Activation Resting Memory Versus Naive CD4 T Cell Differential Role

Home , Activation, Memory T cell, Naive T cell

Differential Role of CTLA-4 in Regulation of Resting Memory Versus Naive CD4 T Cell Activation

This information is current as D. P. Metz, D. L. Farber, T. Taylor and K. Bottomly of September 29, 2021. J Immunol 1998; 161:5855-5861; ; http://www.jimmunol.org/content/161/11/5855 Downloaded from References This article cites 42 articles, 22 of which you can access for free at: http://www.jimmunol.org/content/161/11/5855.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 © 1998 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. Differential Role of CTLA-4 in Regulation of Resting Memory Versus Naive CD4 T Cell Activation1

D. P. Metz,* D. L. Farber,† T. Taylor,* and K. Bottomly2*

Regulation of peripheral T cell responses is critical for preserving self tolerance. Memory T cells have a lower threshold for activation through the TCR and are thought to be less dependent on costimulation than naive T cells, suggesting a requirement for more stringent regulation of memory T cells. We have recently shown that CD4 engagement apart from the TCR results in the inactivation of memory, but not naive, CD4 T cells. We show here that this inhibition requires ligation of CTLA-4, in that blocking CTLA-4-B7 interactions restores memory CD4 T cell responsiveness. Early signaling through CTLA-4 is possible because resting memory, but not naive, CD4 T cells contain intracellular stores of CTLA-4 that are continuously recycled between the cytoplasm and the cell surface. This mechanism ensures that low intensity TCR engagements, which are thought to be important for peripheral T cell longevity, do not cause memory T cell activation but instead raise their threshold for costimulatory signals. Downloaded from This may give memory T cells an extended lifespan with a reduced risk of inappropriate activation. The Journal of Immunology, 1998, 161: 5855–5861.

nteractions of peripheral CD4 T cells with MHC class II- sponses are regulated and how proliferation due to recognition of positive APCs in the absence of exogenous Ags are required self peptide is prevented.

I for long term T cell survival (1, 2). To prevent proliferation While it is established that naive T cell responses to self are http://www.jimmunol.org/ to self peptides, activation of resting T cells has to be stringently controlled by their requirement for costimulation, the role of co- regulated. For naive CD4 T cells, tolerance to self is assured by stimulation in regulating memory T cell responses is still unclear. their need for two signals to proliferate, one via contact of Ag/ It has been shown that unlike naive T cells, resting, Ag-primed MHC with the TCR and the second via contact of B7 molecules CD4 T cells respond to Ag presented by resting B cells that do not expressed on APCs with a costimulatory molecule, such as CD28 express costimulatory molecules (14). Furthermore, in vitro stim- (3, 4). The need for costimulation ensures that naive T cells be- ulated memory CD4 T cells, but not naive T cells, proliferate to come activated only when APCs are loaded with exogenous Ag be- plate-bound anti-CD3 in the absence of additional costimulatory cause resting APCs express few or no B7 molecules (5). In the ab- molecules (15). Several in vivo studies using CTLA4-Ig to block sence of costimulation, naive CD4 T cells become unresponsive (6). costimulation delivered through B7 also indicate that memory T by guest on September 29, 2021 Memory CD4 T cells develop during a primary infection and, cells are costimulation independent (16, 17), while other studies like naive T cells, are resting, long-lived lymphocytes. However, suggest that a secondary response can be blocked by CTLA4-Ig in unlike naive CD4 T cells, memory T cells are previously primed vivo (18). It is possible that the discrepancies in these studies re- cells and have been reported to be more sensitive to triggering sulted from the inability to deliver CTLA4-Ig efﬁciently in the through the TCR than naive T cells (7, 8), to respond more vig- former studies or that the requirement of memory CD4 T cells for orously to Ag in terms of proliferation (9) and cytokine production costimulation depends on the circumstances of activation. (10), and to exhibit altered coupling of proximal signaling events CTLA-4, a CD28 homologue that negatively regulates the pro- to the TCR (11). Furthermore, while naive T cells recirculate be- liferation of effector T cells (19, 20), has recently been shown to tween blood and secondary lymphoid organs, memory CD4 T cells prevent costimulation through CD28 when APCs express limiting are capable of entering nonlymphoid tissue directly from the blood B7 molecules. The importance of CTLA-4 in regulating effector T (12, 13). The altered recirculation pattern enhances the chance of cells is emphasized by the massive lymphoproliferation observed contacting foreign Ag, but also exposes memory T cells to new self in mice that lack CTLA-4 (21, 22). A role for CTLA-4 in the peptides for which they may not have been selected in the thymus. regulation of memory CD4 T cell responses in mice has recently While the altered signaling and recirculation pathways are be- been suggested (23) but has not yet been demonstrated. lieved to contribute to a more vigorous secondary immune re- We have recently shown that CD4 ligation in memory, but not sponse, they also pose the question of how memory T cell re- naive, CD4 T cells leads to unresponsiveness (24, 25). In these studies memory CD4 T cells failed to proliferate to soluble anti- CD3 presented by syngeneic MHC class IIϩ APCs. The response

*Immunobiology Section, Yale University School of Medicine, New Haven, CT was restored when anti-CD3 was presented by APCs that did not 06510; †Department of Microbiology, University of Maryland, College Park, MD express MHC class II or expressed a mutant MHC class II mole- 20910 cule unable to engage CD4. We hypothesized that memory, but not Received for publication May 5, 1998. Accepted for publication July 23, 1998. naive, CD4 T cells fail to proliferate under conditions that do not The costs of publication of this article were defrayed in part by the payment of page recruit CD4 into the TCR complex, but instead allow independent charges. This article must therefore be hereby marked advertisement in accordance CD4-MHC class II interactions, which resemble the inhibition pro- with 18 U.S.C. Section 1734 solely to indicate this fact. duced in response to Ab-mediated CD4 cross-linking. Such a 1 This work was supported by National Institutes of Health Grant CA38350 and the Howard Hughes Medical Institute. mechanism may be important in preventing memory CD4 T cells 2 Address correspondence and reprint requests to Dr. K. Bottomly, 310 Cedar St., LH from proliferating to self peptides that fail to recruit CD4 to the 405, New Haven, CT 06510. TCR interaction site.

To determine the role of costimulation in activation of resting APC preparation memory CD4 T cells, we investigated the role of CD28/CTLA-4 in Spleen cells from BALB/c and RHA␤o/o mice were depleted of T cells with memory T cell responses. We show here that both naive and mem- complement after incubation with anti-Thy.1, anti-CD8, and anti-CD4 ory CD4 T cell proliferation in response to soluble anti-CD3 re- mAbs, and were treated with mitomycin (50 ␮g/ml; Boehringer Mann- quires costimulation through CD28. We further show that under heim, Indianapolis, IN) for 30 min as previously described (15). Resultant Ͼ conditions where CD4-MHC class II interactions occur, memory APC preparations were 95% pure as determined by FACS analysis of Fc␥RII-stained cells. In experiments in which fibroblasts were used as CD4 T cells have a higher threshold for costimulation through APCs, mitomycin treatment (75 ␮g/ml) was conducted for 90 min. CD28 than do naive T cells. Furthermore, the inability of memory CD4 T cells to proliferate in the presence of CD4-MHC class II Proliferation assays interactions is dependent on CTLA-4-B7 interaction early after CD45RBhigh and CD45RBlow CD4 T cells were cultured at 25,000 cells/ activation. CTLA-4 functions uniquely in memory CD4 T cells in well in 96-well flat-bottom tissue culture plates (Costar, Cambridge, MA) that it continuously recycles from intracellular stores to the cell with soluble anti-CD3 (10 ␮g/ml final concentration) and 50,000 splenic surface. We propose that the ability of CTLA-4 signaling to inhibit APCs or 25,000 fibroblasts/well in Eagles’ high amino acid medium sup- plemented with 5% FCS, 50 U/ml penicillin, 50 ␮g/ml streptomycin sul- memory CD4 T cell responses is dependent on CD4 ligation. This fate, and 10 mM HEPES. Titration of APC numbers and the concentration mechanism ensures that low intensity TCR engagement, which of anti-CD3 have been published previously (15, 24). Where indicated, mediates CD4-MHC class II engagement away from the TCR, does anti-CTLA-4 (5 ␮g/ml), anti-CTLA-4 Fab fragments (50 ␮g/ml), anti- not cause memory T cell activation, but instead raises their threshold CD28 (10 ␮g/ml), and/or CTLA4-Ig (8 ␮g/ml) were added to the cultures. for costimulatory signals. This may extend the life span of memory To immobilize anti-CD3 and anti-CD28, the mAbs were diluted to 10 ␮g/ml in 50 mM Tris-HCl (pH 9.5), resuspended at 100 ␮l/well in 96-well Downloaded from CD4 T cells with reduced risk of inappropriate activation. culture plates, and incubated for 2–4 h at 37°C. The wells were washed twice in medium before culturing 25,000 T cells/well. For assessment of proliferation, cultures were pulsed with 1 ␮Ci of [3H]dTdR (6.7 Ci/mmol) Materials and Methods after 72 h of incubation at 37°C. Pulsed plates were harvested 18–24 h later Mice on a Tomtec 96-well plate harvester (Wallac, Gaithersburg, MD). Radio- activity was quantitated by scintillation counting, and results are presented Female BALB/c ByJ mice were obtained from The Jackson Laboratory as the mean of triplicates Ϯ SEM.

(Bar Harbor, ME) and were maintained at the animal facility at Yale Uni- http://www.jimmunol.org/ versity (New Haven, CT). All mice were used at 8–12 wk of age. MHC CTLA-4 immunoblots class II-negative mice were originally obtained from Dr. D. Mathis (Institut de Chimie Biologique, Strasbourg, France) (26) and were bred and main- Naive (CD45RBhigh) and memory (CD45RBlow) CD4 T cells were pre- tained at the Yale animal facility. pared as described above. As a positive control for CTLA-4 expression, naive and memory CD4 T cells were activated with plate-bound anti-CD3 Cell lines and anti-CD28 as described above for 48 h. Cells were lysed in lysis buffer (20 mM HCl (pH 7.5), 150 mM NaCl, 1 mM MgCl2, 1 mM EGTA, con- DAP.3 ﬁbroblasts and DAP.3 transfected with I-Ad (line 44) were gifts taining 10 ␮g/ml leupeptin, 10 ␮g/ml aprotinin, 10 mM PMSF, 50 mM from Dr. R. Germain (27). The ﬁbroblasts were cotransfected with Fc␥RII NaF, 10 mM sodium pyrophosphate, and 1 mM sodium orthovanadate) as by Dr. D. L. Farber as recently described (24). recently described (24). Protein lysates were resolved on a 12% SDS gra-

dient gel, transferred electrophoretically to nitrocellulose, and hybridized by guest on September 29, 2021 Antibodies to anti-CTLA-4 Ab C19 (Santa Cruz Biotechnology) followed by horseradish peroxidase-coupled anti-goat secondary Ab (Santa Cruz Biotech- The following mAbs were purified from supernatants of hybridomas main- nology). Bands were revealed by chemiluminescence using the enhanced tained in this laboratory using standard protein A or protein G affinity chemiluminescence Western blotting detection reagents (Amersham, Ar- chromatography: anti-CD3⑀ (rat IgG2c, C363.29B), anti-CD4 (rat IgG2b, lington Heights, IL) according to the manufacturer’s instructions. GK1.5), anti-CD8 (rat IgG2b, 2.43), anti-CD8 (rat IgG2a, 53-6.72), anti- Thy-1 (rat IgG2c, Y19), anti-I-Ad (mouse IgG2a, 212.A1), and anti-murine CTLA-4 staining Fc␥RII (rat IgG2b, 2.4G2). Purified as well as labeled mAbs to anti- For the detection of intracellular CTLA-4 on resting CD45RB separated CD45RB (rat IgG2a, C353.16A), CTLA-4 (Armenian hamster IgG, UC10- naive and memory CD4 T cells, an intracellular staining kit from Caltag 4F10), isotype control (pooled Armenian hamster IgG, G94-56), CD28 Laboratories (Burlingame, CA) was used. The staining protocol was con- (Syrian hamster IgG, 37.51), CD80 (rat IgG2a, 1G10), and CD86 (rat ducted according to the manufacturer’s recommendations. Cells were an- IgG2a, GL1) were obtained from PharMingen (San Diego, CA). Anti- alyzed by flow cytometry and were gated on low angle and sideways light CTLA-4 (goat polyclonal, C19) for Western blotting and secondary anti- scatter to exclude large, activated lymphocytes. For CTLA-4 expression goat IgG/horseradish peroxidase were purchased from Santa Cruz Biotechnol- upon culture, naive and memory CD4 T cells were incubated with or with- ogy (Santa Cruz, CA). Anti-CD4/Quantum red was purchased from Sigma (St. out plate-bound anti-CD3 and anti-CD28 at 37°C in the presence of either Louis, MO), and CTLA4-Ig was provided by Dr. P. S. Linsley (28). phycoerythrin-labeled anti-CTLA-4 or an isotype control mAb for 3 h. To distinguish between CTLA-4 cell surface staining and staining resulting Naive and memory CD4 T cell preparation from the internalization of CTLA-4/PE3 mAb, parallel cultures were left unstained during the culture period and were cell surface labeled with CD4-positive T cells were isolated from spleens and separated into anti-CTLA-4/PE or the isotype control for 30 min on ice after culture. Cells CD45RBhigh and CD45RBlow fractions by MACS separation as recently were analyzed by flow cytometry for the uptake of CTLA-4 mAbs. described (15). In brief, CD4-positive T cells were isolated from spleens of BALB/c ByJ mice by negative selection using mAbs to CD8, I-Ad, and Fc␥RII followed by incubation with anti-rat IgG-, anti-mouse IgG-, and Results anti-mouse IgM-coated magnetic beads (Collaborative Research, Bedford, Proliferation of memory CD4 T cells to soluble anti-CD3 bound MA). CD4 T cells were collected by magnetic depletion of Ab-labeled cells to FcRs on fibroblasts is costimulation dependent and centrifuged through a 85–62% Percoll gradient to deplete activated T cells and contaminating accessory cells. For MACS separation into Memory CD4 T cells proliferate to plate-bound anti-CD3, while CD45RBhigh and CD45RBlow fractions, CD4 T cells were stained with naive T cells fail to do so, suggesting that memory T cells prolif- biotinylated CD45RB mAb followed by incubation with streptavidin- erate in the absence of costimulation (15). However, both naive coated beads (Miltenyi Biotec, Sunnyvale, CA). CD45RBlow cells were eluted within the MACS column, while the retained CD45RBhigh fraction and memory CD4 T cells respond equally well to soluble anti-CD3 was collected outside the magnetic field. The purity of the two fractions when expressed on fibroblasts (DAP.3) transfected with Fc-␥ re- was verified by FACS analysis (FACScan, using CellQuest software, Bec- ceptor II (Fc␥RII). In this instance naive CD4 T cells are capable ton and Dickinson, Mountain View, CA) after incubation of the cells with FITC-avidin D. CD45RBlow cells were Ͼ95% CD4 positive, while CD45RBhigh cells contained 85–90% CD4-positive T cells. 3 Abbreviations used in this paper: PE, phycoerythrin; Fc␥RII, Fc-␥ receptor II. The Journal of Immunology 5857

assays were performed either in the presence of CTLA4-Ig or in the presence of fibroblasts that had been sorted for a lack of B7.1 expression. Interestingly, proliferation of naive and memory CD4 T cells was equally blocked when stimulated in the presence of B7.1ϩ fibroblasts and CTLA4-Ig (Fig. 1A) or in the presence of B7.1Ϫ fibroblasts (Fig. 1B). The importance of costimulation was further accentuated by the fact that both T cell populations proliferated when Abs to CD28 were added to cultures containing B7.1Ϫ fibroblasts. These data clearly show that naive and memory CD4 T cells are equally dependent on costimulation when anti-CD3 is expressed on FcRs of APCs.

CD28 ligation can overcome memory CD4 T cell unresponsiveness to anti-CD3 in the presence of syngeneic APC/ IIϩ Previous studies have shown that unlike naive CD4 T cells memory T cells fail to proliferate to anti-CD3 in the presence of syn- ϩ ϩ geneic MHC class II APCs (APC/II ). Hyporesponsiveness was Downloaded from associated with a lack of cytokine production (24) and an inability to enter cell cycle (25). The incapacity of memory CD4 T cells to proliferate was dependent on an interaction of CD4-MHC class II, since responses were restored when APCs were derived from MHC class II-negative mice or from ﬁbroblasts transfected with

MHC class II that was mutated at the CD4 binding sites (29). It http://www.jimmunol.org/ FIGURE 1. Memory and naive CD4 T cells fail to proliferate in response was proposed that memory T cells fail to become activated upon to soluble anti-CD3 presented by FcR-transfected fibroblasts in the absence of CD4-MHC class II interaction in situations where weak TCR stim- ϩ costimulation through CD28. A, Proliferation in the presence of B7.1 fibro- ulation occurs and CD4 is not efficiently recruited to the TCR blasts with or without soluble CTLA4-Ig and anti-CD28. B, Proliferative re- complex (25). Since the costimulation requirements of memory sponse of memory and naive CD4 T cells to soluble anti-CD3 in the presence CD4 T cells in the presence of CD4 ligation might differ, we in- or the absence of anti-CD28 presented by fibroblasts lacking B7.1 expression. Assays were conducted in triplicate, and mean values are indicated Ϯ SEM. vestigated the potential role of CD28 in this response. Naive and memory CD4 T cells were stimulated with T-depleted splenocytes from MHC class IIϩ or class IIo mice in the presence or the ab- of proliferating because costimulation is provided by the fibro- sence of anti-CD28. Fig. 2 shows that both memory and naive CD4 by guest on September 29, 2021 blasts in the form of B7.1. To determine whether memory CD4 T T cells proliferated in response to soluble anti-CD3 expressed on cells can proliferate in the absence of costimulation, proliferation APC/IIo in the presence or the absence of anti-CD28 mAb (Fig.

FIGURE 2. Unresponsiveness of memory CD4 T cells to soluble anti-CD3 expressed on syngeneic APC/IIϩ can be reversed by Ab-mediated CD28 ligation. Memory and naive CD4 T cells were stimulated with soluble anti-CD3 presented by APCIIϩ (A) or APC/IIo (B) in the presence or the absence of anti-CD28. Results represent the mean Ϯ SEM from triplicate wells. C, FACS analysis of B7.1 and B7.2 expression on resting T-depleted splenocytes isolated from MHC class IIϩ and MHC class IIo mice. 5858 REGULATION OF MEMORY CD4 T CELLS BY CTLA-4

FIGURE 3. Memory CD4 T cell responses to soluble anti-CD3 in the presence of MHC class II are restored in the presence of high levels of B7.1. Compar- ison of memory and naive CD4 T cell responses to anti-CD3 presented by MHC class IIϩ (A, upper panel) or MHC class IIo (B, upper panel) fibroblasts with low or high levels of B7.1 (A and B, lower pan- els, where shaded histograms represent negative controls). Downloaded from http://www.jimmunol.org/ 2B). In contrast, naive, but not memory, CD4 T cells prolifer- memory CD4 T cells have a greater requirement for costimulation ated in response to anti-CD3 in the presence of resting APC/IIϩ than naive T cells to overcome the inhibitory signal delivered (Fig. 2A). mAb-mediated CD28 ligation restored proliferation through CD4-MHC class II interactions. of memory CD4 T cells in the presence of APC/IIϩ and enhanced the proliferation of naive T cells (Fig. 2A). The differ- ence in the ability of APC/IIϩ and APC/IIo to induce memory Memory, but not naive, CD4 T cells express intracellular stores T cell responses was not due to differences in B7 expression, of CTLA-4 ϩ o since APC/II and APC/II expressed similarly low levels of It has recently been suggested that CTLA-4 signaling increases the by guest on September 29, 2021 B7.1 and B7.2 (Fig. 2C). Although B7 expression was increased threshold for costimulation through B7 molecules in activated T in both APC populations later in the response (data not shown), cells (20, 30). We sought to investigate the possibility that late expression of high B7 on the APC is unlikely to be involved CTLA-4 signaling affected memory CD4 T cells at the initiation of in the initiation of the T cell response. We conclude that in the a response to soluble anti-CD3. CTLA-4 has not been detected on presence of resting APC/IIϩ, but not APC/IIo, the proliferation resting T cells (31, 32). However, resting memory CD4 T cells of memory CD4 T cells in response to anti-CD3 is more de- represent previously activated cells that may retain residual ex- pendent on costimulation than is the proliferation of naive T pression of CTLA-4. We investigated this possibility by staining cells. isolated resting naive and memory CD4 T cells for CTLA-4 ex- MHC class II expression on APCs increases the threshold for pression. No cell surface expression of CTLA-4 was detected on costimulation in memory CD4 T cell responses to anti-CD3 naive or memory CD4 T cells (Fig. 4A). However, anti-CTLA-4 ϩ Western blots prepared from resting and 48-h activated naive and While B7 expression on MHC class II and class IIo APC was the memory CD4 T cell lysates showed several bands between 33–37 same, CD28-B7 interactions were only sufficient to activate mem- kDa in the activated naive and memory T cell populations that ory CD4 T cells when APCs lacked MHC class II. Yet the same were likely to represent differentially glycosylated forms of CD28-B7 interactions induced naive CD4 T cell responses when APCs expressed MHC class II. This suggested that memory CD4 CTLA-4, as has been previously described (33). These bands were T cells needed more costimulation than naive T cells when MHC also visible in resting memory, but not resting naive, CD4 T cells class II was expressed on the APC. To test this hypothesis, we (Fig. 4B), suggesting that resting memory CD4 T cells express sorted MHC class IIo and class IIϩ fibroblasts for low and high intracellular stores of CTLA-4. Despite a Percoll gradient separa- expression of B7.1 and used these sublines to stimulate naive and tion, we cannot rule out that some activated T cells contaminated memory CD4 T cells. Both populations proliferated in response to the resting memory CD4 T cell pool used for Western blotting. To anti-CD3 in the presence of class II-negative fibroblasts indepen- exclude this possibility we stained intracellularly for CTLA-4 and dent of the level of B7.1 expression (Fig. 3B). However, in the analyzed the data by flow cytometry, which allowed us to gate on presence of class II only naive CD4 T cells proliferated in response low angle and sideways light scatter to exclude large activated T to anti-CD3 at low levels of B7.1. In contrast to the responses cells. Similar to the Western blot, resting memory CD4 T cells induced at low levels of B7.1, both naive and memory CD4 T cells revealed a small amount of intracellular CTLA-4 (Fig. 4C). While responded in the presence of high B7.1 (Fig. 3A). Furthermore, the the levels indicated by staining are low, no CTLA-4 is seen in induction of unresponsiveness seen upon CD4/MHC class II in- naive CD4 T cells, and no signal is seen in memory T cells with teractions was abrogated; therefore, CD4 ligation due to MHC the isotype control Ab. The prevalence of intracellular stores of class II expression on APCs increases the threshold for costimu- CTLA-4 in memory T cells suggested that CTLA-4 may be avail- lation on memory, but not naive, CD4 T cells. We concluded that able to resting memory CD4 T cells. The Journal of Immunology 5859

between endosomal compartments and the cell surface (32, 34). To determine whether CTLA-4 is recycled in resting memory T cells or whether TCR ligation induces CTLA-4 recycling, we incubated CD4 T cells in the presence of anti-CTLA-4/PE or an isotype control/PE mAbfor3hat37°C. Cells were stimulated with plate-bound anti- CD3 and anti-CD28 or remained unstimulated. If CTLA-4 cycles between the cell surface and endosomes, then labeled anti-CTLA-4 mAbs added to the cultures will accumulate intracellularly, as has recently been shown (31). To distinguish between reaccumulated CTLA-4 inside the cells and stable cell surface expression, duplicate cultures were labeled with anti-CTLA-4/PE or the isotype control after culture for 3 h. Unlike resting naive T cells, a subpopulation of memory CD4 T cells was CTLA-4/PE positive after incubation with the mAb for 3 h, whereas no staining was observed with the isotype control mAb (Fig. 5A). Expression of CTLA-4 was enhanced when memory T cells were stimulated with plate-bound anti-CD3 and anti- CD28, while CTLA-4 expression was not observed in similarly activated naive CD4 T cells (Fig. 5B). Naive and memory CD4 T cells that were stained with anti-CTLA-4/PE after culture for 3 h did not Downloaded from stain for CTLA-4 (data not shown). The fact that memory CD4 T cells were positive for CTLA-4 when cultured in the presence of the mAb but not when anti-CTLA-4 was added after culture suggested that CTLA-4 continuously recycled between the cytoplasm and the cell FIGURE 4. CD45RB fractionated memory, but not naive, CD4 T cells surface, and that cell surface expression of CTLA-4 was unstable. express small amounts of intracellular CTLA-4. A, Cell surface staining of http://www.jimmunol.org/ naive and memory CD4 T cells with PE-labeled anti-CTLA-4 (gray, solid Blocking CTLA-4 interactions with B7 reinstates memory CD4 T line) or isotype control mAb (black, dotted line). B, Anti-CTLA-4 Western cell responses to soluble anti-CD3 in the presence of B7low blot of cell lysates from resting (lanes 1 and 2) and 48-h activated (lanes APC/IIϩ 3–6) naive (N) and memory (M) CD4 T cells. To visualize CTLA-4 in resting memory CD4 T cell population the membrane was exposed for 4 To determine how CTLA-4 may function in the activation of mem- min (lanes 1–4). Lanes 5 and 6 represent lanes 3 and 4 after 30-s exposure. ory CD4 T cells, we investigated memory CD4 T cell proliferation ϩ C, Intracellular staining of naive and memory CD4 T cells with the same in response to soluble anti-CD3 in the presence of APC/II under mAbs as in A, where anti-CTLA-4/PE-labeled cells are depicted by gray, conditions where CTLA-4-B7 interactions were blocked by Fab of solid lines, and cells labeled with the isotype control/PE mAb are presented anti-CTLA-4. Blocking CTLA-4-B7 interactions had no effect on by guest on September 29, 2021 as dotted black lines. Cell surface and intracellular stainings were analyzed the proliferation of naive CD4 T cells in response to anti-CD3 in by ﬂow cytometry, gated on small lymphocytes by forward and side scatter the presence of APC/IIo or APC/IIϩ (Fig. 6B). However, while to exclude large activated T cells. memory CD4 T cell responses in the presence of APC/IIo were also unaffected by CTLA-4 Fab fragments (Fig 6B), proliferation Resting memory CD4 T cells continuously recycle CTLA-4 in the presence of APC/IIϩ was reinstated when CTLA-4-B7 in- between the cytoplasm and the cell surface teractions were blocked by anti-CTLA-4 Fab (Fig. 6A). These data It has recently been shown that cell surface expression of CTLA-4 is suggest that at limiting amounts of B7.1, CTLA-4 signaling pre- unstable on activated T cells, and that CTLA-4 continuously recycles vents memory CD4 T cell proliferation in the presence of APC/IIϩ

FIGURE 5. In contrast to CD45RBhigh naive CD4 T cells, CD45RBlow memory T cells incorpo- rate anti-CTLA-4 mAb, but not isotype control mAb, after culture for 3 h. Memory (A) and naive (B) CD4 T cells were stained with PE-labeled anti-CTLA-4 (solid line) or an isotype control mAb (dotted line) before culture at 37°C for 3 h. CD4 T cells either remained unstimulated (top panel) or were stimulated with plate- bound anti-CD3 and anti-CD28 (bottom panel). Staining was analyzed by ﬂow cytometry. No staining was observed if T cells were stained after culture for 3 h (data not shown). 5860 REGULATION OF MEMORY CD4 T CELLS BY CTLA-4

It has been shown that the strength of the TCR signal and the need for costimulation are inversely correlated; the weaker the TCR signal, the more CD28 ligation is required (23). Early signaling events oc- curring after T cell stimulation with soluble anti-CD3 presented by APCs are identical with those seen upon antagonistic peptide stimulation (35). In this instance, CD28 ligation is probably required to enhance the strength of the TCR signal. In contrast, plate-bound anti- CD3 and soluble anti-CD3 presented by FcR-transfected CHO cells (15), which induce memory CD4 T cell proliferation in the absence of costimulation, are likely to act as agonists, leading to extensive cross- linking of the TCR, which overrides the need for costimulation in memory, but not naive, CD4 T cells. Therefore, the strength of the TCR signal is probably involved in determining whether memory CD4 T cells proliferate in the absence of costimulation. This is in contrast to naive CD4 T cells, where a strong TCR signal may lower the amount of costimulation that is required but does not allow proliferation in the absence of costimulation. The inhibitory signal delivered by CD4-MHC class II interaction to Downloaded from memory CD4 T cells (24) can be overcome by ligation of CD28 or by blocking CTLA-4-B7 interactions, suggesting that CTLA-4 signaling in memory CD4 T cells prevents costimulation through CD28 under conditions where B7 expression on the APC is limiting. Interestingly, CTLA-4 does not inhibit proliferation of memory CD4 T cells when CD4-MHC class II interactions are removed. This suggests that signaling through both CD4 and CTLA-4 is required to prevent memory http://www.jimmunol.org/ CD4 T cell proliferation. Studies of CTLA-4 recycling in memory CD4 T cells in the presence or the absence of CD4 ligation do not FIGURE 6. Memory CD4 T cell responses to soluble anti-CD3 in the suggest that CD4 influences CTLA-4 cell surface expression within presence of MHC class II can be rescued by inhibiting CTLA-4 signaling. the first few hours of activation (preliminary data, not shown); how- Memory and naive CD4 T cells were stimulated with soluble anti-CD3 ϩ ever, a small change in surface expression may have significant con- presented by B7.1low, MHC class II (A) or MHC class IIo (B) fibroblasts in the presence or the absence of anti-CTLA-4 (Fab). Data are expressed as sequences. It is also possible that CTLA-4 activity is dependent on the mean Ϯ SEM from triplicate wells. Similar results were obtained when CD4 signaling, and that in the absence of CD4 ligation, CTLA-4 whole anti-CTLA-4 mAbs were used instead of Fab (data not shown). proceeds to recirculate but fails to become activated. Although little is by guest on September 29, 2021 known about CTLA-4 function, it has been shown that the cytoplas- mic domain of CTLA-4 contains two SH2 binding domains that re- but not APC/IIo. We conclude that CD4/MHC class II interaction quire phosphorylation for CTLA-4 to signal (19). It is feasible that is essential for CTLA-4-induced unresponsiveness. CD4-associated p56lck phosphorylates CTLA-4-associated SH2 binding domains or that it is required for the activation of enzymes that Discussion associate with these domains upon phosphorylation. The alternative Activation of peripheral T cells is stringently regulated to prevent route, that CTLA-4 signaling is required for CD4 activation, seems proliferation to self. While mechanisms underlying the control of unlikely because it has been shown that CD4-associated lck is con- naive T cell activation have been elucidated using TCR-transgenic stitutively active in the absence of CTLA-4 (36); therefore, lck activity mice, little is still known about the regulation of memory T cell cannot be dependent on CTLA-4 signaling. It is, however, conceiv- activation. In the present study we investigate mechanisms that able that inhibition of proliferation induced by CTLA-4 and inhibition regulate memory CD4 T cell responses. We have previously of that induced by CD4 occur independent from each other, but sig- shown that CD4 ligation by MHC class II inhibits memory, but not naling through both is required to reach the threshold that inhibits naive, T cell proliferation in response to soluble anti-CD3, and we memory T cell proliferation. Although CTLA-4 has a 10-fold higher show here that this inhibition requires CTLA-4. We demonstrate binding affinity for B7 than CD28 (37), it is unlikely to bind all B7 that memory CD4 T cell responses to soluble anti-CD3, which has even at limiting B7 expression because CTLA-4 cell surface expres- been shown to induce a weak TCR signal in the presence of APCs sion at the peak time of expression (48 h after stimulation) is only 5% (23), require costimulation through CD28, suggesting that memory that of CD28 (38). Memory CD4 T cell proliferation may therefore CD4 T cells, like naive T cells, require a second signal for acti- occur at minimal CD28 ligation as long as inhibitory signals through vation in the presence of APCs. We further show that in the pres- CD4 are not delivered. However, this possibility seems unlikely be- ence of CD4 ligation, memory, but not naive, CD4 T cell activa- cause blocking CTLA-4-B7 interactions did not significantly increase tion has an increased requirement for costimulation. This is the proliferative response of memory CD4 T cells in the absence of supported by the fact that memory CD4 T cell responses can be MHC class II on the APC. rescued by Ab-mediated CD28 cross-linking or by high B7.1 ex- It has recently been shown that the lymphoproliferative disor- pression on the APC. Furthermore, the unresponsiveness to anti- ders seen in CTLA-4-negative mice are mediated by CD4-positive, CD3 observed in memory CD4 T cells is mediated by CTLA-4, not CD8-positive, T cells (30). While the role of CD4 in CTLA-4 which continuously recycles in memory, but not resting, naive signaling was not defined, the study suggested that CTLA-4 plays CD4 T cells. These data suggest that under conditions of weak a different role in CD4 vs CD8 T cell homeostasis. TCR stimulation and low B7 expression, CD4 ligation occurs, A role for CD4 in the maintenance of CD4 T cell homeostasis leading to CTLA-4-induced unresponsiveness. may lie in the fact that long term survival of mature peripheral The Journal of Immunology 5861

CD4 T cells is dependent on the expression of MHC class II mol- 17. Lu, P., X. Zhou, S. Chen, M. Moorman, A. Schoneveld, S. Morris, ecules but not on the presence of specific Ag (1, 35, 39). This F. D. Finkelman, P. Linsley, E. Claassen, and W. C. Gause. 1995. Requirement of CTLA-4 counter receptors for IL-4 but not IL-10 elevation during a primary implies that peripheral T cells require TCR engagement with MHC systemic in vivo immune response. J. Immunol. 154:1078. class II loaded with self peptide. Given that memory T cells have 18. Keane-Myers, A., W. C. Gause, P. S. Linsley, S.-J. Chen, and M. Wills-Karp. a lower threshold for proliferation than naive T cells (40), there is 1997. B7-CD28/CTLA-4 costimulatory pathways are required for the develop- ment of T helper cell 2-mediated allergic airway responses to inhaled antigen. a danger that memory CD4 T cells might become activated by the J. Immunol. 158:2042. signals for their survival. It is likely that weakly cross-reactive self 19. Waterhouse, P., L. E. M. Marengere, H.-W. Mittruecker, and T. W. Mak. 1996. peptide, similar to antagonistic peptide or soluble anti-CD3, does CTLA-4, a negative regulator of T-lymphocyte activation. Immunol. Rev. 153: 183. not recruit CD4 into the TCR-MHC class II interaction site. CD4 20. Thompson, C. B., and J. P. Allison. 1997. The emerging role of CTLA-4 as an engagement away from the TCR complex may therefore induce immune attenuator. Immunity 7:445. CTLA-4-mediated unresponsiveness to prevent memory CD4 T 21. Waterhouse, P., J. M. Penninger, E. Timms, A. Wakeham, A. Shahinian, K. P. Lee, cell proliferation to self peptide but allow memory T cell survival. C. B. Thompson, H. Griesser, and T. W. Mak. 1995. Lymphoproliferative disorders with early lethality in mice deficient in CTLA-4. Science 270:985. Furthermore, this model may have implications for HIV-in- 22. Tivol, E. A., F. Borriello, A. N. Schweitzer, W. P. Lynch, J. A. Bluestone, and fected memory T cells, where CD4 is similarly ligated outside the A. H. Sharpe. 1995. Loss of CTLA-4 leads to massive lymphoproliferation and TCR complex by gp120. It has been shown that hyporesponsive fatal multiorgan tissue destruction, revealing a critical negative regulatory role of gp120-positive CD4 T cells can be induced to proliferate when CTLA-4. Immunity 3:541. 23. Chambers, C. A., M. F. Krummel, B. Boitel, A. Hurwitz, T. J. Sullivan, CD28 is ligated by mAb (41, 42). It would be of interest to see S. Fournier, D. Cassell, M. Brunner, and J. P. Allison. 1996. The role of CTLA-4 whether such T cells express CTLA-4. in the regulation and initiation of T-cell responses. Immunol. Rev. 153:27. Even though it is currently not clear how CD4 and CTLA-4 24. Farber, D. L., M. Luqman, O. Acuto, and K. Bottomly. 1995. Control of memory Downloaded from CD4 T cell activation: MHC class II molecules on APCs and CD4 ligation inhibit inhibit memory CD4 T cell responses, it is evident that the two memory but not naive CD4 T cells. Immunity 2:249. mechanisms act in concert. We propose that CTLA-4 signaling is 25. Metz, D. P., D. L. Farber, R. Konig, and K. Bottomly. 1997. Regulation of not just required for down-regulation of already activated T cells, memory CD4 T cell adhesion by CD4-MHC class II interaction. J. Immunol. but that it is a critical regulator of resting memory CD4 T cells. 159:2567. 26. Gosgrove, D., D. Gray, A. Dierich, J. Kaufman, M. Lemeur, C. Benoist, and CTLA-4 inhibits memory CD4 T cell proliferation under condi- D. Mathis. 1991. Mice lacking MHC class II molecules. Cell 66:1051. tions where CD4 is ligated outside the TCR complex. 27. Konig, R., L.-Y. Huang, and R. N. Germain. 1992. MHC class II interaction with http://www.jimmunol.org/ CD4 mediated by a region analogous to the MHC class I site for CD8. Nature 356:796. Acknowledgments 28. Tan, P., C. Anasetti, J. A. Hansen, J. Melrose, M. Brunvand, J. Bradshaw, We thank Dr. I. N. Crispe for critical review of the manuscript, and Patricia J. A. Ledbetter, and P. S. Linsley. 1993. Induction of alloantigen-specific hyporesponsiveness in human T lymphocytes by blocking interaction of CD28 with its Ranney for looking after the mice. natural ligand B7/BB1. J. Exp. Med. 177:165. 29. Konig, R., X. Shen, and R. N. Germain. 1995. Involvement of both MHC class References II ␣ and ␤ chains in CD4 function indicates a role for ordered oligomerization in T cell activation. J. Exp. Med. 182:779. 1. Kirberg, J., A. Berns and H. von Boehmer. 1997. Peripheral T cell survival 30. Chamber, C. A., T. J. Sullivan, and J. P. Allison. 1997. Lymphoproliferation in

requires continual ligation of the T cell receptor to major histocompatibility com- CTLA-4-deficient mice is mediated by costimulation-dependent activation of by guest on September 29, 2021 plex-encoded molecules. J. Exp. Med. 186:1269. CD4ϩ T cells. Immunity 7:885. 2. Takeda, S., H.-R. Rodewald, H. Arakawa, H. Bluethmann, and T. Shimizu. 1996. ϩ 31. Linsley, P. S., J. Bradshaw, J. Greene, R. Peach, K. L. Bennett, and R. S. Mittler. MHC class II molecules are not required for survival of newly generated CD4 1996. Intracellular trafficking of CTLA-4 and focal localization towards sites of T cells, but affect their long-term life span. Immunity 5:217. TCR engagement. Immunity 4:535. 3. Linsley, P., W. Brady, L. Grosmaire, A. Aruffo, N. K. Damle, and J. A. Ledbetter. 1991. Binding of the B cell activation antigen B7 to CD28 costimulates T cell 32. Alegre, M.-L., P. J. Noel, B. J. Eisfelder, E. Chuang, M. R. Clark, S. L. Reiner, proliferation and interleukin 2 mRNA accumulation. J. Exp. Med. 173:721. and C. B. Thompson. 1996. Regulation of surface and intracellular expression of 4. Koulova, L., E. A. Clark, G. Shu, and B. Dupont. 1991. The CD28 Ligand CTLA-4 on mouse T cells. J. Immunol. 157:4762. B7/BB1 provides costimulatory signal for alloactivation of CD4ϩ T cells. J. Exp. 33. Walunas, T. L., D. J. Lenschow, C. Y. Bakker, P. S. Linsley, G. J. Freeman, Med. 173:759. J. M. Green, G. B. Thompson, and J. A. Bluestone. 1994. CTLA-4 can function 5. Lenschow, D. J., G. H.-T. Su, L. A. Zuckerman, N. Nabavi, C. L. Jellis, G. S. Gray, as a negative regulator of T cell activation. Immunity 1:405. J. Miller, and J. A. Bluestone. 1993. Expression and functional significance of an 34. Shiratori, T., S. Miyatake, H. Ohno, C. Nakaseko, K. Isono, J. S. Bonifacino, and additional ligand for CTLA-4. Proc. Natl. Acad. Sci. USA 90:11054. T. Saito. 1997. Tyrosine phosphorylation controls internalization of CTLA-4 by 6. Linsley, P. S., and J. A. Ledbetter. 1993. The role of the CD28 receptor during regulating its interaction with clathrin-associated adaptor complex AP-2. Immu- T cell responses to antigen. Annu. Rev. Immunol. 11:191. nity 6:583. 7. Byrne, J. A., J. L. Butler, and M. D. Cooper. 1988. Differential activation re- 35. Madrenas, J., L. A. Chau, J. Smith, J. A. Bluestone, and R. N. Germain. 1997. The quirements for virgin and memory T cells. J. Immunol. 141:3249. efficiency of CD4 recruitment to ligand-engaged TCR controls the agonist/partial 8. Byrne, J. A., J. L. Butler, E. L. Reinherz, and M. D. Cooper. 1989. Virgin and agonist properties of peptide-MHC molecule ligands. J. Exp. Med. 185:219. memory T cells have different requirements for activation via the CD2 molecule. Int. Immunol. 1:29. 36. Marengere, L. E. M., P. Waterhouse, G. S. Duncan, H.-W. Mittruecker, G.-S. 9. Pihlgren, M., P. M. Dubois, M. Tomkowiak, T. Sjogren, and J. Marvel. 1996. Feng, and T. W. Mak. 1996. Regulation of T cell receptor signaling by tyrosine Resting memory CD8ϩ T cells are hyperreactive to antigenic challenge in vitro. phosphatase SYP association with CTLA-4. Science 272:1170. J. Exp. Med. 184:2141. 37. Van der Merwe, P. A., D. L. Bodian, S. Daenke, P. Linsley, and S. J. Davis. 1997. 10. Bradley, L. M., D. D. Duncan, K. Yoshimoto, and S. L. Swain. 1993. Memory CD80 (B7–1) binds both CD28 and CTLA-4 with a low affinity and very fast effectors: a potent, IL-4-secreting helper T cell population that develops in vivo kinetics. J. Exp. Med. 185:393. after restimulation with antigen. J. Immunol. 150:3119. 38. Linsley, P. S., J. L. Greene, P. Tan, J. Bradshaw, J. A. Ledbetter, C. Anasetti, and 11. Farber, D. L., O. Acuto, and K. Bottomly. 1997. Differential T cell receptor- N. K. Damle. 1992. Coexpression and functional cooperativity of CTLA-4 and mediated signaling in naive and memory CD4 T cells. Eur. J. Immunol. 27:2094. CD28 on activated T lymphocytes. J. Exp. Med. 176:1595. 12. MacKay, C. R., W. L. Marston, and L. Dudler. 1990. Naive and memory T cells 39. Brocker, T. 1997. Survival of mature CD4 T lymphocytes is dependent on major show distinct pathways of lymphocyte recirculation. J. Exp. Med. 171:801. histocompatibility complex class II-expressing dendritic cells. J. Exp. Med. 8:1223. 13. MacKay, C. R., W. L. Marston, L. Dudler, O. Spertini, T. F. Tedder, and 40. Lalvani, A., R. Brookes, S. Hambleton, W. J. Britton, A. V. S. Hill, and W. R. Hein. 1992. Tissue-specific migration pathways by phenotypically distinct A. J. McMichael. 1997. Rapid effector function in CD8 memory T cells. J. Exp. subpopulations of memory T cells. Eur. J. Immunol. 22:887. Med. 186:859. 14. Croft, M., L. M. Bradley, and S. L. Swain. 1994. Naive versus memory CD4 T cell response to antigen. J. Immunol. 152:2675. 41. Tuosto, L., C. Piazza, S. Moretti, A. Modesti, R. Greenlaw, R. Lechler, ϩ G. Lombardi, and E. Piccolella. 1995. Ligation of either CD2 or CD28 rescues 15. Luqman, M., and K. Bottomly. 1992. Activation requirements for CD4 T cells ϩ differing in CD45R expression. J. Immunol. 149:2300. CD4 T cells from HIV-gp120-induced apoptosis. Eur. J. Immunol. 25:2917. 16. Gause, W. C., P. Lu, X. Zhou, S. Chen, K. B. Madden, S. C. Morris, P. S. Linsley, 42. Faith, A., H. Yssel, R. E. O’Hehir, and J. R. Lamb. 1996. Reversal of the inhib- F. D. Finkelman, and J. F. Urban. 1996. H. polygyrus: B7-independence of the itory effects of HIV-gp120 on CD4ϩ T cells by stimulation through the CD28 secondary type 2 response. Exp. Parasitol. 84:264. pathway. Clin. Exp. Immunol. 105:225.