Overexpression of the p80 TNF Receptor Leads to TNF-Dependent Apoptosis, Nuclear Factor-κB Activation, and c-Jun Kinase Activation¹

TNF, a highly pleiotropic cytokine, mediates all its effects through two distinct receptors, referred to as p60 and p80. The existence of a similar extracellular domain (∼29% homology in amino acid sequences) but a dissimilar intracellular domain (ICD)³ has led to the proposal that these two TNF receptors must transduce distinct signals (for references, see 1 . This hypothesis is further strengthened by observations that the ICD of the p60 receptor contains an amino acid sequence of approximately 80 amino acids, referred to as the death domain (DD; due to its involvement in cell death), whereas the p80 receptor does not (2). Moreover, it has been shown that the ICD of the p60 receptor recruits several signaling proteins distinct from that of the p80 receptor (for references, see 3 , including TRADD, FADD, FLICE, RIP, TRAF-2, ICH-1, MADD, TRIP, I-FLICE, sentrin, A-20, and I-TRAF. In contrast, factors associated with N-sphingomyelinase activation, PIP5K, TNF receptor-associated protein-1, and TNF receptor-associated protein-2 bind to the juxtamembrane domain. The ICD of p80 without the DD recruits TRAF-1, TRAF-2, IAP-1, IAP-2, and NF-κB-inducing kinase (3, 4). Despite distinct ICD of the two receptors, TRAF-2 and IAP-1, which bind to the p80 receptor, also bind through secondary interaction to the p60 receptor, but most of the p60 receptor binding proteins do not bind to the p80 receptor. In addition to these proteins, two distinct putative serine/threonine protein kinases have been shown to be recruited by the p60 and p80 receptors (5, 6). The roles of these proteins in TNF signaling are not entirely clear, but it appears that different proteins may mediate different TNF-induced signals. Even though TRADD, FADD, FLICE, RIP, and ICH-1 have been implicated in cell death (7, 8, 9, 10, 11, 12, 13), and RIP, TRADD, and TRAF-2 have been implicated in TNF-induced activation of NF-κB and c-Jun kinase (14, 15, 16, 17), the mechanisms are unknown.

One of the most fundamental unanswered questions in TNF biology is thus what types of signals are mediated through the p80 receptor compared with the p60 receptor. In general, it has been shown that most of the TNF signals are mediated through the p60 receptor. The p80 receptor has been implicated in ligand passing, in cytotoxicity, and in potentiation of p60 receptor-mediated cytotoxicity (18, 19, 20). Recent observations using receptor-specific TNF muteins have indicated that the cytotoxicity and NF-κB activation are mediated entirely through the endogenous p60 receptor (21).

In the present report we compared signaling through the endogenous receptor with that through the recombinant p80 receptor. TNF muteins that bind exclusively to either the p60 or the p80 receptor (22) were employed for these studies. In p60 receptor-specific mutein, residue 32 was altered from arginine to tryptophan, and residue 86 was altered from serine to threonine, whereas in p80 receptor-specific mutein, the carboxyl-terminus at position 143 was mutated from aspartic acid to asparagine and at position 145 from alanine to arginine. The p80 receptor-binding mutein induced apoptosis, activated NF-κB, and induced c-Jun kinase in cells overexpressing the recombinant receptor but not in cells expressing endogenous receptor. Thus, our results demonstrate that even though the ICD of the p80 receptor is distinct from that of the p60 receptor and recruits several distinct proteins, overexpressed p80 receptor can transduce signals similar to those of the endogenous p60 receptor but distinct from those of the endogenous p80 receptor.

RPMI 1640 medium, FCS, penicillin, streptomycin, and trypsin-EDTA were obtained from Life Technologies (Grand Island, NY). Tris, glycine, NaCl, and BSA were obtained from Sigma Chemical (St. Louis, MO). Hygromycin B was purchased from Calbiochem-Novabiochem International (La Jolla, CA). Bacteria-derived recombinant human TNF purified to homogeneity was provided by Genentech (South San Francisco, CA). ³²P-labeled γ-ATP with a sp. act. of 7000 Ci/mmol was obtained from Amersham Life Sciences (Arlington Heights, IL). Recombinant human TNF muteins altered by site-specific mutagenesis to bind to either the p60 or p80 receptor, expressed in bacteria, and purified to homogeneity were supplied by F. Hoffmann-La Roche (Basel, Switzerland). The isolation and characterization of these muteins has been described previously (22). Rabbit anti-JNK polyclonal Ab was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Mouse anti-poly(A)DP ribose polymerase (anti-PARP) Ab was purchased from PharMingen (San Diego, CA). Anti-mouse horseradish peroxidase-conjugated secondary Ab was purchased from Transduction Laboratories (Lexington, KY). Protein A/G-Sepharose beads were obtained from Pierce Chemical (Rockford, IL). GST-Jun_1–79 was expressed in Escherichia coli and purified essentially as previously described (5).

The human histiocytic cell line U-937, human myeloid ML-1a, and human rhabdomyosarcoma cell line KYM-1 were grown in RPMI 1640 medium supplemented with FBS (10%), penicillin (100 μg/ml), and streptomycin (100 μg/ml). Cells were seeded at a density of 1 × 10⁵/ml in T75 flasks (Falcon 3013, Becton Dickinson Labware, Lincoln Park, NJ) containing 20 ml of medium and were grown at 37°C in an atmosphere of 95% air and 5% CO₂. Cell cultures were split every third day. KYM-1 cells were trypsinized and seeded again every third day.

HeLa cells stably transfected with pCDM8 mammalian expression vector containing the full-length TNFRp80 cDNA, as previously described (18), were used for our studies. A plasmid carrying the hygromycin phosphotransferase gene was used as a selection marker and was cotransfected with the pCDM18 plasmid. Transfected cells were routinely grown in MEM containing 10% FBS, penicillin (100 μg/ml), streptomycin (100 μg/ml), and hygromycin (200 μg/ml).

Different clones were generated from the HeLa-p80 cells by two different methods. The first method involved sorting by FACS based on anti-p80 receptor Ab and anti-rabbit FITC-labeled secondary Ab. Depending on the level of p80 expression, four different populations of HeLa-p80 cells were obtained by this method. The second method involved isolation of different HeLa-p80 clones by limited dilution. Briefly, HeLa-p80 cells were diluted to 10 cells/ml and plated at 1 cell/100 μl/well into 96-well plates. The plates were incubated at 37°C for 2 wk, and thereafter, three different clones (no. 1–3) were isolated and further characterized.

The effect of p80 receptor transfection was also analyzed in another cell line, breast carcinoma MCF-7 (expresses only the p60 receptor), and by using a different vector. For this, full-length TNFRp80 cDNA (6) was cloned into pEC1214A(pTet), a modified tetracycline repressor system expression vector (a gift from Dr. Hong-Ji Xu, University of Texas M. D. Anderson Cancer Center, Houston, TX) containing a cloning site for the desired gene that is under the control of the tetracycline repressor, the constitutive expression of tetracycline repressor, and the neomycin resistance gene. MCF-7 cells were transfected with the pTet and pTet-p80 vector using Lipofectamine as described by the manufacturer (Life Technologies). A stable pool of MCF-7 cells harboring either pTet (MCF-7) or pTet-p80 (MCF-7 (p80)) were selected by growing the cells in complete medium containing tetracycline (1 μg/ml) and G418 (600 μg/ml). Induction of p80 receptor occurred after the withdrawal of tetracycline.

The cytotoxic effects of TNF and its muteins on adherent cells (KYM-1 and HeLa) were measured by the crystal violet dye uptake assay. Briefly, 10 × 10³ cells were plated in 0.1 ml of medium in 96-well flat-bottom plates. After an overnight incubation at 37°C, the medium was removed, and different concentrations of TNF were layered in 0.1 ml of fresh medium. After 72 h of incubation at 37°C, the medium was removed, and the viability of cells was monitored by crystal violet staining according to a procedure described previously (23). The cytotoxic effects of TNF on nonadherent cells (U-937) were determined by the amount of [³H]thymidine incorporated by the cells as described previously (24). Briefly, cells were plated at 5000/well in 0.1 ml of medium in 96-well flat-bottom Falcon plates. Different concentrations of TNF were added in an additional 0.1 ml of medium and were incubated at 37°C for 72 h. During the last 6 h before harvesting, [³H]thymidine (5 mCi/mmol; Amersham) was added to each well (0.5 μCi/well), and then cells were harvested with the aid of Filtermate 196 harvester (Packard Instruments, Meriden, CT). Radioactivity bound to the filter was measured in a liquid scintillation counter (model 1600 TR, Packard Instruments).

TNF-induced apoptosis was also examined by proteolytic cleavage of PARP (25). Briefly, cells (1 × 10⁶/ml) were treated with TNF or the muteins for either 2 h (U-937 cells) or 4 h (KYM-1, HeLa and HeLa-p80 cells) in the presence of 1 μg/ml of cycloheximide. After treatment, cell extracts were prepared by incubating the cells for 30 min on ice in 0.05 ml of buffer containing 20 mM HEPES (pH 7.4), 2 mM EDTA, 250 mM NaCl, 0.1% Nonidet P-40, 2 μg/ml leupeptin, 2 μg/ml aprotinin, 1 mM PMSF, 0.5 μg/ml benzamidine, and 1 mM DTT. The lysate was centrifuged, and the supernatant was collected. Cell extract protein (50 μg) was resolved on a 7.5% SDS-polyacrylamide gel, electrotransferred onto a nitrocellulose membrane, blotted first with mouse anti-PARP Ab and then with anti-mouse horseradish peroxidase-conjugated IgG as the secondary Ab, and then detected by chemiluminescence (ECL, Amersham). Apoptosis was represented by the cleavage of 116-kDa PARP into 85- and 41-kDa peptide products (25).

An EMSA to examine NF-κB activation was conducted as described previously (26, 27). Briefly, cells (2 × 10⁶/ml) were treated separately with different concentrations of TNF and its muteins at 37°C for 30 min. Nuclear extracts were then prepared as described previously (27). EMSA was performed by incubating 4 μg of nuclear extracts with 16 fmol of ³²P end-labeled 45-mer double-stranded NF-κB oligonucleotide from the HIV-1 long terminal repeat, 5′-TTGTTACAAGGGACTTTCCGCTGGGGACTTTCCAGGGAGGCGTGG-3′, in the presence of 2 to 5 μg of poly(dI-dC) in a binding buffer (25 mM HEPES (pH 7.9), 0.5 mM EDTA, 0.5 mM DTT, 1% Nonidet P-40, 5% glycerol, and 50 mM NaCl) for 15 min at 37°C. The DNA/protein complex formed was separated from free oligonucleotide on a 7.5% native polyacrylamide gel. A double-stranded mutant oligonucleotide was used as a control to examine the specificity of binding of NF-κB to the DNA. The specificity of binding was also checked by competition with the unlabeled oligonucleotide, by using mutant oligonucleotide, and by supershift of the band with Abs against NF-κB p50 and p65 proteins. The radioactive bands from dried gels were visualized on a PhosphorImager (Molecular Dynamics, Sunnyvale, CA) and quantitated using ImageQuant software.

Receptor binding assays were performed as described previously (28). TNF was labeled with Na¹²⁵I, using the Iodogen method, to an approximate sp. act. of 40 mCi/mg. Binding assays were performed in flexible 96-well plates. Cells (0.5 × 10⁶/0.1 ml) were incubated in a binding medium (RPMI 1640 containing 10% FBS) with ¹²⁵I-labeled TNF in the absence (total binding) or the presence of 100 nM unlabeled ligand (nonspecific binding) for 1 h at 4°C. Thereafter, cells were washed three times with ice-cold medium (PBS containing 0.1% BSA) at 4°C, and the cell-bound radioactivity was determined in a gamma counter (Cobra-AutoGamma, Packard Instrument). All determinations were performed in triplicate. Specific binding of TNF was calculated by subtraction of the amount (counts per minute) of nonspecific binding from the total binding.

To determine the levels of p60 and p80 forms of TNF receptors, affinity-purified Abs specific for each type of receptor were employed (21). Cells (0.5 × 10⁶/0.1 ml) were preincubated with the Ab (5 μg/ml) for 1 h at 4°C and then examined for specific binding of labeled TNF as described above. TNF muteins were also used to compete for TNF binding sites. Specific binding of TNF observed on competition with TNF (p60) or on cells pretreated with anti-p60 Ab would be due to the p80 receptor and vice versa.

The c-jun kinase assay was performed using a modified method as described previously (29). Briefly, after treatment of cells (3 × 10⁶/ml) with TNF for 15 min, cell extracts were prepared by lysing cells in buffer containing 20 mM HEPES (pH 7.4), 2 mM EDTA, 250 mM NaCl, 0.1% Nonidet P-40, 2 μg/ml leupeptin, 2 μg/ml aprotinin, 1 mM PMSF, 0.5 μg/ml benzamidine, and 1 mM DTT. Cell extracts (150–250 μg) were subjected to immunoprecipitation with 0.03 μg of anti-JNK Ab for 30 min at 4°C. Immune complexes were collected by incubation with protein A/G-Sepharose beads for 30 min at 4°C. The beads were collected by centrifugation and washed extensively with lysis buffer (four times, 400 μl each time) and kinase buffer (twice, 400 μl each time; 20 mM HEPES (pH 7.4), 1 mM DTT, and 25 mM NaCl). Kinase assays were performed for 15 min at 30°C with 2 μg GST-Jun_1–79 in 20 μl containing 20 mM HEPES (pH 7.4), 10 mM MgCl₂, 1 mM DTT, and 10 μCi [γ-³²P]ATP. Reactions were stopped with 15 μl of SDS-sample buffer, boiled for 5 min, and subjected to SDS-PAGE. GST-Jun_1–79 was visualized by staining with Coomassie Blue, and the dried gel was analyzed by a PhosphorImager (Molecular Dynamics) and quantitated by ImageQuant Software (Molecular Dynamics).

We examined the role of the p80 receptor in TNF-mediated apoptosis, NF-κB activation, and c-Jun kinase activation. To distinguish the signaling transduced by the p80 receptor from that of p60, we used genetically engineered TNF muteins that bind to either the p60 or the p80 receptor. The results were confirmed by the use of affinity-purified Abs specific to each receptor. We also compared the signaling of the recombinant p80 receptor with that of the endogenous p80 receptor.

We used U-937 and KYM-1 cells, which have been shown to express a high density of high affinity (∼5000 binding sites with a K_d of 0.1 nM) TNF receptors (30). Furthermore, the level of p80 receptors on these cells, determined using receptor-specific Abs, was approximately 3000 sites/cell (30, 31). In the present report we further examined the levels of p60 and p80 receptors by using TNF muteins specific to each receptor as a competitor. TNF (p60) inhibited TNF binding on U-937 cells by approximately 46%, and TNF (p80) inhibited it by 56% (Fig. 1, upper panel), suggesting that they expressed almost equal numbers of p60 and p80 receptors. Like U-937 cells, KYM-1 (Fig. 1, lower panel) cells expressed a high level of TNF receptors, and their binding was equally inhibited by p60- and p80-specific muteins. Compared with p60, U-937 cells expressed slightly lower p80, and KYM-1 cells expressed slightly higher p80 receptor levels.

To examine the role of recombinant p80 receptors, we used HeLa cells transfected with cDNA to overexpress p80 receptor as described previously (18). Radioreceptor results (Fig. 2 A) indicate that nontransfected HeLa cells expressed p60 receptor, as TNF binding could be blocked by anti-p60 Ab only (upper right) and inhibited by TNF (p60) mutein and not by TNF (p80) mutein (upper left). HeLa cells, when transfected with p80 receptor cDNA, bound 20 to 30 times more labeled TNF, and this binding could be blocked primarily by anti-p80 Abs (lower right) and inhibited by TNF (p80) mutein. Anti-p60 Ab or TNF (p60) had a minimal effect on binding of TNF to transfected cells. These results clearly indicate that transfected HeLa cells overexpressed p80 receptors. Scatchard analysis revealed that transfected HeLa cells express approximately 190,000 binding sites/cell (18).

The expression of endogenous and recombinant p80 receptors was also confirmed by Western blot analysis using anti-p80 receptor Abs (Fig. 2 B). U-937, ML-1a, and KYM-1 expressed variable amounts of endogenous p80 receptor protein, with the highest being expressed by KYM-1. As expected, nontransfected HeLa cells expressed no detectable p80 receptor, whereas after transfection these cells expressed a high level. A small amount of a low m.w. band was also found in transfected cells. The origin of this band is unclear at present, but may represent the degraded form of the p80 receptor.

To examine signaling through the endogenous p80 receptor, we examined the cytotoxic responses to TNF, TNF (p60), and TNF (p80). Both TNF and TNF (p60) were cytotoxic to U-937 cells (Fig. 3,A, upper panel) and to KYM-1 cells (Fig. 3 A, lower panel) in a dose-dependent manner. As little as 0.01 ng/ml of the ligand showed a significant effect. TNF (p80) mutein, however, had no effect on these cells, even when used at 200 ng/ml, suggesting that the p80 receptor is inert in these cells. As will be shown later, the lack of this effect could not be attributed to the possibility that TNF (p80) mutein was not biologically active.

We investigated the effect of overexpressed recombinant p80 receptors on TNF-mediated cytotoxicity. TNF had severalfold higher cytotoxic effects on p80-transfected cells than control, endogenous p60-expressing HeLa cells (Fig. 3,B, upper panel). These results imply that the p80 receptor, when overexpressed, could also mediate TNF-induced cell killing. To determine whether TNF could kill the cells without the assistance of the endogenous p60 receptor, we used TNF muteins. Interestingly, we found that p80 receptor transfection also enhanced the activity of the mutein specific to the p60 receptor (Fig. 3,B, middle panel), suggesting a synergistic effect in the signaling mediated through the p60 and p80 receptors. TNF (p80) mutein, which was inactive on cells expressing endogenous receptor, was also inactive on nontransfected HeLa cells (which express only the p60 receptor; Fig. 3 B, bottom panel). However, when examined on cells transfected to overexpress the p80 receptor, TNF (p80) mutein was as active as TNF, indicating that unlike the p60 receptor, the p80 receptor can mediate cytotoxicity only when overexpressed. These results also indicate that p80 not only can mediate cytotoxicity but can also enhance p60-mediated cytotoxicity.

Besides cytotoxicity, which required 72-h treatment to be expressed, we also examined TNF-induced cleavage of PARP, another early characteristic feature of apoptosis (25). After treatment of KYM-1 cells for 4 h, both TNF and the TNF (p60) mutein induced cleavage of PARP; the TNF (p80) mutein, however, had no effect (Fig. 4). Similarly, in U-937 cells the TNF (p60) mutein induced PARP cleavage within 2 h, whereas the TNF (p80) mutein did not (Fig. 4). These results clearly indicate that apoptosis is mediated through the endogenous p60 receptor but not the endogenous p80 receptor. We also found that the TNF (p80) mutein induced PARP cleavage in p80-transfected HeLa cells but not in cells expressing endogenous p60 receptor. Thus, as was the case for cytotoxicity, PARP cleavage was mediated through the recombinant p80 receptor but not through the endogenous receptor.

Both TNF and TNF (p60) activated NF-κB in U-937, ML-1a, and KYM-1 cells, but the TNF (p80) mutein had no effect on any of these cells (Fig. 5,A). Even TNF (p80) at 10 nM (100-fold higher) failed to activate NF-κB in KYM-1 cells (Fig. 5,B). Thus, endogenous p80 receptor appears to be inactive for induction of both apoptosis and NF-κB activation. Both TNF and TNF (p60) activated NF-κB within 30 min in a dose-dependent manner in HeLa cells, whereas TNF (p80) was inactive even when tested at a 100-fold excess (Fig. 5,C, upper panel). In contrast, on p80-transfected cells, TNF (p80) mutein was as active as TNF or TNF (p60) for this cellular response (Fig. 5,C, middle panel). To determine the specificity of the TNF-mediated response, we used anti-p80 Abs, which block endogenously expressed p80-mediated signaling. Serendipitously, we found that when these Abs to the p80 receptor were used to treat cells that overexpress recombinant p80 receptor, NF-κB was activated through the p80 receptor (Fig. 5 C, lower panel). Thus, the Ab acted as a TNF agonist. These results indicate that in contrast to endogenous p80 receptor, recombinant p80 receptor can also activate NF-κB.

Recently, it has been reported that overexpression of TRAF2, which associates with the p80 receptor, can activate c-Jun kinase (13, 15). Therefore, we examined the role of endogenous p80 receptor in mediating c-Jun kinase activation by using TNF muteins. In U-937 cells, both TNF and TNF (p60) activated c-Jun kinase, whereas TNF (p80) had no effect (Fig. 6). Both TNF and TNF (p60) activated JNK on control HeLa cells and transfected cells, whereas TNF (p80) activated JNK only on transfected cells (Fig. 6). These results demonstrate that the p80 receptor can mediate c-Jun kinase activation only when overexpressed. Our results to date suggest that endogenous p80 receptor lacks the ability to mediate apoptosis, NF-κB activation, and c-Jun kinase activation.

Why there are two distinct TNF receptors and whether there is any communication between the two receptors with respect to signaling have not been established. Since both receptors appeared to respond independently of each other in our system, we examined possible cross-talk between the two receptors. Given that our earlier results indicated that these receptors may function in a synergistic manner (Fig. 3,B, middle panel), we examined the effect of anti-p80 Ab on TNF-, TNF (p60)-, and TNF (p80)-mediated cytotoxicity in p80-transfected cells. The results show that anti-p80 blocked not only TNF- and TNF (p80)-mediated cytotoxicity, but also the TNF (p60)-mediated effect (Fig. 7, upper panel). Inhibition of TNF (p60)-induced cytotoxicity by anti-p80 occurred in a dose-dependent manner (Fig. 7, lower panel). The results with PARP cleavage also showed that the TNF (p60)-mediated effect was blocked by anti-p80 (Fig. 4). These results provide further evidence in HeLa-p80 cells that to mediate its effects, TNF (p60) mutein requires the p80 receptor in addition to the p60 receptor.

It is also intriguing to note that the same anti-p80 Ab that displayed an agonistic activity on HeLa-p80 cells for NF-κB activation (Fig. 5, lower panel) and for c-Jun kinase activation (Fig. 6, lower panel) showed antagonistic activity for cytotoxicity (Fig. 7).

All the results shown above were conducted with a single clone of p80-transfected HeLa cells. To ascertain that signaling through p80 receptor is not clone specific, we generated different clones of HeLa-p80 both by FACS sorting and by single cell cloning. HeLa-p80 cells were sorted into different populations based on the level of p80 receptor expression by FACS analysis. The different populations were designated A (untransfected), B (very low), C (low), D (medium), E (high), and F (unsorted). The level of p80 receptor expression on these cells was evaluated by labeled TNF binding (Fig. 8,A) and by Western blot (Fig. 8,B). In cells with medium and high levels of p80 expression (populations D and E), TNF (p80) mutein could induce cytotoxicity (Fig. 8,C), activation of NF-κB (Fig. 8,D), and activation of c-Jun kinase (Fig. 8,E). Thus, these data suggest that there is a direct correlation between increasing expression of the p80 receptor in HeLa cells and its ability to transduce various signals. Similarly, when different clones were isolated by single cell cloning, the level of expression of p80 (Fig. 9,A) again correlated with NF-κB activation (Fig. 9,B) and cytotoxicity (Fig. 9 C) induced by TNF (p80) mutein.

To ascertain that signaling through the recombinant p80 receptor is not restricted to HeLa cells, we also investigated the effect of p80 receptor transfection in MCF-7 cells, which normally express only p60 receptor. For this, MCF-7 cells were stably transfected with an inducible expression vector containing the p80 receptor (MCF-7 (p80)). The p80 receptor expression was induced as evaluated by Western blot (Fig. 10,A) and by labeled TNF binding (Fig. 10,B). While MCF-7 cells were completely resistant to increasing doses of the TNF (p80) mutein, expression of the p80 receptor in MCF-7 (p80) cells caused cytotoxicity in a dose-dependent manner when exposed to TNF (p80) mutein (Fig. 10,C). The expression of the p80 receptor in MCF-7 (p80) also led to activation of NF-κB by the TNF (p80) mutein (Fig. 10 D). These results suggest that overexpression of the p80 receptor leads to signaling not only in HeLa cells, but also in MCF-7 cells.

The aim of this study was to investigate the functional role of the p80 receptor in mediating TNF-induced signaling. To examine the role of endogenous p80 receptors, we used U937 and KYM-1 cells, which are known to express the maximum number of high affinity receptors among cell lines described to date. We also used HeLa cells and MCF-7 (which express only p60 receptor) that had been transfected with p80 receptor cDNA and thus engineered to express higher numbers of receptors than U-937 and KYM-1 cells. In addition, to distinguish the role of the p80 receptor from that of the p60 receptor in TNF-mediated signaling, we used TNF muteins, made by site-specific mutagenesis, to bind either the p60 (mutated at R32W and S86T) or the p80 (mutated at D143N and A145R) receptor. We studied effects on apoptosis and activation of NF-κB and of c-Jun kinase, all previously known to be mediated through the p60 receptor.

In the present report, we demonstrate that the endogenous p80 receptor is incapable of inducing these three signals. However, p80 receptor can, when overexpressed, mediate cytotoxicity, NF-κB activation, and induction of c-Jun kinase in a ligand-dependent manner. In addition, we found that overexpression of p80 receptor potentiates p60 receptor-mediated cytotoxicity in a synergistic manner, and p60 receptor-mediated cytotoxicity was blocked by Abs to the p80 receptor. The p80 Abs, which was antagonistic on endogenous receptor acted as an agonist on the cells overexpressing p80 receptor for NF-κB activation.

Although several reports indicate that the endogenous p60 receptor can mediate cytotoxicity, NF-κB activation, and induction of c-Jun kinase, very few indicate that endogenous p80 receptor can mediate similar responses (31, 32, 33, 34, 35, 36, 37). This has led to suggestions of an alternate role of p80 receptor including ligand passing to the p60 receptor, the receptor for transmembrane form of TNF, and the receptor required for immunomodulation by TNF (19, 32, 33). Only when overexpressed could the p80 receptor signal for all the cellular responses examined. Induction of apoptosis by overexpression of p80 is consistent with the findings of a previous report (18).

Why overexpression of p80 receptor is needed for its signaling, whereas the p60 receptor can signal without overexpression is not clear. It has been shown that receptor clustering is needed for cytokine signaling, so it is possible that receptor clustering is more efficient for the p60 receptor than for the p80 receptor. That the ICD of the p60 receptor contains a DD that itself has a strong tendency to self-associate, whereas the p80 receptor lacks this domain, may explain the differential behavior of the two receptors. This may also explain why overexpression of the p60 receptor leads to ligand-independent signaling, whereas overexpression of the p80 receptor, as noted in our studies, still requires activation by the ligand to signal.

The DD of the p60 receptor has been shown to recruit several proteins that are involved in TNF-mediated cytotoxicity, NF-κB activation, and induction of c-Jun kinase. How p80 receptor signals without the DD binding proteins is not known. TRAF-2 and cellular IAP-1 have been shown to bind to both receptors (38, 39, 40). The potential role of TRAF-2 in NF-κB activation and induction of JNK has been demonstrated (13, 14, 15, 16, 17). However, TRAF-2 plays no role in TNF-induced apoptosis. Thus, how p80 receptor can mediate cytotoxicity independent of the p60 receptor is not clear. Our results also showed that p60 receptor-mediated cytotoxicity is potentiated by p80 receptor (see Fig. 3,B) and that p60 receptor-mediated cytotoxicity is blocked by anti-p80 (Fig. 7). How p80 receptor potentiates the p60 receptor-mediated signaling is not clear. Our results, however, are in agreement with those of a recent report that showed enhancement of p60 receptor-mediated cytotoxicity by p80 receptor (20). The ability of p80 receptor to cooperate with the activation of p60 receptor was found to be due to TRAF-2 (20). Overexpression of TRAF-2 alone, however, does not induce cytotoxicity. The role of TRADD in p60 receptor-mediated cytotoxicity has been demonstrated (14). It is possible that the p80 receptor recruits proteins similar to TRADD that can also mediate cytotoxicity. Alternatively, it is also possible that both TRADD and TRAF2 are needed for cell killing by TNF. Ligand-induced formation of a heterocomplex between p60 and p80 has recently been demonstrated (41). In any case, our results provide evidence of cross-talk not only at the receptor level, but also for downstream signaling.

We showed that when U-937 and KYM-1 cells, which express endogenous p80 receptor, are stimulated with TNF (p80), they do not show activation of NF-κB. However, when the p80 receptor was overexpressed, the TNF (p80) mutein activated NF-κB. This is consistent with the report that overexpression and clustering of TRAF2 can by itself activate NF-κB (13). The p80-mediated activation of NF-κB seen in HeLa-p80 cells was perhaps due to such a clustering of the TRAF2 molecules. Besides TNF (p80), anti-p80 receptor Ab, which was antagonistic to induction of apoptosis (see Fig. 7), activated NF-κB in HeLa-p80 cells (Fig. 5 C, bottom panel). How an antagonistic Ab can become an agonist is not clear. This may be because differential receptor clustering is needed for different signals.

Our results show that p80 receptor can transduce similar signals as the p60 receptor but require a higher level of expression. It is possible that the normal level of p80 receptors present in the cells does signal, but the level is too low to be detectable. Alternatively, a normal level of p80 receptor expression is responsible for long term effects, but is too low to trigger short term effects. Overall, our results indicate that both p60 and p80 receptors can mediate identical signals, but the p60 receptor is more efficient than the p80 receptor. Our results also provide evidence for cross-talk between the two receptors.

We thank Dr. Bing Su for providing plasmid pGST-Jun_1–79.