Cutting Edge: In Contrast to Effector T Cells, CD4+CD25+FoxP3+ Regulatory T Cells Are Highly Susceptible to CD95 Ligand- but Not to TCR-Medi
http://www.100md.com
免疫学杂志 2005年第13期
Abstract
+ regulatory T cells (Treg) suppress T cell function and protect rodents from autoimmune disease. Regulation of Treg during an immune response is of major importance. Enhanced survival of Treg is beneficial in autoimmune disease, whereas increased depletion by apoptosis is advantageous in cancer. We show here that freshly isolated FACS-sorted Treg are highly sensitive toward CD95-mediated apoptosis, whereas other T cell populations are resistant to CD95-induced apoptosis shortly after isolation. In contrast, TCR restimulation of Treg in vitro revealed a reduced sensitivity toward activation-induced cell death compared with CD4+CD25– T cells. Thus, the apoptosis phenotype of Treg is unique in comparison to other T cells, and this might be further explored for novel therapeutic modulations of Treg.
Introduction
The function and manipulation of CD4+CD25+FoxP3+ regulatory T cells (Treg)3 during an immune response have gained a lot of attention (1, 2). Depletion of Treg leads to autoimmunity in mice (1), and dysfunction of Treg has been linked to human autoimmune diseases (3, 4, 5). Whereas therapeutical expansion of Treg may be advantageous in autoimmunity, accumulation of immunosuppressive Treg in tumors could be detrimental (6). Depletion of Treg should be beneficial in cancer, but specific therapeutic tools such as Treg-depleting Abs are limited. Thus, new strategies to modulate survival or apoptosis of Treg are warranted.
The investigation of apoptotic properties of Treg is important, because they may be used to modulate the ratio of Treg to effector T cells (Teff). Taams et al. (7) reported a high susceptibility of human Treg to spontaneous cell death or cytokine-deprivation death, and murine Treg can be depleted due to their susceptibility to cyclophosphamide toxicity or gamma irradiation (8, 9). Conversely, other groups reported apoptosis resistance of murine Treg when cells were treated with dexamethasone or anti-CD95 Ab (10, 11). While information on the apoptosis sensitivity of murine Treg is inconsistent, CD95-mediated apoptosis of Treg has not been studied in humans.
Apoptosis mediated by the interaction of CD95 (Apo-1/Fas) with CD95 ligand (CD95L) is well characterized in T cells (12). CD95 is widely expressed (13), whereas expression of CD95L is tightly regulated (14). Although 20–60% of naive CD4+ T cells express CD95, >90% of them are resistant to CD95-mediated apoptosis (15). However, several days after activation, they become sensitive toward CD95-mediated apoptosis and up-regulate CD95L after TCR restimulation (12). Subsequently, CD95L triggers apoptosis of CD95+ activated T cells, a phenomenon called activation-induced cell death (AICD). AICD is a major mechanism to eliminate the expanded pool of effector lymphocytes during the contraction phase of the immune response and to maintain lymphocyte homeostasis (12).
To further clarify the physiology of Treg, we studied their apoptosis phenotype ex vivo. We show for the first time that freshly isolated Treg are highly sensitive toward CD95L-mediated apoptosis unlike their resistant Teff counterparts. In contrast, we find that Treg are substantially less sensitive to AICD than Teff.
Materials and Methods
Abs and reagents
The mAbs against CD4, CD62L, and CD95L (Nok-1) were obtained from BD Pharmingen, and anti-CD25 Ab from Miltenyi Biotec. CD95L was produced as a leucine zipper-tagged ligand of CD95 (15, 16). The anti-CD3 Ab OKT3 and the agonistic monoclonal anti-CD95 Ab (anti-Apo-1) were purified from hybdridoma supernatants by protein A affinity purification (15). The monoclonal anti-FoxP3 Ab was a kind gift from A. Banham (Nuffield Department of Clinical Laboratory Sciences, University of Oxford, Oxford, U.K.). The pan-caspase inhibitor zVAD-fmk was obtained from Bachem, annexin V Alexa Fluor 488 from Molecular Probes, and propidium iodide (PI) and protein A were obtained from Sigma-Aldrich.
Lymphocyte separation
PBL were obtained from healthy individuals. CD4+CD25+ cells were first enriched using MACS beads (Miltenyi Biotec), and subsequently CD4+CD25high cells were sorted with a FACS-Diva cell sorter (BD Biosciences).
Cell culture and cytotoxicity assays
Freshly isolated T cells were cultured in IL-2 (100 IU) containing ex Vivo-15 medium (Cambrex) supplemented with 1% Glutamax (Invitrogen Life Technologies). For apoptosis induction, T cells were stimulated with 1 μg/ml anti-CD95 Ab and 10 ng/ml protein A or 1/10 dilution of CD95L (15). Unstimulated cells were incubated with an isotype control Ab or CD95L-free control medium yielding spontaneous apoptosis of 15–30% in 20 h. To enhance viability of Treg and Teff, some experiments were performed in wells coated with anti-CD3 Ab (30 μg/ml). To measure AICD, T cells were expanded for 6 days (day 6 T cells) with 0.1 μg/ml anti-CD3 Ab and 1 μg/ml anti-CD28 Ab in combination with irradiated JY feeder cells (kind gift from C. Falk, Institute for Molecular Immunology, GSF National Research Center for Environment and Health, Munich, Germany) and 300 IU of IL-2 (3, 17). For induction of AICD, day 6 T cells were transferred into wells containing 30 μg/ml plate-bound anti-CD3 Ab (pb-anti-CD3) and were cultured for 24 h. Cell death was assessed by annexin V/PI costaining and forward- to side-scatter profile. Specific cell death was calculated as described previously (15): (% experimental cell death – % spontaneous cell death)/(100% – % spontaneous cell death) x 100.
Cell surface staining
Cells were stained with PE-labeled anti-CD25 Ab or anti-CD62L Ab, PE-Cy5-labeled anti-CD4 Ab, and FITC-labeled anti-CD95 Ab, or their respective isotype control Abs, and analyzed with a FACScan cytometer with at least 10,000 Treg or Teff counted.
RNA preparation and quantitative RT-PCR
Total RNA was isolated using the Absolutely RNA Microprep kit (Stratagene) and cDNA was prepared using random oligo(dT) primers (Invitrogen Life Technologies). FoxP3 message expression was quantified by detection of incorporated SYBR Green using the ABI Prism 5700 sequence detector system (Applied Biosystems). The relative expression level was determined by normalization to GAPDH with results presented as fold expresssion of Teff mRNA levels. FoxP3 primer sequences were as follows: 5'-AGC TGG AGT TCC GCA AGA AAC (forward) and 5'-TGT TCG TCC ATC CTC CTT TCC (reverse).
Results
Previous reports using magnetic bead-isolated CD4+CD25+ T cells showed CD95 expression on human and murine CD25+ cells (7, 16). We confirmed these data on CD4+ Treg FACS-sorted for very high CD25 expression. These cells induced strong suppression of both T cell proliferation and cytokine production (data not shown) and expressed 100-fold higher FoxP3 levels than Teff cells (see Fig. 2B). Almost all Treg expressed scurfin protein in the nucleus as determined by immunocytochemistry (data not shown). Naive Teff were sorted by gating on CD4+CD25– T cells, which were negative for scurfin expression and showed extremely low FoxP3 levels (see Fig. 2B). Treg expressed higher levels of CD95 molecules on the cell surface as compared with naive Teff (Fig. 1A). Treg remained CD95 positive during short-term (day 6) and long-term (day 20) in vitro expansion (Fig. 1A).
CD4+CD25+FoxP3+ Treg are sensitive to CD95L-induced apoptosis without TCR prestimulation
To test whether freshly isolated Treg are sensitive to CD95-mediated apoptosis without previous TCR stimulation, we incubated FACS-sorted Treg and Teff with cross-linked anti-CD95 Ab (anti-Apo-1) for 20 h. As reported previously (12), freshly isolated Teff did not die upon addition of anti-CD95 Ab or soluble CD95L. However, Treg treated under the same conditions showed a very high induction of apoptosis to both stimuli (Fig. 1B). We have carefully titrated both CD95L (Fig. 1C) and anti-CD95 Ab and performed experiments side by side with both reagents. Although both reagents yielded similar results, soluble CD95L might be the more physiological mimic to trigger CD95-mediated apoptosis than anti-CD95 Ab. After 4–6 days of TCR stimulation in vitro CD4+ T cells are usually sensitive toward the extrinsic pathway of apoptosis, which is initiated by the binding of CD95L to CD95 and is then executed by a cascade of caspases (12). To test whether CD95L-induced apoptosis of freshly isolated Treg involves these events, we used a neutralizing Ab against CD95L as well as zVAD-fmk as an inhibitor of caspase activity. Both treatments blocked CD95-mediated apoptosis in a concentration-dependent manner (Fig. 1D). Once apoptosis has been triggered, annexin V staining and uptake of PI allows distinguishing between early apoptotic and late apoptotic or necrotic cells. Apoptotic cell death started within the first hour after CD95 triggering and led to 30% early apoptotic cells (annexin+PI–) by 10 h (Fig. 1, E and F). After 20 h, most of the apoptotic cells had lost membrane integrity (annexin V+PI+), resulting in a total of 40–60% specific cell death, which did not increase further until 48 h (data not shown). Similar observations were also made with freshly isolated CD4+CD25+ T cells from murine spleen and lymph nodes demonstrating a consistent phenotype between murine and human Treg (data not shown).
Sensitivity to CD95L-induced apoptosis without TCR prestimulation is a unique feature of CD4+CD25+FoxP3+ Treg
Using high-speed four-way FACS, we simultaneously sorted four different subpopulations from peripheral blood: CD25– Teff, intermediate T cells with low CD25+ expression (Tint CD25+), intermediate T cells with high CD25+ expression (Tint CD25++), and Treg with the highest CD25 expression (Fig. 2A). Quantitative PCR analysis revealed very high FoxP3 mRNA expression in Treg. Low amounts of FoxP3 mRNA were also detectable in the adjacent Tint CD25++ subpopulation, whereas Tint CD25+ and CD25– Teff cells were essentially FoxP3– (Fig. 2B). Similarly, only the Treg subpopulation showed significant suppressive capacity and anergy as determined by proliferation assays (data not shown). Among the four freshly isolated T cell subpopulations, CD95-induced apoptosis was limited to cells with the highest expression of Treg markers (CD25, FoxP3) (Fig. 2C). We suggest that only the CD25++ cells are Treg and that cells within the Tint CD25++ sort gate comprise a mixture of Treg and activated Teff. This "contamination" of Treg in the Tint CD25++ subpopulation could explain the small increase in apoptosis upon CD95 cross-linking.
AICD is diminished in CD4+CD25+FoxP3+ Treg
When previously activated lymphocytes encounter a second TCR signal, they express CD95L and kill each other by AICD. To test whether Treg undergo AICD, we prestimulated CD25++ cells with anti-CD3/anti-CD28 Abs in combination with FcR-bearing, cross-linking feeder cells and IL-2 (3, 17). This activation protocol allowed proliferation of Treg and resulted in expansion of Treg within 6 days of stimulation. At day 6, Treg were restimulated by pb-anti-CD3 Ab for 24 h. Surprisingly, Treg were significantly less sensitive to AICD than Teff, although both cell types showed similar apoptosis sensitivity toward CD95L treatment (Fig. 3B). Because freshly isolated Treg died in response to CD95 stimulation, we tested whether they would also be killed by pb-anti-CD3 Ab. However, stimulation of day 0 Treg with pb-anti-CD3 Ab for 24 h yielded even less cell death compared with Treg cultured in IL-2 medium alone (Fig. 3A). Next, we tested Treg expanded for 20 days (100- to 1000-fold expansion) before restimulation with pb-anti-CD3 Ab. We repeatedly observed CD62L+ and CD62L– cells in anti-CD3/anti-CD28-expanded Treg cultures, which were both suppressive. Restimulation of both subpopulations with pb-anti-CD3 Ab did not induce AICD, whereas CD95 killing was preserved (Fig. 3C). In summary, we found that freshly isolated Treg are highly sensitive to CD95 cross-linking in contrast to their resistant CD4+CD25– Teff counterparts. However Treg are less sensitive to AICD compared with Teff.
Discussion
Upon Ag encounter, T cells are activated, proliferate, and exert their effector functions as Teff (12). Once the Ag is cleared, most of the Teff die and only a few T cells are thought to remain as memory cells. AICD is one of the main mechanisms for T cell contraction in vivo (12, 18). Freshly isolated naive Teff are resistant to CD95-mediated apoptosis in vitro. Four to 6 days after TCR stimulation in vitro, Teff become sensitive toward CD95-mediated apoptosis induced by CD95L or TCR restimulation (12). In contrast, we show for the first time that freshly isolated human Treg are highly sensitive to CD95-mediated apoptosis. We demonstrate that sensitivity of such T cells is a unique feature of CD25++ Treg within the CD4+ T cell compartment.
Papiernik and colleagues (11) observed a resistance of prestimulated CD4+CD25+ T cells derived from C57BL/6 mice toward apoptosis triggered by anti-CD95 Ab. However, this study included neither freshly isolated FACS-sorted T cells nor human T cells. Given the relatively high percentage of Treg dying spontaneously even under conditions of Treg proliferation and expansion (data not shown), some of the CD95-sensitive CD4+CD25+ cells might have died during their prestimulation phase. This might lead to an underestimation of CD95-triggered cell death. In addition, we used CD95L, which might bind and multimerize CD95 more efficiently than the anti-murine CD95 Ab used by Papiernik and colleagues (11). We consistently observed sensitivity of Treg toward CD95-triggered apoptosis in both freshly isolated human Treg (day 0) and short-term-stimulated Treg (days 2–6) as well as in long-term-expanded Treg (day 20).
In vivo, the apparent selective sensitivity of Treg to CD95L might be an important mechanism to eliminate Treg during the acute effector phase of an immune response at a time when Teff are resistant to CD95-mediated apoptosis. In particular, in the presence of a fulminant acute infection, Treg could be detrimental to the organism and a rapid elimination of Treg in danger situations seems appropriate. In this case, Treg might be killed either by soluble CD95L or adjacent membrane-bound CD95L expressed, e.g., on dying infected cells. Reduced elimination of Treg during acute infection might hamper Teff cells to clear the infection as recently shown in a model of leishmaniasis (19, 20).
Our second finding demonstrates a relative resistance of human Treg to AICD. Neither freshly isolated Treg nor Treg activated for 6 days or expanded for 3 wk showed significant AICD upon TCR stimulation or restimulation. In contrast, activated Teff are AICD sensitive (12) and thus die due to the weekly TCR restimulation (3, 21), whereas Treg expand (data not shown). This observation further supports a reduced AICD sensitivity of Treg. Interestingly, Treg also express CD95L mRNA upon TCR restimulation (data not shown). Further studies including quantitative analysis of CD95L mRNA expression and CD95L protein expression might clarify the reason for reduced AICD sensitivity of Treg. Other mechanisms including proteolytical cleavage of CD95L and CD95 into antagonistic products could further explain the relative AICD resistance of Treg (22, 23).
Because T cell responses to pathogens could not only prime T cells for pathogen elimination, but also induce autoaggressive Teff, prolonged survival of AICD-resistant Treg in the critical down-phase of an immune response is conceivable. In support of this idea, Treg have been described to escape from clonal deletion induced by Ag-specific (re)stimulation in vivo (24, 25, 26).
Even in the absence of AICD, Treg numbers might be contracted during the down-phase of the immune response. First, the rapid decline of Teff cells leads to much lower IL-2 levels, which may become limiting, and thus eliminates surplus Treg via death by cytokine deprivation (27). Second, our data offer the possibility that CD95L-expressing Teff might kill neighboring Treg. Although still hypothetical, Teff might keep Treg in check by modulating their survival, whereas Treg might mainly suppress effector function of Teff, resulting in a tightly balanced ratio of both T cell populations. Obviously, further work is warranted to explore the use of the CD95/CD95L system for specifically modulating Treg cell number and function in vivo.
Acknowledgments
Drs. A. Banham, G. Roncador, G. Moldenhauer, and H. Weyd are gratefully acknowledged for their generous gift of Abs, and C. Falk for JY cells. We thank B. Franz, J. Mohr, and Dr. J. Sykora for helping to establish the immunocytochemistry protocol for scurfin; S. Prinz, M. Rutz, I. Hefft, K. Hexel, and M. Scheuermann for technical assistance; Drs. R. Arnold, K. Gu"low, and A. Golks, E. Kleinmann, and E. Pauly for critical reading of the manuscript; C. Frey and C. Fritsch for computational assistance; and Dr. M. Korporal for helpful discussion.
Disclosures
The authors have no financial conflict of interest.
Footnotes
The costs of publication of this article were defrayed in 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.
1 This work was supported by grants from the Gemeinnützige Hertie-Stiftung (1.319.110/01/11 and 1.01.1/04/003), Biogen GmbH, Landesstiftung Baden-Wuerttemberg, and Serono GmbH, and by the Young Investigator Award of the Faculty of Medicine, University of Heidelberg (to B.F.).
2 Address correspondence and reprint requests to Dr. Benedikt Fritzsching or Dr. Peter H. Krammer, Tumor Immunology Program D030, German Cancer Research Center, Im Neuenheimer Feld 280, 69120 Heidelberg, Germany. E-mail addresses: Benedikt.Fritzsching@dkfz.de and P.Krammer@dkfz.de
3 Abbreviations used in this paper: Treg, regulatory T cell; Teff, effector T cell; CD95L, CD95 ligand; AICD, activation-induced cell death; PI, propidium iodide; pb-anti-CD3, plate-bound anti-CD3 Ab.
Received for publication January 24, 2005. Accepted for publication April 27, 2005.
References
Shevach, E. M.. 2000. Regulatory T cells in autoimmunity. Annu. Rev. Immunol. 18: 423-449.
Bach, J. F.. 1995. Organ-specific autoimmunity. Immunol. Today 16: 353-355.
Viglietta, V., C. Baecher-Allan, H. L. Weiner, D. A. Hafler. 2004. Loss of functional suppression by CD4+CD25+ regulatory T cells in patients with multiple sclerosis. J. Exp. Med. 199: 971-979.、
Ehrenstein, M. R., J. G. Evans, A. Singh, S. Moore, G. Warnes, D. A. Isenberg, C. Mauri. 2004. Compromised function of regulatory T cells in rheumatoid arthritis and reversal by anti-TNF therapy. J. Exp. Med. 200: 277-285.、
Kriegel, M. A., T. Lohmann, C. Gabler, N. Blank, J. R. Kalden, H. M. Lorenz. 2004. Defective suppressor function of human CD4+CD25+ regulatory T cells in autoimmune polyglandular syndrome type II. J. Exp. Med. 199: 1285-1291.
Curiel, T. J., G. Coukos, L. Zou, X. Alvarez, P. Cheng, P. Mottram, M. Evdemon-Hogan, J. R. Conejo-Garcia, L. Zhang, M. Burow, et al 2004. Specific recruitment of regulatory T cells in ovarian carcinoma fosters immune privilege and predicts reduced survival. Nat. Med. 10: 942-949.
Taams, L. S., J. Smith, M. H. Rustin, M. Salmon, L. W. Poulter, A. N. Akbar. 2001. Human anergic/suppressive CD4+CD25+ T cells: a highly differentiated and apoptosis-prone population. Eur. J. Immunol. 31: 1122-1131.
Ghiringhelli, F., N. Larmonier, E. Schmitt, A. Parcellier, D. Cathelin, C. Garrido, B. Chauffert, E. Solary, B. Bonnotte, F. Martin. 2004. CD4+CD25+ regulatory T cells suppress tumor immunity but are sensitive to cyclophosphamide which allows immunotherapy of established tumors to be curative. Eur. J. Immunol. 34: 336-344.(Benedikt Fritzsching2,*,,)
+ regulatory T cells (Treg) suppress T cell function and protect rodents from autoimmune disease. Regulation of Treg during an immune response is of major importance. Enhanced survival of Treg is beneficial in autoimmune disease, whereas increased depletion by apoptosis is advantageous in cancer. We show here that freshly isolated FACS-sorted Treg are highly sensitive toward CD95-mediated apoptosis, whereas other T cell populations are resistant to CD95-induced apoptosis shortly after isolation. In contrast, TCR restimulation of Treg in vitro revealed a reduced sensitivity toward activation-induced cell death compared with CD4+CD25– T cells. Thus, the apoptosis phenotype of Treg is unique in comparison to other T cells, and this might be further explored for novel therapeutic modulations of Treg.
Introduction
The function and manipulation of CD4+CD25+FoxP3+ regulatory T cells (Treg)3 during an immune response have gained a lot of attention (1, 2). Depletion of Treg leads to autoimmunity in mice (1), and dysfunction of Treg has been linked to human autoimmune diseases (3, 4, 5). Whereas therapeutical expansion of Treg may be advantageous in autoimmunity, accumulation of immunosuppressive Treg in tumors could be detrimental (6). Depletion of Treg should be beneficial in cancer, but specific therapeutic tools such as Treg-depleting Abs are limited. Thus, new strategies to modulate survival or apoptosis of Treg are warranted.
The investigation of apoptotic properties of Treg is important, because they may be used to modulate the ratio of Treg to effector T cells (Teff). Taams et al. (7) reported a high susceptibility of human Treg to spontaneous cell death or cytokine-deprivation death, and murine Treg can be depleted due to their susceptibility to cyclophosphamide toxicity or gamma irradiation (8, 9). Conversely, other groups reported apoptosis resistance of murine Treg when cells were treated with dexamethasone or anti-CD95 Ab (10, 11). While information on the apoptosis sensitivity of murine Treg is inconsistent, CD95-mediated apoptosis of Treg has not been studied in humans.
Apoptosis mediated by the interaction of CD95 (Apo-1/Fas) with CD95 ligand (CD95L) is well characterized in T cells (12). CD95 is widely expressed (13), whereas expression of CD95L is tightly regulated (14). Although 20–60% of naive CD4+ T cells express CD95, >90% of them are resistant to CD95-mediated apoptosis (15). However, several days after activation, they become sensitive toward CD95-mediated apoptosis and up-regulate CD95L after TCR restimulation (12). Subsequently, CD95L triggers apoptosis of CD95+ activated T cells, a phenomenon called activation-induced cell death (AICD). AICD is a major mechanism to eliminate the expanded pool of effector lymphocytes during the contraction phase of the immune response and to maintain lymphocyte homeostasis (12).
To further clarify the physiology of Treg, we studied their apoptosis phenotype ex vivo. We show for the first time that freshly isolated Treg are highly sensitive toward CD95L-mediated apoptosis unlike their resistant Teff counterparts. In contrast, we find that Treg are substantially less sensitive to AICD than Teff.
Materials and Methods
Abs and reagents
The mAbs against CD4, CD62L, and CD95L (Nok-1) were obtained from BD Pharmingen, and anti-CD25 Ab from Miltenyi Biotec. CD95L was produced as a leucine zipper-tagged ligand of CD95 (15, 16). The anti-CD3 Ab OKT3 and the agonistic monoclonal anti-CD95 Ab (anti-Apo-1) were purified from hybdridoma supernatants by protein A affinity purification (15). The monoclonal anti-FoxP3 Ab was a kind gift from A. Banham (Nuffield Department of Clinical Laboratory Sciences, University of Oxford, Oxford, U.K.). The pan-caspase inhibitor zVAD-fmk was obtained from Bachem, annexin V Alexa Fluor 488 from Molecular Probes, and propidium iodide (PI) and protein A were obtained from Sigma-Aldrich.
Lymphocyte separation
PBL were obtained from healthy individuals. CD4+CD25+ cells were first enriched using MACS beads (Miltenyi Biotec), and subsequently CD4+CD25high cells were sorted with a FACS-Diva cell sorter (BD Biosciences).
Cell culture and cytotoxicity assays
Freshly isolated T cells were cultured in IL-2 (100 IU) containing ex Vivo-15 medium (Cambrex) supplemented with 1% Glutamax (Invitrogen Life Technologies). For apoptosis induction, T cells were stimulated with 1 μg/ml anti-CD95 Ab and 10 ng/ml protein A or 1/10 dilution of CD95L (15). Unstimulated cells were incubated with an isotype control Ab or CD95L-free control medium yielding spontaneous apoptosis of 15–30% in 20 h. To enhance viability of Treg and Teff, some experiments were performed in wells coated with anti-CD3 Ab (30 μg/ml). To measure AICD, T cells were expanded for 6 days (day 6 T cells) with 0.1 μg/ml anti-CD3 Ab and 1 μg/ml anti-CD28 Ab in combination with irradiated JY feeder cells (kind gift from C. Falk, Institute for Molecular Immunology, GSF National Research Center for Environment and Health, Munich, Germany) and 300 IU of IL-2 (3, 17). For induction of AICD, day 6 T cells were transferred into wells containing 30 μg/ml plate-bound anti-CD3 Ab (pb-anti-CD3) and were cultured for 24 h. Cell death was assessed by annexin V/PI costaining and forward- to side-scatter profile. Specific cell death was calculated as described previously (15): (% experimental cell death – % spontaneous cell death)/(100% – % spontaneous cell death) x 100.
Cell surface staining
Cells were stained with PE-labeled anti-CD25 Ab or anti-CD62L Ab, PE-Cy5-labeled anti-CD4 Ab, and FITC-labeled anti-CD95 Ab, or their respective isotype control Abs, and analyzed with a FACScan cytometer with at least 10,000 Treg or Teff counted.
RNA preparation and quantitative RT-PCR
Total RNA was isolated using the Absolutely RNA Microprep kit (Stratagene) and cDNA was prepared using random oligo(dT) primers (Invitrogen Life Technologies). FoxP3 message expression was quantified by detection of incorporated SYBR Green using the ABI Prism 5700 sequence detector system (Applied Biosystems). The relative expression level was determined by normalization to GAPDH with results presented as fold expresssion of Teff mRNA levels. FoxP3 primer sequences were as follows: 5'-AGC TGG AGT TCC GCA AGA AAC (forward) and 5'-TGT TCG TCC ATC CTC CTT TCC (reverse).
Results
Previous reports using magnetic bead-isolated CD4+CD25+ T cells showed CD95 expression on human and murine CD25+ cells (7, 16). We confirmed these data on CD4+ Treg FACS-sorted for very high CD25 expression. These cells induced strong suppression of both T cell proliferation and cytokine production (data not shown) and expressed 100-fold higher FoxP3 levels than Teff cells (see Fig. 2B). Almost all Treg expressed scurfin protein in the nucleus as determined by immunocytochemistry (data not shown). Naive Teff were sorted by gating on CD4+CD25– T cells, which were negative for scurfin expression and showed extremely low FoxP3 levels (see Fig. 2B). Treg expressed higher levels of CD95 molecules on the cell surface as compared with naive Teff (Fig. 1A). Treg remained CD95 positive during short-term (day 6) and long-term (day 20) in vitro expansion (Fig. 1A).
CD4+CD25+FoxP3+ Treg are sensitive to CD95L-induced apoptosis without TCR prestimulation
To test whether freshly isolated Treg are sensitive to CD95-mediated apoptosis without previous TCR stimulation, we incubated FACS-sorted Treg and Teff with cross-linked anti-CD95 Ab (anti-Apo-1) for 20 h. As reported previously (12), freshly isolated Teff did not die upon addition of anti-CD95 Ab or soluble CD95L. However, Treg treated under the same conditions showed a very high induction of apoptosis to both stimuli (Fig. 1B). We have carefully titrated both CD95L (Fig. 1C) and anti-CD95 Ab and performed experiments side by side with both reagents. Although both reagents yielded similar results, soluble CD95L might be the more physiological mimic to trigger CD95-mediated apoptosis than anti-CD95 Ab. After 4–6 days of TCR stimulation in vitro CD4+ T cells are usually sensitive toward the extrinsic pathway of apoptosis, which is initiated by the binding of CD95L to CD95 and is then executed by a cascade of caspases (12). To test whether CD95L-induced apoptosis of freshly isolated Treg involves these events, we used a neutralizing Ab against CD95L as well as zVAD-fmk as an inhibitor of caspase activity. Both treatments blocked CD95-mediated apoptosis in a concentration-dependent manner (Fig. 1D). Once apoptosis has been triggered, annexin V staining and uptake of PI allows distinguishing between early apoptotic and late apoptotic or necrotic cells. Apoptotic cell death started within the first hour after CD95 triggering and led to 30% early apoptotic cells (annexin+PI–) by 10 h (Fig. 1, E and F). After 20 h, most of the apoptotic cells had lost membrane integrity (annexin V+PI+), resulting in a total of 40–60% specific cell death, which did not increase further until 48 h (data not shown). Similar observations were also made with freshly isolated CD4+CD25+ T cells from murine spleen and lymph nodes demonstrating a consistent phenotype between murine and human Treg (data not shown).
Sensitivity to CD95L-induced apoptosis without TCR prestimulation is a unique feature of CD4+CD25+FoxP3+ Treg
Using high-speed four-way FACS, we simultaneously sorted four different subpopulations from peripheral blood: CD25– Teff, intermediate T cells with low CD25+ expression (Tint CD25+), intermediate T cells with high CD25+ expression (Tint CD25++), and Treg with the highest CD25 expression (Fig. 2A). Quantitative PCR analysis revealed very high FoxP3 mRNA expression in Treg. Low amounts of FoxP3 mRNA were also detectable in the adjacent Tint CD25++ subpopulation, whereas Tint CD25+ and CD25– Teff cells were essentially FoxP3– (Fig. 2B). Similarly, only the Treg subpopulation showed significant suppressive capacity and anergy as determined by proliferation assays (data not shown). Among the four freshly isolated T cell subpopulations, CD95-induced apoptosis was limited to cells with the highest expression of Treg markers (CD25, FoxP3) (Fig. 2C). We suggest that only the CD25++ cells are Treg and that cells within the Tint CD25++ sort gate comprise a mixture of Treg and activated Teff. This "contamination" of Treg in the Tint CD25++ subpopulation could explain the small increase in apoptosis upon CD95 cross-linking.
AICD is diminished in CD4+CD25+FoxP3+ Treg
When previously activated lymphocytes encounter a second TCR signal, they express CD95L and kill each other by AICD. To test whether Treg undergo AICD, we prestimulated CD25++ cells with anti-CD3/anti-CD28 Abs in combination with FcR-bearing, cross-linking feeder cells and IL-2 (3, 17). This activation protocol allowed proliferation of Treg and resulted in expansion of Treg within 6 days of stimulation. At day 6, Treg were restimulated by pb-anti-CD3 Ab for 24 h. Surprisingly, Treg were significantly less sensitive to AICD than Teff, although both cell types showed similar apoptosis sensitivity toward CD95L treatment (Fig. 3B). Because freshly isolated Treg died in response to CD95 stimulation, we tested whether they would also be killed by pb-anti-CD3 Ab. However, stimulation of day 0 Treg with pb-anti-CD3 Ab for 24 h yielded even less cell death compared with Treg cultured in IL-2 medium alone (Fig. 3A). Next, we tested Treg expanded for 20 days (100- to 1000-fold expansion) before restimulation with pb-anti-CD3 Ab. We repeatedly observed CD62L+ and CD62L– cells in anti-CD3/anti-CD28-expanded Treg cultures, which were both suppressive. Restimulation of both subpopulations with pb-anti-CD3 Ab did not induce AICD, whereas CD95 killing was preserved (Fig. 3C). In summary, we found that freshly isolated Treg are highly sensitive to CD95 cross-linking in contrast to their resistant CD4+CD25– Teff counterparts. However Treg are less sensitive to AICD compared with Teff.
Discussion
Upon Ag encounter, T cells are activated, proliferate, and exert their effector functions as Teff (12). Once the Ag is cleared, most of the Teff die and only a few T cells are thought to remain as memory cells. AICD is one of the main mechanisms for T cell contraction in vivo (12, 18). Freshly isolated naive Teff are resistant to CD95-mediated apoptosis in vitro. Four to 6 days after TCR stimulation in vitro, Teff become sensitive toward CD95-mediated apoptosis induced by CD95L or TCR restimulation (12). In contrast, we show for the first time that freshly isolated human Treg are highly sensitive to CD95-mediated apoptosis. We demonstrate that sensitivity of such T cells is a unique feature of CD25++ Treg within the CD4+ T cell compartment.
Papiernik and colleagues (11) observed a resistance of prestimulated CD4+CD25+ T cells derived from C57BL/6 mice toward apoptosis triggered by anti-CD95 Ab. However, this study included neither freshly isolated FACS-sorted T cells nor human T cells. Given the relatively high percentage of Treg dying spontaneously even under conditions of Treg proliferation and expansion (data not shown), some of the CD95-sensitive CD4+CD25+ cells might have died during their prestimulation phase. This might lead to an underestimation of CD95-triggered cell death. In addition, we used CD95L, which might bind and multimerize CD95 more efficiently than the anti-murine CD95 Ab used by Papiernik and colleagues (11). We consistently observed sensitivity of Treg toward CD95-triggered apoptosis in both freshly isolated human Treg (day 0) and short-term-stimulated Treg (days 2–6) as well as in long-term-expanded Treg (day 20).
In vivo, the apparent selective sensitivity of Treg to CD95L might be an important mechanism to eliminate Treg during the acute effector phase of an immune response at a time when Teff are resistant to CD95-mediated apoptosis. In particular, in the presence of a fulminant acute infection, Treg could be detrimental to the organism and a rapid elimination of Treg in danger situations seems appropriate. In this case, Treg might be killed either by soluble CD95L or adjacent membrane-bound CD95L expressed, e.g., on dying infected cells. Reduced elimination of Treg during acute infection might hamper Teff cells to clear the infection as recently shown in a model of leishmaniasis (19, 20).
Our second finding demonstrates a relative resistance of human Treg to AICD. Neither freshly isolated Treg nor Treg activated for 6 days or expanded for 3 wk showed significant AICD upon TCR stimulation or restimulation. In contrast, activated Teff are AICD sensitive (12) and thus die due to the weekly TCR restimulation (3, 21), whereas Treg expand (data not shown). This observation further supports a reduced AICD sensitivity of Treg. Interestingly, Treg also express CD95L mRNA upon TCR restimulation (data not shown). Further studies including quantitative analysis of CD95L mRNA expression and CD95L protein expression might clarify the reason for reduced AICD sensitivity of Treg. Other mechanisms including proteolytical cleavage of CD95L and CD95 into antagonistic products could further explain the relative AICD resistance of Treg (22, 23).
Because T cell responses to pathogens could not only prime T cells for pathogen elimination, but also induce autoaggressive Teff, prolonged survival of AICD-resistant Treg in the critical down-phase of an immune response is conceivable. In support of this idea, Treg have been described to escape from clonal deletion induced by Ag-specific (re)stimulation in vivo (24, 25, 26).
Even in the absence of AICD, Treg numbers might be contracted during the down-phase of the immune response. First, the rapid decline of Teff cells leads to much lower IL-2 levels, which may become limiting, and thus eliminates surplus Treg via death by cytokine deprivation (27). Second, our data offer the possibility that CD95L-expressing Teff might kill neighboring Treg. Although still hypothetical, Teff might keep Treg in check by modulating their survival, whereas Treg might mainly suppress effector function of Teff, resulting in a tightly balanced ratio of both T cell populations. Obviously, further work is warranted to explore the use of the CD95/CD95L system for specifically modulating Treg cell number and function in vivo.
Acknowledgments
Drs. A. Banham, G. Roncador, G. Moldenhauer, and H. Weyd are gratefully acknowledged for their generous gift of Abs, and C. Falk for JY cells. We thank B. Franz, J. Mohr, and Dr. J. Sykora for helping to establish the immunocytochemistry protocol for scurfin; S. Prinz, M. Rutz, I. Hefft, K. Hexel, and M. Scheuermann for technical assistance; Drs. R. Arnold, K. Gu"low, and A. Golks, E. Kleinmann, and E. Pauly for critical reading of the manuscript; C. Frey and C. Fritsch for computational assistance; and Dr. M. Korporal for helpful discussion.
Disclosures
The authors have no financial conflict of interest.
Footnotes
The costs of publication of this article were defrayed in 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.
1 This work was supported by grants from the Gemeinnützige Hertie-Stiftung (1.319.110/01/11 and 1.01.1/04/003), Biogen GmbH, Landesstiftung Baden-Wuerttemberg, and Serono GmbH, and by the Young Investigator Award of the Faculty of Medicine, University of Heidelberg (to B.F.).
2 Address correspondence and reprint requests to Dr. Benedikt Fritzsching or Dr. Peter H. Krammer, Tumor Immunology Program D030, German Cancer Research Center, Im Neuenheimer Feld 280, 69120 Heidelberg, Germany. E-mail addresses: Benedikt.Fritzsching@dkfz.de and P.Krammer@dkfz.de
3 Abbreviations used in this paper: Treg, regulatory T cell; Teff, effector T cell; CD95L, CD95 ligand; AICD, activation-induced cell death; PI, propidium iodide; pb-anti-CD3, plate-bound anti-CD3 Ab.
Received for publication January 24, 2005. Accepted for publication April 27, 2005.
References
Shevach, E. M.. 2000. Regulatory T cells in autoimmunity. Annu. Rev. Immunol. 18: 423-449.
Bach, J. F.. 1995. Organ-specific autoimmunity. Immunol. Today 16: 353-355.
Viglietta, V., C. Baecher-Allan, H. L. Weiner, D. A. Hafler. 2004. Loss of functional suppression by CD4+CD25+ regulatory T cells in patients with multiple sclerosis. J. Exp. Med. 199: 971-979.、
Ehrenstein, M. R., J. G. Evans, A. Singh, S. Moore, G. Warnes, D. A. Isenberg, C. Mauri. 2004. Compromised function of regulatory T cells in rheumatoid arthritis and reversal by anti-TNF therapy. J. Exp. Med. 200: 277-285.、
Kriegel, M. A., T. Lohmann, C. Gabler, N. Blank, J. R. Kalden, H. M. Lorenz. 2004. Defective suppressor function of human CD4+CD25+ regulatory T cells in autoimmune polyglandular syndrome type II. J. Exp. Med. 199: 1285-1291.
Curiel, T. J., G. Coukos, L. Zou, X. Alvarez, P. Cheng, P. Mottram, M. Evdemon-Hogan, J. R. Conejo-Garcia, L. Zhang, M. Burow, et al 2004. Specific recruitment of regulatory T cells in ovarian carcinoma fosters immune privilege and predicts reduced survival. Nat. Med. 10: 942-949.
Taams, L. S., J. Smith, M. H. Rustin, M. Salmon, L. W. Poulter, A. N. Akbar. 2001. Human anergic/suppressive CD4+CD25+ T cells: a highly differentiated and apoptosis-prone population. Eur. J. Immunol. 31: 1122-1131.
Ghiringhelli, F., N. Larmonier, E. Schmitt, A. Parcellier, D. Cathelin, C. Garrido, B. Chauffert, E. Solary, B. Bonnotte, F. Martin. 2004. CD4+CD25+ regulatory T cells suppress tumor immunity but are sensitive to cyclophosphamide which allows immunotherapy of established tumors to be curative. Eur. J. Immunol. 34: 336-344.(Benedikt Fritzsching2,*,,)