13. Haskell, C. A., M. D. Cleary, and I. F. Charo. 1999. Molecular uncoupling of fractalkine-mediated cell adhesion and signal transduction: rapid flow arrest of Recently, the second membrane-type chemokine, CXC ligand CX3CR1-expressing cells is independent of G-protein activation. J. Biol. Chem. 16, has been described (37, 38). In humans its receptor CXCR6/ 274:10053. Bonzo is expressed on fractions of CD45ROϩ CD8ϩ and CD4ϩ T 14. Garton, K. J., P. J. Gough, C. P. Blobel, G. Murphy, D. R. Greaves, cells as well as small subsets of CD16ϩ NK cells and ␥␦ T cells P. J. Dempsey, and E. W. Raines. 2001. TACE (ADAM17) mediates the cleavage and shedding of fractalkine (CX3CL1). J. Biol. Chem. 276:37993. and is coordinately regulated with CCR5 (39). A recent study has 15. Furuichi, K., T. Wada, Y. Iwata, N. Sakai, K. Yoshimoto, M. Shimizu, further shown that CXCR6 is expressed by subsets of Th1 and Tc1, K. Kobayashi, K. Takasawa, H. Kida, S. Takeda, et al. 2001. Upregulation of but not by Th2 or Tc2 cells (40). In contrast to the almost complete fractalkine in human crescentic glomerulonephritis. Nephron 87:314. match of surface expression of CX3CR1 and possession of pre- 16. McDermott, D. H., J. S. Colla, C. A. Kleeberger, M. Plankey, P. S. Rosenberg, Downloaded from formed perforin and granzyme B in various lymphocyte lineages, E. D. Smith, P. A. Zimmerman, C. Combadiere, S. F. Leitman, R. A. Kaslow, et ϩ al. 2000. Genetic polymorphism in CX3CR1 and risk of HIV disease. Science including CD8 T cells (Fig. 3), however, only an average 34% of 290:2031. ϩ ϩ granzyme A CD8 T cells were shown to be positive for CXCR6 17. Faure, S., L. Meyer, D. Costagliola, C. Vaneensberghe, E. Genin, B. Autran, (40). Thus, we assume that, like CCR5, CXCR6 is expressed only J. F. Delfraissy, D. H. McDermott, P. M. Murphy, P. Debre, et al. 2000. Rapid progression to AIDS in HIVϩ individuals with a structural variant of the che- on subsets of CX3CR1-expressing cytotoxic effector lymphocytes mokine receptor CX3CR1. Science 287:2274. and on noncytotoxic effector cells. It remains to be seen whether 18. Moatti, D., S. Faure, F. Fumeron, M. Amara, P. Seknadji, D. H. McDermott, http://www.jimmunol.org/ ϩ ϩ there are any differences between the CXCR6 CX3CR1 and P. Debre, M. C. Aumont, P. M. Murphy, D. de Prost, et al. 2001. Polymorphism CXCR6ϪCX3CR1ϩ subsets of cytotoxic lymphocytes in terms of in the fractalkine receptor CX3CR1 as a genetic risk factor for coronary artery disease. Blood 97:1925. their functional property and tissue migration (40, 41). 19. Shresta, S., C. T. Pham, D. A. Thomas, T. A. Graubert, and T. J. Ley. 1998. How In conclusion, we have demonstrated that fractalkine and do cytotoxic lymphocytes kill their targets? Curr. Opin. Immunol. 10:581. CX3CR1 are likely to play dual roles in the recruitment of effector 20. Hasegawa, H., T. Nomura, M. Kohno, N. Tateishi, Y. Suzuki, N. Maeda, lymphocytes with full cytotoxic activity, suggesting important R. Fujisawa, O. Yoshie, and S. Fujita. 2000. Increased chemokine receptor pathophysiological roles of fractalkine in vascular and tissue inju- CCR7/EBI1 expression enhances the infiltration of lymphoid organs by adult T-cell leukemia cells. Blood 95:30. ries (42). CX3CR1 is also valuable as a highly selective surface 21. Granberg, C., K. Blomberg, I. Hemmila, and T. Lovgren. 1988. Determination of by guest on September 26, 2021 marker for terminally differentiated TCE, whose selective dysfunc- cytotoxic T lymphocyte activity by time-resolved fluorometry using europium- tion is likely to be associated with various chronic diseases, such labelled concanavalin A-stimulated cells as targets. J. Immunol. Methods 114: as HIV-1 infection (30). 191. 22. Trowbridge, I. S., and M. L. Thomas. 1994. CD45: an emerging role as a protein tyrosine phosphatase required for lymphocyte activation and development. Annu. Rev. Immunol. 12:85. Acknowledgments 23. Hamann, D., M. T. Roos, and R. A. van Lier. 1999. Faces and phases of human We thank Dr. Yoshimi Takai for constant support and encouragement; Drs. CD8 T-cell development. Immunol. Today 20:177. Hiroshi Kawamoto, Tatsuo Kina, and Yoshimoto Katsura for support with 24. Hamann, D., P. A. Baars, M. H. Rep, B. Hooibrink, S. R. Kerkhof-Garde, M. R. Klein, and R. A. van Lier. 1997. Phenotypic and functional separation of cell sorting; and Dr. Hitoshi Hasegawa for an anti-CCR7 mAb. ϩ memory and effector human CD8 T cells. J. Exp. Med. 186:1407. 25. Arbones, M. L., D. C. Ord, K. Ley, H. Ratech, C. Maynard-Curry, G. Otten, D. J. Capon, and T. F. Tedder. 1994. Lymphocyte homing and leukocyte rolling References and migration are impaired in L-selectin-deficient mice. Immunity 1:247. 1. Butcher, E. C., M. Williams, K. Youngman, L. Rott, and M. Briskin. 1999. 26. Kagi, D., B. Ledermann, K. Burki, R. M. Zinkernagel, and H. Hengartner. 1996. Lymphocyte trafficking and regional immunity. Adv. Immunol. 72:209. Molecular mechanisms of lymphocyte-mediated cytotoxicity and their role in 2. Yoshie, O., T. Imai, and H. Nomiyama. 2001. Chemokines in immunity. Adv. immunological protection and pathogenesis in vivo. Annu. Rev. Immunol. 14:207. Immunol. 78:57. 27. Mentzer, S. J., J. A. Barbosa, and S. J. Burakoff. 1985. T3 monoclonal antibody 3. Sallusto, F., A. Lanzavecchia, and C. R. Mackay. 1998. Chemokines and che- activation of nonspecific cytolysis: a mechanism of CTL inhibition. J. Immunol. mokine receptors in T-cell priming and Th1/Th2-mediated responses. Immunol. 135:34. Today 19:568. 28. Goda, S., T. Imai, O. Yoshie, O. Yoneda, H. Inoue, Y. Nagano, T. Okazaki, 4. Imai, T., M. Nagira, S. Takagi, M. Kakizaki, M. Nishimura, J. Wang, P. W. Gray, H. Imai, E. T. Bloom, N. Domae, et al. 2000. CX3C-chemokine, fractalkine- K. Matsushima, and O. Yoshie. 1999. Selective recruitment of CCR4-bearing enhanced adhesion of THP-1 cells to endothelial cells through integrin-dependent Th2 cells toward antigen-presenting cells by the CC chemokines thymus and and -independent mechanisms. J. Immunol. 164:4313. activation-regulated chemokine and macrophage-derived chemokine. Int. Immu- 29. Yoneda, O., T. Imai, S. Goda, H. Inoue, A. Yamauchi, T. Okazaki, H. Imai, nol. 11:81. O. Yoshie, E. T. Bloom, N. Domae, et al. 2000. Fractalkine-mediated endothelial 5. Yamamoto, J., Y. Adachi, Y. Onoue, Y. S. Adachi, Y. Okabe, T. Itazawa, cell injury by NK cells. J. Immunol. 164:4055. M. Toyoda, T. Seki, M. Morohashi, K. Matsushima, et al. 2000. Differential 30. Campbell, J. J., S. Qin, D. Unutmaz, D. Soler, K. E. Murphy, M. R. Hodge, expression of the chemokine receptors by the Th1- and Th2-type effector popu- ϩ lations within circulating CD4ϩ T cells. J. Leukocyte Biol. 68:568. L. Wu, and E. C. Butcher. 2001. Unique subpopulations of CD56 NK and NK-T 6. Andrew, D. P., N. Ruffing, C. H. Kim, W. Miao, H. Heath, Y. Li, K. Murphy, peripheral blood lymphocytes identified by chemokine receptor expression rep- J. J. Campbell, E. C. Butcher, and L. Wu. 2001. C-C chemokine receptor 4 ertoire. J. Immunol. 166:6477. expression defines a major subset of circulating nonintestinal memory T cells of 31. Wu, Z., E. R. Podack, J. M. McKenzie, K. J. Olsen, and M. Zakarija. 1994. both Th1 and Th2 potential. J. Immunol. 166:103. Perforin expression by thyroid-infiltrating T cells in autoimmune thyroid disease. 7. Bazan, J. F., K. B. Bacon, G. Hardiman, W. Wang, K. Soo, D. Rossi, Clin. Exp. Immunol. 98:470. D. R. Greaves, A. Zlotnik, and T. J. Schall. 1997. A new class of membrane- 32. Muller, S., J. Lory, N. Corazza, G. M. Griffiths, K. Z’Graggen, L. Mazzucchelli, bound chemokine with a CX3C motif. Nature 385:640. A. Kappeler, and C. Mueller. 1998. Activated CD4ϩ and CD8ϩ cytotoxic cells 8. Butcher, E. C. 1991. Leukocyte-endothelial cell recognition: three (or more) steps are present in increased numbers in the intestinal mucosa from patients with to specificity and diversity. Cell 67:1033. active inflammatory bowel disease. Am. J. Pathol. 152:261. 6180 FRACTALKINE AND CX3CR1 IN MIGRATION OF CYTOTOXIC LYMPHOCYTES

33. Rubesa, G., E. R. Podack, J. Sepcic, and D. Rukavina. 1997. Increased perforin 38. Wilbanks, A., S. C. Zondlo, K. Murphy, S. Mak, D. Soler, P. Langdon, expression in multiple sclerosis patients during exacerbation of disease in pe- D. P. Andrew, L. Wu, and M. Briskin. 2001. Expression cloning of the STRL33/ ripheral blood lymphocytes. J. Neuroimmunol. 74:198. BONZO/TYMSTR ligand reveals elements of CC, CXC, and CX3C chemokines. 34. Yasukawa, M., H. Ohminami, J. Arai, Y. Kasahara, Y. Ishida, and S. Fujita. 2000. J. Immunol. 166:5145. Granule exocytosis, and not the fas/fas ligand system, is the main pathway of ϩ ϩ 39. Unutmaz, D., W. Xiang, M. J. Sunshine, J. Campbell, E. Butcher, and cytotoxicity mediated by alloantigen-specific CD4 as well as CD8 cytotoxic T D. R. Littman. 2000. The primate lentiviral receptor Bonzo/STRL33 is coordi- lymphocytes in humans. Blood 95:2352. nately regulated with CCR5 and its expression pattern is conserved between 35. Austrup, F., D. Vestweber, E. Borges, M. Lohning, R. Brauer, U. Herz, H. Renz, human and mouse. J. Immunol. 165:3284. R. Hallmann, A. Scheffold, A. Radbruch, et al. 1997. P- and E-selectin mediate 40. Kim, C. H., E. J. Kunkel, J. Boisvert, B. Johnston, J. J. Campbell, recruitment of T-helper-1 but not T-helper-2 cells into inflamed tissues. Nature M. C. Genovese, H. B. Greenberg, and E. C. Butcher. 2001. Bonzo/CXCR6 385:81. expression defines type 1-polarized T-cell subsets with extralymphoid tissue 36. Sallusto, F., D. Lenig, R. Forster, M. Lipp, and A. Lanzavecchia. 1999. Two homing potential. J. Clin. Invest. 107:595. subsets of memory T lymphocytes with distinct homing potentials and effector functions. Nature 401:708. 41. Masopust, D., V. Vezys, A. L. Marzo, and L. Lefrancois. 2001. Preferential 37. Matloubian, M., A. David, S. Engel, J. E. Ryan, and J. G. Cyster. 2000. A localization of effector memory cells in nonlymphoid tissue. Science 291:2413. transmembrane CXC chemokine is a ligand for HIV-coreceptor Bonzo. Nat. Im- 42. Umehara, H., E. Bloom, T. Okazaki, N. Domae, and T. Imai. 2001. Fractalkine munol. 1:298. and vascular injury. Trends Immunol. 22:602. Downloaded from http://www.jimmunol.org/ by guest on September 26, 2021