Differentiation of Regulatory T Cells 1 Is Induced by CD2 Costimulation Abdelilah Wakkach, Françoise Cottrez and Hervé Groux This information is current as J Immunol 2001; 167:3107-3113; ; of September 29, 2021. doi: 10.4049/jimmunol.167.6.3107 http://www.jimmunol.org/content/167/6/3107 Downloaded from References This article cites 40 articles, 23 of which you can access for free at: http://www.jimmunol.org/content/167/6/3107.full#ref-list-1

Why The JI? Submit online.

http://www.jimmunol.org/ • 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

*average by guest on September 29, 2021 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. Differentiation of Regulatory T Cells 1 Is Induced by CD2 Costimulation1

Abdelilah Wakkach, Franc¸oise Cottrez, and Herve´Groux2

Induction and maintenance of peripheral tolerance is an important phenomenon for the control of homeostasis in the . There is now compelling evidence for CD4؉ T cells that prevent immune pathology, both in and in transplantation. However, the mechanisms involved in the specific differentiation of these T cells are unknown. We had previously shown that repetitive stimulations of naive T cells in the presence of IL-10 induce the differentiation of T regulatory cells 1. We further dissected the mechanism of IL-10 function and demonstrated that IL-10 acts by the down-regulation of most costimulatory molecules without modifying the expression of CD58. Using artificial APCs expressing various costimulatory molecules, we dem- onstrated that, in contrast to other costimulation patterns, costimulation via CD2 alone, in the absence of costimulations through

CD28- or LFA-1, induced anergy in an IL-10-independent pathway along with the differentiation of Ag-specific regulatory Downloaded from T cells. T regulatory cell-1 differentiation via CD2 was very efficient as both high IL-10 secretion and regulatory function were observed after the first stimulation of naive T cells with CD32-CD58 L cells. The possibility to rapidly induce the differentiation of Ag-specific regulatory T cells will certainly accelerate their characterization and their potential use as regulators of T cell- mediated diseases. The Journal of Immunology, 2001, 167: 3107–3113.

uccessful human organ transplantation is complicated by cules CD80 (5) and CD86 (6) and the adhesion molecule CD54. How- http://www.jimmunol.org/ the development of allograft rejection and despite the use ever, IL-10 did not modify CD58/LFA-3 expression on APCs (7). S of immunosuppressive medications true graft tolerance is CD58/LFA-3, which is a ligand of CD2 (8), is a member of the seldom realized. One way to achieve this problem would be to Ig superfamily expressed widely by hemopoietic and nonhemo- specifically inhibit the expansion of alloreactive T cells or to in- poietic cells. CD2, which is also a member of the Ig superfamily, duce the differentiation of regulatory T cells. T cell activation is a is found on , peripheral T , and NK cells. carefully orchestrated process that involves the TCR and a multi- CD2-LFA-3 interaction is considered to contribute to T cell acti- tude of accessory molecules expressed on T cells and APCs. Op- vation by strengthening the adhesion between T cells and APCs or erationally, two types of molecules that modulated TCR-mediated target cells (9), thus helping the TCR to reinforce contact with its T cell activation can be distinguished: 1) accessory molecules such ligands. Besides these adhesion enhancer mechanisms, LFA3-CD2 by guest on September 29, 2021 as LFA-1, which facilitate TCR triggering by promoting T cell- interaction has also been reported to be an important regulator of APC adhesion, and 2) costimulatory molecules, such as CD28, T cell responses. Indeed, in murine models, sequential treatment which enhance T cell activation without affecting the rate of TCR with CD2 was shown to induce tolerance in vivo (10). triggering (1, 2). In this study, we dissected the functional mechanism of IL-10 on We have recently shown that both human and mouse CD4ϩ T the differentiation of Tr1 cells and addressed the importance of the cells, repeatedly stimulated in the presence of IL-10 differentiate costimulatory role of CD2 in the induction of tolerance through the into a new subset of CD4ϩ T cells different from the Th1 and Th2 differentiation of regulatory T cells. Our data demonstrate that co- cells (3). These cells, termed T regulatory cell 1 (Tr1),3 have a stimulation of TCR with the CD2 molecule specifically induced poor proliferative response, secrete high levels of IL-10, and reg- the differentiation of Ag-specific Tr1 cells. Moreover, Tr1 cell dif- ulate Th1 and Th2 immune responses in vivo (3, 4). IL-10 is a ferentiation is not due to a lack of costimulation or poor stimula- , produced by a variety of cells, with important anti- tion as no regulatory T cells are induced after T cell stimulation in inflammatory and immunosuppressive properties. These properties the absence of CD2 mobilization. However, the specific differen- are associated with a reduction of Ag-presenting capacity of APC. tiation of Tr1 cells after CD2 and CD3 stimulation can be reverted Indeed, IL-10 down-regulates the expression of costimulatory mole- by the mobilization of the CD28 molecule by its respective ligands. Materials and Methods Institut National de la Sante« et de la Recherche Me«dicale Unite« 343, Hopital de Cells, Abs, and other reagents l’Archet, Nice, France Received for publication February 13, 2001. Accepted for publication July 3, 2001. Mouse anti-human CD3 (clone X35.7), CD4 (clone O.516), CD8 (clone L533), and CD11b (cloneOKM1), CD20 (clone 1F5) kindly provided by The costs of publication of this article were defrayed in part by the payment of page E. A. Clark (University of Washington, Seattle, WA); CD28 (clone 28.2), charges. This article must therefore be hereby marked advertisement in accordance a gift from Dr. D. Olive (Institut National de la Sante«et de la Recherche with 18 U.S.C. Section 1734 solely to indicate this fact. Me«dicale, Unite«119, Marseille, France); and CD58 (clone TS 2.9), a kind 1 This work was supported by the Association pour la Recherche sur le Cancer, La gift from Dr. B. Haynes (Duke University, Durham, NC) have been ob- Fondation pour la Recherche Me«dicale, and La Ligue Re«gionale Contre le Cancer. tained in our laboratory and described previously (11, 12). Mouse anti- 2 Address correspondence and reprint requests to Dr. Herve«Groux, Institut National human IL-10 (clone JES3-9D7), antihuman IL-2R (BB-10), and human de la Sante«et de la Recherche Me«dicale Unite«343, Hopital de l’Archet, Route de recombinant IL-2 and IL-10 were a kind gift from Dr. R. L. Coffman Saint-Antoine de Ginestiere, 06200 Nice, France. E-mail address: [email protected] (DNAX Research Institute, Palo Alto, CA). The L cell line is a mouse 3 Abbreviations used in this paper: Tr1, T regulatory cell 1; NIP, nitrophenylacetic fibroblast cell line. The L cells transfected with mouse CD32 or human acid. CD80 were a gift from Dr. L. Lanier (DNAX Research Institute). The P815

Copyright © 2001 by The American Association of Immunologists 0022-1767/01/$02.00 3108 ROLE OF CD2 IN Tr1 CELL DIFFERENTIATION is a mouse mastocytoma cell line that expresses Fc␥ receptors. DR1ϩ EBV cloned at one cell per well by flow cytometry (FACSVantage SE; BD lymphoblastoid B cells were a generous gift from Dr. H. Yssel (Institut Biosciences) in 96-wells precoated with anti-CD3 mAb (1 ␮g/ml) in 0.1 M National de la Sante« et de la Recherche Me«dicale, Unite« 454, Montpellier, Tris (pH 9.4; Sigma, St. Louis, MO). After sorting, a mixture of irradiated France). feeder cells (JY, 105/ml, and PBMCs, 106/ml) and 10 U/ml rIL-2 in 100 ␮l was added. The clones were expanded with IL-2 (10 U/ml) and then an- Culture medium alyzed for cytokine secretion after activation with immobilized anti-CD3 (1 ϩ ␮g/ml) and anti-CD28 mAb (1 ␮g/ml) for 48 h. DR1 EBV lymphoblastoid B cells and T cells were cultured in RPMI 1640 (Life Technologies, Cergy Pontoise, France) supplemented with 2 Immunofluorescence analysis mM L-glutamine, 100 U/ml penicillin-streptomycin (both from Life Tech- nologies), and 10% FCS (Boehringer Mannheim, Roche, France). L cell For detection of cell surface Ags, 105 cells were labeled with PE- or FITC- and P815 cell transfectants were cultured in F-12 medium (Life Technol- conjugated mAbs. Cells were incubated for 30 min with the appropriate Ab ogies) supplemented with 10% FCS (Roche, Myelan, France), 2 mM L- at 4¡C in PBS (Life Technologies) with 0.1% BSA (Sigma) and 0.02 mM glutamine (Life Technologies), 100 U/ml penicillin-streptomycin (Life NaN3. After three washes, the labeled cell samples were analyzed on a Technologies), and 50 ␮M 2-ME (Life Technologies). FACScan. Cell purification Determination of lymphokine production PBMCs were prepared by centrifugation over Ficoll-Hypaque (Amersham ␥ ϩ Sandwich ELISAs were used to measure IL-2, IL-4, IL-10, and IFN- as Pharmacia Biotech, Uppsala, Sweden). CD4 T cells were purified by previously described (13). In brief, ELISA plates (Polylab, Strasbourg, negative depletion. Negative purification was performed using a mixture of ϩ France) were coated with the appropriate anti-cytokine Abs (Mah IL-2, Abs directed against non-CD4 T cells: mouse anti-human CD8, CD11b, 8D4, 8, JES3-9D7, and A35, respectively) and incubated at 4¡C overnight. and CD20. Cells were incubated with saturating amounts of Abs for 20 min After incubation, plates were blocked for 30 min at room temperature by Downloaded from at 4¡C. After washing, dynabeads (Dynal, Compie`ne, France) were added adding 150 ␮l of 20% FCS/PBS (Life Technologies) containing 0.04% at 10:1 bead:target cell ratio and incubated for1hat4¡C. Beads and ϩ Tween 20 (Sigma) to each well. Supernatants from in vitro-stimulated contaminating cells were removed by magnetic field. CD4 T cells were CD4ϩ T cells or from activated T cell clones were added at a volume of 50 analyzed by FACStar (BD Biosciences, Le Pont de Claix, France) and ␮l/well. Plates were incubated overnight at 4¡C, then washed, and the revealed to be 95% positive. second-step Ab:nitrophenyl acetic acid-conjugated mAb (GaH IL-2, MP4- 25D2, JES312G8, 1, and B27 for IL-2, IL-4, IL-10, and IFN-␥, respec- Generation and culture of mouse L cell and P815 transfectants tively) was added at 50 ␮l/well. Plates were incubated for1hatroom

A full-length cDNA for human CD58 (LFA-3) was generated from PBMC temperature, then washed, and the anti-nitrophenyl acetic acid-HRP con- http://www.jimmunol.org/ RNA using primers flanked by KpnI and NotI sites. This cDNA was cloned jugate was added to each well. Plates remained at room temperature for 1 h, into pcDNA 3.1/hygro (Invitrogen, San Diego, CA), digested with ScaI after which they were washed and 100 ␮l/well substrate containing 1 (Roche), and transfected into CD32- and CD32-CD80-L cells using stan- mg/ml 2,2Ј-azino-bis (Sigma, St. Quentin, Fallavier, France), 0.0003% dard procedures of Lipofectamine (Life Technologies) to create CD32- H2O2 in Na2HPO4, and 0.05 M citric acid was added. After the substrate CD58-, and CD32-CD58-CD80-L cells. Stable transfectants were purified was developed, applying 50 ␮l of 0.2 M citric acid solution to each well by staining transfected cell populations with mAb against CD58 and then stopped the reaction. The plates were read on an ELISA reader (Lab- performing multiple sterile sorts using a FACSVantage SE (BD Bio- systems iEMS reader; Labsystems, Helsinki, Finland). sciences). Similarly, HLA-DR1 ␣- and ␤-chain cDNA were generated from PBMC RNA isolated from a DR1ϩ individual, cloned into pcDNA3.1 Results

(Invitrogen), and cotransfected into L cells as described above. DR1-L Costimulation with CD58 (LFA-3) alone induced the by guest on September 29, 2021 cells were purified by sorts using a FITC-conjugated anti-HLA-DR1 mAb (Dako, Trappes, France). DR1-CD58-L cells were obtained by transfection differentiation of Tr1-type cells of the DR1-L cells with the cDNA of human CD58, and DR1-CD58-CD80 We have previously shown that repetitive stimulations of CD4ϩ T L cells were obtained by subcloning the cDNA of human CD80 into cells in the presence of IL-10 induces the differentiation of a novel pcDNA3.1/zeo and transfection of DR1-CD58-L cells. Similar procedures were used to obtain CD54- and CD86-L cells as well as CD80- and CD58- subset of regulatory T cells (Tr1) with immunoregulatory proper- P815 cells. ties (3). Preliminary studies have shown that differentiation of Tr1 cells in the presence of IL-10 could not be obtained in the absence Generation of Tr1 cells of APCs (data not shown). To analyze whether IL-10 effects on the CD4ϩ T cells were plated in 96-well plates (BD Biosciences) at a differentiation of Tr1 cells were due to a modification of the Ag- 6 density of 2 ϫ10 cells/ml in a total volume of 200 ␮l of RPMI 1640 presenting capacity of APCs obtained through down-modulation supplemented and stimulated with irradiated CD32-CD58-, CD32-CD80-, CD32-CD54-, CD32-CD86-L cells, or CD80-, CD58-P815 cells (600 rad) of costimulatory molecules, we used CD32-L cells expressing dif- at 5 ϫ104cells/ml in the presence of soluble human anti-CD3 (10 ␮g/ml). ferent costimulatory molecules and analyzed their capacity to in- ϩ For Ag-specific differentiation of Tr1 cells, CD4ϩ T cells were stimulated duce proliferation and cytokine production of purified CD4 T ϩ with irradiated DR1-CD58-L cells, DR1-CD58-CD80-L cells, or DR1 cells in the presence of CD3 mAb. Proliferative response of CD4ϩ EBV lymphoblastoid cells (600 rad). Cytokine synthesis was determined T cells was minimally induced by the cross-link of CD3 mAb on by the analysis of supernatants 48 h after stimulation, and proliferation was measured after a pulse with 0.5 ␮Ci of [3H]thymidine during the final 12 h the surface of L cells expressing the CD32 molecule (Fig. 1A). In of a 3-day culture. contrast, the expression of CD80 or CD86 (data not shown) on the cell surface of L cells induced maximal proliferative response as Test for regulatory function of T cells previously described (14). To a lesser extent, costimulation was CD4ϩ T cells were stimulated with irradiated PBMCs (600 rad) and sol- also obtained by cell surface expression of CD54 (ICAM-1) and uble human anti-CD3 mAb (10 ␮g/ml) in the bottom compartment of a ϩ CD58 (LFA-3) as previously shown (15). Finally, the expression transwell system. The T cells were cocultured with CD4 T cell popula- of CD58 and CD80 act in cooperation to induce proliferation (Fig. tions (previously differentiated with different costimulatory molecules) contained in the upper transwell compartment. After 2 days, the basket was 1A) whereas minimal cooperation is observed between CD54 removed and the proliferative response of the bystander CD4ϩ T cells was and CD80. measured after a pulse with 0.5 ␮Ci of [3H]thymidine during the final 18 h Analysis of released in the supernatants of stimulated of a 3-day culture. CD4ϩ T cells (Table I) shows that stimulation of CD28 by CD80 Cloning of Tr1 cells or CD86 induces high levels of IFN-␥ and IL-2 and some IL-10 and IL-4. Activation in the presence of CD54 induces solely the Human Tr1 cells were generated from CD4ϩ T cells stimulated with irra- diated CD32-CD58-L cell transfectants or CD32-CD80-L cell transfectants secretion of IL-2 as previously described (16). In contrast, activa- used as controls. After 7 days, cells were labeled with an anti-CD4 FITC tion in the presence of CD58 induces low levels of IL-2, no IL-4, and anti-CD25 PE mAbs (BD Biosciences). The CD4ϩCD25ϩ T cells were some IFN-␥, and high levels of IL-10, a cytokine secretion profile The Journal of Immunology 3109 Downloaded from http://www.jimmunol.org/

FIGURE 1. CD2 costimulation induced the differentiation of regulatory T cells (Tr1). A and B, Resting human purified CD4ϩ T cells were stimulated with anti-CD3 mAb (10 ␮g/ml) in the presence of L cells (A) or P815 (B) cell lines transfected with different costimulatory molecules as indicated. Proliferation was measured by [3H]thymidine in the last 12 h of a 72-h culture. Results represent mean Ϯ SD of triplicate culture of one representative experiment of 10. C and D, Resting human purified CD4ϩ T cells (106 cells/ml) were stimulated with anti-CD3 mAb (10 ␮g/ml) and irradiated PBMCs (600 rad, 106 cells/ml) in the bottom compartment of a transwell system. As indicated, the CD4ϩ T cells were cocultured with various CD4ϩ T cell by guest on September 29, 2021 populations (106 cells) contained in the upper transwell compartment and stimulated with anti-CD3 mAbs and irradiated PBMCs (106 cells). After 2 days, the basket was removed and the proliferative response of the bystander CD4ϩ T cells was measured after a pulse with 0.5 ␮Ci of [3H]thymidine during the final 12 h of a 72-h culture. Results represent mean Ϯ SD of triplicate culture of one representative experiment of two.

similar to the one expressed by Tr1 T cell clones (3). Interestingly, slightly enhanced bystander T cell proliferation presumably this peculiar cytokine profile was lost upon coexpression of CD80 through the secretion of IL-2 and/or IL-4. or CD54 molecules which triggered stimulation via their respec- ϩ tive ligands to enhance IL-2 secretion (Table I). CD2 costimulation-induced anergy of CD4 T cells To analyze whether the effect observed with LFA-3 stimulation independently of the secretion of IL-10 could be linked to the cell line used to express the human mole- We had previously shown that IL-10 induces anergy in CD4ϩ T cule, similar experiments were performed with P815 cells trans- cells (13). To analyze whether stimulation of CD4ϩ T cells with fected with CD80, CD54, and CD58. As shown in Fig. 1B and CD58 alone that results in high levels of IL-10 secretion would Table I, CD3 stimulation in the presence of CD80 induces maxi- induce anergy, CD4ϩ T cells were stimulated with soluble CD3 mum proliferation and secretion of IL-2, IFN-␥, and some IL-4 mAb (10 ␮g/ml) in the presence of L cells expressing CD80 or and IL-10. In contrast, activation with CD58 alone induced min- CD58 in the presence or absence of blocking anti-IL-10 mAb, imal proliferation and secretion of minimal amounts of IL-2 and blocking anti-IL-2R, or exogenous IL-2 (Fig. 2). Addition of anti- IL-4, some IFN-␥, and high levels of IL-10, showing that this IL-10 Abs partially increases the proliferative response of CD4ϩ T cytokine profile is related to activation through the CD2 molecule. cells stimulated with CD3 mAbs and L cells expressing CD58 To confirm that CD2 costimulation specifically induced the dif- alone. Blocking IL-10 did not further induce the proliferative re- ferentiation of populations, coculture experiments sponse of CD4ϩ T cells stimulated with CD3 mAbs and L cells were performed to analyze the potential regulatory function of the expressing CD80 in the presence or absence of CD58 (data not different T cell populations obtained after activation with various shown). To analyze whether anergy was induced in these cell cul- costimulation conditions (Fig. 1, C and D). The proliferation of ture conditions, the cells were kept in culture for 10 days. No resting CD4ϩ T cells in response to CD3 mAb and irradiated significant cell death was observed in either culture condition and PBMCs was dramatically reduced following coculture in the trans- comparable cell numbers were harvested at termination of the as- well system with CD3-activated Tr1 cell populations generated in says (data not shown). After this culture period, similarly to culture the presence of CD58 alone. In contrast, CD3 stimulation of other performed in the presence of exogenous IL-10 (13), CD4ϩ T cells T cell populations generated in the presence of other combinations previously stimulated with CD58 alone failed to proliferate in re- of costimulatory molecules had either no significant effect or sponse to cross-linked CD3 mAb stimulation. This unresponsive 3110 ROLE OF CD2 IN Tr1 CELL DIFFERENTIATION

Table I. Cytokine profile of CD4ϩ T cells stimulated with different APCsa

IL-2 (pg/ml) IL-4 (pg/ml) IL-10 (pg/ml) IFN-␥ (pg/ml)

L cells 24 Ϯ 3NDb 46 Ϯ 20 ND L-CD54 1,897 Ϯ 216 186 Ϯ 31 800 Ϯ 48 476 Ϯ 9 L-CD58 251 Ϯ 20 200 Ϯ 29 6,666 Ϯ 209 1,349 Ϯ 17 L-CD80 4,788 Ϯ 325 825 Ϯ 360 2,132 Ϯ 123 7,447 Ϯ 428 L-CD86 3,987 Ϯ 541 936 Ϯ 57 2,057 Ϯ 59 6,212 Ϯ 238 L-CD58/CD80 5,102 Ϯ 329 809 Ϯ 67 4,325 Ϯ 65 7,015 Ϯ 718 L-CD80/CD54 4,879 Ϯ 297 769 Ϯ 59 3,643 Ϯ 85 4,226 Ϯ 86 P815 20 Ϯ 2 210 Ϯ 59 2,053 Ϯ 30 3,070 Ϯ 35 P815-CD54 1,472 Ϯ 124 112 Ϯ 41 2,302 Ϯ 99 3,959 Ϯ 220 P815-CD58 43 Ϯ 3 124 Ϯ 39 10,658 Ϯ 471 4,730 Ϯ 100 P815-CD80 6,459 Ϯ 827 1,025 Ϯ 257 5,728 Ϯ 225 8,241 Ϯ 70

a CD4ϩ T cells (106 cells/ml) were activated with cross-linked anti-CD3 mAbs (10 ␮g/ml) in the presence of L cells (2 ϫ 105/ml) or P815 (106/ml) expressing different costimulatory molecules as indicated. Supernatants were harvested after 24 h and the levels of cytokines were analyzed by ELISA. Results represent mean Ϯ SD of triplicate experiments and are from one representative experiment of six. b ND, Not detected. Downloaded from state could not be reversed by saturating concentrations of exog- not modify the differentiation of Tr1 cells induced by CD2 co- enous IL-2 (200 U/ml) or by anti-CD28 mAbs (10 ␮g/ml; Fig. 2B). stimulation (Fig. 3). Interestingly this anergic state was also induced in CD4ϩ T cells stimulated with CD58 alone in the presence of anti-IL-10 Abs. Tr1 cells induced by CD2 costimulation suppress proliferation ϩ These results suggest that anergy induced in CD4 T cells by of naive T cells

IL-10 was due to an indirect effect of IL-10 that modulates APC http://www.jimmunol.org/ To analyze the functional properties of Tr1 T cell clones induced costimulatory molecules and not to a direct effect of IL-10 on T after stimulation with CD2 and CD3 mobilization, coculture cells. In contrast to regulatory T cell populations obtained after CD2 costimulation, CD4ϩ T cell populations obtained after CD28 costimulation could be fully reactivated by cross-linked CD3 mAb alone or in the presence of soluble CD28 mAb or IL-2 (Fig. 2B). To analyze the importance of IL-2 in anergy induction, experi- ments were performed in the presence of exogenous IL-2 or block- ing anti-IL-2R Abs. The data presented in Fig. 2 clearly show that anergy induction cannot be explained by the lack of IL-2 secretion by guest on September 29, 2021 after CD2 costimulation as anergy was obtained even after exog- enous IL-2 addition. Similarly, even after stimulation in the pres- ence of blocking anti-IL-2R Abs, T cells stimulated in the presence of CD80 were fully responsive. Taken together, these results suggest that the unresponsiveness observed in CD4ϩ T cells stimulated in the presence of CD58 alone is due to the differentiation of regulatory T cells (Tr1) that actively regulated the proliferative response of the T cell population.

CD2 costimulation induces the differentiation of Tr1 T cell clones Isolation and expansion of Ag-specific human Tr1 T cell clones have proven to be difficult. To analyze whether CD58 costimula- tion could be used to easily induce the differentiation of Tr1 T cell clones, human purified CD4ϩ T cells were stimulated with CD3 mAb in the presence of mouse CD32-CD58-L cells in the presence or absence of anti-IL-10 mAb or with CD32-CD80-L cell trans- FIGURE 2. CD2 costimulation induced IL-10-independent anergy in fectants. Seven days later, T cells were collected, washed, stained CD4ϩ T cells. A, [3H]Thymidine incorporation of purified CD4ϩ T cells ϩ ϩ with CD4 and CD25 mAbs, and CD4 CD25 T cells were cloned cultured for 3 days with anti-CD3 mAb (1 ␮g/ml) in the presence of dif- at 1 cell/well in 96-well plates and expanded in the presence of ferent L cells expressing human CD32 either alone (L cells) or in combi- IL-2. After expansion the different clones were stimulated with nation with CD58 (L-CD58) or CD80 (L-CD80). In some conditions, as cross-linked CD3 and soluble CD28 mAbs and their cytokine pro- indicated, cells were cultured in the presence of blocking anti-IL-10 mAb ␮ ␣ ␮ files were analyzed by ELISA after 48 h (Fig. 3). Approximately (9D7, 10 g/ml), blocking anti-IL-2Rr -chain (BB-10, 10 g/ml), or IL-2 (50 U/ml). B, The six different bar sets represent cells that have been half (48%) of the T cell clones stimulated with CD3 and CD58 preactivated as described in A, kept in culture for 10 days, washed, and displayed the cytokine profile of Tr1 T cell clones with high IL-10, restimulated at 106 cells/ml with anti-CD3 mAb (1 ␮g/ml) and irradiated ␥ intermediate IFN- , and low to no IL-4 secretion, whereas T cell PBMCs and either medium alone (Ⅺ, barely seen), cross-linked anti-CD3 clones stimulated with CD3 and CD80 displayed a cytokine profile mAb (u), cross-linked anti-CD3 and anti-CD28 mAb (o), or cross-linked of Th0 or Th1 T cell clones with high to moderate IFN-␥ and some anti-CD3 and IL-2 (50 U/ml) (f). All measurements were done in tripli- IL-4 secretion (Fig. 3). Addition of blocking anti-IL-10 mAb did cate. One representative experiment of two is shown. The Journal of Immunology 3111

FIGURE 3. CD2 costimulation induced the differentiation of Tr1 T cell clones. Purified hu- man CD4ϩ T cells were stimulated with anti- CD3 mAb (10 ␮g/ml) in the presence of L cells expressing CD32 and CD80 or CD58 in the pres- ence or absence of anti-IL-10 mAb (9D7, 10 ␮g/ ml). After 7 days, cells were collected, labeled with anti-CD4-FITC and anti-CD25-PE, and cloned at 1 cell/well by FACS sorting. T cell clones were expanded by repetitive stimulations in the presence of PHA (0.1 ␮g/ml), feeder cells, and rIL-2 (50 U/ml). After expansion, 59 clones in each condition were stimulated with anti-CD3 and anti-CD28 mAbs and the supernatant was collected after 48 h. IL-10, IFN-␥, and IL-4 was determined by ELISA. Results (F) represent mean of duplicate measurements for each clone. Quadrants have been set to discriminate between Th1-, Th2-, and Tr1-type clones. Values have been set accordingly: IFN-␥ (10,000 ng/ml), Downloaded from IL-10 (1,000 ng/ml), and IL-4 (10 ng/ml). Per- centage of clones in each quadrant are shown.

experiments with resting T cells in a transwell system were per- cells (3). We also demonstrated that IL-10 induces anergy in both ϩ ϩ formed using two different Tr1 T cell clones obtained after differ- CD4 (13) and CD8 T cells (7). We further dissected the func- http://www.jimmunol.org/ entiation with L-CD58 cells. To this end, syngeneic naive CD4ϩ T tional role of IL-10 in inducing anergy and regulatory T cell dif- cells were stimulated with CD3 and CD28 mAbs in the presence of ferentiation. Although it has been shown that IL-10 has immuno- Tr1, Th0, or Th1 T cell clones placed in the top basket. As shown suppressive activities directly on purified T cells (17), preliminary in Fig. 4, Tr1 T cell clones suppressed the proliferation of by- experiments have shown that no regulatory T cells were obtained stander resting T cells in response to CD3 and CD28 mAb stim- after stimulation of purified CD4ϩ T cells with CD3 and CD28 ulation. Addition of a combination of anti-IL-10 and anti-TGF-␤ mAbs in the presence of IL-10. These results suggest that IL-10 mAbs partially restored the proliferation of the bystander T cells as primarily acts by modifying APCs. Indeed, IL-10 displays immu- previously described (3). In contrast, coculture experiments with noregulatory functions mainly through the down-regulation of co- Th0 or Th1 T cell clones induced the proliferation of the naive T stimulatory molecules. We (7) and others (18Ð20) have shown that by guest on September 29, 2021 cells as expected (Fig. 4). IL-10 addition on activated human decreases MHC class I and class II expression as well as the expression of co- Ag-specific differentiation of Tr1 T cell clones stimulatory molecules like CD80, CD86, or CD54 (5, 21, 22). To determine whether this protocol could be used to differentiate in vitro Ag-specific Tr1 T cell clones, we generated L cells ex- pressing the two chains of HLA-DR1 along with CD58. Purified CD4ϩ T cells from a non-DR1 donor were stimulated with L cells expressing HLA-DR1 in the presence of CD58 alone or with CD80. As a control, T cells were also stimulated with an HLA- DR1-positive EBV- line. After 7 days, CD4ϩCD25ϩ T cells were cloned as described above. The several Ag-specific T cell clones obtained after stimulation with HLA-DR and CD58 dis- played a cytokine profile of Tr1 T cell clones with high IL-10 secretion. In contrast, all of the T cell clones obtained after stim- ulation with the HLA-DR1ϩ EBV-B cell line or L cells expressing CD80 (data not shown) displayed a cytokine profile of the Th0 or Th1 T cell clones (Table II). FIGURE 4. Tr1 T cell clones inhibited the proliferative response of na- ive CD4ϩ T cells. Purified human CD4ϩ T cells were stimulated with Discussion anti-CD3 (10 ␮g/ml) and anti-CD28 (1 ␮g/ml) mAbs for 3 days in the This study demonstrates that CD2-CD58 interaction induces the bottom compartment of a transwell system. The resting T cells were cocul- differentiation of nonproliferating regulatory T cells secreting high tured with various CD3ϩCD28-activated T cell clones contained in the levels of IL-10 (Tr1). This function is specific for CD2 signaling upper transwell compartment; a Th0 T cell clone (L4) previously differ- as it is not obtained and even inhibited by mobilization of other entiated with L cells alone, two Tr1 T cell clones (58-7, 58-12) differen- costimulatory molecules like CD28 or LFA-1. tiated with L-CD58 cells, and a Th1 T cell clone (80-5) differentiated with L-CD80 cells. After 3 days, the basket was removed and the proliferative We had previously shown that repetitive stimulations of human ϩ response of the CD4 T cells was analyzed by [3H]thymidine incorpora- or mouse T cells in the presence of IL-10 led to the differentiation ϩ tion during the last 12 h of culture. Cultures of human T cells were con- of a regulatory subset of CD4 T cells secreting high levels of ducted in the presence of an isotype-matched control at 20 ␮g/ml (Ⅺ); a IL-10 (3). These Ag-specific Tr1 cells suppress the proliferation of blocking anti-IL-10 mAb at 5 ␮g/ml (9D7, u), a blocking anti-TGF-␤ at 10 ϩ CD4 T cells in response to Ag and prevent experimental colitis ␮g/ml (from Genzyme, Cambridge, MA; o), or a combination of the two ϩ induced in SCID mice by pathogenic CD4 CD45RBhigh splenic T Abs (f). Results represent one of three experiments performed. 3112 ROLE OF CD2 IN Tr1 CELL DIFFERENTIATION

Table II. Differentiation of Ag-specific Tr1 cells after stimulation with HLA-DR/CD58 L cellsa

Donor 1 Donor 2

Stimulation IL-10 (pg/ml) IFN-␥ (pg/ml) IL-10 (pg/ml) IFN-␥ (pg/ml)

DR1-L cells 326 Ϯ 40 13,372 Ϯ 1,500 3,657 Ϯ 324 9,638 Ϯ 63 DR1-CD58L cells 2,315 Ϯ 183 19,403 Ϯ 975 6,402 Ϯ 272 914 Ϯ 322 DR1-B cell line Ͻ40 16,005 Ϯ 123 174 Ϯ 59 4,930 Ϯ 685

a Purified CD4ϩ T cells from two different DR1Ϫ donors were stimulated for 7 days with L cells expressing HLA-DR1 either alone or in combination with CD58. As a control, CD4ϩ T cells were stimulated with DR1ϩ transformed B cells. After 7 days, cells were washed and labeled with CD4-PE and CD25-FITC and cloned at one cell per well using a FACS sorter. T cell clones were expanded as described above and 10 T cell clones in each group were stimulated with DR1ϩ transformed B cells. Cytokine levels were measured by ELISA in supernatants collected 48 h after stimulation. Results represent mean Ϯ SEM for 10 clones in each group of one representative experiment of four.

However, expression of CD58 (LFA-3) was not affected by IL-10 Whereas ligation of the CD2R epitope is linked to the delivery of addition (7). Using artificial APCs expressing selective costimu- activation signals, perturbation of other regions of CD2 can inhibit latory molecules, we mimicked the effect of IL-10 on APCs. We T cells, apparently through a mechanism more complex than sim- demonstrated that stimulation of T cells using APCs expressing ply the disruption of the adhesion function of CD2 (30, 38). For Downloaded from only the CD58 molecule, in the absence of CD80, CD86, and example, T cells obtained from mice treated with a nondepleting CD54, induced the differentiation of regulatory T cells able to CD2 mAb display a markedly reduced proliferative response to suppress the proliferation of bystander T cells through the secre- various stimuli (30). This polyclonal T cell unresponsiveness per- tion of IL-10 and TGF-␤. This mechanism has been indirectly sists long after the full recovery of CD2 cell surface expression and confirmed in two recent manuscripts (23, 24) where the authors the apparent clearance of the mAb and thus cannot be due to the demonstrated that blockade of CD28/CD80-CD86 and CD40- disruption of the interaction of CD2 with its ligands (30). More- http://www.jimmunol.org/ CD154 interactions during primary allogeneic stimulation resulted over, the use of CD2 mAb has been shown to induce long-term in the differentiation of alloantigen-specific regulatory T cells se- tolerance in both mouse and rat models of transplantation. These creting high levels of IL-10 and maintaining anergy. There have observations could be explained by the specific differentiation of been other studies showing that the combination of certain co- regulatory T cells induced by the specific mobilization of the CD2 stimulatory molecules can induce selective cytokine patterns. In molecule. However, in a completely different system using aner- agreement with our results, it has been reported by Parra et al. (25) gized T cell clones, Boussiotis et al. (39) have shown that CD2 that costimulation with CD80, on T cells stimulated with superan- costimulation restored responsiveness to TCR engagement. This tigen (staphylococcal enterotoxin A), resulted in a vigorous re- discrepancy could be explained by the fact that we analyzed on sponse with production of high levels of IL-2 and IFN-␥ and pro- naive T cells the importance of CD2 costimulation on the differ- by guest on September 29, 2021 longed proliferation. In contrast, transient proliferation and low entiation of T cells toward a Tr1 phenotype, whereas in that study levels of IL-2 were seen after CD58 costimulation. CD54 costimu- the role of CD2 costimulation was analyzed on already anergized lation was characterized by a high proliferative response and high fully differentiated T cells. Therefore, it cannot be excluded that levels of IL-2 secretion, whereas secretion of IFN-␥ was not in- CD2 costimulation in the absence of other costimulatory signals duced (16). Moreover, in recent experiments (26), distinct IL-10 has an important role in the differentiation of Tr1 whereas in con- and TNF-␣ profiles depending on either ICAM-1 (CD54), junction with other costimulatory signals it can induce the reversal ICAM-2, or ICAM-3 were used to costimulate human T cells with of T cell anergy. anti-CD3 mAb were observed (26). ICAM-1 costimulation induces Interestingly, differentiation of Tr1 cells induced by CD2 co- higher IL-10 and lower TNF-␣ secretion as compared with stimulation did not require multiple stimulations, as both high ICAM-2 and ICAM-3 (26). IL-10 secretion and regulatory function were observed after the Various roles for CD2 in T cell activation have been proposed, first stimulation of naive T cells with anti-CD3 and CD32-CD58-L including function as an adhesion molecule (15, 27), thereby re- cells. This is in contrast with previous studies showing that the ducing amounts of Ag required for T cell activation (15), as a differentiation of Tr1 cells needs repetitive stimulations by imma- costimulatory molecule (28), or as a direct promoter of T cell ture dendritic cells (40) or IL-10 (3). This discrepancy further sug- activation (29). Moreover, CD2 has been implicated in the induc- gests that Tr1 differentiation is primarily dependent on the stimu- tion of T cell anergy (30) and has been reported to modulate cy- lation of naive T cells with TCR and CD2 mobilization in the tokine production by T cells (31, 32) and to regulate positive se- absence of other costimulatory signals. lection (33). Surprisingly, however, CD2-deficient mice did not Finally, the ability to rapidly differentiate in vitro Ag-specific show an obvious phenotype and could efficiently cope with viral regulatory T cells opens new therapeutic perspectives for the use infections (34), undermining the view that CD2 plays a major role of Tr1 in autoimmune/inflammatory diseases and allogeneic trans- in T cell activation. When stimulated by appropriate mAbs, CD2 plantation. Indeed, in vitro pulsing of modified APCs with self or elicits proliferative responses comparable to those of CD3 mAb alloantigens followed by in vivo injection of the T cell populations and mitogenic lectins (35, 36). A distinctive feature of CD2 mAb- could lead to down-regulation of self/alloreactivity mediated by mediated activation of T cells is the requirements that pairs of both Th1 (3) or Th2 T cells (4). mAbs be used (37). For human CD2, one of the mAbs must be directed against the CD2R epitope, which is poorly expressed by resting T cells but is induced following binding to CD2 of the References second Ab of the pair (35). None of the known physiologic CD2 1. Bachmann, M. F., K. McKall-Faienza, R. Schmits, D. Bouchard, J. Beach, D. E. Speiser, T. W. Mak, and P. S. Ohashi. 1997. Distinct roles for LFA-1 and ligands alone delivers an activation signal, but the combination of CD28 during activation of naive T cells: adhesion versus co-stimulation. Immu- anti-CD2R mAb and CD58 induces T cell proliferation (35). nity 7:549. The Journal of Immunology 3113

2. Viola, A., and A. Lanzavecchia. 1996. T cell activation determined by T cell 22. Chang, C. H., M. Furue, and K. Tamaki. 1995. -1 expression of Langerhans number and tunable thresholds. Science 273:104. cells is up-regulated by proinflammatory cytokines, and is down-regulated by 3. Groux, H., M. Bigler, A. O’Garra, M. Rouleau, S. Antonenko, J. E. de Vries, and interferon-␥ or by -10. Eur. J. Immunol. 25:394. M.-G. Roncarolo. 1997. A CD4ϩ T-cell subset inhibits -specific T cell 23. Koenen, H. J., and I. Joosten. 2000. Blockade of CD86 and CD40 induces al- responses and prevents colitis. Nature 389:737. loantigen-specific immunoregulatory T cells that remain anergic even after re- 4. Cottrez, F., S. D. Hurst, R. L. Coffman, and H. Groux. 2000. T regulatory cells versal of hyporesponsiveness. Blood 95:3153. 1 inhibit a Th2-specific response in vivo. J. Immunol. 165:4848. 24. Van Gool, S. W., J. Vermeiren, K. Rafiq, K. Lorr, M. de Boer, and 5. Willems, F., A. Marchant, J. P. Delville, C. Gerard, A. Delvaux, T. Velu, J. L. Ceuppens. 1999. Blocking CD40-CD154 and CD80/CD86-CD28 interac- M. de Boer, and M. Goldman. 1994. Interleukin-10 inhibits B7 and intercellular tions during primary allogeneic stimulation results in T cell anergy and high adhesion molecule-1 expression on human monocytes. Eur. J. Immunol. 24:1007. IL-10 production. Eur. J. Immunol. 29:2367. 25. Parra, E., A. G. Wingren, G. Hedlund, M. Bjorklund, H. O. Sjogren, T. Kalland, 6. Steinbrink, K., M. Wolfl, H. Jonuleit, J. Knop, and A. H. Enk. 1997. Induction of ϩ tolerance by IL-10-treated dendritic cells. J. Immunol. 159:4772. D. Sansom, and M. Dohlsten. 1994. Costimulation of human CD4 T lympho- 7. Groux, H., M. Bigler, J. E. de Vries, and M. G. Roncarolo. 1998. Inhibitory and cytes with B7 and function-associated antigen-3 results in distinct stimulatory effects of IL-10 on human CD8ϩ T cells. J. Immunol. 160:3188. cell activation profiles. J. Immunol. 153:2479. 8. Selvaraj, P., M. L. Plunkett, M. Dustin, M. E. Sanders, S. Shaw, and 26. Bleijs, D. A., R. de Waal-Malefyt, C. G. Figdor, and Y. van Kooyk. 1999. Co- ␣ T. A. Springer. 1987. The T lymphocyte CD2 binds the cell surface stimulation of T cells results in distinct IL-10 and TNF- cytokine profiles de- ligand LFA-3. Nature 326:400. pendent on binding to ICAM-1, ICAM-2 or ICAM-3. Eur. J. Immunol. 29:2248. 9. Davis, S. J., and P. A. van der Merwe. 1996. The structure and ligand interactions 27. Springer, T. A., M. L. Dustin, T. K. Kishimoto, and S. D. Marlin. 1987. The of CD2: implications for T-cell function. Immunol. Today 17:177. lymphocyte function-associated LFA-1, CD2, and LFA-3 molecules: cell adhe- 10. Punch, J. D., T. Tono, L. Qin, D. K. Bishop, and J. S. Bromberg. 1998. Tolerance sion receptors of the immune system. Annu. Rev. Immunol. 5:223. induction by anti-CD2 plus anti-CD3 monoclonal : evidence for an 28. Bierer, B. E., A. Peterson, J. C. Gorga, S. H. Herrmann, and S. J. Burakoff. 1988. IL-4 requirement. J. Immunol. 161:1156. Synergistic T cell activation via the physiological ligands for CD2 and the T cell receptor. J. Exp. Med. 168:1145. 11. Deckert, M., J. Kubar, D. Zoccola, G. Bernard-Pomier, P. Angelisova, V. Horejsi, 29. Ohno, H., C. Ushiyama, M. Taniguchi, R. N. Germain, and T. Saito. 1991. CD2 and A. Bernard. 1992. CD59 molecule: a second ligand for CD2 in T cell adhe- can mediate TCR/CD3-independent T cell activation. J. Immunol. 146:3742. sion. Eur. J. Immunol. 22:2943. Downloaded from 30. Guckel, B., C. Berek, M. Lutz, P. Altevogt, V. Schirrmacher, and B. A. Kyewski. 12. Groux, H., S. Huet, H. Valentin, D. Pham, and A. Bernard. 1989. Suppressor ϩ 1991. Anti-CD2 antibodies induce T cell unresponsiveness in vivo. J. Exp. Med. effects and cyclic AMP accumulation by the CD29 molecule of CD4 lympho- 174:957. cytes. Nature 339:152. 31. Valentin, H., H. Groux, C. Gelin, I. Chretien, and A. Bernard. 1990. Modulation 13. Groux, H., M. Bigler, J. de Vries, and M.-G. Roncarolo. 1996. Interleukin-10 ϩ of lymphokine release and cytolytic activities by activating peripheral blood lym- induces a long-term antigen-specific anergic state in human CD4 T cells. J. Exp. phocytes via CD2. J. Immunol. 144:875. Med. 184:19. 32. Gollob, J. A., J. Li, E. L. Reinherz, and J. Ritz. 1995. CD2 regulates responsive- 14. Lenschow, D. J., T. L. Walunas, and J. A. Bluestone. 1996. CD28/B7 system of ness of activated T cells to . J. Exp. Med. 182:721. T cell costimulation. Annu. Rev. Immunol. 14:259. 33. Teh, S. J., N. Killeen, A. Tarakhovsky, D. R. Littman, and H. S. Teh. 1997. CD2 http://www.jimmunol.org/ 15. Koyasu, S., T. Lawton, D. Novick, M. A. Recny, R. F. Siliciano, B. P. Wallner, regulates the positive selection and function of antigen-specific CD4ϪCD8ϩ T and E. L. Reinherz. 1990. Role of interaction of CD2 molecules with lymphocyte cells. Blood 89:1308. function- associated antigen 3 in T-cell recognition of nominal antigen. Proc. 34. Killeen, N., S. G. Stuart, and D. R. Littman. 1992. Development and function of Natl. Acad. Sci. USA 87:2603. T cells in mice with a disrupted CD2 gene. EMBO J. 11:4329. 16. Labuda, T., J. Wendt, G. Hedlund, and M. Dohlsten. 1998. ICAM-1 co-stimu- 35. Meuer, S. C., R. E. Hussey, M. Fabbi, D. Fox, O. Acuto, K. A. Fitzgerald, lation induces IL-2 but inhibits IL-10 production in superantigen-activated human J. C. Hodgdon, J. P. Protentis, S. F. Schlossman, and E. L. Reinherz. 1984. An ϩ CD4 T cells. Immunology 94:496. alternative pathway of T-cell activation: a functional role for the 50 kd T11 sheep 17. de Waal Malefyt, R., H. Yssel, and J. E. de Vries. 1993. Direct effects of IL-10 ϩ erythrocyte receptor . Cell 36:897. on subsets of human CD4 T cell clones and resting T cells: specific inhibition 36. Clark, S. J., D. A. Law, D. J. Paterson, M. Puklavec, and A. F. Williams. 1988. of IL-2 production and proliferation. J. Immunol. 150:4754. Activation of rat T lymphocytes by anti-CD2 monoclonal antibodies. J. Exp. 18. de Waal Malefyt, R., J. Haanen, H. Yssel, M.-G. Roncarolo, A. te Velde, Med. 167:1861.

C. Figdor, K. Johnson, R. Kastelein, H. Spits, and J. E. de Vries. 1991. IL-10 and 37. Huet, S., H. Wakasugi, G. Sterkers, J. Gilmour, T. Tursz, L. Boumsell, and by guest on September 29, 2021 v-IL-10 strongly reduce antigen specific human T cell responses by diminishing A. Bernard. 1986. T cell activation via CD2 [T, gp50]: the role of accessory cells the antigen presenting capacity of monocytes via down-regulation of class II in activating resting T cells via CD2. J. Immunol. 137:1420. MHC expression. J. Exp. Med. 174:915. 38. Miller, G. T., P. S. Hochman, W. Meier, R. Tizard, S. A. Bixler, M. D. Rosa, and 19. Peguet-Navarro, J., C. Moulon, C. Caux, C. Dalbiez-Gauthier, J. Banchereau, and B. P. Wallner. 1993. Specific interaction of lymphocyte function-associated an- D. Schmitt. 1994. Interleukin-10 inhibits the primary allogeneic T cell response tigen 3 with CD2 can inhibit T cell responses. J. Exp. Med. 178:211. to human epidermal Langerhans cells. Eur. J. Immunol. 24:884. 39. Boussiotis, V. A., G. J. Freeman, J. D. Griffin, G. S. Gray, J. G. Gribben, and 20. Caux, C., C. Massacrier, B. Vanbervliet, C. Barthelemy, Y.-J. Liu, and L. M. Nadler. 1994. CD2 is involved in maintenance and reversal of human J. Banchereau. 1994. Interleukin 10 inhibits T cell alloreaction induced by human alloantigen-specific clonal anergy. J. Exp. Med. 180:1665. dendritic cells. Int. Immunol. 6:1177. 40. Jonuleit, H., E. Schmitt, G. Schuler, J. Knop, and A. H. Enk. 2000. Induction of 21. Ding, L., P. S. Linsley, L. Y. Huang, R. N. Germain, and E. M. Shevach. 1993. interleukin 10-producing, nonproliferating CD4ϩ T cells with regulatory prop- IL-10 inhibits costimulatory activity by selectively inhibiting the erties by repetitive stimulation with allogeneic immature human dendritic cells. up-regulation of B7 expression. J. Immunol. 151:1224. J. Exp. Med. 192:1213.