Genetic deletion of the tumor necrosis factor receptor p60 or p80 abrogates ligand-mediated activation of nuclear factor-kappa B and of mitogen-activated protein kinases in macrophages.

Tumor necrosis factor (TNF) is a pleiotropic cytokine known to regulate cell growth, viral replication, inflammation, immune system functioning, angiogenesis, and tumorigenesis. These effects are mediated through two different receptors, TNFR1 and TNFR2 (also called p60 and p80, respectively), with p60 receptor being expressed on all cell types and p80 receptor only on cells of the immune system and on endothelial cells. Although the role of p60 receptor in TNF signaling is well established, the role of p80 is less clear. In this report, by using macrophages derived from wild-type mice (having both receptors) and mice in which the gene for either p60 (p60(-/-)), or p80 (p80(-/-)), or both (p60(-/-) p80(-/-)) receptor have been deleted, we have redefined the role of these receptors in TNF-induced activation of nuclear factor (NF)-kappa B and of mitogen-activated protein kinases. TNF activated NF-kappa B in a dose- and time-dependent manner in wild-type macrophages but not in p60(-/-), p80(-/-), or p60(-/-) p80(-/-) macrophages. These results correlated with the I kappa B alpha degradation needed for NF-kappa B activation. We also found that TNF activated c-Jun N-terminal protein kinase in a dose- and time-dependent manner in wild-type macrophages but not in p60(-/-), p80(-/-), or p60(-/-) p80(-/-) macrophages. TNF activated p38 MAPK and p44/p42 MAPK in wild-type but not in p60(-/-), p80(-/-), or p60(-/-) p80(-/-) macrophages. TNF induced the proliferation of wild-type macrophages, but for p60(-/-) and p80(-/-) macrophages proliferation was lower, and in p60(-/-) p80(-/-) it was absent. Overall, our studies suggest that both types of TNF receptors are needed in macrophages for optimum TNF cell signaling.

Tumor necrosis factor (TNF 1 ), a pleiotropic cytokine initially shown to be produced mainly by macrophages, is now known to be produced by many cell types (1). Depending on the cells, TNF induces proliferation, apoptosis, or differentiation; activates various kinases of the mitogen-activated protein kinase (MAPK) family; and induces various transcription factors, including NF-B and AP-1. TNF exerts these biological effects by interacting with two distinct receptors, TNFR1 (p60TNFR or CD 120a) (2, 3) and TNFR2 (p80TNFR or CD 120b) (4). The p60 receptors are expressed in virtually all mammalian cell types, whereas the p80 receptor is expressed only on cells of the immune system and endothelial cells (5,6). Each receptor consists of a ligand-binding extracellular domain, which shares 28% homology with its counterpart, a transmembrane region, and a cytoplasmic domain lacking any distinct homology to its counterpart (2)(3)(4)(5)(6). Through its cytoplasmic death domain (DD), the p60 receptor sequentially recruits TNFR1-associated death domain protein (TRADD), TNF receptor-associated factor (TRAF)-2, receptor-interacting protein, and IB␣ kinase, leading to NF-B activation (7)(8)(9)(10). TRADD can also sequentially recruit Fas-associated death domain protein (FADD) and FADD-like interleukin-1-converting enzyme (ICE), leading to apoptosis (11,12). The recruitment of TRAF2 is also known to be required for activation of JNK, p42/p44 MAPK, and p38 MAPK (13). The p80 receptor, although lacking the death domain, is known to bind TRAF2 through another adapter protein, TRAF1 (14).
The reasons why there are two different TNF receptors and which signals are mediated through the p60 receptor as compared with the p80 receptor are widely debated. Most of the information on the role of each receptor in TNF signaling is based on either receptor overexpression (15,16), the use of receptor-blocking antibodies (17,18), receptor agonist antibodies (19,20), or TNF muteins that bind preferentially to one of the receptors (21,22). These studies have revealed that although most TNF signals are transduced by the p60 receptor, the role of p80 remains unclear. All the model systems used in the above studies were artificial and do not delineate or mimic actual ligand-mediated signaling. There is only one report utilizing mouse embryonic fibroblasts from TNFR knockout animals, which showed that TNFR1 and TNFR2 differentially activate p44/p42 MAPK (23). Human fibroblasts, however, express only the p60 and not the p80 form of the TNF receptors (24). Whereas, macrophages express both forms of the receptor (4,5). Therefore, to investigate TNF signaling mediated through each receptor, we generated macrophage cell lines from wild-type mice and from mice with gene knockouts for the p60 receptor or the p80 receptor or both. Ligand-induced activation of NF-B, JNK, p44/p42 MAPK, and p38 MAPK were examined. Our results suggest that genetic deletion of the either of the TNF receptors abolishes TNF-induced activation of NF-B, JNK, p44/p42 MAPK, p38 MAPK, and cell proliferation.
Cell Lines and Culture-Mice with genetic deletion of p60, p80, and both the TNF receptors have been described (25,26). These mice in which p80 and p60p80 are deleted were obtained from Genentech Inc. (South San Francisco, CA), and that with p60TNFR-deficient were obtained from Jackson Laboratories (Bar Harbor, ME). Immortalized macrophage cell lines were established by infecting the bone marrow of wild-type C57BL/6J mice and TNFR knockout homozygous mice (p60 Ϫ/Ϫ , p80 Ϫ/Ϫ , and p60 Ϫ/Ϫ p80 Ϫ/Ϫ ) with the murine recombinant J2 retrovirus containing the v-myc and v-raf oncogenes as previously described (27,28). Density-gradient-purified bone marrow cells were suspended in supernatant of the J2 virus-producing CRE/J2 cell line (provided by Dr. G. Luca Gusella, NCI, National Institutes of Health) containing Polybrene and granulocyte macrophage-colony stimulating factor (GM-CSF; R&D Systems, Minneapolis, MN). After 7 days of culture, the medium was replaced with complete RPMI 1640 and GM-CSF. The cells were weaned of GM-CSF, whereupon transformants emerged as GM-CSF-independent cell lines. The immortalized cell lines were subsequently cloned by limiting dilution. The lines generated were phenotyped by flow cytometric analysis and were found to be Mac-1 ϩ , F4/80 ϩ , ThB Ϫ , CD4 Ϫ , CD8 Ϫ , and CD40 ϩ (26).
Flow Cytometric Analysis of TNFR Expression-For analysis of TNFR expression, macrophages were harvested by gentle scraping, centrifuged, and resuspended in Dulbecco's phosphate-buffered saline containing 10% fetal bovine serum and 0.1% sodium azide. A three-step labeling procedure was performed using antibodies and reagents purchased from BD-PharMingen (San Diego, CA). The cells were preincubated for 10 min at 4°C with anti-mouse CD16/32 (Fc Block), and then monoclonal hamster IgG anti-mouse TNFR antibodies were added. Following a 1-h incubation at 4°C, the cells were washed and incubated an additional 1 h in a mixture of two mouse anti-hamster IgG monoclonal antibodies. The cells were washed and subjected to a final 1-h incubation at 4°C with phycoerythrin-conjugated streptavidin. The cells were analyzed using a FACS Vantage flow cytometer and CellQuest acquisition and analysis programs (Becton Dickinson, San Jose, CA).
Electrophoretic Mobility Shift Assay-NF-B activation was analyzed by EMSA as described previously (29). In brief, 8-g nuclear extracts prepared from TNF-treated or untreated cells were incubated with 32 P-end-labeled 45-mer double-stranded NF-B oligonucleotide from human immunodeficiency virus-1 long terminal repeat (5Ј-TTGTTACAAGGGACTTTCCGCTGGGGACTTTCCAGGGAGGCGTG-G-3Ј; the underlined sequence indicates binding sites) for 15 min at 37°C, and the DNA⅐protein complex resolved in a 6.6% native polyacrylamide gel. The specificity of binding was examined by competition with unlabeled 100-fold excess oligonucleotide and with mutant oligonucleotide. The composition and specificity of binding was also determined by supershift of the DNA⅐protein complex using specific and irrelevant antibodies. For the supershift experiment, the antibodytreated samples of NF-B were resolved on a 5.0% native gel. The radioactive bands from the dried gels were visualized and quantitated by PhosphorImager (Molecular Dynamics, Sunnyvale, CA) using ImageQuaNT software.
c-Jun Kinase Assay-The c-Jun kinase assay was performed by a modified method as described earlier (30). Briefly, 100 g of whole-cell extract was treated with anti-JNK1 antibodies, and the immune complexes so formed were precipitated with protein A/G-Sepharose beads (Pierce Chemical, Rockford, IL). The kinase assay was performed using washed beads as a source of enzyme and glutathione S-transferase-Jun- Genomic DNA from each of the four cell lines was subjected to PCR analysis with primer pairs that amplify regions of neo, TNFR1, and TNFR2 genes. PCR products were electrophoresed in 1.5% agarose gels containing ethidium bromide. B, flow cytometric analysis of TNFR expression on immortalized macrophage cell lines. One million cells of each of the four cell lines were harvested and antibody-labeled for the analysis of expression of TNFR1 (solid lines) and TNFR2 (dotted lines). Labeling was performed via a three-step stain consisting of sequential incubations with hamster monoclonal anti-TNFR antibodies, followed by mouse anti-hamster IgG, followed by phycoerythrin-conjugated streptavidin. Negative controls, shown as shaded histograms, consisted of cells labeled with second-step antibodies, alone, followed by phycoerythrin-conjugated streptavidin. The histograms shown depict analysis of 10,000 cells and are representative of three separate analyses.
(1-79) as substrate (2 g/sample) in the presence of 10 Ci of [ 32 P]ATP per sample. The kinase reaction was carried out by incubating the above mixture at 30°C in kinase assay buffer for 15 min. The reaction was stopped by adding SDS sample buffer, followed by boiling. Finally, protein was resolved on a10% reducing gel. The radioactive bands of the dried gel were visualized and quantitated by PhosphorImager as described above.
Cell Proliferation Assay-Three thousand cells per 100 l of medium were taken in triplicate wells of 96-well plates and cultured in the presence of different concentrations of mTNF. Six hours before completion of the experiment, cells were pulse-treated with 1 mCi of tritiated thymidine. The proliferation of cells was determined at 72 h by examining the uptake of tritiated thymidine using a Matrix-9600 ␤-counter (Packard Instruments, Downers Grove, IL).

RESULTS
To understand the function of each type of TNF receptor, we isolated macrophage cell lines from mice in which the gene for either of the receptors or both the receptors were deleted. Because human TNF does not bind to the murine p80 receptor (31), we used murine TNF to activate the receptors.
Characterization of Wild-type and Receptor-deficient Cell Lines-To ensure that the cell lines derived from TNF receptor knockout mice are deficient in respective receptors, we performed PCR and FACS analysis. Genomic DNA from representative wild-type and p60-null (p60 Ϫ/Ϫ ), p80-null (p80 Ϫ/Ϫ ), and p60p80-null (p60 Ϫ/Ϫ p80 Ϫ/Ϫ ) cells was analyzed using PCR primers directed against the genes for TNFR1, TNFR2, or the neomycin insert used to disrupt the genes for the two receptors. As shown in Fig. 1A, wild-type macrophages yielded two bands corresponding to TNFR1 and TNFR2 receptors but not the neomycin insert, whereas p60 Ϫ/Ϫ and p80 Ϫ/Ϫ cells gave PCR products for TNFR2 and TNFR1, respectively, plus the neomycin insert. In double knockout cells, only the neomycin insert band was seen.
The PCR data were supplemented with FACS analysis of the cell surface expression of receptors using receptor-specific antibodies. Results in Fig. 1B indicate that wild-type cells express both TNF receptors, although the expression of p60 was low. Neither of these receptors was detected in double-knockout cells. In p60 Ϫ/Ϫ cells only p80 receptors were expressed and vice versa in the case of p80 Ϫ/Ϫ cells (Fig. 1B). Thus, cell lines used in this investigation were authentic and had the expected deletion of the gene product.
Both TNF Receptors Are Needed for Maximum Activation of NF-B-Activation of NFB is one of the earliest events induced by TNF in most mammalian cells. Whether TNF-induced NF-B activation also occurs through individual TNF receptors is not known. To understand the specific role of TNF receptors in ligand-induced NF-B activation, we treated wild-type macrophage cell line and its TNF receptor-deficient variants with mTNF, prepared the nuclear extracts, and analyzed them by EMSA for NF-B. As shown in Fig. 2A, dose-dependent activation of NF-B occurred in wild-type cells. No NF-B was activated in double-knockout cells, and very little was activated in single-knockout cells at higher concentrations of TNF. The activation was higher in p80 Ϫ/Ϫ than that in p60 Ϫ/Ϫ cells.
EMSA of nuclear extracts prepared from TNF-treated cells showed that either anti-p50 or anti-p65 antibodies abrogated/ supershifted the NF-B⅐DNA complex, whereas preimmune sera (PIS) or irrelevant anti-cyclin D1 antibodies had no effect (Fig. 2B). Thus, NF-B induced by TNF in macrophages derived from C57BL/6J mice contained both the p50 and p65 (RelA) subunits. The specificity of the TNF-induced NF-B⅐DNA complex was further confirmed by demonstrating that the binding was disrupted in the presence of a 100-fold excess of unlabeled B-oligonucleotide (Fig. 2B, Cold oligo) but not by mutant oligonucleotide (Mutant oligo).
To determine the kinetics of NF-B activation, all four cell lines were treated with 0.1 nM mTNF for various times, and the nuclear and the cytoplasmic extracts were prepared and analyzed by EMSA for NF-B and by Western blot for IB␣. Maximum activation of NF-B (6.2-fold) was obtained in wild-type cells after 20 min of ligand treatment (Fig. 3A). In p80 Ϫ/Ϫ cells, NF-B activation was extremely poor, whereas no activation was seen either in cells from p60 Ϫ/Ϫ or in cells from double knockout mice (Fig. 3A).
Because TNF-induced degradation of IB␣ is a critical step in the pathway leading to the activation of NF-B, we examined whether time-dependent NF-B activation is correlated with IB␣ degradation. As shown in Fig. 3B, the degradation of IB␣ in wild-type cells was maximum at 20 min, matching kinetics of NF-B activation (Fig. 3A). As expected, no degradation of IB␣ was noticed in p60 Ϫ/Ϫ or p60 Ϫ/Ϫ p80 Ϫ/Ϫ cells, whereas a minor degradation was noted in p80 Ϫ/Ϫ cells. Thus, none of the receptors was sufficient by itself for significant activation of ligand-induced NF-B in macrophages.
Both TNF Receptors Are Needed for Maximum Activation of JNK-JNK activation is another early event transduced by TNF (13). Whether JNK is activated by the individual TNF receptor in response to the ligand is not well understood. To examine that, we first treated wild-type and receptor-deficient cells with different concentrations of mTNF for 15 min and performed a kinase assay. In wild-type cells dose-dependent JNK activation had occurred; the activation was highest (9.6fold) at 10 nM TNF (Fig. 4A). In contrast, TNF failed to activate JNK in macrophages from any of the knockout mice (Fig. 4A). The lower panels in Fig. 4A represent loading controls and indicate that JNK activation in wild-type cells was not due to the expression of JNK1 protein.
To examine the kinetics of JNK activation by TNF, cells were

. Activation of JNK in wildtype and TNF-receptor-knockout macrophage cells by TNF.
A, dose response for JNK activation. Two million wild-type and receptor knockout cells per milliliter were treated with 0 -10 nM TNF for 15 min. One hundred micrograms of whole-cell protein was reacted with JNK1 antibodies and then immunoprecipitated with protein A/G-Sepharose. The beads were washed and subjected to kinase assay as described under "Experiment Procedures" (upper panel). Forty micrograms of the same protein extracts was probed with JNK1 antibodies (lower panel). B, time course of JNK activation. Two million wild-type and receptor-knockout cells per milliliter were treated with 1 nM TNF for 0 -30 min. One hundred micrograms of whole-cell protein was treated with JNK1 antibodies and then immunoprecipitated with protein A/G-Sepharose. The beads were washed and subjected to kinase assay as described under "Experimental Procedures" (upper panel). Forty micrograms of the same protein extracts was probed with JNK1 antibodies (lower panel).

FIG. 5. Activation of p38 MAPK in wild-type and TNF receptor knockout macrophage cells by TNF.
Two million wild-type and receptor knockout cells per milliliter were treated with 1 nM TNF for 0 -60 min. Sixty micrograms of whole-cell protein was resolved on 10% SDS-PAGE gel, electrotransferred to a nitrocellulose membrane, and probed with phospho-p38 MAPK antibodies as described under "Experimental Procedures" (upper panels). The same blot was reprobed with p38 MAPK antibodies (lower panels).
FIG. 6. Activation of p44/p42 MAPK in wild-type and TNF receptor knockout macrophage cells by TNF. Two million wild-type and receptor knockout cells per milliliter were treated with 1 nM TNF for 0 -60 min. Sixty micrograms of whole-cell protein was resolved on 10% SDS-PAGE gel, electrotransferred to a nitrocellulose membrane, and probed with phospho-p44/p42 MAPK antibodies as described under "Experimental Procedures" (upper panels). The same blot was reprobed with ERK2 antibody (lower panels). blot analysis using phospho (Tyr/Thr)-specific p38 MAPK antibodies. These results shows that TNF activated p38 MAPK only in wild-type cells (Fig. 5). The activation was time-dependent with a maximum at 10 min. These results suggest that TNF does not activate p38 MAPK through individual receptors.
Both TNF Receptors Are Needed for Maximum Activation of p44/p42 MAPK-Through the Ras/Raf/MEK cascade, TNF can activate ERK1/2 (32). Which TNF receptor mediates the activation of ERK-MAPK pathway is unclear. To examine ERK1/2, whole-cell extracts from wild-type and receptor knockout cell lines treated with 1 nM TNF for different times were analyzed by Western blot using phospho-p44/p42 MAPK antibodies. In wild-type cells, p42 MAPK was activated within 5 min of the start of TNF treatment, whereas p44 MAPK was minimally activated (Fig. 6). The duration of ERK-MAPK activation was longer than that noted with p38 MAPK, because much of residual activity was retained at 60 min. A transient activation in the initial 5 min appeared in p80 Ϫ/Ϫ cells but not in the other two knockout cells (p60 Ϫ/Ϫ and p60 Ϫ/Ϫ p80 Ϫ/Ϫ ). Thus these results demonstrate that both TNF receptors are needed for optimum ligand-induced activation of p44/p42 MAPK.
Both TNF Receptors Are Needed for Maximum Ligand-induced Proliferation of Cells-TNF induces proliferation in many normal and tumor cells. To dissect the function of both receptors in TNF-induced proliferation, all four cell lines were cultured for 72 h in the presence of different concentrations of TNF, and cellular proliferation was assayed by tritiated thymidine uptake. Results shown in Fig. 7 indicated that TNF induced more than a 4-fold proliferation in wild-type cells, approximately a 2-fold proliferation in p80 Ϫ/Ϫ cells, a 1.5-fold increase in p60 Ϫ/Ϫ cells, and no increase in double knockout cells; the increases were dose-dependent (Fig. 7). These results suggests that both TNF receptors participate in TNF-induced cellular proliferation. DISCUSSION In this report, we used macrophage cell lines derived from mice in which the genes for either one or both of the TNF receptors were deleted. The deletion of either of the TNF receptors abolished ligand-induced signal transduction and abrogated mTNF-induced NF-B activation, IB␣ degradation, and activation of JNK, p38 MAPK, and p44/p42 MAPK. The deletion of the genes for either the p60 or the p80 receptor abolished essentially all cellular responses to TNF, the effect being more pronounced in cells with p60 Ϫ/Ϫ than in cells with p80 Ϫ/Ϫ . For optimum proliferation of macrophages to be induced by exogenous TNF, both receptors were required.
Although there are several reports on the role of the two receptors in TNF signaling using various approaches, our report is the first to investigate the role of each TNF receptor side by side using stable cell lines derived from receptor knockout animals. We have demonstrated that the deletion of either of the TNF receptors eliminates most of the TNF-induced NF-B activation in macrophages. Previously, it was shown that TNFinduced NF-B activation in T lymphocytes and dendritic cells is abolished in p60 receptor-deficient mice (25,33). Another study showed that mouse embryo fibroblasts derived from p60 Ϫ/Ϫ mice fail to activate NF-B in response to TNF (23). NF-B was found to be constitutively active in regenerating liver from wild-type animals but not in the liver from p60 Ϫ/Ϫ animals (34), again suggesting the importance of the p60 receptor in this response.
Interestingly, however, our results indicate that macrophages derived from p80 Ϫ/Ϫ animals also show suppression of NF-B activation in response to TNF, thus suggesting the critical role of this receptor. The role of p80 receptor in TNFinduced NF-B activation is highly controversial. Although most reports indicate that the p80 receptor alone is insufficient to activate NF-B (17,22,23), we have shown previously that overexpression of p80 can mediate ligand-induced NF-B activation (16). Our results are in agreement with a recent report by Abu-Amer et al. (35), who showed that marrow-derived macrophages from p80 Ϫ/Ϫ animals had significantly suppressed TNF-induced NF-B activation.
Our observation that the deletion of either of the receptors abolished TNF-induced NF-B activation, suggests that the two TNF receptors may function synergistically in activating NF-B. This conclusion is consistent with several previous reports from our laboratory and others (16, 22, 36 -38), which showed that one receptor may enhance the effect of the other. For instance NF-B-regulated genes such as ICAM-1, VCAM-1, and iNOS require the activation of both receptors (39 -41). How one receptor can potentiate or synergize with another receptor is not clear. This potentiation may be mediated at the intracellular level through TRAF2, which has been shown to bind to both the receptors (36 -42). Although there are reports that indicate that dominant-negative TRAF2 can suppress TNF-induced NF-B activation (43), others indicate that cells from TRAF2-deficient animals can maintain the TNF-induced NF-B activation (44 -45), whereas receptor-interacting protein mediates the TNF-induced NF-B signal (46). Thus it is possible that the collaboration between two receptors occurs through still other unknown signaling elements.
It is not known whether the type of collaboration between the two receptors specific for macrophages as noted here also occurs in other cell types. Most epithelial cells express primarily FIG. 7. Proliferation of wild-type and TNF receptor knockout macrophage cells induced by TNF. Three thousand wild-type and receptor knockout cells in 0.1 ml were exposed to 0, 0.1, 0.3, 1, 3, 10 nM TNF for 72 h and pulsed with 1 mCi of tritiated thymidine for 6-h. Thymidine uptake was then determined in a ␤ counter as described under "Experimental Procedures." p60 receptor, and yet, TNF can activate NF-B in these cells (4,5,16,22). Most inflammatory cells, on the other hand, express both TNF receptors (4,5). It is possible that the p80 receptor of inflammatory cells contributes to the p60 receptor-mediated NF-B activation, perhaps explaining why in general there is a higher level of TNF-induced NF-B activation in inflammatory cells, such as macrophages and T cells, than in epithelial cells (47). Because NF-B is a pro-inflammatory transcription factor, its enhanced levels activated through both the TNF receptors may contribute to the overall inflammatory response.
Our results indicate that, like NF-B activation, JNK activation also requires the presence of both the TNF receptors. Elimination of either receptor suppressed TNF-induced JNK activation. There is no previous report on the contribution of each of the receptors to TNF-induced JNK activation. TNF can activate JNK in epithelial cells that express only the p60 receptor, and so the role of the p80 receptor has been unclear. We have previously shown that overexpression of the p80 receptor also causes ligand-induced JNK activation. Our results indicate that in macrophages both receptors are needed for JNK activation. It is very likely that this collaboration is mediated through TRAF2, which binds to both the receptors; unlike the mechanism for NF-B, the deletion of TRAF2 abolishes the TNF-induced JNK activation (44,45).
There has been no other report about the role of either receptor in TNF-induced p38 MAPK activation. Our results indicate that TNF activates p38 MAPK in a time-dependent manner in wild-type cells, but no trace of p38 activation was noted in macrophages derived from receptor-deficient animals. Our results clearly demonstrate that both receptors are needed in macrophages for p38 activation. These results parallel those for p44/p42 MAPK (see Fig. 6).
Although most cells undergo apoptosis in response to TNF, some cell types such as normal fibroblasts, B cells, and macrophages proliferate on exposure to TNF (24, 48 -50). The TNFinduced fibroblast proliferation derived from p60 Ϫ/Ϫ animals is abrogated (23), thus suggesting the importance of p60 receptor. Because fibroblasts express only the p60 receptor, the contribution of the p80 receptor to TNF-induced proliferation is unclear. Our results indicated that mTNF induces proliferation of macrophages derived from control animals but proliferation of either p60 Ϫ/Ϫ or p80 Ϫ/Ϫ animals was significantly suppressed. No proliferation was seen in cells from which both receptors had been deleted. These results suggest that both receptors are required for TNF-induced proliferation of macrophages. This may also explain why levels of proliferation in macrophages by TNF was higher (greater than 4-fold) than that noted in fibroblasts (less than 2-fold (23)). Overall, our results clearly demonstrate that, in macrophages, both TNF receptors are needed for optimum activation of NF-B, JNK, p38 MAPK, p44/p42 and MAPK and for cellular proliferation.