CXCL11) This Information Is Current As of October 1, 2021

CXCR3 Internalization Following T Cell-Endothelial Cell Contact: Preferential Role of IFN-Inducible T Cell α Chemoattractant (CXCL11) This information is current as of October 1, 2021. Alain Sauty, Richard A. Colvin, Ludwig Wagner, Sophie Rochat, Francois Spertini and Andrew D. Luster J Immunol 2001; 167:7084-7093; ; doi: 10.4049/jimmunol.167.12.7084 http://www.jimmunol.org/content/167/12/7084 Downloaded from

References This article cites 37 articles, 19 of which you can access for free at: http://www.jimmunol.org/content/167/12/7084.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 October 1, 2021 *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 © 2001 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. CXCR3 Internalization Following T Cell-Endothelial Cell Contact: Preferential Role of IFN-Inducible T Cell ␣ Chemoattractant (CXCL11)1

Alain Sauty,*† Richard A. Colvin,* Ludwig Wagner,* Sophie Rochat,† Francois Spertini,† and Andrew D. Luster2*

Chemokine receptors are rapidly desensitized and internalized following ligand binding, a process that attenuates receptor- mediated responses. However, the physiological settings in which this process occurs are not clear. Therefore, we examined the fate of CXCR3, a chemokine receptor preferentially expressed on activated T cells following contact with endothelial cells. By immunofluorescence microscopy and flow cytometry, we found that CXCR3 was rapidly internalized when T cells were incubated with IFN-␥-activated human saphenous vein endothelial cells (HSVEC), but not with resting HSVEC. Similar results were ob- Downloaded from tained using human CXCR3-transfected murine 300-19 B cells. CXCR3 down-regulation was significantly more pronounced when T cells were in contact with HSVEC than with their supernatants, suggesting that CXCR3 ligands were efficiently displayed on the surface of HSVEC. Using neutralizing mAbs to IFN-induced protein-10 (CXCL10), monokine induced by IFN-␥ (CXCL9), and IFN-inducible T cell ␣ chemoattractant (I-TAC; CXCL11), we found that even though I-TAC was secreted from IFN-␥-activated HSVEC to lower levels than IFN-induced protein-10 or the monokine induced by IFN-␥, it was the principal chemokine responsible for CXCR3 internalization. This correlated with studies using recombinant chemokines, which revealed that I-TAC was the http://www.jimmunol.org/ most potent inducer of CXCR3 down-regulation and of transendothelial migration. Known inhibitors of chemokine-induced chemotaxis, such as pertussis toxin or wortmannin, did not reduce ligand-induced internalization, suggesting that a distinct signal transduction pathway mediates internalization. Our data demonstrate that I-TAC is the physiological inducer of CXCR3 internalization and suggest that chemokine receptor internalization occurs in physiological settings, such as leukocyte contact with an activated endothelium. The Journal of Immunology, 2001, 167: 7084–7093.

ecruitment and activation of T lymphocytes is a prere- unique chemokine receptor, CXCR3. CXCR3 is expressed on ac- quisite step for the development of an adaptive immune tivated T cells, preferentially of the Th1 phenotype, NK cells, and ϩ ϩ response and maintenance of chronic inflammation. Che- on a significant fraction of circulating CD4 and CD8 T cells by guest on October 1, 2021 R ϩ mokines, a superfamily of small molecular mass (8Ð10 kDa) che- (ϳ20Ð40%; Ref. 6Ð8). The majority of peripheral CXCR3 T ␤ moattractant cytokines, play a crucial role in the transendothelial cells express CD45RO (memory T cells) as well as 1 integrins (8) and interstitial migration of lymphocytes during inflammation. which are implicated in the binding of lymphocytes to endothelial Chemokines are divided into the CC, CXC, C, and CX3C families cells and the extracellular matrix (9). In addition, CXCR3 has been (1). Among the CXC members, IFN-induced protein of 10 kDa reported to be expressed on plasmacytoid dendritic cells (10), leu- 3 (IP-10) (CXCL10), monokine induced by IFN-␥ (Mig; CXCL9), kemic B cells (11, 12), eosinophils (13), and dividing microvas- and IFN-inducible T cell ␣ chemoattractant (I-TAC; CXCL11) are cular endothelial cells (14). ␥ ϩ unique in that they are all induced by IFN- in a wide variety of CXCR3 T cells accumulate at sites of Th1-type inflammation cell types, including endothelial cells (2Ð5), and act through a where IFN-␥ is highly expressed, including atherosclerosis (5), sarcoidosis (15), inflammatory bowel diseases (8), and rheumatoid *Division of Rheumatology, Allergy, and Immunology, Center for Immunology and arthritis (8, 16). IP-10 has been found to be highly expressed in a Inflammatory Diseases, Massachusetts General Hospital, Harvard Medical School, number of Th1-type inflammatory diseases, including psoriasis Charlestown, MA 02129; and †Immunology and Allergy Division, University Hos- pital, Lausanne, Switzerland. (17), tuberculoid leprosy (18), sarcoidosis (15), and viral menin- ␥ Received for publication December 20, 2000. Accepted for publication October gitis (19). In addition, our finding that IFN- -stimulated endothe- 4, 2001. lial cells and endothelium from atherosclerotic lesions are a rich The costs of publication of this article were defrayed in part by the payment of page source of IP-10, Mig, and I-TAC suggest that these chemokines charges. This article must therefore be hereby marked advertisement in accordance play an important role in the transendothelial migration and local with 18 U.S.C. Section 1734 solely to indicate this fact. retention of CXCR3ϩ T cells found in atherosclerotic lesions (5). 1 These studies were funded by National Institutes of Health Grants CA69212 and DK50305 (to A.D.L.). A.S. was supported by a grant of Muschamp Foundation. In support of this hypothesis, IP-10 and Mig induce the rapid ad- 2 Address correspondence and reprint requests to Dr. Andrew D. Luster, Center for hesion of IL-2-activated T cells to immobilized VCAM-1 and Immunology and Inflammatory Diseases, Massachusetts General Hospital, Building ICAM-1, and IP-10, Mig, and I-TAC are potent chemotactic 149, 13th Street, Charlestown, MA 02129. E-mail address: [email protected]. agents for activated T cells. Both of these activities are dependent harvard.edu on CXCR3 activation (20). 3 Abbreviations used in this paper: IP-10, IFN-induced protein of 10 kDa; Mig, monokine induced by IFN-␥; I-TAC, IFN-inducible T cell ␣ chemoattractant; PTX, per- Chemokine receptors are seven transmembrane-spanning G pro- tussis toxin; PI3K, phosphoinositol 3-kinase; PKC, protein kinase C; GAM, goat tein-coupled cell surface receptors that are pertussis toxin (PTX)- anti-mouse; HMEC, human microvascular endothelial cells; HSVEC, human saphenous vein endothelial cells; MFI, mean fluorescence intensity; HSPG, heparan sulfate sensitive, indicating that they are linked to the Gi class of hetero- proteoglycan; SDF, stromal cell-derived factor. trimeric G proteins (21). Activation of chemokine receptors

Copyright © 2001 by The American Association of Immunologists 0022-1767/01/$02.00 The Journal of Immunology 7085 induces G protein-dependent inhibition of adenylate cyclase, acti- Coculture experiments vation of phosphoinositol 3-kinase (PI3K), phospholipase C, pro- Human saphenous vein endothelial cells (HSVEC) were used from passage tein kinase C (PKC), and protein kinase A, the generation of ino- two to passage four and were cultured in 0.1% gelatin-precoated wells from sitol triphosphate, and a transient rise in intracellular calcium (22). 6-well plates (Costar, Cambridge, MA). When cells reached conﬂuence, Overall, the biological responses resulting from these different cas- they were treated with or without 100 ng/ml IFN-␥ for 16 h in fresh complete M199 medium. Following removal of the supernatant and two washes cades include rapid activation of integrin-dependent cell adhesion ϩ with prewarmed HBSS, CXCR3 T cells (1 ϫ 106) in 1 ml of complete and directed cell migration (chemotaxis). RPMI were then added per well. HSVEC-conditioned medium (250 ␮l) As is the case in many biological systems, chemokine receptor was used to resuspend a separate aliquot of 0.5 ϫ 106 CXCR3ϩ T cells activation also turns on a program to limit its own responses. In the pelleted in an Eppendorf. After various times, CXCR3ϩ cells were col- chemokine system, this takes the form of receptor desensitization lected by being gently washed off of the endothelial monolayer in 6-well plates or by centrifugation of CXCR3ϩ cells incubated with HSVEC su- and receptor internalization. These two processes likely play an pernatants. Cells were analyzed for CXCR3 surface expression by ﬂow important role in allowing a cell to continuously sense small cytometry. In control experiments, we found that IFN-␥ treatment of T changes in an existing, and perhaps even changing, concentration cells had no direct effect on CXCR3 expression. gradient. Receptor desensitization is a very rapid response, ren- Measurement of CXCR3 surface expression dering a cell transiently unresponsive to a subsequent stimulation ϩ through that receptor, and usually results from uncoupling the re- CXCR3 cells (0.25Ð0.5 ϫ 106) were washed into an ice-cold FACS 2ϩ 2ϩ ceptor from heterotrimeric G proteins (23). Receptor internaliza- buffer (HBSS without Ca and Mg supplemented with 1% BSA, 2% FCS, and 0.1% sodium azide) and resuspended in 90 ␮l of FACS buffer tion is a distinct process where activated receptors are removed and 10 ␮l of FITC-conjugated anti-CXCR3 mAb, PE-conjugated anti- from the cell surface for degradation or recycling and leads to a

CXCR4, or isotype controls. Cells were incubated for 30 min on ice in the Downloaded from more prolonged state of cellular unresponsiveness to the internal- dark. After two washings with the ice-cold FACS buffer, cells were resus- ized receptors’ agonists (24Ð26). The role of receptor internaliza- pended in FACS buffer with propidium iodine (0.3 ␮g/ml, and only when CXCR3 was studied alone) to gate on live cells. FACS analysis was per- tion in the physiological responses induced by chemokines, such as formed on 104 cells using a FACSCaliber flow cytometer (BD Biosciences, transendothelial migration and chemotaxis, is unclear. To begin to San Jose, CA) using CellQuest software (San Jose, CA). Mean fluores- address these questions, we studied the fate of CXCR3 following cence intensity (MFI) values were obtained by subtracting the MFI of the contact between CXCR3ϩ T cells with a resting or IFN-␥-acti- isotype control from the MFI of the positively stained sample. CXCR3 vated endothelium. surface expression was expressed as a percent of baseline expression using http://www.jimmunol.org/ the formula: (MFI of tested cells/MFI of untreated cells) ϫ 100. Materials and Methods Regulation of CXCR3 cell surface expression Reagents ϩ ϩ CXCR3 T cells or CXCR3 300-19 cells (0.25 ϫ 106) were incubated Recombinant human IP-10, I-TAC, Mig, IFN-␥, and IL-2 were obtained with various concentrations of rIP-10, I-TAC, or Mig for different times as from PeproTech (Rocky Hill, NJ). Neutralizing murine mAb to human indicated. In some experiments, T cells were preincubated for 2 h with 1 IP-10 and Mig were purchased from R&D Systems (Minneapolis, MN). ␮M of wortmannin, 100 ␮M of genestein, 10 nM of staurosporin, or Neutralizing murine mAb to human I-TAC was a kind gift from K. Neote 0Ð1000 ng/ml PTX at 37¡C, 5% CO2 before the addition of I-TAC, IP-10, (Pfizer, Groton, CT). FITC-conjugated and nonlabeled mouse anti-human or Mig (250 ng/ml), or PMA (0.1 nM) for an additional 30 min. T cells

CXCR3 mAb (49801.111; IgG1) was purchased from R&D Systems. PE- were also incubated with different concentrations of PMA and calcium by guest on October 1, 2021 conjugated mouse anti-CXCR4 mAb (12G5, IgG2a), irrelevant-purified ionophore for 30 min. At the end of these experiments, ice-cold FACS and FITC-conjugated IgG1, and PE-conjugated IgG2a mouse Abs (isotype buffer was added, then cells were studied for cell surface expression of controls) were obtained from BD PharMingen (San Diego, CA). Texas CXCR3. Red-conjugated goat anti-mouse (GAM) Ab was obtained from Southern Biotechnology Associates (Birmingham, AL). Bispecific anti-CD3/CD8 Indirect immunofluorescence was a kind gift from J. Wong (Massachusetts General Hospital, Boston, CXCR3ϩ T cells or 300-19 cells (107 cells/ml) were incubated with anti- MA; Ref. 27). PTX and calcium ionophore were purchased from Calbio- CXCR3 (10 ␮g/ml in PBS) for 30 min on ice. Cells were then washed chem (La Jolla, CA). Genestein, wortmannin, PMA, staurosporin, and pro- twice in ice-cold PBS and then incubated for 30 min with 500 ng/ml I-TAC pidium iodine were obtained from Sigma-Aldrich (St. Louis, MO). T cell diluted in complete RPMI prewarmed at 37¡C. Alternatively, they were enrichment columns were purchased from R&D Systems. exposed for 30 min to a monolayer of HSVEC treated or not with IFN-␥ Cell isolation and culture for 16 h and washed twice with prewarmed PBS. Cells were then resuspended in the ice-cold FACS buffer, centrifugated, and immediately fixed Human vascular endothelial cells were isolated from saphenous veins from in 3.7% paraformaldehyde for 20 min on ice. After permeabilization with seven different donors by collagenase treatment as described elsewhere (5). 0.1% saponin in FACS buffer, cells were incubated with Texas Red-con- Cells were maintained in M199 medium (BioWhittaker, Walkersville, MD) jugated GAM diluted in 0.1% saponin buffer for 30 min on ice and in supplemented with 100 U/ml penicillin, 100 ␮g/ml streptomycin (Bio- complete darkness. CXCR3 expression was assessed by immunofluores- Whittaker), 5% FCS (Atlanta Biologicals, Norcross, GA), 100 ␮g/ml hep- cence microscopy (Axiophot microscope; Zeiss, Oberkochen, Germany). arin (Sigma-Aldrich), and 50 mg/ml endothelial cell growth factor (Pel- Freez Biologicals, Rogers, AR). Human microvascular endothelial Role of heparin (HMEC)-1 cells, a human microvascular cell line, were cultured in com- Confluent HSVEC were incubated with complete M199 supplemented with plete endothelial growth medium (Clonetics, Walkersville, MD) supple- 0Ð500 ␮g/ml heparin with or without 100 ng/ml IFN-␥ for 16 h. After mented with 5% FCS, gentamicin, amphotericin, bovine brain extract (12 removal of the supernatant, cells were washed twice with prewarmed ␮g/ml) and hydrocortisone (1 ␮g/ml), and epidermal growth factor (10 HBSS (37¡C). HSVEC and their supernatants were examined for their ng/ml). ϩ ability to induce CXCR3 internalization on T cells as mentioned above. In CXCR3 T cells were obtained from different sources. In some exper- another set of experiments, 100 ng/ml IP-10 or I-TAC in either complete iments, PBMC isolated by a Ficoll gradient were incubated with 200 ng/ml RPMI, complete M199 (100 ␮g/ml heparin), or complete M199 without IL-2 and bispecific anti-CD3/CD8 or anti-CD3/CD4 for 4 days in RPMI ϩ heparin were used to resuspend CXCR3 T cells. Following incubation for 1640 (Cellgro and Mediatech, Herndon, VA) supplemented with 10% FCS, 2hat37¡C, 5% CO , T cells were stained for CXCR3 surface expression 100 U/ml penicillin, 100 ␮g/ml streptomycin (Mediatech), and 2 mM of 2 ϩ ϩ ϩ and analyzed by FACS. L-glutamine (Mediatech). Proliferating CXCR3 CD4 or CD8 T cells were kept in IL-2 only for an additional 4Ð10 days. In other experiments, Effect of exogenous CXCR3 ligands added on HSVEC CD3ϩ T cells were derived from PBMC passed through a T cell enrichment column (R&D Systems) and incubated for 4 days with 5 ␮g/ml PHA Confluent HSVEC were incubated with complete M199 (including 100 (Sigma-Aldrich) and 200 ng/ml IL-2 and then in IL-2 only for 4Ð10 days. ␮g/ml heparin) with or without 250 ␮g/ml I-TAC, IP-10, or Mig, or the ϩ Human CXCR3 -transfected 300-19 cells were a gift from B. Moser (The- three of them for1hat37¡C, 5% CO2. After removal of the supernatant, odor Kocher Institute, Bern, Switzerland; Ref. 6). cells were washed twice with prewarmed HBSS (37¡C) to remove unbound 7086 CXCR3 INTERNALIZATION AFTER T CELL-ENDOTHELIAL CELL CONTACT chemokines. CXCR3ϩ T cells were then added onto HSVEC monolayers pore size: 5 ␮m) of a 96-well ChemoTx plate (NeuroProbe, Gaithersburg, or resuspended into HSVEC supernatants and incubated for 30 min before MD). After 96 h, cell adherence and confluence were confirmed by staining CXCR3 surface expression was analyzed by FACS. with Diff-Quick and acetyl/low density lipoprotein-DiI (Biomedical Tech- nologies, Stoughton, MA), 10 ␮g/ml in DMEM 1% BSA. Indicated con- CXCR3 recycling centrations of chemokines were added in the bottom chamber (in 31 ␮lof ϫ 4 ϩ ϫ 6 RPMI/1% BSA). After gentle washing of the HMEC-1 cells, 2.5 10 of CXCR3 T cells (0.75 10 ) were treated with I-TAC (500 ng/ml) in human CXCR3ϩ 300-19 cells in 25 ␮l were added on top of HMEC-1 cells complete RPMI for 30 min (37¡C, 5% CO ). Following two washes with 2 and incubated for5hat37¡C, 5% CO . The filter was then removed and prewarmed (37¡C) HBSS to remove unbound chemokines, they were re- 2 ␮ the migrated cells were counted. Each experiment was performed in suspended in 750 l of complete RPMI for up to 3 h. CXCR3 surface triplicate. expression was determined at baseline in untreated cells (Ϫ30 min) and at times 0, 1, and 3 h following I-TAC stimulation. Data analysis Neutralizing HSVEC chemokine activity Experiments were performed using at least five different donors for HS- VEC and CXCR3ϩ T cells. Results are shown as the mean of at least three When tested for their ability to inhibit CXCR3 internalization induced by Ϯ recombinant chemokines, all neutralizing mAb were effective in a dose- independent experiments SE. Statistical analysis was calculated using a paired Student’s t test and statistical significance was considered with dependent manner. Fifty micrograms per milliliter neutralizing mAb sig- Ͻ nificantly reduced CXCR3 surface disappearance induced by 250 ng/ml p 0.05. Mig (43.4 vs 18%), IP-10 (47 vs 18%), and I-TAC (65 vs 19%). IFN-␥- treated HSVEC supernatants were pretreated with 50 ␮g/ml neutralizing Results anti-Mig, anti-IP-10, or anti-I-TAC mAb or control mAb for1hat37¡C. CXCR3 down-regulation induced by IFN-␥-activated HSVEC CXCR3ϩ T cells were then incubated with these supernatants for 30 min and CXCR3 surface expression examined as described above. The ability

Initial experiments showed that overnight coincubation of freshly Downloaded from of neutralizing mAb to inhibit the activity of HSVEC monolayers was also isolated CD3ϩ T cells with IFN-␥-stimulated, but not resting, HS- studied. After removal of their supernatant and two washes, HSVEC mono- VEC led to the disappearance of CXCR3 surface expression on T layers (in 24-well plates) were treated with 50 ␮g/ml neutralizing mAb or ϩ ϩ ϩ control mAb for1hat37¡C before the addition of CXCR3ϩ T cells. cells (Fig. 1A). We found this for both CD8 and CD4 CXCR3 T cells (data not shown). Time-course experiments showed that Transendothelial migration CXCR3 surface expression decreased from 100 to 30Ð35% within HMEC-1 cells were trypsinized and added (5 ϫ 105 in 50 ␮l of complete the first 15 min of coincubation (Fig. 1B). To assess whether this http://www.jimmunol.org/ endothelial cell growth medium) on the top of the filter (polycarbonate, process was specific for CXCR3, we studied CXCR4 expression by guest on October 1, 2021

FIGURE 1. Cell surface expression of CXCR3 on T cells cocultured with HSVEC. A, Freshly isolated peripheral blood CD3ϩ T cells (1 ϫ 106) were incubated with resting or IFN-␥-activated (100 ng/ml for 24 h) confluent HSVEC. After 16 h, T cells were removed from the HSVEC monolayers and analyzed by FACS for CXCR3 cell surface expression. Percentages in the upper right squares represent the percentage of CXCR3ϩ cells (37.4 vs 2%). B, Kinetics of CXCR3 and CXCR4 expression on T cells cocultured with HSVEC. CXCR3ϩ T cells were incubated with resting or IFN-␥-stimulated HSVEC (100 ng/ml overnight in complete medium). At indicated times, cell surface expression of CXCR3 and CXCR4 was assessed by FACS analysis. Data represent the mean Ϯ SE of three independent experiments. C, Efficacy of HSVEC monolayers and HSVEC supernatants at inducing CXCR3 internalization on T cells. CXCR3ϩ T cells (1 ϫ 106) were cocultured with resting or IFN-␥-treated HSVEC (100 ng/ml, 16 h). Alternatively, 250 ␮l of supernatant ϩ ϫ 6 collected from these same HSVEC cultures were used to resuspend CXCR3 T cells (0.25 10 ). After2hat37¡C and 5% CO2, T cells were removed and analyzed for CXCR3 surface expression by FACS. Compared with respective controls, CXCR3 internalization was greater with HSVEC monolayers (72.3 Ϯ 4.6%) compared with their supernatants (48.3 Ϯ 12.8%), p Ͻ 0.015. Data represent the mean Ϯ SE of four independent experiments. The Journal of Immunology 7087 during these kinetic experiments and found no significant decrease medium with that of direct cell contact at inducing CXCR3 in CXCR4 expression (Fig. 1B). Thus, IFN-␥ activated, but not down-regulation. Coincubation of CXCR3ϩ T cells with IFN-␥- resting, HSVEC quickly and selectively down-regulates CXCR3 activated HSVEC for 2 h induced an 80% decrease in CXCR3 expression on T cells. expression compared with a 25% decrease following coincubation with unstimulated HSVEC controls (Fig. 1C). In contrast, super- Cell contact is more efficient at inducing CXCR3 down- natants collected from these same IFN-␥-activated HSVEC de- regulation than is conditioned medium creased CXCR3 expression on T cells by only 50%, while super- The most likely explanation for selective chemokine receptor in- natants collected from unstimulated HSVEC cultures had no effect ternalization is selective chemokine receptor activation (26). IFN- on CXCR3 expression. Thus, IFN-␥-activated and control HSVEC ␥-activated endothelial cells secrete the three known CXCR3 li- were significantly more potent at inducing CXCR3 down- gands: IP-10, Mig, and I-TAC (5, 28). Following overnight regulation than their respective conditioned media ( p Ͻ 0.015 and culture, we have found ϳ100 ng/ml IP-10, 10 ng/ml Mig, and 1 p Ͻ 0.03, respectively). ng/ml I-TAC in the conditioned medium of IFN-␥ activated HSVEC (28). Although chemokines are secreted from cells, they CXCR3 down-regulation induced by CXCR3 ligands: IP-10, are basic proteins that under physiological conditions have an Mig, and I-TAC affinity for cell surface negatively charged heparan sulfate proteo- Because we recently showed that IFN-␥-stimulated HSVEC pro- glycans (HSPG) and accumulate on the surface of endothelial cells duced IP-10, Mig, and I-TAC, the three known ligands for CXCR3 (29). Therefore, we compared the ability of HSVEC-conditioned (5, 28), we sought to determine whether these chemokines induced Downloaded from http://www.jimmunol.org/ by guest on October 1, 2021

FIGURE 2. Time- and dose-dependent internalization of CXCR3 on T cells and CXCR3ϩ 300-19 cells. A, CXCR3ϩ T cells and CXCR3-transfected 300-19 cells (2 ϫ 106 cells/ml) were incubated in complete RPMI supplemented with 500 ng/ml IP-10, Mig, or I-TAC for the indicated times. B, top panels, CXCR3ϩ T cells and CXCR3ϩ 300-19 cells (2 ϫ 106 cells/ml) were treated with the indicated concentrations of chemokine for 30 min. Propidium iodine (PI) was used to gate on viable cells. Bottom panels, Representative FACS data obtained with T cells treated with varying concentrations of I-TAC (0Ð1000 ng/ml). Values are the mean Ϯ SE of three to four independent experiments. 7088 CXCR3 INTERNALIZATION AFTER T CELL-ENDOTHELIAL CELL CONTACT

CXCR3 internalization on T cell surfaces. CXCR3ϩ T cells and much more potent than IP-10 or Mig, and IP-10 was more potent CXCR3ϩ 300-19 cells were treated with 500 ng/ml IP-10, Mig, or than Mig. Representative results from FACS analysis performed I-TAC for various times and then analyzed by FACS for CXCR3 on T cells treated with different doses of I-TAC are shown in the surface expression (Fig. 2A). We found that maximum receptor bottom panel of Fig. 2B. A similar attenuated pattern was also internalization occurred within the first 15 min with all three li- found with CXCR3ϩ 300-19 cells (Fig. 2B). These results dem- gands, as shown for HSVEC-mediated internalization. For both onstrate that IP-10, Mig, and I-TAC have differential abilities to cell types, I-TAC was more potent than IP-10 or Mig, but this down-regulate CXCR3 internalization on T cells, and that among difference was much greater for T cells compared with CXCR3- these three IFN-␥-inducible CXCR3 agonists, I-TAC is the most transfected 300-19 B cell line. Indeed, CXCR3 expression on T potent inducer of CXCR3 down-regulation. cells diminished by ϳ75Ð85% with I-TAC, by ϳ35Ð40% with IP-10, and by ϳ30% with Mig. In addition, we observed that in Internalization of CXCR3 following ligand binding presence of the ligand, there was no significant CXCR3 reappear- To study CXCR3 internalization, cells were treated with an anti- ance on T cells for up to 120 min. In contrast, CXCR3 reappeared CXCR3 mAb and the fate of CXCR3 was followed after exposure on 300-19 cells by 120 min despite the continuous presence of the to HSVEC cells and rI-TAC. Immunofluorescence microscopy us- ligands (Fig. 2A). The differences in the magnitude and kinetics of ing a Texas Red-conjugated GAM revealed CXCR3 at the cell CXCR3 internalization and recycling in T cells compared with surface and in a perinuclear vesicular distribution in untreated transfected 300-19 B cells is not clear. However, we found that CXCR3ϩ T cells and 300-19 cells (Fig. 3). Following I-TAC ex- simply washing CXCR3ϩ 300-19 cells for flow cytometric anal- posure as well as exposure to IFN-␥-treated HSVEC, an increase ysis or into fresh medium increased CXCR3 expression ϳ1.5-fold in CXCR3 internalization occurred, discernible as a decrease of Downloaded from and ϳ2.5-fold above baseline, respectively. This effect was not cell surface expression and an increase in intracellular accumula- affected by cycloheximide, indicating that these treatments in- tion. In contrast, the cellular distribution of CD4 was unaffected by duced the mobilization of an intracellular pool of receptors in these CXCR3 ligand exposure, demonstrating the specificity of I-TAC- transfected cells (data not shown). induced CXCR3 internalization (data not shown). Dose-response studies were also performed using CXCR3ϩ T cells and CXCR3ϩ 300-19 cells which were treated with 0, 10, CXCR3 recycling following agonist exposure 100, and 1000 ng/ml ligand for 30 min. At Յ10 ng/ml, only a To determine whether CXCR3 is recycled back to the cell surface http://www.jimmunol.org/ minor decrease in CXCR3 expression was seen for all three li- after agonist-induced internalization, CXCR3ϩ T cells were gands on T cells (Fig. 2B). At higher concentrations, I-TAC was treated for 30 min with I-TAC (250 ng/ml) and then washed into by guest on October 1, 2021

FIGURE 3. IFN-␥-treated HSVEC- and I-TAC-induced internalization of CXCR3 on T cells and CXCR3ϩ 300-19 cells. CXCR3ϩ T cells and CXCR3ϩ 300-19 cells were ﬁrst incubated on ice with 10 ␮g/ml anti-CXCR3 mAb and then exposed to resting or IFN-␥-treated HSVEC or to I-TAC (500 ng/ml) for 30 min. Cells where then ﬁxed, permeabilized in 0.1% saponin, and incubated with GAM-Texas Red. In control cells, CXCR3 was found on the cell surface and accumulated into perinuclear vesicles (spontaneous internalization). After exposition to IFN-␥-treated HSVEC and I-TAC, CXCR3 surface expression strongly decreased and increased into perinuclear vesicles. Bottom panels, Anti-CXCR3 mAb was omitted and cells were only incubated with GAM-Texas Red. The Journal of Immunology 7089

Regulation of CXCR3 internalization To begin to dissect the mechanisms responsible for chemokine- induced CXCR3 internalization, we investigated the role of different intracellular signaling pathways using selective inhibitors and activators. Activation of chemokine receptors is known to activate multiple intracellular signaling pathways, including heterotrimeric G proteins, PI3K, tyrosine kinases, and PKC. To investi- gate the role of PI3K and tyrosine kinases, T cells were pretreated for 2 h with wortmannin to inhibit PI3K or genestein to inhibit tyrosine kinases. The expression of CXCR3 was then determined by FACS analysis following 30 min incubation with 250 ng/ml I-TAC or IP-10 (Fig. 5A). Neither wortmannin nor genestein inhibited the ability of I-TAC or IP-10 to induce CXCR3 internal- ϩ ϫ 6 FIGURE 4. CXCR3 recycling on T cells. CXCR3 T cells (2 10 ization, suggesting that PI3K and tyrosine kinases are not mediat- cells/ml) were incubated with or without I-TAC (250 ng/ml) in complete ing the internalization signal. RPMI for 30 min. After two washes with HBSS prewarmed at 37¡Cto remove unbound chemokines, cells were resuspended in fresh prewarmed In addition, because heterotrimeric G proteins of the Gi subclass complete RPMI. CXCR3 expression was determined at baseline and at have been shown to mediate CXCR3-induced chemotaxis and times 0, 1, and 3 h after chemokine removal by washing. Data represent the calcium ﬂux responses, we studied the effects of PTX, a Ϯ Downloaded from mean SE of four independent experiments. selective G␣1 inhibitor, on I-TAC-induced CXCR3 internalization. T cells were therefore pretreated with PTX (250, 500, or 1000 ng/ml) for2hat37¡C and then stimulated with I-TAC (250 ng/ml; Fig. 5B). PTX pretreatment at concentrations that inhibited fresh medium without I-TAC (Fig. 4). CXCR3 surface expression I-TAC- and IP-10-induced chemotaxis and calcium ﬂux did not was assessed at times 0, 1, and 3 h. Following agonist stimulation, inhibit I-TAC- or IP-10-induced CXCR3 internalization. This sug- receptor recycling occurred rapidly and by 3 h, CXCR3 cell sur- gests that agonist-induced CXCR3 internalization is not dependent http://www.jimmunol.org/ Ϯ face expression reached 79 4% of its prestimulation levels. on G␣1 activation. by guest on October 1, 2021

FIGURE 5. Regulation of chemokine-induced internalization of CXCR3 in T cells. CXCR3ϩ T cells (2 ϫ 106 cells/ml) were pretreated for2hat37¡C with (A)1␮M wortmannin or 100 ␮M genestein or (B) PTX (250Ð1000 ng/ml). I-TAC or IP-10 (250 ng/ml) was then added for 30 min at which point CXCR3 surface expression was assessed by FACS analysis. Data represent mean Ϯ SE of three independent experiments. C, CXCR3ϩ T cells (2 ϫ 106 cells/ml) were stimulated with Ca ionophore or PMA at indicated concentrations, then examined for CXCR3 surface expression by FACS analysis. Results are expressed as the mean Ϯ SE of three independent experiments. D, CXCR3ϩ T cells (2 ϫ 106 cells/ml) were pretreated with 10 nM of staurosporin for 2 h prior to PMA (0.1 nM) or IP-10, Mig, or I-TAC (250 ng/ml) were added. CXCR3 surface expression was then determined. Results are expressed as the mean Ϯ SE of four independent experiments. 7090 CXCR3 INTERNALIZATION AFTER T CELL-ENDOTHELIAL CELL CONTACT

Because other chemokine receptors, such as CXCR4 (30) and CCR3 (31) have been shown to be internalized following activation of PKC, we determined whether direct activation of PKC by PMA or calcium ionophore (A12387) would induce CXCR3 internalization. We found that both PKC activators induced a dose- dependent CXCR3 down-regulation (Fig. 5C), suggesting that global activation of PKC can induce CXCR3 internalization in T cells and might be involved in agonist-induced CXCR3 internalization. Therefore, we pretreated CXCR3ϩ T cells for 2 h with 10 ␮M of staurosporin, a PKC inhibitor, prior to exposure to 250 ng/ml I-TAC, Mig, or IP-10 (Fig. 5D). Although staurosporin completely inhibited PMA (nonagonist)-induced CXCR3 internalization, it did not affect the effect of IP-10 and Mig and very mildly affected the effect of I-TAC. Therefore, PKC activation does not seem to be involved in agonist-induced CXCR3 internalization.

Effect of heparin on ligand-induced CXCR3 internalization Chemokines are basic proteins that bind negatively charged gly- Downloaded from cosaminoglycan molecules, such as heparin and HSPG, at physiological salt concentrations. This interaction can significantly modulate the activity of chemokines and influence chemokine binding to their receptors and subsequent signaling. In particular, IP-10 has a relatively high affinity for heparin (25 nM) and can bind to HSPG on endothelial cells. Heparin has been shown to displace an IP-10 alkaline-phosphatase fusion protein bound to http://www.jimmunol.org/ endothelial cells (29). Therefore, we sought to determine whether heparin could affect the down-regulation of CXCR3 induced by incubation of T cells with HSVEC monolayers or their supernatants. To study this, HSVEC were incubated overnight with or without IFN-␥ (100 ng/ml) in medium containing various concentrations of heparin (0Ð500 ␮g/ml). HSVEC were then washed to FIGURE 6. Modulation of CXCR3 internalization in T cells by heparin. remove the heparin, and CXCR3ϩ T cells were added to the A, Confluent HSVEC were cultured overnight with or without IFN-␥ (100 ng/ml) in complete M199 medium supplemented with 0Ð500 ␮g/ml hep-

HSVEC monolayers or resuspended in HSVEC supernatants for by guest on October 1, 2021 ϩ 30 min. We found that heparin, in a dose-dependent manner, was arin. CXCR3 T cells were resuspended in HSVEC supernatants or added onto HSVEC monolayers and incubated for 30 min at 37¡C, 5% CO . able to very modestly decrease CXCR3 internalization on T cells 2 Ϯ CXCR3 expression was then assessed by FACS analysis. Data represent Ͻ ء incubated with HSVEC monolayers or their supernatants (13 Ϯ Ϯ Ͻ Ϯ Ϯ mean SE of three independent experiments. , p 0.03 compared with 3% vs 24 3%, p 0.007, and 33 5% vs 47 10%, NS, for control without heparin. B, rI-TAC and IP-10 (100 ng/ml final) were di- monolayers and supernatants, respectively, at 500 ng/ml heparin) luted in either complete RPMI medium or complete M199 medium with or (Fig. 6A). These results suggest that heparin, at least at these con- without heparin (100 ␮g/ml). These media were then used to incubate ϩ centrations, only moderately displaced CXCR3 ligands from CXCR3 T cells for 30 min at 37¡C and 5% CO2. CXCR3 expression was HSVEC surfaces. then assessed by FACS analysis. Data represent the mean Ϯ SE of three p Ͻ 0.001 compared with ,ءء ;p Ͻ 0.003 ,ء .To further characterize the role of glycosaminoglycans, we pre- independent experiments incubated IP-10 or I-TAC (100 ng/ml) in medium with or without control. heparin (100 ␮g/ml). Interestingly, we found that heparin significantly decreased IP-10- or I-TAC-induced internalization of CXCR3 (Fig. 6B). This is in contrast to experiments using HSVEC supernatants, and suggests that IP-10 or I-TAC secreted from reflection of the amount of biologically active I-TAC bound to HSVEC or sequestered on HSVEC surfaces might be already HSVEC surfaces following IFN-␥ treatment. bound to HSPGs in such a way that does not inhibit their activity on CXCR3. I-TAC is the physiological inducer of CXCR3 internalization To examine whether exogenous chemokines would efficiently Because IP-10, Mig, and I-TAC are produced to varying levels bind to resting HSVEC and induce CXCR3 internalization on T from IFN-␥-activated HSVEC (IP-10 Ͼ Mig Ͼ I-TAC) and ex- cells, IP-10, Mig, and I-TAC were individually added (250 ng/ml) ogenously added recombinant ligands can, to varying degrees, in- to the medium of HSVEC for 1 h and then unbound chemokines duce CXCR3 internalization (I-TAC Ͼ IP-10 Ͼ Mig), we sought were removed by washing. We found that I-TAC, IP-10, and Mig to determine the contribution of these ligands to the ability of remained bound to HSVEC surfaces after washing because they IFN-␥-activated HSVEC monolayers and their supernatants to in- induced CXCR3 sequestration in T cells, with a potency order duce CXCR3 internalization on T cells. This was achieved using (I-TAC Ͼ IP-10 Ͼ Mig) identical to soluble recombinant proteins neutralizing Abs. Anti-IP-10 and anti-Mig mAbs (50 ␮g/ml), (Fig. 7). It is worth noting that the simultaneous addition of all which were able to completely inhibit IP-10- and Mig- (250 ng/ml) three ligands was not significantly different from I-TAC alone and induced CXCR3 down-regulation, had no effect on HSVEC mono- that I-TAC (250 ng/ml) was similar to the effect of IFN-␥-stimu- layer- or supernatant-induced CXCR3 down-regulation (Fig. 8). In lated HSVEC monolayers. These data suggest that the level of contrast, anti-I-TAC mAb (50 ␮g/ml), which was able to inhibit I-TAC found in HSVEC supernatants might not be an accurate I-TAC (250 ng/ml)-induced CXCR3 down-regulation, was able to The Journal of Immunology 7091

FIGURE 7. Exogenous chemokines bind HSVEC and induce CXCR3 internalization on T cells. rI-TAC, IP-10, or Mig (250 ng/ml), or all three were added to the medium of resting HSVEC monolayers for 1 h. After washes, CXCR3ϩ T cells were incubated with chemokine-treated HSVEC for 30 min. CXCR3 expression was then assessed by FACS analysis. Data represent the mean Ϯ SE of four independent experiments. partially inhibit the ability of HSVEC supernatants from down- regulating CXCR3 expression on T cells (Fig. 8). The addition of Downloaded from all three mAbs was no better than the I-TAC mAb alone (Fig. 8). FIGURE 8. Neutralization of CXCR3 ligands reveals a preferential role These results suggest that I-TAC is the physiological inducer of ␥ for I-TAC in HSVEC-mediated CXCR3 internalization. Conﬂuent HSVEC CXCR3 internalization produced by IFN- -activated HSVEC. The monolayers (A) or HSVEC supernatants (B) were incubated with anti-I- fact that anti-I-TAC did not completely inhibit HSVEC-induced TAC, anti-IP-10, anti-Mig, or isotype control (50 ␮g/ml) for1hat37¡C. CXCR3 down-regulation suggests that I-TAC bound to HSPG on CXCR3ϩ T cells were resuspended in HSVEC supernatant (B) or added

HSVEC may not be completely accessible to mAb neutralization, onto HSVEC monolayers (A) and incubated for1hat37¡C, 5% CO2. http://www.jimmunol.org/ or that the surface of HSVEC has an I-TAC independent mecha- CXCR3 expression on T cells was then assessed by FACS analysis. Data nism for inducing CXCR3 internalization. represent the mean Ϯ SE of three independent experiments. Values of p for anti-I-TAC compound to no Ab control. Role of CXCR3 ligands in transendothelial migration The ability of IP-10, Mig, and I-TAC to induce transendothelial migration being performed with filters coated with HMEC-1 en- cell migration of CXCR3ϩ 300-19 cells was compared with the dothelial cells (Fig. 9, A and B). Chemotaxis across uncoated filters ability of these ligands to induce chemotaxis across bare filters. essentially confirmed published studies, demonstrating compara- Both assays used 5 ␮M-pore size filters with transendothelial cell ble chemotaxis for all three ligands. However, in contrast, I-TAC by guest on October 1, 2021

FIGURE 9. CXCR3 ligands induced transendothelial migration. Migration of CXCR3Ð300-19 cells through bare filters (5 ␮m) or through these same filters covered with a confluent monolayer of HMEC-1 was analyzed in ChemoTx 96-well chambers (NeuroProbe). After5hofincubation at 37¡Cin5%

CO2, migration was determined by counting migrating cells in the bottom chambers containing CXCR3 ligands at indicated concentrations. Top panels, Migration through bare filters (left) or through HMEC-1 monolayers (right). Comparison between Mig, IP-10, and I-TAC. Bottom panels, Individual comparison for Mig, IP-10, and I-TAC. छ, ϩHMEC-1 cells; F, ϪHMEC-1 cells. Data is representative of four independent experiments and is presented as the mean Ϯ SE of one representative experiment performed in triplicate. 7092 CXCR3 INTERNALIZATION AFTER T CELL-ENDOTHELIAL CELL CONTACT was markedly more potent and more efficacious at inducing mi- gesting that intracellular signaling pathways that activate PKC gration across endothelial cells. Although the presence of HMEC-1 may lead to agonist-independent internalization of CXCR3, a increased the number of cells that migrated across the filter for all mechanism that was reported for CXCR4 (30) and CCR3 (31). In ligands, this effect was more pronounced for I-TAC, especially at contrast, PKC activation was not significantly involved in agonist- low concentrations. For example, at 10 ng/ml I-TAC there was an induced CXCR3 internalization. ϳ33-fold increase in the number of cells that migrated across We found that anti-I-TAC mAb only partially blocked HSVEC HMEC-1 coated filters compared with bare filters, while HMEC-1 monolayer- and supernatant-mediated CXCR3 internalization in T cells only caused an ϳ5-fold increase to IP-10 and an ϳ6-fold cells and that anti-Mig and anti-I-TAC had no effect on this pro- increase to Mig. cess. This was observed despite the fact that these mAbs can neutralize recombinant protein-induced internalization, chemotaxis, Discussion and calcium flux. The inability of these Abs to neutralize HSVEC- Ligand-induced desensitization and internalization of chemokine induced CXCR3 internalization is unclear but may reflect the fact receptors is thought to be a critical mechanism allowing leukocytes that these chemokines are secreted from cells bound to proteogly- to maintain their capacity to respond to small changes in a che- cans (A. D. Luster and L. Wagner, unpublished observations; Ref. motactic gradient (32). This process has largely been studied by 36) or bound to proteoglycans on cells which limit the ability of the addition of recombinant chemokines to purified leukocytes. the mAbs to neutralize them. Alternatively, additional ligands may Chemokine receptor internalization has not been studied in more be present on HSVEC cells or in their supernatants or there may physiologically relevant settings, such as leukocyte contact with even be ligand-independent mechanisms that contribute to a loss of endothelial cells, and the role of receptor internalization in the CXCR3 surface expression. For example, it has recently been Downloaded from process of transendothelial migration has also been largely ig- shown that CXCR1 and CXCR2 undergo surface cleavage follow- nored. It has recently been demonstrated that under flow condi- ing activation of neutrophils with TNF-␣ and LPS (37), a process tions, stromal cell-derived factor-1 (SDF-1/CXCL12) bound to inhibited by metalloproteinase inhibitors. This process would have HUVEC promoted lymphocyte transendothelial migration, while to be specific for certain chemokine receptors because CXCR4 soluble SDF-1 had little effect on this process (33), suggesting that surface expression on T cells was unaffected by coculture with chemokines bound to endothelial cells provide the critical signal HSVECs while CXCR3 was down-regulated. for transendothelial migration. In the present study, we have found To explore the functional consequences of CXCR3 down-reg- http://www.jimmunol.org/ that CXCR3 is efficiently internalized following T cell contact with ulation, we examined the ability of CXCR3 ligands to induce che- IFN-␥-activated HSVECs and that cell contact is more efficient at motaxis of CXCR3Ð300-19 cells across uncoated filters and filters inducing internalization than conditioned medium collected from coated with endothelial cells. We found that the presence of en- these cells. Furthermore, although IFN-␥-activated HSVEC ex- dothelial cells preferentially increased the potency and efficacy of press more IP-10 and Mig than I-TAC (28), I-TAC is the factor I-TAC-induced transendothelial migration compared with IP-10 present in HSVEC that is responsible for inducing CXCR3 and Mig. Although the reason for this cannot be ascertained from internalization. these experiments, in light of our data, it is interesting to speculate

These findings are consistent with our observation that rI-TAC that preferential receptor internalization induced by I-TAC plays a by guest on October 1, 2021 was the most potent inducer of CXCR3 internalization. This type role in its marked ability to induce transendothelial migration. It of functional hierarchy in terms of chemokine-induced receptor will be interesting to determine whether receptor internalization internalization (I-TAC Ͼ IP-10 Ͼ Mig) has been described for plays a role in transendothelial cell migration. other chemokines, such as CCR5 ligands (amino-oxypentane- Our findings that activated endothelial cells retain functional RANTES Ͼ wild-type RANTES; Ref. 34), CCR3 ligands CXCR3 ligands is consistent with Piali et al. (20), who found that (RANTES Ͼ eotaxin; Ref. 31), and CXCR2 ligands (IL-8 Ͼ T cells adhered to IFN-␥ and TNF-␣-stimulated HUVEC, and that granulocyte chemotactic protein-2 Ͼ neutrophil-activating pep- this effect was partially blocked by an anti-CXCR3 mAb. Our tide-2; Ref. 35). CXCR3 ligands, as well as IFN-␥-activated study suggests that following the induction of chemokine-induced HSVECs, specifically induced CXCR3, but not CXCR4, surface firm adhesion, CXCR3 is internalized on T cells and that this may down-regulation on T cells. This finding indicates that IFN-␥- play a role in transendothelial cell migration. activated HSVEC express CXCR3 ligands but not the CXCR4 ligand SDF-1. It also reflects the fact that CXCR3 agonists are not Acknowledgments globally activating a pathway that down-regulates chemokine We are grateful to Francois Mach and Peter Libby for providing us with receptors but are specifically targeting CXCR3 for internalization HSVECs, Edwin Ades for HMEC-1 cells, and Otto Yang for CD4ϩ T cells. following activation. These data also imply that different chemokines differentially activate the same receptor resulting in different References biological outcomes. 1. Luster, A. D. 1998. Chemokines: chemotactic cytokines that mediate inflamma- Our data also revealed that the signal transduction pathway me- tion. N. Engl. J. Med. 338:436. 2. Luster, A. D., and J. V. Ravetch. 1987. Biochemical characterization of a ␥ diating agonist-induced CXCR3 internalization is distinct from the interferon-inducible cytokine (IP-10). J. Exp. Med. 166:1084. pathway that mediates CXCR3-induced chemotaxis. We found 3. Farber, J. M. 1993. Humig: a new human member of the chemokine family of cytokines. Biochem. Biophys. Res. Comm. 192:223. that wortmannin, as well as PTX, did not inhibit agonist-induced 4. Cole, K. E., C. A. Strick, M. Loetscher, T. J. Paradis, K. T. Ogborne, CXCR3 internalization as was reported for CCR3. This is in con- R. P. Gladue, W. Lin, J. G. Boyd, B. Moser, D. E. Wood, et al. 1998. Interferon trast to agonist-induced chemotaxis, which is inhibited by both inducible T-cell ␣ chemoattractant (I-TAC): a novel non-ELR CXC chemokine with potent activity on activated T-cells through selective high affinity binding to wortmannin and PTX (35). The finding that genestein, a tyrosine CXCR3. J. Exp. Med. 187:2009. kinase inhibitor, could not block CXCR3 internalization is consis- 5. Mach, F., A. Sauty, A. Iarossi, G. Sukhova, K. Neote, P. Libby, and A. Luster. ␥ tent with previous reports on other chemokine receptors, suggest- 1999. The interferon- inducible CXC-chemokines IP-10, Mig and I-TAC are differentially expressed by human atheroma-associated cells: implication for lym- ing that phosphorylation of serine/threonine residues, but not of phocyte recruitment in atherogenesis. J. Clin. Invest. 104:1041. tyrosine residues, by specific G protein-receptor kinases is required 6. Loetscher, M., B. Gerber, P. Loestcher, S. A. Jones, L. Piali, I. Clark-Lewis, M. Baggiolini, and B. Moser. 1996. Chemokine receptor specific for IP-10 and for receptor sequestration. CXCR3 sequestration was induced by Mig: structure, function, and expression in activated T-lymphocytes. J. Exp. Med. direct activation of PKC with PMA and calcium ionophore, sug- 184:963. The Journal of Immunology 7093

7. Loetscher, M., P. Loetscher, N. Brass, E. Meese, and B. Moser. 1998. Lympho- 23. Chuang, T. T., L. Iacovelli, M. Sallese, and A. De Blasi. 1996. G protein-coupled cyte-specific chemokine receptor CXCR3: regulation, chemokine binding and receptors: heterologous regulation of homologous desensitization and its impli- gene localization. Eur. J. Immunol. 28:3696. cations. Trends Pharmacol. Sci 17:416. 8. Qin, S., J. B. Rottman, P. Myers, N. Kassam, M. Weinblatt, M. Loetscher, 24. Aragay, A. M., A. Ruiz-Gomez, P. Penela, S. Sarnago, A. Elorza, A. E. Koch, B. Moser, and C. R. Mackay. 1998. The chemokine receptors M. C. Jimenez-Sainz, and J. F. Mayor. 1998. G protein-coupled receptor kinase CXCR3 and CCR5 mark subsets of T cells associated with certain inflammatory 2 (GRK2): mechanisms of regulation and physiological functions. FEBS Lett. reactions. J. Clin. Invest. 101:746. 430:37. 9. Butcher, E. C., and L. J. Picker. 1996. Lymphocyte homing and homeostasis. 25. Signoret, N., M. M. Rosenkilde, P. J. Klasse, T. W. Schwartz, M. H. Malim, Science 272:60. J. A. Hoxie, and M. Marsh. 1998. Differential regulation of CXCR4 and CCR5 10. Cella, M., D. Jarrossay, F. Facchetti, O. Alebardi, H. Nakajima, A. Lanzavecchia, endocytosis. J. Cell Sci. 111:2819. and M. Colonna. 1999. Plasmacytoid monocytes migrate to inflamed lymph 26. Solari, R., R. E. Offord, S. Remy, J. P. Aubry, T. N. C. Wells, E. Whitehorn, nodes and produce large amounts of type I interferon. Nat. Med. 5:919. T. Oung, and A. E. I. Proudfoot. 1997. Receptor-mediated endocytosis of CC- 11. Trentin, L., C. Agostini, M. Facco, F. Piazza, A. Perin, M. Siviero, C. Gurrieri, chemokines. J. Biol. Chem. 272:9617. S. Galvan, F. Adami, R. Zambello, and G. Semenzato. 1999. The chemokine 27. Wong, J. T., and R. B. Colvin. 1987. Bi-specific monoclonal antibodies: selective receptor CXCR3 is expressed on malignant B cells and mediates chemotaxis. binding and complement fixation to cells that express two different surface an- J. Clin. Invest. 104:115. tigens. J. Immunol. 139:1369. 12. Jones, D., R. J. Benjamin, A. Shahsafaei, and D. M. Dorfman. 2000. The che- 28. Marx, N., F. Mach, A. Sauty, J. H. Leung, M. N. Sarafi, R. M. Ransohoff, mokine receptor CXCR3 is expressed in a subset of B-cell lymphomas and is a P. Libby, J. Plutzky, and A. D. Luster. 2000. Peroxisome proliferator-activated ␥ ␥ marker of B-cell chronic lymphocytic leukemia. Blood 95:627. receptor- activators inhibit IFN- -induced expression of the T cell-active CXC 13. Jinquan, T., C. Jing, H. H. Jacobi, C. M. Reimert, A. Millner, S. Quan, chemokines IP-10, Mig, and I-TAC in human endothelial cells. J. Immunol. 164: J. B. Hansen, S. Dissing, H. J. Malling, P. S. Skov, and L. K. Poulsen. 2000. 6503. CXCR3 expression and activation of eosinophils: role of IFN-␥-inducible pro- 29. Luster, A. D., S. M. Greenberg, and P. Leder. 1995. The IP-10 chemokine binds tein-10 and monokine induced by IFN-␥. J. Immunol. 165:1548. to a specific cell surface heparan sulfate site shared with platelet factor 4 and inhibits endothelial cell proliferation. J. Exp. Med. 182:219. 14. Romagnani, P., F. Annunziato, L. Lasagni, E. Lazzeri, C. Beltrame, M. Fran- 30. Signoret, N., J. Oldbridge, A. Pelchen-Matthews, P. J. Klasse, T. Tran, calanci, M. Uguccioni, G. Galli, L. Cosmi, L. Maurenzig, et al. 2001. Cell cycle- L. F. Brass, M. M. Rosenkilde, T. W. Schwartz, W. Holmes, W. Dallas, et al. dependent expression of CXC chemokine receptor 3 by endothelial cells mediates Downloaded from 1997. Phorbol esters and SDF-1 induce rapid endocytosis and down modulation angiostatic activity. J Clin Invest 107:53. of the chemokine receptor CXCR4. J. Cell Biol. 139:651. 15. Agostini, C., M. Cassatella, R. Sancetta, R. Zambello, L. Trentin, S. Gasperini, 31. Zimmermann, N., J. J. Conkright, and M. E. Rothenberg. 1999. CC chemokine A. Perin, F. Piazza, M. Siviero, A. D. Luster, et al. 1998. Involvement of the receptor-3 undergoes prolonged ligand-induced internalization. J. Biol. Chem. interferon-inducible protein 10 (IP-10) chemokine in sarcoid granulomatous re- 274:12611. actions. J. Immunol. 161:6413. 32. Murphy, P. M. 1994. The molecular biology of leukocyte chemoattractant recep- 16. Loestcher, P., M. Uguccioni, L. Bordoli, M. Baggiolini, B. Moser, C. Chizzolini, tors. Annu. Rev. Immunol. 12:593. and J. M. Dayer. 1998. CCR5 is characteristic of Th1 lymphocytes. Nature 391: 33. Cinamon, G., V. Shinder, and R. Alon. 2001. Shear forces promote lymphocyte

344. migration across vascular endothelium bearing apical chemokines. Nat. Immunol. http://www.jimmunol.org/ 17. Gottlieb, A. B., A. D. Luster, D. N. Posnett, and D. M. Carter. 1988. Detection 2:515. ␥ of a interferon-induced protein IP-10 in psoriatic plaques. J. Exp. Med. 168: 34. Mack, M., B. Luckow, P. J. Nelson, J. Cihak, G. Simmons, P. R. Clapham, 941. N. Signoret, M. Marsh, M. Stangassinger, F. Borlat, et al. 1998. Aminooxypen- 18. Kaplan, G., A. D. Luster, G. Hancock, and Z. A. Cohn. 1987. The expression of tane-RANTES induces CCR5 internalization but inhibits recycling: a novel in- ␥ a interferon-induced protein (IP-10) in delayed immune responses in human hibitory mechanism of HIV infectivity. J. Exp. Med. 187:1215. skin. J. Exp. Med. 166:1098. 35. Feniger-Barish, R., D. Belkin, A. Zaslaver, S. Gal, M. Dori, M. Ran, and 19. Lahrtz, F., L. Piali, D. Nadal, H. W. Pfister, K. S. Spanaus, M. Baggiolini, and A. Ben-Baruch. 2000. GCP-2-induced internalization of IL-8 receptors: hierar- A. Fontana. 1998. Chemotactic activity on mononuclear cells in the cerebrospinal chical relationships between GCP-2 and other ELRϩ-CXC chemokines and fluid of patients with viral meningitis is mediated by interferon-␥ inducible pro- mechanisms regulating CXCR2 internalization and recycling. Blood 95:1551. tein-10 and monocyte chemotactic protein-1. Eur. J. Immunol. 10:2484. 36. Wagner, L., O. O. Yang, E. A. Garcia-Zepeda, Y. Ge, S. A. Kalams, B. D. 20. Piali, L., C. Weber, G. LaRosa, C. R. Mackay, T. A. Springer, I. Clark-Lewis, and Walker, M. S. Pasternack, and A. D. Luster. 1988. ␤-chemokines are released B. Moser. 1998. The chemokine receptor CXCR3 mediates rapid and shear- from HIV-1-specific cytolytic T-cell granules complexed to proteoglycans. Na- by guest on October 1, 2021 resistant adhesion-induction of effector T lymphocytes by the chemokines IP-10 ture 391:908. and Mig. Eur. J. Immunol. 28:961. 37. Khandaker, M. H., G. Mitchell, L. Xu, J. D. Andrews, R. Singh, H. Leung, 21. Neer, E. J. 1995. Heterotrimeric G proteins: organizers of transmembrane signals. J. Madrenas, S. S. Ferguson, R. D. Feldman, and D. J. Kelvin. 1999. Metallo- Cell 80:249. proteinases are involved in lipopolysaccharide- and tumor necrosis factor-␣-me- 22. Ward, S. G., K. Bacon, and J. Westwick. 1998. Chemokines and T lymphocytes: diated regulation of CXCR1 and CXCR2 chemokine receptor expression. Blood more than an attraction. Immunity 9:1. 93:2173.