Vol. 13, 13–18, January 2002 Cell Growth & Differentiation 13

Tumor Necrosis Factor ␣ (TNF-␣) Activates Jak1/Stat3-Stat5B Signaling through TNFR-1 in Human B Cells1

Sebastiano Miscia,2 Marco Marchisio, Alfredo Grilli, one of two specific and ubiquitously expressed receptors:

Valentina Di Valerio, Lucia Centurione, TNFR-1 (CD 120 a, Mr 55,000) and TNFR-2 (CD 120 b, Mr Giuseppe Sabatino, Francesco Garaci, Giorgio Zauli, 75,000). These molecules share homology in their extracel- Ezio Bonvini, and Angela Di Baldassarre lular domains but are dissimilar in their intracellular portion (3, Unit at the Department of Biomorphology [S. M., M. M., 4). They mediate a distinct but, at least in part, overlapping A. G., V. D. V., L. C., A. D. B.] and Section of Neonatology [G. S.], School of Medicine, University of Chieti, 66100 Chieti, Italy; repertoire of responses. TNFR-1 appears to mediate the Dipartimento di diagnostica per immagini e Radiologia Interventistica, majority of TNF actions including cytotoxicity, fibroblast pro- University of Roma Tor Vergata, D0100 Rome, Italy [F. G.]; Section of liferation, and immunological and inflammatory responses (5, Human Anatomy at the Department of Morphology and Embryology, University of Ferrara, 44100 Ferrara, Italy [M. M.]; Department of Human 6), whereas an independent role in the proliferation of mono- Morphology, School of Medicine, University of Trieste, 34127 Trieste, cytes and T lymphocytes, and a function complementary to Italy [G. Z.]; and Laboratory of Immunobiology, Center for Biologistics TNFR-1 in cytotoxicity have been suggested for TNFR-2 Evaluation and Research, Bethesda, Maryland 20814 [E. B.] (7–9). Because the TNF receptors are devoid of enzymatic activity (3, 4), the signal appears to be mediated by the Abstract recruitment of cytosolic proteins through specific protein- The biological actions of tumor necrosis factor ␣ (TNF- protein interaction domains. -induced ag- ␣) are mediated by two cell surface receptors, TNFR-1 gregation enables the interaction of TNFR-1 with the TNF- and TNFR-2. These receptors do not display protein receptor-associated death domain protein, a protein that activity. Nevertheless, an early TNF- couples TNFR-1 to gene modulation through NF␬B activa- induced activation of specific tyrosine kinases has tion, or cell death, through the involvement of caspases (10). been reported as an important cue to the cellular Ligation of TNFR-2 can affect the expression of genes in- response to this cytokine. Here we present evidence volved in immediate early prosurvival responses (11), and that TNF-␣ induces the activation of the cytoplasmic this is also dependent on the activation of NF␬B. The TNFR-2 Janus tyrosine kinases Jak1 and Tyk2 in both human associated factor TRAF-2 is in fact engaged in a number of healthy peripheral and lymphoma B cells. This event protein-protein interactions that lead to the release of NF␬B was accompanied by the recruitment of a specific set from its the cytosolic inhibitor, allowing its nuclear translo- of latent cytosolic transcription factors, Stat3 and cation (12, 13). Whereas several steps of TNF signaling have Stat5b. Furthermore, Jak1 coprecipitated with TNFR-1 been elucidated, many of the molecular events evoked by after TNF-␣ treatment. These data suggest that at least this cytokine are still unclear. For instance, an early TNF- in human B cells this cytokine can exert its biological induced activation of tyrosine kinase activity has been re- effects through the Jak-Stat signaling pathway and ported, although neither TNF receptor displays intrinsic pro- that such signals are initiated through an interaction tein tyrosine kinase activity (3, 4). Furthermore, tyrosine between TNFR-1 and Jak 1. kinases inhibitors suppress TNF-induced phosphorylation of ␬ a Mr 23,000 nuclear protein (14), activation of NF B (15), and DNA fragmentation (16). Additionally, recent evidence in mu- Introduction rine adypocites have provided additional clues to the exist- TNF-␣3 is an endogenous mediator of inflammatory and ence of a tyrosine kinase signaling pathway involved in cer- immune responses able to regulate cell metabolism by mod- tain cellular response to this cytokine (17). The aim of the ulating gene expression and by controlling cell proliferation present study was to investigate, in human B primary and and survival (1, 2). Important progress has been made in lymphoma cells, the involvement of a family of cytoplasmic defining the signaling intermediates of TNF activity. The in- protein tyrosine kinases, the Jaks, in tracellular signal is initiated by the binding of the cytokine to mediated by TNF.

Results Received 5/14/01; revised 10/8/01; accepted 11/9/01. The costs of publication of this article were defrayed in part by the To characterize the susceptibility of primary B and Ramos payment of page charges. This article must therefore be hereby marked cells to the TNF-␣ action, we determined effects on cell advertisement in accordance with 18 U.S.C. Section 1734 solely to indi- viability by trypan blue dye exclusion and on morphology by cate this fact. 1 Supported by “G.D’Annunzio” University 1999/2000 local and Ministero means of semithin sections after scalar doses of TNF-␣ (50, dell’Universita` e Ricerca Scientifica e Tecnologica-COFIN funds. 100, 250, and 500 nM). Table 1 shows that after 16 h of 2 To whom requests for reprints should be addressed, at Dipartimento di treatment with 250 nM TNF-␣, the most effective concentra- Biomorfologia, Via dei Vestini 6, 66100 Chieti, Italy. Phone: 39-871- 3555304; Fax: 39-871-574361; E-mail: [email protected]. tion in our hands, Ͼ 50% of lymphoma and primary B cells 3 The abbreviations used are: TNF, tumor necrosis factor; TNFR-1, tumor failed to exclude trypan blue. The semithin section samples necrosis factor receptor-1; TNFR-2, tumor necrosis factor receptor-2; NF␬B, nuclear factor ␬B; Jak, ; Stat, signal transducer and disclosed the well-known apoptotic morphological changes activator of transcription. that are hallmarks of apoptosis (Fig. 1). To shed on the 14 Jak-Stat Signaling Activation by TNF-␣ in Human B Cells

Table 1 Effect of TNF-␣ (16 hr) on the viability of primary B and Ramos lymphoma cellsa

% of viable % of viable Treatment primary B cells Ramos cells Control 96 Ϯ 2.0 94 Ϯ 3.0 TNF-␣ (50 nM)59Ϯ 2.5 57 Ϯ 2.0 TNF-␣ (100 nM)52Ϯ 3.0 54 Ϯ 2.5 TNF-␣ (250 nM)48Ϯ 4.2 43 Ϯ 5.0 TNF-␣ (500 nM)49Ϯ 3.5 50 Ϯ 4.0 a Data are the mean of five separate experiments Ϯ SD.

Fig. 1. A and B, effect of TNF-␣ (16 h) on the morphology of Ramos cells. Compared with control cells (A), TNF-treated cells display the typical apoptotic rearrangements such as chromatin margination and formation of micronuclei (B). The results presented are representative of at least three separate experiments

Fig. 2. A–D, effect of TNF-␣ the phosphorylation of Jak1 and Tyk2 in Ramos (A and B) and primary B cells (C and D). Cells were labeled with 32 ␮ upstream molecular events related to the biological effects [ P]Pi (100 C/ml) in phosphate-free DMEM for 1 h. After labeling, exerted by TNF-␣ on these cells, we next investigated the washes, and TNF-␣ treatment for the times reported, proteins were im- munoprecipitated and separated as described in “Materials and Meth- possible involvement of Jak proteins. This was first accom- ods.” Autoradiography of the original membranes is shown (AR). In some 32 plished by an in vivo labeling of cells with [ P]Pi followed by experiments, Jak1 and Tyk2 were phosphorylated with an anti-P-tyr an- TNF-␣ treatment and immunoprecipitation of each Jak pro- tibody to specifically detect the tyrosine phosphorylation (A and B, middle panels). Immunoblots of the same membranes incubated respectively tein with specific antisera. The administration of the cytokine with an anti-Jak1 or -Tyk2 demonstrated the equal loading of the induced the phosphorylation of Jak1 and Tyk2 as early as 5 proteins. When reported, the identity of TNFR-1 was achieved by incu- bating the membrane with an anti-TNFR-1 antibody. The results are min in both Ramos (Fig. 2, A and B) and primary B cells (Fig. representative of three independent experiments. 2, C and D). The fact that Jak2 and Jak3 phosphorylation were scarcely or not detected in either cell only influenced the number of viable cells without producing substantial changes in induced Jak1 and Tyk2 phosphorylation (data not 3, A and B). Similar experiments performed on primary B

shown). In Ramos cells we detected a Mr 50,000/60,000 cells demonstrated that TNFR-1 coprecipitated with Jak1 phosphorylated protein coprecipitated with phosphorylated after TNF-␣ treatment (Fig. 3 C), whereas TNFR-2 was nei- Jak1 (but not with Tyk2). We therefore reasoned that this ther phosphorylated nor associated with any phosphoprotein protein could be one of the TNF receptors, namely TNFR-1, (data not shown). To assess whether Jak1 and Tyk2 were which could have recruited Jak 1. Therefore we reacted the activated by TNF-␣ in both cell types we immunoprecipitated same blot with an anti-TNFR-1 antibody, and this demon- Jak1 and Tyk2 from control and TNF-␣-treated Ramos and

strated that the Mr 50,000/60,000 protein was indeed primary B cells, and measured their tyrosine kinase activity TNFR-1 (Fig. 2A). This result was additionally confirmed by by a kinase assay. As shown (Fig. 4, A–D), the activities of immunoprecipitating both TNFR-1 and TNFR-2 receptors both kinases were strongly stimulated by TNF-␣. from 32P-labeled Ramos cells treated with TNF-␣. After sep- The activation of Jak proteins is known to lead to tyrosine aration of the immunoprecipitates through SDS-PAGE and phosphorylation of Stat proteins. Therefore, we screened for transfer onto nitrocellulose and immunoblotting, autoradiog- possible Stat(s) involvement downstream of the TNF-␣- raphy demonstrated that Jak1 was phosphorylated and as- mediated activation of Jak1 and Tyk2. Stat proteins were, sociated with TNFR-1, whereas TNFR-2 was evidently phos- therefore, immunoprecipitated from lysates from untreated phorylated but not associated with any phosphoprotein (Fig. and TNF-␣-treated Ramos cells and, after separation and Cell Growth & Differentiation 15

Fig. 3. A–C, effect of TNF-␣ (5 min) on the phos- phorylation of TNFR-1 and TNFR-2. A and B, Ramos 32 ␮ cells were labeled with [ P]Pi (100 C/ml) in phos- phate-free DMEM for 1 h. After labeling, washes, and TNF-␣ treatment, proteins were immunoprecipi- tated and separated as described in “Materials and Methods.” Autoradiography of the original mem- brane is shown in the top panel (AR). Immunoblots of the same membranes incubated respectively with an anti-TNFR-1 or -TNFR-2 antibody indicated the ϳ equal loading. The identity of the band of Mr 130,000 as Jak1 was achieved by reprobing the TNFR-1 membrane with a specific anti-Jak1 anti- body (bottom panel). Detection of TNFR-1/Jak1 as- sociation in primary B cells (C). Jak1 was immuno- precipitated from lysates either untreated or treated with TNF-␣; immunoprecipitates were sepa- rated by SDS-PAGE and transferred onto nitrocellu- lose membrane, as described in “Materials and Methods.” The membrane was then reacted with an anti-TNFR-1 antibody, and the protein was revealed by enhanced chemiluminescence reaction. The re- sult is representative of three independent experi- ments.

Fig. 4. A–D, effect of TNF-␣ on Jak1 and Tyk2 ty- rosine kinase activity in Ramos cells (A and B) and primary B cells (C and D). Jak1 and Tyk2 were im- munoprecipitated and processed as described in “Materials and Methods.” Kinase activity was meas- ured by autoradiography (in vitro phosphorylation with [␥-32P]ATP).

transfer onto nitrocellulose, their phosphorylation level was phorylated Stats underwent to translocation from the cyto- determined by anti-P-Tyr antibody. Only phosphorylation of plasm to the nuclear compartment, we performed immuno- Stat3 and Stat5b was evident after TNF-␣ stimulation, and fluorescence analyses for both Stat3 and Stat5b at different this was detectable within 5 min of treatment, consistent with times after TNF-␣ treatment. Both Stat3 and Stat5b translo- the activation kinetics of Jak1 and Tyk2, and was clearly cated into the nucleus within 20 min of treatment, and this evident up to 40–60 min of TNF-␣ stimulation (Fig. 5A). The was clearly evident at 40 min of TNF-␣ administration (Fig. 6). expression and/or the phosphorylation of other Stats were either undetected (Stat5a) or not modified when compared with untreated cells (Fig. 5B). Stat3 and Stat5b tyrosine Discussion phosphorylation was also evident in primary B cells after Jaks are cytoplasmic protein tyrosine kinases that mediate TNF-␣ treatment (Fig. 5C). To investigate whether the phos- cytokine signaling by phosphorylating different intracellular 16 Jak-Stat Signaling Activation by TNF-␣ in Human B Cells

Fig. 5. A–C, effect of TNF-␣ (5 min) on Stat phosphorylation in Ramos (A and B), and primary B cells (C). Cells were treated with the cytokine, and the proteins were immunoprecipitated from cell lysates, separated, and transferred onto nitrocellulose as described in “Materials and Methods.” Membranes were incubated with an anti-P-tyr antibody, and the same membranes were reprobed respectively with the corresponding anti-Stat to show the equal loading for each experiment. The results are representative of three independent experiments.

Fig. 6. Immunofluorescence analy- sis of Stat3 and Stat5b subcellular distribution after TNF-␣ treatment (20 min) of Ramos lymphoma cells. In regard to controls, TNF-␣ clearly induces the translocation of Stat3 and Stat5b into the nuclei of which the counterstaining was obtained by means of 4Ј,6-diamidino-2-phenylin- dole. Results are representative of three separate experiments.

targets (18, 19). In mammals, four Jak family members have is available on the signaling pathway(s) that lead to tyrosine been described: Jak1, Jak2, Jak3, and Tyk2. These kinases phosphorylation. Our data demonstrate that the administra- are unique in containing tandem kinase and kinase-like do- tion of TNF-␣ to B human cells induces tyrosine phospho- mains (20, 21). Their activation is promoted by the dimeriza- rylation and activation of the Jaks Jak1 and Tyk2. It is well tion of cytokine receptors on ligand binding. Because previ- known that binding of TNF-␣, active as homotrimer, results in ous studies demonstrated induction of a tyrosine kinase the clustering of monomeric TNF receptors (24, 25). This activity in response to TNF-␣ and because TNF-␣ receptors self-association seems to fit with the model suggested for are devoid of known enzymatic activity, we investigated the Jak activation (19), allowing the juxtaposition and the possible involvement of Jaks in TNF-␣ signaling. The ability transphosphorylation of the kinases. Because TNFR-1 forms of this cytokine to induce cell death in lymphoma cells has homotrimers by associating with itself and with cytoplasmic been already described (22, 23), but thus far little information proteins implicated in cell death, it is conceivable that only Cell Growth & Differentiation 17

one member of the Jak family (i.e., Jak1) is recruited. In this samples exceeding a purity of 95% were used for the ex- context, preliminary results of in vitro binding assays indicate periments. Blood peripheral B lymphocytes and Ramos B an interaction between Jak1 and TNFR-1, yet this observa- lymphoma cells were grown in RPMI 1640 (Mascia Brunelli, tion needs to be additionally explored, because in the ex- Milan, Italy) containing 10% FCS (Mascia Brunelli), 2 mM periments reported in the present paper, the association L-glutamine, 5 mM HEPES (pH 7.3), and penicillin-streptomy- between Jak1 and TNFR-1 appears to be ligand dependent. cin (50 IU/ml and 50 mg/ml, respectively) at 37°C with 5% ␣ Regardless, the finding that both Jak1 and Tyk2 become CO2. The susceptibility of cells to TNF- was monitored by tyrosine phosphorylated and activated suggests that a Jak1/ determining effects on cell viability and morphology by Tyk2 complex is also recruited in the TNFR-1 complex. Given trypan blue exclusion test and electron microscopy analysis, the well-established apoptotic role of TNFR-1, the recruit- respectively, after incubating cells with scalar doses of ment of Jak1 and Tyk2, which are generally involved in cell TNF-␣ (Boehringer Mannheim, Mannheim, Germany) for proliferation or survival, in the TNFR-1 complex appears to 16 h. Once assessed, 250 nM was determined as the most be a paradoxical event, which needs more investigation. In effective dose, and molecular changes were investigated this context it is worth noting that Jak1 and/or Tyk2 have within a time interval ranging from 5 min to 60 min. been demonstrated recently to be involved in the induction of Transmission Electron Microscopy. For semithin sec- either apoptosis or cell growth suppression (26, 27). These tions, cells were fixed in suspension for 30 min with 2.5% findings would argue for different roles of these Jaks that glutaraldehyde in 0.1 M cacodylate buffer (pH 7.2), postfixed could depend on the function of either Jak homology do- in 1% OsO4, and embedded in Spurr resin. mains or of Jak NH2-terminal region that ultimately influence Immunocytochemical Detection of Stat3 and Stat5b. the interaction of the kinase to its cognate cytokine receptor Cells were first spun onto coverslips fixed in 4% paraform- subunit. How TNFR-2 engagement may play a role remains aldehyde/PBS for 10 min at room temperature, washed twice to be determined, because in Ramos cells this receptor with PBS, incubated with PBS for additional 15 min to appears to be phosphorylated but is associated with any quench the remaining paraformaldehyde, and saturated/ phosphoprotein, and in primary B cells TNFR-2 is apparently permeabilized using NET gel solution [150 mM NaCl, 5 mM not involved in TNF signaling. These findings are more in line EDTA, 50 mM Tris-HCl (pH 7.4), 0.05% NP40, 0.25% Carra- with the hypothesis that TNFR-2 is rather essential in TNF- geenan Lambda gelatin, and 0.02% Na azide] for 30 min at mediated cytotoxicity in T cells (8). Alternatively, these find- room temperature. After two washes with Net gel solution, ings might reflect different signaling features between neo- cells were treated with anti-Stat3 or Stat5b rabbit polyclonal plastic cells and primary cells. antibodies (diluted 1:50; Santa Cruz Biotechnology, Santa Activated Jaks phosphorylate members of a family of tran- Cruz, CA) for 60 min at room temperature. After two washes scription factors, the Stats. In mammalian species, seven with Net gel solution, goat antirabbit IgG (GAR-FITC, diluted Stat proteins are known: Stat1, Stat2, Stat3, Stat4, Stat5a, 1:50) was added to the cells and incubated for 45 min at Stat5b, and Stat6 (18, 19, 28–30). Tyrosine phosphorylation room temperature. After three additional washes (two in Net of the receptor by Jaks provide docking sites for Stats. Once gel solution and one in PBS), the nuclei were counterstained phosphorylated, Stats are released from the receptor, dimer- with 4Ј,6-diamidino-2-phenylindole (Sigma Chemical Co.) ize, and translocate into the nucleus where they specifically and mounted in DABCO (Sigma Chemical Co.)-glycerol- bind DNA and activate the transcription of target genes, PBS. The immunostaining specificity for the monitored pro- many of which are relevant for the inflammatory responses teins was confirmed by the absence of any reactivity when: (18, 19, 31, 32). Different ligands act on a restricted sets of (a) primary antibodies directed against the other Stat proteins Jaks and Stats. Our data demonstrate that in human B cells, (diluted 1:50; Santa Cruz Biotechnology) and (b) normal rab- the TNF-␣-induced Jak1 and Tyk2 activation are accompa- bit serum (diluted 1:50) were used. nied by Stat3 and Stat5b tyrosine phosphorylation and trans- 32P-Labeling, Solubilization of Proteins, Immunopre- location into the nucleus, strongly demonstrating that in cipitation, and Immunoblotting. Cells were maintained for these cells, additional elements can control the transcrip- 32 ␮ 3 h in phosphate-free DMEM, labeled with [ P]Pi (100 Ci/ml; tional machinery elicited by TNF-␣. How Stat3 and Stat5b Amersham, Milan, Italy) for 1 h, and stimulated as described activation modulates the transcriptional response of these above. Cells were harvested by centrifugation at 4°C, solubi- human cells and the precise role of these factors in the lized in lysis buffer [60 mM Tris-HCl (pH 7.8), 150 mM NaCl, 5 mM cellular response to TNF-␣ are currently under investigation ␮ EDTA, 10% glycerol, 2 mM Na2Vo4,25mM NaF, 10 g/ml in our laboratories. Nevertheless, our observations suggest leupeptin, 10 ␮g/ml aprotinin, 1 mM 4-(2-aminoethyl)benzene- ␣ that among the different cellular responses evoked by TNF- sulfonyl fluoride, 1 mM TLCK (N-tosyl-L-phenylalanine chloro- in B lymphoid cells, recruitment of Jak1 through TNFR-1and methyl ketone)] and normalized at 400 ␮. Whole cell gene modulation by Stat3 and Stat5b may represent con- lysates were incubated at 4°C for 1 h with 1 ␮g of either rabbit tributing factors. polyclonal anti-Jak, anti-Stat, anti-TNFR-1, or anti-TNFR-2 an- tibodies (all from Santa Cruz Biotechnology) that were coupled Materials and Methods previously to goat antirabbit IgG magnetic beads (Dynal). Im- Cell Culture and Treatment. B lymphocytes were obtained munocomplexes were collected by a magnet, electrophoresed, from peripheral blood of healthy consenting donors by using and transferred onto nitrocellulose. The phosphorylation level of magnetic beads coated with anti-CD 19 monoclonal anti- the different proteins was assessed by autoradiography, body (Dynabeads M-450 CD19; Dynal, Oslo, Norway). Only whereas the specific tyrosine phosphorylation was assessed by 18 Jak-Stat Signaling Activation by TNF-␣ in Human B Cells

means of a mouse monoclonal anti-P-Tyr antibody (PY-99; 13. Wajant, H., Grell, M., and Scheurich, P. TNF receptor associated Santa Cruz Biotechnology). The detection and the equal load- factor in cytokine signaling. Cytokine Growth Factor Rev., 10: 15–26, ing of TNFR-1 or TNFR-2 was assessed by Western blot ex- 1999. 14. Ji, L., Zang, G., and Hirabayashi, Y. Inhibition of tumor necrosis factor periments reacting the same membranes with mouse mono- and ceramide-induced internucleosomal DNA fragmentation by herbimy- clonal anti-TNFR1 and anti-TNFR2 antibodies. cin A in U937. Biochem. Biophys. Res. Commun., 212: 640–647, 1995. In Vitro Kinase Assay. Jak1 and Tyk2 were immunopre- 15. Reddy, S. A. G., Chaturvedi, M., Darnay, B. G., Chan, H., Higuchi, M., cipitated as described above from equal numbers of un- and Aggarwal, B. B. Reconstitution of NF-kB activation induced by tumor treated and TNF-treated cells (2 ϫ 107). Immunoprecipitates necrosis factor requires membrane-associated components: comparison with pathway induced by ceramide. J. Biol. Chem., 269: 25369–25374, were washed once with lysis buffer, once with 0.5 M LiCl and 1994. 20 mM Tris (pH 7.5), and once with distilled water. Samples 16. Ji, L., Zang, G., and Hirabayashi, Y. Tumor necrosis factor increases were resuspended in 30 ml of 20 mM Tris (pH 7.5), 10 mM tyrosine phosphorylation of a 23-kDa nuclear protein in U937 trough MnCl, and 10 mCi of [␥-32P]-ATP (Amersham) per kinase ceramide signaling pathway. Biochem. Biophys. Res. Commun., 215: assay, and incubated at room temperature for 15 min. Sam- 489–496, 1995. ples were then washed with distilled water, and proteins 17. Guo, D. Q., Dumbar, J. D., Yang, C. H., Pfeffer, L. M., and Donner, were separated by SDS-PAGE, transferred onto nitrocellu- D. B. Induction of Jak/STAT signaling by activation of the type 1 TNF receptor. J. Immunol., 160: 2742–2750, 1998. lose, and analyzed by autoradiography. Equal loading of 18. Schindler, C., and Darnell, J. E. Transcriptional responses to polypep- samples was assessed by immunoblotting the same mem- tide ligands: the Jak/STAT pathway. Annu. Rev. Biochem., 64: 621–651, branes with anti-Jak1 and Tyk2 antibodies. 1995. 19. Wilks, A. F., and Harpur, A. G. Cytokine signal transduction and the Jak family of protein tyrosine kinases. Bioessays, 16: 313–318, 1994. References 20. Wilks, A. F., Harpur, A. G., Kurban, R. R., Ralph, S. J., Zurcher, G., and 1. Fiers, W. Tumor necrosis factor: characterization of the molecular, Ziemiecki, A. Two novel protein tyrosine kinases, each with a second cellular and in vivo level. FEBS Lett., 285: 199–212, 1991. phosphotransferase-related catalytic domain define a new class of protein kinase. Mol. Cell. Biol., 11: 2057–2065, 1991. 2. Beutler, B. TNF, immunity and inflammatory disease: lessons of the past decade. J. Investig. Med., 43: 227–235, 1995. 21. Harpur, A. G., Andres, A. C., Ziemiecki, A., Aston, R. R., and Wilks, A. F. JAK2, a third member of the JAK family of protein tyrosine kinases. 3. Loetscher, H., Pan, Y. C., Lahm, H. W., Gentz, R., Brockhaus, M., Oncogene, 7: 1347–1353, 1992. Tabuchi, H., and Lesslauer, W. Molecular cloning and expression of the 22. Di Pietro, R., Santavenere, E., Centurione, L., Stuppia, L., Centurione, human 55kd tumor necrosis factor receptor. Cell, 61: 351–359, 1990. M. A., Vitale, M., and Rana, R. TNF ␣-induced apoptosis in Daudi cells: 4. Smith, C. A., Davis, T., Anderson, D., Solam, L., Beckmann, M. P., multiparametric analysis. Cytokine, 9: 463–468, 1997. Jerzy, R., Dower, S. K., Cusman, D., and Goodwin, R. G. A receptor for 23. Miscia, S., Di Baldassarre, A., and Cataldi, A. Shift of DNA polymerase tumor necrosis factor defines an unusual family of cellular and viral pro- ␣ distribution and activity in apoptotic human leukemia cells. Biochem. teins. Science (Wash. DC), 248: 1019–1023, 1990. Biophys. Res. Commun., 234: 203–208, 1997. 5. Engelmann, H., Holtmann, H., Barkebush, C., Avni, Y. S., Sarov, I., 24. Eck, M. J., and Sprang, S. R. The structure of tumor necrosis factor Nophar, Y., Hadas, E., Leitner, D., and Wallach, D. Antibodies to a soluble ␣ at 2.6 A resolution: implications for receptor binding. J. Biol. Chem., 264: form of a tumor necrosis factor (TNF) receptor have a TNF-like activity. 17595–17605, 1989. J. Biol. Chem., 265: 14497–14504, 1990. 25. Banner, D. W., D’Arcy, A., Janes, W., Gentz, R., Schoenfeld, H. J., 6. Pfeffer, K., Matsuyama, T., Kundig, T. M., Wakeham, A., Kishihara, K., Broger, C., Loetscher, H., and Lesslauer, W. Crystal structure of the Shahinian, A., Wiegmann, K., Ohashi, P. S., Kronke, M., and Mak, T. W. soluble human 55 kd TNF receptor-human TNF␤ complex: implications Mice deficient for the 55 kd tumor necrosis factor receptor are resistant to for TNF receptor activation. Cell, 73: 431–445, 1993. endotoxic shock, yet succumb to L. Monocytogenes infection. Cell, 73: 26. Gamero, A. M., and Larner, A. C. Vanadate facilitates ␣ 457–467, 1993. mediated apoptosis that is dependent on the Jak-Stat pathway. J. Biol. 7. Tartaglia, L. A., Goeddel, D. V., Reinolds, C., Figari, I. S., Weber, R. F., Chem. 276: 13547–13553, 2001. Fendly, B. M., and Palladino, M. A., Jr. Stimulation of human T-cell 27. Nakashima, O., Terada, Y., Hanada, S., Yamamoto, K., Kumehana, proliferation by specific activation of the 75-kDa tumor necrosis factor M., Sasaki K., and Marumo, F. Activated Stat1 suppresses proliferation of receptor. J. Immunol., 155: 4637–4641, 1993. cultured rat mesangial cells. Kidney Int. 57: 2249–2257, 2000. 8. Heller, R. A., Song, K., Fan, N., and Chang, D. J. The p70 tumor 28. Fu, X. Y. A transcription factor with SH2 and SH3 domains is directly ␣ necrosis factor receptor mediates cytotoxicity. Cell, 70: 47–56, 1993. activated by an interferon -induced cytoplasmic protein tyrosine ki- nase(s). Cell, 70: 323–335, 1992. 9. Chan, F. K., and Lenardo, M. J. A crucial role for p80 TNF-R2 in 29. Hou, J., Schindler, U., Henzel, W. J., Ho, T. C., Brasseur, M., and amplifying p60 TNF-R1 apoptosis signals in T lymphocytes. Eur. J. Im- McKnight, S. L. An Interleukin-4-induced transcription factor: IL-4 Stat. munol., 30: 652–660, 2000. Science (Wash. DC), 265: 1701–1706, 1994. 10. Hsu, H., Shu, H. B., Pan, M. G., and Goeddel, D. V. TRADD-TRAF2 30. Leaman, D. W., Leung, S., Li, X., and Stark, G. R. Regulation of and TRADD-FADD interactions define two distinct TNF receptor 1 signal STAT-dependent pathways by growth factors and . FASEB J., transduction pathways. Cell, 84: 299–308, 1996. 10: 1578–1586, 1996. 11. Wang, C. Y., Mayo, M. W., and Korneluk, R. G. NF-␬B antiapoptosis: 31. Darnell, J. E. J., Kerr, I. M., and Stark G. R. Jak-STAT pathways and induction of TRAF 1 and TRAF 2 and c-IAP 1 and c-IAP 2 to suppress transcriptional activation in response to IFNs and other extracellular sig- caspase-8 activation. Science (Wash. DC), 281: 1680–1683, 1998. naling proteins. Science (Wash. DC), 264: 1415–1421, 1994. 12. Cheng, G., and Baltimore, D. TANK, a co-inducer with TRAF2 of TNF- 32. O’Shea, J. J. Jaks, STATs, cytokine signaling and immunoregulation: and CD40L-mediated NF-␬B activation. Genes Dev., 10: 963–973, 1996. are we there yet? Immunity, 7: 1–11, 1997.