Cooperative Signaling by Tumor Necrosis Factor Receptors CD120a (p55) and CD120b (p75) in the Expression of Nitric Oxide and Inducible Nitric Oxide Synthase by Mouse Macrophages*

Tumor necrosis factor-α (TNFα) is recognized by the cell-surface receptors CD120a (p55) and CD120b (p75). In the present study, we have investigated the role of these receptors in the expression of NO2 −, a stable metabolite of nitric oxide, and inducible nitric oxide synthase (iNOS) by mouse macrophages. Specific antibody-mediated aggregation of CD120a (p55) induced NO2 − accumulation in culture supernatants and iNOS mRNA expression in macrophage lysates, whereas cross-linking of CD120b (p75) had a minimal effect. In contrast, simultaneous cross-linking of both receptors led to a marked augmentation in NO2 − and iNOS mRNA expression. Antibody-mediated blockade of CD120a (p55) completely inhibited NO2 − expression in response to TNFα, whereas blockade of CD120b (p75) reduced NO2 − accumulation by ∼50%. Specific ligation of CD120a (p55) with either (i) human TNFα or (ii) by incubation with mouse TNFα following pretreatment of macrophages with blocking concentrations of anti-CD120b (p75) antibody resulted in a similar reduction in NO2 − production in response to TNFα. Quantification of iNOS mRNA, protein, and NO2 −expression during independent and co-ligation of CD120a (p55) and CD120b (p75) indicated that iNOS mRNA and protein expression was transient in nature when CD120a (p55) was cross-linked alone but was prolonged when both receptors were simultaneously cross-linked. In addition, cross-linking both receptors also led to a potentiation of NO2 − accumulation in culture supernatants that was more pronounced at later time points. These findings suggest that while cross-linking of CD120a (p55) is necessary and sufficient for iNOS mRNA and NO2 − expression, CD120b (p75) participates by (i) increasing the sensitivity of the cells to TNFα, probably by “passing” ligand to CD120a (p55), and (ii) initiating a signaling event that results in a more sustained induction of iNOS mRNA and protein and thereby augments the production of nitric oxide.

Tumor necrosis factor-␣ (TNF␣) is recognized by the cell-surface receptors CD120a (p55) and CD120b (p75). In the present study, we have investigated the role of these receptors in the expression of NO 2 ؊ , a stable metabolite of nitric oxide, and inducible nitric oxide synthase (iNOS) by mouse macrophages. Specific antibodymediated aggregation of CD120a (p55) induced NO 2 ؊ accumulation in culture supernatants and iNOS mRNA expression in macrophage lysates, whereas cross-linking of CD120b (p75) had a minimal effect. In contrast, simultaneous cross-linking of both receptors led to a marked augmentation in NO 2 ؊ and iNOS mRNA expression. Antibody-mediated blockade of CD120a (p55) completely inhibited NO 2 ؊ expression in response to TNF␣, whereas blockade of CD120b (p75) reduced NO 2 ؊ accumulation by ϳ50%. Specific ligation of CD120a (p55) with either (i) human TNF␣ or (ii) by incubation with mouse TNF␣ following pretreatment of macrophages with blocking concentrations of anti-CD120b (p75) antibody resulted in a similar reduction in NO 2 ؊ production in response to TNF␣. Quantification of iNOS mRNA, protein, and NO 2 ؊ expression during independent and coligation of CD120a (p55) and CD120b (p75) indicated that iNOS mRNA and protein expression was transient in nature when CD120a (p55) was cross-linked alone but was prolonged when both receptors were simultaneously cross-linked. In addition, cross-linking both receptors also led to a potentiation of NO 2 ؊ accumulation in culture supernatants that was more pronounced at later time points. These findings suggest that while cross-linking of CD120a (p55) is necessary and sufficient for iNOS mRNA and NO 2 ؊ expression, CD120b (p75) participates by (i) increasing the sensitivity of the cells to TNF␣, probably by "passing" ligand to CD120a (p55), and (ii) initiating a signaling event that results in a more sustained induction of iNOS mRNA and protein and thereby augments the production of nitric oxide.
Nitric oxide (NO ⅐ ) 1 is a major effector of macrophage-mediated cytocidal activity against both tumor cells and obligate and facultative intracellular parasites such as Toxoplasma gondii (1), Leishmania major (2,3), Mycobacterium leprae (4), and Listeria monocytogenes (5,6). The production of NO ⅐ is controlled by inducible nitric oxide synthase (iNOS) that catalyzes the oxidation of the guanidino nitrogen of L-arginine to form NO ⅐ and L-citrulline (7). However, in contrast to constitutive nitric oxide synthases of brain and endothelium (8 -10), macrophage iNOS expression results in NO ⅐ production for prolonged periods and at high levels of output. The expression of iNOS is stimulated by a variety of exogenous stimuli including lipopolysaccharide (11) which act to stimulate the transcription of the iNOS gene. However, in many situations it is the cytokines IFN␥ and TNF␣ produced during the host inflammatory and immune responses that act cooperatively to induce iNOS expression (12).
TNF␣ interacts with macrophages and other cell types through two distinct, although related, cell-surface receptors with molecular masses of 55 kDa (CD120a (p55)) and 75 kDa (CD120b (p75)) (13)(14)(15)(16). With the aid of monoclonal and polyclonal antibodies that can function both as mimics and antagonists of TNF␣ actions, it has become clear that most functions elicited by TNF␣ are mediated by cross-linking of CD120a (p55). These functions include cytotoxicity against LM cells (16), acute phase protein synthesis by hepatoma cells (17), and synthesis of the fibroblast progression-type growth factor insulin-like growth factor I by macrophages (IGF-I) (18). Antibodyinduced cross-linking of CD120b (p75) has also been shown to induce certain restricted functions such as thymocyte proliferation (16) and granulocyte/macrophage colony-stimulating factor secretion (19). However, an important function of this receptor appears to be the enhancement of functions initiated by cross-linking of CD120a (p55). Two concepts have been proposed to address the mechanism by which ligation of CD120b (p75) may assist the functions of CD120a (p55). First, by virtue of its higher affinity and more rapid association and dissocia-* This work was supported in part by U. S. Public Health Service Grants HL55549 and SCOR HL56556 from the National Institutes of Health. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18  tion rate constants than CD120a (p55), CD120b (p75) has been proposed to regulate the rate of interaction of TNF␣ with CD120a (p55) thereby enhancing signaling by CD120a (p55) (20,21). Second, cross-linking of CD120b (p75) may generate a signal that may or may not synergize with signals produced during ligation of CD120a (p55). Since both concepts are not mutually exclusive, cross-linking of CD120b (p75) may contribute to the functions elicited by ligation of CD120a (p55) although may not necessarily, by itself, stimulate a cellular response. In this study we have investigated the roles of CD120a (p55) and CD120b (p75) in the induction of NO 2 Ϫ (a stable oxidation product of NO ⅐ ) and iNOS mRNA expression by mouse macrophages and the mechanisms underlying their involvement. Our findings show that whereas cross-linking of CD120a (p55) is necessary and sufficient for iNOS and NO 2 Ϫ expression, additional cross-linking of CD120b (p75) and signaling by this receptor are required for the optimal and prolonged induction of iNOS and NO 2 Ϫ .

EXPERIMENTAL PROCEDURES
Materials-Dulbecco's modified Eagle's medium was obtained from Hazleton Laboratories, Denver, PA, and was supplemented immediately before use with 2 mM L-glutamine, 100 units/ml penicillin, and 100 g/ml streptomycin. Fetal bovine serum was purchased from Irvine Scientific, Santa Ana, CA. 24-Well tissue culture plates and 100-mm tissue culture dishes were purchased from Costar, Cambridge, MA.
[␣-32 P]deoxycytidine 5Ј-triphosphate (Ͼ3000 Ci/mmol) and [␥-32 P]ATP (Ͼ3000 Ci/mmol) were obtained from NEN Life Science Products. Mouse recombinant IFN␥ and TNF␣, rabbit anti-mouse CD120a (p55) and rabbit anti-mouse CD120b (p75) antibodies were generously donated by Genentech, Inc., San Francisco, CA. Human TNF␣ was purchased from Genzyme Diagnostics, Cambridge, MA. All other reagents were of the highest possible purity. C3H/HeJ mice were bred at the Biological Resource Center at National Jewish Medical and Research Center under specific pathogen-free conditions. Recombinant GST-c-Jun 1-79 was kindly provided by Dr. Gary Johnson, National Jewish Medical and Research Center, Denver, CO. Anti-mouse iNOS antibody was purchased from Alexis Biochemicals, San Diego, CA.
Isolation and Culture of Mouse Macrophages-Mouse bone marrowderived macrophages were obtained by culturing femoral and tibial bone marrow from C3H/HeJ mice in growth medium that comprised Dulbecco's modified Eagle's medium supplemented with 2 mM L-glutamine, 100 units/ml penicillin, 100 g/ml streptomycin, 10% (v/v) heatinactivated fetal calf serum, and 10% (v/v) L-cell conditioned medium as a source of macrophage colony-stimulating factor as previously reported (22,23). The bone marrow cells were dispensed at a density of 2.4 ϫ 10 5 nucleated cells/well in 24-well tissue culture plates for experiments in which NO 2 Ϫ expression was determined and in 100-mm diameter tissue culture dishes for RNA analyses. The cultures were maintained under an atmosphere of 10% CO 2 for 5 days to obtain confluent monolayers of macrophages. The cells were incubated with the various stimuli in growth medium.
Quantification of NO 2 Ϫ Accumulation-The accumulation of NO 2 Ϫ in culture supernatants was quantified in 96-well plates using the method described by Ding et al. (12). One hundred-l aliquots of the culture supernatants were dispensed in duplicate into 96-well plates and were mixed with 100 l of a solution composed of 1% (w/v) sulfanilamide, 0.1% (w/v) naphthylethylenediamine hydrochloride, and 2.5% (v/v) H 3 PO 4 . A standard curve consisting of 0.1-5.0 nmol of NaNO 2 per 100-l sample was prepared in growth medium. After incubation at ambient room temperature for 10 min, the absorbances of the wells were quantified at 550 nm in a Biotek Instruments enzyme-linked immunosorbent assay plate reader. The number of cells/well was determined by lysing the cell monolayers in "Zapoglobin" and quantifying the number of released nuclei with a model ZM Coulter counter. Concentrations of NO 2 Ϫ were interpolated from the NaNO 2 standard curve, corrected for the volume of the culture supernatant, and normalized to the number of cells/well. Results are presented as nanomoles of NO 2 Ϫ per 10 6 adherent cells per 24 h. Each experiment was conducted in triplicate and was conducted a minimum of three times. The results presented are the mean Ϯ S.D. of three experiments.
Analysis of iNOS mRNA and Protein Expression-The expression of iNOS mRNA was determined by Northern analysis. The extraction, purification, electrophoresis, and transfer of the RNA to nylon membranes was carried out as described (22,23). Briefly, macrophage mono-layers were extracted with 4 M guanidine isothiocyanate (24), and the RNA was purified by centrifugation through 5.7 M cesium chloride at 100,000 ϫ g for 18 h. Fifteen g of total RNA were electrophoresed under denaturing conditions through a 1.0% (w/v) agarose-formaldehyde gel and then transferred to Nytran membranes. The blots were hybridized with 5 ϫ 10 6 dpm of 32 P-labeled iNOS cDNA probe (8), and autoradiograms were prepared by exposure to Kodak XAR-5 film at Ϫ70°C. The mouse iNOS cDNA probe was kindly provided by Dr. Charles Lowenstein, Johns Hopkins University School of Medicine, Baltimore. For the measurement of iNOS protein, macrophages were lysed in 500 l of Nonidet P-40 lysis buffer (50 mM Tris/HCl buffer, pH 8.0, containing 137 mM NaCl, 10% glycerol, 1% (v/v) Nonidet P-40, 1 mM NaF, 10 g/ml leupeptin, 10 g/ml aprotinin, 2 mM Na 3 VO 4 , and 1 mM phenylmethylsulfonyl fluoride). Samples were then separated by SDS-PAGE through a 7.5% gel, electroblotted onto nitrocellulose membranes, and immunoblotted with anti-mouse iNOS antibody as described previously (26).

NO 2
Ϫ Production by IFN␥ and TNF␣-The production of NO 2 Ϫ by mouse bone marrow-derived macrophages was similar to that previously reported for elicited mouse peritoneal macrophages (12) in that both IFN␥ and TNF␣ were either unable or poorly capable of stimulating NO 2 Ϫ when presented alone but induced abundant levels of NO 2 Ϫ when presented together. As can be seen in Fig. 1, NO 2 Ϫ production was initially detected at a TNF␣ concentration of 0.1 ng/ml, whereas maximal production was achieved at concentrations of TNF␣ of 30 ng/ml or more.

Agonistic Effects of Anti-TNF Receptor Antibodies on NO 2
Ϫ Production-The independent roles of CD120a (p55) and CD120b (p75) in NO 2 Ϫ production were initially addressed by incubating macrophage monolayers with increasing concentrations of polyclonal antibodies directed against each receptor type in the presence and absence of IFN␥. As can be seen in 2, when macrophages were incubated with anti-CD120a (p55) antibody in the presence of IFN␥ (10 units/ml), a concentrationdependent increase in NO 2 Ϫ production was seen that peaked at between 0.5 and 1.0 g/ml antibody. Further increments in the concentration of the anti-CD120a (p55) antibody were found to result in a decline in NO 2 Ϫ production until at 10 g/ml NO 2 Ϫ production was not detected. By contrast, incubation of macrophages with anti-CD120b (p75) antibody either in the presence or absence of IFN␥ did not result in any significant production of NO 2 Ϫ (Fig. 2). However, when macrophages were simultaneously challenged with equivalent amounts of both anti-CD120a (p55) and anti-CD120b (p75), antibodies in the presence of IFN␥, an augmentation in NO 2 Ϫ production was observed compared to when each antibody was added alone (Fig. 2). Exposure of macrophages to either antibody in the absence of IFN␥ failed to stimulate NO 2 Ϫ production (data not shown). Furthermore, incubation of mouse macrophages with non-immune rabbit IgG in the presence or absence of IFN␥ did not result in the induction of NO 2 Ϫ production (data not shown). These data suggest that although ligation of CD120a (p55) is sufficient for NO 2 Ϫ production, cross-linking of CD120b (p75) augmented the response initiated by the antibody-mediated cross-linking of CD120a (p55).
Antagonistic Effects of Anti-CD120a (p55) and Anti-CD120b (p75) Antibodies on TNF␣-induced NO 2 Ϫ Production-Given the specificity of the anti-TNF receptor antibodies, it seemed reasonable to assume that the interaction with their respective antigens will result in homologous cross-linking of each receptor type. Thus, the data presented above suggests that during antibody-mediated cross-linking, CD120b (p75) enhanced the signal generated by cross-linking of CD120a (p55). Furthermore, since TNF␣ was not required in these experiments, hence eliminating any possibility of "ligand-passing," it seemed likely that cross-linking of CD120b (p75) initiated or augmented an independent signaling event. To determine if this conclusion also applied to TNF␣ itself, we independently blocked CD120a (p55) and CD120b (p75) with inhibitory concentrations of the anti-receptor antibodies (10 g/ml) and quantified the effect on the TNF␣-mediated stimulation of NO 2 Ϫ production. Macrophage monolayers were pretreated with either anti-CD120a (p55) antibody or anti-CD120b (p75) antibody for 30 min prior to stimulation with an optimal concentration of TNF␣ (10 ng/ml) for 24 h in the presence of IFN␥ (10 units/ml). Fig. 3 shows that under these conditions anti-CD120a (p55) antibody completely inhibited NO 2 Ϫ production. In addition, anti-CD120b (p75) antibody partially, although significantly (p Ͻ 0.05), blocked NO 2 Ϫ production by TNF␣.
Neither antibody stimulated the production of NO 2 Ϫ when added in the absence of IFN␥ (Fig. 3). Furthermore, non-immune rabbit IgG (10 g/ml) had no effect on the production of NO 2 Ϫ in response to challenge with IFN␥ (10 units/ml) and TNF␣ (10 ng/ml) (data not shown).
To explore further the possibilities that (i) CD120b (p75) may function to "pass" TNF␣ to CD120a (p55) and/or (ii) that CD120b (p75) may initiate a signal to enhance the response of CD120a (p55), macrophage monolayers were pretreated with an inhibitory concentration of anti-CD120b (p75) antibody (10 g/ml) for 30 min prior to stimulation with IFN␥ and increasing concentrations of TNF␣ (0.1-30 ng/ml) to span the entire concentration response curve for NO 2 Ϫ production. If CD120b (p75) was functioning to "pass" ligand to CD120a (p55), the inhibitory effect of the anti-CD120b (p75) antibody should be overcome by increasing the concentration of TNF␣ since the ligand passing phenomenon is based on the higher affinity of CD120b (p75) for TNF␣ compared with CD120b (p55). By contrast, if CD120b (p75) was required in a signaling capacity, the inhibitory effect of the anti-CD120b (p75) antibody would not be overcome by excess TNF␣ under conditions in which both CD120a (p55) and CD120b (p75) were saturated with ligand. As can be seen in Fig. 4, blocking CD120b (p75) function with an inhibitory concentration of the specific antibody resulted in an almost complete cessation of NO 2 Ϫ production at low concentrations of TNF␣. However, importantly, the inhibitory effect of the anti-CD120b (p75) antibody was not overcome with increasing concentrations of TNF␣ and remained significantly blocked by almost 50% (p Ͻ 0.05) when the cells were exposed to a concentration of TNF␣ (30 ng/ml) that maximally stimulated NO 2 Ϫ production (Fig. 4). These findings suggest that although CD120b (p75) may function to pass TNF␣ to CD120a (p55) at low ligand concentrations, it directly contributes to TNF␣-mediated signaling under saturating ligand concentrations.
Effect of Human TNF␣ on NO 2 Ϫ Production-The results obtained above were confirmed by determining the effect of human TNF␣ on the production of NO 2 Ϫ since previous studies have shown human TNF␣ to bind to mouse CD120a (p55) but not to mouse CD120b (p75). Mouse macrophage monolayers were incubated with increasing concentrations of human or mouse TNF␣ (0.01-500 ng/ml) in the presence of IFN␥ (10 units/ml) for 24 h. As can be seen in Fig. 5, incubation of mouse macrophages with mouse TNF␣ resulted in a concentration-dependent increase in NO 2 Ϫ production that was identical to that seen earlier (Fig. 1). However, the induction of NO 2 Ϫ expression in response to human TNF␣ was different in two respects as follows: (i) human TNF␣ was ineffective at low concentrations Ϫ production was only about half that seen with mouse TNF␣ and was not further increased with increasing ligand concentration (p Ͻ 0.05). Thus, the pattern of NO 2 Ϫ expression by selective ligation of CD120a (p55) with human TNF␣ was indistinguishable from that seen in response to mouse TNF␣ when CD120b (p75) function was blocked with specific antibody.
Effect of Anti-TNF Receptor Antibodies on iNOS mRNA and Protein Expression-To investigate further the roles of CD120a (p55) and CD120b (p75) in iNOS induction, we quantified the time course of induction of iNOS expression during independent and concurrent ligation of both receptors to determine the contribution of each receptor to iNOS induction. The results shown in Fig. 6 indicate that ligation of CD120a (p55) alone for 3 h stimulated iNOS mRNA expression to a level that was not different to that seen during concurrent ligation of both receptors. However, by 9 and 24 h post-stimulation, it was apparent that although the level of iNOS mRNA expression was relatively sustained in macrophages in which CD120a (p55) and CD120b (p75) were co-ligated with either the combination of anti-CD120a (p55) and anti-CD120b (p75) antibodies, or with mouse TNF␣, the level of expression progressively declined at these later time points in cells in which CD120a (p55) alone was ligated. Thus, while at 3 h, the level of expression of iNOS mRNA was 98.0 and 99.2 densitometer units for macrophages exposed to anti-CD120a (p55) antibody alone or to both CD120a (p55) and CD120b (p55) antibodies, respectively, the levels at 24 h were 76.1 and 33.3 densitometer units, respectively. These findings were also reproduced at the level of iNOS protein expression. Lysates from macrophages stimulated with mouse or human TNF␣ in the presence of IFN␥ (10 units/ml) for time intervals up to 24 h were analyzed for the presence of iNOS protein by SDS-PAGE followed by immunoblotting with an anti-mouse iNOS antibody. As can be seen in Fig. 7, incubation with mouse TNF␣, a ligand for both CD120a (p55) and CD120b (p75), led to a progressive increase in the expression of iNOS protein that peaked at 9 h and that remained elevated for the remainder of the time course. In contrast, while incubation with human TNF␣, a selective ligand for CD120a (p55), led to a similar rate of increase in iNOS protein during the first 6 h, the response peaked at 9 h but then began to decline. Consistent with both the mRNA data shown in Fig. 6, the level of iNOS protein detected at 24 h was approximately half that seen when both receptors were ligated. It can also be seen in Fig. 7 that while the amount of NO 2 Ϫ detected in culture supernatants was similar at early time points in response to both mouse and human TNF␣, the accumulation was markedly reduced at later time points in cells in which CD120a (p55) had been ligated with human TNF␣ in the absence of ligation of CD120b (p75). Thus, ligation of CD120a (p55) alone resulted in a transient expression of iNOS mRNA and protein, whereas co-ligation of CD120a (p55) and CD120b (p75) resulted in a more prolonged induction of the enzyme.
Effect of Ligation of CD120a (p55) and CD120b (p75) on JNK Activation-To begin to define how co-ligation of CD120a (p55) and CD120b (p75) may augment and prolong the expression of iNOS, we investigated the role of these receptors in the activation of JNK. In previously reported work, we have shown that specific members of the mitogen-activated protein kinase family, i.e. p42 mapk/erk2 , p38 mapk , and p46 JNK , are activated by TNF␣ (26 -28). Although CD120b (p75) plays no apparent role in p42 mapk/erk2 and p38 mapk activation, other than by the phenomenon of ligand passing at low concentrations of TNF␣ (27,28), the role of CD120b (p75) in the activation of JNK in macrophages has not been previously investigated. Therefore, to determine the relative role of each receptor in the activation of JNK, we utilized the differential binding properties of human and mouse TNF␣ to CD120a (p55) and CD120b (p75) to selectively ligate CD120a (p55) alone or to ligate both receptors, respectively, as described earlier. Mouse macrophages were stimulated with increasing concentrations of human or mouse TNF␣ (0.1-100 ng/ml) for 10 min and lysed, and the level of JNK activation was quantified using a solid phase in vitro kinase assay. As can be seen in Fig. 8, ligation of both CD120a (p55) and CD120b (p75) with mouse TNF␣ led to a concentration-dependent activation of JNK. However, al-though ligation of CD120a (p55) alone with human TNF␣ was capable of activating JNK, the responses differed between the two ligands in that the macrophages were (i) approximately 10-fold less sensitive to human TNF␣ than to the mouse cytokine at low concentrations of ligand, and (ii) there was a markedly reduced amplitude of the maximal response to human TNF␣ compared with mouse TNF␣ at high concentrations of ligand. DISCUSSION With the recognition that TNF␣ binds to cells through a binary system of receptors, a significant emphasis has been placed on determining the relative roles of each receptor in the initiation of specific cellular responses. Initial reports showed that each receptor was capable of signaling independent responses in some cell systems. For example, cross-linking of CD120a (p55) with polyclonal antibody was found to stimulate cytotoxicity in mouse LM cells (16), whereas polyclonal anti-CD120b (p75) antibody was shown to stimulate the proliferation of both mouse thymocytes and the cytolytic T-cell line, CT-6 (16). The results of the present study show that crosslinking of CD120a (p55) by either homotrimeric TNF␣ or by stimulatory concentrations of anti-CD120a (p55)-specific antibody is necessary and sufficient for the induction of iNOS mRNA and NO 2 Ϫ production. However, although ligation of CD120b (p75) alone did not stimulate appreciable iNOS mRNA or NO 2 Ϫ expression, simultaneous cross-linking of both CD120a (p55) and CD120b (p75) by stimulatory concentrations of specific antibodies resulted in an augmentation and prolongation of the expression of iNOS mRNA, protein, and NO 2 Ϫ . The conclusion that cross-linking of CD120a (p55) was necessary for iNOS and NO 2 Ϫ expression was supported by three lines of evidence. First, polyclonal antibodies directed against CD120a (p55) were found to stimulate the expression of iNOS mRNA and NO 2 Ϫ in the presence of IFN␥. Second, human TNF␣, which binds to CD120a (p55) but not CD120b (p75) (29),

FIG. 7. Time course of the effect of co-ligation of CD120a (p55) and CD120b (p75) (mouse TNF␣) or ligation of CD120a alone (human TNF␣) on the expression of iNOS protein and NO 2
؊ by mouse macrophages. Cells were co-stimulated with human or mouse TNF␣ (100 ng/ml) and IFN␥ for the times indicated. Culture supernatants were then collected and assayed for NO 2 Ϫ , whereas the cells were lysed and analyzed for iNOS protein expression by immunoblotting. was able to stimulate iNOS protein and NO 2 Ϫ production, albeit to a lesser extent. Third, blockade of anti-CD120a (p55) function at high (antibody excess) concentrations of anti-receptor antibody resulted in an almost complete inhibition of NO 2 Ϫ production in response to TNF␣. The anti-TNF receptor antibodies were less efficient agonists of iNOS mRNA and NO 2 Ϫ production than TNF␣ itself. Moreover, additional cross-linking with goat anti-rabbit IgG F(abЈ) 2 fragments did not increase the level of NO 2 Ϫ production (data not shown). The reason(s) for this finding is unclear although possible explanations include the following: (i) possible steric hindrance by the large IgG molecule leading to inefficient receptor aggregation, (ii) the formation of smaller receptor aggregates (e.g. receptor dimers), or (iii) receptor cross-linking with antibody may not fully mimic all events necessary for nitric oxide synthesis.
Stimulation of NO 2 Ϫ expression by anti-CD120a (p55) antibody or by the combination of anti-CD120a (p55) and anti-CD120b (p75) antibodies was quite strikingly different from other systems in which these antibodies have been applied, e.g. L929 cell cytotoxicity (30,31), in that the concentration-response curve for NO 2 Ϫ expression was bell-shaped. Thus, the pattern of anti-TNF receptor antibody-mediated NO 2 Ϫ production was similar to that observed in other cell systems in which receptor aggregation is an important element of cell signaling, e.g. the high affinity IgE receptor of mast cells. Although speculative, it is conceivable that this difference in the concentration-response curves for L929 cytotoxicity and macrophage NO 2 Ϫ expression may relate to the requirement for recruitment of additional signaling molecules to the intracellular domain of CD120a (p55) for the initiation of iNOS and NO 2 Ϫ expression. Moreover, IFN␥, which is required for iNOS expression but not for L929 cytotoxicity, may participate in this putative event. Direct support for this notion has been provided by the finding using deletional mutants of the CD120a (p55) intracellular domain that the membrane proximal region of the intracellular domain is essential for NO 2 Ϫ production but is not necessary for L929 cytotoxicity (31).
In addition to the essential role of CD120a (p55) in signaling, our results show that CD120b (p75) acts cooperatively with CD120a (p55) in two distinct ways to stimulate optimal induction of iNOS and NO 2 Ϫ expression. At low concentrations (pg/ ng) of TNF␣, mouse TNF␣ was a more efficient stimulus of NO 2 Ϫ production than human TNF␣. Moreover, selective blockade of CD120b (p75) function with a specific antibody substantially inhibited the capacity of low concentrations of mouse TNF␣ to stimulate NO 2 Ϫ production to an extent that resembled the dose-response curve for human TNF␣. These findings are consistent with the ligand passing concept originally developed by Tartaglia and Goeddel (20,21) and which was recently shown to result in the ligand-dependent formation of a shortlived heterocomplex between CD120a (p55) and CD120b (p75) (32).
However, in addition to ligand passing, a second form of cooperation between CD120a (p55) and CD120b (p75) was observed at higher concentrations of TNF␣ and which was characterized in three ways. First, the maximum (plateau) level of NO 2 Ϫ production seen in response to human TNF␣ was only about 50% the level detected in response to mouse TNF␣. Second, blockade of CD120b (p75) function with specific antibody inhibited the production of NO 2 Ϫ in response to mouse TNF␣ to the same level as that observed in response to human TNF␣. Moreover, in both situations, the deficit in NO 2 Ϫ production could not be overcome by increasing the concentration of the cytokine as would be predicted if CD120b (p75) acted solely to pass ligand to CD120a (p55). Third, in the absence of TNF␣, stimulation of mouse macrophages with agonistic concentra-tions of both anti-CD120a (p55) and anti-CD120b (p75) antibodies resulted in an augmentation in the expression of NO 2 Ϫ compared with cells stimulated with anti-CD120a (p55) antibody alone. Furthermore, the level of NO 2 Ϫ production observed in response to cross-linking CD120a (p55) was approximately 50% the level detected following concurrent ligation of both receptors. We interpret these collective findings as suggesting that in addition to its role in ligand passing, cross-linking of CD120b (p75) generates a signal which, while being insufficient or poorly efficient at initiating nitric oxide production, augments the signal generated by cross-linking of CD120a (p55) to mediate optimal synthesis of nitric oxide.
The nature of the augmentation in nitric oxide production was revealed in part by time course studies of the effects of independent and combined ligation of CD120a (p55) and CD120b (p75) with specific agonistic antibodies and through the use of human and mouse TNF␣ on iNOS mRNA and protein expression by macrophages. Ligation of CD120a (p55) alone was shown to result in a relatively transient increase in iNOS mRNA expression, whereas co-ligation of both receptors led to a sustained increase in iNOS mRNA analogous to that seen during exposure to recombinant mouse TNF␣. These findings were substantiated with the finding that ligation of CD120a (p55) with human TNF␣ also led to a transient expression of iNOS protein, whereas ligation of both receptors with mouse TNF␣ led to a more prolonged induction of iNOS protein. Thus, the greater level of NO 2 Ϫ produced as a consequence of coligation of both receptors appears to arise by a more prolonged presence of the enzyme involved in nitric oxide synthesis compared with ligation of CD120a (p55) alone. This conclusion also raises the question of how co-ligation of both receptors may result in a sustained response leading to greater output of NO 2 Ϫ , as opposed to the transient response leading to more modest levels of NO 2 Ϫ output induced by ligation of CD120a (p55) alone.
One possible mechanism investigated herein was that coligation of CD120a (p55) and CD120b (p75) leads to an augmentation in signaling responses. Previously reported work from this laboratory has shown that ligation of CD120a (p55) is sufficient to activate both p42 mapk/erk2 and p38 mapk and that co-ligation of both receptors has no additional effect on the activation of these kinases at saturating concentrations of TNF␣ (27,28). In contrast, in the present study we observed that co-ligation of CD120a (p55) and CD120b (p75) with mouse TNF␣ resulted in a marked increase in the level of JNK activation compared with ligation of CD120a (p55) alone. This finding suggests that signals arising from both receptors may act to increase signal transduction output compared with that arising from ligation of CD120a (p55) alone. Interestingly, recently reported studies have suggested that TRAF2 is a necessary component in the activation of JNK (although not involved in the activation of p42 mapk/erk2 ) (33,34). In addition, TRAF2 has been implicated in signaling from CD120a (p55) through its ability to indirectly bind to the intracellular domain of the receptor via TNF receptor-associated death domain-containing protein (TRADD) (35), and in signaling from CD120b (p75) via its ability to directly interact with the cytoplasmic domain of the receptor (36). Furthermore, NF-B has been implicated in the induction of iNOS expression (37) and is independently activated by ligation of both CD120a (p55) and CD120b (p75). Thus, perhaps the signaling pathways leading to NF-B activation also undergo sustained activation during co-ligation of both receptors.
Synergistic interactions between CD120a (p55) and CD120b (p75) have been reported in other cell types and systems. For example, Weiss and colleagues (38) have shown that CD120b (p75) dramatically enhances CD120a (p55)-mediated cytoxicity in HeLa cells, also through a TRAF2-dependent mechanism; and Kalb and colleagues (39) provided data to suggest that co-ligation of both receptors enhanced fibroblast proliferation over that seen in response to ligation of CD120a (p55) alone. The results of the present study provide new insights into how these receptors may also interact to regulate the expression of inducible nitric oxide synthase and have clear implications with respect to macrophage defense against obligate intracellular pathogens such as T. gondii.