OF
THE
JOURNAL IMMUNOLOGY
CUTTING EDGE Cutting Edge: Two Distinct Motifs within the Fas Ligand Tail Regulate Fas Ligand-Mediated Costimulation1 Mingyi Sun,* Shinhee Lee,* Saoussen Karray,2† Matthieu Levi-Strauss,3† Kristina T. Ames,* and Pamela J. Fink4* The cytoplasmic domain of Fas ligand is sufficient to costimulate CD8ⴙ T cells by driving Fas ligand recruitment into lipid rafts and association with select Src homology 3-containing proteins, activating PI3K and MAPK pathways, mediating nuclear translocation of the transcription factors NFAT and AP-1, and enhancing IFN-␥ production and Ag-specific CD8ⴙ T cell proliferation. We now show that Fas ligand molecules lacking amino acids 45–54 in the proline-rich region of the cytoplasmic domain fail to costimulate but serve as effective death inducers. Death induction and costimulation by Fas ligand are therefore clearly separable functions. Further, upon Fas ligand-mediated costimulation, casein kinase I phosphorylates Fas ligand, in which two conserved casein kinase I binding sites regulate NFAT activation and costimulation. These results help resolve how one molecule can serve as a double-edged immunomodulator by directing discrete biological consequences. The Journal of Immunology, 2007, 179: 5639 –5643.
also transmit a positive reverse signal to costimulate CD8⫹ T cells (6 –10). FasL-mediated costimulation is required for optimal thymocyte maturation (10) and for Ag-driven proliferation of mature T cells in vivo and in vitro (7–9), with a stronger influence on CD8⫹ than on CD4⫹ T cells (8). The cytoplasmic tail of FasL is sufficient to costimulate by the recruitment of FasL into lipid rafts and association with select Src homology 3 (SH3)-containing proteins such as Fyn, PI3K, and Grb2 but not Lck, by enhancement of the phosphorylation of Akt, Erk1/2, JNK, and FasL itself at serine residues, and by activation of the transcription factors NFAT and AP-1 (6, 11, 12). The FasL tail contains two distinct motifs: a striking prolinerich region consisting of several SH3 binding domains and two conserved casein kinase I (CKI) binding sites. Using mutagenesis, we now identify residues within both motifs that are crucial for FasL costimulation but are not required for protein expression or Fas-induced death. These results offer the first evidence that costimulation and death induction are separable functions and demonstrate that two distinct motifs within the FasL cytoplasmic domain regulate costimulation through multiple signaling pathways.
ptimal T cell activation generally requires costimulation in addition to a signal delivered through the TCR. Three main classes of costimulatory molecules have been identified to date: Ig superfamily members (1), TNF receptor (TNFR)5 family members (2), and TNF family members that are newly emerging mediators of costimulation through reverse signaling (3). Although the molecular mechanisms of costimulation delivered by Ig and TNFR family members have been studied extensively (4, 5), the underlying mechanism of costimulation transduced by TNF family members has only recently come under scrutiny (6). Fas ligand (FasL, CD178) is a type II transmembrane protein belonging to the TNF family. Although well characterized for its capacity to deliver a death signal through Fas (CD95), previous studies demonstrate that this TNF family member can
O
Materials and Methods
*Department of Immunology, University of Washington, Seattle, WA 98195; and †Institut National de la Sante´ et de la Recherche Me´dicale (INSERM) Unite´ 580, Hoˆpital Necker, Paris, France
3
Received for publication June 15, 2007. Accepted for publication August 24, 2007. 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 funded by National Institutes of Health Grant AI 44130 (to P.J.F.) and by a predoctoral training grant from the Cancer Research Institute (to M.S.). 2
Current address: INSERM Unite´ 753, Institut Gustave Roussy, Villejuif, France
www.jimmunol.org
Mice Wild-type (WT) C57BL/6 (B6), B6.MRL-Faslpr (lpr), and BALB/cJ mice were purchased from The Jackson Laboratory or bred on site. B6.FasLnull.lpr mice were generated by cross/backcross breeding of B6.lpr to B6.FasLnull mice (13). All mice were used at 6 –9 wk of age and in accordance with the Institutional Animal Care Use Committee guidelines of the University of Washington (Seattle, WA).
Reagents Hamster anti-CD3 (clone 145-2C11), mouse anti-Ly49a (clone A1), and 7-aminoactinomycin D (7-AAD) were purchased from BD Pharmingen. Rabbit anti-Erk1/2, anti-phospho-Erk1/2, and anti-phospho-Akt were obtained from Cell Signaling Technology. Mouse anti-GFP (clone JL-8) was purchased from Clontech Laboratories. Santa Cruz Biotechnology supplied rabbit antiFasL (clone Q-20). Affinity-purified goat anti-mouse IgG and anti-hamster IgG were obtained from Rockland. Molecular Probes supplied streptavidinAF488 and goat anti-rabbit AF546. Current address: INSERM SC11-Orphanet, Paris, France
4
Address correspondence and reprint requests to Dr. Pamela J. Fink, University of Washington, Department of Immunology, I-607H Health Sciences Center, Campus Box 357650, Seattle, WA 98195. E-mail address:
[email protected] 5 Abbreviations used in this paper: TNFR, TNF receptor; 7-AAD, 7-aminoactinomycin D; B6, C57BL/6; CID, chemical inducers of dimerization; CKI, casein kinase I; EC, extracellular; FasL, Fas ligand; FKBP3, three FK506 binding domains; HA, hemagglutinin; lpr, MRL-Faslpr; M, myristoylation-target domain; SH3, Src homology 3; TM, transmembrane domain; WT, wild type. Copyright © 2007 by The American Association of Immunologists, Inc. 0022-1767/07/$2.00
5640
CUTTING EDGE: TWO MOTIFS WITHIN THE FasL TAIL REGULATE COSTIMULATION
Cell lines The generation of H-2d-reactive FasLnull.lpr CTLs and the maintenance of H-2Ld-reactive L3, Jurkat, and Phoenix E packaging cells were described previously (6). To generate stable infectants, LZRS-GFP-IRES-based retroviruses (provided by Dr. Philip Greenberg, University of Washington) expressing target genes were isolated from Phoenix E stable transfectants and concentrated to infect L3 or FasLnull.lpr CTLs 4 days after restimulation.
Recombinant DNA constructs The murine and FasL chemical inducers of dimerization (CID) constructs were generated as described (6). In brief, the myristoylation-target domain (M) is linked to the murine and FasL cytoplasmic domains, three FK506 binding protein domains (FKBP3), and the hemagglutinin (HA) epitope. For the FasL/ Ly49a chimeric construct, the extracellular (EC) and transmembrane (TM) domains of murine Ly49a (also a type II protein) were fused to the cytoplasmic domain of murine FasL. This chimeric molecule lacks the FasL trimerization domain and cannot associate with endogenous FasL molecules. Deletional mutations in the proline-rich region and alanine substitutions in the CKI binding sites of the FasL intracellular domain were performed by PCR on the FasL CID, full-length FasL, or FasL/Ly49a chimeric constructs by using the QuikChange multisite-directed mutagenesis kit (Stratagene) as directed. All DNA constructs were verified by sequence.
CID, immunoblotting, and proliferation assays CID, immunoblotting and [3H]TdR proliferation assays were performed as described (6).
Flow cytometry and microscopy Surface expression of FasL/Ly49a fusion proteins was determined by flow cytometric analysis of L3 infectants stained with biotinylated anti-Ly49a and streptavidin-PE. For monitoring the kinetics of FasL/Ly49a expression, lot-dependent differences in anti-Ly49a brightness were accommodated by normalizing the geometric mean fluorescence intensity of the day 0 time point. To assay killing by FasLnull.lpr CD8⫹ T cell lines stably infected with either WT or mutated FasL molecules, B6.WT and B6.lpr thymocytes were used as targets and incubated with effector CD8⫹ T cells for 4 h in the presence of 4 mM MgCl2 and 4 mM EGTA. Apoptosis of CD4⫹CD8⫹ thymocytes was assessed by 7-AAD staining. To detect the intracellular expression of CID constructs, Jurkat cells were transfected with CID constructs 2 days before fixation with 4% paraformaldehyde and 0.1% saponin and staining with biotinylated anti-HA and rabbit antiFasL, followed by streptavidin-AF488 and goat anti-rabbit AF546, respectively. Stained cells were mounted on slides with Fluoromount-G medium (Southern Biotechnology Associates) and examined under a laser capture microdissection microscope (AS-LCM; Leica Microsystems). Images magnified by ⫻500 were captured with a Hamamatsu Orca ER digital camera using Wasabi software (Hamamatsu) and further processed using Photoshop (Adobe).
In vitro CKI assay L3 cells were lysed and FasL was immunoprecipitated as described (6). The beads with protein complexes were incubated with and without 200 M CKI inhibitor CKI-7 (US Biological) in 20 l of Assay Dilution Buffer I (Upstate) containing 500 ng of recombinant CKI (Upstate) and 10 Ci or [32P]ATP (PerkinElmer) per sample at 30°C for 10 min before boiling for 5 min. Supernatants were blotted onto P81 paper and washed three times with 0.75% phosphoric acid and once with acetone before transfer to a vial containing 5 ml of a scintillation fluid.
Results and Discussion Deleting aa 45–54 in the proline-rich region blocks delivery of a costimulatory signal
Previous results showed that the cytoplasmic domain of FasL is sufficient to mediate costimulation by driving NFAT nuclear translocation and AP-1 activation (6). A CID assay was used to further determine whether the proline-rich region is critical for FasL-mediated signal transduction (6). In brief, the and FasL cytoplasmic tails, lacking both TM and ligand binding EC domains, were targeted to the inner leaflet of the plasma membrane of transfected Jurkat cells through an M motif (Fig. 1A). These membrane-bound domains were then crosslinked by chemical inducers that oligomerize molecules containing
FIGURE 1. Deleting aa 45–54 in the proline-rich region blocks delivery of a costimulatory signal transduced by the cytoplasmic domain of FasL. A, The murine FasL CID construct encodes an N-terminal M motif to target the fusion protein into the plasma membrane, the FasL cytoplasmic domain lacking aa 45–54 in the proline-rich region, FKBP3, and a C-terminal HA epitope. B, The CID constructs are expressed well in Jurkat cells. Intracellular staining with anti-HA (in green) and anti-FasL (in red) indicate the proper expression levels and localization of the CID fusion proteins. Endogenous FasL expressed in the cytoplasm of Jurkat cells was also stained with anti-FasL. C, Jurkat cells cotransfected with combinations of CID constructs and firefly and Renilla luciferase reporters were subjected to CID analysis. The combinations were empty vector alone (V), plus empty vector (), FasL plus empty vector (F), and plus FasL (⫹F). Values are presented as the ratio of emission from the firefly reporter relative to the Renilla transfection control. The ratio of reporter to transfection control for cells incubated with medium alone was 4.0 –2.9 ⫻ 10⫺3. Each signal was normalized to the medium control and represents the mean ⫾ SD of two experiments. Asterisks denote significant differences relative to the -only control. Using the two-tailed Student’s t test on data from the optimal dimerizer concentration of WT transfectants, p ⱕ 0.006 for NFAT, p ⱕ 0.002 for AP-1, and p ⱕ 0.022 for IFN-␥.
FK506 binding domains. To assess the delivery of signals dependent on CID, Jurkat cells were cotransfected with the relevant and FasL CID constructs and firefly luciferase reporters of NFAT, AP-1, and IFN-␥ activity. Cotransfection with a Renilla luciferase reporter served as a transfection control. A deletional mutation lacking aa 45–54 (⌬45–54) was generated to disrupt the putative SH3 binding domains (Fig. 1A) while maintaining the integrity of the neighboring positively charged arginine residues thought to direct FasL into secretory vesicles, thereby enabling rapid activation-dependent cell surface expression in hemopoietic cells (14). Western blots developed with anti-HA indicated cytoplasmic expression of the protein encoded by this construct (not shown); by microscopy, the mutant was properly inserted into the plasma membranes of transfected cells (Fig. 1B). Despite this appropriate expression, co-crosslinking ⌬45–54 with the -chain resulted in impaired activation of the luciferase reporters of NFAT, AP-1, and IFN-␥ promoter activity (Fig. 1C). These data suggest that aa 45–54 in the proline-rich region of the FasL tail are important for FasL-mediated costimulation. Deleting aa 45–54 or aa 45–50 in the FasL tail does not affect FasL surface expression but inhibits FasL-mediated costimulation of CD8⫹ T cell proliferation
The ⌬45–54 mutation was incorporated into a construct (Fig. 2A) encoding a fusion protein of the murine FasL cytoplasmic
The Journal of Immunology
5641
and ⌬45–54 fusion proteins (Fig. 2E). However, crosslinking ⌬45–50 FasL molecules failed to costimulate CD8⫹ T cell proliferation (Fig. 2), indicating that aa 45–50 of the FasL cytoplasmic tail comprise one or more SH3 binding domains required to transmit a costimulatory signal delivered through FasL. Deleting aa 45–54 in the FasL tail inhibits Akt and Erk1/2 activation
FIGURE 2. Deleting aa 45–54 or 45–50 in the FasL tail does not affect FasL surface expression but inhibits FasL-mediated costimulation of CD8⫹ T cell proliferation. A and D, Schematic diagram of the mutated FasL/Ly49a chimeric constructs encoding the murine FasL cytoplasmic domain lacking aa 45–54 (A) or aa 45–50 (D) and Ly49a TM and EC domains. B and E, L3 CTLs infected with the indicated retroviral constructs were cultured with irradiated BALB/c stimulators on day 0 and stained with anti-Ly49a at the indicated times. Stained populations were gated on GFP⫹ cells and the normalized mean fluorescence intensity (MFI) plotted against time after Ag stimulation represents FasL/Ly49a surface expression. C and F, L3 CTL infectants were cultured with 0.5 g/ml plate-bound anti-CD3 alone or with the indicated doses of plate-bound antiLy49a. Proliferation was measured by [3H]TdR uptake in cultures pulsed on day 2. Each result is representative of two independent experiments.
tail and the Ly49a TM and EC domains (6). Ag-dependent, nontransformed, mouse L3 CD8⫹ T cells were infected with a retrovirus carrying this chimeric construct and infectants were identified using the GFP tag on the retroviral vector. Because the proline-rich region of the FasL tail plays an important role in regulating FasL trafficking and surface expression (14), the kinetics of surface expression of both WT and mutated chimeric proteins were examined upon Ag encounter (8). Up-regulation of the chimeric protein on the cell surface occurred after Ag encounter, peaked on day 2, and began to decrease by day 4 (Fig. 2B). Thus, deleting aa 45–54 does not affect the timing or extent of FasL surface expression, revealing that this domain is not required for regulating FasL transport in CD8⫹ T cells. However, ⌬45–54 failed to costimulate the TCR-induced proliferation of L3 infectants (Fig. 2C), whereas crosslinking the WT fusion protein boosted the anti-CD3-induced proliferation of the infectants in an anti-Ly49a dose-dependent manner. These data clearly indicate that aa 45–54 are critical for costimulation and that the defective function of the mutant may be associated with the inhibition of NFAT and AP-1 activation detected by the CID assay (Fig. 1). A more refined deletional mutation (⌬45–50) was incorporated into the FasL/Ly49a chimeric construct (Fig. 2D). Surface expression of ⌬45–50 followed the same kinetics as both WT
To define the impaired costimulation transduced by ⌬45–54, Akt and Erk1/2 phosphorylation were measured after culture of WT and ⌬45–54 L3 infectants with a suboptimal dose of platebound anti-CD3 alone, plate-bound anti-Ly49a alone, or both. Although co-crosslinking TCR and chimeric protein on WT infectants resulted in enhancement of Akt and Erk1/2 phosphorylation, the same treatment did not increase Akt and Erk1/2 activation in cells expressing mutant chimeric proteins (Fig. 3). Thus, aa 45–54 in the FasL tail may associate with the SH3-containing adaptors required for Akt and Erk1/2 activation upon FasL-mediated costimulation. The defective costimulation of CD8⫹ T cell proliferation (Fig. 2) appears to be due to a block of Erk1/2 activation via the PI3K pathway (Fig. 3). Partial dissociation between PI3K and FasL/Ly49a chimeric proteins lacking aa 45–54, as detected by coimmunoprecipitation (not shown), may lead to impaired Akt activation. ⌬45–54 induces normal killing of Fas⫹ target cells
The cytoplasmic domain of FasL has been reported to regulate the localization of FasL in lipid rafts, where it induces death through binding to Fas. However, it is controversial whether the proline-rich domain is required for optimal cytolysis (15– 17). To determine whether FasL lacking aa 45–54 in the tail can
FIGURE 3. Deleting aa 45–54 in the FasL tail inhibits Akt and Erk1/2 activation. L3 transfectants were cultured with medium alone, 0.5 g/ml platebound anti-CD3, 5 g/ml plate-bound anti-Ly49a, or both. The levels of phosphorylated Akt (A), phosphorylated Erk1/2 (B), total Erk1/2, and GFP expression were assessed on the same blot after 20 min of incubation. Each blot was developed first with the relevant anti-phospho-specific Abs and then with anti-Erk and anti-GFP. GFP and total Erk1/2 were used as protein loading controls. The tabulated values represent the relative ratios of the L3 band densities generated from phosphoproteins to those generated from GFP or Erk under each condition, which were further normalized to those generated with medium alone (relative ratios with medium alone were set at 1.0). There were similar levels of phospho-Akt and phospho-Erk1/2 under each condition in cells transfected with vector alone (not shown). Each result is representative of two independent experiments. ␣, Anti; p, phosphorylated.
5642
CUTTING EDGE: TWO MOTIFS WITHIN THE FasL TAIL REGULATE COSTIMULATION
bind to Fas and induce the apoptosis of Fas⫹ cells, FasLnull.lpr CTL lines were stably infected with constructs encoding murine FasL molecules with either WT or mutated cytoplasmic domains. In these CTL lines, the only FasL molecules were those provided by infection, and the absence of Fas avoided both suicide and fratricide. CD4⫹CD8⫹ thymocytes were used as target cells because they express high levels of Fas and are sensitive to Fas-mediated apoptosis (18). In addition to exhibiting efficient surface expression but impaired costimulation for Ag-specific CD8⫹ T cell proliferation (not shown), cells expressing ⌬ 45–54 were as cytolytic as those expressing WT FasL over a range of E: T ratios (Fig. 4A), delivering death signals upon interaction with Fas on WT but not on lpr target cells (Fig. 4B). ⌬45–54 therefore does not influence death induction but selectively blocks FasL-mediated costimulation. This deletional mutation should be useful for generating a knockin mouse to study FasL costimulation in vivo in the absence of lymphoproliferative disease (19). FasL is phosphorylated by CKI and FasL molecules with mutant CKI binding sites mediate defective NFAT activation
Serine residues in the FasL tail can be rapidly phosphorylated when L3 cells are stimulated through the TCR in the presence of plate-bound FasIgG (6), suggesting that CKI binding motifs provide a docking site that influences the costimulatory signal transduced by the FasL cytoplasmic domain. To investigate the role of CKI in FasL phosphorylation, an in intro CKI assay was performed using FasL immunoprecipitated from L3 cells as a substrate. Recombinant CKI can phosphorylate FasL in vitro, and this phosphorylation can be blocked by the CKI inhibitor CKI-7 (Fig. 5A). FasL immunoprecipitated from cells after stimulation with anti-CD3 and FasIgG showed a marked reduction in sensitivity to CKI phosphorylation in vitro (Fig. 5A). This suggests that CKI-sensitive sites are phosphorylated in vivo upon TCR and FasL co-crosslinking and are therefore unavailable to exogenous CKI in vitro. To further identify the in-
FIGURE 4. Cells expressing FasL molecules lacking aa 45–54 in the tail induce Fas-mediated death. A, FasLnull.lpr CD8⫹ T cell lines stably infected with the indicated constructs on day 3 after alloantigenic restimulation were incubated with WT thymocytes at the indicated E: T ratios for 4 h in the presence of EGTA to chelate Ca2⫹ and prevent perforin-mediated cytolysis. Apoptosis of CD4⫹CD8⫹ thymocytes was assessed by 7-AAD staining. B, The indicated effectors were cultured with either WT or lpr thymocytes at an E: T ratio of 8:1. The percentage (%) of the specific lysis of targets by effectors was normalized by subtracting the percentage of 7-AAD⫹ lpr thymocytes from the percentage of 7-AAD⫹ WT thymocytes. Data represent the mean ⫾ SD of three experiments. Using the two-tailed Student’s t test on the data from the WT and ⌬ 45–54 effectors, p ⫽ 0.52.
FIGURE 5. Recombinant CKI phosphorylates FasL and mutating serine/ threonine residues in the CKI binding sites of the FasL cytoplasmic domain inhibits NFAT activation. A, Equal numbers of L3 cells were incubated for 1 h with PBS or 0.72 g/ml plate-bound anti-CD3 plus 5 g/ml plate-bound FasIgG. FasL immunoprecipitated from the cell lysates was used as a CKI substrate in the presence (f) or absence (䡺) of CKI-7 as indicated. Equal amounts of FasL immunoprecipitated under each condition were subjected to SDSPAGE to normalize CKI activity. Data represent the mean ⫾ SD of two experiments. B, The murine FasL CID construct encodes an N-terminal M motif, the FasL cytoplasmic domain with alanine substitutions of serine/threonine residues (S/T3 A) in the CKI binding sites, FKBP3, and a C-terminal HA epitope. C, The localization of CID proteins was detected as described in Fig. 1B. D, The CID assay was performed and analyzed as described in Fig. 1C. Each signal represents the mean ⫾ SD of two independent experiments. Asterisks denote significant differences relative to the -only control. Using the twotailed Student’s t test on the data from the optimal dimerizer concentration, p ⱕ 0.05 (WT) for NFAT, p ⱕ 0.025 (WT) and p ⱕ 0.054 (S/T3 A) for AP-1, and p ⱕ 0.05 (WT) for IFN-␥. Incorporating these four alanine substitutions into the FasL/Ly49a construct resulted in apparent protein degradation (not shown), and the influence of the CKI sites in L3 costimulation could not be further investigated.
fluence of CKI binding sites on FasL costimulation, four point mutations were introduced by alanine substitution of S17, S18, T20, and S21 residues in the two conserved CKI recognition sites and incorporated into the CID construct (Fig. 5B). Western blot analyses indicated the cytoplasmic expression of the mutant protein (not shown), and microscopy showed its proper plasma membrane insertion (Fig. 5C). Drug-induced cocrosslinking of this mutant protein with the -chain enhanced neither NFAT activation or IFN-␥ reporter activity, whereas AP-1 activation was less impaired (Fig. 5D). These data suggest that CKI is responsible for phosphorylating FasL at serine/threonine residues to induce downstream signaling. Because the CKI consensus sequence is conserved in the cytoplasmic domains in 6 of 10 TNF family members that can signal bidirectionally, this result provides insight into the molecular mechanism of reverse signaling. The upstream pathway triggering
The Journal of Immunology
NFAT activation and the readout of this NFAT activation upon FasL costimulation are still not well defined. It is possible that the phosphorylation status of these serine/threonine residues is important for regulating the function of GSK3, a molecule involved in the PI3K pathway and known to promote nuclear export of NFAT (20). In contrast, costimulation through CD28 results in the phosphorylation of tyrosine residues and the further activation of downstream transcription factors (e.g., NF-B) essential for cell survival but not for cell proliferation or IL-2 production (4, 21). These results further demonstrate that the costimulatory pathways transduced by FasL and CD28 are distinct. The existence of costimulation transduced through reverse signaling is well established by now (3, 22, 23). However, a main challenge to studying this signaling pathway is that although the biological effects are clear, the underlying molecular mechanisms have proven difficult to trace. The identification of specific residues within the FasL cytoplasmic domain regulating costimulation confirms that FasL is a positive regulator of CD8⫹ T cells and can be used as a model for studying the costimulatory signaling induced by other TNF family molecules. In addition, the specific residues we identified are not essential for Fas-induced death, indicating that costimulation and death induction are separable functions of FasL involving distinct residues within the FasL tail.
Disclosures The authors have no financial conflict of interest.
References 1. Frauwirth, K. A., and C. B. Thompson. 2002. Activation and inhibition of lymphocytes by costimulation. J. Clin. Invest. 109: 296 –299. 2. Croft, M. 2003. Costimulation of T cells by OX40, 4 –1BB, and CD27. Cytokine Growth Factor Rev. 14: 265–273. 3. Sun, M., and P. J. Fink. 2007. A new class of reverse signaling costimulators belongs to the TNF family. J. Immunol. 179: 4307– 4312. 4. Rudd, C. E., and H. Schneider. 2003. Unifying concepts in CD28, ICOS and CTLA4 co-receptor signalling. Nat. Rev. Immunol. 3: 544 –556.
5643 5. Croft, M. 2003. Co-stimulatory members of the TNFR family: keys to effective T-cell immunity? Nat. Rev. Immunol. 3: 609 – 620. 6. Sun, M., K. T. Ames, I. Suzuki, and P. J. Fink. 2006. The cytoplasmic domain of Fas ligand costimulates TCR signals. J. Immunol. 177: 1481–1491. 7. Suzuki, I., and P. J. Fink. 1998. Maximal proliferation of cytotoxic T lymphocytes requires reverse signaling through Fas ligand. J. Exp. Med. 187: 123–128. 8. Suzuki, I., and P. J. Fink. 2000. The dual functions of Fas ligand in the regulation of peripheral CD8⫹ and CD4⫹ T cells. Proc. Natl. Acad. Sci. USA 97: 1707–1712. 9. Suzuki, I., S. Martin, T. E. Boursalian, C. Beers, and P. J. Fink. 2000. Fas ligand costimulates the in vivo proliferation of CD8⫹ T cells. J. Immunol. 165: 5537–5543. 10. Boursalian, T. E., and P. J. Fink. 2003. Mutation in Fas ligand impairs maturation of thymocytes bearing moderate affinity T cell receptors. J. Exp. Med. 198: 349 –360. 11. Hane, M., B. Lowin, M. Peitsch, K. Becker, and J. Tschopp. 1995. Interaction of peptides derived from the Fas ligand with the Fyn-SH3 domain. FEBS Lett. 373: 265–268. 12. Ghadimi, M. P., R. Sanzenbacher, B. Thiede, J. Wenzel, Q. Jing, M. Plomann, A. Borkhardt, D. Kabelitz, and O. Janssen. 2002. Identification of interaction partners of the cytosolic polyproline region of CD95 ligand (CD178). FEBS Lett. 519: 50 –58. 13. Karray, S., C. Kress, S. Cuvellier, C. Hue-Beauvais, D. Damotte, C. Babinet, and M. Levi-Strauss. 2004. Complete loss of Fas ligand gene causes massive lymphoproliferation and early death, indicating a residual activity of gld allele. J. Immunol. 172: 2118 –2125. 14. Blott, E. J., G. Bossi, R. Clark, M. Zvelebil, and G. M. Griffiths. 2001. Fas ligand is targeted to secretory lysosomes via a proline-rich domain in its cytoplasmic tail. J. Cell Sci. 114: 2405–2416. 15. Jodo, S., V. J. Pidiyar, S. Xiao, A. Furusaki, R. Sharma, T. Koike, and S. T. Ju. 2005. Fas ligand (CD178) cytoplasmic tail is a positive regulator of Fas ligand-mediated cytotoxicity. J. Immunol. 174: 4470 – 4474. 16. Nachbur, U., D. Kassahn, S. Yousefi, D. F. Legler, and T. Brunner. 2006. Posttranscriptional regulation of Fas (CD95) ligand killing activity by lipid rafts. Blood 107: 2790 –2796. 17. Cahuzac, N., W. Baum, V. Kirkin, F. Conchonaud, L. Wawrezinieck, D. Marguet, O. Janssen, M. Zornig, and A. O. Hueber. 2006. Fas ligand is localized to membrane rafts, where it displays increased cell death-inducing activity. Blood 107: 2384 –2391. 18. Ogasawara, J., T. Suda, and S. Nagata. 1995. Selective apoptosis of CD4⫹ CD8⫹ thymocytes by the anti-fas antibody. J. Exp. Med. 181: 485– 491. 19. Nagata, S., and T. Suda. 1995. Fas and Fas ligand: lpr and gld mutations. Immunol Today 16: 39 – 43. 20. Kane, L. P., and A. Weiss. 2003. The PI-3 kinase/Akt pathway and T cell activation: pleiotropic pathways downstream of PIP3. Immunol. Rev. 192: 7–20. 21. Barbulescu, K., C. Becker, J. F. Schlaak, E. Schmitt, K. H. Meyer zum Buschenfelde, and M. F. Neurath. 1998. IL-12 and IL-18 differentially regulate the transcriptional activity of the human IFN-␥ promoter in primary CD4⫹ T lymphocytes. J. Immunol. 160: 3642–3647. 22. Janssen, O., J. Qian, A. Linkermann, and D. Kabelitz. 2003. CD95 ligand– death factor and costimulatory molecule? Cell Death Differ. 10: 1215–1225. 23. Eissner, G., W. Kolch, and P. Scheurich. 2004. Ligands working as receptors: reverse signaling by members of the TNF superfamily enhance the plasticity of the immune system. Cytokine Growth Factor Rev. 15: 353–366.
