The Journal of Immunology The Chemokine Stromal Cell-Derived Factor-1␣ Modulates ␣ ␤ 4 7 Integrin-Mediated Lymphocyte Adhesion to Mucosal Addressin Cell Adhesion Molecule-1 and Fibronectin1 Natalia Wright,* Andre´s Hidalgo,2* Jose´Miguel Rodrı´guez-Frade,† Silvia F. Soriano,† Mario Mellado,† Marisa Parmo-Caban˜as,* Michael J. Briskin,‡ and Joaquin Teixido´3* ␣ ␤ The interaction between the integrin 4 7 and its ligand, mucosal addressin cell adhesion molecule-1, on high endothelial venules represents a key adhesion event during lymphocyte homing to secondary lymphoid tissue. Stromal cell-derived factor-1␣ (SDF-1␣) is a chemokine that attracts T and B lymphocytes and has been hypothesized to be involved in lymphocyte homing. In this work ؉ ␤ ␣ we show that 4 7-mediated adhesion of CD4 T lymphocytes and the RPMI 8866 cell line to mucosal addressin cell adhesion molecule-1 was up-regulated by SDF-1␣ in both static adhesion and cell detachment under shear stress assays. Both naive and ␤ ␣ ␣ ␣ ؉ memory phenotype CD4 T cells were targets of SDF-1 -triggered increased adhesion. In addition, SDF-1 augmented 4 7- dependent adhesion of RPMI 8866 cells to connecting segment-1 of fibronectin. While pertussis toxin totally blocked chemotaxis ␤ ␣ ␣ ؉ of CD4 and RPMI 8866 cells to SDF-1 , enhanced 4 7-dependent adhesion triggered by this chemokine was partially inhibited, ␣ ␣ ␣ indicating the participation of G i-dependent as well as G i-independent signaling. Accordingly, we show that SDF-1 induced ␣ ␣ a rapid and transient association between its receptor CXCR4 and G i, whereas association of pertussis toxin-insensitive G 13 with CXCR4 was slower and of a lesser extent. SDF-1␣ also activated the small GTPases RhoA and Rac1, and inhibition of RhoA ␣ ␤ ␣ activation reduced the up-regulation of 4 7-mediated lymphocyte adhesion in response to SDF-1 , suggesting that activation of RhoA could play an important role in the enhanced adhesion. These data indicate that up-regulation by SDF-1␣ of lymphocyte ␣ ␤ adhesion mediated by 4 7 could contribute to lymphocyte homing to secondary lymphoid tissues. The Journal of Immunology, 2002, 168: 5268–5277. ymphocytes recirculate from blood to secondary lym- longs to the Ig superfamily, is expressed on HEV in Peyer’s phoid tissues and back again into the blood circulation, a patches, mesenteric lymph nodes, and lamina propria venules L process that is of fundamental importance for normal im- within the gut, and its expression represents sites of lymphocyte mune surveillance (1–3). Homing of lymphocytes to lymphoid tis- extravasation into these intestine-associated lymphoid tissues (4, 9, by guest on October 1, 2021. Copyright 2002 Pageant Media Ltd. sues is mediated by selective interactions between adhesion recep- 10). Naive T lymphocytes (CD45RAϩ among CD4ϩ) show rela- ␣ ␤ tors expressed on lymphocytes and their ligands displayed on high tively homogeneous intermediate levels of 4 7 expression, endothelial venules (HEV)4 located on postcapillary venules (4). whereas memory B and T lymphocytes can be subdivided into The integrin ␣ ␤ is a cell adhesion receptor expressed on different ␣ ␤ high ␣ ␤ Ϫ 4 7 4 7 and 4 7 populations (6, 11, 12). Both naive and mem- T and B lymphocyte subsets that mediates their attachment to HEV ␣ ␤ Ϫ ory T cells home efficiently to Peyer’s patches, but 4 7 lym- in mucosa-associated lymphoid tissues by interacting with muco- ␣ ␤ phocytes are excluded (12, 13), indicating that 4 7 expression is sal vascular addressin mucosal addressin cell adhesion molecule-1 associated with homing to intestinal lymphoid tissues. (MAdCAM-1) (5–9). MAdCAM-1, a 60-kDa glycoprotein that be- ␣ ␤ Apart from mediating cell adhesion to MAdCAM-1, 4 7 can also http://classic.jimmunol.org interact with the connecting segment-1 region of fibronectin (CS-1/ ␣ ␤ high FN) (14, 15). This interaction might play an important role in 4 7 † *Department of Immunology, Centro de Investigaciones Biolo´gicas, and Department lymphocyte homing within lymphoid tissues rather than in recruit- of Immunology and Oncology, Centro Nacional de Biotecnologı´a, Madrid, Spain; and ‡Millenium Pharmaceuticals, Inc., Cambridge, MA 02139 ment of lymphocytes from the blood (4). Several amino acids on both ␣ ␤ ␣ ␤ Received for publication July 12, 2001. Accepted for publication March 5, 2002. 4 and 7 subunits that are critical for 4 7-dependent cell adhesion The costs of publication of this article were defrayed in part by the payment of page to MAdCAM-1 and CS-1/FN have recently been identified (16–18). Downloaded from charges. This article must therefore be hereby marked advertisement in accordance After tethering and rolling of lymphocytes on HEV, they rapidly with 18 U.S.C. Section 1734 solely to indicate this fact. ␣ ␤ stick and arrest, a process involving 4 7/MAdCAM-1 interac- 1 This work was supported by Grant SAF99-0057 from Ministerio de Ciencia y Tec- tions and favored by chemokine-triggered integrin activation (19). nologı´a. N.W. is the recipient of a predoctoral fellowship from the Comunidad de Madrid. The stromal cell-derived factor-1␣ (SDF-1␣; CXCL12) is a CXC 2 Current address: Department of Hematology, Mount Sinai School of Medicine, One chemokine that potently attracts lymphocytes (20–22) and exerts Gustave L. Levy Place, New York, NY 10029. chemoattractive and activating functions upon binding to its G 3 Address correspondence and reprint requests to Dr. Joaquin Teixido´, Department of protein-coupled receptor CXCR4, which is expressed on B and T Immunology, Centro de Investigaciones Biolo´gicas, Vela´zquez 144, 28006 Madrid, lymphocytes, including CD4ϩ and CD8ϩ cells (23–26). In addi- Spain. E-mail address: [email protected] tion, CXCR4 acts as a coreceptor for T-tropic HIV, and SDF-1␣ 4 Abbreviations used in this paper: HEV, high endothelial venule; CS-1/FN, connecting segment-1 of fibronectin; F-actin, polymerized actin; PTX, pertussis toxin; SDF-1␣, stro- inhibits T tropic HIV infection (23, 25). mal cell-derived factor-1␣; rh, recombinant human; MAdCAM-1, mucosal addressin cell Previous studies have shown that SDF-1␣ can modulate the ad- adhesion molecule-1; sMAdCAM-1-IgG, soluble human MAdCAM-1-Ig fusion protein; ϩ CHO, Chinese hamster ovary; BCECF-AM, 2Ј,7Ј-bis-(2-carboxyethyl)-5-(and-6)-car- hesive activity of the VLA-4 integrin on CD34 human bone mar- boxyfluorescein acetoxymethyl ester; SLC, secondary lymphoid tissue chemokine. row hemopoietic progenitors, myeloma cells, and T lymphocytes Copyright © 2002 by The American Association of Immunologists 0022-1767/02/$02.00 The Journal of Immunology 5269 (27–30). SDF-1␣ is constitutively expressed on many tissues, in- at a flow rate of 1 dyne/cm2. Flow was then stopped, and cells were allowed cluding secondary lymphoid tissues (20, 31), and is therefore a to settle for different times. Total cells from different fields were counted 2 potential candidate to contribute to lymphocyte homing during re- before flow was restored (0.5–2 dyne/cm ), and cells remaining tightly bound for Ͼ3 min were counted. Data are presented as the percentage of circulation. In the present work we have investigated whether cells remaining bound compared with total cells in each field before rees- ␣ ␤ 4 7-dependent lymphocyte adhesion to MAdCAM-1 and CS- tablishing the flow. 1/FN can be subjected to regulation by SDF-1␣. Modulation of this Actin polymerization and chemotactic assays adhesion could contribute to lymphocyte homing to mucosa-asso- ciated lymphoid tissue. To determine the content of polymerized actin (F-actin), 105 cells per con- dition were permeabilized, fixed, and stained in a single step by addition of a2ϫ solution containing 0.5 mg/ml L-␣-lysophosphatidyl-choline (Sigma- Materials and Methods Aldrich, St. Louis, MO), 8% formaldehyde, and 4 U/ml FITC-phalloidin Cells and Abs (Molecular Probes, Eugene, OR). Cells were incubated at 22°C for 10 min, washed with PBS, and subjected to flow cytometry. For chemotactic as- The human B lymphoblastoid cell line RPMI 8866 was cultured in RPMI says, RPMI 8866 (2 ϫ 105)orCD4ϩ cells (3 ϫ 105)in100␮l adhesion 1640 medium (BioWhittaker, Verviers, Belgium) supplemented with 10% medium were placed in the upper chamber of a Transwell (5-␮m pore size; FBS (BioWhittaker) and antibiotics (complete medium). The Chinese ham- ␮ ␣ ␣ Costar). Then 600 l adhesion medium with or without rhSDF-1 (200 ster ovary (CHO)-MAdCAM-1 transfectants were maintained in -MEM ng/ml) was added to the lower chamber, and cells were allowed to migrate (BioWhittaker) supplemented with 10% FBS and containing 0.8 mg/ml for 3 h at 37°C. Viable migrated cells were counted in a flow cytometer, G418 (Calbiochem, San Diego, CA). Human PBMC were isolated from analyzing each sample in the same predetermined time and flow conditions. buffy coats using a Ficoll density gradient (Biochrom, Berlin, Germany). Where indicated, cells were treated with 500 ng/ml PTX for2hat37°C. After attachment to plastic for 1.5 h at 37°C in complete medium, nonad- hered cells were recovered, and CD4ϩ T lymphocytes were purified with GTPase activity assays the CD4 Positive Isolation kit (Dynal Biotech, Oslo, Norway). Purity was Ͼ99% for each sample, as analyzed by flow cytometry (EPICS XL; We followed essentially the method previously reported (37). The GST- Coulter, Hialeah, FL). The mAbs used in this study included anti-␤ Lia C21 and GST-PAK-CD fusion proteins were generated as previously de- 1 ␣ 1/2.1, anti-CD4 T4, anti-CD45RA RP 2/2.1, P3X63 (all gifts from Dr. F. scribed (38). To determine the effect of SDF-1 on RhoA and Rac1 acti- ␣ ␤ vation, cells were treated with or without 150 ng/ml rhSDF-1␣, washed in Sa´nchez-Madrid, Hospital de la Princesa, Madrid, Spain), anti- 4 7 Act-1 (32), anti-CD45RO (Caltag Laboratories, Burlingame, CA), anti-CXCR4 ice-cold PBS, and incubated for 15 min at 4°C in lysis buffer (37).