Critical Role for the Receptor CXCR6 in Homeostasis and Activation of CD1d-Restricted NKT Cells

This information is current as Elitza Germanov, Linnea Veinotte, Robyn Cullen, Erin of September 27, 2021. Chamberlain, Eugene C. Butcher and Brent Johnston J Immunol 2008; 181:81-91; ; doi: 10.4049/jimmunol.181.1.81 http://www.jimmunol.org/content/181/1/81 Downloaded from

References This article cites 62 articles, 36 of which you can access for free at: http://www.jimmunol.org/content/181/1/81.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

• Fast Publication! 4 weeks from acceptance to publication

by guest on September 27, 2021 *average

Subscription Information about subscribing to The Journal of 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 © 2008 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Critical Role for the CXCR6 in Homeostasis and Activation of CD1d-Restricted NKT Cells1

Elitza Germanov,2* Linnea Veinotte,2* Robyn Cullen,† Erin Chamberlain,* Eugene C. Butcher,§ and Brent Johnston3*†‡

NK T (NKT) cells play important roles in the regulation of diverse immune responses. However, little is known about the mechanisms that regulate homeostasis and activation of these cells. Thymic NKT cells up-regulated the chemokine receptor CXCR6 following positive selection and migrated toward CXCL16 in vitro. However, CXCR6 was not essential for thymic development or maturation. In contrast, liver and lung NKT cells were depleted in CXCR6؉/؊ and CXCR6؊/؊ mice. The reduction in liver and lung NKT cells coincided with an increase in bone marrow NKT cells, suggesting a redistribution of NKT -cells in CXCR6؊/؊ animals. In wild-type mice, CXCL16 neutralization reduced accumulation of mature NK1.1؉, but not imma ؊ ؊ ture NK1.1 NKT cell recent thymic emigrants in the liver. Given that thymic NKT cells are preferentially exported as NK1.1 Downloaded from cells, this suggests an additional role for CXCR6/CXCL16 in maturation or survival of immature liver NKT cells. CXCL16 blockade did not deplete resident NK1.1؉ NKT cells, indicating that CXCR6/CXCL16 are not required to retain mature NKT cells in the liver. production by liver and spleen NKT cells was impaired in CXCR6؊/؊ mice following in vivo stimulation with ␣-galactosylceramide, implicating a novel role for CXCR6 in NKT cell activation. Reduced IFN-␥ production was not due to an intrinsic defect as production was normal following PMA and ionomycin stimulation. Preformed transcripts for IL-4, but not /IFN-␥, were reduced in CXCR6؊/؊ liver NKT cells. These data identify critical roles for CXCR6/CXCL16 in NKT cell activation http://www.jimmunol.org and the regulation of NKT cell homeostasis. The Journal of Immunology, 2008, 181: 81–91.

he NK T (NKT)4 cells comprise a distinct lineage of lym- from other rare NKT cell subsets and TCR diverse T cells that can phocytes that share functional and phenotypic character- up-regulate NK cell markers following activation (1, 2). T istics with NK cells and effector T (1, 2). NKT cells develop in the thymus and are thought to arise from Most NKT cells in mice express an invariant V␣14J␣18 TCR re- uncommitted precursors that randomly rearrange the V␣14J␣18 arrangement paired preferentially with V␤8.2, V␤7, or V␤2 (3), segments (3, 7, 8). Following TCR rearrangement, developing NKT whereas human NKT cells express a homologous V␣24J␣18 re- cells transit through a CD4ϩCD8ϩ double-positive (DP) step (9–11) by guest on September 27, 2021 arrangement coupled with V␤11 (3, 4). Instead of recognizing pep- and are selected via CD1d expressed on other DP (12, tide Ags, these cells recognize glycolipids presented by the non- 13). Positively selected NKT cells expand and undergo a maturation polymorphic MHC-like molecule CD1d and can be detected using process characterized by sequential up-regulation of CD44 and NK1.1 CD1d tetramers loaded with the glycolipid ␣-galactosylceramide (14, 15). Immature CD44ϩ NK1.1Ϫ NKT cells are exported prefer- (␣-GalCer) (5, 6). These prototypical NKT cells are often referred entially from the thymus and colonize peripheral tissues where they ␣ ϩ to as V 14i T cells or invariant NKT cells to differentiate them undergo final maturation into NK1.1 cells (14, 15). In contrast, most NKT cells that up-regulate NK1.1 in the thymus are retained as a mature resident population (16). *Department of Microbiology and Immunology, †Department of Pathology, and ‡De- In the periphery, NKT cells localize preferentially to the liver, partment of Pediatrics, Dalhousie University, Halifax, Nova Scotia, Canada; and §Laboratory of Immunology and Vascular Biology, Department of Pathology, Stan- spleen, and bone marrow where they respond rapidly to activating ford University School of Medicine, Stanford, CA 94305; and the Center for Molec- stimuli by secreting an array of immunoregulatory includ- ular Biology and Medicine, Veterans Affairs Palo Alto Health Care System, Palo Alto, ␥ CA 94304 ing IFN- and lL-4 (6, 17–19). Consequently, NKT cells have been implicated in a number of diverse immunological processes, including Received for publication September 20, 2007. Accepted for publication April 21, 2008. tumor responses, clearance of pathogens, regulation of , ␣ ␣ The costs of publication of this article were defrayed in part by the payment of page and tolerance induction (20–23). Treatment with -GalCer or -Gal- charges. This article must therefore be hereby marked advertisement in accordance Cer-loaded dendritic cells (DC) induces a potent antitumor response with 18 U.S.C. Section 1734 solely to indicate this fact. by stimulating IFN-␥ production (24–26), while ␣-GalCer has also 1 This work was supported by a grant from the Terry Fox Foundation awarded to B.J. been shown to ameliorate autoimmunity through the induction of Th2 by the National Institute of Canada. B.J. holds the Canada Research Chair in Inflammation and . L.V. is the recipient of a fellowship from the Killam responses (27–29). In contrast to the ability of ␣-GalCer to generate Trusts. E.G. is the recipient of a studentship from the Nova Scotia Health Research both Th1 and Th2 cytokines, OCH, a truncated derivative of ␣-Gal- Foundation. R.C. is the recipient of a Cancer Research Training Program Award with funding from the Dalhousie Cancer Research Program. Cer, induces a preferential Th2 response that can suppress experimen- 2 E.G. and L.V. contributed equally to this work. tal autoimmune encephalitis and collagen-induced arthritis (30, 31), whereas the C-glycoside derivative ␣-C-GalCer induces a selective 3 Address correspondence and reprint requests to Dr. Brent Johnston, Department of Microbiology and Immunology, Dalhousie University, Sir Charles Tupper Medical Th1 response and is very effective at preventing tumor metastasis Building, Room 7C, 5850 College Street, Halifax, Nova Scotia, Canada B3H 1X5. (32). Although the nature of the glycolipid Ag influences the NKT E-mail address: [email protected] cell response, costimulatory molecules are also important. CD40/ 4 Abbreviations used in this paper: NKT cell, NK ; ␣-GalCer, ␣-galactosylce- ␥ ramide; DC, ; DN, double-negative; DP, double-positive; eGFP, en- CD40L interactions are critical for inducing IFN- production (33, hanced GFP; FSC, forward scatter. 34), while IL-4 production is regulated by the integrin LFA-1 (35), www.jimmunol.org 82 CXCR6 MEDIATES NKT CELL HOMEOSTASIS AND ACTIVATION Downloaded from http://www.jimmunol.org/ by guest on September 27, 2021

FIGURE 1. Functional expression of CXCR6 on thymic NKT cells. A, NKT cells were stained with loaded CD1d tetramers and fluorochrome-conjugated Abs against CD4 and CD8 to distinguish CD4ϪCD8Ϫ (DN), CD4ϩCD8ϩ (DP), CD4ϩ, and CD8ϩ subsets. Expression of CXCR6 on NKT cells subsets was determined by eGFP expression in CXCR6-eGFP knock-in heterozygotes, or binding of a CXCL16-Fc chimera in wild-type mice. Nonspecific staining cells were excluded using unloaded tetramers and doublets were excluded by FSC-H vs FSC-A gating (data not shown). B and C, Chemotactic migration assays of thymic NKT cells to CXCL16. Fresh unfractionated thymocytes were allowed to migrate through Transwell inserts (5-␮m pore size) for 2 h. B, Wild-type NKT cells were stained with loaded CD1d tetramers and fluorochrome-conjugated Abs against TCR␤, CD4, and CD8 to distinguish subsets. The proportion of migrated cells in each population was calculated as a fraction of the input population (43, 55). C, NKT cell migration of age- and sex-matched CXCR6ϩ/ϩ, CXCR6ϩ/Ϫ, and CXCR6Ϫ/Ϫ littermates in response to CXCL16. D, NKT cells were stained with loaded CD1d tetramers, CXCL16-Fc, and fluorochrome-conjugated Abs against CD44 and NK1.1 to distinguish sequential stages of NKT cell development: CD44ϪNK1.1Ϫ, CD44ϩNK1.1Ϫ, and CD44ϩNK1.1ϩ. E, NKT cells in the liver, lung, and spleen were identified with loaded CD1d tetramers and Ab against TCR␤. CXCL16-Fc was used to examine CXCR6 expression. FACS plots are representative of at least three independent experiments. Chemotaxis assays show the mean Ϯ SEM of six .p Ͻ 0.05 compared with basal migration. †, p Ͻ 0.05 compared with CXCR6ϩ/ϩ ,ء .individual experiments and CD28-B7 interactions are important for generation of both IFN-␥ accumulation in several inflammatory diseases (47– and IL-4 (33, 36). It is likely that other signals contribute to the reg- 51). In addition to being expressed in inflamed tissues (49–52), ulation of cytokine responses generated by Ag-stimulated NKT cells. CXCL16 is expressed at high levels in the liver and lungs, with induce leukocyte adhesion and migration via inter- lower levels in the spleen, thymus, and other tissues (44, 45, 53). actions with seven transmembrane-spanning G- coupled re- Importantly, CXCL16 is expressed as a transmembrane protein on ceptors, facilitating differential homing of specific leukocyte sub- the surface of APCs including DCs, , and B cells (44, sets to normal and inflamed tissues (37, 38). However, chemokines 53). Therefore, in addition to mediating tissue homing, CXCR6- also play important roles in lymphocyte development, differenti- CXCL16 interactions could facilitate activation or costimulation of ation, and effector functions including cytokine polarization (39, NKT cells. 40). Human and mouse NKT cells express several chemokine re- To investigate the role of CXCL16 and CXCR6 in mediating ceptors, including high levels of CXCR6 (41–44). CXCR6, and its NKT cell development, localization, and function, we examined ligand CXCL16, have been shown to mediate the accumulation of NKT cell populations in CXCR6-deficient mice and wild-type NKT cells in the liver (45) and localization of NKT cells to cardiac mice treated with anti-CXCL16 Abs. CXCR6Ϫ/Ϫ mice exhibited allografts (46). CXCR6 and CXCL16 have also been implicated in reductions in liver and lung NKT cells, whereas NKT cells were The Journal of Immunology 83 Downloaded from http://www.jimmunol.org/

FIGURE 2. CXCR6 is not required for thymic NKT cell development. A, Thymocytes from age- and sex-matched CXCR6ϩ/ϩ, CXCR6ϩ/Ϫ, and Ϫ Ϫ CXCR6 / littermates were stained with loaded CD1d tetramers and flu- by guest on September 27, 2021 orochrome-conjugated Abs against TCR␤ and NK1.1 to identify NKT cell FIGURE 3. CXCR6 mediates NKT cell accumulation in the liver and ϩ/ϩ populations (TCR␤ϩNK1.1ϩ and TCR␤ϩCD1d tetramerϩ). B, Total num- lung. A, Liver lymphocytes from age- and sex-matched CXCR6 , ϩ/Ϫ Ϫ/Ϫ bers of thymic NKT cells (TCR␤ϩNK1.1ϩ or TCR␤ϩCD1d tetramerϩ) CXCR6 , and CXCR6 littermates were stained with loaded CD1d were calculated from the proportion of total cells isolated that were iden- tetramers and fluorochrome-conjugated Abs against TCR␤ and NK1.1 to ϩ ϩ ϩ tified as NKT cells by flow cytometry. FACS plots are representative of six identify NKT cell populations (TCR␤ NK1.1 and TCR␤ CD1d tet- ϩ ϩ independent experiments. Total NKT cell numbers are represented as the ramer ). B, Total numbers of invariant NKT cells (TCR␤ CD1d tet- ϩ mean Ϯ SEM of six individual experiments. ramer ) in the liver, spleen, bone marrow, lungs, and peripheral lymph nodes were calculated from the proportion of cells isolated from each tissue that were identified as NKT cells by flow cytometry. FACS plots are rep- increased in the bone marrow. CXCR6/CXCL16 mediated the ac- resentative of at least three independent experiments. Total NKT cell num- Ϯ cumulation of recent thymic emigrants in the liver and also played bers are represented as the mean SEM of six individual experiments. .p Ͻ 0.05 compared with CXCR6ϩ/ϩ ,ء a role in their maturation and/or survival. However, CXCL16 was not required for retention or survival of NKT cells that had pre- viously matured in the liver. Stimulation with ␣-GalCer revealed a performed with approval from University Committee on Laboratory novel role for CXCR6/CXCL16 in mediating Ag-dependent NKT Animals. cell activation as the production of IL-4 and IFN-␥ and other cy- tokines was impaired in CXCR6Ϫ/Ϫ mice. This study has provided Screening of CXCR6-deficient mice evidence that CXCR6 and CXCL16 play a critical role in NKT cell DNA was isolated from tail clippings using a DNeasy Tissue Kit (Qiagen). activation and cytokine generation in addition to their role in NKT PCR amplification was performed using an Eppendorf Mastercycler ep cell localization. thermocycler with specific primers for CXCR6 (forward, 5Ј-TACGATGG GCACTACGAGGGAG; reverse, 5Ј-GCAAAGAAACCAACAGGGAGA CCAC) and eGFP (forward, 5Ј-TGAAGCAGCACGACTTCTTCAAGT Materials and Methods CC; reverse, 5Ј-TCGTCCATGCCGAGAGTGATCC). Amplification was Mice conducted for 40 cycles with: denaturation at 94°C for 45 s, annealing at 60°C for 60 s, and extension at 72°C for 330 s. PCR results correlated with CXCR6-enhanced GFP (eGFP) knock-in mice, generated by replacing the the mean fluorescence intensity of eGFP expressed by peripheral coding exon for CXCR6 with eGFP (52), were backcrossed against CD8ϩ T lymphocytes obtained by submandibular venous puncture (data C57BL/6J mice for 10 generations. Heterozygotes were bred together, and not shown). litters were genotyped by PCR and FACS. Sex- and age-matched wild-type ϩ/ϩ ϩ/Ϫ Ϫ/Ϫ (CXCR6 ), heterozygote (CXCR6 ), and knockout (CXCR6 ) lit- Cell isolation termates were compared in these studies. Additional male and female C57BL/6J mice were obtained from The Jackson Laboratory. Experiments Mice were anesthetized with a ketamine-xylazine mixture and sacrificed by were conducted when mice were 6–12 wk of age. Mice were housed in the cervical dislocation. Blood was drawn into a heparinized syringe by cardiac Carleton Animal Care Facility at Dalhousie University. Experiments were puncture. Leukocytes were isolated from thymus, spleen, bone marrow 84 CXCR6 MEDIATES NKT CELL HOMEOSTASIS AND ACTIVATION Downloaded from http://www.jimmunol.org/ by guest on September 27, 2021

FIGURE 4. CXCL16 mediates accumulation of FITC-labeled recent thymic NKT cell emigrants (RTE) in the liver. Wild-type mice received intrathymic injections of 10 ␮l of FITC (filtered 1 mg/ml solution) or saline. Animals were treated daily with a blocking CXCL16 mAb (i.p. 250 ␮g) or an -matched control Ab. Cells were isolated from the thymus (A), spleen (B), and liver (C) 72 h after thymic injection and analyzed by flow cytometry using loaded CD1d tetramers and fluorochrome-conjugated Abs against TCR␤ and NK1.1. The frequency of NKT cells (CD1d tetramerϩ) and the proportion of NK1.1ϩ and NK1.1Ϫ NKT cells were assessed in the FITCϩ and FITCϪ gates. FACS plots are representative of six individual mice per group. The frequency of NK1.1ϩ and NK1.1Ϫ CD1d tetramerϩ NKT cells was determined in the FITCϩ gate of the thymus (D), spleen (E), and liver (F) and .p Ͻ 0.05 compared with isotype Ab ,ء .presented as the mean Ϯ SEM of six individual mice per group

(femur and tibia), and peripheral lymph nodes (pooled inguinal, axillary, wire mesh, and isolating the lymphocytes on a 33% Percoll (GE Healthcare brachial, and superficial cervical nodes) by mechanical dispersion through Bio-Sciences) gradient. Lung lymphocytes were isolated by digesting lungs a wire mesh followed by hypotonic erythrocyte lysis. Liver lymphocytes in 300 U/ml type VIII collagenase (Sigma-Aldrich) for 30 min as described were obtained by mincing the tissue, mechanically dispersing it through a previously (54). Lymphocytes were resuspended in RPMI 1640 with 10% The Journal of Immunology 85

Ͼ FCS and allowed to recover in a CO2 incubator for 1 h before chemotaxis experiments. Chemotaxis assays Chemotactic migration assays were performed as previously described (43, 55). Briefly, 1.0 ϫ 106 lymphocytes were placed in the upper chamber of Transwell inserts (5-␮m pore size; Corning Costar). Inserts were placed in wells containing medium alone (basal) or medium plus chemokine. The chemokine CXCL16 was purchased from R&D Systems and used at 5 nM, the optimal concentration determined by titration. After2hofmigration, inserts were removed, and polystyrene beads (Polysciences) were added to each well as an internal standard. Three wells were pooled for each con- dition. Migrated leukocyte populations were stained with Abs and CD1d tetramer to identify cell populations by flow cytometry. Chemotaxis was determined by comparing the bead to cell ratios in the migrated and input populations (43, 55). Flow cytometry The following conjugated mAbs were used in various staining protocols: FITC, PE, PerCP, or allophycocyanin-labeled NK1.1 (clone PK136); Cy- Chrome or allophycocyanin-labeled TCR␤ (clone H57-597); FITC, PerCP,

APC, or allophycocyanin-Cy7-labeled CD4 (clone RM4-5); FITC, PE, PE- Downloaded from Cy7, or allophycocyanin-labeled CD8␣ (clone 53-6.7); FITC-labeled CD44 (clone 1M7). A CXCL16-Fc chimera was used to detect expression Ј of CXCR6 (44). The chimera was detected using a PE-labeled F(ab )2 goat anti-human Fc␥ polyclonal Ab (Jackson ImmunoResearch Laboratories). PE- and allophycocyanin-labeled CD1d tetramers loaded with the ␣-Gal- Cer analog PBS57 were obtained from the National Institutes of Health Tetramer Core Facility (Emory Vaccine Center at Yerkes, Atlanta, GA). Unloaded tetramers were used as a control. Four-color flow cytometry was http://www.jimmunol.org/ performed using a two-laser FACSCalibur with BD CellQuest Pro software (BD Biosciences). Isotype-matched control Abs were used to establish placement of gates and quadrants. Five-color flow cytometry was per- formed using a two-laser BD FACSAria sorter with BD FACSDiva soft- ware (BD Biosciences). Doublet discrimination was determined by anal- ysis of forward light scatter (FSC)-H vs FSC-A.

Labeling and tracking of recent thymic emigrants FIGURE 5. Reduced NK1.1 expression on CXCR6Ϫ/Ϫ NKT cells is not In some mice, 10 ␮l of FITC (filtered 1 mg/ml solution; Sigma-Aldrich) due to differential survival. A and B, Tissue lymphocytes were stained with loaded CD1d tetramers and fluorochrome-conjugated Abs against TCR␤ by guest on September 27, 2021 was injected into each lobe of the thymus to detect recent thymic emigrants Ϫ ϩ in the periphery (14, 15). Animals were treated daily, starting 1 day before and NK1.1 to identify immature (NK1.1 ) and mature (NK1.1 ) NKT cell ϩ ϩ thymic injections, with 250-␮g i.p. injections of a blocking CXCL16 mAb populations. A, Representative NK1.1 staining of CXCR6 / and (clone 142417) or an isotype-matched control Ab (clone 54447), both from CXCR6Ϫ/Ϫ liver NKT cells. B, Frequency of NK1.1 expression on liver, R&D Systems. Cells were isolated from thymus and peripheral tissues 72 h spleen, bone marrow, and thymic NKT cell populations in CXCR6ϩ/ϩ and after thymic FITC injection and analyzed by flow cytometry. CXCRϪ/Ϫ mice. C, Tetramer-positive NKT cells were purified from the ϩ/ϩ Ϫ/Ϫ Activation with ␣-GalCer livers of CXCR6 and CXCR mice and cultured for 18 h. Cells were stained with annexin V and 7-aminoactinomycin D to examine the pro- Mice were injected i.p. with 4 ␮gof␣-GalCer (Alexis Biochemicals; in portion of apoptotic NKT cells. Data are presented as the mean Ϯ SEM of .p Ͻ 0.05 compared with CXCR6ϩ/ϩ ,ء .saline with 0.5% Tween 20) or vehicle alone. After 2 and 24 h, blood four individual experiments 0.9% was collected for measurement of serum cytokine levels using the BD Biosciences Cytometric Bead Array (CBA) for mouse Th1/Th2 cyto- kines (IL-2, IL-4, IL-5, IFN-␥, and TNF). Samples were processed accord- ing to manufacturer’s protocols and analyzed using BD CBA software (version 1.4). In some animals, liver and spleen were harvested at2hfor with the following modifications. Cells were lysed in TRIZOL reagent determination of intracellular IL-4 (clone 11B11) and IFN-␥ (clone (Invitrogen) and mixed 4:1 with chloroform. After centrifugation, the aque- XMG1.2) cytokine staining levels using PE conjugated Abs (eBioscience). ous phase of the TRIZOL-chloroform mixture was applied to an RNeasy Samples were processed using the BD Cytofix/Cytoperm kit without in column (Qiagen). The subsequent steps of the RNeasy kit were followed vitro restimulation. according to manufacturer’s instructions. Reverse transcription was per- formed with SuperScript II (Invitrogen). Quantitative real-time PCR was PMA and ionomycin stimulation performed using an ABI Prism 7000 Sequence Detection System (Applied Biosystems). IFN-␥, IL-4, and GAPDH transcripts were measured using Liver and spleen lymphocytes were surface stained with CD1d-tetramers TaqMan assays (IFN-␥, Mm00801778_m1; IL-4, ␤ and Ab against TCR . Cells were then stimulated in culture for 2 h with Mm00445259) and a TaqMan Rodent GAPDH endogenous reference kit ␮ PMA (50 ng/ml; Sigma-Aldrich) and ionomycin (1 g/ml; Sigma-Aldrich) (Applied Biosystems). Data were analyzed with ABI Prism 7000 SDS soft- ␥ followed by intracellular staining for IL-4 and IFN- . ware using a relative standard curve method according to the manufactur- ␥ NKT cell apoptosis er’s protocol. IFN- and IL-4 expression levels were normalized to the endogenous GAPDH control and expressed relative to mRNA levels in Tetramer-positive NKT cells were purified from liver and spleen lympho- NKT cells isolated from the liver of wild-type mice. cytes using a BD FACSAria cell sorter. Cells were cultured in 96-well plates with RPMI 1640 with 10% FCS. Cells were stained at different times Statistical analysis with annexin V and 7-aminoactinomycin D (BD Biosciences) to differen- Values are reported as means Ϯ SEM except for real time PCR which is tiate apoptotic and necrotic NKT cells. reported as means Ϯ SD. A Mann-Whitney U test was used for compar- Real-time PCR isons between two groups. Comparisons between multiple data groups were performed by nonparametric ANOVA (Kruskal-Wallis test) followed Tetramer-positive NKT cells were purified from liver lymphocytes using a by Dunn’s posttest for multiple comparisons. Statistical significance was BD FACSAria cell sorter. RNA was isolated using an RNeasy kit (Qiagen) set at p Ͻ 0.05. 86 CXCR6 MEDIATES NKT CELL HOMEOSTASIS AND ACTIVATION

FIGURE 6. CXCR6Ϫ/Ϫ NKT cells are impaired in their generation of cytokines after stimulation with ␣-GalCer. Mice were treated i.p. with the NKT cell stimulator a-GalCer (4 ␮g/mouse). A and B, Two hours after ␣-GalCer treatment, spleen and liver lymphocytes were isolated and stained with loaded CD1d tetramers and fluorochrome-conjugated Abs against TCR␤ and NK1.1 to identify CD1d-restricted NKT cells. Cells were fixed and permeabilized for intracellular staining Downloaded from with Abs against IFN-␥, IL-4, or an isotype control (IgG1). A, Representative intracellular cytokine staining of liver NKT cells (gated on TCRbϩ CD1d tetramerϩ cells) from CXCR6ϩ/ϩ and CXCR6Ϫ/Ϫ mice. B, Pro- portion of NK1.1ϩ and NK1.1Ϫ liver and spleen NKT cells positive for intracellular staining with IL-4 and IFN-␥. Data are presented as mean Ϯ SEM of six in- http://www.jimmunol.org/ p Ͻ 0.05 compared with ,ء .dividual experiments CXCR6ϩ/ϩ mice. C, Serum levels of IL-2, IL-4, IFN-␥ and TNF were measured 2 and 24 h after ␣-GalCer stimulation using a cytometric bead array assay. Data are presented as the mean Ϯ SEM of seven individual p Ͻ 0.05 compared with control. †,pϽ ,ء .experiments 0.05 compared with CXCR6ϩ/ϩ. by guest on September 27, 2021

Results NKT cell gate consisted of doublets and cells binding tetramer Thymic NKT cells up-regulate CXCR6 following positive nonspecifically. These cells were excluded from our analysis selection using unloaded tetramers and doublet-exclusion gating (FSC-H vs FSC-A). Very few CD4ϩ or DN NKT cells were present as We and others have reported that NKT cells in blood and periph- doublets or bound unloaded tetramers (data not shown). In func- eral tissues express high levels of the chemokine receptor CXCR6 ϩ (41–44). To examine the expression of CXCR6 on CD1d-re- tional chemotaxis assays, CD4 and DN NKT cells from wild- stricted NKT cells in the thymus, heterozygous mice with a tar- type mice migrated in response to the CXCR6 ligand, CXCL16 geted replacement of CXCR6 by eGFP were analyzed by flow (Fig. 1B). In addition, migration to CXCL16 was absent in Ϫ/Ϫ cytometry. The CXCR6-eGFP reporter was not detected on CXCR6 mice, indicating that CXCR6 is the only receptor CD4ϩCD8ϩ DP NKT cells (Fig. 1A), which are thought to be for CXCL16 expressed by these cells (Fig. 1C). Furthermore, undergoing positive selection (9–11). However, CXCR6-eGFP we were able to demonstrate tight regulation between CXCR6 ϩ was detected at high levels on positively selected CD4ϩ and and eGFP expression as eGFP NKT cells in heterozygous Ϫ CD4ϪCD8Ϫ double-negative (DN) NKT cells. In mice, CD8ϩ mice migrated in response to CXCL16, whereas eGFP NKT CD1d-restricted NKT cells are absent due to negative selection or cells did not (data not shown). To further elucidate the timing down-regulation of CD8 (3, 7). The few residual CD8ϩ NKT of CXCR6 up-regulation following positive selection, we ex- cells that could be detected did not express CXCR6-eGFP. Ex- amined binding of the CXCL16-Fc chimera to NKT cells at pression of eGFP in heterozygotes correlated with staining in different stages of development. NKT cells acquired CXCR6 wild-type mice using a CXCL16-Fc chimera that binds CXCR6 expression coincident with the up-regulation of CD44, but be- (Fig. 1A). As reported by Pellicci et al. (15), Ͼ80% of the DP fore acquisition of NK1.1 expression (Fig. 1D). Following The Journal of Immunology 87 export into the periphery, CXCR6 expression was maintained at high levels on NKT cells (Fig. 1E). CXCR6 is not required for thymic NKT cell development Previously, CXCL16, the ligand for CXCR6, has been detected in the medullary regions of the mouse thymus, suggesting that it could play a role in development (44). However, anal- ysis of age- and sex-matched littermates from the mating of CXCR6-eGFP heterozygotes did not reveal differences in either the proportion or total number of NK1.1ϩ (NK1.1ϩTCR␤ϩ)or CD1d-restricted NKT cells present in the thymus of CXCR6ϩ/ϩ, CXCR6ϩ/Ϫ, or CXCR6Ϫ/Ϫ animals (Fig. 2). This indicates that CXCR6 is not required for NKT cell development in the thymus, and that it could be more important for NKT cell homing or func- tion in the periphery.

CXCR6 mediates NKT cell localization in the liver FIGURE 7. Stimulation of NKT cells with PMA and ionomycin. Liver ϩ/ϩ Ϫ/Ϫ Analysis of NK1.1ϩ (NK1.1ϩTCR␤ϩ) and CD1d-restricted NKT and spleen lymphocytes were isolated from CXCR6 and CXCR6 mice and stained with loaded CD1d tetramers and a fluorochrome-conju- cells in the periphery of CXCR6-deficient mice revealed a signif- Downloaded from gated Ab against TCR␤ to identify CD1d-restricted NKT cells. Cells were icant reduction in the proportion and total number of NKT cells in stimulated in culture with PMA and ionomycin for 2 h followed by fixation ϩ/Ϫ Ϫ/Ϫ the liver of CXCR6 and CXCR6 mice (Fig. 3). NKT cell and permeabilization for intracellular staining with Abs against IFN-␥, populations were reduced in the liver of heterozygotes by 40–50%, IL-4, or an isotype control (IgG1). The frequency of IFN-␥ϩ (A) and IL-4ϩ Ϫ Ϫ whereas CXCR6 / mice exhibited reductions of 75–85%. (B) NKT cells (gated on TCR␤ϩ CD1d tetramerϩ cells) was examined by CXCR6ϩ/Ϫ and CXCR6Ϫ/Ϫ mice also exhibited a reduction in the flow cytometry. Data are presented as the mean Ϯ SEM of four individual Ͻ ϩ/ϩ ء number of NKT cells in the lung (Fig. 3B). These results are con- experiments. , p 0.05 compared with CXCR6 . http://www.jimmunol.org/ sistent with the liver and lung containing high levels of CXCL16 message and protein (44, 45, 53). NKT cell numbers in CXCR6Ϫ/Ϫ mice were increased in the bone marrow (Fig. 3B), suggesting a redistribution of cells from the liver and/or lung. igrated from the thymus but likely plays a role in the maturation or There were no significant differences in NKT cell numbers in the survival of immature NKT cells in the liver rather than a role in spleen and lymph nodes (Fig. 3B). homing. An alternate possibility is that CXCR6/CXCL16 selec- tively mediates homing of NK1.1ϩ thymic emigrants. However, CXCL16 mediates accumulation of thymic NKT cells in the liver mature NK1.1ϩ NKT cells represent a minority of thymic emi- Expression of CXCL16 has been reported in the spleen and liver grants in young mice (16). by guest on September 27, 2021 (44, 45, 53). To examine the role of CXCL16 in the homing of CXCR6 deficiency impairs maturation of NKT cells NKT cells to the periphery, we used a function blocking mAb against CXCL16. Mice were treated with intrathymic FITC injec- In our thymic emigration experiments, it became apparent that tions to enable detection of recent thymic emigrants in the periph- there was a preferential defect in the accumulation of mature ϩ ery (14, 15). Animals were given daily i.p. injections of neutral- CD1d-restricted NK1.1 NKT cells in the liver of mice treated izing CXCL16 Ab or an isotype control starting 1 day before FITC with anti-CXCL16 mAbs. We then examined the expression of Ϫ Ϫ injection. Tissues and cells were isolated 72 h after thymic injec- NK1.1 on the residual NKT cells present in the liver of CXCR6 / tion for determination of the effects of Ab treatment on NKT cell mice. Consistent with the anti-CXCL16 treatment experiments, the Ϫ Ϫ distribution. ELISA revealed that high levels of CXCL16 Ab re- remaining NKT cells in the liver of CXCR6 / mice were pri- Ϫ mained in the serum 24 h after the last Ab treatment (data not marily NK1.1 . In wild-type mice, 75–85% of the liver NKT cells ϩ shown). FITC labeling of thymic NKT cells was equivalent in both were NK1.1 (Fig. 5, A and B). In contrast, only 25% of the the anti-CXCL16 and isotype-treated mice, and there were no ef- residual NKT cells in CXCR6-deficient mice expressed NK1.1. fects of anti-CXCL16 treatment on thymic NKT cell levels (Fig. 4, Impaired maturation was not limited to the liver NKT cells as the A and D). Similarly, there were no differences in accumulation of proportion of NKT cells expressing NK1.1 was also reduced in the recent NKT cell emigrants in the spleen (Fig. 4, B and E). In spleen and bone marrow (Fig. 5B) even though NKT cell numbers contrast, treatment with anti-CXCL16 Ab significantly reduced the were not reduced (Fig. 3B). Unlike Geissmann et al. (45), we did accumulation of FITCϩ NKT cells in the liver (Fig. 4, C and F). not observe differences in the apoptosis of cultured liver NKT cells ϩ ϩ Ϫ Ϫ The reduction in FITCϩ liver NKT cells was observed in the from CXCR6 / and CXCR6 / mice (Fig. 5C), suggesting that NK1.1ϩ population of CD1d-restricted NKT cells, which is more differential NKT cell survival does not mediate the reduced num- Ϫ Ϫ mature than the NK1.1Ϫ subset that is exported from the thymus ber of NKT cells in the liver of CXCR6 / mice. (14, 15). In contrast, anti-CXCL16 treatments had no significant effect on the resident FITCϪ NKT cells in the liver (Fig. 4C and CXCR6 deficiency impairs cytokine production by activated data not shown), suggesting that CXCR6/CXCL16 interactions are NKT cells not required for retention or survival of resident NKT cells that As cytokine production capacity of thymic NKT cells has been have already matured in the liver. Furthermore, CXCR6/CXCL16 shown to change with maturity (14, 15), we investigated intracel- interactions do not appear to be required for maintenance of NK1.1 lular IL-4 and IFN-␥ production as well as serum cytokine levels expression given that NK1.1 expression was maintained on resi- following in vivo stimulation with ␣-GalCer (i.p. 4 ␮g). Two hours dent FITCϪ NKT cells following anti-CXCL16 treatment (Fig. 4C after ␣-GalCer treatment, a large proportion of wild-type NK1.1ϩ and data not shown). These results indicate that CXCL16 is im- and NK1.1Ϫ NKT cells from liver and spleen stained positive for portant for the accumulation of NKT cells that have recently em- the generation of intracellular IL-4 and IFN-␥ (Fig. 6, A and B). 88 CXCR6 MEDIATES NKT CELL HOMEOSTASIS AND ACTIVATION

block the homing of immature thymic NK1.1Ϫ NKT cells to the liver but significantly reduced the accumulation of mature NK1.1ϩ recent thymic emigrants and/or acquisition of the mature NK1.1ϩ NKT cell phenotype by recent thymic emigrants. The potential roles for CXCR6/CXCL16 in homing, retention, maturation, or survival of NKT cells do not need to be mutually exclusive, and it is quite likely that CXCR6/CXCL16 are playing roles in several or all of these processes. Our experiments also identified a novel role FIGURE 8. Examination of preformed cytokine mRNA transcripts in CXCR6ϩ/ϩ and CXCR6Ϫ/Ϫ NKT cells. Preformed mRNA transcripts were for CXCR6 in Ag-dependent NKT cell activation. After in vivo measured by quantitative real-time PCR in populations of liver NKT cells treatment with the NKT cell Ag ␣-GalCer, NKT cells in the liver Ϫ Ϫ purified from CXCR6ϩ/ϩ and CXCR6Ϫ/Ϫ mice by FACS. IFN-␥ and IL-4 and spleen of CXCR6 / mice exhibited a significant impairment expression levels were normalized to the endogenous GAPDH control and in their ability to generate intracellular IL-4 and IFN-␥, and the expressed relative to mRNA levels in NKT cell samples isolated from the levels of serum cytokines were severely reduced at 2 and 24 h ϩ/ϩ Ϯ liver of CXCR6 mice. Data are presented as the mean SD of four posttreatment. The reduction in IL-4 production could be partly p Ͻ 0.05 compared ,ء .independent experiments performed in duplicate ϩ ϩ explained by a reduction in the number of preformed IL-4 mRNA with CXCR6 / . transcripts in CXCR6Ϫ/Ϫ NKT cells. However, there were no sig- nificant differences in the quantity of preformed IFN-␥ transcripts in CXCR6Ϫ/Ϫ NKT cells. These results implicate a critical role for This is consistent with recent findings demonstrating that periph- Downloaded from eral NK1.1Ϫ NKT cells exhibit a mature cytokine profile (56). In CXCR6/CXCL16 in Ag-dependent NKT cell stimulation and cy- contrast, the proportion of NK1.1ϩ and NK1.1Ϫ NKT cells gen- tokine production in addition to their role in NKT cell homeostasis. erating IL-4 and IFN-␥ was significantly reduced in CXCR6Ϫ/Ϫ We found that the expression of CXCR6 on thymic NKT cells mice (Fig. 6, A and B). Similarly, serum levels of IL-2, IL-4, and is up-regulated following positive selection (Fig. 1A). CXCR6 was IFN-␥ were significantly reduced in CXCR6Ϫ/Ϫ mice 2 h after not detected on many DP NKT cells but was expressed on 60– ␣ ϩ -GalCer treatment (Fig. 6C). At 24 h, serum levels of IL-2 and 80% of CD4 and DN NKT cells. Although, NKT cells proceed http://www.jimmunol.org/ IL-4 had reduced to baseline levels in all mice. Serum IFN-␥ in- through a DP developmental step (9–11), the population detected creased further at 24 h, but CXCR6Ϫ/Ϫ mice continued to exhibit by flow cytometry may consist of doublets of NKT cells bound to impaired IFN-␥ production compared with wild-type mice. Serum DP thymocytes or cells that bind tetramer nonspecifically (15). levels of TNF (Fig. 6C) and IL-5 (data not shown) increased to Indeed, the majority of DP NKT cells were detected as doublets or similar levels in CXCR6ϩ/ϩ and CXCR6Ϫ/Ϫ mice. bound to unloaded CD1d tetramers. However, few of the specific To determine whether CXCR6 deficiency results in an intrinsic DP singlets expressed CXCR6 compared with high levels on the ϩ ϩ defect in NKT cell activation, we stimulated CXCR6 / and CD4ϩ and DN NKT cells, suggesting that CXCR6 is up-regulated Ϫ Ϫ CXCR6 / splenocytes and liver lymphocytes in culture with after the DP step. Further support for the up-regulation of CXCR6 PMA and ionomycin (Fig. 7). Intracellular cytokine staining for after thymic NKT cell selection comes from the observation that by guest on September 27, 2021 ϩ/ϩ Ϫ/Ϫ IFN-␥ was equivalent in CXCR6 and CXCR6 NKT cells, CXCR6 is up-regulated concomitantly with CD44 expression. implicating a direct role for CXCR6 in induction of this cytokine. NKT cells up-regulate CD44 following selection but before NK1.1 Intracellular staining for IL-4 was reduced in liver but not spleen is expressed (11, 14). Importantly, CXCR6 was up-regulated ϩ/ϩ NKT cells from CXCR6 mice, suggesting a direct role for before final maturation into NK1.1ϩ NKT cells (Fig. 1D). Because Ϫ/Ϫ CXCR6 as well as an intrinsic developmental defect in CXCR6 the majority of NKT cells exported from the thymus leave before liver NKT cells. up-regulating NK1.1 (14, 15), the expression of CXCR6 before NKT cells contain preformed cytokine transcripts that contrib- this step would be consistent with a role for CXCR6 in mediating ute to their ability to rapidly generate cytokines after stimulation homing or function of recent thymic emigrants in the periphery. (17, 57). To investigate whether impaired cytokine production by We and others have shown previously that CXCR6 is expressed CXCR6Ϫ/Ϫ NKT cells was related to reduced levels of constitutive on subsets of T cells in addition to NKT cells (41–44, 47–53). cytokine mRNA, we examined IL-4 and IFN-␥ transcripts in wild- Ϫ Ϫ However, chemotaxis of peripheral NKT cells to CXCL16 is very type and CXCR6 / NKT cells using quantitative real time PCR low (43), and migration of CXCR6ϩ T cells to CXCL16 requires (Fig. 8). There were no significant differences in the expression of ϩ ϩ Ϫ Ϫ cell activation (44). In contrast, we now show that resting thymic preformed IFN-␥ transcripts by CXCR6 / and CXCR6 / NKT cells. In contrast, liver CXCR6Ϫ/Ϫ NKT cells expressed lower NKT cells exhibited significant migration in response to CXCL16, levels of preformed IL-4 transcripts. suggesting that these cells are more likely to respond to ligand in vivo and CXCR6/CXCL16 could mediate their peripheral tissue Discussion distribution before responses are down-regulated. The basal accu- In this study, we examined the role of CXCR6 in the development, mulation of NKT cells in the liver and lungs of wild-type mice is distribution, and activation of CD1d-restricted NKT cells. Func- consistent with the high levels of CXCL16 expression in these tional CXCR6 expression was up-regulated on developing thymic tissues (44, 45, 53). In addition to playing a role in basal NKT cell NKT cells following positive selection but before acquisition of localization, Jiang et al. (46) have reported that CXCL16 mediates NK1.1 expression. Although CXCR6 was not required for thymic accumulation of NKT cells in cardiac allografts. However, it is NKT cell development, it played a vital role in the accumulation unclear whether the infiltrating NKT cells in this allograft model of NKT cells in the liver and lungs. The redistribution of NKT reflect the preferential accumulation of recent thymic emigrants cells to bone marrow of CXCR6Ϫ/Ϫ mice suggests a role for expressing functional CXCR6 receptor, or results from activation CXCR6 in NKT cell homing or retention in the liver and lungs. signals that mobilize NKT cells from the periphery and up-regulate However, CXCR6 and CXCL16 also appear to be important in CXCR6 function, leading to CXCL16-dependent homing to the mediating maturation and/or survival of liver NKT cells soon after allograft. A role for CXCR6/CXCL16 in basal NKT cell tissue their localization to the liver. Anti-CXCL16 treatment did not distribution is confirmed by the reduction of NKT cells in the liver The Journal of Immunology 89 and lung of CXCR6Ϫ/Ϫ mice (Fig. 3). Furthermore, the redistri- mediating survival or retention of immature NKT cells. Geissmann bution of NKT cells to the bone marrow in CXCR6Ϫ/Ϫ mice sup- et al. (45) reported that CXCR6Ϫ/Ϫ NKT cells exhibited enhanced ports a role for CXCR6 in NKT cell homing or retention in the apoptosis in culture. Although this is an attractive explanation, we liver and lungs. did not observe differences in the apoptosis (annexin V staining) of In contrast to our findings, Geissmann et al. (45) reported that cultured CXCR6ϩ/ϩ and CXCR6Ϫ/Ϫ NKT cells (Fig. 5C). This is ϩ Ϫ NKT cell numbers were normal in the liver of CXCR6 / mice. consistent with the finding that apoptosis is not increased in NKT The reason for this discrepancy is unclear because animals used in cells from CXCL16Ϫ/Ϫ mice (59). In addition, anti-CXCL16 treat- both studies were generated from the same founder mice (52). One ment did not decrease survival of resident liver NKT cells as cell possibility is that Geissmann et al. (45) used mice backcrossed as numbers remained the same as in isotype-treated mice. few as three generations. We used mice backcrossed for 10 gen- It is unclear why CXCR6/CXCL16 would be important for mat- erations and compared age- and sex-matched littermates. Our stud- uration/survival in the liver but not other tissues. Although there ϩ/Ϫ ies indicate that CXCR6 mice are not equivalent to wild-type were small reductions in the proportion of NK1.1ϩ NKT cells in mice and should not be used in the place of wild-type animals. the spleen and bone marrow, there were no reductions in the num- We found that CXCR6 also played a role in basal NKT cell ber of NKT cells in these tissues. It is possible that the requirement accumulation in the lungs. In contrast, Meyer et al. (58) demon- for CXCR6 in the maintenance of liver or lung NKT cells is strated that CCR4 mediates NKT cell recruitment the lung and unique. Although few phenotypic differences have been detected ␣ airways following intranasal stimulation with -GalCer. CCR4 in mouse NKT cells, it has been shown that NKT cells from dif- also mediated basal localization of NKT cells to the airways, but ferent tissue sites are functionally distinct in their ability to prevent not the lung. Therefore, it is likely that CXCR6 and CCR4 mediate tumor metastasis (60). Therefore, the liver NKT cells may repre- Downloaded from homeostasis of NKT cells in different lung compartments under sent a unique subset that is more dependent on CXCR6 for basal conditions. However, in the inflamed lung, CCR4 appears NK1.1ϩ NKT cell maturation. Alternately, it is possible that sufficient for NKT cell recruitment given that lung inflammation Ϫ/Ϫ Ϫ/Ϫ CXCR6 is required for maturation/survival in the liver environ- was ablated in CCR4 but not CXCR6 mice (58). ment but dispensable in other tissues. In addition to alterations in NKT cell tissue distribution, there ϩ CXCR6/CXCL16 were also found to play a critical role in NKT were differences in the number of mature NK1.1 NKT cells in cell activation. NKT cells are potently activated by CD1d-depen- http://www.jimmunol.org/ CXCR6Ϫ/Ϫ mice. This is most likely due to a role for CXCR6/ dent presentation of the glycolipid Ag ␣-GalCer, resulting in the CXCL16 in NKT cell maturation. We observed a reduction in rapid production of immunoregulatory cytokines (6, 17–19). After NK1.1 expression on the residual liver NKT cells in CXCR6Ϫ/Ϫ ␣-GalCer stimulation, we observed deficiencies in both intracel- mice (Fig. 5). Also, the NKT cell populations in the spleen and Ϫ/Ϫ lular cytokine staining in NKT cells and serum cytokine levels in bone marrow of CXCR6 mice exhibited small reductions in the Ϫ Ϫ CXCR6 / mice (Fig. 6). To determine whether CXCR6 plays a proportion of NK1.1-expressing cells even though NKT cell num- direct role in NKT cell stimulation vs an intrinsic developmental bers were normal or increased, respectively. Reduced NK1.1 ex- Ϫ Ϫ defect in CXCR6 / NKT cells, we also stimulated NKT cells pression is unlikely due to the disruption of the NK1.1 allele or ␥ ϩ with PMA and ionomycin. Induction of IFN- was not impaired in by guest on September 27, 2021 loss during backcrossing because staining for NK1.1 NK cells Ϫ/Ϫ Ϫ Ϫ CXCR6 NKT cells following stimulation with PMA and iono- was not reduced in CXCR6 / mice (Fig. 3A), and NKT cells in mycin (Fig. 7), implicating a direct role for CXCR6 in Ag-depen- the wild-type littermates generated from the breeding of ϩ Ϫ dent NKT cell activation. Production of IL-4 by splenic NKT cells CXCR6 / mice all had normal levels of NK1.1 expression. It is was also dependent on a direct role for CXCR6. However, there also unlikely that CXCR6/CXCL16 interactions are required for Ϫ/Ϫ maintenance of NK1.1 expression on mature liver NKT cells be- was an intrinsic deficiency in CXCR6 liver NKT cells as IL-4 cause resident liver NKT cells did not down-regulate NK1.1 ex- induction was still reduced following stimulation with PMA and C ionomycin. This is likely related to a reduction in preformed IL-4 pression following anti-CXCL16 treatment (Fig. 4 ). Rather, these Ϫ/Ϫ data are consistent with a role for CXCR6 in maturation of most transcripts in CXCR6 liver NKT cells (Fig. 8), and suggests an NKT cells in the liver and a subset of NKT cells in the spleen and additional role for CXCR6 in the developmental regulation of pre- bone marrow. A role for maturation in the liver is also supported formed IL-4 transcripts in liver NKT cells. The recent work of by the observation that preformed IL-4 mRNA transcripts were Shimaoka et al. (59) also supports a direct role for CXCR6/ reduced in CXCR6Ϫ/Ϫ NKT cells (Fig. 8). Also, anti-CXCL16 CXCL16 in NKT cell activation. This group found similar defi- ϩ ␣ treatment reduced accumulation of mature NK1.1 NKT cells ciencies in liver NKT cell numbers and -GalCer-induced cyto- Ϫ/Ϫ from recent thymic emigrants (Fig. 4, C and F). However, kine production in CXCL16 mice. The expression of CXCL16 CXCL16 blockade did not reduce the number of mature resident on DCs was critical for optimal cytokine production by cultured NKT cells in the liver (Fig. 4C), indicating that CXCR6/CXCL16 NKT cells in response to ␣-GalCer stimulation. are dispensable for the maintenance and retention of mature NKT ␣-GalCer needs to be presented via CD1d on APCs to activate cells in the liver. NKT cells (19, 25, 61). Because CXCL16 is expressed as a trans- Although a role for CXCR6/CXCL16 in NKT cell maturation is membrane molecule on activated DCs and macrophages (44, 53), likely, other factors may contribute to the reduced NK1.1ϩ NKT it is possible that CXCR6/CXCL16 interactions mediate cell-cell cell numbers in CXCR6Ϫ/Ϫ mice. The increased frequency of adhesion and/or provide costimulatory signals that up-regulate NK1.1Ϫ NKT cells in the spleen and bone marrow could reflect NKT cell function. Indeed, optimal NKT cell activation requires increased distribution of immature recent thymic emigrants to bidirectional signals between NKT cells and DC. In previous stud- these tissues in the absence of CXCR6. Alternatively, CXCR6/ ies, CD40/CD40L interactions and DC-derived IL-12 have been CXCL16 may influence NKT cell survival in the liver. In anti- shown to be critical for the production of IFN-␥ by NKT cells (33, CXCL16-treated mice, there was no increase in the frequency of 34), whereas CD28/B7 interactions are required for the generation NK1.1Ϫ NKT cells within the FITCϩ pool, which would be ex- of both IL-4 and IFN-␥ (33, 36). The requirement for CXCR6 to pected if there was only an impairment in NKT cell maturation. obtain optimal NKT cell activation is consistent with the observa- Similarly, NK1.1Ϫ NKT cells did not accumulate in the liver of tion that chemokine receptor interactions help facilitate activation CXCR6Ϫ/Ϫ mice (Fig. 3). This would favor a role for CXCR6 in of T cells during Ag presentation (40, 62). 90 CXCR6 MEDIATES NKT CELL HOMEOSTASIS AND ACTIVATION

In conclusion, we have characterized the contribution of 20. Seino, K., S. Motohashi, T. Fujisawa, T. Nakayama, and M. Taniguchi. 2006. CXCR6 in NKT cell development, homeostasis and activation. We Natural killer T cell-mediated antitumor immune responses and their clinical applications. Cancer Sci. 97: 807–812. revealed an important role for CXCR6 not only in the control of 21. Tupin, E., Y. Kinjo, and M. Kronenberg. 2007. The unique role of natural killer NKT cell accumulation in the liver and lungs but also in the ac- T cells in the response to microorganisms. Nat. Rev. Microbiol. 5: 405–417. ␣ tivation of NKT cells and regulation of cytokine production. In the 22. Van Kaer, L. 2005. -Galactosylceramide therapy for autoimmune diseases: prospects and obstacles. Nat. Rev. Immunol. 5: 31–42. future, it will be important to characterize the exact role of 23. Nowak, M., and J. Stein-Streilein. 2007. Invariant NKT cells and tolerance. Int. CXCR6/CXCL16 in NKT cell interactions with APCs. Because Rev. Immunol. 26: 95–119. interactions between CXCR6 and CXCL16 appear to be critical for 24. Kawano, T., J. Cui, Y. Koezuka, I. Toura, Y. Kaneko, H. Sato, E. Kondo, M. Harada, H. Koseki, T. Nakayama, et al. 1998. Natural killer-like nonspecific NKT cell function, they could act as therapeutic targets to enhance tumor cell lysis mediated by specific ligand-activated V␣14 NKT cells. Proc. NKT cell responses against tumors and pathogens or regulate tol- Natl. Acad. Sci. USA 95: 5690–5693. 25. Toura, I., T. Kawano, Y. Akutsu, T. Nakayama, T. Ochiai, and M. Taniguchi. erance induction and autoimmunity. 1999. Cutting edge: inhibition of experimental tumor metastasis by dendritic cells pulsed with ␣-galactosylceramide. J. Immunol. 163: 2387–2391. Acknowledgments 26. Hayakawa, Y., S. Rovero, G. Forni, and M. J. Smyth. 2003. ␣-Galactosylceram- ide (KRN7000) suppression of chemical- and oncogene-dependent carcinogene- We thank Dr. D. R. Littman (New York University School of Medicine, sis. Proc. Natl. Acad. Sci. USA 100: 9464–9469. New York, NY) for providing CXCR6-GFP knock-in mice and Dr. Me- 27. Sharif, S., G. A. Arreaza, P. Zucker, Q. S. Mi, J. Sondhi, O. V. Naidenko, hrdad Matloubian (University of California San Francisco, San Francisco, M. Kronenberg, Y. Koezuka, T. L. Delovitch, J. M. Gombert, et al. 2001. Acti- CA) for the gift of CXCL16-Fc chimera. vation of natural killer T cells by ␣-galactosylceramide treatment prevents the onset and recurrence of autoimmune type 1 . Nat. Med. 7: 1057–1062. 28. Jahng, A. W., I. Maricic, B. Pedersen, N. Burdin, O. Naidenko, M. Kronenberg,

Disclosures Y. Koezuka, and V. Kumar. 2001. Activation of natural killer T cells potentiates Downloaded from Eugene C. Butcher is a Director of Bio-Seek Inc., Burlingame, CA. or prevents experimental autoimmune encephalomyelitis. J. Exp. Med. 194: 1789–1799. 29. Saubermann, L. J., P. Beck, Y. P. De Jong, R. S. Pitman, M. S. Ryan, H. S. Kim, References M. Exley, S. Snapper, S. P. Balk, S. J. Hagen, et al. 2000. Activation of natural 1. Godfrey, D. I., H. R. MacDonald, M. Kronenberg, M. J. Smyth, and K. L. Van. killer T cells by ␣-galactosylceramide in the presence of CD1d provides protec- 2004. NKT cells: what’s in a name? Nat. Rev. Immunol. 4: 231–237. tion against colitis in mice. Gastroenterology 119: 119–128. 2. Godfrey, D. I., K. J. Hammond, L. D. Poulton, M. J. Smyth, and A. G. Baxter. 30. Chiba, A., S. Oki, K. Miyamoto, H. Hashimoto, T. Yamamura, and S. Miyake. 2000. NKT cells: facts, functions and fallacies. Immunol. Today 21: 573–583. 2004. Suppression of collagen-induced arthritis by natural killer T cell activation http://www.jimmunol.org/ 3. Lantz, O., and A. Bendelac. 1994. An invariant T cell receptor ␣ chain is used by with OCH, a sphingosine-truncated analog of ␣-galactosylceramide. Arthritis a unique subset of major histocompatibility complex class I-specific CD4ϩ and Rheum. 50: 305–313. CD4Ϫ8Ϫ T cells in mice and humans. J. Exp. Med. 180: 1097–1106. 31. Miyamoto, K., S. Miyake, and T. Yamamura. 2001. A synthetic glycolipid pre- 4. Dellabona, P., E. Padovan, G. Casorati, M. Brockhaus, and A. Lanzavecchia. vents autoimmune encephalomyelitis by inducing TH2 bias of natural killer T 1994. An invariant V␣24-J␣Q/V␤11 T cell receptor is expressed in all individ- cells. Nature 413: 531–534. uals by clonally expanded CD4Ϫ8Ϫ T cells. J. Exp. Med. 180: 1171–1176. 32. Fujii, S., K. Shimizu, H. Hemmi, M. Fukui, A. J. Bonito, G. Chen, R. W. Franck, 5. Benlagha, K., A. Weiss, A. Beavis, L. Teyton, and A. Bendelac. 2000. In vivo M. Tsuji, and R. M. Steinman. 2006. Glycolipid ␣-C-galactosylceramide is a identification of glycolipid -specific T cells using fluorescent CD1d tet- distinct inducer of dendritic cell function during innate and adaptive immune ramers. J. Exp. Med. 191: 1895–1903. responses of mice. Proc. Natl. Acad. Sci. USA 103: 11252–11257. 6. Matsuda, J. L., O. V. Naidenko, L. Gapin, T. Nakayama, M. Taniguchi, 33. Hayakawa, Y., K. Takeda, H. Yagita, L. Van Kaer, I. Saiki, and K. Okumura. C. R. Wang, Y. Koezuka, and M. Kronenberg. 2000. Tracking the response of 2001. Differential regulation of Th1 and Th2 functions of NKT cells by CD28 and

natural killer T cells to a glycolipid antigen using CD1d tetramers. J. Exp. Med. CD40 costimulatory pathways. J. Immunol. 166: 6012–6018. by guest on September 27, 2021 192: 741–754. 34. Kitamura, H., K. Iwakabe, T. Yahata, S. Nishimura, A. Ohta, Y. Ohmi, M. Sato, 7. Kronenberg, M. 2005. Toward an understanding of NKT cell biology: progress K. Takeda, K. Okumura, L. Van Kaer, et al. 1999. The natural killer T (NKT) cell and paradoxes. Annu. Rev. Immunol. 23: 877–900. ligand ␣-galactosylceramide demonstrates its immunopotentiating effect by in- 8. Shimamura, M., T. Ohteki, U. Beutner, and H. R. MacDonald. 1997. Lack of ducing interleukin (IL)-12 production by dendritic cells and IL-12 receptor ex- ϩ directed V␣14-J␣281 rearrangements in NK1 T cells. Eur. J. Immunol. 27: pression on NKT cells. J. Exp. Med. 189: 1121–1128. 1576–1579. 35. Matsumoto, G., E. Kubota, Y. Omi, U. Lee, and J. M. Penninger. 2004. Essential 9. Gapin, L., J. L. Matsuda, C. D. Surh, and M. Kronenberg. 2001. NKT cells derive role of LFA-1 in activating Th2-like responses by ␣-galactosylceramide-activated from double-positive thymocytes that are positively selected by CD1d. Nat. Im- NKT cells. J. Immunol. 173: 4976–4984. munol. 2: 971–978. 36. Ikarashi, Y., R. Mikami, A. Bendelac, M. Terme, N. Chaput, M. Terada, T. Tursz, 10. Dao, T., D. Guo, A. Ploss, A. Stolzer, C. Saylor, T. E. Boursalian, J. S. Im, and E. Angevin, F. A. Lemonnier, H. Wakasugi, and L. Zitvogel. 2001. Dendritic cell D. B. Sant’Angelo. 2004. Development of CD1d-restricted NKT cells in the maturation overrules H-2D-mediated natural killer T (NKT) cell inhibition: crit- mouse thymus. Eur. J. Immunol. 34: 3542–3552. ical role for B7 in CD1d-dependent NKT cell interferon ␥ production. J. Exp. 11. Benlagha, K., D. G. Wei, J. Veiga, L. Teyton, and A. Bendelac. 2005. Charac- Med. 194: 1179–1186. terization of the early stages of thymic NKT cell development. J. Exp. Med. 202: 37. Johnston, B., and E. C. Butcher. 2002. Chemokines in rapid leukocyte adhesion 485–492. triggering and migration. Semin. Immunol. 14: 83–92. 12. Coles, M. C., and D. H. Raulet. 2000. NK1.1ϩ T cells in the liver arise in the 38. Ebert, L. M., P. Schaerli, and B. Moser. 2005. Chemokine-mediated control of T thymus and are selected by interactions with class I molecules on CD4ϩCD8ϩ cell traffic in lymphoid and peripheral tissues. Mol. Immunol. 42: 799–809. cells. J. Immunol. 164: 2412–2418. 39. Campbell, D. J., C. H. Kim, and E. C. Butcher. 2003. Chemokines in the systemic 13. Wei, D. G., H. Lee, S. H. Park, L. Beaudoin, L. Teyton, A. Lehuen, and organization of immunity. Immunol. Rev. 195: 58–71. A. Bendelac. 2005. Expansion and long-range differentiation of the NKT cell lineage in mice expressing CD1d exclusively on cortical thymocytes. J. Exp. 40. Viola, A., R. L. Contento, and B. Molon. 2006. T cells and their partners: the Med. 202: 239–248. chemokine dating agency. Trends Immunol. 27: 421–427. 14. Benlagha, K., T. Kyin, A. Beavis, L. Teyton, and A. Bendelac. 2002. A thymic 41. Kim, C. H., B. Johnston, and E. C. Butcher. 2002. Trafficking machinery of NKT precursor to the NK T cell lineage. Science 296: 553–555. cells: shared and differential chemokine receptor expression among ␣ ϩ ␤ ϩ 15. Pellicci, D. G., K. J. Hammond, A. P. Uldrich, A. G. Baxter, M. J. Smyth, and V 24 V 11 NKT cell subsets with distinct cytokine-producing capacity. D. I. Godfrey. 2002. A natural killer T (NKT) cell developmental pathway in- Blood 100: 11–16. volving a thymus-dependent NK1.1ϪCD4ϩ CD1d-dependent precursor stage. 42. Thomas, S. Y., R. Hou, J. E. Boyson, T. K. Means, C. Hess, D. P. Olson, J. Exp. Med. 195: 835–844. J. L. Strominger, M. B. Brenner, J. E. Gumperz, S. B. Wilson, and A. D. Luster. 16. Berzins, S. P., F. W. McNab, C. M. Jones, M. J. Smyth, and D. I. Godfrey. 2006. 2003. CD1d-restricted NKT cells express a chemokine receptor profile indicative Long-term retention of mature NK1.1ϩ NKT cells in the thymus. J. Immunol. of Th1-type inflammatory homing cells. J. Immunol. 171: 2571–2580. 176: 4059–4065. 43. Johnston, B., C. H. Kim, D. Soler, M. Emoto, and E. C. Butcher. 2003. Differ- 17. Matsuda, J. L., L. Gapin, J. L. Baron, S. Sidobre, D. B. Stetson, M. Mohrs, ential chemokine responses and homing patterns of murine TCR␣␤ NKT cell R. M. Locksley, and M. Kronenberg. 2003. Mouse V␣14i natural killer T cells subsets. J. Immunol. 171: 2960–2969. are resistant to cytokine polarization in vivo. Proc. Natl. Acad. Sci. USA 100: 44. Matloubian, M., A. David, S. Engel, J. E. Ryan, and J. G. Cyster. 2000. A 8395–8400. transmembrane CXC chemokine is a ligand for HIV-coreceptor Bonzo. Nat. Im- 18. Crowe, N. Y., A. P. Uldrich, K. Kyparissoudis, K. J. Hammond, Y. Hayakawa, munol. 1: 298–304. S. Sidobre, R. Keating, M. Kronenberg, M. J. Smyth, and D. I. Godfrey. 2003. 45. Geissmann, F., T. O. Cameron, S. Sidobre, N. Manlongat, M. Kronenberg, Glycolipid antigen drives rapid expansion and sustained cytokine production by M. J. Briskin, M. L. Dustin, and D. R. Littman. 2005. Intravascular immune ϩ NK T cells. J. Immunol. 171: 4020–4027. surveillance by CXCR6 NKT cells patrolling liver sinusoids. PLoS Biol. 19. Fujii, S., K. Shimizu, M. Kronenberg, and R. M. Steinman. 2002. Prolonged 3: e113. IFN-␥-producing NKT response induced with ␣-galactosylceramide-loaded DCs. 46. Jiang, X., T. Shimaoka, S. Kojo, M. Harada, H. Watarai, H. Wakao, N. Ohkohchi, Nat. Immunol. 3: 867–874. S. Yonehara, M. Taniguchi, and K. Seino. 2005. Cutting edge: critical role of The Journal of Immunology 91

CXCL16/CXCR6 in NKT cell trafficking in allograft tolerance. J. Immunol. 175: thelial chemokine/CCL28) selectively attracts IgA plasmablasts. J. Immunol. 2051–2055. 170: 3799–3805. 47. Kim, C. H., E. J. Kunkel, J. Boisvert, B. Johnston, J. J. Campbell, 55. Campbell, J. J., E. P. Bowman, K. Murphy, K. R. Youngman, M. A. Siani, M. C. Genovese, H. B. Greenberg, and E. C. Butcher. 2001. Bonzo/CXCR6 D. A. Thompson, L. Wu, A. Zlotnik, and E. C. Butcher. 1998. 6-C-kine (SLC), expression defines type 1-polarized T-cell subsets with extralymphoid tissue a lymphocyte adhesion-triggering chemokine expressed by high endothelium, is homing potential. J. Clin. Invest. 107: 595–601. an agonist for the MIP-3␤ receptor CCR7. J. Cell Biol. 141: 1053–1059. 48. Sato, T., H. Thorlacius, B. Johnston, T. L. Staton, W. Xiang, D. R. Littman, and 56. McNab, F. W., D. G. Pellicci, K. Field, G. Besra, M. J. Smyth, D. I. Godfrey, and E. C. Butcher. 2005. Role for CXCR6 in recruitment of activated CD8ϩ lym- S. P. Berzins. 2007. Peripheral NK1.1 NKT cells are mature and functionally phocytes to inflamed liver. J. Immunol. 174: 277–283. distinct from their thymic counterparts. J. Immunol. 179: 6630–6637. 49. Fukumoto, N., T. Shimaoka, H. Fujimura, S. Sakoda, M. Tanaka, T. Kita, and 57. Stetson, D. B., M. Mohrs, R. L. Reinhardt, J. L. Baron, Z. E. Wang, L. Gapin, S. Yonehara. 2004. Critical roles of CXC chemokine ligand 16/scavenger recep- M. Kronenberg, and R. M. Locksley. 2003. Constitutive cytokine mRNAs mark tor that binds phosphatidylserine and oxidized lipoprotein in the pathogenesis of natural killer (NK) and NK T cells poised for rapid effector function. J. Exp. Med. both acute and adoptive transfer experimental autoimmune encephalomyelitis. 198: 1069–1076. J. Immunol. 173: 1620–1627. 58. Meyer, E. H., M. A. Wurbel, T. L. Staton, M. Pichavant, M. J. Kan, P. B. Savage, R. H. DeKruyff, E. C. Butcher, J. J. Campbell, and D. T. Umetsu. 2007. iNKT 50. Agostini, C., A. Cabrelle, F. Calabrese, M. Bortoli, E. Scquizzato, S. Carraro, cells require CCR4 to localize to the airways and to induce airway hyperreac- M. Miorin, B. Beghe, L. Trentin, R. Zambello, et al. 2005. Role for CXCR6 and tivity. J. Immunol. 179: 4661–4671. its ligand CXCL16 in the pathogenesis of T-cell alveolitis in sarcoidosis. 59. Shimaoka, T., K. Seino, N. Kume, M. Minami, C. Nishime, M. Suematsu, Am. J. Respir. Crit. Care Med. 172: 1290–1298. T. Kita, M. Taniguchi, K. Matsushima, and S. Yonehara. 2007. Critical role for 51. Nanki, T., T. Shimaoka, K. Hayashida, K. Taniguchi, S. Yonehara, and CXC chemokine ligand 16 (SR-PSOX) in Th1 response mediated by NKT cells. N. Miyasaka. 2005. Pathogenic role of the CXCL16-CXCR6 pathway in rheu- J. Immunol. 179: 8172–8179. matoid arthritis. Arthritis Rheum. 52: 3004–3014. 60. Crowe, N. Y., J. M. Coquet, S. P. Berzins, K. Kyparissoudis, R. Keating, 52. Unutmaz, D., W. Xiang, M. J. Sunshine, J. Campbell, E. Butcher, and D. G. Pellicci, Y. Hayakawa, D. I. Godfrey, and M. J. Smyth. 2005. Differential D. R. Littman. 2000. The primate lentiviral receptor Bonzo/STRL33 is coordi- antitumor immunity mediated by NKT cell subsets in vivo. J. Exp. Med. 202: nately regulated with CCR5 and its expression pattern is conserved between 1279–1288. human and mouse. J. Immunol. 165: 3284–3292. 61. Trobonjaca, Z., F. Leithauser, P. Moller, R. Schirmbeck, and J. Reimann. 2001. Downloaded from 53. Wilbanks, A., S. C. Zondlo, K. Murphy, S. Mak, D. Soler, P. Langdon, Activating immunity in the liver. I. Liver dendritic cells (but not hepatocytes) are D. P. Andrew, L. Wu, and M. Briskin. 2001. Expression cloning of the STRL33/ potent activators of IFN-␥ release by liver NKT cells. J. Immunol. 167: BONZO/TYMSTR ligand reveals elements of CC, CXC, and CX3C chemokines. 1413–1422. J. Immunol. 166: 5145–5154. 62. Friedman, R. S., J. Jacobelli, and M. F. Krummel. 2006. Surface-bound chemo- 54. Lazarus, N. H., E. J. Kunkel, B. Johnston, E. Wilson, K. R. Youngman, and kines capture and prime T cells for synapse formation. Nat. Immunol. 7: E. C. Butcher. 2003. A common mucosal chemokine (mucosae-associated epi- 1101–1108. http://www.jimmunol.org/ by guest on September 27, 2021