Subcellular compartmentalization of FADD as a new level of regulation in death receptor signaling Niko Fo ¨ ger 1 , Silvia Bulfone-Paus 1 , Andrew C. Chan 2 and Kyeong-Hee Lee 1 1 Department of Immunology and Cell Biology, Research Center Borstel, Leibniz Center for Medicine and Biosciences, Germany 2 Department of Immunology, Genentech, Inc., San Francisco, CA, USA Introduction CD95 (Fas ⁄ Apo-1 ⁄ TNFRSF6) is a prototypic death receptor belonging to the tumor necrosis factor recep- tor superfamily. CD95 is expressed on the surface of cells as preassociated homotrimers and, upon CD95L binding, undergoes a conformational change to reveal its cytoplasmic death domain (DD) to favor homotypic interactions with other DD-containing proteins. Fas- associated protein with DD (FADD) is the most proxi- mal adaptor molecule transmitting the death signal mediated by CD95 [1]. As a DD-containing and death effector domain-containing proapoptotic adaptor molecule, FADD is essential to recruit the initiator caspases-8 and -10 to instigate formation of the death- inducing signal complex (DISC), which mediates death receptor-induced apoptosis [2,3]. Expression of a dominant-negative form of FADD, consisting of the N-terminal DD only, impairs the relay of the apoptotic signal from death receptors [4]. Moreover, FADD- deficient mice display profound defects in apoptotic pathways, particularly in the immune system [5]. FADD is a multifunctional protein that, in addition to its prominent role in cell death, has also been implicated in the regulation of cell survival ⁄ proliferation and cell cycle progression, as well as embryonic develop- ment [5–7]. In our previous work, we demonstrated that CD95 internalization plays a role in CD95-induced apoptosis [8]. Upon ligand binding, CD95 is internalized and delivered to endosomal compartments, which then serve as major sites for CD95-mediated DISC forma- tion and caspase-8 activation. Given that the key role of FADD in apoptotic signaling is efficient DISC Keywords apoptosis; CD95; compartmentalization; FADD; nuclear trafficking Correspondence K H. Lee, Department of Immunology and Cell Biology, Research Center Borstel, Leibniz Center for Medicine and Biosciences, Parkallee 22, 23845 Borstel, Germany Fax: +49 4537 1884904 Tel: +49 4537 188585 E-mail: klee@fz-borstel.de (Received 30 April 2009, accepted 4 June 2009) doi:10.1111/j.1742-4658.2009.07134.x Fas-associated protein with death domain (FADD) is an essential adaptor protein in death receptor-mediated signal transduction. During apoptotic signaling, FADD functions in the cytoplasm, where it couples activated receptors with initiator caspase-8. However, in resting cells, FADD is pre- dominantly stored in the nucleus. In this study, we examined the modalities of FADD intracellular trafficking. We demonstrate that, upon CD95 acti- vation, FADD redistributes from the nucleus to the cytoplasm. This induc- ible nuclear–cytoplasmic translocation of FADD is independent of CD95 internalization, formation of the death-inducing signaling complex, and caspase-8 activation. In contrast to nuclear export of FADD, its subse- quent recruitment and accumulation at endosomes containing internalized CD95 requires a caspase-8-dependent feedback loop. These data indicate the existence of differential pathways directing FADD nuclear export and cytoplasmic trafficking, and identify subcellular compartmentalization of FADD as a novel regulatory mechanism in death receptor signaling. Abbreviations BFA, brefeldin A; DAPI, 4¢,6-diamidino-2-phenylindole; DD, death domain; DISC, death-inducing signaling complex; EEA-1, early endosome antigen 1; FADD, Fas-associated protein with death domain; GFP, green fluorescent protein; INP54p, Saccharomyces cerevisiae inositol polyphosphate 5-phosphatase; PtdIns(4,5)P 2, phosphatidylinositol 4,5-bisphosphate. 4256 FEBS Journal 276 (2009) 4256–4265 ª 2009 The Authors Journal compilation ª 2009 FEBS assembly at endosomal structures, FADD is expected to function within the cytoplasm. However, FADD carries strong nuclear localization and nuclear export signals, and has been reported to primarily localize to the nucleus in a variety of different cell types [9–12]. This raises the question of how a predomi- nantly nuclear protein such as FADD is involved in DISC formation occurring at endosomes in the cytoplasm. Here, we demonstrate that CD95 stimulation induces translocation of nuclear FADD to the cyto- plasm. Employing a combination of biochemical, cell biological and genetic methods, we investigated the role of ‘classic’ apoptotic signal transduction events in the nuclear–cytoplasmic relocalization of FADD and its subsequent recruitment to endosomal compart- ments, where FADD promotes efficient DISC forma- tion. The regulation of the subcellular localization of FADD adds a new level of complexity to the apoptotic signaling cascade. Results Nuclear–cytoplasmic redistribution of FADD in response to CD95L stimulation To explore whether FADD shuttles between the nucleus and the cytoplasm in response to an apoptotic stimulus, we analyzed the subcellular distribution of FADD in resting versus CD95L-treated BJAB cells, a human B-cell Burkitt’s lymphoma cell line (Fig. 1A). In agreement with previous reports on other cell lines, FADD colocalizes with the nuclear stain 4¢,6-diamidi- no-2-phenylindole (DAPI) in resting BJAB cells, as well as in human peripheral blood CD4 + T-lympho- cytes, indicating preferential nuclear localization of FADD (Fig. 1A,B, left panels). In response to CD95 receptor triggering, however, FADD redistributed from a predominantly nuclear to a nuclear and cyto- plasmic pattern. In BJAB cells, within 5 min of CD95L treatment, a significant proportion of FADD relocalized from the nucleus to the cytoplasm and exhibited dispersed fine punctuate patterns in the cyto- plasm (Fig. 1A, middle panel). These structures became more pronounced and enlarged at 15–30 min after CD95L stimulation (Fig. 1A, right panel). A similar redistribution of FADD was also observed in human peripheral blood CD4 + T-lymphocytes (Fig. 1B, right panel). These observations indicate that FADD undergoes regulated redistribution from the nucleus to the cyto- plasm in response to CD95 triggering. Notably, we did not observe recruitment of FADD to the plasma mem- brane, but, instead, FADD relocalized to vesicular structures in the cytoplasm. This specific vesicular localization of FADD is probably due to functional association of FADD with internalized CD95, which predominantly occurs at endosomal compartments and constitutes an essential step in CD95-mediated apop- totic signaling [8]. 0 minCD95L: 15 min FADD B FADD/DAPI 0 minCD95L: 5 min 15 min A FADD/DAPI FADD Fluorescence intensity Low High Fig. 1. Nuclear–cytoplasmic translocation of FADD in response to CD95 stimulation. (A) BJAB cells were stimulated with CD95L for the indicated times. Cells were stained for FADD (red), and nuclei were counterstained with DAPI (blue). Overlay fluorescence is shown in the upper panel. Quantitative image analysis with relative pixel intensities recorded for FADD fluorescence signals is shown in the lower panel. (B) Activated human peripheral blood CD4 + T-cells were stimulated with CD95L for 15 min. Single FADD staining (upper panel, red) and overlay fluorescence (lower panel) of FADD and DAPI are shown. Fluorescence images were generated by deconvolution microscopy. The data shown are representative of > 150 cells analyzed. N. Fo ¨ ger et al. FADD trafficking and CD95 signaling FEBS Journal 276 (2009) 4256–4265 ª 2009 The Authors Journal compilation ª 2009 FEBS 4257 Expression of a plasma membrane-localized phosphatidylinositol 4,5-bisphosphate [PtdIns(4,5)P 2 ]-specific 5¢-phosphatase inhibits CD95 endocytosis and apoptosis, but not the nuclear–cytoplasmic translocation of FADD As FADD translocation from the nucleus to the cyto- plasm occurred within 2–5 min following CD95L stim- ulation, prior to significant CD95 internalization, we analyzed whether FADD translocation required CD95 internalization. To this end, we utilized Saccharomy- ces cerevisiae inositol polyphosphate 5-phosphatase (INP54p), an enzyme that hydrolyzes PtdIns(4,5)P 2 to phosphatidylinositol 4-phosphate [13]. Cellular levels of PtdIns(4,5)P 2 are tightly regulated, and it plays impor- tant roles in a multitude of cellular functions, including clathrin-mediated endocytosis [14–16]. Expression of a green fluorescent protein (GFP)-tagged plasma mem- brane-targeted INP54p (FynC–GFP–INP54p) in BJAB cells specifically reduces PtdIns(4,5)P 2 levels in the plasma membrane, and results in the inhibition of CD95L-induced CD95 receptor endocytosis and apop- tosis [8] (Fig. 2A,B). BJAB cells transfected with FynC–GFP–INP54p did, however, still relocalize FADD from the nucleus to the cytoplasm in response to CD95L stimulation (Fig. 2C). Whereas the overall degree of the CD95L-induced FADD nuclear–cytoplas- mic translocation was similar between FynC–GFP– INP54p + cells and control cells, the pattern of FADD staining was qualitatively distinct in FynC–GFP– INP54p + cells. At 15 min following CD95 activation, FynC–GFP–INP54p + cells (Fig. 2C, middle panel) showed only a diffuse staining pattern of cytoplasmic FADD and did not exhibit the intense coalescence of FADD with larger endocytic structures that is observed in FynC–GFP–INP54p ) cells (Fig. 2C, right panel). This may reflect a lack of internalized CD95 to concen- trate FADD within endocytic vesicles. 0 CD95L: 0 min 15 min 15 min 10 20 30 40 50 60 70 80 90 100 GFP – GFP + GFP – GFP + FynC–GFP–INP54 FynC–GFP A 10 0 10 1 10 2 10 3 10 4 10 0 10 1 10 2 10 3 10 4 FynC–GFP+ FynC –GFP–INP54+ Annexin V Cell count B CD95 internalization (%) C GFP FADD DAPI FADD Low Hi g h Fluorescence intensity 6 5 4 3 1 2 Fig. 2. FADD translocation into the cytoplasm is independent of CD95 internalization. (A, B) BJAB cells transiently expressing FynC–GFP– INP54p, a PtdIns(4,5)P 2 -specific 5¢-phosphatase–GFP fusion construct, or the control construct FynC–GFP were analyzed for CD95 internali- zation (A) and apoptosis (B) following CD95L stimulation for 30 min (A) and 6 h (B), respectively. (A) The remaining surface CD95 was detected by FACS analysis, and the percentage of CD95 downregulation was calculated for the GFP + and GFP ) populations. (B) Apoptosis in GFP + (red) cells was assessed by annexin V staining and FACS analysis. Nonstimulated cells are shown in gray. The data shown are representative of three experiments. (C) BJAB cells transiently transfected with FynC–GFP–INP54p were stimulated with CD95L for 0 min (1, 4) and 15 min (2, 3, 5, 6). Panels 1, 2, 4, 5 show FynC–GFP–INP54p-expressing cells (GFP + ), and FynC–GFP–INP54p-non-expressing cells are shown in the right panel (3, 6). Cells were stained for DAPI (blue) and FADD (red). Overlay fluorescence is shown in the upper panel (1–3), and quantitative image analysis of CD95 fluorescence signals is shown in the lower panel (4–6). The data shown are representa- tive of > 50 cells analyzed. FADD trafficking and CD95 signaling N. Fo ¨ ger et al. 4258 FEBS Journal 276 (2009) 4256–4265 ª 2009 The Authors Journal compilation ª 2009 FEBS CD95 internalization promotes endosomal targeting of FADD To investigate whether internalized CD95 provides a docking signal to recruit FADD to endosomes, we analyzed the subcellular localization of FADD and CD95 in CD95L-activated FynC–GFP–INP54p + and FynC–GFP–INP54p ) BJAB cells. Following stimula- tion for 30 min with CD95L, colocalization of cyto- plasmic FADD with internalized CD95 was readily detected at intracellular compartments in FynC– GFP–INP54p ) cells (Fig. 3A, panels 5–8). In con- trast, in CD95-activated but endocytosis-defective FynC–GFP–INP54p + BJAB cells, CD95 had formed microaggregates in the plasma membrane, and no significant colocalization between FADD and CD95 was observed, although FADD could be readily detected in the cytoplasm (Fig. 3A, panels 1–4). There was minimal overlap of staining for FADD with the early endosome marker early endosome antigen 1 (EEA-1) in resting cells (Fig. 3B, panels 1–3). Overlap of staining for FADD and EEA-1 was, however, readily detected in CD95L-stimulated control FynC–GFP–INP54p ) cells (Fig. 3B, panels 9–11), whereas in FynC–GFP–INP54p + BJAB cells, FADD largely failed to accumulate at EEA-1 + en- dosomes (Fig. 3B, panels 5–8). An internalization-defective CD95 mutant disrupts apoptotic signaling but still induces FADD nuclear–cytoplasmic translocation To further analyze the interrelationship between CD95 internalization and FADD nuclear–cytoplasmic relo- calization, we specifically interfered with CD95 receptor endocytosis by employing the internalization- defective CD95(Y291F) mutant [8]. The ability of this mutant form of CD95, in which Tyr291 within the consensus AP-2-binding motif of CD95 has been mutated to Phe, to internalize in murine A20 B-lym- phoma cells following stimulation with a mAb against human CD95 (CH-11) was significantly reduced as compared to wild-type CD95 (Fig. 4A). Concomi- tantly, the ability of CD95(Y291F)-expressing cells to activate caspase-8 in response to CD95 stimulation was similarly compromised (Fig. 4C). However, despite the relative inability of CD95(Y291F) to internalize and to induce classic proximal apoptotic signaling B FADD EEA-1 FADD / EEA-1 FADD / DAPI 1 2 3 8 4 7 6 5 9 10 11 12 CD95L 0 min (GFP +) 30 min (GFP +) 30 min (GFP–) A GFP CD95 FADD CD95 / FADD 4 3 2 1 CD95L 30 min 30 min 8 7 6 5 Fig. 3. CD95 internalization promotes endosomal targeting of FADD. (A) BJAB cells were transfected with FynC–GFP– INP54p and stimulated with CD95L for 30 min. Cells were stained for CD95 (red) and FADD (blue). Panels 1–4 represent a FynC–GFP–INP54p-expressing (GFP + ) cell, and panels 5–8 show a FynC–GFP–INP54p- non-expressing (GFP ) ) cell. The data shown are representative of > 50 cells analyzed. (B) BJAB cells were transfected with FynC–GFP–INP54p and stimulated with CD95L for 0 min (1–4) and 30 min (5–12). Cells were stained for FADD (green), EEA-1 (red), and DAPI (blue). Individual and merged fluorescence images were obtained by deconvolution microscopy. FynC–GFP– INP54p-expressing cells (GFP + ) are shown in panels 1–8, and a FynC–GFP–INP54p-non- expressing cell (GFP ) ) is shown in pan- els 9–12. The data shown are representative of > 100 cells analyzed. N. Fo ¨ ger et al. FADD trafficking and CD95 signaling FEBS Journal 276 (2009) 4256–4265 ª 2009 The Authors Journal compilation ª 2009 FEBS 4259 events, stimulation of CD95(Y291F) still induced nuclear–cytoplasmic relocalization of FADD (Fig. 4A,B). FADD was preferentially localized within the nucleus of resting cells expressing CD95(Y291F). In response to CD95 stimulation, FADD exhibited a nuclear and cytoplasmic distribution in cells expressing either wild- type CD95 or CD95(Y291F). However, whereas in wild-type human CD95-expressing cells FADD concentrated and colocalized with internalized CD95 at EEA-1-positive endosomal compartments, in cells expressing the internalization mutant CD95(Y291F) FADD remained in a diffuse cytoplasmic pattern and showed no significant colocalization with EEA-1. The data on the nuclear–cytoplasmic relocalization of FADD, as observed by deconvolution microscopy, were further confirmed by biochemical subcellular frac- tionation experiments. Little to no FADD protein was detected in the cytoplasmic fraction of nonstimulated cells transfected with either wild-type human CD95 or CD95(Y291F) (Fig. 4D, lanes 3 and 6). Triggering of human CD95 for 15–30 min induced a significant increase in the amount of FADD in the cytoplasmic fraction of cells expressing wild-type CD95 (Fig. 4D, lanes 4 and 5). A similar increase in cytoplasmic FADD was also observed in CD95(Y291F)-expressing cells stimulated with antibody against human CD95 (Fig. 4D, lanes 7 and 8). Together, these data indicate that CD95L-induced FADD translocation to the cytoplasm occurs indepen- dently of CD95 internalization. However, internalized CD95 then probably serves as a scaffold to amplify and ⁄ or stabilize FADD assembly at endosomal com- partments. Inhibition of caspase-8 activation allows for transient nuclear–cytoplasmic shuttling of FADD and results in the recycling of CD95 To further investigate whether inhibition of apoptotic signaling affects the subcellular localization of CD95 and ⁄ or FADD, BJAB cells were treated with the caspase-8 inhibitor z-IETD, and FADD localization B Y291F Y291F WT FADD EEA-1 FADD/EEA-1 1 2 3 6 5 4 7 8 9 C WT Y291F WB: Cas-8 hCD95 CD95 : 0’ 15’ 30’ 60’ 0’ 15’ 30’ 60’ CD95: 0’ 0’ 0’ 15’ 30’ 30’ 0’ 15’ 1 2 3 4 5 6 7 8 WB: FADD Laminin GDI-Rho WT Y291F WT Y291F Nuclear Cytoplasmic D 1 2 3 4 5 6 7 8 A Y291F WT FADD CD95 CD95/FADD 1 2 3 6 5 4 Fig. 4. Cytoplasmic translocation of FADD in cells expressing the internalization mutant of human CD95 (hCD95). (A) A20 cells expressing the internalization mutant hCD95(Y291F) (1–3) or wild- type (WT) hCD95 (4–6) were activated for 30 min with mAb against hCD95 (CH-11). Cells were subsequently stained for FADD (green), CD95 (red), and DAPI (blue). (B) A20 cells expressing hCD95(Y291F) (1–6) or wild-type hCD95 (7–9) were activated with CH-11 for 0 min (1–3) or 30 min (4–9). Cells were stained for FADD (green), EEA-1 (red), and DAPI (blue). Images were obtained by deconvolution microscopy. The data shown are representative of > 60 cells analyzed. (C) A20 cells were transfected with wild-type hCD95 or hCD95(Y291F) and stimulated with biotinylated CH-11 for the indicated times. Human CD95-associated signaling complexes were isolated using streptavidin-conjugated beads. Association of caspase-8 with hCD95 was analyzed by immunoblotting for cas- pase-8 and hCD95. (D) A20 cells were transfected with wild-type hCD95 or hCD95(Y291F) and stimulated with CH-11 for the indi- cated times. Nuclear (lanes 1 and 2) and cytoplasmic (lanes 3–8) fractions were prepared from total cellular lysates and were immu- noblotted using antibody against FADD. Effective separation of nuclear and cytoplasmic fractions was controlled for by immuno- blotting for laminin (nuclear marker) and GDI-Rho (cytosolic marker). FADD trafficking and CD95 signaling N. Fo ¨ ger et al. 4260 FEBS Journal 276 (2009) 4256–4265 ª 2009 The Authors Journal compilation ª 2009 FEBS was investigated. In unstimulated BJAB cells either treated or not treated with the caspase-8 inhibitor z- IETD, FADD was predominantly detected in the nucleus (Fig. 5A, panels 1–3 and 13–15). Within 2 min of stimulation with CD95L, FADD could readily be detected in the cytoplasm of z-IETD-treated cells (Fig. 5A, panels 16–18), as in control cells. In untreated control cells, FADD remained in the cyto- plasm after 30 and 60 min of CD95 stimulation, and cells started to exhibit signs of apoptosis (Fig. 5A, panels 7–12). In contrast, in z-IETD-treated cells, which do not undergo apoptosis, significant amounts of cytoplasmic FADD could only be detected within 30 min of CD95L stimulation (Fig. 5A, panels 19–20). At 60 min, only minimal amounts of FADD had remained in the cytoplasm of z-IETD-treated cells (Fig. 5A, panels 22–24). Thus, inhibition of caspase-8 activation does not affect the initial nuclear–cytoplas- mic translocation of FADD; however, FADD relocal- ization to the cytoplasm is not persistent under these conditions. Whether, in the absence of caspase-8 acti- vation, FADD shuttles back to the nucleus or is degraded in the cytoplasm remains to be investigated. As treatment of cells with caspase inhibitors has been reported to be required for CD95 internalization following receptor activation [17], we next analyzed the kinetics with which caspase inhibition may affect receptor internalization. Treatment of BJAB cells with the inhibitors z-IETD (caspase-8 selective), z-VAD (a general caspase inhibitor) or z-DEVD (caspase-3 selec- tive) did not affect ligand-mediated CD95 internaliza- tion at 15 min and had moderate effects at 30 min as compared to untreated cells (Fig. 5B,C). Between 30 min and 60 min, control cells further downregulated CD95, whereas in cells treated with caspase inhibitors an increase in CD95 surface expression was observed. These kinetics were further supported by microscopy studies, in which CD95 was detected within the cyto- plasm within 30 min following CD95L stimulation, even in the presence of z-IETD (Fig. 5A, panel 20). At 60 min following CD95L stimulation, when CD95 had maximally internalized and cells already demonstrated morphological changes associated with apoptosis (Fig. 5A, panel 11), CD95 was detected almost exclu- sively at the cell surface in cells treated with caspase inhibitors (Fig. 5A, panel 23; Fig. 5B,C), as previously reported [17]. To analyze the potential contributions of CD95 recycling to the plasma membrane, cells were treated with brefeldin A (BFA), a fungal metabolite that blocks protein transport from the endoplasmic reticu- lum to the Golgi and protein recycling, in the presence or absence of z-VAD. Whereas cells incubated with z-VAD alone again demonstrated significant down- regulation of surface CD95 expression at 30 min followed by an increase at 60 min, cells treated with z-VAD and BFA continued to downregulate CD95 without any subsequent increase in surface CD95 expression (Fig. 5D,E). Thus, CD95 internalization following receptor engagement is not dependent on caspase activation, and a significant proportion of the surface expression of CD95 observed at 30 min and 60 min following receptor engagement in the presence of caspase inhibitors appears to be a consequence of CD95 receptor recycling when cells are unable to undergo apoptosis. Microscopic analysis of CD95- stimulated cells treated with both BFA and the cas- pase-8 inhibitor z-IETD showed that CD95 largely accumulated in the cytoplasm and significant amounts of FADD localized to the cytoplasm, but CD95 and FADD failed to interact with each other under these conditions (Fig. 5F). These data indicate that nuclear– cytoplasmic shuttling of FADD is independent of cas- pase-8 activity. Further recruitment of FADD to CD95-containing endosomal compartments, however, seems to require an activation loop involving active caspase-8. Discussion FADD is an essential adaptor protein in the CD95- mediated apoptotic signaling cascade that couples activated receptors with the activation of initiator caspase-8 [1,18,19]. Here, we demonstrate that, in response to CD95 receptor activation, a significant amount of FADD relocalizes from the nucleus to the cytoplasm. Our data indicate that CD95 receptor triggering induces membrane proximal signals to induce nuclear export of FADD that are independent of CD95 inter- nalization and ‘classic’ apoptotic signaling events, such as DISC formation and caspase-8 activation. We employed two different experimental systems to inhibit CD95 internalization: modulation of PtdIns(4,5)P 2 levels by INP54p, and the internalization mutant CD95(Y291F). In these systems, CD95-induced DISC formation, caspase-8 activation and apoptosis are severely compromised [8], whereas CD95 triggering still induces translocation of FADD from the nucleus to the cytoplasm. Subsequent recruitment and concentration of FADD to endosomal compartments, where DISC is stabilized and amplified, however, requires CD95 inter- nalization. Consequently, in endocytosis-defective cells, FADD did not accumulate at endosomal structures in response to CD95 stimulation, but exhibited more dif- fuse localization in the cytoplasm. Thus, internalized N. Fo ¨ ger et al. FADD trafficking and CD95 signaling FEBS Journal 276 (2009) 4256–4265 ª 2009 The Authors Journal compilation ª 2009 FEBS 4261 0 10 20 30 40 50 60 70 80 90 0 204060 No inhibitor z-IETD (Cas-8) z-VAD (general) z-DEVD (Cas-3 & Cas-7) (min) MFI No inhibitor z-IETD z-VAD z-DEVD Cell count CD95 0 min 5 min 30 min CD95 Cell count z-VAD z-VAD + BFA 0 min 5 min 30 min BFA No inhibitor 0 20 40 60 80 100 120 0 204060 No inhibitor BFA z-VAD z-VAD + BFA (min) MFI FADD CD95 CD95/FADD CD95L (30 min) z-IETD BFA CD95L (30 min) BFA z-IETD (caspase-8 inhibitor)No inhibitor CD95L A BC D E F 0 min 2 min 30 min 60 min FADD CD95 CD95/FADD FADD CD95 CD95/FADD 123 456 789 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Fig. 5. FADD translocation is independent of caspase-8 activation. (A) BJAB cells were stimulated with CD95L for the indicated times in the absence (left, 1–12) or presence (right, 13–24) of 50 l M caspase-8 inhibitor z-IETD. Cells were stained for FADD (green), CD95 (red), and DAPI (blue). Images were obtained by deconvolution microscopy. The data shown are representative of > 30 cells analyzed. (B, C) BJAB cells were pretreated with the caspase inhibitor zIETD-fmk, zVAD-fmk or zDEVD-fmk for 1 h. Cells were then stimulated with CD95L for the indicated times, and surface CD95 expression was assessed by FACS. Changes in mean fluorescence intensity (MFI) are quantified in (C). (D, E) BJAB cells were pretreated with either BFA (10 lgÆmL )1 ), 50 lM z-VAD-fmk or both for 30 min. Cells were then stimulated with CD95L for the indicated times, and surface CD95 expression was assessed by FACS (D). Changes in MFI are quantified in (E). The data shown are representative of three independent experiments. (F) BJAB cells were stimulated with CD95L for 30 min in the presence of BFA (10 lgÆmL )1 ) (upper panel) or with the combination of BFA (10 lgÆmL )1 ) and 50 lM caspase-8 inhibitor z-IETD (lower panel). Cells were stained for FADD (green), CD95 (red), and DAPI (blue). Individual and merged fluorescence images were obtained by deconvolution micros- copy. The data shown are representative of > 50 cells analyzed. FADD trafficking and CD95 signaling N. Fo ¨ ger et al. 4262 FEBS Journal 276 (2009) 4256–4265 ª 2009 The Authors Journal compilation ª 2009 FEBS CD95 within the endosome appears to provide a local- izing signal for further recruitment of FADD. This specific recruitment of FADD to internalized CD95 is, however, severely compromised in the presence of a caspase-8 inhibitor, even when accumulation of internalized CD95 is forced by treatment of cells with BFA. Hence, CD95 internalization is required, but is not sufficient, for endosomal accumulation of FADD. Noteworthy, in BJAB cells treated with caspase-8 inhibitors, internalized CD95 appears to recycle to the cell surface, and CD95-induced FADD shuttling to the cytoplasm is only of a transient nature. Our data suggest a sequential model of signaling in which CD95 receptor activation generates early signals at the plasma membrane that lead to the translocation of nuclear FADD to the cytoplasm. In a process that depends on a positive feedback loop involving caspase- 8 activation, cytoplasmic FADD is then further recruited to internalized CD95 at endosomal struc- tures, leading to efficient DISC assembly and amplifi- cation and eventually to apoptotic cell death. Nuclear localization of FADD can be regulated by phosphorylation at Ser194, which is required for the interaction of FADD with the nuclear–cytoplasmic transport receptor exportin-5 [10]. The phosphoryla- tion of FADD does not, however, appear to play a significant role in the induction of apoptosis by CD95 [20], but is, rather, involved in the nonapoptotic functions of FADD, such as regulation of cell cycle progression [21,22]. Another signaling event potentially involved in the translocation of FADD from the nucleus to the cytoplasm is CD95-induced generation of ceramide. A recent report has implicated ceramide in the regulation of nucleocytoplasmic trafficking in smooth muscle cells [23]. It is currently unclear whether CD95-induced ceramide exhibits a similar regulatory function during apoptosis. Also, whether or not CD95-mediated ceramide generation, like CD95- mediated FADD translocation, is independent of caspase-8 activation is still controversial [24–26]. Thus, the molecular mechanisms involved in the regulation of FADD subcellular localization during apoptotic signaling await further investigation. What is the biological function of nuclear FADD and its nuclear–cytoplasmic translocation? Functional DISC assembly and activation of caspase-8 is generally considered to be a ‘point of no return’ in the apoptotic signaling cascade. Thus, trapping FADD in the nucleus and away from the cytoplasm, where the other compo- nents of DISC can be found, may serve as a safety mechanism to protect cells from unwanted spontaneous DISC formation and apoptosis. Mutation of the nuclear export signal within FADD, such that FADD is retained within the nucleus, reduces the death- inducing efficacy of FADD. Only upon specific CD95-induced signals does FADD relocalize to the cytoplasm, promoting CD95–FADD association, which in turn leads to DISC assembly, caspase-8 activation, and apoptotic cell death. In addition, nuclear FADD may be involved in other, nonapoptotic functions of FADD, such as the control of cell cycling and prolifer- ation of lymphoid cells or embryonic development [5,7,21,27]. Nuclear FADD has also been implicated in genome surveillance through its association with the DNA repair molecule MBD4 [10]. Like FADD, the tumor necrosis factor receptor 1-associated DD- containing adaptor protein TRADD also rapidly shut- tles between the nucleus and the cytoplasm. Whereas cytoplasmic TRADD mediates apoptosis through FADD and caspase-8 activation, nuclear TRADD acts through a mitochondrial apoptosis pathway [28]. Our study provides, for the first time, experimental evidence for the regulation of nuclear cytoplasmic shut- tling of FADD by CD95-mediated signals, suggesting a new level of regulation in death receptor signaling. As the specific relocalization of FADD from the nucleus to the cytoplasm is independent of CD95 receptor internalization, DISC assembly at endosomes and cas- pase activation, our data indicate that CD95 triggering induces additional, plasma membrane proximal signals. The elucidation of the molecular pathways involved in connecting CD95 signaling to the compartmentaliza- tion of FADD will help us to better understand the regulatory mechanisms in death receptor signaling and may lead to new avenues in apoptosis research. Experimental procedures Cells Human Burkitt lymphoma BJAB cells and murine A20 B-lymphoma cells were cultured in RPMI-1640 supple- mented with 10% fetal bovine serum, penicillin ⁄ streptomycin (50 lgÆmL )1 each) and 2 mml-glutamine (RPMI standard medium). Cells were maintained in 5% CO2 at 37 °C. CD4 + human peripheral blood T-lymphocytes were isolated from heparinized blood of healthy donors with the Rosette Sep Kit (Stem Cell Technologies, Vancouver, Canada) and subsequent Ficoll-Hypaque density centrifugation. Freshly isolated CD4+ human peripheral blood T-lymphocytes were activated with mAbs against CD3 (1 lgÆmL )1 , UCHT1; BD Pharmingen, Franklin Lakes, NJ, USA) and CD28 (5 lgÆmL )1 , CD28.2; BD Pharmingen), and maintained in RPMI-1640 standard medium containing recombinant human interleukin-2 (R&D Systems, Minneapolis, MN, USA; 25 UÆmL )1 ). N. Fo ¨ ger et al. FADD trafficking and CD95 signaling FEBS Journal 276 (2009) 4256–4265 ª 2009 The Authors Journal compilation ª 2009 FEBS 4263 DNA constructs and transfection The DNA constructs have been described previously [8]. The catalytic domain of INP54p was cloned into the modi- fied pEGFP-C1 (Clontech, Mountain View, CA, USA) vector following the C-terminus of GFP. The first 10 amino acids of Fyn were engineered in frame N-terminal to GFP (FynC–GFP–INP54p). Human CD95 was inserted into pcDNA4 ⁄ TO (Invitrogen, Carlsbad, CA, USA), and the specific amino acid mutation (Y291F) was generated using the QuickChange site-directed mutagenesis kit (Stratagene, La Jolla, CA, USA). Plasmids were transfected using the Nucleofector (Lonza, Ko ¨ ln, Germany) transfection system according to the manufacturer’s instructions. Cell stimulation and apoptosis assay For induction of apoptosis, cells were cultured with 50 ngÆmL )1 recombinant human CD95L (AXXORA, Lo ¨ rrach, Germany) or 200 ngÆmL )1 antibody against human CD95 (CH-11) for the time periods described in the figure legends. Apoptosis was determined by annexin V ⁄ 7-AAD staining according to the manufacturer’s instructions (BD Pharmingen). Apoptotic cells were quantified on a FACS Calibur flow cytometer and analyzed using cellquest software (Becton Dickinson, Franklin Lakes, NJ, USA). CD95 receptor downregulation Cells were incubated with CD95L on ice for 30 min in the presence or absence of caspase inhibitors (Biozol, Eching, Germany) and ⁄ or BFA (Epicenter Technologies, Madison, WI, USA). Cells were then stimulated by subjecting them to a temperature of 37 °C for the time periods described in the figure legends. Stimulation-induced internalization was ter- minated by adding ice-cold 0.5% azide containing RPMI medium and placing the cells on ice. Nonspecific interactions were blocked by preincubation with isotype-matched IgG 1 , and cell surface CD95 was stained with a mAb against human CD95 (DX2; BD Pharmingen) on ice. Cells were then fixed with 2% paraformaldehyde for analysis by flow cytometry. Alternatively, cells were stimulated with Alexa 647-labeled CH-11 at 37 °C and analyzed by fluorescence microscopy. Immunofluorescence microscopy Cells were fixed with 4% PFA and permeabilized with either 0.2% Triton X-100 for detection of FADD and CD95 or 0.2% Triton X-100 and 0.2% sodium citrate for EEA-1 detection. Immunofluorescence labeling was performed according to standard procedures, using specific mAbs against FADD [clone 1 (Becton Dickinson) or clone A66-2 (BD Pharmingen)], CD95 (CH-11; MBL, Woburn, MA, USA), and EEA-1 (clone 14; Becton Dickinson). All primary antibodies were directly labeled with Alexa 488, Alexa 546, or Alexa 647, or biotinylated according to the manufac- turer’s recommendations (Invitrogen). To block nonspecific staining, cells were preincubated with isotype-matched mouse IgG 1 or IgG 2a prior to staining with specific antibodies. Alexa 546-conjugated or Alexa 647-conjugated streptavidin and DAPI were purchased from Invitrogen. Images were obtained using a deconvolution microscope (Applied Precision, Issaquah, WA, USA) equipped with inverted fluorescence optics and a CCD camera. Deconvo- luted images from 60 z-serial sections were subsequently generated by softworx software (Applied Precision). Quantitative analysis of images to determine relative pixel values of fluorescence intensity was performed using iVision software (Biovision Technologies, Exton, PA, USA). Immunoprecipitation and western blotting Cells were stimulated for the indicated times with 500 ngÆmL )1 CH-11 (MBL) at 37 °C, and lysed with buffer containing 50 mm Tris ⁄ HCl (pH 7.4), 150 mm NaCl, 1% NP-40, 1 mm Na 3 Vo 4 ,10mm NaF, and complete protease inhibitor cocktail (Boehringer, Mannheim, Germany). To isolate the CD95-associated signaling complex, cell lysates were immunoprecipitated using specific antibody against the DD of human CD95 (G254-274; BD Pharmingen) and pro- tein A ⁄ G plus agarose (Thermo Fisher Scientific, Rockford, IL, USA). Immunoprecipitates were subjected to western blot analysis using antibodies against human CD95 (C20; Santa Cruz Biotechnology, Heidelberg, Germany), FADD [clone 1F7 (Millipore, Schwalbach, Germany) or H-181 (Santa Cruz Biotechnology)], caspase-8 (C15; Alexis Bio- chemicals, Farmingdale, NY, USA), Laminin A ⁄ C (clone 14; Millipore), and GDI-Rho (clone 16; BD Pharmingen). Membrane fractionation A20 cells expressing either full-length human CD95 or mutant human CD95(Y291F) were incubated with 500 ngÆmL )1 antibody against human CD95 (CH11; MBL) for the indicated times at 37 °C. Stimulation was termi- nated by adding ice-cold homogenization buffer (BioVision, Mountain View, CA, USA) containing 0.5% azide. Nuclear and cytoplasmic membrane fractions were subsequently separated using a nuclear ⁄ cytosol protein extraction kit (BioVision), according to the manufacturer’s instructions. References 1 Chinnaiyan AM, O’Rourke K, Tewari M & Dixit VM (1995) FADD, a novel death domain-containing protein, interacts with the death domain of Fas and initiates apoptosis. Cell 81, 505–512. FADD trafficking and CD95 signaling N. Fo ¨ ger et al. 4264 FEBS Journal 276 (2009) 4256–4265 ª 2009 The Authors Journal compilation ª 2009 FEBS 2 Ashkenazi A & Dixit VM (1998) Death receptors: signaling and modulation. Science 281, 1305–1308. 3 Tibbetts MD, Zheng L & Lenardo MJ (2003) The death effector domain protein family: regulators of cellular homeostasis. Nat Immun 4, 404–409. 4 Chinnaiyan AM, Tepper CG, Seldin MF, O’Rourke K, Kischkel FC, Hellbardt S, Krammer PH, Peter ME & Dixit VM (1996) FADD ⁄ MORT1 is a common media- tor of CD95 (Fas ⁄ APO-1) and tumor necrosis factor receptor-induced apoptosis. J Biol Chem 271, 4961– 4965. 5 Zhang J, Cado D, Chen A, Kabra NH & Winoto A (1998) Fas-mediated apoptosis and activation-induced T-cell proliferation are defective in mice lacking FADD ⁄ Mort1. Nature 392, 296–300. 6 Newton K, Harris AW, Bath ML, Smith KG & Strasser A (1998) A dominant interfering mutant of FADD ⁄ MORT1 enhances deletion of autoreactive thymocytes and inhibits proliferation of mature T lymphocytes. EMBO J 17, 706–718. 7 Walsh CM, Wen BG, Chinnaiyan AM, O’Rourke K, Dixit VM & Hedrick SM (1998) A role for FADD in T cell activation and development. Immunity 8, 439–449. 8 Lee KH, Feig C, Tchikov V, Schickel R, Hallas C, Schu ¨ tze S, Peter ME & Chan AC (2006) The role of receptor internalization in CD95 signaling. EMBO J 25, 1009–1023. 9Go ´ mez-Angelats M & Cidlowski J (2003) Molecular evidence for the nuclear localization of FADD. Cell Death Differ 10, 791–797. 10 Screaton RA, Kiessling S, Sansom OJ, Millar CB, Maddison K, Bird A, Clarke AR & Frisch SM (2003) Fas-associated death domain protein interacts with methyl-CpG binding domain protein 4: a potential link between genome surveillance and apoptosis. Proc Natl Acad Sci USA 100, 5211–5216. 11 Bhojani MS, Chen G, Ross BD, Beer DG & Rehemtul- la A (2005) Nuclear localized phosphorylated FADD induces cell proliferation and is associated with aggres- sive lung cancer. Cell Cycle 4, 1478–1481. 12 Chen G, Bhojani MS, Heaford AC, Chang DC, Laxman B, Thomas DG, Griffin LB, Yu J, Coppola JM, Giordano TJ et al. (2005) Phosphorylated FADD induces NF-kappaB, perturbs cell cycle, and is associ- ated with poor outcome in lung adenocarcinomas. Proc Natl Acad Sci USA 102, 12507–12512. 13 Stolz LE, Huynh CV, Thorner J & York JD (1998) Identification and characterization of an essential family of inositol polyphosphate 5-phosphatases (INP51, INP52 and INP53 gene products) in the yeast Saccharo- myces cerevisiae. Genetics 148, 1715–1729. 14 Itoh T, Koshiba S, Kigawa T, Kikuchi A, Yokoyama S & Takenawa T (2001) Role of the ENTH domain in phosphatidylinositol-4,5-bisphosphate binding and endocytosis. Science 291, 1047–1051. 15 Itoh T & Takenawa T (2004) Regulation of endocytosis by phosphatidylinositol 4,5-bisphosphate and ENTH proteins. Curr Top Microbiol Immunol 282, 31–47. 16 Jost M, Simpson F, Kavran JM, Lemmon MA & Schmid SL (1998) Phosphatidylinositol-4,5-bisphosphate is required for endocytic coated vesicle formation. Curr Biol 8, 1399–1402. 17 Algeciras-Schimnich A & Peter ME (2003) Actin depen- dent CD95 internalization is specific for type I cells. FEBS Lett 546, 185–188. 18 Boldin MP, Goncharov TM, Goltsev YV & Wallach D (1996) Involvement of MACH, a novel MORT1 ⁄ FADD-interacting protease, in Fas ⁄ APO-1- and TNF receptor-induced cell death. Cell 85, 803–815. 19 Kischkel FC, Hellbardt S, Behrmann I, Germer M, Pawlita M, Krammer PH & Peter ME (1995) Cytotox- icity-dependent APO-1 (Fas ⁄ CD95)-associated proteins form a death-inducing signaling complex (DISC) with the receptor. EMBO J 14, 5579–5588. 20 Scaffidi C, Volkland J, Blomberg I, Hoffmann I, Kr amme r PH & Peter ME (2000) Phosphorylation of FADD ⁄ MORT1 at serine 194 and association with a 70-kDa cell cycle-regulated protein kinase. J Immunol 164, 1236–1242. 21 Alappat EC, Feig C, Boyerinas B, Volkland J, Samuels M, Murmann AE, Thorburn A, Kidd VJ, Slaughter CA, Osborn SL et al. (2005) Phosphorylation of FADD at serine 194 by CKIalpha regulates its nonapoptotic activities. Mol Cell 19, 321–332. 22 Alappat EC, Volkland J & Peter ME (2003) Cell cycle effects by C-FADD depend on its C-terminal phosphor- ylation site. J Biol Chem 278, 41585–41588. 23 Faustino RS, Cheung P, Richard MN, Dibrov E, Kneesch AL, Deniset JF, Chahine MN, Lee K, Black- wood D & Pierce GN (2008) Ceramide regulation of nuclear protein import. J Lipid Res 49, 654–662. 24 Grullich C, Sullards MC, Fuks Z, Merrill AH & Kolesnick R (2000) CD95(Fas ⁄ APO-1) signals ceramide generation independent of the effector stage of apopto- sis. J Biol Chem 275, 8650–8656. 25 Grassme H, Jekle A, Riehle A, Schwarz H, Berger J, Sandhoff K, Kolesnick R & Gulbins E (2001) CD95 signaling via ceramide-rich membrane rafts. J Biol Chem 276, 20589–20596. 26 Wagenknecht B, Roth W, Gulbins E, Wolburg H & Weller M (2001) C2-ceramide signaling in glioma cells: synergistic enhancement of CD95-mediated, caspase- dependent apoptosis. Cell Death Differ 8, 595–602. 27 Kabra NH, Kang C, Hsing LC, Zhang J & Winoto A (2001) T cell-specific FADD-deficient mice: FADD is required for early T cell development. Proc Natl Acad Sci USA 98, 6307–6312. 28 Bender LM, Morgan MJ, Thomas LR, Liu ZG & Thorburn A (2005) The adaptor protein TRADD acti- vates distinct mechanisms of apoptosis from the nucleus and the cytoplasm. Cell Death Differ 12, 473–481. N. Fo ¨ ger et al. FADD trafficking and CD95 signaling FEBS Journal 276 (2009) 4256–4265 ª 2009 The Authors Journal compilation ª 2009 FEBS 4265 . Subcellular compartmentalization of FADD as a new level of regulation in death receptor signaling Niko Fo ¨ ger 1 , Silvia Bulfone-Paus 1 , Andrew. independent of CD95 internalization, formation of the death- inducing signaling complex, and caspase-8 activation. In contrast to nuclear export of FADD, its