The Phosphatidylinositol 3-Kinase-Akt Pathway Limits Lipopolysaccharide Activation of Signaling Pathways and Expression of Inflammatory Mediators in Human Monocytic Cells*

Monocytes and macrophages express cytokines and procoagulant molecules in various inflammatory diseases. In sepsis, lipopolysaccharide (LPS) from Gram-negative bacteria induces tumor necrosis factor-alpha (TNF-α) and tissue factor (TF) in monocytic cells via the activation of the transcription factors Egr-1, AP-1, and nuclear factor-κB. However, the signaling pathways that negatively regulate LPS-induced TNF-α and TF expression in monocytic cells are currently unknown. We report that inhibition of the phosphatidylinositol 3-kinase (PI3K)-Akt pathway enhances LPS-induced activation of the mitogen-activated protein kinase pathways (ERK1/2, p38, and JNK) and the downstream targets AP-1 and Egr-1. In addition, inhibition of PI3K-Akt enhanced LPS-induced nuclear translocation of nuclear factor-κB and prevented Akt-dependent inactivation of glycogen synthase kinase-β, which increased the transactivational activity of p65. We propose that the activation of the PI3K-Akt pathway in human monocytes limits the LPS induction of TNF-α and TF expression. Our study provides new insight into the inhibitory mechanism by which the PI3K-Akt pathway ensures transient expression of these potent inflammatory mediators.

However, mechanisms and signaling pathways that limit the magnitude of the induction of these genes are poorly understood.
Recent evidence suggests that activation of phosphatidylinositol 3-kinase (PI3K), a ubiquitous lipid-modifying enzyme, may modulate positively acting signaling pathways. PI3K is a heterodimeric protein consisting of a p85 regulatory subunit and a p110 catalytic subunit. LPS stimulation of monocytes/ macrophages activates the PI3K pathway (16 -18), although the steps between the CD14⅐TLR4⅐MD2 complex and activation of PI3K have not been characterized. Activation of PI3K appears to occur via phosphorylation of tyrosine residues in the Src homology 2 domain of p85, which permits docking of PI3K to the plasma membrane and allows allosteric modifications that increase its catalytic activity (19 -21). Activated PI3K catalyzes the phosphorylation of membrane inositol lipids and the accumulation of phosphatidylinositol 3,4,5-trisphosphate and its phospholipid phosphatase product phosphatidylinositol 3,4-bisphosphate in the membrane. These membrane changes allow docking of the lipid kinases phosphatidylinositol-dependent kinase 1 and protein kinase B/Akt. After membrane recruitment Akt is activated by dual phosphorylation of Ser 473 and Thr 308 by phosphatidylinositol-dependent kinase 1 and possibly phosphatidylinositol-dependent kinase 1-related kinase-2 (22,23).
The PI3K-Akt pathway has been shown to regulate negatively NF-B and the expression of inflammatory genes. Wortmannin, a specific inhibitor of PI3K, enhanced LPS-induced nitric-oxide synthase in murine peritoneal macrophages (18), and activation of PI3K-Akt suppressed LPS-induced lipoprotein lipase expression in J774 macrophages (24). Induction of nitric-oxide synthase in C6 glial cells and rat primary astrocytes was also regulated negatively by activation of PI3K (17), and a constitutively active PI3K inhibited induction of nitricoxide synthase gene expression in human astrocytes (25). Angiopoeitin-1, a potent activator of PI3K, negatively regulated vascular endothelial growth factor-and TNF-␣-induced TF expression in endothelial cells (26). Finally, in endothelial cells the PI3K-Akt pathway limited vascular endothelial growth factor activation of the p38 MAPK pathway and TF gene expression (27).
In contrast to studies showing that the PI3K-Akt pathway negatively regulates expression of inflammatory genes in macrophages, a recent study demonstrated that the PI3K-Akt pathway positively regulated nuclear factor (NF)-B-dependent gene expression in HepG2 cells via phosphorylation and increased transactivation activity of p65 (28). Overexpression of a constitutively active form of Akt also increased NF-B-dependent gene expression in 3T3 fibroblasts via the activation of I-B kinase and the p38 MAPK (29). Activation of PI3K-Akt has been implicated in playing a pivotal role in cytokine-induced transcriptional activation of NF-B-and AP-1-dependent gene expression and in inhibiting apoptosis (30 -34). Finally, activation of the TLR2 receptor in human monocytic cells by Gram-positive bacteria activated the PI3K-Akt pathway and increased the transactivation activity of p65 (35).
The PI3K-Akt pathway has been shown to regulate negatively many kinases including Raf-1 and glycogen synthase kinase (GSK)-3␤, which mediate induction of inflammatory genes. We and others have shown that LPS-induced TNF-␣ and TF expression in monocytic cells is mediated, in part, via the activation of the Raf-MEK-ERK1/2 pathway (2,6,13,15). Activation of Akt has been shown to regulate negatively the serine/threonine kinase Raf-1 and the downstream MEK-ERK1/2 signaling pathway (36,37). Akt induces an inhibitory phosphorylation of Ser 259 in the N-terminal CR2 domain of Raf-1, which increases its association with 14-3-3 protein and masks the accessibility of residues in the kinase domain of Raf-1 necessary for its activation (38). Dephosphorylation of Ser 259 and phosphorylation of Ser 338 and Tyr 341 in the Cterminal kinase domain are required for Raf-1 activation and its interaction with downstream substrates (39,40).
GSK-3␤ is another serine/threonine kinase that is inhibited by Akt-dependent phosphorylation. Akt phosphorylates Ser 9 in the N terminus of GSK-3␤ and inactivates the kinase (41). The phenotype of GSK-3␤ Ϫ/Ϫ embryos is similar to that of RelA Ϫ/Ϫ embryos, suggesting that GSK-3␤ may regulate the transactivational activity of p65 (42). LiCl is a potent inhibitor of GSK-3␤ and inactivates GSK-3␤ by inducing its phosphorylation on Ser 9 in its N terminus (43). Indeed, the effect of LiCl on wingless signaling in wild-type cells mimicked the phenotype observed in GSK-3␤ null cells (41). The role of GSK-3␤ in the regulation of inflammatory genes in monocytes is currently undefined.
Our study demonstrates that inhibition of the PI3K-Akt pathway enhances LPS-induced TNF-␣ and TF gene expression via increased activation of Egr-1-, AP-1, and NF-B. Inhibition of PI3K also enhanced TNF-␣ and TF gene expression, in part, by increasing the transactivational activity of p65 by inhibiting Akt-dependent inactivation of GSK-3␤. Therefore, activation of the PI3K-Akt pathway in LPS-treated human monocytes and THP-1 cells limits the induction of TNF-␣ and TF gene expression. PBMCs and THP-1 cells were stimulated with either 10 ng/ml or 10 g/ml LPS, respectively. C, THP-1 cells were left untreated or preincubated with 20 or 50 mM LiCl or 50 mM NaCl for 60 min before stimulation with 10 g/ml LPS for 5 h. TNF-␣ antigen in cell culture supernatants was determined by ELISA, and TF activity was determined using a one-step clotting assay. These data represent the mean Ϯ S.D. from three independent experiments. Black bars indicate TNF-␣ levels, and white bars indicate TF levels.
TNF-␣ ELISA-To study the effect of the PI3K inhibitors on TNF-␣ production, THP-1 cells or PBMCs (1 ϫ 10 6 ) were preincubated with 100 nM wortmannin or 10 M LY294002 for 1 h at 37°C before the addition of LPS for 5 h at 37°C. The effect of inhibition of GSK-3␤ on TNF-␣ release was studied by preincubating THP-1 cells with 20 or 50 mM LiCl for 1 h at 37°C before stimulating with LPS for 5 h at 37°C. 20 or 50 mM NaCl served as an osmolarity control for LiCl. TNF-␣ protein levels were measured using a commercial ELISA kit (R&D Systems, MN).
TF Activity-THP-1 or PBMC cell pellets (1 ϫ 10 6 ) were solubilized at 37°C for 15 min using 15 mM octyl-␤-D-glucopyranoside. TF activity in cell lysates was measured using a one-stage clotting assay as described by Morrissey et al. (45) with the PT program on the Start 4 clotting machine (Diagnostica Stago, Asnieres, France). Clotting times were converted to milliunits of TF activity by comparison with a standard curve established with purified human brain TF.
Western Blotting-Whole cell lysates and cytosolic and nuclear extracts were prepared from THP-1 cells (5 ϫ 10 6 ) (16,45). Protein concentrations were measured using a Bio-Rad protein assay kit. Proteins were separated by SDS-PAGE and transferred to Hybond-enhanced chemiluminescence membrane (Amersham Biosciences). Activation of Akt, ERK1/2, p38, and JNK was assessed using a 1:1,000 dilution of anti-phosphospecific antibodies (New England Biolabs). Inactivated GSK-3␤ was detected using a 1:1,000 dilution of an antibody that recognizes GSK-3␤ phosphorylated at Ser 9 . Activation of Raf-1 was monitored using a 1:2,000 dilution of an antibody that recognizes Raf-1 phosphorylated at Ser 338 (39, 40) (United Biotechnology Inc., PA). When whole cell or cytosolic extracts were used, blots were stripped and reprobed using a 1:1,000 dilution of antibodies against the nonphosphorylated forms of each protein to monitor protein loading. Levels of p65 were monitored in the nuclear extracts using a 1:1,000 dilution of an anti-N-terminal RelA antibody (Santa Cruz Biotechnology). Egr-1 was visualized in nuclear extracts using a 1:1,000 dilution of an anti-Egr-1 antibody (Santa Cruz Biotechnology). To ensure equal protein loading, blots with nuclear extracts were stripped and reprobed with a 1:1000 dilution of an anti-histone antibody (Santa Cruz Biotechnology).
Plasmids-pTF-LUC contains 2,106 bp of the human TF promoter. pTNF-␣-LUC contains 615 bp of the human TNF-␣ promoter, and pEgr-1-LUC contains 1,200 bp of the murine Egr-1 promoter. p(B) 5 -LUC contains five copies of an NF-B site, and p(AP-1) 4 -LUC contains four copies of an AP-1 site. These sites were cloned upstream of the minimal simian virus 40 (SV40) promoter expressing the firefly luciferase (LUC) reporter gene in pGL2-Promoter (Promega, Madison, WI) (47). A plasmid expressing dominant-negative (dnAkt) (S308A/S473A) was kindly provided by G. Bokoch (The Scripps Research Institute, La Jolla, CA). The control plasmid pFA-CMV expresses the GAL4 DNA binding domain alone and was obtained from Stratagene. pFR-Luc (pGAL4-LUC) contains 5XGAL4 binding sites upstream of a minimal promoter. pGAL4-p65 contains the transactivation domain (amino acids 386 -551) of p65 fused to the DNA binding portion of GAL4 (35). pcDNA3 (Invitrogen) was used as a control plasmid for transfections when expression plasmids were used.
Transfections-THP-1 cells were transfected using DEAE-dextran (14). After transfection, cells were incubated in complete medium for 46 h at 37°C before stimulating with 10 g/ml LPS for 5 h at 37°C. In some experiments cells were incubated with 100 nM wortmannin or 10 M LY294002 for 1 h at 37°C before stimulation with LPS. In other experiments cells were incubated with 50 mM LiCl or 50 mM NaCl for 1 h at 37°C before stimulation with LPS. Cell lysates were assayed for luciferase activity as described in the manufacturer's protocol (Promega) using a Monolight 2010 luminometer (Analytical Luminescence Laboratory, San Diego, CA). Cells were cotransfected with pRLTK, which expresses Renilla luciferase (Promega). Renilla luciferase was measured according to the manufacturer's protocol (Promega) and used to normalize the activity of the firefly luciferase. Data Analysis-The number of experiments analyzed is indicated in each figure. Band intensity was quantified by densitometric analyses using a Personal Densitometer (Molecular Dynamics, Sunnyvale, CA) and ImageQuant software. Data were collected using a minimum of three experiments and used to calculate the mean Ϯ S.D. Statistical significance was calculated using an unpaired Student's t test and was considered significant at p values Յ 0.05.

The PI3K-Akt Pathway Inhibits LPS-induced TNF-␣ and TF Protein Expression in Human Monocytic
Cells-We examined the role of the PI3K-Akt pathway in LPS induction of TNF-␣ and TF in human monocytic cells using two different pharmacological inhibitors (LY294002 and wortmannin) that block the activation of PI3K by different mechanisms. LPS induction of TNF-␣ and TF in PBMCs was measured in the presence and absence of LY294002 (Fig. 1A). LY294002 enhanced LPS induction of TNF-␣ and TF expression 2.0-and 3.3-fold, respectively. THP-1 cells represent a well established model of human monocytes. LY294002 also enhanced LPS induction of TNF-␣ and TF expression in THP-1 cells 2.8-and 2.0-fold, respectively (Fig. 1B). Similar results were observed in THP-1 cells with wortmannin (Fig. 1B). These results indicate that the FIG. 4. LPS-induced phosphorylation of various kinases in the presence of either LY294002 or LiCl. A, THP-1 cells were preincubated with either 10 M LY294002 or vehicle for 60 min before stimulation with 10 g/ml LPS for the various times as indicated. Whole cell lysates were prepared, and phospho-Akt levels were measured by Western blotting using an anti-phospho-Ser 473 antibody. Whole cell lysates (B-D) or cytosolic extracts (E and F) were analyzed by Western blotting using an anti-phospho-Ser 338 Raf-1 antibody (B), an anti-phospho-p38 antibody (C), an anti-phospho-Ser 9 GSK-3␤ antibody (D), an anti-phospho-ERK1/2 antibody (E), and an anti-phospho-JNK antibody (F). THP-1 cells were preincubated without or with 50 nM LiCl for 60 min before stimulation with 10 g/ml LPS for various times. Whole cell lysates were analyzed by Western blotting using an anti-phospho-Ser 9 GSK-3␤ antibody (G). Each blot was stripped and reprobed with an antibody against the nonphosphorylated form of each kinase to monitor loading. The blots are representative of three independent experiments.

FIG. 5. Inhibition of PI3K enhances LPS-induced DNA binding of NF-B and AP-1 and Egr-1 expression.
A, THP-1 cells were preincubated with either 10 M LY294002 or vehicle for 60 min before stimulation with 10 g/ml LPS for the various times indicated. p65 was detected in nuclear extracts by Western blotting using the N-terminal anti-p65 antibody. The blots were stripped and reprobed with antihistone antibody to monitor loading. B, electrophoretic mobility shift assays were performed by incubating nuclear extracts with doublestranded radiolabeled oligonucleotide containing an NF-B site, an AP-1 site, or an Sp1 site. C, THP-1 cells were preincubated with either 50 nM LiCl or vehicle for 60 min before stimulation with LPS. Levels of nuclear NF-B and Sp1 were determined by ELISA. D, total RNA was extracted from THP-1 cells preincubated with either 10 M LY294002 or vehicle for 60 min before stimulation with LPS at 37°C for various times. Egr-1 mRNA levels were determined by Northern blotting. The membrane was reprobed with a glyceraldehyde 3-phosphate dehydrogenase (G3PDH) probe to assess loading. Egr-1 protein levels were analyzed at 2 h by Western blotting using an anti-Egr-1 antibody.
PI3K-Akt pathway is a negative regulator of LPS-induced TNF-␣ and TF expression in human monocytic cells. Because the response to LY294002 was similar in PBMCs and THP-1 cells, we used THP-1 cells to determine further the mechanism by which activation of PI3K-Akt negatively regulates LPSinduced TNF-␣ and TF expression.

GSK-3␤ Positively Regulates TNF-␣ and TF Protein Expression in LPS-stimulated THP-1 Cells-GSK-3␤
is a kinase that is negatively regulated by Akt-dependent phosphorylation. However, the role of GSK-3␤ in the regulation of inflammatory genes is currently unknown. We wished to determine whether GSK-3␤ played a role in LPS induction of TNF-␣ and TF gene expression. We used LiCl to inhibit GSK-3␤ kinase activity (41,42). LiCl reduced LPS-induced TNF-␣ and TF expression in a dose-dependent manner (Fig. 1C). NaCl was used as an osmolarity control. These results indicate that GSK-3␤ positively regulates TNF-␣ and TF expression in LPS-stimulated monocytic cells and may be a potential target for negative regulation by the PI3K-Akt signaling pathway.

Activation of the PI3K-Akt Pathway Negatively Regulates LPS Induction of TNF-␣ and TF Gene Expression-
The role of the PI3K-Akt pathway in the LPS induction of TNF-␣ and TF mRNA expression was determined by Northern blot analysis (Fig. 2). LPS induced maximal levels of TNF-␣ and TF mRNAs at 1 h. LY294002 enhanced the LPS induction of TNF-␣ at 1 h by 4.2-fold. However, LY294002 had only a minor effect on LPS-induced levels of TF mRNA at 1 h but increased TF mRNA expression at 2 h by 5.8-fold. The slower migrating band represents a differentially spliced, nonfunctional TF transcript (48). These results indicate that activation of PI3K limits the LPS induction of TNF-␣ and TF mRNA expression in monocytic cells.
Next, we evaluated the role of PI3K-Akt signaling in the LPS-induced TNF-␣ and TF promoter activity. Wortmannin significantly enhanced LPS-induced TNF-␣ (p ϭ 0.0008) and TF (p ϭ 0.037) promoter activity (Fig. 3A). Importantly, cotransfection of cells with a plasmid expressing dnAkt also enhanced LPS induction of TNF-␣ and TF promoter activity  (Fig. 3B). The results using a pharmacological inhibitor and dnAkt indicate that LPS-induced TNF-␣ and TF gene expression is negatively regulated by activation of Akt at the level of transcription.
Inhibition of PI3K Blocks GSK-3␤ Inactivation-Because inactivation of GSK-3␤ by LiCl inhibited TNF-␣ and TF gene expression, we monitored activation of GSK-3␤ in the same extracts used to study the activation of the MAPKs. LPS induced a time-dependent phosphorylation of GSK-3␤ on Ser 9 (inactivation) which correlated with the kinetics of Akt activation by LPS (Fig. 4D). Preincubation of cells with LY294002 completely abrogated LPS-induced inactivation of GSK-3␤, thereby retaining the kinase in its active (dephosphorylated) state (Fig. 4D). These results demonstrated that GSK-3␤ is inactivated in a PI3K-dependent manner in LPS-stimulated THP-1 cells. Inactivation of GSK-3␤ may limit the LPS induction of TNF-␣ and TF expression.
To ascertain further the role of GSK-3␤ in LPS-stimulated TNF-␣ and TF expression in THP-1 cells, we used LiCl to inhibit GSK-3␤. Pretreatment of cells with 50 mM LiCl, a dose at which TNF-␣ and TF expression was significantly inhibited, had no effect on LPS-induced activation of ERK1/2, p38, or JNK MAPK (not shown). In contrast, LiCl enhanced LPSinduced phosphorylation and inactivation of GSK-3␤ (Fig. 4G). These results indicate that the inhibitory effect of LiCl on LPS-induced TNF-␣ and TF expression is mediated by inactivation of GSK-3␤ and is not mediated by effects on the MAPK pathways.
Inhibition of PI3K Enhances LPS-induced DNA Binding of NF-B and AP-1-It is well documented that NF-B/Rel and AP-1 transcription factors play a major role in the LPS induction of TNF-␣ and TF expression in human monocytic cells (3,12,14). First, we examined the effect of LPS on nuclear translocation of p65 and NF-B in the presence and absence of LY294002. The level of p65 was determined in nuclear extracts by Western blot analysis (Fig. 5A). LPS induced nuclear translocation of p65 that was maximal at 1 h. LY294002 enhanced LPS-stimulated p65/Rel nuclear translocation. Next, we examined nuclear translocation of NF-B by electrophoretic mobility shift assays. Similar to the Western blot analysis, LPS induced nuclear translocation and binding of NF-B with a peak at 1 h (Fig. 5B). NF-B binding was increased when cells were pretreated with LY294002 before stimulation with LPS (Fig. 5B). Preincubation of cells with LY294002 also enhanced LPS-stimulated AP-1 binding at 2 and 4 h (Fig. 5B). Inhibition of PI3K had a small stimulatory effect on both basal NF-B and AP-1 binding (Fig. 5B). Binding of Sp1 to a prototypic site was used as a control to show that the different nuclear extracts had similar amounts of protein. However, LiCl did not affect LPSinduced nuclear translocation of NF-B (Fig. 5C). Binding of Sp1 was used as a control.
Inhibition of PI3K-Akt Enhances LPS-induced Egr-1 Expression-We and others have demonstrated that the TNF-␣ and TF promoters are regulated in LPS-stimulated human monocytic cells by the coordinated binding of NF-B, AP-1, and Egr-1 transcription factors (13,15). Because Egr-1 is synthesized de novo in LPS-stimulated monocytic cells, we evaluated the effect of LY294002 on LPS-induced Egr-1 gene expression. LY294002 enhanced LPS-induced Egr-1 mRNA expression at 1 h by 2.1-fold (Fig. 5D). Preincubation with LY294002 also increased LPS-induced Egr-1 protein expression at 2 h (Fig.   FIG. 7. LPS-induced transcriptional activity of p65 is regulated by GSK-3␤. A, THP-1 cells were cotransfected with 5 g of pGAL4-p65TA and 1.5 g of pGAL4-LUC. 46 h post-transfection cells were preincubated with either control or with 100 nM wortmannin for 60 min before stimulation with 10 g/ml LPS for 5 h at 37°C. Luciferase activity in cell lysates was determined, and results are expressed as -fold induction. B, THP-1 cells were transfected as above. 46 h posttransfection cells were preincubated with either 50 mM LiCl, 50 mM NaCl, or left untreated (control) before stimulation with LPS for 5 h at 37°C. Luciferase activity in cell lysates was determined, and results are expressed as -fold induction. Data (mean Ϯ S.D.) are shown for three independent experiments. 5D). These results demonstrate that the negative regulation of the Raf-1-MEK-ERK1/2 pathway by LPS-induced activation of Akt also limits Egr-1 expression, a target gene of the pathway.
LPS-induced Transcriptional Activity of p65 Is Negatively Regulated by the PI3K-Akt Pathway via Inactivation of GSK-3␤-We have shown that inhibition of the PI3K-Akt pathway blocks the Akt-dependent inhibitory phosphorylation of GSK-3␤, whereas LiCl increases the inactivation of GSK-3␤. We determined whether the level of activation of GSK-3␤ correlated with the transcriptional activity of p65 using a GAL4-p65 chimeric protein that contains the transactivation domain of p65 (Fig. 7). THP-1 cells were cotransfected with the reporter plasmid pGAL4-Luc and the expression plasmid pGAL4-p65. Wortmannin enhanced LPS induction of the transcriptional activity of p65. In contrast, inactivation of GSK-3␤ by LiCl reduced p65-dependent transcription. Treatment of cells with an equimolar concentration of NaCl had no effect on LPSinduced luciferase activity and served as a control for osmolarity. These results demonstrate that LPS-induced activation of the PI3K-Akt pathway inactivates GSK-3␤ and reduces the transactivation activity of p65. DISCUSSION Our study demonstrates that LPS-induced expression of TNF-␣ and TF in human monocytic cells is regulated by both positive and negative pathways. We provide multiple lines of evidence and new insights to support the contention that LPSinduced activation of the PI3K-Akt pathway produces a "limiting" effect on TNF-␣ and TF gene expression in PBMCs and THP-1 monocytic cells. We show that LPS-induced activation of PI3K-Akt negatively regulates the transcription factors Egr-1, AP-1, and NF-B. The net inhibitory effect on the activation of all three transcription factors reduces LPS induction of TNF-␣ and TF expression in monocytic cells, which is a key cell type in sepsis.
This study demonstrates that LPS-induced activation of PI3K-Akt in monocytic cells negatively regulates Raf-1. Two targets of the Raf-1 pathway (ERK1/2 and Egr-1) were also negatively regulated by LPS-induced activation of PI3K-Akt. Egr-1 is required for maximal induction of both TNF-␣ and TF in human monocytic cells treated with LPS (13,15). Enhancement of LPS-induced TF mRNA expression in the presence of LY294002 was delayed (2 h) relative to the enhancement of TNF-␣ mRNA expression (1 h). This difference may be the result of a greater contribution of Egr-1 to the induction of TF gene expression because the TF promoter contains three Egr-1 sites. We also found that inhibition of PI3K enhanced LPSinduced activation of p38 and JNK. Our data are consistent with a recent study demonstrating that inhibition of the PI3K-Akt pathway in endothelial cells enhanced TF expression by increasing the activation of p38 (27). We found that LY294002 and dnAkt increased LPS-induced AP-1-dependent gene expression. Gratton and colleagues (49) have shown recently that Akt-dependent phosphorylation of MEKK3 reduces its kinase activity and inhibits the MEKK3/6-p38 pathway. These data suggest that Akt can negatively regulate multiple signaling pathways.
Studies on the role of the PI3K-Akt pathway in NF-B-dependent gene expression are controversial. The PI3K-Akt pathway has been shown to act both positively and negatively on NF-B-dependent gene expression. These differences may reflect the use of different cell types and different agonists. In addition, overexpression of a constitutively active form of Akt may override normal regulatory pathways. Our study demonstrates that the PI3K-Akt pathway negatively regulates NF-B in LPS-stimulated monocytic cells. Inhibition of PI3K-Akt enhanced LPS-induced nuclear translocation of p65, increased NF-B binding, and increased NF-B-dependent gene expression.
Recent studies showed that both TLR2 and TLR4 signaling activates the PI3K-Akt pathway in human monocytic cells and macrophages (35,50). We show that LPS-TLR4 signaling in human monocytic cells activates Akt. Importantly, macrophages exhibited different patterns of cellular responses after stimulation with TLR2 and TLR4 agonists, which suggested that different intracellular signaling pathways were activated by TLR2 and TLR4 (50 -54). In human monocytic cells, TLR2dependent activation of the PI3K-Akt pathway positively regulated the transactivational activity of p65 (35). In contrast, we show that TLR4-dependent activation of the PI3K-Akt pathway negatively regulates the transactivational activity of p65. These differences probably reflect the activation of different signaling pathways that modulate the effect of the PI3K-Akt pathway on the transactivational activity of p65.
The increased transactivational activity of p65 in the presence of wortmannin correlated with the inhibition of LPSinduced, Akt-dependent inactivation of GSK-3␤. In parallel, we showed that inhibition of GSK-3␤ with LiCl reduced LPS induction of TNF-␣ and TF in monocytic cells, suggesting that GSK-3␤ positively regulates these genes by increasing NF-B activity. LiCl did not affect LPS-induced nuclear translocation of NF-B but decreased the transactivational activity of p65. Therefore, inhibition of GSK-3␤ at later times via Akt-dependent phosphorylation may represent at least one mechanism by which monocytic cells limit the expression of NF-B-dependent genes.
Several kinases have been implicated in the control of p65 transcriptional activity, but the most compelling data are derived from studies of various knockout mice. GSK-3␤ and T2K are both necessary for TNF-␣ and IL-1 signaling, whereas NIK is selective to the LT-␤R pathway (42,(55)(56)(57). Interestingly, protein kinase C deficiency impairs p65 transcriptional activity in response to TNF-␣, interleukin-1, and lymphotoxin-␤, which suggests that protein kinase C may be downstream of GSK-3␤, NF-B-inducing kinase, and T2K (58). Indeed, protein kinase C is the only kinase that has been shown to interact directly with p65.
A model of LPS induction of TNF-␣ and TF in monocytic cells is shown in Fig. 8. The current study demonstrates that LPS stimulation of monocytic cells leads to an activation of the PI3K-Akt pathway, which inactivates MAPK pathways (ERK1/2, p38, and JNK) and the NF-B pathway by phosphorylation of Raf-1, I-B kinase, GSK-3␤, and other upstream kinases, such as MEKK3. Inhibition of these pathways limits the activation of the transcription factors NF-B, AP-1, and Egr-1, all of which cooperatively regulate TNF-␣ and TF gene expression. Thus, the PI3K-Akt pathway imposes a "braking" mechanism to limit the expression of TNF-␣ and TF in LPSstimulated monocytes and ensure transient expression of these inflammatory mediators.