In this study, we show that the costimulation of CD9 on naive T cells during TCR ... costimulation led to the induction of apoptosis of once-activated T cells.
CD9-Mediated Costimulation of TCR-Triggered Naive T Cells Leads to Activation Followed by Apoptosis’ Xu-Guang Tai,* Kazuhito Toyooka,* Yumi Yashiro,* Ryo Abe,+ Cheung-Seog Park,* Toshiyuki Hamaoka,* Michiko Kobayashi,* Steven Neben,% and Hiromi Fujiwara** The induction of full activation or death in TCR-triggered Tcells depends largely on whether appropriate costimulatorysignals are provided. In this study, we show that the costimulation of CD9 on naive T cells during TCR stimulation results in transient, albeit potent, activation followed by apoptosis, rather than full activation. Anti-CD9 mAb synergized with suboptimal doses of anti-CD3 mAb in inducing T cell activation. t3H]TdR incorporation determined 2 days after CD9 costimulation wasas potent as that induced by CD28 costimulation. In contrast to progressive T cell proliferation induced by CD28 costimulation, CD9 costimulation led to the induction of apoptosis of once-activated T cells. Although IL-2R expression was induced significantly earlier and to a greater degree after CD9 costimulation than after CD28 costimulation, CD9 costimulation only transiently produced a small amount of 11-2 and induced apparentlylow levels of bcl-x, compared with those observed in CD28 costimulation. Addition of rlL-2 to cultures of CD9 costimulation induced strikingly enhanced expression ofbcl proteins, especially of bcI-xL, and protected TCR-stimulated T cells from apoptosis.These data indicate that CD9-mediated costimulation of TCRtriggered naiveT cells leads to activation followed by apoptosis as the result of failure to generate a positive signal for sufficient levelsof11-2 production. The Journal of Immunology, 1997, 159: 3799-3807.
T
cell activation is initiated by the interaction of Ag-specific TCR with processed Ag peptides plus MHC molecules expressed on APCs ( l), While this is the primary signal in T cell activation, complete activation of the T cell requires a second (costimulatory) signal that is generated by other receptor-ligand interactions between the APC and the T cell (1-3). Recent studies have revealed that naive T cells die by apoptosis when activated by TCR engagement alone in the absence of costimulatory signals (4-7). There are multiple receptor-ligand interactions that take place between the T cell and the APC to deliver effective costimulatory signals (8). Many studies have demonstrated that CD28 is theprincipal costimulatory receptor for T cell activation (9, IO). Costimulation of T cells has been shown to affect various aspects of T cell activation (9). CD28 costimulation markedly enhances the production of lymphokines, especially of IL-2 necessary for growth, by T cells (11 , 12). Simultaneously, the signal through CD28 can directly induce CD25 (IL-2R a-chain) expression on virgin murine T cells by an IL-2-independent mechanism (13), resulting in T cell proliferation. In addition, the CD28 signaling pathway appears to *Biomedical Research Center, Osaka University Medical School, Yamadaoka, Suita, Osaka, Japan; ‘Research Institute for Biological Sciences, Science University of Tokyo, Chiba, Japan; Immune Cell Biology Program, Naval Medical Research Institute, Bethesda, MO 20889; and *Genetics Institute, Cambridge, M A 02140 Received forpubllcationJanuary 21, 1997.
21, 1997.Acceptedforpublication
July
The costs of publication of this article were defrayed i n part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
’
This work was supported by Grants-in-Aid for Scientific Research from the Ministry of Education, Science, and Culture, Japan, and in part by Naval Medical Research and Development Command Grant M0095003.1412. The views expressed in this article are those of the authors and do notreflect the officialpolicy and position of Department of the Navy, Department of Defense, or the United States government. Address correspondence and reprint requests to Dr. HiromiFujiwara, Biomedical Research Center, Osaka University MedicalSchool, 2-2, Yamada-oka, Suits, Osaka 565, Japan. Copyright 0 1997 by The American Association of Immunologists
have an important role in the survival of naive T cells after their TCR are stimulated. Studies from two laboratories have demonstrated that CD28 costimulation not only enhances the proliferative expansion of naive T cells activated through the TCR, but also by inducing enhanced expression of bcl-x, protein, augments the intrinsic ability of these activated T cells to resist apoptosis and to survive during a primary T cell activation (5-7). Thus, the expression of both extrinsic growth factor (IL-2) and intrinsic cell survival factors (bcl-x,) appears to play a role in the survival of TCRstimulated T cells. Despite a seemingly absolute requirement of CD28 for T cell activation and apoptosis prevention, it is also true that T cells from CD28-deficient mice can mount immune responses in vitro and in vivo, although at a lower magnitude than T cells from wild-type mice (14, 15). This suggests that there may exist costimulatory interactions between the T cell and the APC that function independently of CD28. In fact, various cell-bound and soluble molecules have been shown to serve as costimulatory receptors or to provide costimulatory function. These include CD43 (16), CD44 (17), CD5 (18), and IL-1/IL-6 (19). Recently, we have also found that the CD9 molecule, a member of the transmembrane 4 superfamily (TM4SF)3 (20), is expressed on most mature T cells and can deliver a costimulatory signal in terms of enhanced [3H]TdR incorporation (21). Surprisingly, the initial activation (increased t3H]TdR uptake) of TCR-stimulated naive T cells by CD9 costimulation was as potent as that induced by CD28 costimulation. However, it is possible that molecules other than CD28 with quantitatively comparable costimulatory capacity may deliver qualitatively different signals depending on the end point measured. In this work, we have examined the nature of CD9 costimulation in comparison with that of CD28 costimulation. We show that CD9 costimulation elicits the activation (as measured by [3H]TdR uptake) of naive T cells like CD28 costimulation, but unlike CD28
’Abbreviations used in this paper:
TM4SF, transmembrane 4 superfamily; PI,
propldium iodide; SN, supernatant, 0022-1 767/97/$02.00
3800
costimulation induces apoptosis of these once activated T cells instead of complete activation (cellular proliferation). This is due to a failure to produce IL-2 progressively after CD9 costimulation and consequently to express sufficient amounts of bcl-x, protein. Thus, this study shows the fundamentally different nature of CD9 costimulation from that of CD28 costimulation.
Materials and Methods Mice C57BL/6 mice were purchased from Shizuoka Laboratory Animal Center (Hamamatsu, Japan) and used at 7 to 9 wk of age.
Reagents Anti-CD3 (145-2C11) (22), anti-CD9 (9D3) (21), anti-CD28 (Pv-I) (23), anti-IL-2R a-chain (7D4) (24), anti-CD4 (American Type Culture Collection (ATCC) clone GK1.5; ATCC, Rockville, MD), and anti-CD8 (ATCC clone 2.43) mAbs were purified from ascitic fluids of hybridomas producing each mAb. Anti-bcl-x capable of recognizing bcl-x, protein and antibcl-2 mAbs were purchased from Transduction Laboratories (Lexington, KY) and PharMingen (San Diego, CA), respectively.
Preparation of a purified T cell population Lymph node cells were depleted of B cells by immunomagnetic negative selection, as described (21). Briefly, lymph node cells were incubated with Advanced Magnetic particles bound to goat anti-mouse Ig (Advanced Magnetics, Cambridge, MA). Surface Ig-negative cells were obtained by removing cell-bound magnetic particles with a rare earth magnet (Advanced Magnetics). This was followed by complement-mediated cytotoxic treatment with anti-Ia mAb (34-5-3s). Purity of the resulting population was checked by flow microfluorometry with anti-CD3. Purified T cells were consistently >98% CD3 positive.
Treatment of T cells with anti-CD4/CD8+ C Purified T cells were treated with anti-CD4 or anti-CD8 mAb in the presence of rabbit C, as described (25).
T cell culture for proliferation, IL-2 production, If -2R induction, and detection of apoptosis and bcl proteins mAbs were diluted to various concentrations in PBS and immobilized to individual wells of 96-well flat-bottom microculture plates (Coming 25860; Coming Glass Works, Coming, NY) in a final volume of 0.1 ml, or 24-well culture plates (Coming 25820) in a volume of 1 ml. After 3 h, solutions were discarded and plates were washed with PBS twice. Purified T cells were cultured in 0.2 ml of RPMI 1640 medium supplemented with 10% FBS and 2-ME at 1 or 2 X lo5 cells/well of mAb-immobilized 96well microculture plates in a humidified atmosphere at 5% CO, at 37°C for various days. The cultures were harvested after an 8-h pulse with 20 kBq/ well of [3H]TdR. Results were calculated from uptake of 13H]TdR and expressed as the mean cpm C SE of triplicate cultures. For assays other than [3H]TdR incorporation, purified T cells (2 X 106/well) were cultured in mAb-immobilized 24-well culture plates in a volume of 2 ml. Culture supernatants (SN) and cells were harvested after various times in culture. Culture SN were assessed for IL-2 activity. Cells were subjected to the detection of IL-2R, DNA fragmentation, and bcl proteins.
Assay systems for IL-2 activity IL-2 activity was assessed by the ability to support the proliferation of the IL-2-dependent T cell line, CTLL-2. CTLL-2 ( 104/well)were cultured with the SN in a volume of 0.2 ml in 96-well flat-bottom microplates (Coming 25860) for24hat 37°C. Proliferation was assessed by the uptake of [3H]TdR during a 4-h pulse with 20 kBq/well of 13H]TdR. The absolute concentrations of lymphokines were determined by extrapolation from each standard curve that was generated by using known amounts of rIL-2.
Immunofluorescence staining and flow microfluorometry Anti-IL-2R (7D4) mAbs were biotinylated in our laboratory. Cells (1 X IO6) were stained with biotinylated 7D4 mAb, followed by FITC-conjugated streptavidln (Becton Dickinson, Mountain View, CA). Cells were analyzed on a FACScan (Becton Dickinson).
Measurement of apoptosis Apoptosis was determined by DNA fragmentation analysis after agarose gel electrophoresis. T cells (1 X 106/sample) harvested after cultures were
CD9-MEDIATED APOPTOSIS IN T CELLS washed in PBS and lysed, and their DNA was isolated as previously described (26). DNA was electrophoresed in 2% agarose gels and stained with 0.1 pg/ml of ethidium bromide. DNA fragmentation was also examined as follows: cells were fixed and stained with propidium iodide (PI), and DNAfragmentation (the proportion of apoptotic cells) was quantitated by determination of hypodiploid areas in PI-staining profiles used for cell cycle analysis (27).
Western blot analysis T cells were activated as described above forindicated times. At each time point, 5 X IO5 viable cells from each of the culture conditions were isolated and lysed in 2X SDS-sample buffer containing 6% 2-ME, and then boiled for 5 min. Proteins were separated on a 12% SDS-polyacrylamide gel, and then transferred to nitrocellulose. The membranes were blocked in PBSI Tween-20 containing 5% skim milk, and incubated with anti-bcl-x, Abs in the same medium, according to the manufacturer’s instructions. After washing, the membranes were treated with sheep anti-mouse IgG-horseradish peroxidase conjugate at l/t0,000 dilution (Amersham, Buckinghamshire, UK). Blots were developed using the ECL chemiluminescence reagents (Amersham; HP7 9NA). For bcl-2 detection, the same membranes were incubated with anti-bcl-2 Abs, and complexes were detected with protein A-horseradish peroxidase conjugate at 1/10,000 dilution (Amersham).
Results CD9 costimulation elicits the initial activation of naive T cells, but fails to induce their complete activation
Our earlier study demonstrated that the 9D3 (anti-CD9) mAb provides potent T cell costimulation in the absence of APC (21). We confirmed this fact (Fig. I). Purified C57BL16 (B6) T cells were cultured for 1 to 3 days in wells containing anti-CD9 (9D3) or anti-CD28 coimmobilized with a suboptimal dose of anti-CD3 mAb in the absence of APC. The anti-CD9 mAb strikingly increased [3H]TdR uptake of purified B6 T cells. The magnitude of CD9 costimulation was comparable with that of CD28 costimulation when evaluated on day 2 or 3 (Fig. 1A). Figure 1B shows that an increased [3H]TdR uptake is observed for both CD4+ and CD8+ subsets of naive T cells. In the course of the above experiments, we found a morphologic difference in cellular responses between CD9 and CD28 costimulation. Blastoid T cells were seen in both CD28 and CD9costimulation cultures. While most T cells (both blastoid and small cells) were observed as viable in CD28 costimulation, cultures of CD9 costimulation contained a significant number of small apoptotic cells among blastoid cells (data not shown). Figure 2 compares the cell recovery between CD9 and CD28 costimulation cultures. In contrast to a progressive increase in the viable cell recovery after CD28 costimulation, the number of viable cells decreased and the number of dead cells increased from day 2 to day 3 in CD9 costimulation cultures. These results indicate that CD9 costimulation can elicit the transient T cell activation, as detected by l3H]TdR uptake, but fails to induce full activation (T cell proliferation). hduction of apoptosis in TCR-stimulated naive T cells following CD9 costimulation
To determine whether a decrease in the viable cell recovery following CD9 costimulation is due to the induction of cell death (apoptosis), we examined DNA from costimulation cultures of antiCD3 plus anti-CD9 or anti-CD28 (Fig. 3). Purified naive T cells stimulated with a suboptimal dose of anti-CD3 alone displayed a significant degree of DNA fragmentation. The levels of DNA fragmentation were quite low in T cells from CD28 costimulation CUItures. In contrast, CD9 costimulation induced a high level of DNA fragmentation. Figure 3 also shows that DNA fragmentation by CD9 costimulation is induced in both CD4+ and CD8+ T cell subsets. Taken collectively, the results suggest that CD9 costimulation elicits apoptotic cell death of initially activated T cells instead of inducing the proliferative expansion of these cells.
3801
The Journal of I m m u n o l o g y I
I
A
-s-
tr-CD3 a-CD3+a-CDQ u-CD3+a-CD28
4- tr-CD3
-
-f-
