Activation of p21-Activated Kinase 1-Nuclear Factor κB Signaling by Kaposi’s Sarcoma-Associated Herpes Virus G Protein-Coupled Receptor during Cellular Transformation

The Kaposi’s sarcoma-associated herpes virus (KSHV) is a γ-2 herpes virus that is implicated in the pathogenesis of malignancies including Kaposi’s sarcoma (KS), primary effusion B-cell lymphomas, and multicentric Castleman’s disease (1, 2, 3). KS lesions are composed of diverse cell types, including infiltrating inflammatory cells, endothelial cells, and fusiform or spindle-shaped cells (4). The pathogenic mechanisms by which infection with KSHV leads to disease are still unclear. KSHV encodes several viral oncogenes, one of which, a chemokine-like G protein-coupled receptor (KSHV-GPCR), has been strongly implicated in KS-mediated pathogenesis (5). The KSHV-GPCR transforms rodent fibroblast cells, causes tumors in nude mice, and stimulates angiogenesis by inducing expression and secretion of vascular endothelial growth factor (VEGF; Refs. 5, 6, 7). Transgenic mice expressing the KSHV-GPCR develop highly vascularized KS-like tumors, further supporting the contribution of the KSHV-GPCR to the pathogenesis of KS. Expression of the KSHV-GPCR activates a number of signal-transducing pathways (6), but the relevance of these pathways to KSHV-GPCR transformation is still poorly understood.

KSHV-GPCR is a member of the CXC chemokine G protein-linked receptor family, with significant homology to the CXCR2 receptor for IL-8 (8). In contrast to the ligand dependence of many other GPCRs, including other CXC chemokine receptors, the KSHV-GPCR signals constitutively in a ligand-independent manner (7). Most chemokine receptors signal primarily through the Gα_i pathway (9). In contrast, KSHV-GPCR predominantly initiates a G_q signaling pathway to activate phospholipase C, thus increasing phosphatidyl-inositol turnover (7, 10, 11). KSHV-GPCR also activates the major mitogen-activated protein kinase (MAPK) pathways, including notably extracellular signal-regulated kinase (ERK) signaling, and has been recently shown to activate transcription of nuclear factor κB (NFκB; Refs. 12, 13, 14, 15, 16). However, a detailed knowledge of the signaling pathways leading from the KSHV-GPCR to the activation of these downstream signaling effectors still remains elusive.

The NFκB transcription factor is critically involved in cellular growth and transformation (17, 18, 19, 20, 21), making its activation by the KSHV-GPCR of particular interest. In the absence of activation, NFκB is retained in the cytoplasm due to a heterodimeric interaction with an inhibitory protein known as the inhibitor of κB (IκB; Refs. 22, 23, 24, 25). Destruction of IκB is required for the activation and nuclear translocation of NFκB and the subsequent transactivation of NFκB target genes. Phosphorylation of IκB on serine 32 and serine 36 by the IκB kinases inhibitor of IκB kinase (IKΚ) α and IKΚβ is an important initiation signal for IκB degradation and NFκB release (26, 27). Based on knockout studies in mice, IKΚβ appears to be more important than IKKα for controlling NFκB activity in response to cytokines and other ligands (28). NFκB activation stimulates both cell survival and cell proliferation, and NFκB signaling has been shown to be essential for oncogenes such as Ras and Raf to transform cells (29).

Consideration of known signaling pathways controlled by the oncoprotein Ras provides a useful context for the analysis of KSHV-GPCR control of NF-κB and the relevance of this control pathway to KSHV-GPCR-dependent cell transformation. Ras and its effectors, including the Rho family GTPases Rac1 and Cdc42, induce the phosphorylation of serine 32 and serine 36 of IκBα and cause the ubiquitination of the IκB complex, which releases the IκB-bound NFκB to translocate to the nucleus (30). Rac1 and Cdc42 share several downstream effectors, including the p21-activated kinase 1 (Pak1; Ref. 31). Pak1 mediates Rac1/Cdc42-stimulated cytoskeletal signals (32) and has also been show to play a critical role in the control of cell survival, cell proliferation, and cell transformation by oncogenes including Ras (33). In many different cell types, Pak1 regulates the ERK cascade. It may also regulate p38 and c-Jun NH₂-terminal kinase signaling, but this signal, unlike the ERK signal, is probably not important for anchorage-dependent cell growth and survival (34, 35). Significantly, recent studies have shown that Ras also requires Pak1 to activate NFκB (36, 37).

Like Ras, GPCRs are dependent on Rho GTPase-controlled signaling pathways (38). Several GPCRs, such as Mas, G2A, and PAR-1, mediate cell transformation through the Rho family proteins Rac1 and RhoA (39, 40). In this work, we demonstrate KSHV-GPCR activation of Rac1 and Pak1 and show that Pak1 is essential for KSHV-GPCR-induced anchorage-dependent growth. We additionally show that the activation of NFκB by KSHV-GPCR proceeds through Pak1-induced phosphorylation of IKKβ, and we demonstrate that NFκB activation is also essential for KSHV-GPCR-mediated cell transformation. Our work thus provides direct evidence for an important role for Pak1 as a downstream signaling molecule involved in the KSHV-GPCR to NFκB cell transformation pathway.

cDNA expression plasmids using the CMV promoter to express myc-tagged Pak1, PakR299, and RacN17 based on the plasmid pCMV6M (a modified version of pCMV5) have been described elsewhere (41). The pCMV6M-PakL107F and pEBG-PIDL107F expression plasmids were a kind gift from J. Chernoff (Fox Chase Cancer Center, Philadelphia, PA).³ Mutating leucine 107 to phenylalanine leads to strong activation of full-length Pak1, and this activity is independent of the Rac1 and Cdc42 GTPases (42), even though binding to these proteins is maintained. The 83–149-amino acid fragment of Pak1, termed the Pak inhibition domain (PID), inhibits autophosphorylation and substrate phosphorylation, whereas the L107F mutated fragment is strongly reduced in its inhibitory potency (42). To generate the pCMV6M-Pak-PID construct, the PID (amino acids 83–149) of human wild-type Pak1 was amplified from wild-type Pak1 cDNA by PCR as a BamHI-EcoRI insert using 5′-CGCCGCGGATCCCACACAATTCATGTCGGTTTTGATGC-3′ and 5′-CGCCGCGAATTCTGACTTATCTGTAAAGCTCATGTATTTCTGGC-3′ as primers and subcloned into the pCMV6M vector. The pCEFL-AU5-KSHV-GPCR expression plasmid was a gift from S. Gutkind (NIH, Bethesda, MD; Ref. 10). The pZIP-NeoSV (x) 1-Human K-Ras4B expression plasmid was a gift from C. Der (University of North Carolina, Chapel Hill, NC; Ref. 43). The pNFκB-Luc reporter plasmid containing five copies of the enhancer element of NFκΒ (ΤGGGGACTTTCCGC) and the pTK-Luc reporter plasmid were a gift from D. Manning (University of Pennsylvania, Philadelphia, PA; Ref. 44). The pCMV-IκBαM and pEGFP-C4 plasmids were purchased from BD Biosciences Clontech (Palo Alto, CA).

Pertussis toxin (PTX), human IFN-γ-inducible protein-10 (IP-10), and wortmannin were purchased from Sigma-Aldrich (St. Louis, MO). Human monokine induced by IFN-γ (MIG) was purchased from PharMingen (Palo Alto, CA), and the inhibitors BAY-11-7082, LY294002, PD98059, and U0186 were purchased from Calbiochem (La Jolla, CA). Monoclonal antibodies (MAbs) against Ras, Rac1, and Cdc42 were purchased from Transduction Lab (Carlsbad, CA). The AU5 MAb against the AU5-tag on KSHV-GPCR was purchased from Research Diagnostics, Inc. (Flanders, NJ), and the anti-myc-tag MAb 9E10 was purchased from Calbiochem. The polyclonal antibodies against Pak1 (N-20), Pak2, and Pak3 were purchased from Santa Cruz Biotechnology Lab (Santa Cruz, CA). All other antibodies were purchased from Cell Signaling Technology, Inc. (Beverly, MA). Recombinant human tumor necrosis factor-α (TNF-α) was purchased from R&D Systems, Inc. (Minneapolis, MN).

NIH3T3 and Rat-1 cells were grown at 37°C in 5% CO₂, 95% air in high-glucose (4.5 g/liter) DMEM purchased from Fisher Scientific (Pittsburgh, PA), supplemented with 10% fetal bovine serum, penicillin (100 U/ml), and streptomycin (100 μg/ml). To establish stable cell lines expressing KSHV-GPCR, the pCEFL-AU5-KSHV-GPCR plasmid was transfected into NIH3T3 cells using LipofectAMINE Plus reagent (Life Technologies, Inc., Carlsbad, CA) according to the manufacturer’s protocol. Forty-eight h after the addition of DNA, the transfected cells were selected in growth medium containing 1 mg/ml Geneticin (G418; Life Technologies, Inc., Grand Island, NY). Protein expression levels were determined by Western blot analysis of G418-selected cell lysates using the anti-AU5 for cells expressing KSHV-GPCR. Blots were visualized with the procedure outlined in the enhanced chemiluminescence kit from Amersham Biosciences (Piscataway, NJ). The K-Ras-transformed stable cell line has been described elsewhere (35).

NIH3T3 cells were transfected with plasmids specified in “Results” together with pEGFP-C4 plasmid DNA expressing the green fluorescence protein (GFP) at a ratio of 10:1, using LipofectAMINE Plus according to the manufacturer’s protocol. Transfection efficiency was assayed by checking for green fluorescence of the GFP protein under a Nikon N6000 fluorescence microscope and was typically 75–90%. Forty-eight h post-transfection, cell lysates were prepared by washing cells with cold PBS followed by lysis in 1× lysis buffer containing 40 mm 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (pH 7.4), 1% NP40, 100 mm NaCl, 1 mm EDTA, 25 mm NaF, 1 mm sodium orthovanadate, 10 μg/ml leupeptin, and 10 μg/ml aprotinin. Samples were centrifuged at 12,000 × g for 10 min at 4°C, and the supernatants were collected. Western blot analysis using appropriate antibody was done using 50 μg of the total cell lysate.

NIH3T3 and Rat-1 cells were transfected as described above. Forty-eight h post-transfection, cells were split 1:5 into 100-mm-diameter dishes. The cells were refed twice a week with fresh growth medium. Cell foci were scored 14–18 days post-transfection by fixing the cells in a 10% acetic acid:10% methanol solution and staining the dishes with 0.4% crystal violet in 10% ethanol. Soft agar assays were performed as described previously (45). In brief, after transfection, cells (10⁵) were plated on 60-mm dishes. After 21–24 days, colonies were scored by inspection under a Nikon DIAPhot microscope using phase contrast and also after staining the dishes with 0.5 mg/ml 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (Sigma-Aldrich) overnight at 37°C. For drug inhibition assays, the soft agar assay was done as described previously (46). In brief, transfected cells were suspended in DMEM containing 0.41% Bacto agar supplemented with 10% fetal bovine serum, in the presence or absence of specific inhibitors (10 nm wortmannin, 25 μm PD98059, 25 μm U0126, 5 mg/ml PTX, 10 mm IP-10, 10 mm MIG, and 5 mg/ml BAY-11-7082). The cell suspension was then overlaid onto a hard agar base composed of DMEM containing 0.72% Bacto agar supplemented with fetal bovine serum in 60-mm plates. Fetal bovine serum (10%)-DMEM with or without the specific inhibitor was added every 3 days to the plates. For wortmannin, which has a short half-life, medium with the drug was added every 8 h. The colonies were scored after 15 days as described above.

NIH3T3 cells were transfected with plasmid-expressing KSHV-GPCR or vector only for either transient or stable expression, and cells were lysed as described above. Small aliquots (20 μg) of lysate were removed for determination of Rac1 and Cdc42 expression by Western blotting. The remainder of the lysate (250 μg) was incubated with glutathione S-transferase (GST) or GST-Pak-binding domain (GST-PBD; a gift from Dr. Chernoff) and glutathione-Sepharose 4B beads (Upstate Biotechnology, Inc., Lake Placid, NY) at 4°C for 1 h (47). The beads were washed three times with 1× lysis buffer and pelleted. Pelleted activated Rac1 bound to GST-PBD and total Rac1 within the lysates was detected by Western blotting using a MAb against Rac1. The blots were then stripped using Restore Western blot stripping buffer (Pierce, Rockford, IL) and reprobed for activated Cdc42 and total Cdc42 protein expression using anti-Cdc42 MAb. To obtain a Rac1 activation profile, NIH3T3 cells expressing the vector alone were cultured with medium containing no serum for 16 h; treated for 0, 2.5, 5, or 10 min with insulin (100 nm); lysed; and assayed for Rac1 activation using the GST-PBD Sepharose beads.

For experiments with the inhibitory compounds, PTX (500 μg/ml), human IP-10 (100 nm), and human MIG (100 nm) cells expressing sustained KSHV-GPCR were cultured with medium containing no serum for 16 h; treated with the above compounds for 3 h; lysed; and analyzed for the presence of activated Rac1 and Pak1. Phosphorylation status of IκBα and IKK was also determined (see protocol below).

Exogenous Pak1 was immunoprecipitated from lysates of NIH3T3 cells transfected with a construct expressing myc-tagged Pak1 by incubation with anti-myc antibody and protein A-agarose for 2 h at 4°C (35). Samples were incubated for 30 min at 4°C, washed three times with 1× lysis buffer and twice with 2× phosphorylation buffer [10 mm MgCl₂, 40 mm 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (pH 7.4)], and then incubated with 5 μg of myelin basic protein (Sigma-Aldrich) for 5 min on ice. Kinase assays were initiated by the addition of 10 μCi of [γ-³²P]ATP (3,000 Ci/mmol) and 20 μm (final concentration) ATP, followed by incubation for 10 min at 22°C. Reactions were stopped by the addition of 2× SDS sample buffer and heating to 95°C. Reaction products were resolved by 12.5% SDS-PAGE and visualized by autoradiography.

To measure endogenous Pak1 activity, NIH3T3 cells were transfected with the appropriate plasmids as described above. The cells were incubated for 24 h to allow protein expression and then serum starved for 16–18 h before lysing. Pak1 was immunoprecipitated from 500 μg of cell lysate by incubating with 2 μg of rabbit anti-Pak antibody (N-20) and protein A-Sepharose for at least 2 h at 4°C on a rotating platform. The immunoprecipitates were pelleted by centrifugation and washed three times with 1× lysis buffer. The immunoprecipitated Pak1 was solubilized with 2× SDS sample buffer, subjected to 12.5% SDS-PAGE, and transferred to nitrocellulose. Phosphorylated Pak1 activity was detected by using phosphospecific Pak1 antibody directed against threonine 423 of wild-type Pak1. Total Pak1 was detected by Western blotting using rabbit polyclonal anti-Pak1 antibody (N-20).

EMSAs were performed using a EMSA “Gel-Shift” kit from Panomics, Inc. (Redwood City, CA). In brief, 100-mm dishes of NIH3T3 stable cell lines expressing vector alone or KSHV-GPCR were starved overnight, along with stable KSHV-GPCR cell-line transiently transfected with plasmids expressing RacN17, Pak-PID, and PIDL107F. Nuclear extracts were prepared the following day using the manufacturer’s protocol for the Panomics nuclear extraction kit. The nuclear extracts were then incubated with a biotin-tagged NFκB probe (5′-AGTTGAGGGGACTTTCCCAGGC) along with nonspecific and specific competitors supplied with the kit where indicated for 30 min at 15°C, according to the manufacturer’s protocol. The extracts were electrophoresed using a 5% polyacrylamide gel at 4°C and transferred to S&S NYTRAN nylon membrane (Schleicher & Schuell, Keene, NH). The blots were developed using the detection kit provided in the EMSA kit and visualized with an enhanced chemiluminescence detection system (Amersham Biosciences).

For measurements of NFκB reporter gene activity, NIH3T3 cell lines stably expressing either vector or KSHV-GPCR or K-Ras were transfected with the cis-reporter plasmid NFκB-Luc (Stratagene, La Jolla, CA), which contains a NFκB-driven firefly luciferase gene, together with pTK-Luc (Promega, Madison, WI), which contains the Renilla luciferase gene under control of the herpes simplex virus thymidine kinase promoter (44). Twenty-four h post-transfection, cells were washed with ice-cold PBS and lysed for measurement of luciferase activities using reagents of the Dual-Luciferase Reporter Assay System (Promega). Measurements were performed with a Turner Designs (Sunnyvale, CA) luminometer using a 10-s setting. Firefly luciferase activities were normalized to Renilla luciferase activities. For experiments with the Pak-PID, the stable KSHV-GPCR and K-Ras cell lines were transfected with Pak-PID along with the luciferase reporter plasmids, and 24 h post-transfection, cells were lysed and assayed for NFκB activity. For experiments with the inhibitory compounds BAY-11-7082 (5 ng/ml), wortmannin (10 nm), PD98059 (30 μm), U0186 (30 μm), and PTX (500 μg/ml), KSHV-GPCR and K-Ras stable cell lines were treated for 30 min with these compounds 24 h after transfection with the luciferase reporter plasmids, lysed, and analyzed for NFκB activity. For experiments with human TNF-α, cells were treated with 30 ng/ml human TNF-α for 40 min, lysed, and analyzed for NFκB activity.

NIH3T3 cells stably expressing KSHV-GPCR were transiently transfected with either the dominant-negative RacN17 or Pak-PID plasmids along with pEGFP-C4 plasmid as described above. Twenty-four h after transfection, the cells were serum starved for another 24 h and then lysed. Fifty μg of total cell lysate were used for immunoblotting using phosphospecific antibodies against IκBα and IKK.

To quantitate data, for Western analysis, densitometric analysis of the enhanced chemiluminescence-exposed blots was done using the NIH ImageJ (version 1.24o) software. For Pak1 kinase assays, the relative levels of myelin basic protein phosphorylation were quantitated based on direct analysis of radioactive signal using a Storm 840 PhosphorImager (Molecular Dynamics, Carlsbad, CA) and IO software version 2.

Statistical differences between different groups were determined using the ANOVA and Tukey-Kramer multiple comparisons test for significance. The data are presented as mean ± SE for at least three separate determinations for each treatment.

We hypothesized that KSHV-GPCR transformation of NIH3T3 cells involves the Pak1 signaling cascade. To evaluate the role of Pak1 and its activation by Rac1 in KSHV-GPCR-induced cell transformation, we transfected NIH3T3 and Rat1 fibroblast cells with KSHV-GPCR by itself or with a set of Rac1 and Pak1 derivatives. These included (a) wild-type Pak1 (Pak1); (b) constitutively active Pak1 (PakL107F; this mutation interrupts an intramolecular Pak1 interaction responsible for Pak inactivation; Ref. 42); (c) a kinase dead Pak1 that functions as a dominant negative (PakR299; Ref. 48); (d) a PID truncation of Pak1 that also functions as a dominant negative (Pak-PID, containing amino acids 83–149 of Pak1; Ref. 42); (e) a Pak-PID with an inserted L107F mutation, which eliminates its dominant-negative function (PID-L107F); and (f) a dominant-negative Rac1, which prevents activation of Pak1 (RacN17; Ref. 41). For comparison, we also transformed cells with constitutively active PakL107F by itself.

We measured transformation by assessing growth on soft agar. Plates transfected with KSHV-GPCR alone had ∼150 colonies, a 30-fold increase over plates transfected with vector (Fig. 1, A and B). KSHV-GPCR-dependent colony formation was strikingly inhibited by all plasmids expressing proteins that inhibited Pak1 activation, with a reduction of 92% (RacN17), 84% (PakR299), and 89% (Pak-PID) observed (Fig. 1,B). The specificity of the inhibitory effect was further confirmed by the fact that the PID-L107F derivative did not inhibit KSHV-GPCR transformation (Fig. 1,B, contrast Lanes 6 and 7). Finally, coexpression of the constitutively activated PakL107F with KSHV-GPCR resulted in a minor increase in transformation efficiency versus the KSHV-GPCR alone (Fig. 1,B, Lanes 2 and 3), whereas transformation with PakL107F alone resulted in induction of colonies almost equivalent to that seen with the KSHV-GPCR (Fig. 1 B, Lanes 2 and 8). These results are compatible with the idea that endogenous levels of Pak1 are sufficient to mediate KSHV-GPCR signaling for transformation. They also suggest that a substantial portion of KSHV-GPCR signaling proceeds through Pak1 and involves full activation of Pak1, such that activation of Pak1 by other (e.g., mutational) means does little to augment KSHV-GPCR transformation.

For comparison, NIH3T3 cells and Rat 1 cells were also cotransfected with K-Ras alone or with PakR299 and RacN17. A partial inhibition of K-Ras-mediated anchorage-dependent growth in NIH3T3 cells was seen with PakR299 (52%), and an almost complete inhibition was observed on cotransfection with RacN17 (95%; Fig. 1 B). Comparable results were obtained with Rat1 transfectants (not shown). Together, these results indicate that KSHV-GPCR-induced transformation requires both Rac1 and Pak1 to a degree comparable with that seen with K-Ras transformation.

Separately, NIH3T3 cells transfected as described above with a subset of the described constructs were also cultured for 14–18 days to assess the dependence of focus formation on the Pak1 signaling pathway. In the plates transfected with KSHV-GPCR alone, 60–75 foci developed versus ∼6 foci in NIH3T3 cells transfected with vector (Fig. 1,C). Notably, cotransfection of the dominant negatives PakR299 or Pak-PID with KSHV-GPCR significantly reduced the number of foci observed (77 and 81% fewer foci, respectively) whereas PID-L107F had no effect. In contrast, cotransfection of constitutively active Pak1 enhanced transformation only a small degree over KSHV-GPCR alone (Fig. 1 C). These results were in agreement with the soft agar studies.

If KSHV-GPCR activation of a Pak signaling cascade is required for KSHV-GPCR induction of cell transformation, we note three different Pak1 family members (Pak1, Pak2, and Pak3) with highly related amino acid sequences and overlapping signaling function have been described previously (32), and each may potentially be a KSHV-GPCR target. We directly assessed expression of these three forms of Pak in NIH3T3 cells and, for contrast, in Rat1 cells (Fig. 2 A). Antibody to Pak1 yielded a strong signal in both cell types; Pak2 was abundant in Rat1 cells but not NIH3T3 cells; and Pak3 was poorly detected in both cell types, suggesting limited expression of Pak3. Based on these results, we conclude that the predominant isoform of Pak expressed in NIH3T3 cells is Pak1, with very low levels of Pak2 and Pak3. Hence, we concentrated the rest of our studies on the Pak1 isoform.

Pak1 is known to be activated through interaction with activated forms of the G proteins Rac1 and Cdc42 (49). We therefore assessed activation of Pak1, Rac1, and Cdc42 by the KSHV-GPCR in NIH3T3 fibroblasts. KSHV-GPCR was transiently transfected into NIH3T3 cells. Forty-eight h post-transfection, the cells were harvested, and the total cell lysate was incubated with GST fused to the Pak-PBD (amino acids 67–149 of Pak, which selectively interacts with activated GTP-bound Rac1 and Cdc42) or to GST alone and purified by pull down with glutathione-Sepharose beads. Only the KSHV-GPCR-transfected cells contained activated Rac1 and Cdc42, although both the vector-transfected and KSHV-GPCR-transfected cells expressed equal amounts of these proteins (Fig. 2,B). KSHV-GPCR activation of Cdc42 was not as robust as its activation of Rac1 (∼50% less; Fig. 2 B). This suggests that KSHV-GPCR activates Pak1 using both the small G proteins Rac1 and Cdc42, similar to the situation with Ras activation of Pak1 (33), but that Rac1 may be the predominant intermediary.

We next examined the KSHV-GPCR activation of Pak1. Because endogenous levels of Pak1 are too low to permit analysis of activation status, NIH3T3 cells were cotransfected with Pak1 along with KSHV-GPCR or vector. Forty-eight h after transfection, Pak1 was immunoprecipitated and tested for its activation status. KSHV-GPCR stimulated activation of Pak1, as detected by Pak1 phosphorylation of myelin basic protein substrate, about 16-fold over the over basal levels (Fig. 2,C). To determine whether Rac1 was required to mediate the signal from KSHV-GPCR to Pak1, we assessed whether RacN17 or the Pak-PID inhibited activation of Pak1 by KSHV-GPCR. Cells were transfected with Pak1 and with vector or KSHV-GPCR alone or KSHV-GPCR with either RacN17 or the Pak-PID. Although an anti-Pak1 immunoprecipitate from KSHV-GPCR-expressing cells had a substantial amount of activated Pak1 phosphorylated at threonine 423, the immunoprecipitates from cells cotransfected with RacN17 or the Pak-PID construct showed about 11-fold less phosphorylated Pak1, reduced to basal uninduced levels (Fig. 2 D). Together, these results indicated that efficient KSHV-GPCR activation of Pak1 requires Rac1.

To facilitate detailed analysis of KSHV-GPCR cell transformation, stable cell lines were developed in the NIH3T3 background that expressed parental vector or KSHV-GPCR (Fig. 3,A). As preliminary characterization, the cell lines were tested in cell transformation assays as described above, using both focus formation and soft agar colony assays. The stable cell lines formed both foci and colonies, which could be inhibited by cotransfection with PakR299, Pak-PID, and RacN17, similar to the transiently transfected NIH3T3 cells (data not shown). Stable expression of KSHV-GPCR in these cell lines activated both Rac1 and Pak1, similar to transient expression of KSHV-GPCR (Fig. 3, B and C, Lanes 1 and 2 in each). We then used these NIH3T3 stable cell lines to perform a series of experiments to study the signals from KSHV-GPCR through Pak1.

To confirm the specificity of Rac1 activation by KSHV-GPCR, we pretreated NIH3T3 cells expressing KSHV-GPCR with IP-10 (a KSHV-GPCR specific inverse agonist) and MIG (a compound structurally related to IP-10, but without KSHV-GPCR inhibiting activity) before assaying for Rac1 or Pak1. IP-10 inhibited both Rac1 and Pak1 activation by KSHV-GPCR, whereas MIG was ineffective (Fig. 3, B and C), indicating that the specificity of Rac1-Pak1 activation observed in our assay system is through KSHV-GPCR signaling. Although KSHV-GPCR predominantly signals through the effector G_q, some reports indicate that it signals to Gα_i as well (13, 14), hence cells were also treated with PTX, a Gα_i-specific inhibitor, and similarly assayed. PTX significantly reduced KSHV-GPCR activation of Rac1 and Pak1, implicating the involvement of Gα_i proteins in this signaling axis, a point explored further below. It was also of interest to determine whether the degree of Rac activation induced by KSHV-GPCR represents a partially or fully activated Rac. For comparison, we assayed for Rac1 activation in NIH3T3 cells after stimulation with insulin, a potent Rac1 activator. In our assay system, insulin stimulated transient activation of Rac1 with normal kinetics, showing a transient peak at 5 min (Fig. 3,D). KSHV-GPCR-induced Rac1 activation was comparable with, if not stronger than, this level of activation (Fig. 3, compare B and D). Together, these results confirmed the importance of KSHV-GPCR to Rac1 to Pak1 signaling for KSHV-GPCR transformation.

Activation of NFκB is a rapid event that occurs within minutes after stimulation and results in transcription of genes that favorably alter the cellular environment to promote cell survival and proliferation. Hence, up-regulation of NFκB is a common strategy among viruses. HIV-1, HTLV-1, HBV, HCV, and herpes viruses have been shown to activate the NFκB pathway (50), whereas in some contexts, Pak1 has been shown to promote NFκB activation (37). To determine whether NFκB activation is essential for KSHV-GPCR-mediated cellular transformation, we performed soft agar assays with the cells expressing KSHV-GPCR cotransfected with a dominant-negative derivative of IκB. This mutant, IκBM, has two serine to alanine mutations at residues 32 and 36 (the sites of IKK phosphorylation), making it resistant to degradation and thus causing a constitutive blockade of the NFκB signaling pathway. As an alternative approach, we treated cells with BAY-11-7082, which specifically blocks TNF-α-mediated degradation of IκB; whereas as a reference, we also used the IP-10 and MIG compounds tested previously for KSHV-GPCR inhibition of Pak1 and Rac1 activation (Fig. 3, B and C). The inhibitor BAY-11-7082 or the dominant-negative IκB mutant resulted in an almost total inhibition of the formation of colonies by KSHV-GPCR (Fig. 4 A) to a degree even greater than that seen with the KSHV-GPCR inhibitor IP-10, whereas the negative control MIG had no effect on colony formation. These results clearly demonstrated a significant role of NFκB in cell transformation induced by KSHV-GPCR.

We next wished to determine whether KSHV-GPCR regulates activation of NFκB through Pak1. We performed two assays to measure activation of NFκB by KSHV-GPCR. As one approach, we performed the EMSAs to check for the DNA-binding ability of the NFκB protein to its DNA consensus site in KSHV-GPCR cells. Stable expression of KSHV-GPCR increased the ability of NFκB to bind the DNA probe over binding levels seen with vector controls (Fig. 4,B). This increase was inhibited by cotransfection with RacN17 and Pak-PID mutants, as well as by treatment with the IκB inhibitor, BAY-11-7082, used as a control (Fig. 4,B). In contrast, transfection with inactive PID derivative, PIDL107F, had no effect on the ability of KSHV-GPCR to induce binding of NFκB to its DNA-binding motif (Fig. 4 B).

As a second approach, we performed dual-luciferase assays in the stable NIH3T3 cells expressing KSHV-GPCR or vector. These cells were cotransfected with a firefly luciferase reporter plasmid under the control of the NFκB promoter and a Renilla luciferase reporter plasmid under the control of thymidine kinase promoter as an internal control for transfection efficiency. As seen in Fig. 4,C, a 7-fold increase in NFκB activity was observed in cells expressing KSHV-GPCR versus cells with vector alone. Similarly constructed NIH3T3 cell lines expressing K-Ras, used as a positive control for NFκB activation, showed a 6-fold increase in NFκB activity over cells with vector alone, comparable with that seen with KSHV-GPCR. Cotransfection of the Pak-PID into these cells along with the luciferase reporters caused a ∼4-fold inhibition of NFκB activity in the KSHV-GPCR cell line, and a ∼11-fold inhibition in the K-Ras cell line, to levels seen with vector alone (Fig. 4,C). Notably, transient transfection with the constitutively active Pak1 (PakL107F) alone showed a similar activation of NFκB (∼6-fold) as cells expressing KSHV-GPCR or K-Ras (Fig. 4,C), suggesting that Pak1 acts as a primary conduit of signaling to NFκB. Finally, as a control, we determined that treatment of KSHV-GPCR cell lines with the IκB inhibitor BAY-11-7082 completely abolished the NFκB activation by KSHV-GPCR (Fig. 4 C). Together, these data indicated that KSHV-GPCR activation of Pak1 is a major conduit of signaling to NFκB.

Because phosphorylation of IκB is a necessary step in the activation of NFκB, we examined the status of IκB in cells expressing KSHV-GPCR. Of the IκBα and IκBβ proteins, IκBα is much more widely studied, and excellent phosphospecific anti-IκBα antibodies are available; hence, we focused analysis on examination of IκBα phosphorylation status. The total cell lysate of cells expressing either the vector or KSHV-GPCR was immunoblotted with phosphospecific antibody directed against the serine residue at position 32 of IκBα subunit. Cells expressing KSHV-GPCR showed a significantly increased amount of phosphorylated IκBα over cells expressing the vector alone (Fig. 5,A). Pretreatment of these cells with PTX or IP-10, both inhibitors of KSHV-GPCR signaling, inhibited phosphorylation of IκBα at serine 32 (Fig. 5,A, i), as did transfection of cells with RacN17 and the PAK-PID dominant-negative mutants of the downstream effector molecules, Rac1 and Pak1 (Fig. 5 A, ii). These results strongly support the idea that Pak1 is involved in KSHV-GPCR signaling to NFκB through IκBα, although involvement of IκBβ cannot be ruled out.

Because most NFκΒ activating signals converge on the IKK signalosome complex, we next evaluated IKK phosphorylation status. IKK is a multisubunit complex, containing two catalytic subunits, IKKα and IKKβ, and a regulatory subunit, which is IKKγ (51). Although both IKKα and IKKβ phosphorylate IκΒ, in the case of microbial and viral infections, it has been established that IKKβ becomes phosphorylated at serine 181, activating the protein to phosphorylate IκΒ at serine 32 and 36 (52). Using antibodies specific to phosphorylated IKK, we detected an ∼7–8-fold increase in activated IKKβ in cells expressing KSHV-GPCR over cells expressing vector (Fig. 5,B). In agreement with our earlier experiments, the amount of activated IKKβ was significantly reduced in cells expressing KSHV-GPCR either pretreated with PTX (70%) or IP-10 (80%; Fig. 5,B, i) or cotransfected with RacN17 (75%) or Pak-PID (90%; Fig. 5, B, ii, and C), as compared with cells expressing KSHV-GPCR. No significant activation of IKKα was observed in the cells under any assay conditions. These results conclusively demonstrated the importance of Pak1 in KSHV-GPCR signaling to NFκB and placed Pak1 activity upstream of IKK.

KSHV-GPCR activates several signaling pathways that might contribute to its ability to activate NFκB either through or independent of Pak1 activation, including the MAPK [MAP/ERK kinase (MEK)/ERK] and PI3K pathways (13, 15, 16, 53). Notably, our data presented above showing PTX inhibition of KSHV-GPCR-dependent Rac1 activation suggest involvement of Gα_i proteins, which can signal to PI3K, in KSHV-GPCR activation of Rac1 and Pak1 (54). To examine the contribution of these other proteins to the KSHV-GPCR to NFκB cell transformation signaling pathway, we investigated the role of the pharmacological agents known to specifically inhibit the Gα_i, PI3K, and MAPK pathways in KSHV-GPCR-mediated cellular transformation. We did this by carrying out soft agar assays with cells expressing KSHV-GPCR treated with PTX (Gα), wortmannin (PI3K), PD98059, or U0126 (MEK/MAPK). We observed fewer colonies with KSHV-GPCR-expressing cells treated with wortmannin, PD98059, U0126, or PTX than the colonies formed by untreated cells expressing the KSHV-GPCR (Fig. 6,A). An inhibition of 82, 75, 68, and 76% of colony formation in soft agar was observed on treatment with PTX, wortmannin, PD98059, and U0126, respectively, indicating that both PI3K and MEK1 signaling contributed to KSHV-GPCR transformation. However, none of these levels of reduction approached those seen with inhibition of Pak1 signaling (Fig. 1 B; data not shown), excluding, for example, PI3K activation as the sole signaling conduit from KSHV-GPCR to Pak1.

Next, we assessed NFκB-dependent transcription in the presence or absence of the same signaling inhibitors. NIH3T3 cells expressing either KSHV-GPCR or K-Ras were transfected with the NFκB promoter-firefly and thymidine kinase promoter-Renilla luciferase reporter plasmids, and 24 h post-transfection, the cells were pretreated for 30 min with compounds targeting Gα_i (PTX), PI3K (wortmannin), or MEK1/2 to ERK (PD98059 and U0126) signaling before assay. Although cells treated with PTX or wortmannin showed a ∼3-fold decrease in NFκB activity over untreated cells expressing KSHV-GPCR (Fig. 6,B), cells treated with PD98059 and U0126 showed marginal or no inhibition of NFκB activity versus that of untreated cells (Fig. 6,B). In contrast, K-Ras-dependent activation of NFκB was inhibited by both wortmannin and PD98059 (∼6-fold and ∼3-fold, respectively; Fig. 6 B). Thus, whereas the KSHV-GPCR- and Ras-transformed cells at least partially required PI3K, only Ras required MEK for NFκB activation.

Finally, it is known that many viruses use the TNF-α pathway to activate NFκB (55). To determine the role of TNF-α in the KSHV-GPCR-Pak1-NFκB signaling axis, we investigated whether TNF-α treatment could enhance KSHV-GPCR-dependent NFκB activation. Although TNF-α could increase NFκB activation by KSHV-GPCR only marginally (Fig. 6,C), it could overcome the inhibition of this signaling caused by RacN17 and Pak-PID by about 50–60% (Fig. 6 C). These results suggest that KSHV-GPCR signaling to NFκB through Pak1 does not require activation of TNF-α, but TNF-α can augment KSHV-GPCR regulation of NFκB through a separate pathway.

Prior studies have addressed signaling from the KSHV-GPCR based on control of end points such as increased vascular endothelial growth factor (56), enhanced cell survival (13), or increased cytokine secretion (57). This is the first study to analyze the signaling requirements for KSHV-GPCR-induced cell transformation. In this work, we have delineated a direct signaling pathway from KSHV-GPCR through Rac1, Pak1, and IKK, culminating in transcriptional activation through NFκB. Furthermore, we have demonstrated a key role for Pak1 activation of NFκB in KSHV-GPCR-mediated cell transformation. Thus, this study extends the known involvement of Pak1 in cell transformation to GPCR oncogenes.

Based on the data amassed in this study, we propose the signaling pathway shown in Fig. 7. In this pathway, KSHV-GPCR induces the activation of Rac1 and Cdc42, most probably through Gα_i, which further promotes the activation of Pak1. This activation cascade may also involve PI3K, because Gα_i has been shown to induce PI3K (58, 59), which stimulates Rac1 (60, 61). Because both PTX and wortmannin inhibit KSHV-GPCR activation of Pak1 (Fig. 3,C) and cell transformation (Fig. 6 A), this pathway could be active in KSHV-GPCR signaling. PI3K also stimulates AKT, which can contribute to Pak1 activation to promote cell proliferation and survival (60). Of relevance, a number of studies have noted that the HIV protein nef interacts and activates Pak1 to promote c-Jun NH₂-terminal kinase activation and increased viral production (62). Both Rac1 to Pak1 and PI3K to Pak1 activation pathways are essential for nef-mediated pathogenesis of HIV (63, 64). Regulation of Pak1 through HIV-nef is also an attractive possibility, because coinfection of HIV is a frequent occurrence in KS patients. Pak1, once activated, further disseminates this signal either directly through NIK or indirectly to IKK leading to phosphorylation of IκB and subsequent activation of NFκB. Finally, activation of RhoA as a consequence of Gα_i activation may augment NFκB stimulation. This use of multiple pathways to activate NFκB would be analogous to the strategy used by Ras to signal to NFκB through the PI3K-Akt-IKKα as well as Raf-Mek-IKKβ pathways (65). After activation of NFκB, it is also reasonable to suppose initiation of a cytokine/autocrine loop involving TNF-α, reinforcing virus maintenance of NFκB activation.

The signaling axis whereby the heterotrimeric G proteins transduce signals from the GPCRs to the MAPK signaling cascades through the Ras, Rac1, and Cdc42 GTPases is evolutionarily ancient (66). For example, in the budding yeast Saccharomyces cerevisiae, the Ste2p mating pheromone GPCR signals through Cdc42p and Ste20p to activate downstream MAPKs Ste11p and Ste7p (67). Pak1 is the mammalian homologue of the Ste20p kinase. In yeast, Ste20p binds the βγ subunits directly in addition to Cdc42, a mechanism that is also conserved for mammalian Pak1 (68). The βγ subunits have been proposed to recruit Pak1 to the membrane where, independent of its kinase activity, it serves as a scaffold to recruit PIX, a GDP/GTP exchange factor. PIX, in turn, reactivates Rac1 and Cdc42 (69, 70). Thus, Pak may act upstream of Rac1 when stimulated by GPCRs, forming a feedback loop that further magnifies potential Pak1 effects. Together, the contributions to Pak activation from Cdc42, Akt, and βγ subunits may explain why we reproducibly observe stronger inhibition of GPCR signaling to components of the NFκB signaling pathway by the Pak PID than by dominant-negative Rac (Fig. 5).

Moving downstream, our data indicate that Pak1-induced phosphorylation of IKKs is a critical step in mediating NFκB activation (Fig. 5), although it does not assign IKKs as direct Pak1 substrates. Although NFκB activation has been demonstrated through both PI3K-Akt and Pak1 in a variety of studies (37), the critical targets of Pak1 in this pathway remain elusive. Pak1 has been reported to transduce signals from Ras, Raf, or Rac1 to NFκB without activating IKKα or IKKβ. However, dominant-negative mutants of both IKKα and IKKβ block the signals from Pak1 to NFκB, suggesting that both are necessary for Pak signaling (37). In studies by Pati et al. (53), dominant-negative mutants of IKKβ and IκBα completely inhibited NFκB activation by KSHV-GPCR in a KS endothelial cell line, whereas a dominant-negative IKKα mutant only partially inhibited NFκB activation. Our data suggest a mode of regulation of NFκB activity by Pak1, wherein upon activation by Pak1 or a related kinase, IKKβ phosphorylates IκBα downstream of Pak1. One possibility, as suggested by Frost et al. (37), is that Pak1 regulates NIK, the protein kinase upstream of IKKα or IKKβ (Fig. 7). Placement of Pak1 at the level of NIK is also consistent with data from several other studies (37, 71, 72). In summary, our study indicates that Gα_i and possibly other heterotrimeric proteins transduce the signal from KSHV-GPCR to the small G protein Rac1 and its effector Pak1, then through IKKβ, IκBα, and NFκB, leading to transformation of NIH3T3 cells. Interestingly, as Pak inhibitors block both Ras and KSHV-GPCR signals to NFκB, Pak1 lies downstream of both. This suggests a convergence of multiple signals through Pak.

NFκB is activated by a diverse set of GPCRs, including receptors for thrombin, platelet-activating factor, and prostaglandin E₂ (73) besides KSHV-GPCR, although the signaling pathways connecting GPCRs and NFκB have not been well-defined in these different systems. Some of the possible pathways put forward include the Ras and the Rho GTPase family pathway. Activation of Ras, and consequently Raf and/or pp90rsk, by the βγ subunits of the heterotrimeric G proteins leads to increased NFκB activity (29). Mutationally active forms of three different Rho GTPases (Rac1, Cdc42, and RhoA) activate NFκB (74). The Ras effectors PI3K, Akt, and Pak1 are also potentially involved in NFκB activation, but the importance of these proteins has not been clearly established in these other systems. Recent reports have shown that KSHV-GPCR enhancement of cell survival and increased cytokine secretion requires NFκB. However, the signaling requirements thus far demonstrated for cell survival and cytokine production are nonequivalent, even though both involve NFκB induction. KSHV-GPCR up-regulation of IL-8 production is dependent on activation of the RhoA GTPase (12), whereas enhancement of cell survival has been studied in the context of the signaling from PI3k to Akt to NFκB (13). Our study has not addressed the possible role of RhoA activity in KSHV-GPCR transformation or NFκB activation, but contribution of RhoA to these processes remains a formal possibility (Fig. 7).

Most viruses activate NFκB to stimulate the production of cytokines such as TNF-α, which in its turn sets up an autocrine loop to keep high levels of constitutive expression of NFκB. The existence of such a loop in KSHV-GPCR-NFκB signaling cannot be ruled out. However, our experiments have established that TNF-α can induce NFκB activation under conditions in which the Pak1 signaling pathway is suppressed, suggesting that TNF-α signals, in part, through a Pak-independent pathway (Fig. 6 C). Together with reports of the involvement of RhoA in KSHV-GPCR activation of NFκB and the data we present herein, it is attractive to postulate that KSHV uses multiple pathways to activate NFκB, ensuring constitutive expression of NFκB-dependent antiapoptotic and growth-promoting genes in the host cell.

The results presented here indicate that signals to Pak and NFκB contribute to cell transformation by KSHV-GPCR. Both PI3K and MAPK pathways are known to be involved in cell transformation signaling. Consequently, we directly examined whether PI3K or MAPK participated in KSHV-GPCR signaling using specific inhibitors. The PI3K inhibitor wortmannin blocked both NFκB activation and cell transformation, whereas PD98059, the MAPK pathway inhibitor, inhibited only cell transformation by KSHV-GPCR but not NFκB activation (Fig. 6, A and B). These data suggest that KSHV-GPCR requires at least two different pathways for cellular transformation, one involving PI3K and/or Pak1-NFκB and the other involving MEK-MAPK but not NFκB. On the other hand, Ras stimulation of NFκB was blocked by both inhibitors (Fig. 6 B), suggesting, as previously observed, that PI3K and MAPK signals are both required for NFκB stimulation by Ras.

Activation of Pak1 by the KSHV-GPCR oncogene corroborates growing reports of involvement of Pak1 and its isoforms in anchorage-independent cell growth and morphological changes in the cell during cell transformation (32, 75). For example, several studies have demonstrated the active role of Pak1 in the increased cell motility, invasion, and angiogenesis exhibited by breast cancer cells as well as in neurofibromatosis 1-mediated cell transformation (34, 76, 77). Pak1 also induces cytoskeletal rearrangements, c-Jun NH₂-terminal kinase activation, and increased viral production in conjunction with the Nef protein of HIV-1 (62) as described previously. In fact, the Nef protein of HIV also participates in the regulation of NFκΒ activity (78) and cellular transformation (79, 80), and Pak1 comes to mind as a common signaling molecule for both HIV-nef and KSHV-GPCR. Epidemiological studies have suggested that KSHV is one of the major determinants in the development of KS (81, 82). All stages of KS, from the initial patch lesions to the more advanced plaque and nodular stages, are characterized by common immunological and histopathological findings, including angiogenesis and an increased concentration of proinflammatory cytokines. The late nodular stage is characterized by outgrowth of tumorous oligoclonal spindle cells and true sarcoma formation with invasion of the underlying dermal structure. Based on the ability of Pak1 to affect signaling involved in survival, immune response, cell migration, and angiogenesis (32), activation of Pak1 has the potential to contribute to all these characteristic KS disorders.

We thank Shrikrishna Dadke, Ya Zhuo, and Guolei Zhou for helpful discussions and comments on the paper.