TNF Activates Syk Protein Tyrosine Kinase Leading to TNF-Induced MAPK Activation, NF-κB Activation, and Apoptosis¹

TNF, a proapoptotic and an inflammatory cytokine, is produced primarily by the macrophages. It mediates its effects through two distinct receptors, TNFR p60 (also called TNFR1 or p55) and TNFR p80 (also called TNFR2 or p75). TNFR p60 is expressed on all cell types, whereas p80 is expressed primarily on hemopoietic cells and endothelial cells. Most TNF signaling has been ascribed to the p60 receptor, whereas the role of the p80 receptor is controversial. It is thought that p60 receptor mediates apoptosis and p80 receptor mediates proliferation (1). Others have suggested that the p80 receptor mediates its signaling through the p60 receptor. That soluble TNF is a ligand for p60 receptor and that transmembrane TNF is a ligand for the p80 receptor has also been demonstrated (2, 3, 4).

Recent studies with genetic deletion of TNF receptors have shown that both receptors are needed for optimum signaling (5, 6). When TNF binds with the receptor, it activates early events (within minutes), which include activation of NF-κB, JNK, p38 MAPK, and p44/p42 MAPK (7, 8, 9), and late events (within days), which include induction of apoptosis (10). Sequential recruitment of TNFR-associated death domain protein (TRADD), Fas-associated death domain protein, and FLICE by the p60 receptor leads to apoptosis; recruitment of TRADD, receptor-interacting protein, and IκB kinase-β leads to NF-κB activation; and recruitment of TRADD, TRAF2, and MAPK kinase (MKK)7 leads to JNK activation (7, 11, 12, 13, 14, 15, 16, 17). The role of protein-tyrosine kinases in TNF signaling, however, is not well understood.

Proline-rich tyrosine kinase (Pyk2)³ (18, 19) and p53/Lyn p56 (20) have roles in TNF-induced neutrophil activation; Etk/Bmx in TNF-induced endothelial cell migration and angiogenesis; c-Src in TNF-induced NF-κB activation in the marrow macrophages (21, 22); and TNF in regulation of gap junctions in airway epithelial cells (23). Pyk2 has also been implicated in TNF-induced JNK activation (24). In neutrophils, TNF has been shown to promote the association of Pyk2 to spleen tyrosine kinase (Syk) (19) and activate Syk, one of the downstream targets of Pyk2 (25).

Syk is a 72-kDa nonreceptor protein kinase that serves as a key regulator of multiple biochemical signal transduction events and has high homology to ZAP70 protein-tyrosine kinase (26). This kinase is ubiquitously expressed in hemopoietic cells and regulates proliferation, differentiation, and apoptosis. In B cells, phospholipase Cγ2 and PI3K are key targets of Syk tyrosine phosphorylation (27). Phosphorylation of phospholipase Cγ2 by Syk results in downstream activation of p44/p42 MAPK and JNK kinases (28), whereas PI3K phosphorylation by Syk mediates Akt activity (29). Recent evidence indicates that Syk is also expressed in nonhemopoietic cells, including hepatocytes (30), colon cancer cells (31), and breast cancer cells (32).

Recently, we showed that Syk plays a critical role in H₂O₂-induced NF-κB activation (33). What role Syk plays in TNF-induced signaling, however, is not known. We investigated whether TNF can activate Syk in hemopoietic and nonhemopoietic cells and whether Syk activation is involved in TNF-induced NF-κB, JNK, p38 MAPK, p44/p42 MAPK, and apoptosis. Our results indicate that TNF rapidly and potently activates Syk and that the deletion of Syk down-regulates TNF-induced activation of JNK, p38 MAPK, p44/p42 MAPK, and NF-κB but enhances apoptosis.

Bacteria-derived human rTNF, purified to homogeneity with a specific activity of 5 × 10⁷ U/mg, was kindly provided by Genentech (South San Francisco, CA). Penicillin, streptomycin, IMDM, DMEM, RPMI 1640, FBS, and LipofectAMINE 2000 were obtained from InVitrogen (Carlsbad, CA). Ab against β-actin and piceatannol were purchased from Sigma-Aldrich (St. Louis, MO). Abs anti-phosphotyrosine (PY99), ZAP70, JNK1, IκBα, PARP, and annexin V staining kit were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). Abs against phospho-IκBα, phospho-p38 MAPK, p38 MAPK, phospho-p44/p42 MAPK, and p44/p42 MAPK were purchased from Cell Signaling (Beverly, MA). Anti-Syk and anti-phosphotyrosine (4G10) Abs were purchased from Neo Markers (Fremont, CA). Wild-type Syk-cDNA, as described previously (32), was kindly provided by Dr. S. C. Mueller (Georgetown University Medical School, Washington, DC). Anti-TNFR1 mAb (htr-9) was kindly provided by Dr. M. Brockhaus (F. Hoffmann-LaRoche, Basel, Switzerland). Anti-TNFR2 mAb (MR2-1) was kindly provided by Dr. W. A. Buurman (Maastricht University, Maastricht, The Netherlands). ,

The generation of siRNA-Syk plasmid as previously described (33) was kindly supplied by Dr. S. Singh (Imgenex, San Diego, CA). The sequence (bp 555–575) used was TCGAGCAGACATGGAACCTGCAGGGGAGTACTGCCCTGCAGGTTCCATGTCTGCTTTTT (binding sequences are in bold letters, stem loop sequences are in italics, and the SalI cloning overhang site is underlined).

Jurkat (human T cells), JCaM1 (Lck and Syk-deficient), U937, KBM-5 (human myeloid), and HeLa (human epithelial) cells were obtained from the American Type Culture Collection (Manassas, VA). The characterization of JCaM1 cells has been previously reported (34). JCaM1/lck (lck-reconstituted JCaM1 cells), were kindly supplied by Dr. A. Weiss (University of California, San Francisco, CA). Jurkat, JCaM1, JCaM1/lck, and U937 cells were cultured in RPMI 1640 with 10% FBS; KBM-5 cells were cultured in IMDM supplemented with 15% FBS; HeLa cells were cultured in DMEM supplemented with 10% FBS with 100 U/ml penicillin and 100 μg/ml streptomycin.

To determine the levels of protein expression in cytoplasm and nuclear extracts, we prepared each extract (35) from TNF-treated cells and fractionated each by SDS-PAGE. After electrophoresis, the proteins were electrotransferred to nitrocellulose membranes, blotted with each Ab, and detected by ECL regent (Amersham, Piscataway, NJ). The density of the bands was measured using a personal densitometer scan version 1.30 and Imagequant software version 3.3 (Molecular Dynamics, Sunnyvale, CA).

The JNK assay was performed as described previously (36). Briefly, cells were lysed for 30 min on ice in whole cell extraction buffer (20 mM HEPES (pH 7.9), 50 mM NaCl, 1% Nonidet P-40, 2 mM EDTA, 0.5 mM EGTA, 2 μg/ml aprotinin, 2 μg/ml leupeptin, 0.5 mM PMSF, and 2 mM sodium orthovanadate). Lysate containing 200 μg of proteins in extraction buffer was incubated with 1 μg/ml anti-JNK1 Ab overnight, followed by treatment with protein A/G-Sepharose beads. After a 2-h incubation, the beads were washed with lysis buffer and then assayed in kinase assay mixture containing 50 mM HEPES (pH 7.4), 20 mM MgCl₂, 2 mM DTT, 20 μCi [γ-³²P]ATP, 10 μM unlabeled ATP, and 2 μg of substrate GST-c-Jun(1–59). After incubation at 30°C for 30 min, the reaction was terminated by boiling with SDS sample buffer for 5 min. Finally, the protein was resolved on a 10% SDS-PAGE, the gel was dried, and the radioactive bands were visualized by PhosphorImager (Molecular Dynamics). To determine the total amounts of JNK1, 30 μg of the whole cell extract were resolved on 7.5% SDS-PAGE, electrotransferred to a nitrocellulose membrane, and then blotted with anti-JNK1 Ab.

To determine NF-κB activation, we performed EMSA as described (37). Briefly, nuclear extracts prepared from TNF-treated cells (2 × 10⁶/ml) were incubated with ³²P-end-labeled 45-mer double-stranded NF-κB oligonucleotides (10 μg of protein with 16 fmol of DNA) from the HIV long terminal repeat, 5′-TTGTTACAAGGGACTTTCCGCTGGGGACTTTCCAGGGAGGCGTGG-3′ (boldface indicates NF-κB binding sites) for 30 min at 37°C, and the DNA-protein complex formed was separated from free oligonucleotides on 6.6% native polyacrylamide gels. A double-stranded mutated oligonucleotides, 5′-TTGTTACAACTCACTTTCCGCTGCTCACTTTCCAGGGAGGCGTGG-3′, was used to examine the specificity of binding of NF-κB to the DNA. The dried gels were visualized, and radioactive bands were quantitated by a PhosphorImager using Imagequant software.

Cells were lysed for 30 min on ice in whole cell extraction buffer (20 mM HEPES (pH 7.9), 50 mM NaCl, 1% Nonidet P-40, 2 mM EDTA, 0.5 mM EGTA, 2 μg/ml aprotinin, 2 μg/ml leupeptin, 0.5 mM PMSF, and 2 mM sodium orthovanadate). Lysate containing 500 μg of proteins in extraction buffer were incubated with 1 μg/ml Abs overnight. The immunocomplex was precipitated using protein A/G-Sepharose beads for 1 h at 4°C. Beads were washed with extraction buffer and resuspended in SDS sample buffer, boiled for 5 min, and fractionated in SDS-PAGE.

To determine the role of Syk protein tyrosine kinase in TNF-induced NF-κB activation, 2 × 10⁵ cells were transfected either with siRNA-Syk or with Syk-Wt. Two micrograms plasmid in each case was diluted in 200 μl of DMEM (without serum and antibiotics) and then mixed with 200 μl of DMEM containing 4 μl of LipofectAMINE 2000. This mixture was added to the cells and incubated for 4 h. After incubation, cells were washed with RPMI 1640, repeated the transfection procedure, washed again, and then maintained for an additional 48 h before using for experiments. Transfection efficiency was found to be 50–60% as examined by β-galactosidase transfection (data not shown). After transfection, cells were found to be 80–90% viable based on trypan blue dye exclusion method. Dead cells were removed by centrifugation before each experiment.

To determine apoptosis, annexin V-positive cells were examined using an annexin V staining kit purchased from Santa Cruz Biotechnology. TNF-treated cells were washed in PBS, resuspended in 100 μl of binding buffer containing FITC-conjugated annexin V, and analyzed by a FACSCalibur flow cytometer (BD Biosciences, San Jose, CA).

TNF-induced apoptosis was also determined by TUNEL assay using In Situ Cell Death Detection reagent (Roche Applied Science, Indianapolis, IN). Briefly, 1 × 10⁵cells were treated with 1 nM TNF for 16 h at 37°C. Thereafter, cells were plated on a poly-l-lysine-coated glass slide by centrifugation using a Cytospin 4 (Thermoshendon, Pittsburg, PA), air-dried, fixed with 4% paraformaldehyde, and permeabilized with 0.1% Triton X-100 in 0.1% sodium citrate. After washing, cells were incubated with reaction mixture for 60 min at 37°C. Stained cells were mounted with mounting medium (Sigma-Aldrich) and analyzed under a fluorescence microscope (Labophot 2; Nikon, Melville, NY).

In the present study, we investigated the role of Syk in TNF-induced activation of JNK, p44/p42 MAPK and p38 MAPK, and NF-κB and in apoptosis. We used Syk-selective metabolic inhibitors, Syk-deficient cell lines, Syk-overexpressing cells, and Syk-down-regulated cells for this study.

To determine the effect of TNF on the activation of Syk protein tyrosine kinase in Jurkat cells, cells were treated with TNF for different times, whole cell extracts were immunoprecipitated with anti-Syk Ab, and then subjected to Western blot using anti-phosphotyrosine Ab. The results in Fig. 1,A shows that TNF induced activation of Syk in a time-dependent manner; the activation can be seen as early as 5 s, reached maximum at 15 s, and decreased thereafter. TNF had no effect on the expression of Syk protein (Fig. 1,A, bottom). TNF also activated Syk in a dose-dependent manner, starting at 10 pM and continuing to increase at 1000 pM (Fig. 1,B). To determine the cell type specificity of TNF-induced Syk activation, myeloid (U937 and KBM-5) and epithelial (HeLa) cells were treated with TNF for the indicated times and examined for Syk activation. TNF activated Syk in all cell lines (Fig. 1,C) but was less pronounced than that observed in Jurkat cells (Fig. 1 A).

Whether TNF induces tyrosine phosphorylation of proteins other than Syk, was also investigated. To determine this, whole cell extracts from TNF-treated cells were resolved on SDS-PAGE and then analyzed by Western blot using anti-phosphotyrosine Ab (Fig. 1,D). The results show that TNF induced tyrosine phosphorylation in a dose-dependent manner of only one protein with a molecular mass of ∼72 kDa; this size corresponds to that of Syk protein. ZAP70 is another protein-tyrosine kinase with high homology to Syk protein (38). Whether TNF also activates ZAP70 was examined. To determine this, cells were treated with TNF for indicated times, and whole cell extracts were prepared, immunoprecipitated with anti-ZAP70 Ab, and then subjected to Western blot analysis using anti-phosphotyrosine Ab (Fig. 1,E). These results indicate that TNF does not induce the phosphorylation of ZAP70, suggesting that Syk is the only target of TNF-induced tyrosine phosphorylation. The Western blot analysis shows that Jurkat cells do express ZAP70 protein (Fig. 1 E).

It has been shown that piceatannol is a specific inhibitor of Syk protein-tyrosine kinase (39). To determine whether TNF-induced Syk activation is inhibited by piceatannol, Jurkat cells were incubated with piceatannol for 1 h and then treated with TNF for different times. Piceatannol completely inhibited TNF-induced activation of Syk (Fig. 2 A). This experiments further support the notion that TNF can activate Syk.

Whether piceatannol inhibits the tyrosine phosphorylation of proteins other than Syk, was also examined. To determine this, piceatannol-pretreated cells were treated with TNF for different times, the whole cell extracts were prepared and resolved on SDS-PAGE, and Western blot analysis was performed using anti-phosphotyrosine Ab (Fig. 2 B). These results clearly show that piceatannol primarily inhibits the TNF-induced phosphorylation of a single protein at 72 kDa, corresponding to that of Syk.

Because TNF-induced Syk activation occurs in seconds, we wondered whether it would affect TNF-induced JNK activation. To determine this, piceatannol-pretreated Jurkat cells were stimulated with 0.1 nM TNF for different times and then examined for JNK. As shown in Fig. 2 C, TNF induced activation of JNK as early as 5 min, and it reached maximum at 15 min. Treatment with piceatannol completely inhibited the TNF-induced JNK activation.

Because TNF is a potent activator of p38 MAPK and p44/p42 MAPK, we also investigated the role of Syk activation in activation of these MAPKs. Cells were treated as indicated above and examined for activation of p38 MAPK and p44/p42 MAPK using phospho-specific Abs. As shown in Fig. 2,D, TNF induced activation of p38 MAPK as early as 5 min, and it reached maximum at 15 min. Treatment with piceatannol again completely inhibited the TNF-induced activation of p38 MAPK. Similarly, TNF-induced activation of p44/p42 MAPK in Jurkat cells, and treatment with piceatannol suppressed TNF-induced activation of p44/p42 MAPK (Fig. 2 E). Thus, TNF-induced Syk activation regulated TNF-induced activation of JNK, p38 MAPK, and p44/p42 MAPK.

Metabolic inhibitors, although convenient, are not always specific. Therefore, it is possible that the inhibitory effect of piceatannol on TNF signaling was independent of inhibition of Syk activation. Thus, we compared the effect of TNF on Jurkat cells with JCaM1 cells, a clone of Jurkat cells genetically deficient in nonreceptor Syk protein-tyrosine kinase (40). As shown in Fig. 3,A, Jurkat cells express Syk protein but JCaM1 cells do not. As expected, TNF induced Syk activation in Jurkat cells but not in JCaM1 cells (Fig. 3,B). We next compared the Syk-expressing (Jurkat) and Syk-deficient (JCaM1) cells for TNF-induced activation of JNK, p38 MAPK, and p44/p42 MAPK. As can be seen, TNF activated JNK (Fig. 3,C), p38 MAPK (Fig. 3,D), and p44/p42 MAPK (Fig. 3 E) in Jurkat cells, but activation was minimal in JCaM1 cells, again suggesting that Syk activation was needed for the activation of these MAPKs.

Besides Syk protein, JCaM1 cells have been shown to lack lck protein (Fig. 3,A). To exclude the role of lck in TNF signaling, we used JCaM1 cells reconstituted with lck protein (Fig. 3,A). The treatment of lck-reconstituted cells with TNF failed to activate Syk (Fig. 3,F), JNK (Fig. 3,G), p38 MAPK (Fig. 3,H), or p44/p42 MAPK (Fig. 3 I). These results indicate that lck plays no role in TNF signaling.

TNF is one of the most potent activators of NF-κB. Whether Syk is also involved in TNF-induced NF-κB activation was investigated. As shown in Fig. 4,A (left), TNF activated NF-κB in a time-dependent manner, and this correlated with IκBα (an inhibitor of NF-κB) phosphorylation and degradation. Pretreatment of cells with piceatannol suppressed the TNF-induced NF-κB activation, IκBα phosphorylation, and IκBα degradation (Fig. 4 A, right).

We then examined the Syk-deficient JCaM1 and JCaM1/lck cells for TNF-induced NF-κB activation. As shown in Fig. 4,B, TNF activated NF-κB in JCaM1 and JCaM1/lck cells, but the maximum activation was ∼61 and 54% of that of Jurkat cells, respectively. TNF also induced IκBα phosphorylation and degradation in JCaM1 and JCaM1/lck cells (Fig. 4,B, bottom). Fig. 4,B shows that JCaM1 cells that lack both Syk and lck (Fig. 4,B, left) or lack only Syk (reconstituted with lck; Fig. 4 B, right) have decreased NF-κB activation. These results indicate that Syk plays a role in TNF-induced NF-κB activation, as it does for MAPKs. The level of TNF-induced NF-κB activation in JCaM1 cells found here and that reported previously (33) appears to vary perhaps due to the metabolic state of the cells.

To further confirm the role of Syk protein-tyrosine kinase in TNF-induced NF-κB activation, we transiently transfected the Jurkat cells with the wild-type Syk-cDNA-containing expression plasmid. As shown in Fig. 5, these cells showed an increase in expression of Syk protein (Fig. 5,A), in parallel with enhanced TNF-induced NF-κB activation (Fig. 5 B). Thus, these results suggest that Syk protein-tyrosine kinase may play a role in also TNF-induced NF-κB activation.

We also used siRNA to suppress Syk protein expression (41). For this, we transiently transfected siRNA-Syk into Jurkat cells. siRNA-Syk suppressed the expression of Syk protein (Fig. 5,C) and diminished the TNF-induced NF-κB activation (Fig. 5 D). These results again suggest that Syk plays an important role in NF-κB activation induced by TNF.

Activation of NF-κB inhibits and suppression of NF-κB stimulates TNF-induced apoptosis (42, 43, 44, 45). Whether increased activation of NF-κB by overexpression of Syk or reduction of NF-κB by down-regulation of Syk protein expression affects TNF-induced apoptosis was investigated. As shown in Fig. 6,A, Jurkat cells were sensitive to TNF-induced cytotoxicity, and overexpression of Syk protein by Syk-cDNA in these cells suppressed the TNF-induced cytotoxicity (Fig. 6,A). Similarly, the suppression of Syk protein by siRNA-Syk potentiated the TNF-induced cytotoxicity (Fig. 6,A). These results were further confirmed by using piceatannol. As shown in Fig. 6,B, pretreatment of cells with piceatannol enhanced the TNF-induced cytotoxicity. We also examined the effect of Syk on TNF-induced apoptosis by annexin V staining (Fig. 6, C and D), TUNEL assay (Fig. 6,E), and PARP cleavage (Fig. 6,F). All these results indicated that overexpression of Syk protein by Syk-Wt cDNA decreases TNF-induced apoptosis and reduction of Syk protein by siRNA increases the TNF-induced apoptosis. JCaM1 that lack Syk are also more sensitive than wild-type or Syk-transfected cells to TNF-induced apoptosis (Fig. 6,D). Thus, whereas TNF-induced NF-κB activation paralleled Syk protein expression (Fig. 5), the TNF-induced apoptosis showed a reciprocal relationship.

Whether TNF-induced JNK and p38 MAPK are affected by modulation of Syk expression was also investigated. Results in Fig. 7,A show that TNF induced JNK activation in Jurkat cells; this activation was enhanced by overexpression of Syk and decreased by transfection of cells with Syk siRNA. Similarly, results in Fig. 7 B show that TNF induced p38 MAPK activation in Jurkat cells; this activation was enhanced by overexpression of Syk and decreased by transfection of cells with Syk siRNA. Thus, these results demonstrate that Syk can also regulate the TNF-induced JNK and p38 MAPK activation pathway.

Whether p38 MAPK or JNK play a role in TNF-induced apoptosis was also investigated by using SP600125, a specific JNK inhibitor, and SB205380, a specific p38 MAPK inhibitor. Annexin V staining results showed that TNF induced apoptosis in wild-type Jurkat cells, and syk overexpression inhibited the apoptosis (Fig. 7 C). Pretreatment of cells with JNK inhibitor or p38 MAPK inhibitor enhanced the TNF-induced apoptosis indicating the antiapoptotic role of these kinases.

How Syk is activated by TNF and modulates TNF signaling was further investigated. It is possible that the Syk protein physically interacts with TNF receptor or receptor-associated proteins. To explore this possibility, Jurkat cells were treated with 1 nM TNF for different times. Whole cell extracts were immunoprecipitated with anti-Syk Ab and then analyzed by Western blotting anti-TNFR1 or anti-TNFR2 Abs. The results show that TNF induced the association of Syk protein with TNFR1 (Fig. 8,A1) and with TNFR2 (Fig. 8 A2). No association was noted in the absence of TNF. The maximum signal appeared at 60 s after the ligand treatment with TNFR1 and 30 s after that with TNFR2. These results suggest that Syk protein does associate with the TNFR in a ligand-dependent manner.

To confirm this association, we performed a reciprocal experiment in which TNF-treated whole cell extracts were immunoprecipitated with anti-TNFR1 or TNFR2 Abs and then blotted with anti-Syk Ab. These results also show that Syk protein associates both with TNFR1 (Fig. 8,B1) and TNFR2 (Fig. 8 B2).

To further confirm the recruitment of Syk protein-tyrosine kinase into the TNF receptor complex, we transiently transfected the Jurkat cells with the wild-type Syk-cDNA expression plasmid. As shown in Fig. 8,C1, these cells showed a TNF-induced recruitment of Syk into TNF receptor complex (Fig. 8,C1, top and middle), in parallel with enhanced expression of Syk protein (Fig. 8,C1, bottom). We also used siRNA to suppress Syk protein expression (41). For this, we transiently transfected siRNA-Syk into Jurkat cells. siRNA-Syk suppressed the expression of Syk protein (Fig. 8,C2, bottom) and diminished the TNF-induced recruitment of Syk into TNF receptor complex (Fig. 8 C2, top and middle). Thus, these results strongly suggest that Syk protein-tyrosine kinase is recruited into the TNF receptor complex in a ligand-dependent manner.

In this report, we investigated the role of Syk protein tyrosine kinase in TNF-induced signaling. We present from four different lines of evidence indicating that TNF can activate Syk in a wide variety of cells and that this activation modulates TNF-induced activation of JNK, p38 MAPK, p44/p42 MAPK, NF-κB activation, and apoptosis. The four different approaches were the use of metabolic inhibitors, Syk-deleted cells, overexpression of Syk protein by Syk-cDNA transfection, and down-regulation of Syk protein expression by siRNA-Syk. We also present evidence that Syk protein interacts with the TNF receptors.

Previously, we have shown that piceatannol can block TNF-induced NF-κB activation in both Syk-expressing as well as Syk-deleted JCaM-1 cells (54). Although other investigators have reported that piceatannol is a specific inhibitor of Syk, studies based on piceatannol alone are not sufficient to implicate Syk in cytokine signaling (Oliver et al., Ref. 39). Therefore, we have used other approaches including Syk-deleted cells, Syk-overexpressing cells, Syk siRNA, and activation of Syk activity by TNF to implicate Syk in the TNF signaling leading to NF-κB activation.

TNF activated Syk in three different cell lines. Although TNF has been shown to activate protein-tyrosine kinase, c-Src (21, 22, 23), p53/p56 lyn (20), Etk/Bmx (46), and PyK2 (18, 19, 24), this is the first report to indicate that TNF can activate a protein-tyrosine kinase, Syk, in different cells. TNF did not activate ZAP70, another T cell-specific kinase homologous to Syk. How TNF activates Syk is not clear. Because TNFR1 can recruit PI3K (47) and inhibition of PI3K by wortmannin blocks the activation of Syk and Syk-associated kinase Pyk2 (19), it is possible that TNF-induced activation of Syk is through PI3K. Because TNF-induced PI3K activation could be seen only after 15 min of TNF treatment (47) and TNF-induced Syk activation occurred within 30 s, it is unlikely that PI3K activation preceded Syk activation. The role of Pyk2 in NF-κB activation has been established (55). Whether TNF activates NF-κB through sequential activation of Pyk2 and Syk is not clear.

How Syk activation mediates TNF-induced activation of JNK and other MAPKs is not clear. TNF has been shown to activate the Syk-associated kinase Pyk2, which has been linked with JNK activation (24). It has been shown that Src and adapter proteins Grb2, crk, and p130^Cas are downstream mediators of Pyk2 leading to the activation of p44/p42 MAPK and JNK (48). The association of Grb2 directly with TNFR1 has been reported (49). Thus, it is possible that TNF-induced Syk activation recruits Pyk2, which in turn recruits Grb2, leading to MAPK activation. TNF has also been shown to activate c-Raf-1 kinase (50, 51). The sequential interaction of TNFR1, Syk, Grb2, SOS, Ras, and Raf-1 kinase can lead to MAPK activation. It has been shown that Raf-1 leads to only p44/p42 MAPK activation and that JNK activation is initiated by mitogen-activated protein kinase kinase kinase (52). MKK7 and MKK3 have been implicated in TNF-induced JNK and p38 MAPK activation, respectively. Our results indicate that Syk lies upstream of all three major MAPKs activated by TNF.

Our results indicate that Syk activation also plays a role in TNF-induced NF-κB activation. The role of c-Src in TNF-induced NF-κB activation has been demonstrated (21, 23). It has been shown that c-Src causes the tyrosine phosphorylation of IκB kinase-β needed for TNF-induced NF-κB activation (23). Whether TNF-induced Syk can also directly induce the phosphorylation of IκBα directly or through activation of c-Src is not clear at present. Abu-Amer et al. (53) showed that c-Src could directly phosphorylate tyrosine 42 in IκBα. It is unlikely that tyrosine 42 phosphorylation of IκBα by Syk could enhance TNF-induced NF-κB activation, because this phosphorylation has been implicated in suppression of TNF-induced NF-κB activation.

We showed that an increase in expression of Syk enhanced TNF-induced NF-κB activation, and this correlated with suppression of TNF-induced cytotoxicity. These results are consistent with previous reports showing that NF-κB activation negatively regulates TNF-induced apoptosis (42, 43, 44, 45). Our results also demonstrate that down-regulating Syk expression decreased NF-κB activation and enhanced TNF-induced cytotoxicity.

We also show that Syk interacted with both TNFR1 and TNFR2, which was enhanced by up-regulation of Syk using Syk-cDNA and suppressed by down-regulation of Syk using siRNA-Syk. Whether TNFR interacts with Syk protein directly or indirectly through some intermediate is very difficult to demonstrate at the whole cell level, especially when the association between TNFR and Syk is TNF dependent. The increase in Syk expression with Syk cDNA and decrease with Syk siRNA showed proportionately increase or decrease in the recruitment of the TNF receptor. These results, although not proof, do suggest that the interaction between TNFR and Syk is direct. This interaction was, however, dependent on TNF. Protein-tyrosine kinase and Etk/Bmx have been shown to associate with TNFR2 in endothelial cells (46). This interaction, however, was found to be TNF independent. Thus, overall our results indicate that a Syk can be activated by TNF and that this activation positively regulates TNF-induced activation of MAPKs and NF-κB and negatively regulates TNF-induced apoptosis.