JOURNAL TRANSCRIPT
Immunity, Vol. 14, 715–725, June, 2001, Copyright 2001 by Cell Press
TGF- Released by Apoptotic T Cells Contributes to an Immunosuppressive Milieu WanJun Chen,1 Mark E. Frank, Wenwen Jin, and Sharon M. Wahl1 Cellular Immunology Section Oral Infection and Immunity Branch National Institute of Dental and Craniofacial Research National Institutes of Health Bethesda, Maryland 20892
Summary T cell apoptosis is critical to development and homeostasis of the immune system. The most salient feature of apoptosis is the lack of an attendant inflammatory response or tissue damage. Here, we present evidence that apoptotic T cells release TGF-, thereby contributing to an immunosuppressive milieu. Apoptotic T cells released not only latent but also bio-active TGF-. Nonetheless, TGF- transcription was not upregulated, suggesting release of existing rather than synthesis of new TGF-. Localized within the intracellular membrane-bound compartment, which includes mitochondria, TGF- was redistributed into the cytosol following loss of mitochondrial membrane potential. TGF- secreted from apoptotic T cells inhibited proinflammatory cytokine production by activated macrophages to foster immune suppression. These findings broaden the potential mechanisms whereby induction of immune tolerance or deficiency occurs through T cell deletion. Introduction T cell apoptosis is critical to the development and homeostasis of the immune system (Lenardo et al., 1999). Although deletion of antigen-specific T cells has been considered as one of the most important mechanisms in immune tolerance and/or unresponsiveness, paradoxes to this paradigm are often recognized. For example, in transplants, induction of apoptosis in the responding T cells by costimulator block plus rapamycin promotes tolerance (Li et al., 1999; Wells et al., 1999), but it is unclear why newly maturing T cells do not simply replace these deleted T cells and eliminate the graft after cessation of treatment (Ferguson and Green, 1999). Additionally, the death of T cells specific for tumor or tumorassociated antigen by Fas-dependent or -independent pathways seemed a plausible explanation for immune escape of neoplastic cells (Bodey et al., 1999; Gastman et al., 1999; Kiessling et al., 1999), yet proximal T cells often appear incapable of promoting antitumor immunity. In another example, the immune deficiency associated with HIV-1 infection is mainly attributed to CD4⫹ T cell death, but residual HIV-1 infected and/or uninfected CD4⫹ T cells may be unresponsive. Moreover, recent studies indicate that the localized death of lymphoid cells within the eye downregulates systemic immunity 1
Correspondence:
[email protected] (W.C.), smwahl@dir. nidcr.nih.gov (S.M.W.)
(Griffith et al., 1996), suggesting a possible soluble inhibitory molecule(s) in this tolerance induction. Mutation of individual genes in mice represents a valuable tool to study the function of the targeted genes. That similar phenotypes often manifest among mice deficient in quite different genes has been perplexing. In this regard, deletion of the genes for Fas death receptors or Fas ligand (FasL) (Nagata, 1997) results in dysregulation of T cell apoptosis, prolonged life span of potential autoreactive T cells, inflammation, autoimmune-like syndrome, and finally an early demise. Furthermore, mutation of IL-2 (Sadlack et al., 1993) also causes uncontrolled T cell proliferation (Sadlack et al., 1995), since IL-2-mediated signals are prerequisites for efficient apoptosis of activated effector T cells (Lenardo et al., 1999; Lenardo, 1991; Van Parijs et al., 1999). Both IL-2 and T cell receptor (TCR) (Mombaerts et al., 1993) null mice develop similar inflammatory diseases, including inflammatory bowel disease (IBD) and untimely death. However, the inflammatory response of these mutated mice may not be solely due to the prolonged life span of autoreactive T cells, since deletion of the IL-10 gene (Kuhn et al., 1993) resulted in a similar phenotype with IBD, without a decrease in T cell apoptosis. Even more paradoxically, mutation of the TGF-1 gene, which actually enhances T cell apoptosis (W. C. et al.,unpublished data), drives massive and lethal inflammation (Kulkarni et al., 1993; Shull et al., 1992). Thus, despite deletion of distinct immunoregulatory genes, there must be some shared mechanisms involved in the pathogenesis among these mice. The most salient feature of apoptosis is the lack of an attendant inflammatory response or tissue damage. This was originally thought to be a passive process and attributed to the rapid phagocytosis of apoptotic cells before lysis, thereby preventing the release of inflammatory or damaging factors. Recent evidence however has indicated that apoptotic cells may be actively involved in suppressing inflammatory responses by inducing the anti-inflammatory cytokines, IL-10 (Gao et al., 1998; Voll et al., 1997) and TGF- (Fadok et al., 1998). Previous studies (Fadok et al., 1998) and our own data (W.C. et al., unpublished data) identified macrophages as one of the cellular sources of TGF- triggered by ingestion of dying cells. Whether apoptotic T cells produce or secrete TGF- to contribute to this suppressive milieu has, however, not yet been determined. In this study, we present evidence that T cells express TGF- within their intracellular membrane-bound compartment that includes the mitochondria, release it during apoptosis, and activate TGF- through a ROS-dependent mechanism. TGF- released from apoptotic T cells likely contributes to resolution of inflammatory and immune responses and/or the generation of tolerance. Results Apoptotic T Cells Release TGF- To study whether apoptotic T cells secrete TGF-, we first used a well-defined activation-induced cell death
Immunity 716
Figure 1. Apoptotic Peripheral T Cells Release TGF- (A–C) Preactivated CD4⫹ T cells (2 ⫻ 105 per well in 96-well plate) were exposed to immobilized anti-CD3 antibody (10 g/ml) in serum free medium (X-Vivo-20) (solid bar) or medium only (hatched bar) for 24 and 48 hr. Cell death was determined by trypan blue exclusion assay (A). Cell-free supernatants were collected for the determination of total TGF- (B) and active TGF- (C) by a TGF-1 Emax ELISA kit. Data represent the mean ⫾ SD of duplicate wells of ELISA plates and are representative of at least three experiments. *, undetectable. (D–F) Preactivated CD4⫹ T cells were incubated with anti-Fas antibody (10 g/ml) for 30 min at 4⬚C. Following extensive washes, cells (2 ⫻ 105 per well) were added into culture plates with immobilized goat anti-hamster IgG antibody (10 g/ml) or restimulated with plate-coated anti-CD3 mAb (10 g/ml) for 48 hr. Cell viability was determined by trypan blue exclusion and/or 7-AAD staining. (D) Specific cell death induced by anti-Fas and anti-CD3 antibodies. The spontaneous apoptosis in medium only cultures was 25%–35%. Data are calculated as 100% ⫻ (1-(viable cells in antibody treated cultures/viable cells in medium control cultures)). Total TGF- (E) and active TGF- (F) in the culture supernatants. The experiment was repeated twice with similar results.
(AICD) system in peripheral T cells (Van Parijs et al., 1999). Murine spleen cells were stimulated with ConA for 72 hr, and CD4⫹ T cells were then isolated. After incubation with recombinant IL-2 overnight, the CD4⫹ T cells were restimulated or not with plate-coated antiCD3 mAb in serum-free medium (X-Vivo 20) for 24 and 48 hr. By 24 hr (Figure 1A), more than 90% of CD4⫹ T cells were killed by anti-CD3 compared to 20%–30% in untreated cultures. When supernatants from these cultures were tested for TGF-, significantly increased levels of total TGF- (acid treated) were detected in antiCD3 treated cultures, although the preactivated cells without anti-CD3 restimulation also released detectable TGF- (Figure 1B). Unexpectedly, in the anti-CD3induced apoptotic cultures, biologically active TGF- (without acid treatment) was also detected, although the level was lower than total TGF- (Figure 1C). The induction of active TGF- appeared not due to the CD3/ TCR stimulation alone, since parallel treatment of freshly isolated naive/resting CD4⫹ T cells failed to lead to an increase in active TGF- (Chen et al., 1998b) (data not shown). These results demonstrated that AICD of T cells initiated release of latent as well as active TGF-. To extend these studies, we induced T cell apoptosis by triggering death receptor Fas and evaluated TGF- secretion. Since Fas-induced cell death does not occur until late in the time course of intensive activation of normal T cells (Boise and Thompson, 1997; Nagata and
Golstein, 1995; Van Parijs et al., 1996), CD4⫹ T cells were first stimulated with anti-CD3 mAb in the presence of irradiated syngeneic spleen antigen- presenting cells (APCs) for 3 days. After incubation with rIL-2 overnight, CD4⫹ T cells were induced to undergo apoptosis with anti-Fas antibody crosslinked by plate-coated goat antihamster antibody (Van Parijs et al., 1998). By 48 hr, 20%–40% increase in apoptotic T cells was found in anti-Fas treated compared to untreated (medium only) cultures (Figure 1D). High levels of total TGF- were observed in anti-Fas treated culture supernatants, which exceeded the levels in parallel anti-CD3 treated cell cultures (Figure 1E), although the latter induced more than 90% of T cell death (Figure 1D). However, anti-Fas antibody failed to cause a measurable increase in secreted bio-active TGF- compared to T cells cultured in medium only (Figure 1F), which was distinct from apoptosis induced by anti-CD3 antibody in which both total and active TGF- were secreted (Figure 1). Apoptotic Thymocytes Induced by ␥-Irradiation and Dexamethasone Release TGF- In addition to peripheral T cells, apoptosis induced using two well-defined thymocyte apoptosis induction regimens, ␥-irradiation (1200 rad), or dexamethasone (1 M), also augmented TGF- levels. Either ␥-irradiation or dexamethasone resulted in the majority (80%–90% and 98%–99%, respectively) of thymocytes becoming apo-
Apoptotic Cells Release TGF- 717
cell cultures (data not shown). Thus, it appears that as apoptotic death progresses, T cells, in spite of different induction pathways, release TGF-, although the quantity and extent of TGF- activation may vary. Release versus Synthesis of TGF- We used two approaches to address whether TGF- in apoptotic T cell cultures was released from intracellular existing protein or programmed cell death promoted new TGF- synthesis. First, freshly isolated thymocytes were irradiated (1200 rad) and cultured for 2 hr, which induced the majority of cells to initiate apoptosis pathways. By RT-PCR, no increase in TGF-1 mRNA was observed in irradiated dying T cells compared to untreated thymocytes (Figure 3A). Secondly, TGF-1 mRNA of preactivated (ConA⫹IL-2) peripheral CD4⫹ T cells programmed to die by anti-CD3 or anti-Fas antibody restimulation was examined by RNase protection assay (RPA) (Chen et al., 1998a). No conclusive increase in TGF- mRNA expression was found in the anti-CD3 treated T cells (2 hr) compared to the untreated control cells (Figure 3B), despite the dramatic release of TGF- protein in the treated cultures (Figure 1). These data suggest that apoptotic T cells release intracellular existing TGF- rather than synthesize new protein.
Figure 2. Thymocyte Apoptosis and Secretion of TGF- Thymocytes were isolated and pooled from B6/129 mice (4–6 weeks, three to five mice per experiment) and resuspended in X-Vivo 20 medium (1 ⫻ 106 per well) for culture as described in the Experimental Procedures. (A). Apoptosis of thymocytes was quantified by staining with 7-AAD and FACScan analysis. Total (B) and active (C) TGF- in the supernatants was determined by ELISA. Data are presented as mean ⫾ SD of duplicate cultures. Med, medium; Rad, ␥-irradiation; Dex, dexamethasone; and *, undetectable.
ptotic by 16 hr in vitro (Figure 2A). Apoptosis was confirmed by 7-AAD staining (Figure 2A) (Chen et al., 1998c) as well as DNA fragmentation (data not shown) and was associated with increases in total (Figure 2B) and active (Figure 2C) TGF- in culture supernatants. In parallel, thymocytes treated with plate-coated anti-CD3 (10 g/ ml) also died (Figure 2A) and released enhanced total (Figure 2B) and active (Figure 2C) TGF-. Since 7-AAD and trypan blue staining primarily detect late apoptotic cells, thymocytes were next irradiated and cultured for only 2–4 hr, when the majority of the cells were in the early stage of apoptosis and were Annexin-V positive but 7-AAD negative. Analysis of the supernatants revealed no increase in TGF- in these early apoptotic
Intracellular Source of TGF- A previous report demonstrated TGF- within the mitochondria of rodent liver and heart cells (Heine et al., 1991), prompting us to separate the membrane-bound compartment including mitochondria from the cytosol of T cells for assay of TGF-. Several-fold higher levels of total TGF- were measured in the preparation containing mitochondria and/or other vesicles than in the cytosol of freshly isolated T cells by Western blot and ELISA (Figure 3C, G Fresh). Moreover, by immunofluorescence, TGF- staining could be colocalized with a mitochondrial-specific dye (Figures 3D–3F). Following treatment of thymocytes with ␥-irradiation (1200 rad) and culture at 37⬚C to induce apoptosis, we measured total TGF- in the mitochondria-enriched and cytosol preparations (Figure 3G) by ELISA. Without culture after irradiation (Figure 3G, 0 hr), most of the detectable TGF- remained associated with the mitochondria or membrane-bound vesicles that copurify with mitochondria. However, mitochondrial-associated TGF- of irradiated cells significantly decreased within 4 hr and continued to decrease for the 16 hr tested. By comparison, TGF- associated with cytosol showed a corresponding increase (Figure 3G). Despite the continuing decrease within the mitochondria-enriched preparation, TGF- in cytosol remained relatively stable from 4–14 hr resulting in a reduction (ⵑ400 pg/ml per 4 ⫻ 106 cells) of the amount of intracellular TGF- by the end of 16 hr, consistent with the levels of TGF- released into the supernatants (ⵑ100 pg/ml per 1 ⫻ 106 thymocytes) (Figure 2B). Association between Mitochondrial Membrane Potential (⌬m) and TGF- Collapse of mitochondrial membrane potential (⌬m) is a hallmark of cells committed to cell death (Kroemer, 1998). The loss of TGF- from the mitochondria-enriched compartment (Figure 3G) suggested an association with loss of ⌬m. First, compared to controls (Figure 4A), thymo-
Immunity 718
Figure 3. TGF- mRNA and Protein Levels in Apoptotic and Viable T Cells (A and B) TGF- mRNA levels. Freshly isolated thymocytes (pooled from three to five mice) were treated with plate-coated antiCD3 mAb (10 g/ml), ␥-irradiation (1200 rad), or medium only at 37⬚C, 5% CO2 for 2 hr. RNA was isolated and assessed for TGF-1 by RTPCR (A). Thymocytes of TGF-1 null mice (⫺/⫺) were included as a negative control for RT-PCR. The experiment was repeated twice with similar results. Rad, ␥-irradiation. (B) Assessment of TGF- mRNA by RPA. Preactivated peripheral CD4⫹ T cells were restimulated with anti-Fas (10 g/ml) cross-linked by plate-coated goat anti-hamster IgG (10 g/ ml) or with plate-coated anti-CD3 antibodies at 37⬚C, 5% CO2 for 2 hr. RNA was isolated and multiple template mCK3b set was used. (C) Western blot analysis of TGF-1. The mitochondria-enriched, membrane-bound compartment and cytosol were isolated from pooled fresh thymocytes (five mice), and 20 g of each protein pool was loaded into each lane. TGF-1 LAP was detected with antiLAP (TGF-1) antibody (R&D Systems). Mt, mitochondrial preparation; and Cyto, cytosol. (D–F) Immunofluorescence staining for TGF-1 and mitochondria. Spleen T cells were stained with mitochondrial-specific dye MitoTracker Red and chicken anti-TGF-1 antibody followed by Cy2-conjugated F(ab)2 donkey antichicken IgY as indicated. Cells were viewed under the confocal fluorescence microscope (⫻100). (D) shows TGF- green immunostaining. (E) shows mitochondrial red fluorescence. (F) represents an overlay of images (D) and (E), in which yellow indicates the colocalization of TGF- with the mitochondrial marker. The figure is a representative experiment. (G) Total TGF- levels within mitochondriaenriched preparations and the kinetics of intracellular redistribution of TGF- from the vesicles/mitochondria to the cytosol in apoptotic thymocytes. The thymocytes (4 ⫻ 106 cells) were irradiated (1200 rad) and cultured for indicated lengths of time. The untreated thymocytes are shown as Fresh.
cytes induced to die showed a corresponding loss of ⌬m as determined by staining with DiOC6, a dye taken up by mitochondria (Hildeman et al., 1999; Kroemer, 1998) (Figure 4A). Secondly, when thymocytes were irradiated, cultured, and the kinetics of their ⌬m measured (Figure 4B), a significant loss of ⌬m as determined by the mean fluorescence intensity (MFI) of DiOC6 staining was detected within 1 hr (Figure 4B), preceding the decrease in the mitochondria-associated TGF- (4–5 hr) (Figure 3G). Over 5 hr, the ⌬m of thymocytes continued to decline (Figure 4B), consistent with the kinetics of mitochondria-associated TGF- loss (Figure 3G) and implicating an association of collapse of ⌬m with TGF- release. Activation of Apoptotic T Cell Latent TGF- Is Associated with Reactive Oxygen Species TGF- is normally secreted in a latent form, L-TGF-, and the latency associated peptide (LAP) in the N-terminal region must be cleaved or dissociated to produce
active TGF- (Khalil, 1999). To discern how TGF- was activated during T cell apoptosis, we focused on reactive oxygen species (ROS) that have been implicated both in programmed cell death involving the loss of ⌬m (Hildeman et al., 1999) and in the activation of L-TGF- (Barcellos-Hoff and Dix, 1996). Intracellular ROS can be quantified by staining with dihydroethidium (Hildeman et al., 1999). Whether induced by anti-CD3, ␥-irradiation, or dexamethasone, apoptotic thymocytes showed increased percent of ROS positive cells (Figure 5A), which corresponded to decreased ⌬m (Figure 4A) and increased cell death (Figure 2A). Similar results were found in apoptotic peripheral CD4⫹ T cells induced by anti-CD3 mAb, but anti-Fas treatment failed to significantly increase dihydroethidium positive cells (data not shown). To test whether the ROS was involved in the loss of ⌬m and cell death and/or activation of latent TGF-, we included a superoxide dismutase mimetic, Mn(III) tetrakis (5,10,15, 20-benzoic acid) porphyrin (MnTBAP) (Hildeman et al., 1999) that inhibits ROS, in the anti-CD3-induced CD4⫹
Apoptotic Cells Release TGF- 719
Figure 4. Loss of ⌬m in Apoptotic Thymocytes (A) Cells were treated (16 hr) as indicated. ⌬m was determined by intracellular staining with DiOC6 and FACScan analysis. The shown percentages represent DiOC6low cells with reduced ⌬m. (B) The kinetics of ⌬m loss during apoptosis. Thymocyte apoptosis was induced with ␥-irradiation (1200 rad), and cells were cultured for 1–5 hr. The changes in mean fluorescence intensity (MFI) of DiOC6 cellular staining were measured by FACScan. The MFI of untreated thymocytes without DiOC6 staining was 38, which was used as a negative control.
T cell apoptosis culture. MnTBAP failed to protect and/ or rescue T cell apoptosis induced by anti-CD3 (Figure 5B), nor could it protect ⌬m in dying cells (data not shown), implying that ROS may not be directly responsible for T cell death mediated by CD3/TCR activation, but rather a consequence of the collapse of ⌬m (Kroemer, 1998). MnTBAP, however, almost completely abrogated the biologically active TGF- found in supernatants of apoptotic cells (Figure 5C). Thus, ROS from apoptotic T cells appear responsible, at least in AICD, for activation of latent TGF-. Apoptotic T Cell TGF- Suppresses Macrophage Inflammatory Cytokines Murine macrophages were stimulated with bacterial lipopolysaccharide (LPS) in the presence or absence of apoptotic thymocytes. LPS induced high levels of TNF␣ in macrophages (Figures 6A and 6B). In the presence of apoptotic cells, the TNF␣ production was reduced when determined either by intracellular cytokine staining (Figure 6A) or, more dramatically, by measurement of the cytokine in the culture supernatants (Figure 6B). Anti-
TGF- antibody added to the cocultures restored TNF␣ production (data not shown), confirming a role for TGF- in the suppression of LPS-induced macrophage TNF␣. Since macrophages also produce and/or activate TGF- upon phagocytosis of apoptotic cells (Fadok et al., 1998) (data not shown), it was necessary to determine which population, T cells or macrophages, contributed to the pool of active TGF- in cocultures. By comparing levels of active TGF- between TGF-1 null (Kulkarni et al., 1993; Shull et al., 1992) and wild-type macrophages exposed to TGF-1(⫹/⫹) apoptotic thymocytes, we confirmed that apoptotic T cells in fact released active TGF- (Figure 6C), since the supernatants of TGF- null macrophages plus wild-type apoptotic T cells contained detectable levels of active TGF-. Thus, apoptotic T cells, in addition to the macrophages clearing these cells, release TGF-, which contributes to the downregulation of macrophage inflammatory functions and likely also impairs viable T cell activation. Apoptotic Liver Cells Release Active TGF- In Vivo To establish whether the release of active TGF- was unique to suicidal T cells, we assayed apoptosis of liver
Immunity 720
Figure 5. ROS Is Associated with the Activation of Latent TGF- (A) Measurement of intracellular ROS by staining with dihydroethidium. Thymocytes were treated as indicated for 16 hr, stained with dihydroethidium, and analyzed by FACScan. The percentages represent dihydroethidiumhi (ROShi) cells. (B and C) MnTABP failed to prevent TCR-mediated T cell death but abrogated active TGF- in culture supernatants. Spleen cells were preactivated with ConA for 3 days, followed by IL-2 incubation overnight. Purified CD4⫹ T cells (2 ⫻ 105 per well) were restimulated with or without plate-coated anti-CD3 mAb (10 g/ml) in the presence or absence of MnTABP (100 M) for 48 hr. The viable cells were counted by trypan blue exclusion (B), and active TGF- in the supernatants was determined by ELISA (C). Med, medium, ␣CD3, anti-CD3 antibody; and *, undetectable.
cells and circulating levels of active TGF- in a welldefined mouse model (Ogasawara et al., 1993). In this model, massive apoptosis of liver cells is induced by injection of anti-Fas antibody (Jo2, 10 g per mouse) (Figures 7A and 7B) without dramatically affecting other organs in normal mice (Ogasawara et al., 1993). Significantly, sera of anti-Fas antibody injected mice showed much higher levels of active TGF- than that in untreated controls (Figure 7E, wild-type). This was further verified by the fact that injection of the same anti-Fas antibody into Fas null mice (B6.MRL-Faslpr) failed to increase active TGF- in sera (Figure 7E, Fas⫺/⫺), since no apoptosis of hepatocytes was induced (Figures 7C and 7D). Thus, apoptotic hepatocytes also release TGF- in vivo indicating that apoptotic cell release of TGF- is not a T cell restricted event, but rather a more general phenomenon. Discussion Evidence is emerging that complex active suppression, rather than simple passive lack of inflammation, is in-
volved during programmed cell death and the processes of apoptotic cell clearance (Fadok et al., 1998; Gao et al., 1998; Voll et al., 1997). We provide here direct evidence that apoptotic T cells release latent as well as active TGF-, which contributes to this anti-inflammatory milieu. Several unprecedented conclusions can be drawn from the current study. First, once programmed to become apoptotic, whether induced by TCR restimulation, death receptor signaling, or nonspecific ␥-irradiation, T cells intend to release TGF-. Variations in TGF- levels from apoptotic T cells may reflect not only subtypes and numbers of apoptotic cells but also the strength and features of specific death signals. One of the noted findings is that anti-Fas antibody (Jo2) triggers release of higher amounts of total TGF- than anti-CD3, although the former induces less T cell death than the latter. These data indicate that Fas-mediated signaling may have an alternative/additional pathway to engage TGF- secretion. Second, release rather than synthesis of TGF- during the apoptotic process is the most plausible explanation for the appearance of TGF- in the
Apoptotic Cells Release TGF- 721
Figure 6. TGF- Derived from Apoptotic T Cells Contributes to the Deactivation of Inflammatory Response (A) Intracellular staining of TNF␣ in LPS (10 ng/ml) activated macrophages. Macrophages were dual-stained with PE-conjugated anti-CD14 for surface marker and FITC-conjugated anti-TNF␣ for intracellular cytokine. Cells were analyzed by FACScan. (a) Negative control with the isotypic antibodies; (b) macrophages were treated with medium only (X-Vivo-20); (c) macrophages were treated with LPS; (d) macrophages preincubated with apoptotic thymocytes (1 ⫻ 106 per 2 ⫻ 105 macrophages) were stimulated with LPS. MFI, mean fluorescence intensity. (B) TNF␣ levels in the supernatants were determined by ELISA. Med, medium; and Apo: apoptotic thymocytes (1 ⫻ 106 per well). Data are presented as mean ⫾ SD of duplicate cultures. (C) Active TGF- in the supernatants of macrophage and apoptotic thymocyte cocultures. Macrophages (2 ⫻ 105 per well); Med, medium; and Apo, apoptotic thymocytes (1 ⫻ 106 per well). Data are presented as mean ⫾ SD of duplicate cultures.
supernatants, since there is no compelling evidence for increased transcription of TGF- in the apoptotic T cells. Third, TGF- was associated with a membrane-bound intracellular compartment and redistributed into the cytosol and medium following the collapse of ⌬m during apoptosis. Our data show that higher TGF- is found in mitochondria and/or in vesicles that copurify with mitochondria than in the cytosol of T cells, and is consistent with the previous finding that TGF- is localized within liver and heart cell mitochondria (Heine et al., 1991; Sporn, 1999). The mitochondrion has been referred to as the common effector of cell death (Kroemer, 1998; Kroemer et al., 1997) even though the “private” or selective death signaling may vary, and the collapse of ⌬m represents the no-return point in cell death. The kinetics of redistribution of intracellular TGF- and its close association with the reduction of ⌬m during T cell apoptosis favor TGF- release as a consequence of the collapse of ⌬m. It has been noted, at least in apoptotic thymocytes induced by irradiation, that only a portion of intracellular TGF- is released into the supernatants. How intracellular TGF- is released into the external milieu of apoptotic cells is unknown, albeit changes of
membrane permeability may influence this redistribution. Fourth, that apoptotic T cells release not only latent but also active TGF- emphasizes the contribution of apoptotic cells themselves in preventing and/or suppressing the potential inflammatory response, since only bio-active TGF- can function through the TGF- receptor signaling pathway (Derynck et al., 1998; Gorelik and Flavell, 2000; Letterio and Roberts, 1998; Massague, 1998; Yang et al., 1999). The fact that MnTBAP, an antagonist of ROS, abrogates nearly all active TGF- but fails to prevent the reduction of ⌬m and rescue T cells from death has important implications: ROS is involved in activating latent TGF- during T cell apoptosis; and accumulated ROS, at least in anti-CD3-induced cell death, is a consequence rather than a cause of the collapse of ⌬m. Finally, active TGF- released from apoptotic T cells clearly contributes to suppressed production of proinflammatory cytokines by macrophages and/or bystander immune reactive T cells, although the “no inflammation” state during apoptosis and clearance of dead cells may involve complex processes and multiple cell participation (Fadok et al., 1998; Gao et al., 1998; Griffith et al., 1996; Voll et al., 1997). In short, the data
Immunity 722
Figure 7. Apoptotic Liver Cells Release Active TGF- In Vivo (A–D) Massive apoptosis of liver cells in B6 wild-type (B) but not Fas null (D) mice injected with anti-Fas antibody (10 g per mouse). Livers were harvested 2–6 hr after antibody injection and fixed with 10% neutral buffered formalin. Tissues from mice injected with PBS (A and C) or anti-Fas (B and D) were stained with H&E and examined under light microscope (original magnification, ⫻63). Arrows in (A), (C), and (D) indicate typical hepatocyte nuclei. Following anti-Fas mAb there is massive death of hepatocytes in wild-type livers with dense pyknotic or broken nuclei (arrows in [B]) but not in Fas null livers (D). R, red blood cells. (E) Active TGF- in sera. Sera were collected from the mice 2–6 hr after anti-Fas and pooled from each group (three to five mice). Sera were diluted (1:20) with the sample buffer before assay (without acid activation) for TGF-1 by ELISA.
shown in this study have clearly broadened the potential mechanisms underlying immune tolerance and/or immune deficiency associated with T cell apoptosis. The balance between sufficient apoptosis of “dangerous” and/or damaged lymphocytes and adequate levels of anti-inflammatory cytokines during their clearance is a prerequisite to maintain normal homeostasis and
function of the immune system. On the one hand, as in Fas or FasL (Nagata, 1997; Nagata and Golstein, 1995), or IL-2 (Klebb et al., 1996; Sadlack et al., 1993, 1995) deleted mice, the most important pathway to induce apoptosis of T cells is abrogated. Subsequently, T cell activation and proliferation go unchecked and neither T cells, unable to commit to suicide, nor macrophages,
Apoptotic Cells Release TGF- 723
without apoptotic T cells to digest, can control the response through releasing anti-inflammatory cytokines such as TGF- and/or IL-10. On the other hand, as in TGF-1 null (Kulkarni et al., 1993; Shull et al., 1992) and/ or perhaps in IL-10 mutated mice (Kuhn et al., 1993), the ability to produce these crucial anti-inflammatory cytokines by T cells and macrophages is lacking, although these mice have normal (Kuhn et al., 1993) or even increased T cell apoptosis (W.C. et al., unpublished data). The net effect in each of these null mice may be a deficiency in negative regulatory molecules, and the mice unavoidably develop uncontrolled inflammation and “autoimmune”-like diseases. The present study may provide insight for exploring the mechanisms of immune tolerance, immune escape, and/or deficiency in clinical situations. Both apoptotic T cells and phagocytic macrophages may contribute to a state of immune unresponsiveness via secretion of TGF- and/or IL-10. This complex immunoregulatory process may favor acceptance of transplanted organs (Li et al., 1999; Wells et al., 1999) or tolerance induction in response to high-dose antigen administration in experimental autoimmune disease models (Chen and Wahl, 1999; Chen et al., 1995, 1998a; Strober et al., 1997), together with the induction of Th3-like cells (Chen et al., 1995). The escape of tumor cells from immune surveillance may involve not only Fas-mediated and/or other death pathways for killing tumor-specific T cells (Bodey et al., 1999; Gastman et al., 1999; Kiessling et al., 1999), but also the concomitant release of immunosuppressive cytokines. In HIV-1 infection, particularly in the later stage of AIDS, death of CD4⫹ T cells as well as the consequent release of immunosuppressive cytokines such as TGF- (Hu et al., 1996; Kekow et al., 1990; Sher et al., 1992; Wahl et al., 1991; Zauli et al., 1996) may underly the exhaustive immune deficiency that occurs even when some uninfected CD4⫹ T cells still remain. By refocusing our investigation on the apoptotic cell in addition to the previously determined phagocytic cell as the source of the immunosuppressive activity, it may become possible to manipulate host offense and defense. Experimental Procedures Mice TGF-1(⫺/⫺) (C57BL/6xSv129) mice were generated by disruption of the TGF-1 gene in murine embryonic stem cells by homologous recombination (Kulkarni et al., 1993), and their wild-type littermates were bred and maintained in a specific pathogen-free rodent facility at the National Institute of Dental and Craniofacial Research, NIH. Mice homozygous for Fas mutation (B6.MRL-Faslpr) and their wildtype controls were purchased from The Jackson Laboratory (Bar Harbor, ME). Antibodies and Reagents Hamster anti-murine CD3 (Clone 145-2C11, NA/LE), Fas (Clone Jo2, NA/LE), purified IgG isotypic control antibody (Clone G235-2356, NA/LE), and Annexin V-PE Apoptosis Kit were purchased from PharMingen (San Diego, CA). Rat anti-murine FITC-CD4 (clone CT-CD4) and the respective isotypic control mAbs were purchased from CalTag Laboratory (San Francisco, CA). Mouse anti-TGF-1, 2, 3 mAb was from R&D Systems (Minneapolis, MN). Goat anti-hamster IgG (H⫹L) antibody was from Pierce (Rockford, IL). Recombinant IL-2 was purchased from Boehringer Mannheim (Indianapolis, IN). 7-aminoactinomycin D(7-AAD) and Mn(III) tetrakis (5,10,15,20-benzoic acid) porphyrin (MnTBAP) were purchased from Calbiochem (La Jolla,
CA). Dye 3,3⬘-dihexyloxacarbocyanine iodide (DiOC6) and dihydroethidium were from Molecular Probes (Eugene, OR). Preparation of CD4ⴙ T Cells Spleens and lymph nodes were harvested and the tissues were gently minced in complete DMEM containing 10% FBS (BioWhittaker, Walkersville, MD) as described (Chen et al., 1998b). Cells were then passed through a cell strainer (Becton Dickinson, Franklin Lakes, NJ) and red blood cells lysed with ACK lysing buffer (BioWhittaker). For CD4⫹ T cell isolation, cells were purified by using a mouse CD4⫹ T Cell Column System (R&D Systems). By FACS analysis, the purity of CD4⫹ T cells was usually ⬎95% with no detectable B cells or monocytes. In some experiments, ␥-irradiated T cell depletedspleen cells (3000 rad) were used as antigen-presenting cells (APCs). Induction of Peripheral T Cell Apoptosis In Vitro For AICD through the CD3/TCR complex, spleen cells (1 ⫻ 106/ml) were cultured in complete DMEM with ConA (2 g/ml) (Sigma) for 3 days. CD4⫹ T cells were then column purified (R&D Systems) and re-incubated with recombinant rIL-2 (50 u/ml) overnight. T cells (2 ⫻ 105/well) were then restimulated with anti-CD3 mAb (10 g/ml) precoated onto 96-well plates (Costar, Cambridge, MA) in serum free X-Vivo 20 culture medium (BioWhittaker) for 24 and 48 hr. For cell death induced by Fas cross-linking, purified CD4⫹ T cells were stimulated with anti-CD3 mAb (0.5 g/ml) in the presence of syngeneic APCs for 3 days. Viable cells were incubated with hamster antiFas mAb (Jo2) at 4⬚C for 30 min. After being washed, cells (2 ⫻ 105/ well) were then added into the plates precoated with goat antihamster antibody (10 g/ml). Cell-free supernatants were collected at 48 hr for detection of cytokines. Cell viability was assayed with trypan blue exclusion and/or 7-AAD staining (Chen et al., 1998b). Induction of Thymocyte Apoptosis In Vitro Thymocytes were cultured with plate-coated anti-CD3 mAb (10 g/ ml) or with dexamethasone (1 M). For apoptosis induced by ␥-irradiation, thymocytes were irradiated (1200 rad) before being added into the plates. Cells were then cultured in serum-free X-Vivo 20 medium for 16 hr. As parallel controls, thymocytes were either cultured with medium at 37⬚C or at 4⬚C. Cell viability was determined with 7-AAD staining or trypan blue exclusion. In some experiments, early apoptotic cells were detected by Annexin-V-PE, together with 7-AAD (Annexin-V⫹ 7-AAD⫺). Culture supernatants were collected at 16–24 hr for cytokine measurement. Cytokine ELISAs Active TGF-1 in the culture supernatants was determined directly by the TGF-1 Emax ImmunoAssay System (Promega Corp., Madison, WI). For total TGF-1 measurement, samples were first treated with 1N HCl (1 l per 50 l supernatant) for 15 min at room temperature and neutralized with an equal amount of 1N NaOH before analysis in the ELISA. Serum was usually diluted 1:20 with the kit sample buffer before TGF-1 assay. Quantitative analysis for TNF␣ was performed using the ELISA kit purchased from Biosource International (Gaithersburg, MD). Assessment of TGF- mRNA Total RNA was isolated from cells using a modified guanidinium isothiocyanate method (RNeasy Kit, Qiagen, Chatsworth, CA). Semiquantitative reverse transcriptase-PCR (RT-PCR) for TGF-1 mRNA was performed as described (McCartney-Francis et al., 1997). HPRT, a housekeeping gene, was used to verify cDNA synthesis from each RNA sample and to enable comparison of RNA levels between different samples. For RNase protection assays (RPA) (Chen et al., 1998a), multiple template mCK3b set (PharMingen) was used, and radiolabeled probes were generated using (33P)dUTP (New England Nuclear, Boston, MA). Mitochondria and Cytosol Preparations Mitochondria were isolated according to the published protocol (Yang et al., 1997). In brief, the T cell pellets were washed with icecold PBS and resuspended with five volumes of buffer A (20 mM Hepes-KOH [pH 7.5], 10 mM KCl, 1.5 mM MgCl2, 1 mM sodium EDTA, 1 mM sodium EGTA, 1 mM dithiothreitol, and 0.1 mM phenyl-
Immunity 724
methylsulfonyl fluoride) containing 250 mM sucrose (thymocytes: 100 ⫻ 106 cells in 500 l; peripheral T cells: 15–20 ⫻ 106 cells in 200–250 l). The cells were homogenized with 10 stroke of a Teflon homogenizer, and the homogenates were centrifuged twice at 750g for 10 min at 4⬚C. The pellets were discarded and the supernatants were centrifuged at 10,000g for 15 min at 4⬚C. The resulting mitochondrial-enriched pellets were washed and resuspended in the original volumes of buffer A and frozen at -80⬚C. The supernatants of the 10,000g spin were further centrifuged at 100,000g for 1 hr at 4⬚C, and the resulting supernatants (cytosol) were frozen at ⫺80⬚C for further experiments. The samples (20 l) were diluted (1:5) with the sample dilution buffer included in the TGF-1 ELISA kit before analysis for TGF-1.
Western Blot Aliquots of mitochondrial preparation and cytosol (20 g protein) were subjected to 12.5% SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to a nitrocellulose filter (Schleicher & Schull, Keene, NH). The nitrocellulose membrane was blocked for 1 hr by using 5% instant skim milk powder in TBS (20 mM Tris-HCl [pH 7.6], 137 mM NaCl, and 0.1% Trition X-100). The filter was probed by an antibody to LAP-TGF-1 (anti-rhLAP, 0.2 g/ml, R&D Systems). After washing, the nitrocellulose membrane was incubated with horseradish peroxidase-conjugated Protein A as recommended by the manufacturer. Finally, the washed blots were developed using an enhanced chemiluminescence detection system (Amersham) and recorded on an autoradiograph (Kodak X-Omat film).
Immunofluorescence Staining Freshly isolated spleen T cells were incubated with mitochondrionselective dye MitoTracker Red CMXRos (50 nM) (Molecular Probes, Eugene, OR) in complete DMEM at 37⬚C for 40 min. After washing with warm phosphate-buffered saline (PBS), cells were fixed with Cytofix/Cytoperm (Cytofix/Cytoperm Plus Kit, PharMingen) at 4⬚C for 20 min. Cells (1 ⫻ 106) were washed with Perm/Wash Buffer (PharMingen) and incubated with chicken anti-human TGF-1 (10 g, R&D Systems) at 4⬚C for 30 min. Cells were washed twice with Perm/Wash Buffer and stained with Cy2-conjugated F(ab)2 donkey anti-chicken IgY antibody (Jackson ImmunoResearch Laboratories, West Grove, PA) at 4⬚C for 30 min. For negative staining controls, MitoTracker Red stained cells were incubated with secondary Cy2conjugated antibody only without adding anti-TGF-1 antibody. After two washes, cells were resuspended in staining buffer (PBSBSA[0.5%]-Az[0.1%]) and subjected to confocal fluorescence microscopy.
Assessment of Effects of Apoptotic T Cells on Macrophage TNF␣ Production Peritoneal macrophages were collected by lavage with ice-cold PBS without Ca2⫹ and Mg2⫹ and washed twice with PBS before resuspension in complete DMEM. Macrophages were adhered (2 ⫻ 105 per well) in a 96-well flat-bottom plate in prewarmed DMEM for 2 hr at 37⬚C, 5% CO2, and nonadherent cells washed off with X-Vivo-20. To some cultures, apoptotic thymocytes (1 ⫻ 106) that had been irradiated and cultured for 3 hr at 37⬚C were added overnight. LPS (10 ng/ml) was added as indicated. Supernatants were collected at 24 hr for cytokine ELISAs. For detection of intracellular TNF␣, macrophages preincubated overnight with or without apoptotic thymocytes were stimulated with LPS (10 ng/ml) for 4–6 hr at 37⬚C, 5% CO2 in the presence of GolgiPlug (1 l/ml, PharMingen) to block secretion. The macrophages were first stained with PE-anti-murine CD14 mAb (PharMingen) on ice for 30 min. After washing, the cells were stained for intracellular TNF␣ with FITC-anti-murine TNF␣ mAb using the Cytofix/Cytoperm Plus kit (PharMingen) following the manufacturer’s protocol. PE-or FITC- labeled isotypic antibody controls for surface and intracellular staining were also used before analysis by FACScan with the software Lysis II. Acknowledgments We thank G. McGrady for maintaining the mice and Drs. N. McCartney-Francis and Ke-jian Lei for critical review of the manuscript. We also thank Dr. W. Swaim for confocal microscopy. Received September 18, 2000; revised April 4, 2001. References Barcellos-Hoff, M.H., and Dix, T.A. (1996). Redox-mediated activation of latent transforming growth factor-beta 1. Mol. Endocrinol. 10, 1077–1083. Bodey, B., Bodey, B., Jr., Siegel, S.E., and Kaiser, H.E. (1999). Fas (Apo-1, CD95) receptor expression in childhood astrocytomas. Is it a marker of the major apoptotic pathway or a signaling receptor for immune escape of neoplastic cells? In Vivo 13, 357–373. Boise, L.H., and Thompson, C.B. (1997). Bcl-x(L) can inhibit apoptosis in cells that have undergone Fas-induced protease activation. Proc. Natl. Acad. Sci. USA 94, 3759–3764. Chen, W.J., and Wahl, S.M. (1999). Manipulation of TGF-beta to control autoimmune and chronic inflammatory diseases. Microbes Infect. 1, 1367–1380.
Detection of ⌬m and ROS in T Cells To detect changes in ⌬m, cells were resuspended in culture medium containing 25 nM DiOC6 for 40 min at 37⬚C as described (Hildeman et al., 1999) for FACS analysis. For determination of intracellular ROS, cells were incubated with dihydroethidium (2.5 M) in culture medium at 37⬚C for 40 min. FITC- or PE- conjugated anti-CD4 costaining with these dyes was performed at 4⬚C, prior to incubation with the dyes. The histograms of DiOC6 were displayed on FL-1 or of dihydroethidium on FL-2.
Chen, Y., Inobe, J., Marks, R., Gonnella, P., Kuchroo, V.K., and Weiner, H.L. (1995). Peripheral deletion of antigen-reactive T cells in oral tolerance. Nature 376, 177–180.
Neutralization of ROS with MnTBAP in Cell Culture Inhibition of ROS in cultures was carried out by incubating the cells in the presence of MnTBAP (100 M) as described (Hildeman et al., 1999). Cell viability was assayed as described above and compared with the parallel medium only (control) treated cells. Supernatants were collected for cytokine ELISAs.
Chen, W.J., Sayegh, M.H., and Khoury, S.J. (1998c). Mechanisms of acquired thymic tolerance in vivo: intrathymic injection of antigen induces apoptosis of thymocytes and peripheral T cell anergy. J. Immunol. 160, 1504–1508.
Injection of Anti-Fas Antibody (Jo2) In Vivo Normal and Fas(⫺/⫺) mice were injected intraperitoneally (i.p.) with Jo2 (10 g per mouse). All wild-type mice, but not Fas deleted mice, died by 6 hr (Ogasawara et al., 1993) (data not shown). Sera were collected for the determination of TGF- at 2–6 hr after antibody injection. The livers were fixed with buffered formalin (10%), paraffinembedded, and processed for histology.
Chen, W.J., Jin, W., Cook, M., Weiner, H.L., and Wahl, S.M. (1998a). Oral delivery of group A streptococcal cell walls augments circulating TGF- and suppresses SCW arthritis. J. Immunol. 161, 6297– 6304. Chen, W.J., Jin, W.W., and Wahl, S.M. (1998b). Engagement of CTLA-4 induces transforming growth factor- production by murine CD4⫹ T cells. J. Exp. Med. 188, 1849–1857.
Derynck, R., Zhang, Y., and Feng, X.H. (1998). Smads: transcriptional activators of TGF-beta responses. Cell 95, 737–740. Fadok, V.A., Bratton, D.L., Konowal, A., Freed, P.W., Westcott, J.Y., and Henson, P.M. (1998). Macrophages that have ingested apoptotic cells in vitro inhibit proinflammatory cytokine production through autocrine/paracrine mechanisms involving TGF-beta, PGE2, and PAF. J. Clin. Invest. 101, 890–898. Ferguson, T.A., and Green, D.R. (1999). T cells are just dying to accept grafts. Nat. Med. 5, 1231–1232. Gao, Y., Herndon, J.M., Zhang, H., Griffith, T.S., and Ferguson, T.A.
Apoptotic Cells Release TGF- 725
(1998). Antiinflammatory effects of CD95 ligand (FasL)-induced apoptosis. J. Exp. Med. 188, 887–896.
matory bowel disease in T cell receptor mutant mice. Cell 75, 274–282.
Gastman, B.R., Atarshi, Y., Reichert, T.E., Saito, T., Balkir, L., Rabinowich, H., and Whiteside, T.L. (1999). Fas ligand is expressed on human squamous cell carcinomas of the head and neck, and it promotes apoptosis of T lymphocytes. Cancer Res. 59, 5356–5364.
Nagata, S. (1997). Apoptosis by death factor. Cell 88, 355–365.
Gorelik, L., and Flavell, R.A. (2000). Abrogation of TGF-beta signaling in T cells leads to spontaneous T cell differentiation and autoimmune disease. Immunity 12, 171–181. Griffith, T.S., Yu, X., Herndon, J.M., Green, D.R., and Ferguson, T.A. (1996). CD95-induced apoptosis of lymphocytes in an immune privileged site induces immunological tolerance. Immunity 5, 7–16. Heine, U.I., Burmester, J.K., Flanders, K.C., Danielpour, D., Munoz, E.F., Roberts, A.B., and Sporn, M.B. (1991). Location of transforming growth factor-1 in mitochondria of murine heart and liver. Cell Regulation 2, 467–477. Hildeman, D.A., Mitchell, T., Teague, T.K., Hanson, P., Day, B.J., Kappler, J., and Marrack, P.C. (1999). Reactive oxygen species regulate activation-induced T cell apoptosis. Immunity 10, 735–744. Hu, R., Oyaizu, N., Than, S., Kalyanaraman, V.S., Wang, X.P., and Pahwa, S. (1996). HIV-1 gp160 induces transforming growth factorbeta production in human PBMC. Clin. Immunol. Immunopathol. 80, 283–289. Kekow, J., Wachsman, W., McCutchan, J.A., Cronin, M., Carson, D.A., and Lotz, M. (1990). Transforming growth factor beta and noncytopathic mechanisms of immunodeficiency in human immunodeficiency virus infection. Proc. Natl. Acad. Sci. USA 87, 8321–8325. Khalil, N. (1999). TGF-beta: from latent to active. Microbes Infect. 1, 1255–1263. Kiessling, R., Wasserman, K., Horiguchi, S., Kono, K., Sjoberg, J., Pisa, P., and Petersson, M. (1999). Tumor-induced immune dysfunction. Cancer Immunol. Immunother. 48, 353–362. Klebb, G., Autenrieth, I.B., Haber, H., Gillert, E., Sadlack, B., Smith, K.A., and Horak, I. (1996). Interleukin-2 is indispensable for development of immunological self-tolerance. Clin. Immunol. Immunopathol. 81, 282–286. Kroemer, G. (1998). The mitochondrion as an integrator/coordinator of cell death pathways. Cell Death Differ. 5, 547. Kroemer, G., Zamzami, N., and Susin, S.A. (1997). Mitochondrial control of apoptosis. Immunol. Today 18, 44–51. Kuhn, R., Lohler, J., Rennick, D., Rajewsky, K., and Muller, W. (1993). Interleukin-10-deficient mice develop chronic enterocolitis. Cell 75, 263–274. Kulkarni, A.B., Huh, C.-H., Becker, D., Gerser, A., Lyght, M., Flanders, K.C., Roberts, A.B., Sporn, M.B., Ward, J.M., and Karlsson, S. (1993). Transforming growth factor- null mutation in mice causes excessive inflammatory response and early death. Proc. Natl. Acad. Sci. USA 90, 770–774. Lenardo, M., Chan, K.M., Hornung, F., McFarland, H., Siegel, R., Wang, J., and Zheng, L. (1999). Mature T lymphocyte apoptosis— immune regulation in a dynamic and unpredictable antigenic environment. Annu. Rev. Immunol. 17, 221–253. Lenardo, M.J. (1991). Interleukin-2 programs mouse alpha beta T lymphocytes for apoptosis. Nature 353, 858–861. Letterio, J.J., and Roberts, A.B. (1998). Regulation of immune responses by TGF-. Annu. Rev. Immunol. 16, 137–161. Li, Y., Li, X.C., Zheng, X.X., Wells, A.D., Turka, L.A., and Strom, T.B. (1999). Blocking both signal 1 and signal 2 of T-cell activation prevents apoptosis of alloreactive T cells and induction of peripheral allograft tolerance. Nat. Med. 5, 1298–1302. Massague, J. (1998). TGF- signal transduction. Annu. Rev. Biochem. 67, 753–791. McCartney-Francis, N.L., Mizel, D.E., Frazier-Jessen, M., Kulkarni, A.B., McCarthy, J.B., and Wahl, S.M. (1997). Lacrimal gland inflammation is responsible for ocular pathology in TGF-beta 1 null mice. Am. J. Pathol. 151, 1281–1288. Mombaerts, P., Mizoguchi, E., Grusby, M.J., Glimcher, L.H., Bhan, A.K., and Tonegawa, S. (1993). Spontaneous development of inflam-
Nagata, S., and Golstein, P. (1995). The Fas death factor. Science 267, 1449–1456. Ogasawara, J., Watanabe-Fukunaga, R., Adachi, M., Matsuzawa, A., Kasugai, T., Kitamura, Y., Itoh, N., Suda, T., and Nagata, S. (1993). Lethal effect of the anti-Fas antibody in mice. Nature 364, 806–809. Sadlack, B., Lohler, J., Schorle, H., Klebb, G., Haber, H., Sickel, E., Noelle, R.J., and Horak, I. (1995). Generalized autoimmune disease in interleukin-2-deficient mice is triggered by an uncontrolled activation and proliferation of CD4⫹ T cells. Eur. J. Immunol. 25, 3053– 3059. Sadlack, B., Merz, H., Schorle, H., Schimpl, A., Feller, A.C., and Horak, I. (1993). Ulcerative colitis-like disease in mice with a disrupted interleukin-2 gene. Cell 75, 253–261. Sher, A., Gazzinelli, R.T., Oswald, I.P., Clerici, M., Kullberg, M., Pearce, E.J., Berzofsky, J.A., Mosmann, T.R., James, S.L., and Morse, H.C., III (1992). Role of T-cell derived cytokines in the downregulation of immune responses in parasitic and retroviral infection. Immunol. Rev. 127, 183–204. Shull, M.M., Ormsby, I., Kier, A.B., Pawlowski, S., Diebold, R.J.M., Yin, R., Allen, Sidman, C., Proetzel, B., calvin, D., Annuniziata, N., and Doeschman, T. (1992). Targeted disruption of the mouse transforming growth factor-1 gene results in multifocal inflammatory disease. Nature 359, 693–699. Sporn, M.B. (1999). TGF-beta: 20 years and counting. Microbes Infect. 1, 1251–1253. Strober, W., Kesall, B., Fuss, I., Marth, T., Ludviksson, B., Ehrhardt, R., and Neurath, M. (1997). Reciprocal IFN-␥ and TGF- response regulate the occurrence of mucosal inflammation. Immunol. Today 18, 61–64. Van Parijs, L., Ibraghimov, A., and Abbas, A.K. (1996). The roles of costimulation and Fas in T cell apoptosis and peripheral tolerance. Immunity 4, 321–328. Van Parijs, L., Biuckians, A., and Abbas, A.K. (1998). Functional roles of Fas and Bcl-2-regulated apoptosis of T lymphocytes. J. Immunol. 160, 2065–2071. Van Parijs, L., Refaeli, Y., Lord, J.D., Nelson, B.H., Abbas, A.K., and Baltimore, D. (1999). Uncoupling IL-2 signals that regulate T cell proliferation, survival, and Fas-mediated activation-induced cell death. Immunity 11, 281–288. Voll, R.E., Herrmann, E.A., Roth, C., Stach, C., Kalden, J.R., and Girkontaite, I. (1997). Immunosuppressive effects of apoptotic cells. Nature 390, 350–351. Wahl, S.M., Allen, J.B., McCartney-Francis, N., Morganti-Kossmann, M.C., Kossmann, T., Ellingsworth, L., Mai, U.E., Mergenhagen, S.E., and Orenstein, J.M. (1991). Macrophage- and astrocyte-derived transforming growth factor beta as a mediator of central nervous system dysfunction in acquired immune deficiency syndrome. J. Exp. Med. 173, 981–991. Wells, A.D., Li, X.C., Li, Y., Walsh, M.C., Zheng, X.X., Wu, Z., Nunez, G., Tang, A., Sayegh, M., Hancock, W.W., et al. (1999). Requirement for T-cell apoptosis in the induction of peripheral transplantation tolerance. Nat. Med. 5, 1303–1307. Yang, J., Liu, X., Bhalla, K., Kim, C.N., Ibrado, A.M., Cai, J., Peng, T.I., Jones, D.P., and Wang, X. (1997). Prevention of apoptosis by Bcl-2: release of cytochrome c from mitochondria blocked. Science 275, 1129–1132. Yang, X., Letterio, J.J., Lechleider, R.J., Chen, L., Hayman, R., Gu, H., Roberts, A.B., and Deng, C. (1999). Targeted disruption of SMAD3 results in impaired mucosal immunity and diminished T cell responsiveness to TGF-beta. EMBO J. 18, 1280–1291. Zauli, G., Vitale, M., Gibellini, D., and Capitani, S. (1996). Inhibition of purified CD34⫹ hematopoietic progenitor cells by human immunodeficiency virus 1 or gp120 mediated by endogenous transforming growth factor beta 1. J. Exp. Med. 183, 99–108.