Phosphorylation of FADD/ MORT1 at Serine 194 and Association with a 70-kDa Cell Cycle-Regulated Protein Kinase

Cell lines

The Burkitt lymphoma BJAB was cultured in RPMI 1640, 10% FCS, 50 mM HEPES, and 50 μg/ml gentamycin. The embryonic kidney cell line 293T was cultured in DMEM, 10% FCS, 100 mM HEPES, and 50 μg/ml gentamycin.

Abs and reagents

For generation of anti-FADD mAbs, BALB/c mice were immunized four times by injection of 300 μg of purified GST-FADD. Spleen cells from immunized animals were fused with the Ag8 myeloma. 2 wk after fusion culture supernatants from wells positive for growth were tested in an ELISA with His-FADD as coated Ag. Hybridomas that produced anti-FADD mAbs were cloned several times by limited dilution yielding subclones positive for the desired Ab. The Ab used in this study was 1C4 (IgG1). The other anti-FADD mAb (IgG1) used was purchased from Transduction Laboratories (Lexington, Kentucky). The mouse mAb anti-APO-1 (IgG3, k, anti-CD95) recognizes an epitope on the extracellular part of APO-1 (21). The HRP-conjugated goat anti-mouse IgG1 Ab was obtained from Southern Biotechnology Associates (Birmingham, AL). Calf intestine phosphatase was purchased from New England Biolabs (Boston, MA).

Fusion proteins, plasmids, and generation of FADD mutant constructs

Using standard PCR and cloning techniques the following fusion proteins were generated: 6-His-tagged FADD, GST-FADD, GST-N-FADD (amino acids 1–117), and GST-C-FADD (amino acids 80–208). Fusion proteins were expressed and purified as previously described (3). The pcDNA3 plasmid containing FADD or C-FADD cDNA were a kind gift of Dr. Dixit (Genentech, San Francisco, CA). C-terminal deletion mutants of FADD and C-FADD were generated by PCR using the primer TATATCTCGAGTTACCTGTTCTGGAGGT CACGG and a primer binding in the 5′ region of the vector and subcloned into pcDNA3 (Invitrogen, San Diego, CA). The FADD S194A mutant was generated using the primer GTGGGGCCATGGCCCCGATGTCATGG and the corresponding complementary primer applied in a QuikChange Mutagenesis Kit (Stratagene, La Jolla, CA).

Metabolic labeling of cells

³⁵S metabolic labeling of BJAB cell was performed essentially as described before (6). For in vivo labeling with [³²P]orthophosphate, 1.5 × 10⁷ Jurkat or BJAB cells transfected with empty vector or C-FADD were washed three times with phosphate-free medium (DMEM; Life Technologies, Rockville, MD), resuspended in 3 ml of phosphate-free medium supplemented with 10% dialyzed FCS and 0.5 mCi [³²P]orthophosphate and incubated for 4 h at 37°C.

Cell cycle arrest

BJAB cells were treated with 4 mM hydroxyurea (Sigma, St. Louis, MO) or 100 ng/ml nocodazole (Sigma) for 20 h. Cells were washed with PBS once and harvested for analysis by flow cytometry and protein analysis as described below.

Immunoprecipitation and Western blotting

Immunoprecipitation of the CD95 DISC was essentially done as described elsewhere (22, 23). For immunoprecipitation of FADD, ³⁵S-labeled cells were lysed in lysis buffer (30 mM Tris/HCl, pH 7.4, 150 mM NaCl, 1 mM PMSF, and small peptide inhibitors (3), 1% Triton X-100 (Serva, Bay City, MI) and 10% glycerol) and incubated for 1 h at 4°C with 10 μg of 1C4 or anti-FADD precoupled to anti-mouse IgG1 beads (Sigma). Immunoprecipitates were analyzed by two-dimensional IEF-SDS-PAGE as described previously (3) followed by autoradiography. For Western blot detection of FADD, cellular lysates equivalent to 10⁶ cells were separated by one or two-dimensional gel electrophoresis and blotted onto a nitrocellulose membrane (Hybond C; Amersham, Arlington Heights, IL). Blots were blocked with 2% BSA (Boehringer Mannheim, Mannheim, Germany) in TPBS (0.05% Tween 20 in PBS) for 1 h at room temperature. After washing, blots were incubated with anti-FADD mAb (1 μg/ml in TPBS) for 16 h at 4°C. After incubation with secondary Ab (1/20,000 in TPBS), blots were developed using enhanced chemiluminescence (Amersham) following the manufacturers instructions.

Kinase assays

For the in vitro kinase assay, immunoprecipitated AU1-tagged FADD from transiently transfected 293T cells or 10 μg of GST fusion proteins precoupled to reduced glutathione (GSH) beads were incubated with cellular lysates from 10⁷ BJAB cells for 2 h at 4°C. After extensive washing in lysis buffer, beads were resuspended in kinase buffer (75 mM Tris, pH 7.5, 1% Nonidet P-40, 10 mM MnCl₂, 10 mM MgCl₂, 10% glycerin, and 100 mM NaCl) containing 0.8 mBq [γ-³²P]ATP and incubated for 20 min at 25°C. Beads were washed once in lysis buffer, and proteins were separated by 12% SDS-PAGE and analyzed by autoradiography and Coomassie blue staining. In-gel kinase assays were performed as described (24). Briefly, cellular lysate from 3 × 10⁷ cells was sequentially incubated with the indicated GST fusion proteins coupled to GSH beads. Beads were washed extensively, and proteins were separated by SDS-PAGE containing 60 μg/ml copolymerized His-FADD or no substrate.

Transfection of cells

BJAB cells were transfected by electroporation (960 μF, 200 V) using a Gene Pulser (Bio-Rad, Richmond, CA), and single clones were selected in 96-well plates using 4 mg/ml G418 (Sigma) containing medium. Positive clones were identified by Western blot analysis. Transient transfection of 293T cells was done using the calcium phosphate precipitation method.

Flow cytometry

A total of 1 × 10⁶ control or hydroxyurea- or nocodazole-treated cells were incubated with propidium iodide (25), and DNA content was measured by FACScan analysis (Becton Dickinson, San Diego, CA). Ten thousand events were analyzed for each sample.

ELISA

A total of 100 μl of FADD C-terminal peptides as described in legend to Fig. 4⇓A (10 μg/ml, diluted in 0.2 M bicarbonate buffer, pH 9.6) were immobilized on microtiter plates at 4°C for 16 h. Plates were then blocked with RPMI 1640 supplemented with 10% FCS for 1 h at 25°C. After washing three times with TPBS, 10 μg/ml of the anti-FADD mAb 1C4 was added to the first row and further diluted 1:1. After incubation for 1 h at 25°C, the plates were washed three times and incubated with HRP-coupled anti-IgG1 secondary Ab (1:5000 in TPBS) for 1 h at 25°C. Subsequently, plates were washed three times and developed with o-phenylenediamine (Sigma; 1 mg/ml, diluted in 0.1 M sodium citrate buffer, pH 5) and 1 μl/ml perhydrol (Merck, West Point, PA). Reaction was stopped by adding 3 N sulfuric acid.

Cytotoxicity-dependent APO-1 associated protein (CAP)-1 and -2 represent differentially phosphorylated FADD

We have recently shown that triggering of CD95 leads to the recruitment of a set of signaling molecules (designated CAP1–6) forming the DISC (3, 10). These signaling molecules can be coimmunoprecipitated with activated CD95 and visualized on a two-dimensional gel (Fig. 1⇓A, left panel). CAP1 and CAP2 were both identified as FADD (3), whereby CAP2 represents a serine phosphorylated form. This was also confirmed by immunoprecipitation with an anti-FADD mAb from in vivo-phosphorylated cells (Fig. 1⇓B, left panel). The specific phosphorylation was further confirmed by treatment of immunoprecipitated CAP2 with calf intestinal phosphatase. This treatment resulted in a conversion of CAP2 into CAP1 in a two-dimensional Western blot analysis (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 1.

CAP1 and CAP2 represent FADD that is differentially phosphorylated at the C terminus. A, BJAB cells stably transfected with empty vector (control) (left panel) or a C-FADD vector (right panel) were labeled with [³⁵S]methionine and stimulated with 2 μg/ml anti-CD95 for 10 min at 37°C. Cells were lysed, and CD95 DISC was immunoprecipitated using protein A-Sepharose. The immunoprecipitate was analyzed by two-dimensional isoelectric focusing (IEF) SDS-PAGE and autoradiography. Migration positions of the FADD forms CAP1 and CAP2, the caspase-8 forms CAP3–6, and two FADD forms lacking the first 79 aa (ΔCAP1 and ΔCAP2) are indicated. The stippled box in the left panel labels the area of the gel shown in the right panel and in B. B, Cells as described in A were labeled with [³²P]orthophosphate, and FADD was immunoprecipitated using an anti-FADD mAb coupled to protein A-Sepharose. Immunoprecipitates were analyzed as in A.

C-FADD is a deletion mutant of FADD, lacking the N-terminal DED (see Fig. 3⇓). We have previously shown that in stably transfected cells C-FADD is recruited to the activated CD95 receptor, competing for the binding of endogenous FADD to the DISC and thereby inhibiting apoptosis (6). To determine whether C-FADD would also be specifically phosphorylated, we tested BJAB cells stably expressing C-FADD on two-dimensional gels (Fig. 1⇑, right panels). In metabolically ³⁵S-labeled cells, DISC-associated C-FADD also appeared predominantly as two distinct spots (ΔCAP1/2), indicating that it is phosphorylated in a similar way as the wild-type molecule (Fig. 1⇑B, right panel). To confirm this assumption, we metabolically labeled cells stably expressing C-FADD with [³²P]orthophosphate to specifically visualize phosphorylated proteins. In an anti-FADD immunoprecipitation from these in vivo-labeled cells, only ΔCAP2 could be visualized (Fig. 1⇑B, right panel), identifying the ΔCAP2 spot as phosphorylated C-FADD. Analysis of the immunoprecipitated DISC from these cells also showed that both DISC-associated CAP2 and ΔCAP2 were ³²P-labeled (data not shown), suggesting that both full-length FADD and C-FADD are phosphorylated in the same way.

Identification of a phosphorylation site-specific anti-FADD mAb

The results so far indicated that a kinase phosphorylates FADD at its C-terminal half, affecting the migration of the phosphorylated molecule on two-dimensional gels. However, only part of FADD is phosphorylated as nonphosphorylated FADD (CAP1) or C-FADD (ΔCAP1) could also be detected in all cells. To test if only the C-terminal half of FADD would be a substrate for phosphorylation, we generated fusion proteins of GST with either FADD, the N-terminal (GST-N-FADD), or the C-terminal half of FADD (GST-C-FADD). Incubation of cellular lysates with these fusion proteins led to the association of a kinase that phosphorylated GST-FADD and GST-C-FADD in an in vitro kinase assay (Fig. 2⇓A, lane 5 and 6). GST-N-FADD was not phosphorylated in this assay (Fig. 2⇓A, lane 6). The data demonstrated that the C-terminal half of FADD was phosphorylated by a kinase that associated with FADD and C-FADD in vitro.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 2.

Characterization of a phosphorylation site-specific anti-FADD mAb. A, Indicated GST fusion proteins were incubated with cellular lysates, washed, and an in vitro kinase assay was performed. Samples were analyzed by SDS-PAGE, Coomassie blue staining (lanes 1–3), and autoradiography (lanes 4–6). Lanes 7–9, A total of 50 ng of the indicated GST fusion proteins were analyzed by Western blotting using 1C4 anti-FADD mAb. As a control, 5 μg of the fusion proteins analyzed by SDS-PAGE and Coomassie blue staining established equal protein loading (data not shown). B, FADD was immunoprecipitated from BJAB cells labeled with [³⁵S]methionine/cysteine using commercial anti-FADD (upper left) or 1C4 anti-FADD (upper right) mAb and analyzed by two-dimensional IEF-SDS-PAGE and autoradiography. In addition, cellular lysates equivalent to 10⁶ BJAB cells were separated by two-dimensional IEF-SDS-PAGE and analyzed by Western blotting using commercial anti-FADD (lower left) or 1C4 anti-FADD (lower right) mAb. Arrow heads indicate the migration position of CAP1 and CAP2.

To further characterize the phosphorylation site of FADD we generated a panel of mAbs against this molecule. The Ab 1C4 recognized only C-FADD but not N-FADD in a Western blot analysis (Fig. 2⇑A, lanes 8 and 9). Interestingly, this Ab only detected unphosphorylated FADD (CAP1) when used in an immunoprecipitation experiment (Fig. 2⇑B, upper right panel), whereas a control anti-FADD mAb detected both, nonphosphorylated (CAP1) and phosphorylated (CAP2) FADD (Fig. 2⇑B, upper left panel). The same result was seen when both Abs were used in a two-dimensional Western blot analysis of FADD in cellular lysates (Fig. 2⇑B, lower panels). This suggests that the 1C4 mAb recognizes FADD only in its unphosphorylated form. Experiments so far have localized the phosphorylation site of FADD to its C-terminal half. Therefore, we conclude that the 1C4 mAb reacts with an epitope within the C terminus of unphosphorylated FADD.

Phosphorylation of FADD occurs at a C-terminal serine cluster

We have previously shown that FADD is phosphorylated at serine residues (3). FADD contains two serine clusters, one located at the N and one at the C terminus of the molecule (Fig. 3⇓A). To test if the C-terminal serine cluster indeed contains the phosphorylation site, we generated deletion mutants of FADD and C-FADD lacking 19 aa at the C terminus including all seven C-terminal serine residues (designated FADDΔC and C-FADDΔC) (Fig. 3⇓A). These molecules were expressed in 293T cells and analyzed by Western blotting using an anti-FADD mAb (Fig. 3⇓B, lanes 1–4). In this assay, phosphorylated and nonphosphorylated FADD can be distinguished by their different mobilities on the gel by giving rise to two distinct bands for each FADD and C-FADD (Fig. 3⇓B, lanes 1 and 2). The mutants lacking the C-terminal serine cluster appeared only as single bands, suggesting that they were no longer phosphorylated in vivo (Fig. 3⇓B, lanes 3 and 4). To confirm this, we also used the 1C4 mAb for Western blot detection as this Ab recognized the phosphorylation site of FADD (Fig. 3⇓B, lanes 5–8). Both C-terminal deletion mutants were not detected by this Ab, whereas nonphosphorylated species of FADD and C-FADD could be readily visualized. These experiments identify the C-terminal serine cluster of FADD as the site of phosphorylation and establish 1C4 as a phosphorylation site-specific Ab.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 3.

FADD is phosphorylated at a C-terminal serine cluster. A, Overview of all generated FADD deletion mutants used in this study. Numbers indicate amino acid positions. Location of serine residues are indicated by asterisks. B, Indicated FADD constructs were expressed in 293T cells, and lysates were analyzed by Western blotting using commercial anti-FADD (lanes 1–4) or 1C4 anti-FADD (lanes 5–8) mAb. Migration positions of the different FADD mutants are indicated.

Identification of the FADD phosphorylation site

Because CAP1 and CAP2 were also found in the DISC of murine L929 cells (10), we assumed that the phosphorylation site was conserved between mouse and man. Of the seven serine residues located at the C terminus of FADD, five are conserved between the two species (Fig. 4⇓A). To map the phosphorylation site exactly, we generated five peptides in which these five serine residues were individually replaced by glutamic acid to mimic phosphorylated serine. Then, 1C4 was tested in an ELISA experiment for reactivity for these peptides. Three of the tested peptides (Pep S200E, S203E, S205E) were efficiently recognized by 1C4 (Fig. 4⇓B), and peptide Pep S190E gave a weak signal with 1C4. However, Pep S194E was not recognized at all by the 1C4 Ab, suggesting that phosphorylation of serine 194 prevents binding of 1C4 (Fig. 4⇓A). To confirm this conclusion we generated a C-FADD molecule in which we replaced serine 194 by alanine. A Western blot analysis of cellular lysates of 293T cells transiently transfected with either C-FADD or C-FADD S194A demonstrated that C-FADD S194A was not recognized by 1C4 anymore (Fig. 4⇓C), confirming the results of the peptide ELISA. The mutated C-FADD did also not serve as a kinase substrate in an in vitro kinase assay (Fig. 4⇓D), indicating that serine 194 is the only site in the molecule that can be phosphorylated. To finally prove that FADD lacking just one hydroxyl group (after an serine-alanine replacement) would not be phosphorylated in vivo anymore, a BJAB cell line was generated stably expressing C-FADD S194A (Fig. 4⇓E). A Western blot analysis with the commercial anti-FADD Ab that recognizes both unphosphorylated and phosphorylated FADD showed that in this transfectant C-FADD S194A only migrated at the position of unphosphorylated C-FADD (Fig. 4⇓E, lane 2). Taken together, the data demonstrate that in vivo FADD is specifically phosphorylated at only one serine residue at position 194.

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 4.

In vivo and in vitro FADD is phosphorylated at serine 194. A, Alignment of the 21 C-terminal amino acids of human and murine FADD with the peptides in which the conserved serine residues were individually replaced by glutamic acid. B, ELISA on the mutant peptides shown in A with the anti-FADD mAb 1C4. C, Western blot analysis of whole-cell lysates from stable BJAB-C-FADD transfectants (Lys), 293T cells transiently expressing C-FADD (WT), C-FADD in which serine 194 was replaced by alanine (S194A), or control vector (C). All bands correspond to wild-type or mutated C-FADD. The upper band in this blot comigrates with CAP1; the lower band represents a degradation product. D, In vitro kinase assay on C-FADD (WT) or C-FADD S194A (S194A) immunoprecipitated from transiently transfected 293T cells. A Western blot using the commercial anti-FADD Ab is shown below. E, Western blot of whole-cell lysates from 10⁶ BJAB cells stably expressing either C-FADD (WT) or C-FADD S194A (S194A).

Phosphorylation of FADD by an associated 70-kDa kinase

To characterize the kinase that phosphorylates FADD, we performed an in-gel kinase assay. In this assay, proteins associated with the GST-FADD constructs after incubation with cellular lysates from BJAB cells were separated by SDS-PAGE followed by a denaturing step. Subsequently, proteins within the gel were slowly renatured, and kinases were identified by incubating the gel in a kinase buffer containing [γ-³²P]ATP. To control for specificity, we used a gel without (Fig. 5⇓A, lanes 1–4) or with (Fig. 5⇓A, lanes 5–8) FADD as copolymerized substrate. Using this assay, we identified a kinase of ∼70 kDa that associated only with FADD and C-FADD and that was specific for the phosphorylation of the copolymerized substrate FADD (Fig. 5⇓A, lanes 7 and 8). A phosphorylated band of the same size was also found to associate with endogenous FADD in in vivo-phosphorylated BJAB cells (Fig. 5⇓B).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 5.

FADD is phosphorylated by an associated 70-kDa kinase. A, Cellular lysates were sequentially incubated with GST, GST-N-FADD, GST-C-FADD, and GST-FADD beads. After washing, associated proteins were separated by SDS-PAGE without (lanes 1–4) or with (lanes 5–8) His-FADD copolymerized into the gel matrix as a substrate and analyzed by an in-gel kinase assay. Arrowhead indicates the migration position of a 70-kDa FADD-specific kinase. The band at 90 kDa in all lanes represents unspecific background, which was also seen after precipitation with GSH beads alone (data not shown). The 200-kDa band seen only in the precipitates using GST or GST-N-FADD represents a background band that was removed during the sequential precipitation with beads coupled to various GST fusion proteins. B, FADD was immunoprecipitated from lysates of in vivo-phosphorylated Jurkat cells using the commercial anti-FADD mAb (lane 2). As a control, caspase-8 was immunoprecipitated with an anti-caspase-8 Ab (lane 1).

Phosphorylation of FADD depends on the cell cycle

As FADD plays an essential role in the signal transduction of CD95, we investigated the role of the phosphorylation of FADD in this function of the molecule. It has been shown that overexpression of FADD results in apoptosis induction, presumably by nonphysiological aggregation of the molecule (26). We could not detect any significant difference in apoptosis induction by transient overexpression of FADD or FADDΔC in 293T cells (data not shown). As this system may not reflect the physiological function of FADD, we generated cells stably expressing C-FADD and C-FADDΔC and analyzed them for their sensitivity to CD95-mediated apoptosis. Consistent with our previous data (6), C-FADD-expressing cells were resistant to CD95-induced apoptosis. Cells expressing similar amounts of C-FADDΔC also showed the same degree of apoptosis resistance. Similar to C-FADD, C-FADDΔC was also recruited to the activated CD95 receptor competing for binding of endogenous FADD (data not shown). These data suggest that phosphorylation of FADD does not play a major role in the induction of apoptosis by CD95.

In addition to its role in death receptor signaling, FADD has been implicated in T cell proliferation (12, 16, 17, 18). The mechanism of this action remains to be defined. However, it was shown that this activity of FADD was not detectable in constantly cycling tumor cells (16). These published data would indicate that in addition to its function in death receptor-mediated apoptosis, FADD is important for cell cycle entry of resting T cells. Furthermore, it was then shown that this inhibition requires functional p53 because in transgenic mice expressing FADD-DN in the T cell compartment the resistance of T cells to mitotic expansion was prevented in the p53 knockout background (17). These data suggest that one activity of FADD may be linked to regulation of the cell cycle.

Therefore, we tested whether phosphorylation of FADD would be subject to cell cycle regulation. BJAB vector transfectants (Fig. 6⇓A, lanes 1–3) and BJAB stably expressing C-FADD (Fig. 6⇓A, lanes 4–6) were arrested at G₁/S or G₂/M transitions of the cell cycle by treating the cells for 24 h with hydroxyurea or nocodazole, respectively. FACS analysis confirmed that both treatments resulted in arrested cells (Fig. 6⇓A, upper panels). Analysis of the FADD phosphorylation status showed that cells arrested at G₁/S with hydroxyurea predominantly contained unphosphorylated FADD and C-FADD, respectively. In contrast, in cells arrested at G₂/M, both FADD and C-FADD were quantitatively phosphorylated. The same result was obtained when cell were arrested at G₁/S with a double thymidine block or at G₂/M with taxol (data not shown).

• Download figure
• Open in new tab
• Download powerpoint

FIGURE 6.

The 70-kDa FADD kinase is cell cycle regulated. A, BJAB (lanes 1–3) or BJAB stably expressing C-FADD (lanes 4–6) were left untreated (/) or treated with hydroxyurea (HU) or nocodazole (N) for 24 h before analysis in a Western blot experiment using the commercial anti-FADD mAb on whole-cell lysates. B, In vitro kinase assay with C-FADD immunoprecipitated from transiently transfected 293T cells after incubation of the immunoprecipitate with detergent-solubilized extracts from untreated (/), hydroxyurea-arrested (HU), or nocodazole-arrested (N) BJAB cells. In addition to an associated phosphoprotein of 70 kDa, a phosphorylated band at 150 kDa was also detected.

To test whether the 70-kDa kinase that associates with FADD would also preferentially be active in G₂/M-arrested cells, we used FADD immunoprecipitated from transiently transfected 293T cells in an in vitro kinase assay after incubation with cellular lysates from arrested BJAB cells (Fig. 6⇑B). Again, in addition to phosphorylated FADD, an associated phosphorylated protein band of 70 kDa was detected. This band was almost absent in the cells arrested with hydroxyurea (Fig. 6⇑B, lane 2).

We then tested whether classical members of the cdk/cdc family would accept FADD as a substrate. We used recombinant cdc2/cyclin A, cdk2/cyclin A, cdc2/cyclin B1, and cdk2/cyclin E to phosphorylate FADD in in vitro kinase assays. However, none of these kinases that are active at the G₂/M boundary could phosphorylate FADD (data not shown). This is consistent with the fact that the putative FADD kinase identified in an in-gel kinase assay that associated with FADD and accepted FADD as an in vitro substrate was 70 kDa in size, whereas all of the cdc/cdk kinases are in the range of 33–35 kDa. Treatment of cells with wortmannin did not change the CAP1-to-CAP2 ratio, making it unlikely that phosphatidylinositol 3′-kinase is involved in the regulation of FADD phosphorylation. In addition, we can rule out protein kinase C because treatment with PMA did not increase the amount of phosphorylated FADD. Furthermore, recombinant casein kinase II also failed to phosphorylate FADD in vitro (data not shown). In summary, our data suggest that during progression during the cell cycle FADD undergoes specific phosphorylation somewhere between G₂ and M phase at serine 194 by an unknown cell cycle-regulated kinase of 70 kDa.