Báo cáo khoa học: Subcellular compartmentalization of FADD as a new level of regulation in death receptor signaling pdf

This induc-ible nuclear–cytoplasmic translocation of FADD is independent of CD95 internalization, formation of the death-inducing signaling complex, and caspase-8 activation.. Results Nu

Subcellular compartmentalization of FADD as a new level

of regulation in death receptor signaling

Niko Fo¨ger1, Silvia Bulfone-Paus1, Andrew C Chan2and Kyeong-Hee Lee1

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 activaforma-tion 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)P2,phosphatidylinositol 4,5-bisphosphate.

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 min

FADD

B

FADD / DAPI

0 min

A

FADD / DAPI

FADD

Fluorescence intensity

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.

Expression of a plasma membrane-localized

phosphatidylinositol 4,5-bisphosphate

[PtdIns(4,5)P2]-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)P2 to

phosphatidylinositol 4-phosphate [13] Cellular levels of

PtdIns(4,5)P2are 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)P2 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

10 20 30 40 50 60 70 80 90 100

A

100 101 102 103 104

100 101 102 103 104

FynC – GFP+

FynC – GFP – INP54+

Annexin V

B

C

GFP FADD DAPI

FADD

Fluorescence intensity

6 5

4

3

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.

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, colocalizastimula-tion 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

8

4

7 6

5

CD95L

0 min (GFP +)

30 min (GFP +)

30 min (GFP–)

4 3

2

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.

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

WB:

Cas-8

hCD95

CD95: 0’ 0’ 0’ 15’ 30’ 0’ 15’ 30’

1 2 3 4 5 6 7 8

WB:

FADD Laminin GDI-Rho

WT Y291F

Nuclear Cytoplasmic

D

1 2 3 4 5 6 7 8

A

Y291F

WT

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).

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)P2 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

0 10 20 30 40 50 60 70 80 90

No inhibitor z-IETD (Cas-8) z-VAD (general) z-DEVD (Cas-3 & Cas-7)

(min)

No inhibitor

z-IETD

z-VAD

z-DEVD

CD95

0 min

5 min

30 min

CD95

z-VAD

z-VAD + BFA

0 min

5 min

30 min

BFA

No inhibitor

0

20

40

60

80

100

120

No inhibitor BFA z-VAD z-VAD + BFA

(min)

FADD CD95 CD95 / FADD

CD95L (30 min) z-IETD BFA

CD95L (30 min) BFA

z-IETD (caspase-8 inhibitor)

No inhibitor

CD95L

A

0 min

2 min

30 min

60 min

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 l M 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 l M 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.

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 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 mm l-glutamine (RPMI standard medium) Cells were maintained in 5% CO2 at 37C 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)

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)1antibody 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 37C 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 IgG1,

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 37C 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 IgG1 or IgG2a 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)1CH-11 (MBL) at 37C, and lysed with buffer containing 50 mm Tris⁄ HCl (pH 7.4), 150 mm NaCl, 1% NP-40, 1 mm Na3Vo4, 10 mm 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)1antibody against human CD95 (CH11; MBL) for the indicated times at 37C 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

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

9 Go´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 phosphoryRehemtul-lated 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, Krammer

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