HCMV pUS28 initiates pro-migratory signaling via activation of Pyk2 kinase
© Vomaske et al; licensee BioMed Central Ltd. 2010
Received: 16 April 2010
Accepted: 7 December 2010
Published: 7 December 2010
Skip to main content
© Vomaske et al; licensee BioMed Central Ltd. 2010
Received: 16 April 2010
Accepted: 7 December 2010
Published: 7 December 2010
Human Cytomegalovirus (HCMV) has been implicated in the acceleration of vascular disease and chronic allograft rejection. Recently, the virus has been associated with glioblastoma and other tumors. We have previously shown that the HCMV-encoded chemokine receptor pUS28 mediates smooth muscle cell (SMC) and macrophage motility and this activity has been implicated in the acceleration of vascular disease. pUS28 induced SMC migration involves the activation of the protein tyrosine kinases (PTKs) Src and Focal adhesion kinase as well as the small GTPase RhoA. The PTK Pyk2 has been shown to play a role in cellular migration and formation of cancer, especially glioblastoma. The role of Pyk2 in pUS28 signaling and migration are unknown.
In the current study, we examined the involvement of the PTK Pyk2 in pUS28-induced cellular motility. We utilized in vitro migration of SMC to determine the requirements for Pyk2 in pUS28 pro-migratory signaling. We performed biochemical analysis of Pyk2 signaling in response to pUS28 activation to determine the mechanisms involved in pUS28 migration. We performed mass spectrometric analysis of Pyk2 complexes to identify novel Pyk2 binding partners.
Expression of a mutant form of Pyk2 lacking the autophosphorylation site (Tyr-402) blocks pUS28-mediated SMC migration in response to CCL5, while the kinase-inactive Pyk2 mutant failed to elicit the same negative effect on migration. pUS28 stimulation with CCL5 results in ligand-dependent and calcium-dependent phosphorylation of Pyk2 Tyr-402 and induced the formation of an active Pyk2 kinase complex containing several novel Pyk2 binding proteins. Expression of the autophosphorylation null mutant Pyk2 F402Y did not abrogate the formation of an active Pyk2 kinase complex, but instead prevented pUS28-mediated activation of RhoA. Additionally, pUS28 activated RhoA via Pyk2 in the U373 glioblastoma cells. Interestingly, the Pyk2 kinase complex in U373 contained several proteins known to participate in glioma tumorigenesis.
These findings represent the first demonstration that pUS28 signals through Pyk2 and that this PTK participates in pUS28-mediated cellular motility via activation of RhoA. Furthermore, these results provide a potential mechanistic link between HCMV-pUS28 and glioblastoma cell activation.
Human cytomegalovirus (HCMV) is a beta-herpesvirus able to establish a life-long persistent infection after primary infection has been cleared. Although infection is ubiquitous in the human population, persistent HCMV infection is commonly asymptomatic in healthy, immunocompetent individuals. However, HCMV causes significant disease in immunosuppressed patients and, despite effective antiviral therapies, HCMV infection is still a significant problem in congenital disease and in bone marrow transplant recipients. Additionally, HCMV has been associated with long-term diseases including the vascular diseases atherosclerosis, restenosis following angioplasty, and chronic allograft rejection following solid organ transplantation [1–4]. Recently, HCMV has been detected in human glioblastomas, and it has been suggested that HCMV exacterbates the progression of this disease . Chemokines and their receptors have been identified as key mediators in chronic inflammatory processes that attend the development of vascular diseases and play a role in tumor development. Herpesviruses manipulate the host chemokine system by regulating the expression of host chemokines and chemokine receptors as well as by encoding chemokine and chemokine receptor homologues. Indeed, HCMV encodes a CXC chemokine (UL146), a putative CC chemokine (UL128), and four potential chemokine receptors (US27, US28, UL33 and UL78). pUS28 is one of the best characterized chemokine receptors and has been implicated in the development of long-term pathologies associated with HCMV infection, including vascular disease and malignancies [6–10].
pUS28 contains homology to the CC-chemokine receptors  and binds to a broad spectrum of CC-chemokine ligands, including RANTES/CCL5, MCP-1/CCL2, MCP-3/CCL7, and MIP-1β/CCL4. pUS28 is unusual in that it is able to bind the CX3C-chemokine Fractalkine/CX3CL1 with high affinity in addition to binding a variety of CC-chemokines [12, 13]. Although the N-terminal 22 amino acids of pUS28 have been shown to be essential for binding of both chemokine classes, CC-chemokines fail to compete with CX3CL1 for receptor occupation. Mutagenesis studies of pUS28 indicate that different N-terminal residues are critical for binding to each chemokine class [12, 14]. Chemokine binding to pUS28 initiates a variety of cellular signaling pathways [15–18]. Additionally, pUS28 has been shown to signal constitutively in the absence of ligand binding in several transformed cell lines . pUS28 mediated migration of arterial SMC requires binding of the CC-chemokines CCL5 or CCL2 , while CX3CL1 inhibits pUS28-mediated SMC migration via the CC-chemokines . However, CX3CL1 does actively promote the robust migration of pUS28 expressing macrophages suggesting that pUS28 has a dual function dependent upon the cell type it is expressed in and the local chemokine environment of that cell . The binding of each chemokine ligand class (CC- vs. CX3C-) results in differential signaling at the level of G-protein coupling to pUS28 . In fact, pUS28 binding to the CC-chemokines promotes SMC migration through the activation of Gα12/13 and the small G-protein RhoA. pUS28-mediated SMC migration can be blocked by expression of a dominant negative RhoA or RhoA effector associated kinase-1 (ROCK-1) .
pUS28 signals through the non-receptor tyrosine kinases focal adhesion kinase (FAK) and Src in a ligand-dependent manner, and this activity is required for induction of pUS28-mediated SMC migration. . FAK is comprised of a central kinase domain flanked on one side by an N-terminal FERM (erythrocyte band 4.1-ezrin-radixin-moesin) domain, which is involved in linking FAK to integrins and growth factor receptors and provides regulation of the Tyr-397 autophosphorylation site [21, 22]. Additionally, the FERM domain regulates FAK kinase activity via direct interaction with the kinase domain thereby blocking the access of substrates to the catalytic cleft . The F.A.T. (focal adhesion targeting) domain is located C-terminal of the central kinase domain and is comprised of multiple protein-protein interaction motifs. FAK tyrosine phosphorylation following cellular stimulation is enhanced by its association with Src-family PTKs at Tyr-397. pUS28-mediated SMC migration is sensitive to the Src inhibitor PP2 and mutation of the FAK autophosphorylation site at Tyr-397 blocks SMC migration in a dominant-negative fashion. Conversely, a kinase-negative FAK mutant (FAK R454K) had no effect on pUS28-mediated SMC migration  suggesting that FAK acts as a protein scaffold rather than an active kinase in the signaling cascade leading to SMC migration.
Proline-rich Tyrosine kinase 2 (Pyk2), also a non-receptor tyrosine kinase, is highly related to FAK and plays an important role in cell motility . Recently, Pyk2 has been shown to be involved in the CCR5-mediated chemotaxis of dendritic cells via binding of HIV-1 gp120 . In addition, Pyk2 is critical for Angiotensin II (AngII)-mediated migration of vascular SMC (VSMC), in which Pyk2 mediates activation of RhoA and its effector kinase ROCK via phosphorylation of PDZ-Rho-GEF [26, 27].
FAK and Pyk2 share an overall 45% sequence homology with 60% identity within the catalytic domain and also have analogous sites for tyrosine phosphorylation and Src binding . Despite their striking sequence similarity, it is becoming increasingly clear that these molecules play very different roles in signaling cascades leading to cellular migration. Pyk2 is expressed in cells of the brain, hematopoetic cells, osteoblasts, some types of epithelium, SMC and fibroblasts . Unlike FAK, Pyk2 activation has a significant dependence on intracellular calcium levels [28, 29]. Calcium-mediated regulation of Pyk2 activity proceeds via binding of Ca2+/calmodulin to the Pyk2 FERM domain, allowing homodimerization and autophosphorylation at Tyr-402 . Autophosphorylation of Pyk2 is believed to be a critical initial step in the activation of Pyk2 via the recruitment of Src to the phosphorylated Tyr-402 site, resulting in further tyrosine phosphorylation of Pyk2 [31, 32]. However, Pyk2 activation can proceed by both Src-dependent [31, 33–35] and Src-independent [36, 37] mechanisms. Additionally, Pyk2-mediated phosphorylation of paxillin has been demonstrated to be independent of Tyr-402 autophosphorylation in SYF cells . Collectively these data demonstrate that Tyr-402 autophosphorylation is not absolutely required for Pyk2 activity. While Pyk2 can partially compensate for the lack of FAK in fibroblasts derived from FAK-/- mouse embryos, Pyk2 alone is not sufficient to reconstitute a migratory phenotype in these cells . Pyk2 and FAK interact in signaling events leading to cellular migration [38–41], but the relationship between these two proteins remains poorly defined and appears to be highly signal and cell context-specific Additionally, phosphorylation kinetics differ between the two proteins .
Here, we examined the involvement of Pyk2 in pUS28-induced signaling and SMC migration. Pyk2 is activated in SMC expressing pUS28 following the addition of ligand in a calcium-dependent manner. Expression of the autophosphorylation site mutant of Pyk2 (F402Y) blocks SMC migration in a dominant-negative fashion by preventing RhoA activation, while the kinase-inactive mutant has no effect on migration. Pyk2 forms an active kinase complex containing several novel Pyk2 binding proteins in CCL5 stimulated SMC expressing pUS28. Suprisingly, the autophosphorylation null mutant Pyk2 F402Y did not abrogate the formation of an active Pyk2 kinase complex but did prevent the pUS28-mediated activation of RhoA. Lastly, we demonstrate the pUS28-mediated activation of RhoA via Pyk2 also occurs in U373 glioblastoma cells. The Pyk2 kinase complex in U373 is distinct from the complex observed in RSMC and contains a number of proteins known to participate in glioma tumorigenesis. These results provide a potential mechanistic link between pUS28 signaling and glioma pathogenesis.
The life-extended telomerized pulmonary artery SMC line, PAT1  were maintained in Medium 199 supplemented with 20% fetal calf serum (FCS) and penicillin-streptomycin-L-glutamine (PSG; Gibco). For migration experiments described below, PAT1 cells were utilized between passage 5 and 30 post-telomerization. Mouse FAK-/- fibroblasts were maintained on gelatin coated culture dishes in Dulbecco's modified Eagle's Medium (DMEM) supplemented with 10% FCS, PSG, non-essential amino acids (Cellgro), and G418 (Sigma; 500 μg/ml) as previously described [34, 42]. FAK-/- cells used in experiments were between passage 5 and 15. Primary F344 rat SMC (RSMC) were maintained in Dulbecco's modified Eagle's Medium (DMEM) with 10% FCS and PSG. RSMC were used between passage 5 and 20. U373 glioblastoma cells were maintained in DMEM supplemented with 10% FCS and PSG. U373 were used between passage 5 and 20.
Recombinant human CCL5 was purchased from R&D Systems and BAPTA-AM was from Sigma-Aldrich (A1076). Anti-Pyk2 (N-19), anti-phospho-402-Pyk2, anti-myc (9E10), anti-RhoA and anti-HA (F-7) antibodies were purchased from Santa Cruz Biotechnology. Secondary anti-mouse, anti-goat and anti-rabbit horseradish peroxidase (HRP)-conjugated antibodies were purchased from Amersham. Goat anti-mouse Alexa Fluor-488 was purchased from Molecular Probes (A11017).
The adenovirus expressing pUS28-HA was previously described [20, 43, 44]. Myc-tagged Pyk2 WT, Pyk2 A457K (kinase inactive mutant) or Pyk2 F402Y (autophosphorylation mutant)  were constructed by cloning of the cDNA into pAdTet7, which contains the tet-responsive enhancer within a minimal CMV promoter followed by the SV40 late poly(A) cassette. Recombinant adenoviruses were produced by co-transfection of 293 cells expressing the Cre-recombinase with pAdTet7 constructs and adenovirus DNA (Ad5-Ψ5, an E1A/E3-deleted adenovirus genome) . Recombinant adenoviruses were expanded on 293-Cre cells and the bulk stocks were titered on 293 cells by limiting dilution. Gene expression was driven by co-infection with an adenovirus expressing the Tet-off transactivator (Ad-Trans) .
Cell migration assays were performed as previously described . PAT1 cells were infected with HCMV (MOI 10) for 2 hrs followed by co-infection with Ad-Trans and Ad-Pyk2 (WT, A457, F402Y) at MOI of 50 for an additional 2 hrs. Subsequently, the transwells were transferred to new 12-well plates containing fresh medium with CCL5 (50 ng/ml) added to the lower chamber. Cells migrating into the lower chamber were counted at 48-72 hours post-infection (hpi) using a Nikon TE300 microscope at magnification 10×. Experiments were performed in at least triplicate wells and ten random fields were read in each well. The average number of cells per well was determined by multiplying the average number of cells per 10× field by the number of fields per well. Mean and standard deviation were calculated. Pyk2 recombinant protein levels were monitored by western blotting and equalized by adjusting the adenoviral vector MOI accordingly.
Subcellular localization of adenovirus-expressed proteins was visualized by fixing adenovirus-infected SMC in 4-well chamber slides (Nunc Lab-Tek) at 24 hpi in 2% paraformaldehyde for 15 min. Cells were then permeablized with PBS containing 0.2% Saponin and 0.02% BSA (sapPBS) for 15 min. Anti-myc primary antibody was diluted 1:500 in sapPBS and cells were incubated for 1 hr at room temperature (r.t.), washed twice with sapPBS, and incubated with goat anti-mouse Alexa Fluor-488 diluted 1:1000 in sapPBS for 30 min at r.t. After two more washes in sapPBS cells were visualized on a Deltavision deconvolution fluorescence microscope (Applied Precision).
FAK-/- or U373 cells plated onto 6-well culture dishes at 50% confluence were serum-starved for 6 hrs. The cells were infected with Ad-Trans and Ad-pUS28 at MOI 150 with Ad-Pyk2 or Ad-Pyk2 F402Y at MOI 50 and placed in serum free medium. For calcium chelation assays, cells were pre-treated with 50 μM BAPTA-AM for 30 min prior to stimulation. At 16 hpi, the cells were stimulated with CCL5 (40 ng/ml) and then scraped in 2× Laemmeli's Sample Buffer. Unboiled samples were loaded on 10% SDS-PAGE. The gels were transferred to Immobilon-P membranes then blocked in PBS containing 3% milk + 0.1% Tween 20 for 15 min at r.t. The primary anti-Phospho-Y402 Pyk2 rabbit polyclonal antibody was added at 1:1000 dilution in blocking buffer for 1 hr at r.t. The blots were washed with TBS-Tween buffer (10 mM Tris pH 7.2, 100 mM NaCl, 0.2% Tween-20). The secondary antibody (goat anti-rabbit conjugated to HRP) was added at 1:20,000 dilution in blocking buffer for 30 min at r.t. After washing 3 times with TBS-Tween buffer and incubation with ECL reagents, the blots were visualized by autoradiography. Blots were dried to inactivate the HRP secondary reagent and then reprobed as above with anti-Pyk2 goat polyclonal antibody at 1:1000 dilution in blocking buffer followed by secondary donkey anti-goat conjugated to HRP at 1:100,000 dilution.
FAK-/- or U373 cells plated onto 6-well culture dishes at 50% confluence were serum-starved for 6 hrs. The cells were infected with Ad-Trans only or co-infected with Ad-Trans and Ad-pUS28 at MOI 150 with or without Ad-Pyk2 or Ad-Pyk2 F402Y at MOI 50 and placed in serum free medium. After 16 hrs, the cells were stimulated with CCL5 (40 ng/ml) and then scraped in equal volumes modified RIPA buffer (10 mM Tris pH7.4, 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, 1% Sodium Deoxycholate, 0.1% SDS) containing protease inhibitor cocktail (100 mM AEBSF, 80 μM aprotinin, 1.5 mM E-64, 2 mM leupeptin hemisulfate, and 1 mM pepstatin A) at 0 (unstimulated), 2.5, 5, or 10 minutes post-stimulation. Cell lysates were homogenized via sonication, then incubated with 20 μl Rho Assay Reagent (Rhotekin-RBD-GST Agarose; Upstate) for 1 hr at 4°C, then washed 3 times with 1 ml of RIPA. The final bead pellet was resuspended in 40 μl of 2× Laemmli's sample buffer, boiled and then run on 10% SDS-PAGE. The gels were transferred to Immobilon-P membranes then blocked in blocking buffer (PBS containing 3% Milk + 0.1% Tween 20) for 15 min at r.t. The primary anti-RhoA mouse monoclonal antibody was added at 1:1000 dilution in blocking buffer for 1 hr at r.t. The blots were washed with TBS-Tween buffer. The secondary antibody (goat anti-mouse conjugated to HRP) was added at 1:40,000 dilution in blocking buffer for 30 min at r.t. After washing 3 times with TBS-Tween buffer and incubation with ECL reagents, the blots were visualized by autoradiography.
In Vitro Kinase Assays. In vitro kinase assays were performed on on immunoprecipitated Pyk2 from F344 aortic RSMC. Cells were plated in 10 cm culture dishes and serum starved for 24 hrs. The RSMC were co-infected with Ad-Trans and Ad-Pyk2 or Ad-Pyk2 F402Y at an MOI of 50 with or without Ad-pUS28 at an MOI of 150. After 16 hrs, the cells were stimulated with recombinant human CCL5 (40 ng/ml) and then harvested at times 0 (unstimulated), 2.5, 5, 10, and 15 minutes post ligand addition. Cells were rinsed in PBS and lysed in modified RIPA containing protease inhibitor cocktail and 200 μM sodium orthovanadate. Total Pyk2 was immunoprecipitated using mouse anti-myc tag monoclonal antibody and Protein-A/G conjugated agarose beads (Santa Cruz Biotechnology). Precipitation reactions were washed one time in Triton-only lysis buffer (modified RIPA without sodium deoxycholate and SDS), two times in HNTG buffer (50 mM HEPES, 150 mM NaCl, 1% Triton, 10% Glycerol, pH 7.4), and two times in kinase buffer (20 mM HEPES, 10 mM MgCl2, 10 mM MnCl2, 150 mM NaCl, 10% Glycerol pH 7.4) and then resuspended in 50 μl kinase buffer plus 10 μCi 32P-γ-ATP (Perkin-Elmer). The kinase reaction was allowed to proceed for 30 minutes at r.t. then stopped by the addition of 2× Laemmeli's sample buffer. After boiling for 5 minutes, the reactions were run on 10% SDS-PAGE and transferred to Immobilon-P membranes. The incorporation of 32P-g-ATP was visualized via autoradiography. After allowing radioactivity to decay, the membranes were further analyzed via western blot as above for the presence of Pyk2-myc and pUS28-HA.
Proteomics analysis was performed on immunoprecipitated Pyk2 from F344 aortic RSMC or U373 glioblastoma cells. Cells were plated in 15 cm culture dishes. The RSMC were serum starved for 24 hrs prior to co-infection with Ad-Trans and Ad-Pyk2 at an MOI of 50 with or without Ad-pUS28 at an MOI of 150. After 16 hrs, the cells were stimulated with recombinant human CCL5 (40 ng/ml) and then triplicate dishes were harvested at times 0 (unstimulated), 5, 10, 15, 30 and 60 minutes post ligand addition. Cells were rinsed in PBS and lysed in modified RIPA containing protease inhibitor cocktail and 200 μM sodium orthovanadate. Total Pyk2 was immunoprecipitated using mouse anti-myc tag monoclonal antibody and Protein-A/G conjugated agarose beads (Santa Cruz Biotechnology). Precipitation reactions were washed twice with RIPA. Pyk2 and associated proteins were eluted with 200 μl of Pierce Gentle Ag/Ab Elution buffer. Triplicate eluates were combined and the samples were sequenced by tandem mass spectrometry analysis at the EMSL user facility at Pacific Northwest National Laboratories.
Proteins from the Pyk2 complexes were desalted and concentrated with Amicon ultra filtration units (Millipore, Billerica, MA). Protein concentration was determined using BCA protein assay (Thermo Scientific) and the concentrated protein was denatured by the addition of trifluoroethanol (TFE) to a final concentration of 50% and heating to 37°C for 60 min. Denatured proteins were reduced with DTT (2 mM final concentration) and diluted fivefold with 50 mM NH4HCO3. Methylated, sequencing-grade porcine trypsin (Promega, Madison, WI) was added at a substrate-to-enzyme ratio of 50:1 (mass to mass) and incubated at 37°C for 3 hrs. The peptides were concentrated with a speed vac and stored at 80°C until analysis.
Peptide samples were analyzed using an automated custom built capillary LC system containing a four capillary column system (Livesay et al, Anal Chem 2008, vol 80, page 294). Eluate from the LC was coupled directly to a hybrid linear ion-trap-orbitrap (LTQ_Orbitrap, Thermo Electron Corp.). The reverse-phase capillary column was prepared by slurry-packing 3-micron Jupiter C18 bonded particles (Phenomenex, Torrence, CA) into a 65 cm long, 75-micron-inner diameter fused silica capillary (Polymicro Technologies, Phoenix, AZ). After peptide loading onto the column, the mobile phase was held at 100% A (0.05% trifluoroacetic acid (TFA) and 0.2% acetic acid in water) for 20 min, followed by a linear gradient from 0 to 70% buffer B (0.1% TFA in 90% acetonitrile, 10% water) over 80 min with a flow rate of ~500 nL/min. Orbitrap spectra (AGC 1 × 106) were collected from 400-2000 m/z at a resolution of 100 k followed by data-dependent ion trap tandem mass spectra (AGC 1 × 104) of the six most abundant ions using a collision energy of 35%. The heated capillary was maintained at 200°C, and the ESI voltage was held at 2.2 kV.
The SEQUEST algorithm was run on each of the datasets against the human.fasta from the International Protein Index database (version 3.54, 75,419 entries, released January 2009). MS/MS peaks were generated by "extract_msn.exe," part of the SEQUEST software package. A peptide was considered to be a match by using a conservative criteria set developed previously by Yates and coworkers . Briefly, peptides were retained if they met the following criteria: 1). SEQUEST DelCN value of ≥0.1 and 2). SEQUEST correlation score (Xcorr) > 1.9 for charge state 1+ and fully tryptic; Xcorr > 2.2 for charge state 2+ and fully or partially tryptic; Xcorr > 3 for charge state 2+ regardless of tryptic state; Xcorr > 3.75 for charge state 3+ for fully or partially tryptic.
In order to further characterize the role of Pyk2 in pUS28-mediated cellular migration, we examined phosphorylation of Pyk2 at Y402 in Ad-pUS28 infected mouse fibroblasts lacking FAK (FAK-/-). FAK-/- cells are known to have elevated levels of endogenous Pyk2, partially compensating for the lack of FAK [34, 47]. We have utilized FAK-/- for pUS28 signaling assays in previous studies [17, 18]. FAK-/- exhibit low baseline activation of pro-migratory signaling molecules and are therefore a highly inducible system in which to study these signaling events. Furthermore, using this system we can study pUS28 signaling to Pyk2 independent of FAK.
Pyk2 binding partners in pUS28-expressing RSMC
Cytoskeleton Structure and Dynamics
β-actin, α1-actin, F-actin capping proteins-α1, -α2*, and -β, Gelsolin [51, 72, 88, 89], Myosin-9, -10, -11, Myosin light chain-6, Tropomodulin-3, Tropomyosin-α1, α3, α4 -β [72, 90, 91], Tubulin-β-2C, Vimentin 
Inflammation and Immunity
Immunoglobulin lambda light chain , NFκB p105
We tested the participation of Pyk2 in the pUS28-mediated activation of RhoA in this system. Similar to results obtained in FAK-/- fibroblasts, pUS28-specific activation of RhoA is readily observed in U373. pUS28-expressing U373 showed 49% activation of RhoA at 5 min post addition of CCL5. Adenovirus expression of WT Pyk2 resulted in similar activation of RhoA, but altered the timing of RhoA activation such that the peak of RhoA activation was seen at 2.5 min post addition of CCL5. Overexpression of the Pyk2 F402Y mutant prevents pUS28-mediated activation of RhoA resulting in less than 1% activation of RhoA throughout the timecourse of stimulation (Figure 6B).
Pyk2 binding partners in pUS28-expressing U373
Cytoskeleton Structure and Dynamics
Inflammation and Immunity
Immunoglobulin lambda light chain , NFκB p105
Migration of vascular SMC is induced by a variety of stimuli including growth factors, cytokines, extracellular matrix (ECM) components and peptide hormones. Pro-migratory signaling in SMC is best understood in the context of platelet-derived growth factor (PDGF) binding through receptor tyrosine kinases PDGFR-α and β. PDGF-mediated migration in SMC is characterized by Src-dependent activation of FAK, RhoA mediated activation of ROCK and subsequent phosphorylation and inhibition of myosin light chain phosphatase as well as activation of various other effector molecules via ERK, p38MAPK and JNK MAPK families . Angiotensin II (Ang II) is another well-studied pro-migratory signal in SMC. Ang II binding to the GPCR angiotensin receptor type 1 (AT-1) initiates the only pro-migratory signaling cascade in SMC previously reported to involve Pyk2 . In this signaling context, Pyk2 is activated by PKCδ  and signals to Rho/ROCK via association with PDZ-RhoGEF . Additionally, Pyk2 has been identified as an essential signaling scaffold linking GPCR signaling to MAP kinase activation via Src activation upon binding to Tyr-402 autophosphorylated Pyk2 . Activation of JNK and ERK MAP kinase pathways are necessary for Ang II-mediated SMC migration .
The results presented herein represent the first example of a chemokine-mediated pro-migratory signaling cascade in VSMC involving Pyk2. We have previously demonstrated that chemokine binding to the viral chemokine receptor pUS28 causes SMC migration via a Src-mediated activation of FAK and ERK , and the Gα12-mediated activation of RhoA and ROCK . These results refine our understanding of the pUS28 receptor-proximal events leading to the activation of RhoA via the ligand-dependent phosphorylation of Pyk2 at Tyr-402. Pyk2 kinase activity is dispensable for pUS28-mediated migration in SMC, suggesting that Pyk2 participates primarily as a signaling scaffold in pUS28 pro-migratory signaling. We have previously demonstrated that pUS28-induced SMC migration is sensitive to the Src inhibitor PP-2 and that FAK functionally interacts with Src in pUS28-expressing cells stimulated with CCL5. We further demonstrated that Src activation is upstream of FAK in this pathway . Similarly, pUS28 can signal to Pyk2 in FAK-/- fibroblasts suggesting that Pyk2 activation is upstream or independent of FAK in this signaling cascade. Our proteomics analysis of Pyk2 binding partners in the context of pUS28 signaling revealed several novel proteins that shed new insight into the function of Pyk2 in pro-migratory signaling. In particular, we show that Pyk2 interacts with several tropomyosin (TM) isoforms and the TM interacting protein tropomodulin isoform 3 (Tmod3) in response to pUS28 signaling. These proteins are well-characterized regulators of actin polymerization and cytoskeletal dynamics . Tmod3 is broadly expressed in cells with dynamic actin filament structures and is unique among tropomodulin proteins in that, in addition to regulating the capping of actin filament pointed ends, it binds free actin monomers and may regulate the nucleation of new actin filaments . Our discovery of an association of Pyk2 with TM isoforms is novel and particularly intriguing. TM proteins regulate the structure and specific properties of actin filaments but the mechanisms of TM regulation remain obscure . One possible mechanism of TM regulation is phosphorylation. Indeed, the phosphorlyation of TM1 in response to ERK signaling in endothelial cells increases the formation of both actin stress fibers and focal adhesions contributing to increased cellular contractility . Our results suggest that the Pyk2 signaling complex participates in cytoskeleton remodeling via modification of Tmod and TM proteins. Taken together, these results suggest that Pyk2 acts together with Src, FAK and other adaptor molecules to promote pro-migratory signaling and that Pyk2 phosphorylation at Tyr-402, is required for the downstream activation of RhoA. Our finding that the scaffolding activity of Pyk2 is separable from its kinase activity is novel in the literature. In the context of AngII-mediated SMC migration as well as migration of endothelial cells, Pyk2 kinase activity is an essential component of pro-migratory signaling [26, 74].
In this paper, we explore the contribution of pUS28 signaling to Pyk2 in migration processes associated with HCMV-mediated vascular disease. However, HCMV has been implicated in other disease processes in which the induction of cellular migration may play a role. In particular, HCMV infection has been associated with several human malignancies including prostate and colon cancers as well as glioblastoma multiforme (GBM) . GBM is the most common and most malignant type of brain cancer and currently has no effective treatment . HCMV DNA is present in over 90% of human malignant gliomas [1, 77], suggesting a role for HCMV in glioma oncogenesis. A recent report has strengthened the connection between HCMV and GBM by demonstrating that vaccination against the HCMV tegument protein pp65 can improve survival in GBM patients . However, the molecular mechanisms of HCMV participation in glioma development remain unclear.
pUS28 signaling has recently been reported to produce a transformed phenotype in transfected NIH-3T3 cells characterized by increased proliferative capacity, and increased expression of the pro-angiogenic marker VEGF. In addition, the study demonstrated that WT HCMV infection of a human glioblastoma cell line (U373) induces VEGF expression while a pUS28 deletion mutant of HCMV fails to induce VEGF upregulation in glioblastoma cells  implicating pUS28 signaling in glioma pathogenesis. However, this study did not explore the contribution of pUS28 ligand binding to the observed phenotypes.
In addition to its role in vascular remodelling, Pyk2 has been identified as a pro-metastatic and pro-angiogenic molecule in several cancer models including small cell lung cancer [79, 80], prostate cancer [81–83] and glioblastoma [41, 58–60]. A number of elegant studies by Lipinski et. al. have established roles for Pyk2 and FAK in glioma tumorigenesis both in vitro and in vivo. In these studies, Pyk2 is demonstrated to promote migration and invasion of glioma cells while FAK controls their proliferation. The authors demonstrate that the migratory capacity of individual glioma cell lines can be correlated to the relative levels of Pyk2 expression and activation. Perhaps the most compelling evidence for the role of Pyk2 and FAK in glioma tumor progression utilized an in vivo xenograft model of glioblastoma in mice. In this study, Pyk2 deficient cells were observed to invade only locally and survival was significantly increased when mice were engrafted with Pyk2 or FAK deficient glioma cells compared to mice engrafted with wild-type glioma cells . Our current study and previous work characterizing the pUS28-mediated activation of FAK suggest a mechanism for acceleration of glioma tumor formation by HCMV via pUS28-mediated signaling through the critical mediators of glioma tumor progression, Pyk2 and FAK. Our results are consistent with existing literature in which the Pyk2 FERM domain is established as a target for reducing glioma cell migration [58, 84]. Pyk2 undergoes dimerization and autophosphorylation upon FERM domain interaction with Ca2+/calmodulin . Therefore inhibition of the Pyk2 FERM domain by mutation or antibody targeting may prevent dimerization and phosphorylation of Pyk2 at Tyr-402 thereby preventing the activation of RhoA.
We observed the formation of a cell type specific Pyk2 complex in response to pUS28 signaling in U373 glioblastoma cells. This complex included several proteins previously implicated in glioma tumorigenesis, but never before identified as Pyk2 binding partners. Our data suggests at least two novel mechanisms for the participation of Pyk2 and pUS28 in glioma tumorigenesis. First, we identified TCF-4 and AF-17 binding to Pyk2 in response to pUS28 signaling. Both proteins are members of the Wnt/β-catenin signaling pathway that has recently been linked to glioma progression [65, 66]. Indeed, siRNA mediated inhibition of this pathway decreases glioma proliferation and invasion capacity . Interestingly, Pyk2 has been shown to phosphorylate β-catenin in endothelial cells  suggesting that Pyk2 may be involved in the activation of canonical β-catenin signaling in multiple cell types. Second, we discovered that pUS28 signaling caused Pyk2 to associate with three members of the Hsp70 family of molecular chaperone proteins. Interestingly, all three of these proteins are known to be specifically upregulated in glioma tissue and play various roles in the progression of glioma tumorigenesis [61–63] as well as other cancer models . Mortalin, in particular, is functionally regulated via tyrosine phosphorylation . Our results suggest that Pyk2, or kinases associated with Pyk2 may play a role in regulating the function of multiple Hsp70-family chaperones, thereby contributing to the proliferation and survival of tumor cells.
These findings refine our understanding of the pro-migratory signaling pathways activated by pUS28 in SMC and establish Pyk2 as an attractive target for the treatment of CMV-mediated vascular disease because it acts early in the pUS28 pro-migratory signaling cascade and its expression is restricted to a few cell types, reducing the opportunities for off-target effects. Further, these results suggest a role for HCMV pUS28 in pathologies associated with aberrant activation of Pyk2 including some malignancies.
In this report, we determined the involvement of Pyk2 in pUS28 signaling and cellular migration (Figure 7). We found that Pyk2 autophosphorylation activity, but not kinase activity, is necessary for HCMV-mediated SMC migration. Furthermore, pUS28 signaling causes ligand-dependent phosphorylation of Pyk2 at Tyr-402 independent of pUS28 signaling to FAK. Pyk2 is incorporated into an active kinase complex in CCL5 stimulated SMC expressing pUS28. Pyk2 associates with a number of other proteins in response to pUS28 signaling including kinases, signaling molecules, cytoskeletal proteins, and molecular chaperones. We determined that although the Pyk2 autophosphorylation mutant (F402Y) does not prevent the formation of an activated kinase complex, it exerts a dominant negative effect on SMC migration by preventing the pUS28-mediated activation of the small G-protein RhoA. Importantly, the effect of pUS28 activating Pyk2 is not cell-type specific as we observed comparable activation of Pyk2 and RhoA via pUS28 signaling in U373 glioblastoma cells. This finding suggests that pUS28 signaling to Pyk2 may be important in glioma cell motility. In addition to establishing Pyk2 as a potential target in the treatment of HCMV-mediated vascular disease, and establishing a number of novel Pyk2-associated proteins, these results provide a potential link between pUS28 and CMV-associated malignancy.
The authors would like to thank David D. Schlaepfer for his generous contribution of the cDNA clones used to make the Pyk2 adenoviruses. This work was supported in part by the National Institutes of Health (NIH) grant HL083194 (to DNS) and HL088603 (to JAN). The proteomics work was performed as part of a Science Theme User Proposal from the Pacific Northwest National Labs William R. Wiley Environmental Molecular Science Laboratory (EMSL) under the theme of Biological Interactions and Dynamics (Project #30468). EMSL is a national scientific user facility sponsored by the Department of Energy's Office of Biological and Environmental Research and located at Pacific Northwest National Laboratory
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (<url>http://creativecommons.org/licenses/by/2.0</url>), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.