Stimulation of NFκB Activity by Multiple Signaling Pathways Requires PAK1

The p21-activated kinase (PAK1) is a serine-threonine protein kinase that is activated by binding to the Rho family small G proteins Rac and Cdc42hs. Both Rac and Cdc42hs have been shown to regulate the activity of the transcription factor NFκB. Here we show that expression of active Ras, Raf-1, or Rac1 in fibroblasts stimulates NFκB in a PAK1-dependent manner and that expression of active PAK1 can stimulate NFκB on its own. Similarly, in macrophages activation of NFκB as well as transcription from the tumor necrosis factor α promoter depends on PAK1. In these cells lipopolysaccharide is a potent activator of PAK1 kinase activity. We also demonstrate that expression of active PAK1 stimulates the nuclear translocation of the p65 subunit of NFκB but does not activate the inhibitor of κB kinases α or β. These data demonstrate that PAK1 is a crucial signaling molecule involved in NFκB activation by multiple stimuli.

NFB 1 is a transcription factor that is critically involved in cellular growth and transformation, the suppression of apoptosis, and the response to inflammatory stimuli (1)(2)(3)(4)(5)(6). It consists of homo-and heterodimers of members of the Rel family of transcription factors (1,2). The most frequently studied form of NFB consists of two proteins, p50 (NFB1) and p65 (RelA). In unstimulated cells, this heterodimer is retained in the cytoplasm by an inhibitory protein known as the inhibitor of B (IB). In response to most stimuli that activate NFB, IB becomes phosphorylated, ubiquitinated, and subsequently degraded by the 26 S proteosome (7,8). Once free of IB, NFB translocates to the nucleus and activates the transcription of target genes. The transcriptional activity of NFB is also controlled by phosphorylation of dimer subunits. For example, both p50 and p65 are phosphorylated in cells, and phosphorylation of p65 has been shown to positively regulate its transcriptional activity (9 -14).
The phosphorylation of IB that leads to its degradation occurs on two conserved serine residues within its N terminus (serines 32 and 36 in IB␣) (15). Two kinases have been identified and cloned that will phosphorylate both of these sites in cells, namely IB kinase ␣ and IB kinase ␤ (IKK␣ and IKK␤, respectively) (16 -20). Based on knockout studies in mice, IKK␤ seems to be more important than IKK␣ for controlling NFB activity in response to cytokines and other ligands (21)(22)(23)(24)(25). Both of these kinases are present in a high molecular weight protein complex that contains at least two distinct scaffolding proteins (IKK␥ and IKK complex-associated protein), as well as NFB, IB, and other proteins (26 -29). The activities of IKK␣ and IKK␤ are controlled by the related kinases NFB-interacting kinase (NIK) and the mitogen-activated protein/extracellular signal-regulated kinase kinase kinases 1-3 (MEKK1-3) (30 -36). Other kinases that control IKK activity may yet be identified. Recent studies have shown that two upstream regulators of NIK are the transforming growth factor-␤-activated kinase 1 and Cot/Tpl2, whereas MEKK1 has been shown to be regulated by the viral protein Tax (37)(38)(39)(40). The relative importance of each of these pathways to the regulation of IKK activity in response to different stimuli remains unclear.
Pro-inflammatory cytokines such as interleukin 1␤ (IL1␤) and tumor necrosis factor ␣ (TNF␣), as well as bacterial endotoxins such as lipopolysaccharide (LPS), signal to the IKK complex through the activation of specific receptors on the plasma membrane. In the case of IL1␤, its receptor (IL1␤RI) signals to an associated complex consisting of the IL1␤-accessory protein, MyD88, and two related interleukin 1 receptorassociated kinase proteins (41,42). The activated receptor complex signals to the IKK complex through a scaffolding protein called TRAF6 (43,44). LPS induces signaling following binding to a glycosyl phosphatidylinositol-linked, membrane-associated protein called CD14. Recently, a member of the toll family of receptors known as toll receptor 4 was identified as a receptor for LPS (45). Another toll family member, toll receptor 2, has also been identified as possible LPS receptor (46). These receptors have significant homology to the IL1␤RI and in fact require many of the known IL1␤RI-associated molecules for efficient signaling (47)(48)(49)(50).
Among the first cells to become activated following exposure to LPS are macrophages. Once activated, macrophages secrete pro-inflammatory cytokines such as IL1␤, TNF␣, and IL6. The signaling pathways required for these events are not well de-fined but are known to result in the activation of three different MAPK cascades as well as an increase in NFB activity (51). It is also clear that activation of NFB is necessary for transcription of the genes encoding these cytokines (48).
Recent studies have shown that the Rho family small G proteins Rac1, Cdc42hrs, and RhoA are capable of activating NFB in various cell types (52)(53)(54)(55). In addition, expression of dominant negative forms of Rac1 and RhoA block NFB activation by IL1␤, TNF␣, and bradykinin. These observations indicate that Rho family small G proteins are involved in regulating NFB activation following cytokine stimulation.
All small G proteins mediate signaling through the activation of specific effector proteins. Both Rac1 and Cdc42hs share a number of common effectors, including the p21-activated kinases (PAKs) (56). To date four PAKs have been cloned, and in the case of PAKs 1-3, binding of GTP-liganded Rac or Cdc42hs promotes their activation by relieving an autoinhibitory constraint (56 -59). Once activated, a number of cellular phenotypes have been attributed to PAK activity, including activation of the extracellular signal-regulated kinase (ERK), JNK, and p38 MAPK cascades (PAKs 1-4), regulation of cytoskeletal organization (PAKs 1, 2, and 4), and regulation of apoptosis (PAK2) (56, 58 -60). A number of extracellular stimuli activate PAK1, including exposure to IL1␤ in epithelial cells and T cell receptor ligation in T cells (61,62). This suggests that PAK1 may play a role in the immune response.
In this study we show that PAK1 mediates NFB activation by Ras, Raf-1, and Rac1 and that expression of an active form of PAK1 is capable of stimulating NFB activity on its own. In addition, we show that active PAK1 stimulates the nuclear translocation of the p65 subunit of NFB in the apparent absence of IKK␣ or IKK␤ activation. We also demonstrate that in mouse macrophages, PAK1 activity is stimulated by LPS and is required for efficient NFB activation and TNF␣ transcription. These results identify PAK1 as an important regulator of NFB in both fibroblasts and macrophages.
Cell Culture and Transfections-NIH 3T3 cells were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% calf serum, 1% L-glutamine, and 100 units/ml penicillin/streptomycin. HEK 293 cells (293 cells) and REF52 cells were grown in DMEM supplemented with 10% fetal bovine serum, 1% L-glutamine, and 100 units/ml penicillin/streptomycin. Prior to transfection, the cells were placed in fresh growth medium. Both cell lines were transfected by calcium phosphate precipitation (63). Twenty hours after transfection, the medium was replaced with DMEM plus 0.5% calf serum (NIH 3T3 cells) or DMEM without serum (293 cells). The cells were then allowed to incubate for another 24 h. RAW 264.7 cells were grown in DMEM containing 10% fetal bovine serum (endotoxin-free), 50 units/ml penicillin, 50 g/ml streptomycin, 2 mM L-glutamine. These cells were transfected using the Profection DEAE-dextran transfection system (Promega) according to the manufacturer's protocol. Twenty-four hours after transfection, the cells were treated with LPS (1 g/ml) or diluent for 6 h and then harvested.
Reporter Assays-Transfected cells were washed once with cold phosphate-buffered saline (PBS) and scraped into luciferase lysis buffer (50 mM Tris-HCl (pH 8.0), 70 mM K 2 HPO 4 , 0.1% Nonidet P-40, 2 mM MgCl 2 , 1 mM dithiothreitol, 20 g/ml aprotinin, 10 g/ml pepstatin A, and 10 g/ml leupeptin). The lysates were rapidly mixed for 10 s, and insoluble material was pelleted by centrifugation at 4°C. The supernatant was removed and either assayed immediately or flash frozen in liquid nitrogen and stored at Ϫ80°C. Firefly luciferase assays were performed according to the manufacturer's protocol using a Luciferase Assay kit (Promega) and a Turner luminometer. For promoter activation assays in NIH3T3 cells, transfection efficiency was monitored by assaying for ␤-galactosidase activity derived from a cotransfected, constitutively active ␤-galactosidase expression plasmid. ␤-Galactosidase activity was determined essentially as described (65). In RAW 264.7 cells, transfection efficiency was determined by measuring the activity of expressed Renilla firefly luciferase derived from the transfection of pRL-TK (Promega). Renilla luciferase assays were performed at the same time as the firefly luciferase assays using a Dual Luciferase Assay kit (Promega) and the Turner luminometer. Chloramphenicol acetyltransferase assays were performed as described previously (51).
Immunoprecipitation and Kinase Assays-To assay the activity of transfected IKK␣ or IKK␤, 293 cells were washed once with cold PBS and lysed with Triton lysis buffer (20 mM Tris-HCl (pH 7.5), 100 mM NaCl, 0.5% Triton X-100, 80 mM ␤-glycerophosphate, 0.5 mM sodium orthovanadate, 1 mM EDTA, 20 g/ml aprotinin, 10 g/ml pepstatin A, 10 g/ml leupeptin, 1 mM phenylmethylsulfonyl fluoride). Insoluble material was pelleted by centrifugation, and supernatants were removed, flash-frozen in liquid nitrogen, and stored at Ϫ80°C. Equal amounts of epitope-tagged IKK proteins were used for immunoprecipitation based on prior immunoblotting. IKKs were immunoprecipitated by incubation for 2 h at 4°C with antibodies directed against the appropriate epitope tags and protein A-Sepharose (Amersham Pharmacia Biotech). Each immunoprecipitate was washed three times with 20 mM Tris-HCl (pH 7.5), 500 mM NaCl, and once with 20 mM Tris-HCl (pH 8.0) and was divided in two parts. One-half was tested for immunoprecipitated IKK␣ or IKK␤ protein by immunoblotting. The other half was assayed for kinase activity toward GST-IB-(1-54). Kinase assays were performed at 30°C for 30 min in buffer containing 20 mM Tris-HCl (pH 7.5), 10 mM MgCl 2 , 1 mM dithiothreitol, 100 M ATP, 3 M GST-IB-(1-54), and 4500 cpm/pmol [␥-32 P]ATP. Reactions were stopped by adding Laemmli sample buffer and were resolved by SDS-polyacrylamide gel electrophoresis followed by Coomassie Blue staining. Phosphate incorporated into GST-IB-(1-54) was measured by excising the Coomassie Blue-stained bands and counting by liquid scintillation.
For assaying endogenous PAK1 activity in RAW 264.7 cells, the cells were stimulated with 1 g/ml LPS for increasing amounts of time, washed once with cold PBS, and lysed in Triton lysis buffer. PAK1 was immunoprecipitated from 1 mg of cell lysate/sample by incubating with 1 g of rabbit anti-PAK␣ antibody (N20) and protein A-Sepharose for at least 2 h at 4°C on a rotating platform. The immunoprecipitates were then pelleted by centrifugation and washed three times with 20 mM Tris-HCl (pH 7.5), 1 M NaCl and once with 50 mM Tris-HCl (pH 8.0), 1 mM dithiothreitol, 10 mM MgCl 2 . The immunoprecipitate was assayed for kinase activity using myelin basic protein (MBP) as a substrate. MBP kinase assays were carried out essentially as described (59). After the kinase reaction the pellet containing the immunoprecipitated PAK1 was solubilized with Laemmli sample buffer, subjected to SDS-polyacrylamide gel electrophoresis, and transferred to nitrocellulose. Immunoprecipitated PAK1 was detected by Western blotting using the rabbit anti-PAK1 antibody (N-20).
Microinjection Assays-REF52 cells were plated on glass coverslips in 10% fetal bovine serum and injected with expression vectors for p65 and Myc-epitope-tagged versions of either V 12 Rac1, PAK1 165, or wt NIK. All expression vectors were injected into the nucleus at 0.25 mg/ml using an Eppendorf 5171 microinjector mounted on a Zeiss Axiophot S135 inverted microscope. Either 4 or 16 h after injection the cells were fixed with 3.7% formaldehyde in PBS for 5 min at 37°C and permeabilized with 0.2% Triton X-100 in PBS for 10 min at room temperature. Expressed p65 RelA was detected with rabbit anti-p65 diluted to 2 g/ml in PBS ϩ 0.05% Tween 20 (PBST). Myc-tagged proteins were detected with mouse anti-Myc diluted to 0.4 g/ml in PBST. Fixed cells were incubated with these antibodies at 37°C for 1 h followed by three 5-min washes with PBST. The cells were then incubated with rhodamine-conjugated donkey anti-rabbit (Jackson Labs) and fluorescein isothiocyanate-conjugated donkey anti-mouse for 1 h at 37°C followed by three 5-min washes with PBST and once with distilled water. Epiflouresence was detected with a Zeiss Axiovert S135 microscope fitted with a Photometrics cooled CCD camera.

RESULTS
PAK1 Activity Is Required for NFB Activation by Ras, Raf-1, and Rac1-Ras, Raf-1, and Rac1 have all been shown to stimulate NFB activity (53,66,67). In the case of Ras and Raf-1, this activation may to be critical to their ability to promote cell growth. On the other hand, NFB activation by Rho family small G proteins has been suggested to be important for the response to pro-inflammatory cytokines. Because PAK1 is a Rac and Cdc42hs effector that is involved in the Ras-dependent activation of Raf-1, we tested whether PAK1 activity is required for activation of NFB by these molecules. Thus, NIH 3T3 cells were cotransfected with an NFB-luciferase reporter plasmid, active forms of Ras (V 12 H-Ras), Raf-1 (Raf BXB), or Rac1 (V 12 Rac1), and either a control vector or dominant negative (dn) PAK1. This dominant negative form of PAK1 consists of the catalytic domain of PAK1 (amino acids 232-544) containing a point mutation that renders it inactive (K298A). Because it lacks the N-terminal regulatory domain, it cannot bind to Rac1 or Cdc42hs. As shown in Fig. 1A, expression of active Ras, Raf-1, or Rac1 stimulated NFB activity in these cells. In the case of Ras, NFB activation was only partially blocked by dominant negative PAK1. Because Ras activates NFB through multiple signaling pathways, this likely indicates that only one of these pathways relies on PAK1 (66). On the other hand, dominant negative PAK1 was more effective at blocking NFB activation by active Rac1 and completely blocked activation by active Raf-1. This suggests that PAK1 plays a more important role in NFB activation stimulated by Raf-1 and Rac1. For all three proteins, dominant negative PAK1 blocked NFB activation slightly less well than dominant negative NIK, and coexpression of dominant negative NIK and dominant negative PAK1 did not result in a greater degree of inhibition of NFB activation (data not shown). Because NIK phosphorylates and activates the IB kinases IKK␣ and IKK␤, this suggests that NFB activation by Ras, Raf-1, or Rac1 ultimately depends on the phosphorylation of IB.
To test if PAK1 functions downstream of proteins known to mediate NFB activation by cytokines, dominant negative PAK1 was coexpressed with increasing amounts of the adaptor proteins TRAF2 and TRAF6. TRAFs link cytokine receptor activation to the IB kinase activation cascade, and overexpression of wild type TRAF proteins stimulates NFB activity (43,44,68,69). As shown in Fig. 1B, dominant negative PAK1 did not significantly inhibit NFB activation by TRAF2 or TRAF6. This fits with the observation that expression of these TRAF proteins in cells does not activate PAK1. 2 Consistent with previous observations, dominant negative NIK partially inhibited NFB activation by TRAF2 (30,70). Taken together, these data indicate that PAK1 activity is required for NFB activation stimulated by active Raf-1 or Rac1 but is not involved in NFB activation stimulated by TRAF2 or TRAF6 overexpression.
PAK1 Stimulates NFB, but Does Not Activate Either IKK␣ or IKK␤-To test whether expression of active PAK1 can stimulate NFB activity, NIH3T3 cells were cotransfected with the NFB reporter and a constitutively active, N-terminal truncation mutant of PAK1 (PAK1 165) (59). As shown in Fig. 2A, the expression of active PAK1 stimulated NFB activity to levels comparable to those stimulated by active Rac1 (Fig. 1A). This activation was specific for NFB, because PAK1 165 did not significantly activate an activator protein-1-luciferase reporter construct (data not shown). NFB activation depended on PAK1 kinase activity, because the kinase-inactive version of this protein did not stimulate NFB activity (data not shown). This suggests that PAK1 selectively activates NFB. In addition, activation of NFB by active PAK1 was inhibited by coexpression of dominant negative forms of NIK, IKK␣, and IKK␤, suggesting that PAK1-mediated NFB activation depended on the phosphorylation of IB. On the other hand, NFB activation stimulated by active PAK1 was not inhibited by two different dominant negative forms of MEKK1 (Fig. 2A) 1. A, activation of NFB by active Ras, Raf-1, or Rac1 requires PAK1 activity. NIH 3T3 cells were cotransfected with an NFB-luciferase reporter vector and either a control vector or constitutively active Ras (V 12 H-Ras), active Raf-1 (Raf BXB), or active Rac1 (V 12 Rac1). The cells were also cotransfected with dominant negative (dn) versions of PAK1 or NIK. Twenty hours after transfection, the cells were placed in 0.5% calf serum and allowed to incubate for another 24 h. The cells were then harvested and assayed for luciferase activity. Transfection efficiency was monitored by assaying for ␤-galactosidase activity derived from a cotransfected, constitutively active ␤-galactosidase expression vector. Shown is the average of four independent experiments. Error bars denote the standard error of the mean. Fold activation refers to the increase in luciferase activity over that found in cells transfected with the reporter and a control plasmid. B, NIH 3T3 cells were cotransfected with the NFB-luciferase reporter and increasing amounts of either wild type TRAF2 or wild type TRAF6. The cells were also cotransfected with either a control vector, dn PAK1, or dn NIK. NFB activation was determined as described (A). TRAF2 or TRAF6 (Fig. 1B), we tested whether activation of NFB by PAK1 represents a separate pathway leading to NFB activation. If so, coexpression of active PAK1 with TRAF2 or TRAF6 should activate NFB to a greater degree than expression of either protein alone. As shown in Fig. 2B, coexpression of constitutively active PAK1 with increasing amounts of TRAF2 or TRAF6 stimulated a higher level of NFB activity than that observed with either TRAF protein alone. This is consistent with data showing that dominant negative PAK1 expression does not block NFB activation stimulated by TRAF2 or TRAF6 (Fig. 1B) and suggests that PAK1 activates NFB by a pathway distinct from that activated by TRAF2 or TRAF6.
Because NFB activation by PAK1 was blocked by the expression of dominant negative forms of NIK, IKK␣, and IKK␤, we examined whether PAK1 stimulated the kinase activity of either IKK␣ or IKK␤. Thus, 293 cells were cotransfected with epitope-tagged forms of IKK␣ or IKK␤ and either a control vector, active PAK1 (PAK1 165), active Rac1 (V 12 Rac1), wild type NIK (wt NIK), or wild type MEKK1 (wt MEKK1). The next day the cells were placed in starvation medium, and after 24 h they were lysed. IKK proteins were then immunoprecipitated and tested for kinase activity using bacterially expressed GST-IB-(1-54) as a substrate. As shown in Fig. 3, neither IKK␣ nor IKK␤ were activated by coexpression of active Rac1 or active PAK1. Under these circumstances, NIK and MEKK1 were strong activators of IKK␣ and IKK␤ activity, respectively. Similar results were found in mouse macrophages (data not shown). PAK1 165 expression also did not stimulate the activity of endogenous IKK␣ or IKK␤ in 293 cells (data not shown).
In separate experiments, coexpression of active Rac1 or PAK1 with epitope-tagged NIK did not increase the activity of the immunoprecipitated NIK toward recombinant IKK␣ or IKK␤ (data not shown). Furthermore, recombinant PAK1 did not phosphorylate recombinant IKK␣ or IKK␤ purified from Sf9 cells and did not stimulate the phosphorylation in vivo of either wild type or kinase-inactive NIK expressed in 293 cells (data not shown). Thus, neither Rac1 nor PAK1 appear to stimulate the activity of NIK, IKK␣, or IKK␤.
To test whether active PAK1 promotes nuclear translocation of NFB by a mechanism other than activation of IKK␣ or IKK␤, we microinjected REF52 fibroblasts with expression vectors for p65 (RelA) and active Rac1 (V 12 Rac1), active PAK1 (PAK1 165), or wild type NIK (wt NIK). p65 is the transcriptional activation component of the most common form of the NFB heterodimer (1,2). Four or sixteen hours after injection, the cells were fixed, and the cellular localization of the expressed p65 was determined by indirect immunofluorescence. As shown in Fig. 4A, expression of either active Rac1, active PAK1, or wild type NIK for 4 h stimulated the translocation of p65 to the nucleus. In the cells expressing active PAK1, this was accompanied by a pronounced retraction of the cell membrane, consistent with previously observed effects of PAK1 on cell morphoplogy (59). Determination of the percentage of cells showing nuclear staining for p65 showed that V 12 Rac1, PAK1 165, and wild type NIK were similarly capable of stimulating the nuclear translocation of p65 (Fig. 4C). At 16 h after injection, cells expressing active PAK1 or NIK still showed nuclear localization of p65 (Fig. 4, B and D). These data indicate that FIG. 2. A, constitutively active PAK1 stimulates NFB. NIH 3T3 cells were cotransfected with the NFB-luciferase reporter and either a control vector or constitutively active PAK1 (PAK1 165). The cells were also transfected with dn versions of NIK, the MEKK1 catalytic domain, full-length MEKK1, IKK␣, or IKK␤. Fold activation refers to the fold increase in luciferase activity as compared with cells cotransfected with the NFB-luciferase reporter and a control vector. Error bars represent the standard error of the mean from at least three independent experiments. B, active PAK1 potentiates NFB activation by wild type TRAF2 or wild type TRAF6. NIH 3T3 cells were cotransfected with the NFB-luciferase reporter, a control vector or PAK1 165, and increasing amounts of either TRAF2 or TRAF6. Activation of NFB was determined as described.
FIG. 3. PAK1 does not activate either IKK␣ or IKK␤. 293 cells were transfected with epitope-tagged, wild type IKK␣ (A) or IKK␤ (B) and either a control vector or increasing amounts of wild type NIK, PAK1 165, wild type MEKK1, or V 12 Rac1 (Fig. 3B only). After transfection the cells were serum-starved. Twenty-four hours later IKK proteins were immunoprecipitated from 0.5% Triton X-100 soluble cell lysates using the appropriate anti-epitope antibodies. IKK activities were examined by monitoring their kinase activity toward GST-IB- . Shown in the top panel of A and B are autoradiograms of phosphorylated GST-IB- . In the bottom panels are immunoblots for immunoprecipitated IKK␣ and IKK␤. Shown are representative experiments from at least three independent experiments. PAK1 stimulates the nuclear translocation of NFB, most likely in the absence of significant activation of IKK␣ or IKK␤ activity. Translocation of NFB to the nucleus in the absence of IKK activation has been observed previously (71)(72)(73). Thus, the finding that dominant negative forms of IKK␣ and IKK␤ are capable of blocking NFB activation by active PAK1 (Fig. 2A) may reflect the ability of these molecules to form tight complexes with NFB and thereby preclude its nuclear localization stimulated by other mechanisms.

PAK1 Functions within the LPS Signaling Pathway to Activate NFB-LPS is a potent activator of NFB in macrophages.
Because PAK1 is required for efficient NFB activation in fibroblasts, we tested whether endogenous PAK1 is activated by LPS in the mouse macrophage cell line RAW 264.7. As shown in Fig. 5, exposure of these cells to LPS led to a timedependent increase in PAK1 activity, as determined by the ability of immunoprecipitated PAK1 to phosphorylate MBP. This activity peaked at 30 min (22-fold) and was decreasing by 60 min (14-fold). These data indicate that LPS is an efficient activator of PAK1 in mouse macrophages.
To test whether PAK1 is required for stimulation of NFB activity by LPS, RAW 264.7 cells were cotransfected with the NFB-luciferase reporter and either a control vector, constitutively active PAK1 (PAK1 165), or dn PAK1. As shown in Fig.  6A, exposure to LPS stimulated NFB activity ϳ5-fold in vector-transfected cells. In nonstimulated cells, expression of PAK1 165 activated the NFB reporter to levels equivalent to those in LPS-stimulated cells (5-fold), and in the presence of LPS this activation was much greater (15-fold). In addition, expression of dominant negative PAK1 reduced NFB activation by LPS by half. These data indicate that PAK1 functions within the NFB activation pathway stimulated by LPS in these cells.
Exposure of macrophages to LPS activates the transcription of many cytokine genes in an NFB-dependent manner, including TNF␣ (51). To test whether PAK1 is involved in the regulation of a physiological promoter dependent on NFB activity, RAW 264.7 cells were cotransfected with a murine TNF␣ promoter-chloramphenicol acetyltransferase reporter construct. This TNF␣ promoter contains five NFB binding sites, one of which is similar to that found in our synthetic NFB-luciferase promoter, as well as binding sites for other transcription factors. These cells were also transfected with a control vector, constitutively active PAK1 (PAK1 165), or dn PAK1. As shown in Fig. 6B  cates that LPS-stimulated TNF␣ transcription requires PAK1 activity. Interestingly, expression of constitutively active PAK1 did not activate this promoter nor did it potentiate activation by LPS. This suggests that active PAK1 expression alone is not sufficient for stimulation of TNF␣ transcription. This lack of activation may reflect the fact that this promoter contains a number of enhancers in addition to the characterized NFB binding sites and that LPS-stimulated TNF␣ transcription requires the activation of one or more transcription factors in addition to NFB (51). DISCUSSION The data presented in this study demonstrate that expression of constitutively active PAK1 stimulates NFB activity in fibroblasts and macrophages. Furthermore, PAK1 activation is required for stimulation of NFB activity by Rac1 and Raf-1 in fibroblasts and by LPS in mouse macrophages. PAK1 appears to stimulate NFB activity independently of TRAF2 or TRAF6 and does not activate NIK, IKK␣, or IKK␤. Nevertheless, in microinjection experiments PAK1 stimulates the nuclear translocation of the p65 subunit of NFB. We have also shown LPS to be a potent activator of PAK1 in mouse macrophages and that PAK1 activation is necessary for full activation of the murine TNF␣ promoter. This indicates that PAK1 is an impor-tant regulator of NFB activity in multiple cell types.
NFB activation is controlled on multiple levels, including through regulation of its subcellular localization. In its inactive form, NFB is sequestered in the cytoplasm by the IB family of proteins. Most stimuli that activate NFB do so by stimulating the phosphorylation of IB on two sites within its N terminus that leads to its degradation, thereby allowing NFB to translocate to the nucleus and activate transcription. The IB kinases IKK␣ and IKK␤ phosphorylate these sites in response to most ligands (15). Thus, the finding that PAK1 stimulates the activity of endogenous NFB as well as the nuclear translocation of expressed p65 but does not activate either IKK␣ or IKK␤ is unusual. The mechanism by which PAK1 does this is not clear. It may be that PAK1 stimulates the activity of an IB kinase other than IKK␣ or IKK␤. It is also possible that PAK1 stimulates NFB translocation through an IKK-independent mechanism. In this regard, both UV-C exposure and treatment with pervanadate have been shown to stimulate NFB translocation in the absence of IKK activation (71)(72)(73). Additionally, it is possible that PAK1 functions in the NFB activation pathway in a manner loosely analogous to that of its homolog in yeast, STE20. In Saccharomyces cerevisiae, STE20 regulates pheromone-dependent mitogen-activated protein kinase activation by a complex mechanism that depends on the scaffolding protein STE5 as well as the MAP3K STE11 (74). In a similar manner, PAK1 may regulate the association of NIK, the IKKs, IB, or NFB with the scaffolding proteins IKK complex-associated protein or IKK␥. In this regard, we have found that expression of active PAK1 reduces the coprecipitation of IKK␤ with NIK from cells. 2 Thus, perhaps PAK1 stimulates NFB activity by altering the kinetics of association between NFB-activating components in the cell. If this were the case, one could envision how expression of dominant interfering forms of NIK or an IKK might block NFB activation by PAK1, because their expression would affect the association of the endogenous kinases with the NFB scaffold. Future studies will be directed at determining the mechanism by which PAK1 stimulates NFB activation.
The requirement for PAK1 activity in the activation of NFB by Raf-1, and to a lesser degree Ras, may have implications for cellular transformation. An increasing body of evidence suggests that NFB activation is crucial for cellular transformation. For example, Ras-dependent transformation of fibroblasts is inhibited by the expression of dominant negative IB (6,75). Similarly, transformation of Rat-1 fibroblasts by Ras also requires PAK1 activity (76). Thus, activation of NFB by PAK1 may be one way in which it contributes to cellular transformation.
The finding that PAK1 is an LPS-regulated kinase and that this activity is required for LPS-mediated NFB activation is also potentially important. LPS causes septic shock in mammals following bacterial infection. The initial response to this type of infection occurs in macrophages, which react by producing pro-inflammatory cytokines such as TNF␣ and IL1␤. The release of these cytokines then causes a massive immune response in the animal. The increase in transcription of both TNF␣ and IL1␤ in response to LPS is controlled in part by NFB. LPS initiates signaling by binding to one or more toll family receptors. These receptors resemble the type 1 IL1␤ receptor in structure and appear to use some of the same signaling molecules. The observation that PAK1 controls NFB activation in response to LPS defines another link in the signaling pathway leading to LPS-dependent transcriptional activation. Future efforts will be directed at understanding the mechanism by which LPS stimulates PAK1 activity and how PAK1, in turn, controls NFB activation.
FIG. 6. LPS activates NFB in mouse macrophages in a PAK1dependent manner. RAW 264.7 cells were cotransfected with the NFB-luciferase reporter and either a control vector, dn PAK1, or constitutively active PAK1 (PAK1 165). The cells were then either stimulated or not with LPS (1 g/ml) for 6 h. Shown is the average of three independent experiments. Error bars refer to the standard error of the mean. B, PAK1 activity is required for activation of the human TNF␣ promoter in mouse macrophages. RAW 264.7 cells were cotransfected with the TNF␣pro-chloramphenicol acetyltransferase reporter and a control vector, dn PAK1, or constitutively active PAK1 (PAK1 165). The cells were either stimulated or not with LPS (1 g/ml) for 6 h, harvested, and assayed for chloramphenicol acetyltransferase activity. Shown is the average of five independent experiments. Error bars denote the standard error of the mean.