Trogocytosis of CD80 and CD86 by Induced Regulatory T Cells

RESEARCH ARTICLE

Trogocytosis of CD80 and CD86 by induced regulatory T cells

Peng Gu1,2,4, Julia Fang Gao1,3,4, Cheryl A D’Souza1,3,4, Aleksandra Kowalczyk1, Kuang-Yen Chou2 and Li Zhang1,3

Trogocytosis is a process which involves the transfer of membrane fragments and cell surface proteins between cells. Various types of T cells have been shown to be able to acquire membrane-bound proteins from antigen-presenting cells and their functions can be modulated following trogocytosis. However, it is not known whether induced regulatory T cells (iTregs) can undergo trogocytosis, and if so, what the functional consequences of this process might entail. In this study, we show that iTregs can be generated from CD802/2CD862/2 double knockout (DKO) mice. Using flow cytometry and confocal fluorescence microscopy, we demonstrate that iTregs generated from DKO mice are able to acquire both CD80 and CD86 from mature dendritic cells (mDCs) and that the acquisition of CD86 occurs to a higher extent than that of CD80. Furthermore, we found that after co-incubation with iTregs, dendritic cells (DCs) downregulate their surface expression of CD80 and CD86. The trogocytosis of both CD80 and CD86 occurs in a cytotoxic T lymphocyte-associated antigen-4 (CTLA-4), CD28 and programmed death ligand-1 (PDL1)-independent manner. Importantly, we showed that iTregs that acquired CD86 from mDCs expressed higher activation markers and their ability to suppress naive CD41 T-cell proliferation was enhanced, compared to iTregs that did not acquire CD86. These data demonstrate, for the first time, that iTregs can acquire CD80 and CD86 from mDCs, and the acquisition of CD86 may enhance their suppressive function. These findings provide novel understanding of the interaction between iTregs and DCs, suggesting that trogocytosis may play a significant role in iTreg-mediated immune suppression. Cellular & Molecular Immunology (2012) 9, 136–146; doi:10.1038/cmi.2011.62; published online 6 February 2012

Keywords: CD80; CD86; CTLA-4; iTregs; trogocytosis

INTRODUCTION inhibiting IL-2 production and cell cycle progression.11 While both Regulatory T cells (Tregs) are a vital subpopulation of T cells and are CD28 and CTLA-4 can bind CD80 and CD86, CD80 is the major essential for maintaining peripheral tolerance, preventing autoim- ligand mediating CTLA-4 localization and CD86 is the main ligand mune diseases and limiting chronic inflammatory diseases.1 Natural for CD28 concentration at the immunological synapse.12 Dendritic CD41CD251FoxP31 Tregs (nTregs) develop in the thymus, while cells (DCs) are professional APCs and their costimulatory molecule adaptive or induced Tregs (iTregs) are generated in the periphery from expression levels, including CD80 and CD86, are critical for eliciting conventional CD41CD252FoxP32 T cells.2 CD41CD251Foxp31 T-cell responses. One of the mechanisms by which Tregs suppress iTregs can also be generated in vitro by stimulation of naive CD4 T immune responses is to inhibit DCs. Onishi et al.13 found that T-cell cells in the presence of transforming growth factor-b (TGF-b),3,4 which receptor (TCR)-stimulated nTregs from DO11.10 transgenic mice induces Foxp3 expression.5 Once induced, iTregs can regulate immune were able to downregulate CD80 and CD86 expression on DCs in a responses both in vitro and in vivo.5–9 Whereas the nature and the lymphocyte function-associated antigen-1 (LFA-1)- and CTLA-4- mechanisms involved in nTreg-mediated suppression have been dependent manner. Qureshi et al.14 demonstrated that both effector studied extensively, iTreg development and function remain elusive. T cells and Tregs can downregulate CD80 and CD86 expression by Cytotoxic T lymphocyte-associated antigen-4 (CTLA-4) expressed CTLA4-mediated trans-endocytosis. Whether iTregs can downregu- on Tregs is a key molecule involved in controlling their suppressive late CD80 and CD86 expression through trogocytosis is not known. function, and together with CD28, helps regulate immune responses. Trogocytosis is the transfer of membrane-associated proteins Both CTLA-4 and CD28 can bind to the B7 family members CD80 and together with lipid fractions of the cell membrane. It is a widespread CD86 expressed on the surface of antigen-presenting cells (APCs). phenomenon which endows acceptor cells with novel functions.15 CD28 enhances T-cell proliferation by increasing the transcription Trogocytosis has been demonstrated by a variety of cell types,16 with and the mRNA stability of IL-2, as well as by upregulating the anti- both stimulatory and suppressive effects on immune responses. So far, apoptotic protein Bcl-xL.10 CTLA-4 restricts T-cell activation by most of the work on trogocytosis has been carried out on effector T

1University of Toronto Transplant Institute, Toronto General Research Institute, University Health Network, Toronto, Ontario, Canada; 2Shanghai Institute of Immunology, Shanghai Jiaotong University School of Medicine, Shanghai, China and 3Department of Immunology, University of Toronto, Toronto, Ontario, Canada 4 These authors contributed equally to this paper. Correspondence: Dr L Zhang, Toronto General Hospital Research Institute, Toronto Medical Discovery Tower 2-807, 101 College Street, Toronto, Ontario M5G 1L7, Canada. E-mail: [email protected] Received 4 October 2011; revised 4 January 2012; accepted 5 January 2012 Trogocytosis of CD80 and CD86 by iTregs PGuet al 137 cells. To the best of our knowledge, no previous study has analyzed blocking purified anti-CTLA4 (UC10-4F10-11) were purchased from CD80 and CD86 acquisition by iTregs. In this study, we sought to BD Biosciences Pharmingen (Mississauga, Ontario). better characterize iTreg generation, trogocytosis and function. To examine trogocytosis of CD80 and CD86, we generated iTregs from BMDCs 2/2 2/2 CD80 CD86 double knockout (DKO) mice. We demonstrate C57BL/6 wild-type (WT) mice, CD45.11 mice or CD802/2 single KO that iTregs can acquire both CD80 and CD86 from mature bone mice were used to isolate BMDCs. Briefly, femur and tibia bones were marrow-derived dendritic cells (BMDCs) and express the acquired dissected and flushed with RPMI media. Red blood cells were lysed by molecules on their surface. Co-incubation of mature dendritic cells incubation in red blood cell Lysis buffer (0.14 M NH4Cl, 0.02 M Tris (mDCs) with iTregs resulted in downregulation of their CD80 and Base, pH 7.2) and the remaining cells were pelleted at 400g for 5 min. CD86 expression. Furthermore, trogocytosis of CD80 and CD86 Cells were plated at 23106 cells per 100 mm dish in 10 ml RPMI-1640 occurs independently of CTLA-4 or CD28 expression on iTregs. complete medium containing 100–200 U/ml (520 ng/ml) rm GM- More importantly, we found that iTregs that are able to acquire CSF. At day 3, another 10 ml complete medium containing 100– CD86 showed enhanced suppressive function. These data reveal, for 200 U/ml rm GM-CSF were added to the plates. At days 6 and 8, a the first time, that trogocytosis of costimulatory molecules from APCs 50% media swap was done with fresh media containing 100–200 U/ml may have biological significance in relation to Treg function. rm GM-CSF. LPS (0.1 mg/ml) was added to the complete medium for the last 24 h. MATERIALS AND METHODS Mice In vitro induction of iTregs 1 C57BL/6 (Thy1.2 ), B6.PL-Th1a/CyJ (congenic to C57BL/6, C57BL/6 WT or CD802/2CD862/2 DKOmicewereusedfortheisola- 1 a b 1 Thy1.1 ) and B6.SJL-Ptprc Pepc /BoyJ (CD45.1 ) mice were tion of CD41CD252 T cells. The cells were then induced in vitro to obtained from The Jackson Laboratory (Bar Harbor, Maine). produce CD41CD251Foxp31 cells (iTregs) as previously described.8 2/2 2/2 2/2 CD80 CD86 DKO mice and CD80 knockout (KO) mice Briefly, splenocytes and lymph node cells were incubated with PE- (all on C57BL/6 background) were kindly provided by Dr David conjugated anti-CD8, PE-anti-CD25, PE-anti-B220, PE-anti-CD11b, Kelvin (University Health Network). All mice were housed in specific PE-anti-NK1.1 and PE-anti-TER119 Abs for 20 min at 4 uCandwashed pathogen-free conditions at the University Health Network mouse once. Cells were then resuspended in buffer (0.5% bovine serum albu- facility (Toronto, Canada). All experiments were approved by the min, 2 mM EDTA in phosphate-buffered saline, pH 7.2) and incubated University Health Network Animal Care Committee. with anti-PE microbeads for 20 min at 4 uC and washed once. CD41CD252 cells were purified by LD column. For the induction of Antibodies (Abs) and reagents iTregs, these cells were cultured in 24-well plates coated with 10 mg/ml Reagents used in this study included: recombinant murine (rm) anti-CD3 mAb, in the presence of 2 mg/ml soluble anti-CD28 Ab, 200 U/ granulocyte–macrophage colony-stimulating factor (GM-CSF; ml IL-2 and 5 ng/ml TGF-b. Cells were cultured for 5 days and the ProSpec, East Brunswick, New Jersey), RPMI-1640 media (GIBCO, expression of Foxp3 and CD25 was evaluated by flow cytometry. Burlington, Ontario), lipopolysaccharide (LPS; Sigma, Oakville, Ontario), antiphycoerythrin (anti-PE) microbeads (Miltenyi Acquisition experiments Biotec, Boston, Massachusetts), LD columns (Miltenyi Biotec), IL- WT or DKO iTregs (1.53105 cells/well) were cocultured with BMDCs, 2 (Chiron Corp., Emeryville, California), TGF-b (R&D Systems, at a DC/iTreg ratio of 2 : 1, in 96-well round bottom plates with 50 U/ Burlington, Ontario), propidium iodide (Sigma-Aldrich), eFluor ml IL-2. At various time points after coculture, cells were collected and 450 (eBioscience, San Diego, California), PKH26 (Sigma-Aldrich) immunostained and used either for flow cytometry or confocal micro- and carboxyfluorescein diacetate, succinimidyl ester (CFSE; scopy analysis. To examine acquisition of CD80 and CD86 by iTregs Invitrogen, Burlington, Ontario). Cells were fixed and permeabilized after coculture with DCs, plots were gated on PI2CD11c2CD41 cells using BD Cytofix/Cytoperm Plus Fixation/Permeabilization kit (BD) and CD80 and CD86 staining within this gate was examined. In block- and FoxP3 Staining Buffer Set (eBioscience). ing assays, iTregs were pre-incubated with either 20 mg/ml anti-CTLA- The following Abs were purchased from BioLegend (San Diego, 4 Ab, 20 mg/ml anti-CD28 Ab or 20 mg/ml anti-PDL1 Ab, 1 h prior to California): PE-conjugated anti-CD8 (53-6.7), PE-anti-CD25 (PC61), coculture. PE-anti-CD11b (M1/70), PE-anti-NK1.1 (PK136), PE-anti-TER119 (TER-119), PE-anti-CD4 (GK1.5), PE-anti-CD11c (N418), PE-CD274/ Flow cytometry programmed death ligand-1 (PDL1) (10F.9G2), fluorescein isothiocya- Samples were stained with the appropriate Abs and sample acquisition nate (FITC)-conjugated anti-CD25 (PC61), FITC-anti-CD80 (16- 10A1), FITC-anti-CD86 (GL-1), FITC-anti-CD11c (N418), allophyco- was performed on a Becton Dickinson LSRII and a Beckman Coulter cyanin (APC)-conjugated anti-CD90.1/Thy1.1 (OX-7), APC-anti- Cytomics FC500 cytometer and data were analyzed using FlowJo soft- CD11c (N418), PE/Cy5-conjugated anti-CD4 (GK1.5), Alexa Fluor ware (TreeStar Inc., Ashland, Oregon). 488-conjugated anti-CD3 (17A2), Alexa Fluor 647-conjugated anti- CD11c (N418), purified antimouse CD28, anti-CD3 (2C11), blocking Cell sorting 2/2 LEAF purified antimouse CD152/CTLA-4 (UC10-4B9), blocking LEAF DKO iTregs were cocultured with CD80 BMDCs in 96-well round purified antimouse CD28 and LEAF purified antimouse CD274/PDL1 bottom plates with 50 U/ml IL-2 for 24 h. Cells were collected and (10F.9G2), purified syrian hamster IgG isotype control. The following stained with FITC-conjugated anti-CD11c, PE-conjugated anti-CD86 Abs were purchased from eBioscience: PE-anti-B220 (RA3-6B2), PE- and PE/Cy5-conjugated anti-CD4. The DKO iTregs which expressed a anti-Foxp3 (FJK-16s), PE/Cy7-anti-CD3 (145-2C11), PE-anti-CD80 high level of acquired CD86 and the iTregs which did not acquire (16-10A1), PE-CTLA-4 (UC10-4B9), PE-CD69 (H1.2F3), PE-CD44 CD86 were sorted by using a Becton Dickinson FACS Aria II cell (IM7) and APC-CD45.1 (A20). PE-conjugated anti-CD86 (GL-1) and sorter.

Cellular & Molecular Immunology Trogocytosis of CD80 and CD86 by iTregs PGuet al 138

Confocal microscopy CD80 and CD86 in WT iTregs after coculture with mDCs, we used Cells were washed with phosphate-buffered saline and stained with iTregs generated from CD802/2CD862/2 DKO mice. As shown in Alexa Fluor 488-conjugated anti-CD3 (17A2), Alexa Fluor 647-conjugated Figure 2b, the expression of CD80 and CD86 on DKO iTregs was anti-CD11c (N418) and PE-conjugated anti-CD80 (16-10A1) or CD86 also increased after co-incubation with mDCs, suggesting acquisi- (GL1) Abs. Cells were plated onto chamber slides and visualized live using tion of these molecules. an Olympus Fluoview 1000 Laser Scanning Confocal Microscope. To further confirm this finding, we examined this phenomenon by using confocal microscopy. DKO iTregs and mDCs were cocul- Suppression assay tured for 24 h and stained with Alexa Fluor 647-anti-CD11c, Alexa CD41CD252 T cells (Thy1.11) were purified from naive B6.PL-Th1a/ Fluor 488-anti-CD3 and PE-anti-CD80 or CD86. DKO iTregs CyJ mice, labeled with CFSE, and cultured (13105 cells/well) with showed diffuse staining for CD3 while only a small membrane 1 mg/ml anti-CD3 Ab in the presence of irradiated syngeneic spleno- patch stained positive for either CD80 or CD86 (Figure 2c). cytes (23104 cells/well). Varying numbers of CD41CD251Foxp31 Superimposition of the red PE signal on the green Alexa Fluor iTreg from WT or DKO C57BL/6 mice were added to the culture at 488 signal resulted in yellow staining on DKO iTregs, suggesting suppressor to responder ratios 1 : 0.5, 1 : 1, 1 : 2, 1 : 4 and 1 : 8. After 4 colocalization of CD3 and CD80 or CD86. These results dem- days of culture, proliferation of responder cells was measured by gat- onstrate that CD80 and CD86 are physically acquired by DKO ing on Thy1.11 cells and assessing their CFSE signal by flow cytometry. iTregs from mDCs through trogocytosis. To further understand the nature of iTreg trogocytosis, we investi- Statistical analysis gated the time course of CD80 and CD86 acquisition by iTregs. DKO Data are presented as the mean6SEM per group. Statistical analysis iTregs were cocultured with mDCs, and CD80 and CD86 expression was made for multiple comparisons using Student’s t-test. A P value of on iTregs was measured by flow cytometry at various time points after ,0.05 was considered to be statistically significant. coculture (Figure 3a). After 2 h of incubation, DKO iTregs had acquired both CD80 and CD86; however, a higher amount of CD86 RESULTS than CD80 was present on iTreg membranes (Figure 3b). During the 1 1 1 24-h time course, DKO iTregs continued to acquire both CD80 and Functional CD4 CD25 Foxp3 iTregs can be generated in vitro 2 2 2 2 CD86, with higher acquisition of CD86 observed at all time points from CD80 / CD86 / DKO mice examined. It is known that T cells can express CD80 and CD86 on their surface 17 Because trogocytosis is known to involve transfer of both proteins after activation. In order to avoid the potential influence of endo- genously expressed CD80 and CD86, we chose to study iTreg trogo- and membrane fragments, we prestained mDCs with the membrane cytosis of CD80 and CD86 using CD802/2CD862/2 DKO mice. dye PKH26 prior to coculture with iTregs, and then measured the However, mice that lack CD80 and CD86 show a dramatic reduction amount of dye transferred to iTregs after coculture. In contrast to in nTreg frequency in the thymus,18 indicating that these molecules the steady increase in CD80 and CD86 on iTregs, transfer of mem- are important for nTreg development. Whether iTreg generation and brane fragments from DCs to iTregs was very high at 6 h of coculture, function would be affected by the CD80/CD86 DKO was not known. but then decreased from 6 to 24 h of coculture (Figure 3c). This To address this question, we first determined whether equivalent suggests that CD80 and CD86 are not passively transferred from numbers of functional CD41CD251Foxp31 iTregs could be gener- DCs to iTregs along with membrane fragments, but that a specific ated from CD41CD252 T cells obtained from CD802/2CD862/2 interaction may be involved. DKO mice and WT mice. We found that similar percentages of CD41CD251Foxp31 T cells could be generated from WT and DKO CD80 and CD86 acquisition by iTregs is not CTLA-4-, CD28- or naive CD41 T cells, and on average, 60% of cells in culture were PDL1-dependent CD41CD251Foxp31 (Figure 1a). To evaluate if the DKO iTregs were Tregs express high levels of CTLA-423 which is known to bind to CD80 functional, iTregs were generated from WT and DKO mice (Thy1.21) and CD86, and is critical for Treg function.24 Qureshi et al.14 recently and cocultured with purified naive Thy1.11CD41 T cells at varying showed that CTLA-4 is critical for human and mouse CD41CD251 suppressor to responder ratios (Figure 1b). Suppression of prolifera- Tregs to acquire CD86 from APCs through trans-endocytosis. tion of Thy1.11CD41 T cells was measured by flow cytometry. As However, we found that nTregs isolated from ctla4 gene KO mice shown in Figure 1b and c, both WT and DKO iTregs could profoundly could acquire both CD80 and CD86 and express these acquired mole- suppress the proliferation of responder cells in a dose-dependent man- cules on their surface at the same levels as seen in WT B6 mice ner. The results demonstrate that, unlike that seen for nTregs, CD80 (Figure 4a), suggesting that CTLA-4 is not required for nTregs to and CD86 expression does not affect iTreg generation and function. acquire CD80/CD86 from mDCs. Because CTLA-4 is required for the induction of iTregs,25 it was not possible to examine iTregs from iTregs acquire CD80 and CD86 by trogocytosis after coculture with these mice. Therefore, we assessed the role of CTLA-4 in iTreg trogo- mDCs cytosis by blocking studies. To this end, iTregs were pre-incubated Several reports have shown that effector T cells are able to undergo with anti-CTLA-4 mAbs to block CTLA-4, and then cocultured with trogocytosis of costimulatory molecules after coculture with mDCs. We found that this treatment did not prevent acquisition of APCs.19–22 To determine if iTregs shared this ability, iTregs were CD80/CD86 by iTreg (Figure 4b and d). CD28 is another major mole- cocultured with mDCs and the level of CD80 and CD86 expression cule that is expressed on iTregs and can bind to CD80/CD86, and has was measured on iTregs cultured alone, or after coculture with been shown to be involved in trogocytosis of CD80 by effector T mDCs. To examine CD80 and CD86 acquisition by iTregs, plots cells.19 Again, we found that trogocytosis of CD80 and CD86 still were gated on PI2CD11c2CD41 cells. As shown in Figure 2a, the occurred in the presence of anti-CD28 mAbs (Figure 4c and d). amounts of CD80 and CD86 on WT iTregs were increased after co- Therefore, in our system, the acquisition of CD80 and CD86 incubation with mDCs. To rule out endogenous upregulation of by iTregs occurs by a CTLA-4- and CD28-independent mechanism.

Cellular & Molecular Immunology Trogocytosis of CD80 and CD86 by iTregs PGuet al 139

Figure 1 In vitro generation of CD41CD251Foxp31 T cells from murine naive T cells. (a) The expression of Foxp3 and CD25 was evaluated in CD41CD252 cells isolated from either C57BL/6 (WT) mice or from CD802/2CD862/2 DKO mice, activated for 5 days with plate-bound anti-CD3 Ab, soluble anti-CD28 Ab and IL-2 in the presence of TGF-b. Left: WT iTreg. Right: CD802/2CD862/2 DKO iTreg. (b) Dose-dependent suppression of T-cell proliferation by CD41CD251Foxp31 iTreg cells. CD41CD252 T cells (Thy1.11) were purified from naive B6.PL-Th1a/CyJ mice, labeled with CFSE, and cultured (13105 cells/well) with 1 mg/ml anti-CD3 Ab in the presence of irradiated syngeneic splenocytes (23104 cells/well). Varying numbers of CD41CD251Foxp31 iTreg from WT or DKO C57BL/6 mice were added to the culture at suppressor to responder ratios 1 : 0.5, 1 : 1, 1 : 2, 1 : 4 and 1 : 8. After 4 days of culture, proliferation of responder cells was measured by gating on Thy1.11 cells and assessing their CFSE signal by flow cytometry. Numbers represent the percentage of undivided cells. (2) control: CFSE labeled CD41CD252 T cells without any stimulation or CD41CD251Foxp31 T cells. (1) contol: CFSE labeled CD41CD252 T cells were cultured with anti-CD3 Ab in the presence of irradiated APCs without CD41CD251Foxp31 T cells. (c) Undivided populations were analyzed by determining the suppression ratio obtained with the formula: suppression ratio5(experiment group2positive control)/experiment group3100%. The data shown are representative of at least three different experiments. Ab, antibody; APC, antigen-presenting cell; CFSE, carboxyfluorescein diacetate, succinimidyl ester; DKO, double knockout; iTreg, induced regulatory T cell; WT, wild-type.

Cellular & Molecular Immunology Trogocytosis of CD80 and CD86 by iTregs PGuet al 140

Figure 2 Flow cytometry analysis of CD80 and CD86 acquired by (a)WTiTregsor(b) DKO iTregs. iTregs were isolated from C57BL/6 (WT) mice or from CD802/2CD862/2 DKO mice and were cultured alone or cocultured with mDCs from C57BL/6 mice at a DC/iTreg ratio of 2 : 1. Cells were stained with PI, CD11c, CD4, CD80 and CD86. Shaded graph: CD80 or CD86 expressed on iTregs alone. Line graph: CD80 or CD86 expressed on iTregs after co-incubation with BMDCs. The plots shown are gated on PI2CD11c2CD41 cells and are representative of at least nine different experiments. (c) Confocal microscopy analysis of CD80 and CD86 acquired by DKO iTregs after co- incubation with BMDCs. (i) Green: Alexa Fluor 488-conjugated anti-CD3. (ii) Red: PE-conjugated anti-CD80 (top) or CD86 (bottom). (iii) Merged image of (i) and (ii). BMDC, bone marrow-derived dendritic cell; DC, dendritic cell; DKO, double knockout; iTreg, induced regulatory T cell; mDC, mature dendritic cell; WT, wild-type.

Since PDL1 has been shown to be expressed on Tregs26 and to interact iTregs which acquired CD86 show stronger suppression and increased with CD80,27 we tested the role of this protein in iTreg acquisition. The levelsofactivationmarkersthaniTregswhichdidnotacquireCD86 use of a PDL1 blocking Ab on iTregs prior to coculture with DCs did Next, we determined the consequence of trogocytosis on iTregs. not block acquisition of either CD80 or CD86 (Figure 4d), suggesting Although both CD80 and CD86 could be acquired by DKO iTregs, PDL1 is not involved in our system. Although expression of CTLA-4, CD86 was acquired to a higher extent than CD80 (Figure 3). To better CD28 and PDL1 could be easily detected on iTregs (Figure 4e), none of characterize the differences between the iTregs which acquired and the blocking Abs affected trogocytosis. those that did not acquire CD86, we compared the suppressive

Cellular & Molecular Immunology Trogocytosis of CD80 and CD86 by iTregs PGuet al 141

Figure 3 (a) Dynamics of CD80 and CD86 acquisition by DKO iTregs after co-incubation with BMDCs. DKO iTregs were cultured alone or cocultured with mDCs during a time course from 2 to 24 h. Shaded graph: CD80 or CD86 expressed on DKO iTregs alone. Line graph: CD80 or CD86 expressed on DKO iTregs after coculture with BMDCs. (b) The MFI percentage increase of CD80 and CD86 acquired by DKO iTregs after co-incubation with BMDCs. Percentage increase of MFI5

(MFIafter co-incubation2MFIpre-incubation )/MFIpre-incubation3100%. (c) Amount of PKH26 acquired by DKO iTregs after co-incubation with PKH26-labeled BMDCs. MFI percent change of acquired PKH26 is shown. BMDC, bone marrow-derived dendritic cell; DKO, double knockout; iTreg, induced regulatory T cell; mDC, mature dendritic cell; MFI, median fluorescence intensity. capacity of each group. DKO iTregs were incubated with CD802/2 CD41 T cells as responders and the sorted iTregs (Thy1.21) as sup- mDCs and then sorted according to the level of CD86 acquisition. To pressors. As shown in Figure 5a, both the iTregs which acquired CD86 evaluate the suppressive capacity of each group towards responder T and the iTregs which did not acquire CD86 were able to suppress the cells, a suppression assay was carried out using purified Thy1.11 naive proliferation of responder cells. Interestingly, iTregs which had

Cellular & Molecular Immunology Trogocytosis of CD80 and CD86 by iTregs PGuet al 142

Figure 4 (a) Acquisition of CD80 and CD86 by CTLA-42/2 nTregs. Naturally occurring CD41CD251Foxp31 nTregs were isolated from CTLA-4 KO mice and cultured alone or cocultured with mDCs from C57BL/6 mice. The amount of CD80 and CD86 on nTregs cultured alone (shaded graph) or after coculture with mDCs (line graph) was evaluated by flow cytometry. (b) CD41CD251Foxp31 iTregs were incubated with 20 mg/ml CTLA-4 Ab or (c)20mg/ml CD28 Ab to block the function of either CTLA- 4 or CD28, respectively, 1 h prior to coculture with mDCs. Surface and intracellular levels of CD86 and CD80 on iTregs cultured alone (dark gray shaded graph) or on iTregs cultured with DCs without blocking Ab (solid line), with blocking Ab (dotted line) or with isotype Ab (light gray shaded graph) are shown. Data are representative of 3–5 experiments with three mice per experiment. (d) MFI values of CD86 and CD80 expressed on iTregs after culture with DCs, with and without blocking Abs and isotype (iso) control Abs. The error bars represent SEM. None of the blocking Abs produced a significant change in the level of CD80 and CD86 acquired by iTregs (one- way ANOVA with Bonferroni’s post hoc test). (e) Levels of CTLA-4, CD28 and PDL1 were analyzed on iTregs (line graph). Shaded graph5FMO control. Ab, antibody; CTLA-4, cytotoxic T lymphocyte-associated antigen-4; DC, dendritic cell; FMO, fluorescence-minus-one; iTreg, induced regulatory T cell; KO, knockout; mDC, mature dendritic cell; MFI, median fluorescence intensity; nTreg, natural regulatory T cell; PDL1, programmed death ligand-1.

Cellular & Molecular Immunology Trogocytosis of CD80 and CD86 by iTregs PGuet al 143

Figure 5 (a) Comparison of suppressive capacity between iTregs which had acquired CD86 and iTregs which had not acquired CD86. DKO iTregs were cocultured with CD802/2 mDCs and then sorted according to the level of CD86 acquisition. CD41CD252 T cells (Thy1.11) were purified from naive B6.PL-Th1a/CyJ mice, labeled with CFSE, and cultured (13105 cells/well) with 1 mg/ml anti-CD3 Ab in the presence of irradiated syngeneic splenocytes (23104 cells/well). Varying numbers of the sorted DKO iTreg were added to the culture at suppressor to responder ratios 1 : 1 and 1 : 2. After 4 days of culture, proliferation of responder cells was measured by gating on Thy1.11 cells and assessing their CFSE signal by flow cytometry. Numbers represent the percentage of undivided cells. (2) control: CFSE labeled CD41CD252 T cells without any stimulation or iTregs. (1) contol: CFSE labeled CD41CD252 T cells were cultured with anti-CD3 Ab in the presence of irradiated APC without iTregs. (B) The percent of undivided cells is shown from suppression assays carried out in (a). The data shown are representative of two different experiments, each carried out in duplicate, using 2–3 mice in each experiment. **P,0.05, one-way ANOVA and Bonferroni’s multiple comparison test. Ab, antibody; APC, antigen- presenting cell; CFSE, carboxyfluorescein diacetate, succinimidyl ester; DKO, double knockout; iTreg, induced regulatory T cell; mDC, mature dendritic cell. acquired CD86 displayed a greater suppressive capacity towards CD41 not acquire CD86 (Figure 6). Interestingly, the population of iTregs T cells than iTregs which had not acquired CD86 (Figure 5b). These which showed acquisition of CD86 displayed significantly higher data suggest that acquisition of CD86 from mDCs can enhance iTreg levels of CD44 and CD69 (Figure 6a and c) than the population of suppressive function. iTregs which did not acquire CD86. Levels of CTLA4 were also To examine the basis for the difference in suppressive capacity increased on iTregs which acquired CD86 (Figure 6b) compared observed between the two groups, we compared levels of activation to iTregs which did not acquire CD86; however, this increase did markers between iTregs which acquired CD86 with those that did not reach statistical significance.

Cellular & Molecular Immunology Trogocytosis of CD80 and CD86 by iTregs PGuet al 144

Downregulation of CD80 and CD86 on mDCs after coculture with iTregs nTregs have been shown to be able to inhibit maturation of DCs13 or downregulate CD80 and CD86 expression on mDCs through trans-endocytosis.14 To determine whether iTregs have any effect on mDC CD80 and CD86 expression, iTregs were cocultured with bone marrow-derived, LPS-matured DCs, which express high levels of CD80 and CD86. CD80 and CD86 expression on mDCs was determined by flow cytometry after co-incubation. As shown in Figure 7, significant downregulation of both CD80 and CD86 on mDCs was observed after 24 h of coculture with iTregs. These data indicate that iTregs can downregulate CD80 and CD86 expression on mDCs.

DISCUSSION CD80 and CD86 have been shown to be important for the thymic development and peripheral homeostasis of nTregs in mice.28 In contrast, CD41CD251Foxp31 iTregs could be induced from CD802/2CD862/2 DKO cells to a similar extent as from WT cells (Figure 1a). It was previously shown that Tregs cocultured with mDCs from CD80-/CD86-deficient mice showed reduced survival, and B7 expression on DCs was deemed necessary for Treg mainten- ance.18 However, this same study showed that normal thymic development and peripheral distribution of conventional T cells were not affected in CD80-/CD86-deficient mice. In accordance with this, we were able to generate functional iTregs from CD802/2CD862/2 DKO mice (Figure 1b and c), indicating that, unlike nTregs, the induction and function of iTregs is not dependant on these molecules.

Figure 6 Characterization of activation markers on iTregs following acquisition of CD86. iTregs were cultured alone (i) or cocultured with mDCs from CD45.11 mice Figure 7 Downregulation of CD80 and CD86 expression on DCs by iTregs. BM- (ii), and the level of CD4 and CD86 were analyzed by flow cytometry. Cells were derived LPS-matured DCs from B6 mice, which express high levels of CD80 and gatedoneF4502CD45.12CD41 cells. Cells that stained positive for CD4 only (Q1) CD86, were cultured alone or cocultured together with iTregs from B6 mice at after coculture represent iTregs that did not acquire CD86, while cells that were iTreg/DC ratio52 : 1. After 24 h, cultures were stained with PI, anti-CD4, anti- double positive for CD4 and CD86 (Q2) after coculture represent iTregs that CD11c, anti-CD80 and anti-CD86 mAbs, and then analyzed by flow cytometry. acquired CD86. (iii) The level of (a)CD44,(b)CTLA-4and(c)CD69wasmeasured (a, c) The plots shown are gated on PI2CD42CD11c1 cells (shaded graph: CD80 on iTregs after coculture with mDCs. Histograms show levels of activation markers or CD86 expression on DCs cultured alone; solid line: CD80 or CD86 expression on iTregs that did not acquire CD86 (Q1 from (ii), shaded graph) and on iTregs that on DCs after coculture with iTregs; dotted line: isotype control). These experi- did acquire CD86 (Q2 from (ii), line graph). (iv) The MFI from three independent ments were repeated at least three times each with two or three mice and similar experiments, each with three pooled mice, for activation markers on iTregs that results were obtained. The bar graphs show the MFI of CD80 (b) and CD86 (d). acquired and that did not acquire CD86 is shown. The error bars represent SEM. The error bars represent SEM. 0.01,*P,0.05, Student’s t-test. BM, bone mar- *P,0.05, One-way ANOVA with Bonferroni’s post hoc test. iTreg, induced regu- row; DC, dendritic cell; iTreg, induced regulatory T cell; LPS, lipopolysaccharide; latory T cell; mDC, mature dendritic cell; MFI, median fluorescence intensity. mAb, monoclonal antibody; MFI, median fluorescence intensity.

Cellular & Molecular Immunology Trogocytosis of CD80 and CD86 by iTregs PGuet al 145

Intercellular transfer of cell surface membrane fragments and pro- acquire CD80 and CD86. Because CTLA-4 is required for the induc- teins has been shown to result in altered cell functions,19,29,30 and thus tion of iTregs,25 we could not use CTLA-4-deficient mice to examine could play an important role in modulating immune responses. the role of CTLA-4 in iTreg trogocytosis. Alternatively, we addressed Several cell types, including T and B lymphocytes, natural killer cells this question by blocking studies and found that blocking CTLA-4 on and cd T cells15,16 have been shown to acquire cell surface molecules, iTreg with anti-CTLA-4 Ab did not affect iTreg trogocytosis of CD80 including peptide–major histocompatibility complex (MHC), costi- and CD86 (Figure 4b). In the report by Qureshi et al.,14 the authors mulatory and adhesion molecules.31 However, to the best of our describe a trans-endocytosis of CD80 and CD86 and have deemed this knowledge, the ability of iTregs to undergo this process was not process to be different from trogocytosis, as they did not detect cell known. In this study, we found that after coculture of iTregs with surface acquisition. In our study, however, iTreg acquisition of CD80/ mDCs, iTregs expressed CD80 and CD86 on their surface CD86 was not affected when an endocytosis inhibitor was used (data (Figure 2a). The observed expression of CD80 and CD86 on iTregs not shown), indicating that it is unlikely that endocytosis is a mech- was not due to upregulation of these molecules as iTregs that were anism of acquisition by iTregs. Because the study by Qureshi et al.14 generated from CD802/2CD862/2 DKO mice also expressed these used transfected cells overexpressing costimulatory molecules, the molecules after co-incubation with mDCs (Figure 2b). Confocal mechanisms involved may differ from our system where naturally microscopy for cell surface expression of CD80 and CD86 indicated occurring levels of CD80 and CD86 were examined on primary cells. that both proteins were expressed on small localized areas on iTregs Furthermore, co-incubation with anti-CD28 (Figure 4c) or anti-PDL1 from DKO mice, further confirming trogocytosis as the mechanism of (Figure 4d) Abs, which show clear binding to iTregs (Figure 4e) also expression. These data provide clear evidence that iTregs are capable of failed to block acquisition of CD80 and CD86 by iTregs. These data acquiring CD80 and CD86 from mDCs. Since pretreatment of iTregs suggest that iTregs may acquire CD80 and CD86 via a different mech- with an endocytosis inhibitor did not alter iTreg trogocytosis (data not anism which requires further study. shown), this suggests that the acquired molecules were not interna- The functional consequences of trogocytosis can have a stimulatory lized by this process. Furthermore, we found that CD80 and CD86 or suppressive effect on immune responses depending on the cell type were acquired by iTregs very quickly and iTreg acquisition increased and acquired protein involved. Sabzevari et al.19 have shown that naive from 2 to 24 h of coculture with mDCs. In addition, CD86 was CD41 T cells which acquired CD80 upon activation were capable of acquired to a greater extent by iTregs than CD80 (Figure 3b). The acting as APCs. In another study, CD41 T cells could differentiate into reason for this difference is not known. Although CD80 and CD86 suppressive cells when they acquired HLA-G.35 Previously, we36 and share the same receptors, they have distinct binding affinities and others37 have shown that both mouse and human CD31CD42CD82 kinetics18 and have been shown to have differential functions.32,33 double negative Tregs were able to acquire MHC–peptides from APCs Levels of both proteins were equally expressed on mDCs prior and use the acquired MHC–peptides to trap and kill activated CD41 to coculture (data not shown), ruling out increased availability of or CD81 T effectors in an antigen-specific manner. Here, we found CD86 for trogocytosis. that the ability of iTregs to suppress naive CD41 T-cell proliferation Transfer of cell surface proteins between cells has been shown to was greatly enhanced after iTregs acquired CD86 from mDCs com- occur by various mechanisms, including through several cell surface pared to iTregs that did not acquire CD86 (Figure 5). Thus, trogocy- receptors.22 We found that the kinetics of CD80 and CD86 acquisition tosis by iTregs enhances the suppressive function of these cells. by iTregs were different to that of the membrane dye PKH26. Whereas Because there are likely several cell surface proteins that are transferred the CD80 and CD86 expression levels on iTregs continued to rise, the during trogocytosis, we cannot rule out acquisition of other molecules PKH26 level declined with time (Figure 3c). These data suggest that the in playing a role in increasing iTreg suppression. transfer of CD80 and CD86 from mDCs to iTregs might be mediated To examine the basis for the difference in suppressive capacity through a specific receptor–ligand interaction rather than by passive observed between iTregs which had acquired and iTregs which transfer of membrane fragments. Previous studies have described that had not acquired CD86, several activation markers were examined stimulation of TCR significantly increased T-cell trogocytosis.19,34 We (Figure 6). Levels of activation markers were increased in the group found that addition of anti-CD3 mAb to iTreg–DC coculture resulted of iTregs which had acquired CD86, indicating that these cells were in in only a slight increase in iTreg trogocytosis of CD80 and CD86 (data a more activated functional state. In support of our findings, Rosenits not shown). Since the induction of iTregs with anti-CD28 and anti- et al.38 showed that trogocytosis-positive T cells displayed higher levels CD3 Abs produces already activated T cells, and adding anti-CD3 Ab of CD69 and a more activated phenotype than trogocytosis-negative T during coculture did not have a significant impact on iTreg trogocy- cells. Differences in cell activation status may contribute to the func- tosis, this suggests that anti-CD3 stimulation may be important for T- tional differences in suppressive capacity observed between the two cell activation rather than having a direct effect on T-cell trogocytosis. groups of cells. Although we do not know the exact mechanism of iTregs express high levels of CTLA-4 and CD28, and both can bind trogocytosis-enhanced iTreg suppression in our system, it is possible to CD80 and CD86. There have been conflicting reports regarding the that acquisition of costimulatory molecules may facilitate cell-to-cell role of these receptors in the acquisition of CD80 and CD86 by recipi- contact and lead to enhanced suppression. Zhou et al.39 have demon- ent cells. Sabzevari et al.19 demonstrated a TCR- and CD28-depend- strated that nTregs with acquired peptide–MHC-II molecules show ent, but CTLA-4-independent mechanism of acquisition of CD80 enhanced suppression to target CD41 T cells. They postulated that by CD41 T cells. Onishi et al.13 reported that nTregs downregulate acquired ligands on nTregs can attract target cells to facilitate suppres- CD80/CD86 expression in a CTLA-4- and LFA-1-dependent manner. sion mediated by other effector molecules. In our study, the increased Qureshi et al.14 showed a CD28-independent, but CTLA4-dependent levels of CD44, which plays a role in cell adhesion, on iTregs which had mechanism of acquisition of CD80 and CD86 by both T effectors and acquired CD86, may also contribute to the subsequent enhanced sup- Tregs. We found that nTregs generated from CTLA-4-deficient mice pressive function. were able to acquire high levels of both CD80 and CD86 from mDCs In addition to enhancing iTreg suppressive function, we also (Figure 4a), indicating that CTLA-4 is not required for nTregs to observed a significant downregulation of CD80/CD86 expression

Cellular & Molecular Immunology Trogocytosis of CD80 and CD86 by iTregs PGuet al 146

on DCs after coculture with iTregs (Figure 7). As the levels of 18 Zeng M, Guinet E, Nouri-Shirazi M. B7-1 and B7-2 differentially control peripheral homeostasis of CD41CD251Foxp31 regulatory T cells. Transpl Immunol 2009; 20: surface expression of these costimulatory molecules are critical for 171–179. mDCs to activate T cells, these data suggest that iTregs may suppress 19 Sabzevari H, Kantor J, Jaigirdar A, Tagaya Y, Naramura M, Hodge J et al. Acquisition of immune responses via controlling the function of DCs. Whether CD80 (B7-1) by T cells. J Immunol 2001; 166: 2505–2513. 20 Tatari-Calderone Z, Semnani RT, Nutman TB, Schlom J, Sabzevari H. Acquisition of trogocytosis of CD80/CD86 by iTregs and downregulation of CD80 by human T cells at early stages of activation: functional involvement of CD80 CD80 and CD86 on mDCs is a cause–effect or an epiphenomenon acquisition in T cell to T cell interaction. J Immunol 2002; 169: 6162–6169. 21 Zhou J, Tagaya Y, Tolouei-Semnani R, Schlom J, Sabzevari H. Physiological relevance requires further investigation. of antigen presentasome (APS), an acquired MHC/costimulatory complex, in the In conclusion, the studies reported here demonstrate for the first sustained activation of CD41 T cells in the absence of APCs. Blood 2005; 105: time that iTregs are able to acquire CD80 and CD86 from mDCs in a 3238–3246. 22 Hwang I, Huang JF, Kishimoto H, Brunmark A, Peterson PA, Jackson MR et al. T cells CTLA-4- and CD28-independent manner and the acquired costimu- can use either T cell receptor or CD28 receptors to absorb and internalize cell surface latory molecules confer increased suppressive function and may play molecules derived from antigen-presenting cells. J Exp Med 2000; 191: 1137–1148. an important role in regulating immune responses by iTregs. 23 Takahashi T, Tagami T, Yamazaki S, Uede T, Shimizu J, Sakaguchi N et al. Immunologic self-tolerance maintained by CD251CD41 regulatory T cells constitutively expressing cytotoxic T lymphocyte-associated antigen 4. J Exp Med 2000; 192: 303–310. 24 Wing K, Onishi Y, Prieto-Martin P, Yamaguchi T, Miyara M, Fehervari Z et al. CTLA-4 control over Foxp31 regulatory T cell function. Science 2008; 322: 271–275. 1 Vignali D. How many mechanisms do regulatory T cells need? Eur J Immunol 2008; 25 Zheng SG, Wang JH, Stohl W, Kim KS, Gray JD, Horwitz DA. TGF-beta requires CTLA-4 38: 908–911. early after T cell activation to induce FoxP3 and generate adaptive CD41CD251 2 Curotto de Lafaille MA, Lafaille JJ. Natural and adaptive foxp31 regulatory T cells: regulatory cells. J Immunol 2006; 176: 3321–3329. more of the same or a division of labor? Immunity 2009; 30: 626–635. 26 Sandner SE, Clarkson MR, Salama AD, Sanchez-Fueyo A, Domenig C, Habicht A et al. 3 Zheng SG, Wang JH, Gray JD, Soucier H, Horwitz DA. Natural and induced Role of the programmed death-1 pathway in regulation of alloimmune responses in CD41CD251 cells educate CD41. J Immunol 2004; 172: 5213–5221. vivo. J Immunol 2005; 174: 3408–3415. 4 Fantini MC, Becker C, Monteleone G, Pallone F, Galle PR, Neurath MF. Cutting edge: 27 Butte MJ, Keir ME, Phamduy TB, Sharpe AH, Freeman GJ. Programmed death-1 TGF-beta induces a regulatory phenotype in CD41. J Immunol 2004; 172: 5149– ligand 1 interacts specifically with the B7-1 costimulatory molecule to inhibit T cell 5153. responses. Immunity 2007; 27: 111–122. 5 Chen W, Jin W, Hardegen N, Lei KJ, Li L, Marinos N et al. Conversion of peripheral 28 Salomon B, Lenschow DJ, Rhee L, Ashourian N, Singh B, Sharpe A et al. B7/CD28 CD41. J Exp Med 2003; 198: 1875–1886. costimulation is essential for the homeostasis of the CD41CD251 immunoregulatory 6 Cobbold SP, Castejon R, Adams E, Zelenika D, Graca L, Humm S et al. Induction of T cells that control autoimmune diabetes. Immunity 2000; 12: 431–440. foxP31 regulatory T cells in the periphery of T cell receptor transgenic mice tolerized to transplants. J Immunol 2004; 172: 6003–6010. 29 Cox JH, McMichael AJ, Screaton GR, Xu XN. CTLs target Th cells that acquire 7 Fantini MC, Becker C, Tubbe I, Nikolaev A, Lehr HA, Galle P et al. Transforming growth bystander MHC class I-peptide complex from APCs. J Immunol 2007; 179: 830–836. 1 30 LeMaoult J, Krawice-Radanne I, Dausset J, Carosella ED. HLA-G1-expressing antigen- factor beta induced FoxP3 regulatory T cells suppress Th1 mediated experimental 1 colitis. Gut 2006; 55: 671–680. presenting cells induce immunosuppressive CD4 T cells. Proc Natl Acad Sci USA 8 Fantini MC, Dominitzki S, Rizzo A, Neurath MF, Becker C. In vitro generation of 2004; 101: 7064–7069. CD41CD251 regulatory cells from murine naive T cells. Nat Protoc 2007; 2: 31 Davis DM. Intercellular transfer of cell-surface proteins is common and can affect 1789–1794. many stages of an immune response. Nat Rev Immunol 2007; 7: 238–243. 9 Bluestone JA, Abbas AK. Natural versus adaptive regulatory T cells. Nat Rev Immunol 32 Lumsden JM, Roberts JM, Harris NL, Peach RJ, Ronchese F. Differential requirement 2003; 3: 253–257. for CD80 and CD80/CD86-dependent costimulation in the lung immune response to 10 Acuto O, Michel F. CD28-mediated co-stimulation: a quantitative support for TCR an influenza virus infection. J Immunol 2000; 164: 79–85. signalling. Nat Rev Immunol 2003; 3: 939–951. 33 Balbo P, Silvestri M, Rossi GA, Crimi E, Burastero SE. Differential role of CD80 and 11 Egen JG, Allison JP. Cytotoxic T lymphocyte antigen-4 accumulation in the CD86 on alveolar macrophages in the presentation of allergen to T lymphocytes in immunological synapse is regulated by TCR signal strength. Immunity 2002; 16: asthma. Clin Exp Allergy 2001; 31: 625–636. 23–35. 34 Bahcheli D, Hay V, Nadeau JL, Piccirillo CA. Transfer of cell membrane components 1 1 1 12 Pentcheva-Hoang T, Egen JG, Wojnoonski K, Allison JP. B7-1 and B7-2 selectively via trogocytosis occurs in CD4 Foxp3 CD25 regulatory T-cell contact-dependent recruit CTLA-4 and CD28 to the immunological synapse. Immunity 2004; 21: 401– suppression. Autoimmunity 2011; 44: 607–615. 413. 35 LeMaoult J, Caumartin J, Daouya M, Favier B, Le RS, Gonzalez A et al.Immune 13 Onishi Y, Fehervari Z, Yamaguchi T, Sakaguchi S. Foxp31 natural regulatory T cells regulation by pretenders: cell-to-cell transfers of HLA-G make effector T cells act as preferentially form aggregates on dendritic cells in vitro and actively inhibit their regulatory cells. Blood 2007; 109: 2040–2048. maturation. Proc Natl Acad Sci USA 2008; 105: 10113–10118. 36 Zhang ZX, Yang L, Young KJ, DuTemple B, Zhang L. Identification of a previously 14 Qureshi OS, Zheng Y, Nakamura K, Attridge K, Manzotti C, Schmidt EM et al. Trans- unknown antigen-specific regulatory T cell and its mechanism of suppression. Nat endocytosis of CD80 and CD86: a molecular basis for the cell-extrinsic function of Med 2000; 6: 782–789. CTLA-4. Science 2011; 332: 600–603. 37 Fischer K, Voelkl S, Heymann J, Przybylski GK, Mondal K, Laumer M et al. Isolation 15 Joly E, Hudrisier D. What is trogocytosis and what is its purpose? Nat Immunol 2003; and characterization of human antigen-specific TCR alpha beta1 CD42CD82 double- 4: 815. negative regulatory T cells. Blood 2005; 105: 2828–2835. 16 Ahmed KA, Munegowda MA, Xie Y, Xiang J. Intercellular trogocytosis plays an 38 Rosenits K, Keppler SJ, Vucikuja S, Aichele P. T cells acquire cell surface important role in modulation of immune responses. Cell Mol Immunol 2008; 5: determinants of APC via in vivo trogocytosis during viral infections. Eur J Immunol 261–269. 2010; 40: 3450–3457. 17 Mukherjee S, Maiti PK, Nandi D. Role of CD80, CD86, and CTLA4 on mouse CD41 T 39 Zhou G, Ding ZC, Fu J, Levitsky HI. Presentation of acquired peptide–MHC class II lymphocytes in enhancing cell-cycle progression and survival after activation with ligands by CD41 regulatory T cells or helper cells differentially regulates antigen- PMA and ionomycin. J Leukoc Biol 2002; 72: 921–931. specific CD41 T cell response. J Immunol 2011; 186: 2148–2155.

Cellular & Molecular Immunology