Activation of NF-κB by Bradykinin through a Gαq- and Gβγ-dependent Pathway That Involves Phosphoinositide 3-Kinase and Akt*

Recent work has suggested a role for the serine/threonine kinase Akt and IκB kinases (IKKs) in nuclear factor (NF)-κB activation. In this study, the involvement of these components in NF-κB activation through a G protein-coupled pathway was examined using transfected HeLa cells that express the B2-type bradykinin (BK) receptor. The function of IKK2, and to a lesser extent, IKK1, was suggested by BK-induced activation of their kinase activities and by the ability of their dominant negative mutants to inhibit BK-induced NF-κB activation. BK-induced NF-κB activation and IKK2 activity were markedly inhibited by RGS3T, a regulator of G protein signaling that inhibits Gαq, and by two Gβγ scavengers. Co-expression of Gαq potentiated BK-induced NF-κB activation, whereas co-expression of either an activated Gαq(Q209L) or Gβ1γ2 induced IKK2 activity and NF-κB activation without BK stimulation. BK-induced NF-κB activation was partially blocked by LY294002 and by a dominant negative mutant of phosphoinositide 3-kinase (PI3K), suggesting that PI3K is a downstream effector of Gαq and Gβ1γ2 for NF-κB activation. Furthermore, BK could activate the PI3K downstream kinase Akt, whereas a catalytically inactive mutant of Akt inhibited BK-induced NF-κB activation. Taken together, these findings suggest that BK utilizes a signaling pathway that involves Gαq, Gβ1γ2, PI3K, Akt, and IKK for NF-κB activation.

G protein-coupled receptors (GPCRs), 1 characterized by their heptahelical structure, constitute a large family of cell surface receptors (1). These receptors are activated by a diverse array of external stimuli including hormones, neurotransmitters, sensory stimuli, chemoattractants and growth factors. GPCRs transduce signals through coupling to a collection of heterotri-meric G proteins, thus generating a broad spectrum of physiological responses (2). Increasing evidence indicates that GPCRs actively regulate transcription and gene expression events (3). The signaling pathways connecting GPCRs to several kinase cascades, including the mitogen-activated protein kinases ERK1 and ERK2, p38, stress-activated protein kinase, and the c-Jun kinase, have recently been elucidated. Activation of these kinases leads to the expression of c-fos and c-jun, components of the transcription factor activated protein-1 (AP-1), as well as other transcription factors such as MEF and ATF2 (4). An increasing number of GPCRs also have been shown to activate nuclear factor (NF)-B, thereby regulating the expression of a wide array of inducible genes. GPCRs that have been identified for their NF-B-activating functions respond to a variety of agonists, including leukocyte chemoattractants (5)(6)(7)(8)(9), thrombin (10 -12), substance P (13), endothelin (14), 5-hydroxytryptamine (15), lysophosphatidic acid (16), and bradykinin (17). These findings suggest the presence of multiple signaling pathways for NF-B activation by GPCRs.
NF-B is a dimeric, ubiquitously expressed transcription factor that plays a critical role in regulating inducible gene expression in immune and inflammatory responses (18,19). The target genes that are regulated by NF-B include cytokines, chemokines, cell adhesion molecules, growth factors, and immunoreceptors (20). In most cells, NF-B proteins exist in the cytoplasm in an inactive complex bound to the IB family of inhibitory proteins. Various stimuli can induce rapid phosphorylation, ubiquitinylation, and degradation of IB␣, resulting in nuclear translocation of NF-B proteins and transcription activation. A key regulatory step in this pathway is the activation of a high molecular weight IB kinase (IKK) complex, in which catalysis is believed to be carried out by multiple kinases including IKK1 (IKK␣) and IKK2 (IKK␤) (19). Much effort has been made in understanding the signal transduction pathways that regulate NF-B activation in response to proinflammatory cytokines such as tumor necrosis factor ␣ (TNF␣) and interleukin-1␤ (IL-1␤). NF-B activation by these inflammatory cytokines is initiated by the intracellular signaling molecules TNF receptor-associated factors (TRAF2 and TRAF6), and may involve transforming growth factor ␤-activated kinase 1 (TAK1), NF-B-inducing kinase (NIK), and IKKs (19,20). More recent studies suggest an additional NF-B activation pathway consisting of the phosphoinositide 3-kinases (PI3K) and its downstream kinase Akt (also termed protein kinase B, or PKB) (21)(22)(23). These findings demonstrate the presence of parallel signaling pathways that converge at the point of IKK activation.
The GPCRs known to activate NF-B couple to different G proteins, including G␣ q , G␣ 13 , G␣ i , and G␣ 16 . However, a de-tailed mechanism underlying GPCR-induced NF-B activation, including the relative contribution of G␣ and G␤␥ proteins and their downstream effectors, has not been delineated. The B2type bradykinin receptor (B2BKR) is known to couple to multiple G␣ proteins (24,25) and has been reported to activate NF-B (17). This study employs B2BKR in a transient transfection model to investigate the specific involvement of G␣ and G␤␥ proteins in NF-B activation. In addition, as a preliminary effort to delineate the signaling pathways downstream of heterotrimeric G proteins, we demonstrate that stimulation of NF-B by this receptor is primarily mediated by IKK2 in a process that involves PI3K and Akt.

EXPERIMENTAL PROCEDURES
Materials-Bradykinin was purchased from Sigma (St. Louis, MO). Pertussis toxin (PTX) was from List Laboratories (Campbell, CA). LY294002 was obtained from CalBiochem (San Diego, CA). Reagents for luciferase assays and the plasmid pCMV␤ were purchased from Promega (Madison, WI). A luciferase reporter plasmid was constructed by ligation of three B binding sequences (5Ј-AGGGGACTTTCCCA-3Ј) in tandem into the pGL2-Luc vector (Promega) upstream from the firefly luciferase cDNA. The green fluorescent protein (GFP) expression vector EGFP-N1 was from CLONTECH (Palo Alto, CA). Plasmids containing cDNA inserts for wild-type and constitutively activated G␣ proteins were gifts from Drs. Cindy Knall and Gary Johnson (National Jewish Center, Denver, CO), and their characterization was detailed previously (26). The G␤ 1 and G␥ 2 constructs were gifts from Dr. Tatyana Voyno-Yasenetskaya (University of Illinois, Chicago). Expression plasmids for RGS3 and RGS3T were kindly provided by Dr. Nickolai Dulin (University of Illinois, Chicago) and were described previously (27). The expression vectors of G␤␥ scavengers, bovine transducin, and T8␤ARK-myc with a ␤ARK carboxyl-terminal fragment were kindly provided by Dr. Heidi Hamm (Northwestern University, Chicago) and Dr. Sivio Gutkind (National Institutes of Health, Bethesda, MD), respectively. The expression vector of the dominant negative mutant of p85 (⌬p85) was prepared as described previously (28). The constructs of wild-type, constitutively activated, and kinase-deficient mutants of Akt/ PKB were described elsewhere (29). The expression vector of a dominant negative IB␣ (IB␣m) was a gift from Dr. Inder Verma (The Salk Institute, La Jolla, CA). Preparation of the wild-type and dominant negative mutants of IKK1 and IKK2 in an expression vector were described previously (30,31). The human B2BKR expression construct was prepared by polymerase chain reaction cloning of the entire coding sequence into the pRK5 expression vector (PharMingen, San Diego, CA).
Cell Culture, Transfection, and Luciferase Assay-HeLa cells were maintained in Dulbecco's modified Eagle's medium containing 10% heat-inactivated fetal bovine serum, 2 mM L-glutamine, 100 international units/ml penicillin, and 50 g/ml streptomycin. One day prior to transfection, the cells were seeded in 6-well cell culture plates to give a final density of 40 -60% confluency (ϳ5 ϫ 10 5 cells/well). Cells were transfected with the B-luciferase reporter, EGFP-N1, B2BKR expression vector, and other expression constructs as indicated in the text and figure legends. Transient transfection was performed using Lipo-fectAMINE Plus reagent (Life Technologies, Inc.), with 0.2 g of each DNA construct except as indicated in the figure legends. When necessary, additional DNA (pCMV␤) was added to make total DNA of 1 g/well. 24 h after transfection, cells were starved in serum-free medium for 16 h and then stimulated with control solvent, 10 nM BK, 50 ng/ml TNF␣, or 40 ng/ml IL-1␤ for 5 h. The B-directed expression of firefly luciferase was determined using luciferase assay reagents from Promega, and the resultant luciferase activities were measured with a Femtomaster FB12 luminometer (Berthold Detection Systems, Pforzheim, Germany). Relative transfection efficiency and protein expression levels in each sample were determined by measurement of fluorescent intensity of the co-transfected GFP in a spectrofluorometer (Photon Technologies International, Monmouth Junction, NJ), with excitation and emission wavelengths of 488 and 507 nm, respectively. Luciferase activities were then normalized against the level of GFP expression to minimize differences in transfection and protein expression efficiency among samples. Cell surface expression of B2BKR was detected by flow cytometry (Becton Dickinson, San Jose, CA) using fluorescein isothiocyanate-labeled BK (10 min at 4°C; NEN Life Science Products) and by immunoblotting (see below). B2BKR expression levels were not altered by co-transfection of the various constructs used in this study. Unless otherwise indicated in the figure legends, all data were collected from three independent experiments, each in duplicate. Normalized data were plotted using Prism software (Version 2.0; GraphPad, San Diego, CA).
Protein Extraction and Immunoprecipitation-HeLa cells were grown in 100-mm cell culture dishes. Transient transfection, culture, serum starvation, and agonist stimulation were done as described above. Cells were then rinsed once with ice-cold phosphate-buffered saline and incubated for 20 min at 4°C in 1 ml of lysis buffer containing 1% Nonidet P-40, 50 mM HEPES (pH 7.6), 100 mM NaCl, 10% glycerol, 1 mM EDTA, 20 mM ␤-glycerophosphate, 20 mM p-nitrophenylphosphate, 1 mM sodium orthovanadate, 1 mM NaF, and 1ϫ Protease Inhibitor Mixture Set I (CalBiochem). Cellular debris was removed by centrifugation at 14,000 ϫ g for 20 min. Cell lysates were incubated for 3 h at 4°C with 30 l of anti-HA-agarose beads (Santa Cruz Biotechnology, Santa Cruz, CA) or anti-FLAG-agarose beads (Sigma), then rinsed 3 times with 0.5 ml of lysis buffer, and eluted with 30 l of HA peptide (Santa Cruz) or FLAG peptide (Sigma). Aliquots of the eluates were used for immunoblot and in vitro phosphorylation assays.
Immunoblot Analysis and Kinase Activity Assays-Immunoblot analysis was performed with anti-HA monoclonal antibody (Santa Cruz), anti-FLAG monoclonal antibody (Sigma), or anti-phospho-Akt polyclonal antibody (Ser473; New England Biolabs, Beverly, MA). The antibodies were visualized with horseradish peroxidase-conjugated goat anti-mouse or anti-rabbit Ig (CalBiochem) using enhanced chemiluminescence (Pierce). Quantitative measurement of phosphorylated Akt was conducted using ImageQuant software (Molecular Dynamics, Mountain View, CA). For the detection of protein expression of the transfected B2BKR and IB␣m, total cell lysates were immunoblotted with an anti-B2BKR monoclonal antibody (Transduction Laboratories, Lexington, KY) and an anti-IB␣ antibody (New England Biolabs), respectively.
For IB kinase assays, HeLa cells in 100-mm dish were transiently transfected with 4 g of B2BKR and 1 g of HA-tagged IKK1 or FLAG-tagged IKK2. Cell culture and serum starvation conditions were the same as in the luciferase assays. Cells were then stimulated with BK or TNF␣ and lysed. IKKs were immunoprecipitated with anti-HAor anti-FLAG-agarose beads and then eluted with HA or FLAG peptide, respectively. The IKK activity assay was performed with 3 l of eluates and 2 g of GST-IB␣ (1-54) in 15 l of kinase buffer containing 20 mM Tris-HCl (pH 7.6), 20 mM MgCl 2 , 1 mM EDTA, 20 mM ␤-glycerophosphate, 20 mM p-nitrophenylphosphate, 1 mM sodium orthovanadate, 0.4 mM phenylmethylsulfonyl fluoride, 1 mM ATP, 20 mM creatine phosphate, 1ϫ Protease Inhibitor Mixture Set I (CalBiochem), and 5 Ci of [␥-32 P]ATP (10 mCi/ml, 6,000 Ci/mmol; Amersham Pharmacia Biotech), at 37°C for 30 min. Samples were subsequently analyzed by 10% SDS-polyacrylamide gel electrophoresis and autoradiography. The expression vector for GST-IB␣ (1-54) was constructed, and the proteins were expressed in Escherichia coli as described previously (32).
For the Akt kinase assay, HeLa cells in 100-mm dishes were transfected with 4 g of B2BKR and 1 g of HA-tagged Akt, cultured and serum-starved as described above, and stimulated with 10 nM BK or 50 ng/ml TNF␣ for indicated time periods. Cells were lysed, and Akt was immunoprecipitated with anti-HA-agarose beads and then eluted with HA peptide. Akt kinase activity assay was performed with 15 l of eluates and 0.4 g of Akt-specific substrate peptide (RPRAATF, Upstate Biotechnology, Lake Placid, NY) in 15 l of 2ϫ kinase buffer at 37°C for 20 min. Samples were subsequently transferred to P81 phosphocellulose squares (Upstate Biotechnology), rinsed 10 times with 50 ml of 0.75% phosphoric acid, dehydrated in acetone, and dissolved in 5 ml of scintillation fluid. The resultant radioactivity was determined with a Beckman LS 3801 scintillation counter.

RESULTS
The Function of IKK1 and IKK2 in BK-induced NF-B Activation-HeLa cells were transiently transfected with a Bdirected luciferase reporter and an expression vector containing the human B2BKR cDNA. Mock (vector)-transfected cells did not respond to BK (up to 100 nM) in luciferase reporter assay (Fig. 1A) because the untransfected HeLa cells do not express B2BKR ( Fig. 1C and data not shown). In B2BKRtransfected cells, BK induced B-directed luciferase activities in a dose-dependent manner from 0.1 to 500 nM (Fig. 1A). To determine whether BK-induced NF-B activation requires IB␣ phosphorylation, a dominant negative mouse IB␣ (IB␣m) devoid of inducible phosphorylation (33) was co-expressed in the HeLa cells. IB␣m abolished BK-induced NF-B activation as well as that induced by TNF␣ and IL-1␤ (Fig. 1B). Expression of B2BKR was not affected by the co-transfection of IB␣m, as evidenced by Western blot analysis (Fig. 1C) and flow cytometry (data not shown). These results indicate that phosphorylation of IB␣ at Ser 32 and Ser 36 is critical for BKinduced NF-B activation.
Receptors for several cytokines and lipopolysaccharide preferentially utilize IKK2 for NF-B activation (30,31). We examined whether this is also the case for BK-induced NF-B activation. The K44M mutants of IKK1 and IKK2, which had been shown to block cytokine-induced NF-B activation, were coexpressed in the transfected HeLa cells. As shown in Fig. 2A, expression of IKK1.DN reduced BK-stimulated NF-B activation by approximately 25%. A more potent inhibition (ϳ60%) was obtained when IKK2.DN was co-transfected into the HeLa cells. When used together, these two dominant negative constructs produced further inhibition of BK-induced NF-B activation. The inhibition of BK-induced NF-B activation was not due to the reduced expression of B2BKR, because co-transfec-tion of IKK1.DN or IKK2.DN did not affect the expression of B2BKR (Fig. 2B). Similar levels of inhibition were observed in the control cells that were stimulated with TNF␣, suggesting that BK, like TNF␣, preferentially activates IKK2. This notion was further supported by IKK assays that employed GST-IB␣ (1-54) as a substrate. Results shown in Fig. 2, C and D, indicate that, although BK activated both IKK1 and IKK2 with peak activation at 15 min, the IKK1 activity was transient, whereas the IKK2 activity was more sustained. These results combined suggest an important role of IKK2 in BK-induced NF-B activation. IKK2 assay was used in subsequent experiments as an additional measure for G protein-mediated functions.
G␣ q and G␤␥ Mediate BK-induced Activation of IKK2 and NF-B-B2BKR couples to heterotrimeric G proteins of the G␣ i , G␣ q/11 , and G␣ 13 classes (24,25). In addition, BK may also couple to G␣ s under certain circumstances (34). To determine which G␣ protein(s) is responsible for BK-induced NF-B activation, the transfected HeLa cells were first treated with PTX. Our results showed that PTX, at concentrations sufficient to block fMet-Leu-Phe induced calcium mobilization and phosphatidylinositol 4,5-bisphosphate hydrolysis (data not shown), did not alter NF-B activation by BK, TNF␣, or IL-1␤ (Fig. 3A). Consistent with this finding, PTX did not affect IKK2 activation induced by either BK or TNF␣ (Fig. 3B). We hypothesized that PTX-insensitive G␣ proteins are responsible for BK-induced NF-B activation. To evaluate this hypothesis, the wildtype G␣ q , G␣ 13 , and G␣ i2 were separately co-expressed in HeLa cells. In the absence of BK stimulation, these G␣ proteins produced small increases in B-directed luciferase activity (Fig. 4A). Following BK stimulation, the cells expressing G␣ q exhibited an ϳ75% increase over the level of luciferase activity induced by BK alone, whereas a smaller (ϳ20%) increase was observed in the cells expressing G␣ 13 . In contrast, G␣ i2 reduced the level of BK-stimulated luciferase activity by ϳ20%.
Previous studies have shown that G␣ 12 and G␣ 13 activate serum responsive factor through the downstream effector RhoA (35)(36)(37). G␣ 13 presumably can also activate NF-B through this mechanism, since GTPases of the Rho family have been implicated in NF-B activation (38,39). In this study we focused on G␣ q because it had not been previously shown to stimulate NF-B activation and because expression of G␣ q produced a potent enhancement of the BK-induced NF-B activation (Fig.  3A). To further examine the role of G␣ q , HeLa cells were cotransfected with expression vectors encoding RGS3, a regulator of G protein signaling, and its shortened form, RGS3T (27,40). RGS proteins can function as GTPase-activating proteins (GAP), which facilitate the conversion of G␣⅐GTP to G␣⅐GDP and thereby negatively regulate heterotrimeric G protein activation (41,42). RGS3T has been shown to be a more potent inhibitor of G␣ q than RGS3 (40). Our results demonstrated that

FIG. 3. Effect of PTX on BK-induced NF-B activation.
HeLa cells were transiently transfected as described in Fig. 1. 24 h after transfection, cells were treated with control solvent (CTL) or with 100 ng/ml PTX in serum-free medium for 16 h. The cells were then stimulated with control solvent (Ϫ), 10 nM BK, 50 ng/ml TNF␣, or 40 ng/ml IL-1␤ for 5 h prior to luciferase assay (A). PTX was present in the medium during stimulation. The relative luciferase activities (RLA) were measured as fold induction. B, IKK2 assay. Cells were transfected and treated as above, except that stimulation was carried out for 15 min (with 10 nM BK) and 10 min (with 50 ng/ml TNF␣). The expressed IKK2 was immunoprecipitated from cell lysate, eluted with FLAG peptide, and used for immunoblotting and kinase activity assays. The 32 Plabeled (top) and Coomassie Blue-stained (middle) substrates as well as the immunoblot of IKK2.WT (bottom) are shown. WT, wild type.

FIG. 4. A role of G␣ q in BK-induced NF-B activation.
A, G␣ q potentiates BK-induced NF-B activation in luciferase assays. HeLa cells were transiently transfected with 0.2 g each of the expression plasmids for G␣ q , G␣ 13 , and G␣ i2 . After culture and serum starvation, the cells were stimulated with control solvent (CTL) or 5 nM BK for 5 h prior to luciferase assay. Relative luciferase activities (RLA) presented are the means and S.D. from two independent experiments, each with duplicate measurements. B, HeLa cells were transfected as above, without (open bars) or with (solid bars) a RGS3T expression construct (0.2 g). The effects of RGS3T on the luciferase activities induced by BK (10 nM), TNF␣ (50 ng/ml), and IL-1␤ (40 ng/ml) were measured and expressed as fold induction. C, the effect of RGS3T on IKK2 kinase activity in cells stimulated with BK (10 nM, 15 min) or TNF␣ (50 ng/ml, 10 min). Various concentrations of RGS3T expression plasmid were used as indicated. GST-IB␣-(1-54) was used as a substrate in the kinase activity assay. The 32 P-labeled (top) and Coomassie Blue-stained (middle) substrates as well as the immunoblot of wild-type IKK2 (IKK2.WT, bottom) are shown. co-expression of RGS3T markedly reduced BK-stimulated luciferase activity (Fig. 4B). In contrast, the luciferase activities induced by TNF␣ and IL-1␤ were only slightly affected by RGS3T (Fig. 4B). Although less potent, RGS3 also inhibited BK-induced luciferase activity by ϳ60% (data not shown). Consistent with these findings, the BK-induced IKK2 activity was inhibited by the co-expressed RGS3T in a dose-dependent manner (Fig. 4C). The lack of inhibition of TNF␣-induced IKK2 activity indicates the specificity of RGS3T in G protein-mediated NF-B activation (Fig. 4C).
Inhibition of G␣ q by RGS3T favors accumulation of G␣q⅐GDP, an inactivated form of G␣ q that complexes with G␤␥ proteins and limits their functions (41). Thus the observed inhibitory effect of RGS3T and the potentiation effect of G␣ q in the transfected HeLa cells may be a function of G␣ q , G␤␥, or both. That G␤␥ proteins play a direct role in NF-B activation was suggested by the co-expression of transducin or a carboxylterminal fragment of ␤ARK (T8␤ARK) (43) in the transfected cells. These ␤␥ scavengers reduced BK-stimulated luciferase activity by approximately 65%, while having little effect on NF-B activation by TNF␣ and IL-1␤ (Fig. 5, A and B). In agreement with the luciferase reporter data, transducin (Fig.  5C) and T8␤ARK (not shown) also reduced BK-stimulated IKK2 activity without affecting TNF␣-induced IKK2 activity.
The direct involvement of G␣ q and G␤␥ in NF-B activation was investigated by expression of G␤ 1 ␥ 2 and a G␣ q mutant (Q209L) devoid of GTPase activity in the HeLa cells. Expression of G␣ q (Q209L) alone resulted in a dramatic increase in B-directed luciferase activity without BK stimulation (Fig.  6A). Similarly, an ϳ17-fold increase in luciferase activity was observed when G␤ 1 ␥ 2 was co-expressed in the HeLa cells. Stimulation of NF-B activation by G␣ q (Q209L) and G␤ 1 ␥ 2 proteins was reflected at the level of IKK activation, as both proteins stimulated the kinase activity of IKK2 (Fig. 6B). These findings suggest that both G␣ q and G␤ 1 ␥ 2 have a direct function in BK-induced activation of IKK2 and NF-B.
PI3K and Akt Are Involved in BK-induced NF-B Activation Downstream of G␣ q and G␤␥-PI3K is known to be an effector of heterotrimeric G proteins (44,45). Recent studies suggest that PI3K and its downstream serine/threonine kinase Akt play an important role in NF-B activation by TNF␣ and platelet-derived growth factor (23,46). We examined activated PI3K and Akt constructs for their functions in NF-B activation in the transfected HeLa cells. Expression of a myristoylated p110 protein (p110.myr) induced an ϳ20-fold increase of B-directed luciferase activity (Fig. 7A). A much smaller increase was seen in cells expressing a myristoylated Akt (Akt.myr), but Akt.myr augmented the effect of p110.myr. A partial inhibition of the induced B activity was obtained by co-expression of IKK2.DN. Taken together, these findings suggest that in HeLa cells, PI3K and Akt can positively regulate NF-B activation.
To investigate the requirement of PI3K in BK-induced NF-B activation in transfected HeLa cells, we examined the effect of LY294002, a PI3K inhibitor (47). Our results demonstrated that both the BK-induced and TNF␣-induced NF-B activation were dramatically inhibited by treatment with 50 M LY294002 prior to agonist stimulation (Fig. 7B). Co-expression of a dominant negative mutant of p85 (⌬p85) (28), a regulatory subunit of PI3K, partially inhibited NF-B activation induced by BK, TNF␣, or IL-1␤ (Fig. 7C). Similarly, the Bdirected luciferase activities induced by expression of G␣ q (Q209L) and G␤ 1 ␥ 2 were susceptible to inhibition by ⌬p85 (Fig. 7D). These data confirm a previous finding that PI3K is involved in BK-induced NF-B activation (48).
To determine whether Akt is a downstream effector of PI3K in the transfected HeLa cells, we first measured changes in the phosphorylation level of Akt and its kinase activity in response to BK. As shown in Fig. 8A, BK stimulated a rapid Akt phosphorylation that peaked at 15 min. Using an Akt substrate peptide as substrate, we found that BK also induced the kinase activity of Akt in a time-dependent manner, with peak kinase activity observed at 15 min (Fig. 8B). The kinetics of Akt activation paralleled that of IKK2 activation (Fig. 2). In both experiments, TNF␣ activated Akt to a lesser extent than did BK.
A kinase-deficient Akt mutant (Akt.KD) has been shown to block the effect of Akt in the inhibition of cell death (49). We examined whether this Akt mutant (K179M) also affected the BK-stimulated B activity. Data shown in Fig. 8C indicate that Akt.KD partially inhibited BK-induced NF-B activation in a dose-dependent manner, while having no obvious effect on the expression level of B2BKR. Furthermore, Akt.KD also inhib- ited the luciferase activities induced by G␣ q (Q209L) and by G␤ 1 ␥ 2 (Fig. 8D). These findings combined indicate that Akt is a component of the BK signaling pathway that leads to NF-B activation. However, Akt.KD was not able to completely abolish the induced luciferase activity by either BK or the G proteins, suggesting the presence of other parallel signaling pathways for BK-induced NF-B activation.

DISCUSSION
The transcription factor NF-B plays a critical role in regulating inducible gene expression in immune and inflammatory responses as well as in growth and development (18 -20). Substantial progress has been made in understanding the signal transduction pathways regulating activation of NF-B in response to proinflammatory cytokines such as TNF␣ and IL-1␤. Increasing evidence indicates that ligands for GPCRs can also activate NF-B; however, the signaling components involved in this process remain to be characterized. BK, a pluripotent peptide, has been shown to activate NF-B, which contributes to its proinflammatory functions in tissue injury and allergy (17). The present study sought to identify the signaling components involved in BK-induced NF-B activation in transfected HeLa cells, which have been widely used for delineation of the signaling pathways for TNF␣-and IL-1␤-mediated NF-B activation (30,50,51).
We first determined the potential involvement of the IB kinases IKK1 and IKK2 in BK-induced NF-B activation. IKKs have been established as a point of convergence for most proinflammatory cytokines that activate NF-B (20). We found that both IKK1 and IKK2 activities were induced by BK in a timedependent manner. In addition, co-expression of the dominant negative mutants of IKK1 and IKK2 inhibited BK-induced NF-B activation. Our results, together with a recently published study (52), provide direct evidence for the involvement of IKKs in GPCR-mediated NF-B activation. Notably, IKK1 activation by BK was transient, while IKK2 activation was sustained. In addition, IKK2.DN exhibited a more pronounced inhibitory effects on BK-induced NF-B activation than IKK1.DN. Our observations are consistent with data obtained from gene targeting studies that suggest IKK2, but not IKK1, plays a major role in the induction of NF-B activity in response to inflammatory stimuli (53)(54)(55)(56).
A large number of GPCRs, including B2BKR, couple to multiple G proteins. To determine which G protein(s) is responsible for BK-induced NF-B activation, we performed co-transfection experiments with three individual G␣ proteins. Expression of G␣ q potently enhanced BK-induced NF-B activation. Furthermore, expression of an activated G␣ q stimulated IKK2 activity as well as NF-B activation in the absence of agonist. These results, combined with the lack of inhibition by pertussis toxin, suggest G␣ q as a primary G protein that couples B2BKR to NF-B activation. G␣ q may directly activate one of its downstream effectors for this function; alternatively, the G␤␥ proteins released from G␣ q following BK stimulation may also activate NF-B. This latter possibility was supported by our data showing that expression of G␤ 1 ␥ 2 strongly stimulated IKK2 activity and NF-B activation. Because agonist-stimulated activation of GPCRs is accompanied by release of free G␤␥ proteins, this function of G␤␥ may be a widely used mech-FIG. 6. Activation of NF-B by G␣ q (Q209L) and G␤ 1 ␥ 2 . A, induction of B-directed luciferase activity in cells transfected with an activated G␣ q (Q209L), and with G␤ 1 ␥ 2 . After culture and serum starvation, luciferase reporter assay was conducted without BK stimulation. In some samples, the dominant negative IKK2 (IKK2.DN) construct was co-transfected. B, activation of IKK2 by G␣ q (Q209L) and G␤ 1 ␥ 2 . HeLa cells were transiently transfected with 2 g of B2BKR or 1 g of FLAG-tagged wild-tupe IKK2 (IKK2.WT) in the absence (Ϫ) or presence (ϩ) of 2 g of G␣ q (Q209L) or G␤ 1 ␥ 2 . IKK2 assay was conducted following BK stimulation (5 nM for 15 min). The 32 P-labeled (top) and Coomassie Blue-stained (middle) substrates as well as immunoblot of IKK2.WT (bottom) are shown. HeLa cells were transiently transfected without (CTL) or with ⌬p85 and with other plasmids for the luciferase assays as described above. The cells were stimulated with control solvent (Ϫ), BK (10 nM), TNF␣ (50 ng/ml), or IL-1␤ (40 ng/ml) for 5 h prior to luciferase assay. D, inhibition of G␣ q (Q209L)-or G␤ 1 ␥2-induced NF-B activation by ⌬p85. The experiment was done similarly to that described in Fig. 7A, except for the inclusion of G␣ q (Q209L) and G␤ 1 ␥ 2 (0.2 g each) in the transfection procedure. The induced luciferase activities were measured in the absence of BK. Maximal response (100%) was measured in the absence of ⌬p85. Data for all experiments are presented as the means and S.D. from at least two experiments, each with duplicate measurements. anism for GPCR-stimulated NF-B activation. However, many GPCRs that generate free G␤ 1 ␥ 2 proteins upon activation do not mediate NF-B activation. Other receptors, such as the fMet-Leu-Phe receptor and C5a receptor, activate NF-B only in certain cell types (7,8). Thus, the G␤␥ proteins released during the activation of these receptors are not sufficient to activate NF-B, and additional signaling components may be required. Alternatively, there may be a negative regulatory mechanism that prevents NF-B activation by some GPCRs. Our finding that RGS3T inhibits BK-induced NF-B activation provides a first line evidence that the RGS proteins can negatively regulate GPCR-mediated transcription activation.
In addition to G␣ q and G␤ 1 ␥ 2 , we have shown that co-expression of G␣ 13 slightly enhanced BK-stimulated NF-B activation. This finding is compatible with the function of G␣ 13 in activating the small GTPase RhoA (57)(58)(59)(60), which has been demonstrated to stimulate NF-B activation (38,39). A previous study using A549 lung epithelial cells demonstrated partial inhibition of BK-induced NF-B activation by a dominant negative RhoA construct and by the C3 exoenzyme from Clostridium botulinum (61), suggesting a function for RhoA in BKinduced NF-B activation. However, the function of G␣ 13 in BK-induced NF-B activation was not investigated, and a subsequent work demonstrated inhibition of NF-B activation by pertussis toxin in A549 cells, suggesting the involvement of G␣ i (48). Results obtained from the current work do not support a positive role of G␣ i in BK-induced NF-B activation in HeLa cells, although we cannot exclude the possibility that the G␤ 1 ␥ 2 proteins released from G␣ i mediate NF-B activation in A549 cells.
Our data suggest that PI3K is a downstream effector of G␣ q and ␤ 1 ␥ 2 and that it is partially responsible for BK-induced NF-B activation in the transfected HeLa cells. PI3K is activated by many GPCRs, and its function in NF-B activation has been suggested in recent studies (9,23,46,48). PI3K may stimulate NF-B activation through a downstream serine/threonine kinase Akt, as shown recently in TNF␣-and plateletderived growth factor-induced NF-B activation (23,46). Our results demonstrated that BK could induce Akt phosphorylation and stimulate its kinase activity. The time course of the induced Akt activation is compatible with that of NF-B activation in BK stimulated cells. A kinase-deficient mutant of Akt partially blocked BK-induced NF-B activation in a dose-dependent manner. Furthermore, Akt.KD also inhibited NF-B activation by G␤ 1 ␥ 2 and by an activated G␣ q . Together, these results suggest that Akt is a downstream effector of PI3K that plays a role in BK-induced NF-B activation in HeLa cells. Activation of Akt by PI3K may require additional intermediate kinases, such as phosphoinositide-dependent kinase 1 (PDK1) (62). Whether these kinases are necessary for BK-induced NF-B activation remains to be investigated.
In summary, the current study provides direct evidence that GPCR-mediated NF-B activation shares part of the signaling mechanisms with that for TNF␣, including the preferential activation of IKK2 and a recently identified pathway involving PI3K and Akt. This study also demonstrated that both G␣ q and G␤ 1 ␥ 2 proteins actively participate in BK-induced NF-B activation and that GPCR-mediated NF-B activation can be regulated by RGS proteins. These signaling mechanisms are not found in TNF␣-induced NF-B activation pathway and are therefore unique to receptors that couple to heterotrimeric G proteins. Taken together, our results suggest that the BK sig- A and B, HeLa cells transfected with HA-tagged Akt were stimulated with 10 nM BK or 50 ng/ml TNF␣ for the indicated time periods. Cells were lysed, and Akt was immunoprecipitated with anti-HA-agarose beads and then eluted with HA peptide. Aliquots of the eluates were used in assays for immunoblotting (A, bottom panels) and quantitation (A, top, bar graph) and for kinase activity measurement (B). An anti-Phospho-Akt antibody (Ser 473 ) and an anti-HA antibody were used to detect the phosphorylated Akt (Akt-P) and total Akt, respectively (A). C, a kinase-deficient Akt (Akt.KD) inhibits NF-B activation by BK. HeLa cells were transfected with the relevant plasmids. 16 h after transfection, serum was removed from culture medium, and the cells were stimulated with BK (10 nM). Expression of the transfected B2BKR and Akt.KD (tagged with HA) were detected by immunoblotting analysis. CTL, control; conc., concentration. In D, the cells were transfected with G␣ q (Q209L) or G␤ 1 ␥ 2 (0.2 g each) with or without Akt.KD (2 g). Inhibition by Akt.KD on the induced relative luciferase activities (RLA) was determined as described above. Data presented are from three independent experiments, each with duplicate measurements.

FIG. 8. Involvement of Akt in BK-induced NF-B activation. In
naling pathway leading to NF-B activation involves B2BKR, G␣ q , G␤ 1 ␥ 2 , PI3K, Akt, and IKK2. However, because inhibition of PI3K and Akt only partially blocked, but did not completely abolish, NF-B activation by BK, G␣ q , and G␤ 1 ␥ 2 , the current study does not exclude the presence of other parallel signaling pathways. Identification of these other signaling components and pathways in future studies will be necessary for a more complete understanding of how GPCRs activate NF-B.