Activation of Phosphatidylinositol 3-Kinase, Protein Kinase B, and p70 S6 Kinases in Lipopolysaccharide-Stimulated Raw 264.7 Cells: Differential Effects of Rapamycin, Ly294002, and Wortmannin on Nitric Oxide Production¹

Bacterial LPS is a potent activator of macrophage functions, which include the production of TNF-α, IL-1, IL-6, and nitric oxide (NO)³ (1, 2). The signaling pathways consequent upon stimulation are beginning to be unraveled; however, understanding of these is far from complete. LPS is believed to mediate its effects through the CD14 receptor (3) and requires additionally a serum-derived LPS binding protein. The CD14 receptor lacks intrinsic kinase activity and was proposed to transduce signaling through the p53/56^lyn protein tyrosine kinase (4). However, recent work casts doubt on this step, as macrophages without Hck, Fgr, and Lyn demonstrated unimpaired responses to LPS (5).

Normally, phosphorylation of receptor tyrosine kinases enables binding with molecules containing SH2 groups such as the adaptor subunit of the PI 3-kinase. These is an important class of lipid kinases that catalyze the phosphorylation of phosphoinositides at the D-3 hydroxyl of the inositol ring generating PI 3-phosphate, PI 3,4-biphosphate, and PI 3,4,5-triphosphate (6, 7). It is noteworthy that PI 3-kinase is activated by LPS and appears to associate with p53/56^lyn in human monocytes (8). This event would implicate the activation of downstream signaling molecules such as p70 S6 kinase, but this has not been shown in the monocyte/macrophage system. However, PKCζ activation has been demonstrated to occur by a PI 3-kinase-dependent mechanism using specific inhibitors and a dominant-negative p85 mutant (9).

Multiple cytokines have been demonstrated to activate PI 3-kinase in a variety of systems, including IL-2, IL-4, IL-10, IL-13, granulocyte-macrophage CSF, and steel factor (10, 11, 12, 13). Investigation of the functional importance of this pathway for hemopoietic cells has revealed an interesting regulatory role for this kinase in the immune response. Pharmacologic intervention with wortmannin and Ly294002 (PI 3-kinase inhibitors) upon granulocyte-macrophage CSF/TNF/FMLP-activated neutrophils indicates a function for PI 3-kinase in superoxide generation, platelet-activating factor release, and migration (14). However, cytokines mentioned previously, such as IL-10, are clearly involved in negatively regulating immune cell function (11). Furthermore, activated PI 3-kinase inhibited T cell receptor-mediated NF-AT induction (15).

The downstream effector molecules for PI 3-kinase remain to be clarified, but the current literature favors a model in which PKB (the cellular homologue of the transforming oncogene, v-akt) lies below PI 3-kinase and upstream of p70 S6 kinase (16, 17). Activation of PKB is recognized to occur through a wortmannin-sensitive phosphorylation of PKB at the Thr³⁰⁸ and Ser⁴⁷³ sites (18) and appears to involve membrane translocation of the kinase (19, 20). Recently, the kinase responsible for Thr³⁰⁸ phosphorylation has been characterized and has been designated PDK1 for PI 3,4,5-triphosphate-dependent protein kinase 1 (21). From a functional perspective, PKB is itself an important molecule in insulin signaling as well as in prevention of apoptosis (22, 23).

The target of rapamycin (known as TOR or FRAP for FKBP12-rapamycin-associated protein) lies either below PKB or on a parallel pathway (24) and is also involved in the activation of p70 S6 kinase (25). It has demonstrated kinase activity in terms of both autophosphorylation and phosphorylating the PHAS-1 protein, which is also known as eIF-4E binding protein and functions as a translational repressor. Recent work indicates that FRAP is a proline-directed protein kinase (26, 27, 28).

The p70 S6 kinase is the kinase responsible for phosphorylation of the S6 protein in response to various stimuli, including GH and insulin. Its activation is complex, requiring multiple serine and threonine phosphorylation events. Intriguingly, recent work has indicated that PDK1 can phosphorylate and activate the p70 S6 kinase (29). Although its functions remain to be clarified (demonstrated indirectly through the use of rapamycin), they include cell cycle control, transcription, and translation initiation (30). To date, the best evidence for a direct functional role involves transcriptional regulation. Rapamycin blocks the Ser¹¹⁷ phosphorylation of the cAMP-responsive modulator (CREMτ) in response to serum, which is the site phosphorylated by p70 S6 kinase in vivo (31).

The biochemical regulation of the production of biologically active molecules by hemopoietic cells is presently not well understood. This is likely to be important, as some of the signaling proteins involved may well be useful targets for therapeutic intervention. We have previously shown that in Raw 264.7 cells LPS activates multiple mitogen-activated protein kinase family members, such as p44 Erk1, p42 Erk2, p46/p54 JNK/SAPK, and p38 Hog (32). JNK/SAPK has been demonstrated to have a role in the translation of TNF-α (33). Additionally, p38 Hog has been demonstrated to be involved in the expression of the inducible nitric oxide synthase (iNOS) gene in mouse astrocytes and primary glial cultures (34, 35). However, there are limited data concerning the involvement of the PI 3-kinase and its putative effector molecules in similar systems. One such example is that histamine secretion in basophils has been demonstrated to be PI 3-kinase dependent (36). Additionally, rapamycin has been demonstrated to inhibit IFN-γ induced by IL-1 in the murine T cell lymphoma, YAC-1 (37). We therefore hypothesized that activation of the PI 3-kinase/PKB/p70 S6 kinase pathway is involved in the production of NO. The latter is produced by the catalytic transformation of l-arginine to citrulline by iNOS, which in Raw macrophage-like cells is a process requiring protein tyrosine phosphorylation (38). Posttranslational modification of iNOS through its tyrosine phosphorylation has recently been demonstrated to be an early event (39). NO is an important biologic molecule with tissue-specific effects that in macrophages include microbicidal and tumoricidal activities as well as apoptosis (3, 40). It also has been demonstrated to be directly correlated with the severity of pathologic conditions, such as AIDS dementia, bronchial asthma, and inflammatory bowel disease (41).

In this study we provide evidence for the activation of individual components of this pathway by LPS in a macrophage cell line and the involvement of an Ly294002- and rapamycin-sensitive protein kinase in the generation of nitric oxide through a possible phosphorylation-dependent mechanism of iNOS itself.

LPS (Escherichia coli serotype 055:B5) and MTT were obtained from Sigma (St. Louis, MO). Abs for p70 S6 kinase (NT), p85 of PI 3-kinase, and the pleckstrin homology domain of PKB as well as the S6 protein peptide substrate and histone H2B were obtained from Upstate Biotechnology (Lake Placid, NY). Ab to iNOS was obtained from Santa Cruz Biotechnology (Santa Cruz, CA). The lipid substrate phosphatidylinositol was obtained from Avanti Polar Lipids (Alabaster, Atlanta, GA). Goat anti-rabbit IgG was obtained from Bio-Rad (Richmond, CA). The inhibitors used were wortmannin, Ly290042, and rapamycin (all from Calbiochem, San Diego, CA). Protein A-Sepharose beads were obtained from Pharmacia (Piscataway, NJ). [γ-³²P]ATP was purchased from DuPont (Wilmington, DE). The TLC plates were obtained from EM Science (Gibbstown, NJ).

Raw 264.7 cells were cultured and stimulated as previously described with minor modifications (32). Briefly 1 × 10⁶ cells were seeded into 100-mm petri dishes containing 10 ml of DMEM supplemented with 10% heat-inactivated FCS and grown at 37°C. After 40 h, by which time they had reached 90% confluence, the cells were switched to serum-free medium and were stimulated with either LPS or PMA 6 h later. For PI 3-kinase activity determinations, the cells were serum starved overnight. Where inhibitors were employed, a preincubation time of 1 h was allowed. Following stimulation the cells were washed with ice-cold PBS, scraped into a 15-ml Falcon tube using a rubber policeman, sedimented by centrifugation, and sonicated for 20 s after the addition of homogenization buffer containing 20 mM MOPS, 50 mM β-glycerophosphate, 5 mM EGTA, 50 mM NaF, 1 mM sodium vanadate, and 1 mM PMSF. For the PI 3-kinase assays the lysis buffer consisted of 20 mM Tris-HCl (pH 8.0), 137 mM NaCl, 1% Triton X-100, 2 mM EDTA, 1 mM sodium vanadate, 1 mM sodium molybdate, and 10% glycerol. Before use of either buffer the following protease inhibitors were added: 40 μg/ml PMSF, 0.5 μg/ml leupeptin, and 2 μg/ml aprotinin.

Aliquots of crude extracts were resolved using 11% SDS-PAGE (42) and transferred to nitrocellulose membrane using a Bio-Rad transblot apparatus at 300 mA for 3 h. The blots were blocked with 5% BSA overnight and washed three times with TTBS (20 mM Tris-HCl (pH 7.4), 250 mM NaCl, and 0.05% Tween 20), the appropriate primary Ab was applied for 4 h, blots were rewashed with TTBS, the secondary Ab was applied for 1 h, and blots were rewashed and developed using enhanced chemiluminescence for horseradish peroxidase-conjugated secondary Abs according to the manufacturer’s recommendations (ECL kit, Amersham, Arlington Heights, IL).

Five hundred micrograms of soluble protein were added to 200 μl of an IP buffer containing 50 mM Tris-HCl (pH 7.4), 100 mM NaCl, 50 mM NaF, 5 mM EDTA, and 3% Nonidet P-40. Five microliters of Ab was added, and the tubes were placed on a Labquake shaker for 1 h at 4°C. Then, 30 μl of a 1/1 slurry of protein A-Sepharose beads was added, and the rotation was continued for 2 h at 4°C. Following this the beads were washed twice with IP buffer and twice with 12.5 mM β-glycerol phosphate, 20 mM MOPS (pH 7.2), 5 mM EGTA, 7.5 mM MgCl₂, 50 mM NaF, and 0.25 mM DTT. Then the beads were resuspended and subjected to an immune complex kinase assay as described below.

Ten microliters of the substrate mixture, either histone H2B (1 mg/ml) or the S6 protein-based peptide substrate (0.5 mg/ml) AKRRRLSSL-RASTSKSESSQK (43) in assay dilution buffer (20 mM MOPS (pH 7.2), 25 mM β-glycerophosphate, 20 mM MgCl₂, 5 mM EGTA, 2 mM EDTA, 1 mM DTT, and 1 mM sodium vanadate), were added to the washed beads, and the reaction was conducted as described above for 20 min using 5 μl of the ATP mixture (250 μM ATP and 1 μCi of [γ-³²P]ATP). Twenty microliters were then spotted onto P81 filter paper, washed extensively with 1% phosphoric acid, and analyzed in a scintillation counter after addition of 250 μl of scintillation fluid. Thirty microliters of 2× sample buffer were added to the beads, and after boiling for 5 min the sample was resolved by 11% SDS-PAGE and subjected to immunoblotting.

This was performed essentially as previously described (13). Briefly detergent lysates were immunoprecipitated with the p85-PI 3-kinase Ab for 3 h, and the beads were washed twice with lysis buffer and three times with Tris-HCl (pH 7.4). Subsequently, the PI was prepared by drying with nitrogen and resuspending in 10 μl of 30 mM HEPES. This was added to the washed beads, and the tube was left on ice for 10 min. Then 40 μl of kinase buffer (30 mM HEPES, 30 mM MgCl₂, 50 μM ATP, 200 μM adenosine, and 10 μCi of [γ-³²P]ATP) were added to each tube, and the reaction was allowed to proceed at room temperature for 15 min. The reaction was stopped with 0.1 N HCl, and the lipids were extracted with 200 μl of chloroform/methanol (1/1). The products were separated on potassium oxalate-pretreated TLC plates by developing with chloroform, methanol, water, and 30% ammonium hydroxide (112/88/19/6). After drying, the plates were exposed to autoradiography, and the phosphorylated products were quantified by excising the spot and scintillation counting.

Raw 264.7 cells were treated with the indicated inhibitors in the absence and the presence of LPS, and culture supernatants collected at 24, 48, and 72 h were centrifuged at 600 × g for 10 min. Nitrite levels were measured by addition of 50 μl of the Griess reagent (1.5% sulfanilamide and 0.1% naphthylethylenediamine dihydrochloride in 2.5% H₃PO₄) to 50 μl of culture supernatant in 96-well plates, leaving in the dark for 10 min, and measuring the color intensity with an automated microtiter plate reader at 550 nm.

Raw cells were cultured in 60-mm petri dishes for 24 h in the presence and the absence of LPS, Ly294002, rapamycin, and wortmannin. After cell lysis as described above the iNOS protein was immunoprecipitated from equal amounts of crude lysate using a rabbit polyclonal Ab. The samples were then resolved on 9% SDS-PAGE and probed with the same Ab, and the bands were detected using ECL.

Raw cells were grown in 96-well plates at a concentration of 5000 cells/well for 24 h, and then MTT or [³H]]thymidine was added. The plates were washed with PBS 4 h later, and scintillation fluid was added for the DNA synthesis assay and quantitated for radioactivity in a scintillation counter (1450 Microbeta, Wallac, Gaithersburg, MD). For the MTT assay 0.1 N HCl was added in absolute propanol, and the absorbance was measured spectrophotometrically at a wavelength of 570 nm. The results are representative of three experiments performed in quadruplicate and are expressed as absorbance units for the MTT assay and counts per minute for the [³H]thymidine uptake assay.

Raw 264.7 cells were cultured under the conditions outlined above. They were switched to phosphate-free medium containing 0.5 mCi of [³²P]phosphate on the second day, and the protein kinase inhibitors were added 30 min before adding LPS. After 18 h at 37°C the cells were lysed in homogenization buffer, and immunoprecipitations were conducted for the iNOS protein as outlined above.

Although activation of p70 S6 kinase has been previously demonstrated in the liver of endotoxic rabbits (44), this has not been described in LPS-stimulated macrophages. The initial objective was therefore to ascertain whether LPS activated p70 S6 kinase in Raw 264.7 cells. The phosphotransferase activity of p70 S6 kinase toward a peptide substrate was assayed following immunoprecipitation from the crude lysates of cells incubated for 0–60 min with LPS. Optimal activation of p70 S6 kinase was observed after 30 min of LPS treatment as shown in Fig. 1. This corresponded to the timing of the maximal electromobility band shift shown in Fig. 1 B. These results were obtained with two different Abs directed at the N and C termini of p70 S6 kinase.

PMA treatment of diverse cells induces a robust activation of p70 S6 kinase via PKC. As shown in Fig. 2,B, LPS and PMA both elicited a fivefold activation of S6 kinase after a 30-min stimulation. This was accompanied by a retardation of the electromobility of the S6 kinase consistent with its hyperphosphorylation (Fig. 2,A). When the FRAP inhibitor rapamycin and the PI 3-kinase inhibitor Ly294002 were coincubated with LPS or PMA in the medium of the Raw cells, there was a complete inhibition of both the phosphorylation and the activation of S6 kinase. Intriguingly, wortmannin, an alternative PI 3-kinase inhibitor, did not completely abolish the activation of p70 S6 kinase by LPS (Fig. 2,B), effecting approximately a 50% reduction of the kinase activity. This indicated that wortmannin and Ly294002 may act via additional targets besides PI 3-kinase. Previous work has indicated that wortmannin (in supra-PI 3-kinase inhibitory concentrations) and Ly294002 are capable of inhibiting the autokinase activity of FRAP (26). Our observations are compatible with a FRAP isoform in Raw cells that is more sensitive to the effects of Ly294002 than to wortmannin. Significantly, in another study (25) FRAP was found to be resistant to the effects of wortmannin, lending further weight to this possibility. Ultimately, characterization of the FRAP isoform in macrophages will clarify this problem. The existence of a PI 3-kinase isoform that is more sensitive to inhibition by Ly294002 than to wortmannin is unlikely, since novel isoforms, such as the PI 3-kinase activated by Gβγ subunits (45), retain sensitivity to these inhibitors. The PMA activation of p70 S6 kinase was also relatively insensitive to intervention with wortmannin (Fig. 2), yet it was clearly inhibited by Ly294002, lending further support to Ly294002 possibly behaving like a FRAP inhibitor (in addition to its PI 3-kinase inhibitory activity).

To clarify whether the p85-associated PI 3-kinase was, in fact, sensitive to inhibition by the standard inhibitors, the next step was to assess the activation of PI 3-kinase in this system. This was conducted by immunoprecipitation and assessment of the in vitro lipid phosphotransferase activity. As shown in Fig. 3, A and B, there was a rapid and sustained activation of PI 3-kinase with LPS. This finding was in accordance with previous results in human monocytes, where LPS led to an activation of PI 3-kinase and its association with the tyrosine kinase Lyn (8). Wortmannin and Ly294002 both inhibited PI 3-kinase activity following 10 min of LPS (Fig. 3 C). These findings indicated that PI 3-kinase did exhibit activation in response to LPS in Raw 264.7 cells and that this activation was sensitive to both of the standard PI 3-kinase inhibitors. Therefore, the incomplete inhibition of p70 S6 kinase with wortmannin was unlikely to be due to a differential sensitivity of the PI 3-kinase to this agent and Ly294002.

The current model of p70 S6 kinase activation has implicated the involvement of PKB upstream, so its role was investigated. The Ab used for this analysis was directed at the pleckstrin homology domain present at the N terminus, which is conserved among all the known isoforms of PKB. Fig. 4 shows that while LPS was able to activate PKB approximately twofold, the effects of PI 3-kinase inhibitors on this activation were surprisingly modest. In control experiments with murine keratinocytes using insulin, a fivefold activation of PKB was observed, with attenuation of this response when the cells were pretreated with wortmannin (data not shown). Although the magnitude of the LPS-mediated activation was modest, PKB activation was PI 3-kinase independent in this system and probably did not contribute significantly to activation of p70 S6 kinase. Two recent observations make these data compatible with current understanding of the regulation of this kinase. Firstly, PKB can be activated following cellular stresses in a PI 3-kinase-independent fashion (46). Secondly, cAMP also activates PKB via a wortmannin-insensitive mechanism (47). Furthermore, taken in conjunction with the data in Fig. 2, a model in which FRAP, rather than PKB, lies upstream of p70 S6 kinase appears more plausible. The observation of the absence of PKB activation by PMA makes it highly unlikely that PKC is involved in its regulation.

Having demonstrated that biochemical activation of the signaling molecules under investigation does indeed occur in macrophages in response to LPS, and that their regulation occurred somewhat differently from the generally accepted model for insulin signaling, the next step was to investigate the effects of wortmannin, Ly294002 and rapamycin on NO production. NO is a functionally important molecule that can be deleterious in certain situations when produced in excess. Its production has been well characterized in Raw cells and correlates very well with the appearance of iNOS (39). The effects of the inhibitors on NO production by Raw cells were assessed using the Griess reagent. Fig. 5 shows that LPS effects a robust production of NO over a period of 72 h. Examining the effects of the inhibitors revealed that while Ly294002 and rapamycin were able to abolish this response almost completely, wortmannin was ineffective, even when it was added again after 12 h of incubation, because of its relative instability in aqueous solutions. In fact, wortmannin paradoxically led to a small increase in its production. The observed changes were maintained for a total duration of 72 h without further addition of inhibitor for either Ly294002 or rapamycin and daily addition of wortmannin. This inhibitory effect on NO production by Ly294002 and rapamycin was not due to killing of the cells by an apoptosis-promoting effect, as cell viability, assessed by trypan blue exclusion, at 24 h was unchanged among the various treatment groups.

In view of the critical role of PI 3-kinase and PKB in cell survival, we sought to further validate this by performing both [³H]thymidine uptake and MTT assays, which are measures of cellular proliferation and mitochondrial activity, respectively. The results, which are shown in Table I, confirm that the effects of the inhibitors could not be attributed simply to a nonspecific toxicity. More specifically, LPS induced an increase in mitochondrial activity (MTT) that was not influenced by any of the inhibitors (a vs b, p < 0.005). Predictably, LPS mediated a negative influence on cell proliferation as assessed by [³H]thymidine uptake (a vs b, p < 0.0005), which persisted to an extent with Ly294002, but was abolished using either rapamycin or wortmannin. Furthermore, the relative insensitivity of PKB activity to these inhibitors argues against an apoptosis-promoting effect. These data indicate a specific effect of both Ly294002 and rapamycin on pathways directly involved in the production of NO.

The role of iNOS protein induction was investigated by directly immunoprecipitating the protein from macrophage extracts stimulated for 24 h with LPS in the absence and the presence of Ly294002, rapamycin, and wortmannin. Fig. 6 shows that LPS stimulation resulted in a predictable increase in the amount of iNOS protein, and that this was largely unaffected by any of the inhibitors used in the study. This was a surprising finding given that the production of NO is currently believed to be due to induction of the iNOS protein. This invoked a posttranslational mechanism as being the step critically modulated by both Ly294002 and rapamycin.

To address whether a phosphorylation step was involved in the regulation of iNOS activity, an in vivo labeling study was performed. Raw 264.7 cells were cultured in medium containing 0.5 mCi of orthophosphate with the appropriate inhibitors and LPS. After harvesting the cells, immunoprecipitation of the iNOS protein was conducted, and the samples were resolved on 9% SDS-PAGE. Fig. 7 indicates that the protein was phosphorylated upon stimulation of Raw 264.7 cells with LPS. To our knowledge this is the first time that this has been demonstrated and implicates iNOS phosphorylation as an important potential means of its regulation. Of further significance is the observation that phosphorylation of iNOS was abolished in response to pretreatment with both Ly294002 and rapamycin in concert with the inhibition of NO production by these agents. Wortmannin pretreatment had no effect on the iNOS phosphorylation, in keeping with its inability to inhibit NO production. These observations indicate that FRAP or its downstream effector(s) is involved in the regulation of iNOS activity.

Previous work has examined the roles played by protein tyrosine phosphorylation and activation of the mitogen-activated protein kinase pathway consequent upon LPS stimulation of macrophages (48). Protein tyrosine kinase activity generally and p38 Hog kinase activity specifically have been implicated in the involvement of this pathway in TNF and IL-1 production (49, 50). Protein tyrosine kinase inhibitors have been demonstrated to block LPS-induced NO production (38). However, the identity of the molecules involved remains unknown; as discussed previously, characterization of function in macrophages lacking Hck, Fgr, and Lyn demonstrated unimpaired responses to LPS (5). One of the possible explanations for these seemingly irreconcilable observations is that genistein inhibited other protein kinases.

The role of LPS in the activation of the PI 3-kinase PKB and p70 S6 kinase and the influence of rapamycin and PI 3-kinase inhibitors on this activation and on nitric oxide production have never been formally investigated. The findings of our study indicate that all the components of this pathway are activated in response to stimulation with LPS; however, the use of pharmacologic inhibitors indicates that this activation does not proceed along the previously accepted model established for insulin signaling. This is exemplified by two observations. Firstly, PKB activation with LPS is uninfluenced by PI 3-kinase inhibitors. Secondly, p70S6 kinase activation is only partially abolished by wortmannin in response to both LPS and PMA, whereas both Ly294002 and rapamycin completely inhibited this activation with both agonists. When the effects of these inhibitors on NO production were examined, a similar pattern was observed, in that both Ly294002 and rapamycin completely inhibited NO production, but wortmannin did not have any effect. The explanation for the differential response to wortmannin and Ly294002 is not immediately clear, but it indicates that different kinases may be involved. In this regard, the incomplete inhibition of p70 S6 kinase activation by wortmannin indicates that PI 3-kinase-independent mechanisms may be playing a role. In support of this, cardiomyocytes exhibit a relative insensitivity to wortmannin of p70 S6 kinase activation in response to sodium arsenite (51), and additionally, endothelin-mediated activation of p70 S6 kinase in bovine airway smooth muscle cells is wortmannin insensitive (52).

The discovery that there are multiple regulatory subunits of the PI 3-kinase (53) with different responses to insulin activation indicates that these (i.e., p50α and p55α) may play a role in LPS signaling. Additionally, the identification of the p101 regulatory subunit mediating signaling through Gβγ subunits (45) serves to emphasize that the regulation of these lipid signaling molecules will probably be more complex than is generally appreciated. In the current study, however, the p85-associated in vitro PI 3-kinase activity was clearly sensitive to inhibition by both inhibitors.

Although the question of wortmannin stability is an important one, several lines of evidence in different cell systems indicate that this may have been overemphasized. A study in chicken macrophages clearly showed an inhibitory effect of wortmannin on LPS-induced NO production (54). In an intestinal cell line, it was clearly able to attenuate the inhibitory effect of IL-13 on TNF/IFN-γ/IL-1-mediated NO production (12).

PKC is also an important downstream effector of PI 3-kinase (24, 55). It has been implicated in both the LPS and the IFN-γ-mediated production of NO (56, 57, 58). In Raw cells, PMA is unable to stimulate NO production unless there is overexpression of PKCε (59). As PMA produced a robust activation of p70 S6 kinase in the present study (Fig. 1), our findings indicate that this activation is insufficient for production of NO. An upstream kinase is probably required for this response in a manner analogous to the mechanism proposed for the insulin-mediated phosphorylation of PHAS-1 (60).

Intriguingly, a strong case is made for the involvement of FRAP in the LPS-mediated production of NO. A differential sensitivity of FRAP to Ly294002 and wortmannin in this particular system could be an explanation for the lack of a functional effect for the latter. This is possible, as one study clearly demonstrated that FRAP kinase activity was wortmannin insensitive (25). More importantly, however, the findings from this study indicate that rapamycin and Ly294002 affect NO production at a site distal to the induction of the iNOS protein. A schematic diagram (Fig. 8) summarizes the novel findings of this study, the most pertinent of which are the unlikelihood that PI 3-kinase regulates FRAP in this system and that FRAP or a downstream effector is responsible for the phosphorylation and consequently the activation of iNOS leading to the production of NO.

In conclusion, this study has provided novel insights into LPS signaling in Raw 264.7 cells and has identified FRAP as a potentially important target for immunosuppression where the abnormal production of NO is judged to be deleterious, such as in inflammatory bowel disease. Alternatively, an additional (other than T cell proliferation (61)) explanation is offered for the immunosuppressive effects of rapamycin, one that involves the inhibition of NO production by the macrophage cell system. Finally, iNOS phosphorylation has been demonstrated for the first time in vivo, and its activity (NO production) has been correlated with the effects of Ly294002, rapamycin, and wortmannin.

The technical assistance provided by E. Leung was appreciated as well as helpful discussions with Dr. J. Sanghera.