QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) All Versions of this Article: mol.104.007799v1 67/6/2102    most recent Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Peavy, R. D. Articles by Hepler, J. R. Search for Related Content PubMed PubMed Citation Articles by Peavy, R. D. Articles by Hepler, J. R.

Differential Effects of Gq, G14, and G15 on Vascular Smooth Muscle Cell Survival and Gene Expression Profiles

Department of Pharmacology (R.D.P., K.B.H., A.L., R.B.F., K.X., C.J.L., T.J.M., J.R.H.) and Biomolecular Computing Resource (T.T.L., K.G.), Emory University School of Medicine, Atlanta, Georgia

Received September 30, 2004; accepted March 23, 2005

	Abstract

Top Abstract Materials and Methods Results Discussion References

 
Gq family members (Gq, G11, G14, and G15/16) stimulate phospholipase C (PLC) and inositol lipid signaling but differ markedly in amino acid sequence and tissue distribution predicting unappreciated functional diversity. To examine functional differences, we compared the signaling properties of Gq, G14, and G15 and their cellular responses in vascular smooth muscle cells (VSMC). Constitutively active forms of Gq, G14, or G15 elicit markedly different responses when introduced to VSMC. Whereas each G stimulated PLC to similar extents when expressed at equal protein levels, Gq and G14 but not G15 initiated profound cell death within 48 h. This response was the result of activation of apoptotic pathways, because Gq and G14, but not G15, stimulated caspase-3 activation and did not alter phospho-Akt, a regulator of cell survival pathways. Gq and G14 stimulate nuclear factor of activated T cell (NFAT) activation in VSMC, but G-induced cell death seems independent of PKC, InsP₃/Ca²⁺, and NFAT, in that pharmacological inhibitors of these pathways did not block cell death. Gene expression analysis indicates that Gq, G14, and G15 each elicit markedly different profiles of altered gene sets in VSMC after 24 h. Whereas all three G stimulated changes (2-fold) in 50 shared mRNA, Gq and G14 (but not G15) stimulated changes in 221 shared mRNA, many of which are reported to be pro-apoptotic and/or involved with TNF- signaling. We were surprised to find that each G also stimulated changes in nonoverlapping G-specific gene sets. These findings demonstrate that Gq family members activate both overlapping and distinct signaling pathways and are more functionally diverse than previously thought.

Although Gq class members share a capacity to activate PLC, they also differ markedly in their biochemical properties and tissue distribution. Gq/11, G14, and G15/16 exhibit limited amino acid sequence identity, with only 57% overall (compared with 90% for Gs and 85% for Gi) and only 30% within the first 40 amino acids. Gq and G11 are the most similar of this class and are expressed fairly ubiquitously (Strathmann and Simon, 1990). G14 has a more limited expression pattern (kidney, liver, lung, testis), whereas G15/16 is limited to hematopoietic tissue (Amatruda et al., 1991; Wilkie et al., 1991).

Despite these differences, established models suggest that Gq family members are functionally redundant and that cellular responses as a result of activation of linked receptors are caused by PLC and downstream calcium/PKC pathways. However, growing evidence indicates that many receptors and G-mediated effects do not involve inositol lipid signaling. For example, several Gq/11-linked receptors have been shown to stimulate mitogen-activated protein kinase pathways independent of activation of inositol lipid signaling, including 1A-adrenergic, thrombin, and metabotropic glutamate 5 receptors (Berts et al., 1999; Seo et al., 2000; Peavy et al., 2001). G subunits released from G-GTP also regulate a growing list of signaling proteins (Gutkind, 1998). Furthermore, introduction of constitutively active Gq family members elicit diverse responses in cell growth, survival, and differentiation depending on the identity of the G and the cell type involved. Together, these findings indicate receptors and linked G proteins activate multiple parallel signaling pathways.

Consistent with this diversity of cellular responses, additional binding partners have been reported for Gq distinct from PLC including RGS proteins, Rho guanine nucleotide exchange factor, GRK2, and Bruton's tyrosine kinase (Bence et al., 1997; Heximer et al., 1997; Carman et al., 1999; Sagi et al., 2001). Whereas RGS proteins clearly modulate the signaling capacity of target G, they and other binding partners may also serve as novel G effectors to activate parallel downstream signaling proteins and pathways. Together, these findings predict that Gq family members activate both overlapping and distinct signaling pathways resulting in diverse cellular responses. However, the relative contribution of PLC versus distinct pathways to signaling by Gq family members is unknown.

To examine signaling diversity among Gq family members, we compared the signaling properties of Gq, G14, and G15 and their cellular responses in vascular smooth muscle cells (VSMC). We report that Gq, G14, and G15 elicited markedly different global cellular responses when expressed in VSMC. Whereas each G stimulated PLC to nearly identical extents when expressed at equal protein levels, Gq and G14, but not G15, initiated cell death. This cell death was caused by activation of apoptotic pathways because Gq and G14, but not G15, stimulated caspase-3 activation. Gq and G14 also activated NFAT, but G-induced cell death is independent of PKC and inositol triphosphate/Ca²⁺ and NFAT because pharmacological inhibitors of these pathways did not block cell death. Gene expression studies indicate that Gq, G14, and G15 each elicit surprisingly different profiles of altered gene sets consisting of both overlapping and distinct sets of mRNA that increase or decrease. Gq and G14, but not G15, stimulated changes in 221 shared genes, many with reported roles in apoptosis. Taken together, these findings suggest that Gq family members activate both overlapping and distinct signaling pathways in VSMC, independent of PLC activation, to elicit unique cellular responses.

	Materials and Methods

Top Abstract Materials and Methods Results Discussion References

 
Materials. The luciferase assay system and the CytoTox 96 nonradioactive cytotoxicity assay kit to measure lactate dehydrogenase (LDH) were purchased from Promega (Madison, WI). myo-[³H]inositol was purchased from PerkinElmer Life and Analytical Sciences (Boston, MA). Fura-2 acetoxymethyl ester, Hoechst 33342 dye, rhodamine 110, and bis-L-aspartic acid amide fluorescent caspase substrate were purchased from Molecular Probes (Eugene, OR). The caspase-3, Akt, and phosphospecific Akt antibodies were purchased from Cell Signaling Technology (Beverly, MA). The monoclonal Glu-Glu ("EE") antibody was purchased from Covance Research Products (Princeton, NJ). Horseradish peroxidase-conjugated goat anti-rabbit IgG was obtained from Bio-Rad (Hercules, CA). Horseradish peroxidase-conjugated goat anti-mouse IgG was obtained from Jackson Immunoresearch Laboratories (West Grove, PA). Restriction enzymes and modifying enzymes were purchased from Stratagene (La Jolla, CA) and New England Biolabs (Beverly, MA). 2-APB and BIS I were purchased from Calbiochem (San Diego, CA). Cyclosporin A was a gift from the laboratory of Grace K. Pavlath (Emory University). Recombinant human PDGF-AB was purchased from Invitrogen (Carlsbad, CA). Media and supplements were purchased from Invitrogen, Mediatech (Herndon, VA), and Atlanta Biologicals (Norcross, GA). Other chemicals and reagents were purchased from Sigma-Aldrich (St. Louis, MO) and Fisher Scientific (Fair Lawn, NJ).

Cell Culture. A continuous line of rat vascular smooth muscle cells (VSMC) obtained from R. W. Alexander (Emory University) were maintained in DMEM with Pen-Strep (100 U/ml penicillin and 100 µM streptomycin) and 10% fetal bovine serum (FBS) in a 37°C humidified incubator with 5% CO₂. Phoenix retroviral producer cells (American Type Culture Collection, Manassas, VA) were also grown in DMEM supplemented with 10% FBS and Pen-Strep.

Plasmid Constructions. The retroviral expression vectors pTJ66 and pCL1 have been described previously (Murphy et al., 2002). The first set of constructs was created by removing the human Gq(Q209L), G14(Q205L), and G15(Q212L) inserts from GNA0Q0EIC0, GNA140EIC0, and GNA150EIC0 obtained from Guthrie cDNA Resource Center (Sayre, PA). Each construct contained an internal Glu-Glu (EE) epitope tag. Amino acids 171–176 within Gq corresponding to AYLPTQ were mutated to EYMPTE, and the corresponding amino acids within G14 and G15 were mutated similarly to generate the internal EE epitope tag. The Gq(Q209L), G14(Q205L), and G15(Q212L) inserts were removed by sequential digests with XhoI and then PmeI and inserted into pTJ66, which had been opened with BclI, blunted with Klenow, then cut again with XhoI to yield compatible sites. The second set of constructs was created by removing the Gq(Q209L) and G15(Q212L) inserts from GNA0Q0EIC0 and GNA150EIC0 with HindIII and PmeI digestion and by removing the G14(Q205L) insert from GNA140EIC0 with BglII, followed by a partial cut by HindIII. The recovered Gq(Q209L), G14(Q205L), and G15(Q212L) inserts were further subcloned into retroviral vector pCL1, which had been cut open with MluI, blunted with Klenow, and then cut with HindIII.

Retroviral Production and VSMC Infection. Protocols for the transient, helper-virus free production of nonreplicating recombinant retroviruses and VSMC infection have been described previously (Murphy et al., 2002), except that the Phoenix-Ampho producer cell line was used for the work described in this report. Infectious retroviral supernatants were prepared from Phoenix producer cells transfected with retroviral plasmids encoding Gq(Q/L), G14(Q/L), G15(Q/L), or empty vector alone for control samples and used fresh or stored at –80°C for future use.

LDH Measurements. VSMC in 24-well plates were infected with retroviral expression vectors containing Gq(Q/L), G14(Q/L), and G15(Q/L) or vector only. Immediately after completing the infection protocol, viral supernatant medium was replaced with phenol red-free DMEM supplemented with 10% FBS and Pen-Strep (500 µl/well). Appropriate concentrations of inhibitors or antagonists or vehicle were added to the VSMC, and 50-µl samples were removed from each well and transferred to 96-well plates and stored at –20°C. At 24-h intervals, additional 50-µl samples were removed and stored over a 5-day period. At each sampling, 50-µl aliquots with inhibitor or antagonist or vehicle were added to the wells to replenish the volume of medium. At the end of the experiments, the samples were measured using the CytoTox 96 nonradioactive cytotoxicity assay kit (Promega, Madison, WI) using a Molecular Devices Thermomax microplate reader (Molecular Devices Corp., Sunnyvale, CA).

[³H]Inositol Phosphate Production. VSMC were infected with retroviruses encoding Gq(Q/L), G14(Q/L), G15(Q/L), or empty virus as control, and then incubated overnight in DMEM/10% FBS and Pen-Strep with 4 µCi/ml [³H]inositol. Cells were placed in HEPES-buffered DMEM, pH 7.4, supplemented with 10 mM LiCl at 37°C, and [³H]inositol phosphates were isolated by ion exchange chromatography as described previously (Heximer et al., 1997).

Calcium Mobilization in VSMC. VSMC grown on coverslips were infected with Gq(Q/L), G14(Q/L), G15(Q/L), or empty virus only and then incubated with 5 µM Fura2-acetoxymethyl ester for 30 min at room temperature. Cells were then placed on a microscope stage for imaging of intracellular calcium concentration. External solution contained 150 mM NaCl, 10 mM HEPES, 3 mM KCl, 2 mM CaCl₂, 2 mM MgCl₂, 5.5 mM glucose, pH 7.3; osmolarity, 325 mOsM. Intensity images of 510 nm wavelength were taken at 340 nm and 380 nm excitation wavelengths, and the two resulting images were taken from individual cells for ratio calculations. Axon Imaging Workbench (version 2.2.1; Axon Instruments, Union City, CA) was used for acquisition of intensity images and conversion to ratios.

Detection of Caspase-3 Activity and Chromatin Condensation. Twenty-four hours after infection with the retroviral expression vectors, VSMC were washed twice with buffer (150 mM NaCl, 10 mM HEPES, 3 mM KCl, 2 mM CaCl₂, 2 mM MgCl₂, and 5.5 mM glucose, pH 7.3), then incubated with the fluorescent caspase-3 substrate, the bis-L-aspartic acid of rhodamine 110 (1.5 µg/ml), for 30 min. The VSMC were washed again with buffer and observed under an Olympus IX51 inverted microscope. Images were captured using Image-Pro Plus software (version 4.5.1; Media Cybernetics, Silver Spring, MD) from the microscope equipped with an Olympus Q Color 3 camera through a Chroma 41001 (fluorescein isothiocyanate/EGFP/BODIPY/Fluo 3/DiO) fluorescent filter set (Chroma Technology Corp., Brattleboro, VT). For images of chromatin condensation, the infected VSMC were washed twice with buffer, then incubated with Hoechst 33342 dye (10 µg/ml) for 15 min. The VSMC were washed again with buffer, and images were captured as described above through a Chroma 31000 (4,6-diamidino-2-phenylindole/Hoechst/7-amino-4-methylcoumarin-3-acetic acid) filter set.

Transcription Factor Luciferase Reporter. The retroviral NFAT-specific luciferase reporter vector has been described previously (Murphy et al., 2002). VSMCs stably expressing the NFAT luciferase reporter were grown in 24-well plates and infected with one round of Gq(Q/L), G14(Q/L), G15(Q/L), or empty vector alone for control samples, as described above. Twenty-four hours after infection, VSMC were serum-starved overnight. VSMC were lysed, and luciferase was activity monitored on a Turner Designs luminometer (Sunnyvale, CA) using a luciferase assay system kit (Promega).

SDS-Polyacrylamide Gel Electrophoresis and Immunoblot Analysis. Whole-cell extracts of VSMC were prepared by lysing the cells in a buffer (50 mM Tris-HCl, 50 mM NaCl, 5 mM EDTA, 10 mM EGTA, 1 mM Na₃VO₄, 2 mM Na₄P₂O₇·10H₂O, 10 µg/ml leupeptin, 2 µg/ml aprotinin, and 1% Triton X-100) collecting and adding Laemmli sample buffer, sonicating briefly, and boiling for 5 min. Samples were loaded and resolved on SDS-polyacrylamide gels and transferred to nitrocellulose. For detection of the Glu-Glu (EE) epitope tag, the membranes were blocked for 1 h at room temperature with 5% nonfat milk in TBST and probed with the monoclonal anti-Glu-Glu (EE) antibody (1:2000 in 5% non-fat milk/TBST) overnight at 4°C. Membranes were washed with TBST and incubated for 1 h in horseradish peroxidase-conjugated goat anti-mouse IgG (1: 20,000 in TBST). Membranes were again washed in TBST, and immunoreactive proteins were detected by chemiluminescence. All other immunoblots were performed using an identical protocol except for the indicated specific antisera. For detection of caspase-3, membranes were probed with the polyclonal caspase-3 antibody (1: 1000 in 5% nonfat milk-TBST) followed by horseradish peroxidase-conjugated goat anti-rabbit IgG (1:10,000 in 5% nonfat milk-TBST). For detection of Akt, membranes were probed with the polyclonal Akt antibody (1:1000 in 5% nonfat milk-TBST) followed by horseradish peroxidase-conjugated goat anti-rabbit IgG (1:10,000 in 5% nonfat milk-TBST). For detection of phospho-Akt, membranes were probed with the polyclonal phospho-Akt antibody (1:1000 in TBST) followed by horseradish peroxidase-conjugated goat anti-rabbit IgG (1:10,000 in 5% nonfat milk-TBST).

Statistical Treatment of Data. Significant differences among treated and control samples were determined using one-way analysis of variance with Tukey's and Dunnett's post tests. Significant differences between conditions (p < 0.05) are indicated in each figure and legend, where appropriate.

Total RNA Isolation and Gene Microarray Studies. VSM cells (grown to confluence in six-well, 35-mm dishes) were infected with empty retrovirus (control) or retroviruses encoding Gq(Q/L), G14(Q/L), or G15(Q/L) (experimental). After 24 h, total RNA was extracted from cells in some wells using TRIzol according to the manufacturer's protocol (Invitrogen). Remaining sister cells continued to grow and were either observed for an additional 48 h (72 h total) to confirm cell death or were harvested and immunoblotted (anti-EE) to confirm expression of G-EE. The quality and amount of RNA were confirmed by ethidium bromide staining of an agarose gel and by spectrophotometry. Total RNA samples from three separate identical experiments were then sent to the UCLA/NHLBI Shared Microarray Facility for processing and analysis. The quality of RNA samples was again assessed at the Microarray Facility to determine suitability for analysis. Samples (cDNA) were generated and hybridized with a GeneChip array (Affymetrix High Density Rat Genome 230-2.0) containing 31,042 oligonucleotides representative of the entire rat genome. For each gene sequence, 11 different complimentary oligonucleotide pairs were present on the array to normalize differences in transcriptional levels (as detected by intensity values) of each gene (http://www.affymetrix.com). Microarrays were processed using the Affymetrix recommended protocol for hybridization, washing, and staining, and results were quantitated using the GeneChip Scanner 3000 High Resolution Scanner and GeneChip Operating Software from Affymetrix. Resulting microarray data sets (CEL files) were analyzed at the Emory Biomolecular and Computing Resource using the R-Bioconducting package (http://www.bioconductor.org) analysis for Affymetrix arrays (Gautier et al., 2004). Data sets were loaded into the R-Bioconductor package and resulting RMA (Robust Multichip Analysis) values (Bolstad et al., 2003; Irizarry et al., 2003) were generated using the following parameters: RMA background correction method, quantile normalization method, PM (perfect match) only values, and the median polish summary method for signal calculation (Irizarry et al., 2003). RMA values were imported into GeneSpring v6.2 (Silicon Genetics, Redwood City, CA) and expression values for each gene were normalized across chips to the median value of each gene. In each experiment, -fold change values (experimental versus control) were determined, and a cutoff value of 2.0 was arbitrarily applied to ascertain genes that were differentially expressed across experiments and between the conditions within each experiment (Gq, G14, or G15 versus control). Gene ID, annotation (where known), -fold change for each condition, average fold change, and range were reported for each condition. Lists of affected genes were sorted by an increase or decrease in fold change (versus control) for each condition. Gene lists generated were compared to determine which gene(s) overlapped in each replicate experiment, and instances in which genes changed less than 2-fold in one or more experiments are not reported. Venn diagrams were used to sort differential expression in one or multiple experimental conditions (overlapping and nonoverlapping).

	Results

Top Abstract Materials and Methods Results Discussion References

 
Gq, G14, and G15 Have Differential Effects on Vascular Smooth Muscle Cell Survival. To study signaling characteristics of the Gq family of G proteins, we used a retroviral infection method to introduce constitutively active (GTPase-deficient, Q/L) mutants of Gq, G14, and G15 into rat vascular smooth muscle cells (VSMC). Differences in cell morphology and viability were evident within 24 h after infection in cells expressing Gq(Q/L) and G14(Q/L), but not those expressing G15(Q/L) or control retroviral vector only (Fig. 1, top). These changes in morphology became more profound over time, and nearly all Gq- and G14-infected cells were dead after 120 h. Cells infected with wild-type (not constitutively active) Gq, G14, and G15 grew normally and were not different from control, and C2C12 mouse myoblasts infected with Gq(Q/L), G14(Q/L), or G15(Q/L) did not die (data not shown). We measured cytosolic LDH released into the culture medium as a method for the quantification of cell death (Fig. 1, bottom). At 72 h, increases in LDH levels were 2- to 3-fold greater for cells infected with Gq(Q/L) and G14(Q/L) than for cells infected with G15(Q/L) or control retroviral vector only. We propose that distinct and divergent signaling pathways from Gq, G14, and G15 mediate the selective cell death seen in VSMC.

View larger version (62K): [in this window] [in a new window]   Fig. 1. Constitutively active Gq and G14, but not G15, cause cell death. Rat VSMCs infected with retroviruses to express constitutively active mutants of Gq (Q209L), G14(Q205L), and G15(Q212L), or empty virus only were observed after infection, and samples of growth medium were taken at 24-h intervals for measurement of released cytosolic LDH as an indicator of cell death. Top, VSMCs expressing Gq(Q/L) and G14(Q/L) exhibit morphology characteristic of cell death at 24 h after infection. Bottom, LDH activity increases more in cells infected with Gq(Q/L) and G14(Q/L) than in those infected with G15(Q/L) or vector. Each point represents the mean ± S.E.M. of three experiments performed in duplicate.

Gq(Q/L), G14(Q/L), and G15(Q/L) Stimulate Inositol Lipid Signaling to Similar Extents. Gq family members link cell surface receptors to phosphoinositide hydrolysis and are reported to activate phospholipase C (PLC) equally in cell-free recombinant systems (Hepler et al., 1993) and intact cells (Lee et al., 1992). We measured the capacity of Gq(Q/L), G14(Q/L), and G15(Q/L) to stimulate phosphoinositide hydrolysis in rat VSMC by the accumulation of radioactively labeled inositol phosphates in the presence of lithium chloride. Cells infected 24 h earlier with Gq(Q/L), G14(Q/L), or G15(Q/L) produced similar levels of PI hydrolysis (Fig. 2, top). We confirmed the expression of each mutant by immunoblotting with an antibody to detect an internal glutamate-glutamate (EE) tag introduced into Gq(Q/L), G14(Q/L), and G15(Q/L) constructs. Equal samples of lysates from cells infected at the same time as those used for the PI hydrolysis experiments showed similar levels of expression of each G protein (Fig. 2, top).

View larger version (21K): [in this window] [in a new window]   Fig. 2. Constitutively active Gq, G14, and G15 stimulate phosphoinositide hydrolysis in rat vascular smooth muscle cells. Rat VSMCs infected with retroviruses encoding Gq(Q209L), G14(Q205L), G15(Q212L), or empty virus only were labeled overnight with [3H]inositol, and the accumulation of [3H]inositol phosphates in the presence of lithium chloride was measured 24 h after infection. Top, when expressed in VSMC, Gq(Q/L), G14(Q/L), and G15(Q/L) stimulated phosphoinositide hydrolysis at similar levels over 30 min in the presence of lithium chloride. An immunoblot shows the expression of each G using an antibody to detect the EE-epitope tag present in each G protein. The graph represents the mean ± S.E.M. of three experiments performed in triplicate, expressed as the fold increase of basal (vector only) activity. Bottom, Gq(Q/L), G14(Q/L), and G15(Q/L) exhibit similar kinetics for the stimulation of phosphoinositide hydrolysis. The figure shows a representative time course experiment performed in duplicate.

Gq(Q/L) and G14(Q/L) Activate Caspase-3. Cell death may result from general cytotoxicity (necrosis), activation of programmed cell death (apoptosis) pathways, or inhibition of cell survival pathways. To test whether cell death in the infected VSMC was apoptotic, we used a fluorogenic substrate (N-benzyloxycarbonyl-DEVD-Rhodamine 110; Molecular Probes) containing the recognition site for caspase-3 (DEVD), a central component of apoptotic cell death pathways. Cleavage of the substrate by activated caspase-3 produces a fluorescent product (Molecular Probes). Twenty-four hours after infection with Gq(Q/L), G14(Q/L), G15(Q/L), or vector only, VSMCs were incubated with the Rhodamine 110 substrate and observed by fluorescent microscopy. Cells infected with Gq(Q/L) and G14(Q/L) showed extensive fluorescence, indicative of caspase-3 activation. Cells infected with G15(Q/L) or vector only showed little or no fluorescence (Fig. 3, middle). We also incubated infected VSMCs with the nuclear stain Hoechst 33342 (Molecular Probes) to detect nuclear chromatin condensation as another marker for apoptotic cell death. Cells infected with Gq(Q/L) and G14(Q/L) exhibited condensed nuclei, but cells infected with G15(Q/L), or vector alone did not (Fig. 3, bottom). As an additional measure of caspase-3 activation, samples of lysates from infected cells were subjected to immunoblotting with a caspase-3 antibody (Fig. 4). Cells infected with Gq(Q/L) and G14(Q/L) showed the appearance of a distinct band at 16 kDa corresponding to the large fragment of caspase-3, generated by proteolytic cleavage of the full-length precursor (Nicholson et al., 1995). Cells infected with G15(Q/L) or vector only did not show the appearance of the fragment (Fig. 4). Based on these results, we conclude that Gq(Q/L) and G14(Q/L), but not G15(Q/L), mediate cell death through an apoptotic pathway involving activation of caspase-3.

View larger version (72K): [in this window] [in a new window]   Fig. 3. Constitutively active Gq and G14, but not G15, increase caspase-3 activity in intact VSM cells. Caspase-3 activity in rat VSMC infected with Gq(Q209L), G14(Q205L), G15(Q212L), or empty virus only was observed 24 h after infection. Gq(Q/L)- and G14(Q/L)-infected cells displayed morphology indicative of cell death (top). The cell-permeable fluorescent substrate for caspase-3, Rhodamine-110 (Molecular Probes) shows elevated activity in VSMC infected with Gq(Q/L) and G14(Q/L), but not G15(Q/L) or vector only (middle). The cell-permeable fluorescent nuclear stain, Hoechst 33342 (Molecular Probes), shows chromatin condensation in nuclei (arrows) of VSMC infected with Gq(Q/L) and G14(Q/L), but not G15(Q/L), or empty virus only (bottom).

View larger version (52K): [in this window] [in a new window]   Fig. 4. Constitutively active Gq and G14, but not G15, increase caspase-3 activity in intact VSM cells. Caspase-3 activity in rat VSMC infected with Gq(Q209L), G14(Q205L), G15(Q212L), and empty virus only, was observed 24 h after infection. Immunoblot analysis of cell lysates prepared 24 h after infection from VSMC. Cells infected with Gq(Q/L) and G14(Q/L) show cleavage of caspase-3 (appearance of 16.5 kDa fragment) indicative of elevated activity.

View larger version (29K): [in this window] [in a new window]   Fig. 5. Akt phosphorylation is unchanged by constitutively activated Gq, G14, or G15 in rat vascular smooth muscle cells. Rat VSMC, infected with retroviruses encoding Gq(Q209L), G14(Q205L), G15(Q212L), or empty vector were lysed at 12 and 24 h after infection, treated for 30 min with PDGF or vehicle, then prepared for immunoblotting with a phospho-specific Akt antibody. No changes were observed in basal or stimulated levels of Akt phosphorylation between VSMC infected with Gq(Q/L), G14Q/L) or G15(Q/L) and vector only.

View larger version (46K): [in this window] [in a new window]   Fig. 6. Blocking protein kinase C and inositol trisphosphate receptors does not inhibit Gq(Q/L)- and G14(Q/L)-induced cell death. Rat VSMC, infected with Gq(Q209L), G14 (Q205L), and G15 (Q212L), or vector only were incubated in growth medium containing the PKC inhibitor BIS, the IP3 receptor antagonist 2-APB, or vehicle. Samples of growth medium were taken at 24-h intervals for measurement of released cytosolic LDH. At 72 h, no differences in LDH activity were detected between treated and untreated conditions. The graph represents the mean ± S.E.M. of three experiments performed in duplicate.

Inhibition of Gq(Q/L) and G14(Q/L) Activation of NFAT Does not Prevent VSMC Death. In considering possible downstream signals mediating cell death through alteration of gene expression, we examined changes in activity of transcription factors often associated with stimulation of G protein-coupled receptors linked to inositol lipid and Ca²⁺/PKC pathways. The transcription factor NFAT has been reported to be a component of Gq-mediated antiapoptotic pathways in cardiomyocytes (Pu et al., 2003). NFAT activation is initiated through calcium release and activation of calmodulin and subsequent activation of calcineurin. Calcineurin binding to NFAT dephosphorylates and causes activated NFAT to translocate to the cell nucleus, where it begins transcription of target genes. We used a VSMC line stably infected with an NFAT-specific luciferase reporter to determine whether differences existed in the ability of Gq(Q/L), G14(Q/L), and G15(Q/L) to stimulate NFAT (Boss et al., 1998). We infected the NFAT-VSMC with Gq(Q/L), G14(Q/L), G15(Q/L), or vector only. After 24 h, cells were serum-starved, and luciferase activity was measured 24 h later. Gq(Q/L) and G14(Q/L) increased NFAT-specific luciferase activity significantly more than vector only, but G15(Q/L) did not increase NFAT activity significantly differently from vector only (Fig. 7, top). We also tested whether cyclosporin A (CsA), a calcineurin inhibitor, blocked NFAT activation in these cells. Supplementing the growth medium of VSMC with CsA did not alter the cell death induced by infection with Gq(Q/L) or G14(Q/L). The capacity of Gq(Q/L) and G14(Q/L) to activate NFAT is greater than G15(Q/L) in VSMC, but activation of NFAT does not seem to alter the outcome of cell death (Fig. 7, bottom).

View larger version (17K): [in this window] [in a new window]   Fig. 7. Inhibition of NFAT stimulated by Gq(Q/L) and G14(Q/L) in rat vascular smooth muscle cells does not prevent G-induced cell death VSMC lines stably expressing an NFAT-specific luciferase reporter were infected with Gq(Q209L), G14(Q205L), G15(Q212L), or empty virus only. After 24 h, infected cells were serum-starved for an additional 24 h (48 h after infection) and assayed for NFAT activity. Top, VSMC infected with Gq(Q/L) and G14(Q/L), showed elevated NFAT activity significantly different from those infected with vector only. VSMC infected with G15(Q/L) was not significantly different from those infected with empty virus only. The graph represents the mean ± S.E.M. of six experiments performed in quadruplicate. (*, p < 0.05). Bottom, rat VSMC infected with retroviruses encoding Gq(Q/L), G14(Q/L), G15(Q/L), or empty virus only were incubated in growth medium containing the calcineurin inhibitor CsA or vehicle. Samples of growth medium were taken at 24-h intervals for measurement of released cytosolic LDH. At 72 h, no differences in LDH activity were detected between treated and untreated conditions. The graph represents the mean ± S.E.M. of three experiments performed in duplicate.

Results of these cDNA microarray analysis are summarized in Fig. 8 and Tables 1, 2, 3, 4, 5. Of 31,042 genes examined, a total of 422 mRNAs changed (2-fold or greater increase or decrease) in each experiment with VSMC expressing Gq(Q/L) after 24 h. In parallel samples, 341 mRNAs and 74 mRNAs changed in cells expressing G14(Q/L) or G15(Q/L). Figure 8 illustrates expressed gene sets that changed and were either shared (overlapping) or distinct (nonoverlapping) among the three G. All three G shared 50 genes that changed (23 increase, 27 decrease), and those genes with known identities and their reported functions are listed in Table 1. The mRNAs that changed most robustly and were shared by all three G are cyclooxygenase 2 and a muscle cytoskeletal protein (Krp1/sarcosin); mRNA for the interleukin 1 receptor antagonist (Il1rn) decreased markedly. We were surprised to find that each G also stimulated changes in distinct gene sets. Gq(Q/L) stimulated changes in 150 unique mRNAs (74 increased, 76 decreased) that did not change with G14(Q/L) or G15(Q/L). Of these, only 37 are defined and are listed with their reported functions in Table 2 [the remainder are undefined expressed sequence tags (ESTs)]. The mRNA most robustly expressed (22-fold) in cells expressing Gq(Q/L) encodes Gq, although we cannot rule out the possibility that this reflects cross-reactivity between the probes and retrovirally expressed recombinant Gq. G14(Q/L) stimulated changes in 70 mRNAs (45 increased and 25 decreased) not changed by Gq(Q/L) or G15(Q/L). G15(Q/L) stimulated changes in 18 genes (16 increased and 2 decreased) not changed by Gq(Q/L) or G14(Q/L). Identified gene sets that changed uniquely in response to either G14(Q/L) or G15(Q/L) and their reported functions are listed in Table 3.

View larger version (22K): [in this window] [in a new window]   Fig. 8. Gene expression profiles of VSMC after 24 h infection with Gq(Q/L), G14(Q/L), or G15(Q/L). VSMC were infected with Gq(Q209L), G14(Q205L), G15(Q212L), or virus carrying empty vector (control). After 24 h, cells were harvested and total RNA was recovered, or sister cells were cultured for an additional 72 h and observed. Total RNA derived from cultures in which sister samples were observed to both express G (by EE-antibody) and to progress to cell death after 72 h [in the case of Gq(Q/L) and G14(Q/L)] were submitted for DNA micrroarray analysis (n = 3 for each condition) and data processing as described under Materials and Methods. A, Venn diagrams showing the number of genes that increased 2-fold or more over control for each G in each experiment. B, Venn diagram showing the number of genes that decreased 2-fold or more over control for each G in each experiment.

View this table: [in this window] [in a new window]   TABLE 1 Defined genes up- or down-regulated by Gq, G14 and G15 in VSMC

View this table: [in this window] [in a new window]   TABLE 2 Genes that are up- or down-regulated by Gq, but not G14 or G15

View this table: [in this window] [in a new window]   TABLE 3 Genes that are up- or down-regulated by either G14 or G15, but not Gqa

View this table: [in this window] [in a new window]   TABLE 4 Defined genes up-regulated by Gq and G14, but not by G15, in VSMC Bold type indicates genes with reported roles in apoptosis and/or TNF signaling (see text).

View this table: [in this window] [in a new window]   TABLE 5 Defined genes down-regulated by Gq and G14, but not by G15, in VSMC Bold indicates genes with reported roles in apoptosis and/or TNF signaling (see text). Listed are defined annotated genes (11) of 85 total that changed.

Gq(Q/L) and G14(Q/L) Stimulate Changes in a Shared Set of Genes, Some with Established Roles in Apoptosis and TNF- Signaling. Most germane to the present study, 221 mRNAs changed (137 increased, 84 decreased) in response to Gq(Q/L) and G14(Q/L), but not G15(Q/L) (Fig. 8, Tables 4 and 5). When comparing the samples using a three-way cluster analysis, changes in gene sets were most similar for Gq(Q/L) and G14(Q/L) and more divergent for G15(Q/L) (data not shown). Of the mRNAs that changed, only a subset are defined with annotations (38 that increased, 11 that decreased); the rest are undefined ESTs. It is noteworthy that several of these gene products are reported to be pro-apoptotic in various cell lines and are presented in bold type in Table 4. The individual mRNA that increased most robustly (12-fold) in response to Gq(Q/L) and G14(Q/L) was TNF- stimulated gene 6 (TSG-6), which (as its name implies) is strongly up-regulated by TNF- and binds to the extracellular matrix sugar hyaluronan. Other mRNAs reported to be strongly up-regulated by TNF- that are also up-regulated by Gq(Q/L) and G14(Q/L) include the enzyme that produces hyaluronan, hyaluronan synthase (Has2); the multidrug resistance gene (Mdr1a), which is induced during cell stress; and the Urokinase/Plasminogen receptor (uPAR-1). Gene products that are robustly up-regulated and pro-apoptotic and also either stimulate TNF- synthesis or mediate TNF- signaling include activating transcription factor 3 (Atf3/LRF1), a cAMP response element-binding protein-like nuclear transcription factor, the TGF--like cytokine bone morphogenic protein 2 (BMP2), mitogen-activated protein kinase kinase kinase 8 (Map3K8;Tpl2;Cot), the nuclear transcription factor NGFI-B/Nurr77, the oxidized LDL-receptor 1 (LOX-1), and connexin 37. The mRNA most robustly down-regulated by Gq(Q/L) and G14(Q/L) was the receptor for angiotensin II (AT1A-R) (Table 5).

TNF- Alone or in Combination with Calcium Ionophore Does Not Stimulate VSMC Death. Because Gq(Q/L) and G14(Q/L) stimulated an increase in numerous mRNAs reported to be pro-apoptotic and/or involved with TNF- signaling, we tested whether TNF-, either alone or in the presence of increased intracellular calcium, stimulated VSMC death. Cells treated with increasing concentrations of TNF- alone (0–100 ng/ml) exhibited no cell death, even at concentrations (30–100 ng/ml) reported to stimulate cell death in other sensitive cells (data not shown). In separate experiments, we tested whether TNF- acted synergistically with calcium to stimulate VSM cell death. As expected, ionomycin alone was cytotoxic at high concentrations (0.1–1 µM), whereas intermediate and low concentrations of ionomycin were not (data not shown). Addition of high concentrations (100 ng/ml) of TNF- did not alter the potency of ionomycin-induced cell death (data not shown), suggesting that TNF- does not act synergistically with ionomycin to shift the sensitivity of VSMC to cell death at low or modest intracellular [Ca²⁺].

	Discussion

Top Abstract Materials and Methods Results Discussion References

 
Gq family exhibit striking differences in sequence homology and tissue distribution that predicts unappreciated differences in cellular functions. Consistent with this idea, we found that Gq, G14, and G15 exert different effects on VSMC survival and gene expression patterns. Constitutively activated forms of Gq and G14, but not G15 or control, stimulate caspase-3 activation and apoptotic cell death when expressed in VSMC. These effects are specific to the cell type involved and the activation state of G, because VSMCs expressing inactive G or myoblasts expressing active Gq, G14, and G15 do not undergo cell death. Each G elicits a markedly different profile of altered gene sets in VSMC. Whereas all three G stimulated changes in a shared set of mRNA, Gq and G14 (but not G15) stimulated changes in a much larger shared set of mRNA, several of which are reported to be pro-apoptotic and involved with TNF- signaling. We were surprised to find that each G also stimulated changes in nonoverlapping G-specific gene sets. These findings demonstrate that Gq family members are functionally diverse and activate both overlapping and distinct signaling pathways.

Established models indicate that Gq family members exert their cellular actions by activating PLC. We found that constitutively active Gq, G14, and G15 stimulate accumulation of inositol phosphates equally in both the kinetics and magnitude of response. PLC activity is sustained until cell death is complete with Gq(Q/L) and G14(Q/L) and for viable cells expressing G15(Q/L) over the same period (data not shown). Despite the accumulation of inositol phosphates, intracellular calcium levels were suppressed and no different for any of the G-expressing or control cells after 24 h, suggesting possible depletion of internal calcium stores. Consistent with this idea, we found that activation of endogenous Gq/11-coupled purinergic receptors in G-expressing cells resulted in markedly reduced calcium responses (data not shown). Similar desensitization of calcium signaling was reported in other cells (SLCC, Swiss-3T3, and different VSMC) after G16(Q/L) stimulated phosphoinositide hydrolysis (Qian et al., 1994; Heasley et al., 1996b; Lobaugh et al., 1996; Higashita et al., 1997). Even though the three G stimulated inositol triphosphate/Ca²⁺ similarly, their effects on VSMC survival differed, suggesting that Gq- and G14-initiated cell death is not mediated by components of the PKC/IP₃-Ca²⁺ signal pathway alone. In support of this, inhibition of PKC activity, IP₃ binding to its receptor, or Ca²⁺ stimulated activation of calcineurin/NFAT failed to block G-induced apoptosis.

Our report adds to others showing differences among the Gq family members in their binding partners and cellular responses. For example, G15 and G16 are "promiscuous" in coupling to various GPCR compared with the selective Gq and G11 (Offermanns and Simon, 1995), and G16 does not interact with GRK2 like other Gq family members (Day et al., 2003). Many reports demonstrate that Gq family members activate distinct downstream signaling events leading to different cell fates. Constitutively active G16 is a more potent stimulator of nuclear factor B than Gq in HeLa cells (Yang et al., 2001), and inhibits growth of SCL carcinoma cells and Swiss 3T3 cells (Qian et al., 1994; Heasley et al., 1996b). Constitutively active Gq and G16 induce cell differentiation and activate JNKs in PC-12 cells (Heasley et al., 1996a). Gq family members also have been linked to activation of caspases and apoptotic pathways in various cell types. In COS-7 and CHO cells, constitutively active Gq causes apoptosis (Althoefer et al., 1997). In cardiomyocytes, constitutively active Gq causes apoptosis through cytochrome c release from mitochondria and caspase activation (Adams et al., 2000). This Gq-induced apoptosis is mediated through inhibition of the PI3 kinase/Akt cell survival pathway characterized by depletion of phosphatidylinositol 4,5-bisphosphate availability, generation of phosphatidylinositol 3,4,5-trisphosphate and dephosphorylation of Akt (Howes et al., 2003). In contrast, we find no differences in resting or stimulated Akt phosphorylation in VSMC expressing constitutively active G subunits. However, in Rat-1 fibroblasts, active Gq inhibits PI3 kinase activity upstream from Akt, suggesting an interaction between Gq and PI3 kinase (Ballou et al., 2003). A subsequent report demonstrated that constitutively active Gq stimulates GSK-3 in HEK cells through inhibition of PI3 kinase and activation of Csk tyrosine kinase but independent of PLC and apparently not involving Akt (Fan et al., 2003). Activation of Gq family members also results in cell fates other than death. Constitutively active Gq stimulates cell transformation of NIH-3T3 cells and differentiation of PC-12 cells (De Vivo et al., 1992; Heasley et al., 1996a). We find that myoblasts expressing Gq(Q/L), G14(Q/L), and G15(Q/L) do not undergo cell death. Taken together with our experiments comparing G14, these findings suggest that Gq family members recruit overlapping and distinct signaling pathways leading to distinct cell-specific effects on cell fate.

To examine differences in cellular responses to Gq family members, we analyzed expressed gene sets in VSMC. Although all 31,042 genes of the rat genome were screened, less than half of the genes that changed in response to G are defined (the rest are undefined ESTs), which provides only a partial picture of the expression profiles. Nevertheless, important (albeit incomplete) information was obtained. During the first 24 h, Gq, G14, and G15 each elicit changes in both overlapping (shared) and distinct (i.e., G-specific) gene sets but exhibit surprisingly different gene profiles. The overlapping sets of genes that changed included those shared by all three G and by two of the three G. Of the latter, Gq and G14 stimulated changes in a large set of shared genes (221), whereas G15 exhibited little or no overlap with either Gq (one shared) or G14 (five shared). This suggests that Gq and G14 are functionally related and is consistent with our findings that Gq and G14, but not G15, stimulated VSMC death. Most of the shared genes robustly up-regulated (5-fold or more) by both Gq and G14 have reported roles in apoptosis and/or TNF signaling. In particular, transcription factors Atf3/LRF1 and NGFI-B/Nurr77, the cytokine bone morphogenic protein 2 (Bmp2), the Ser/Thr kinase Map3K8, and the membrane channel protein connexin 37 are each reported to be pro-apoptotic (Patriotis et al., 2001; Watanabe et al., 2001; Zhang et al., 2001; Hartman et al., 2004; Kume and Kita, 2004; Seul et al., 2004) and up-regulated by TNF or to mediate its actions (van Rijen et al., 1998; Dumitru et al., 2000; Mashima et al., 2001; Gruber et al., 2003; Hofnagel et al., 2004; Inoue et al., 2004). Other robustly up-regulated genes, including the hyaluronan binding protein; TSG-6; the enzyme that makes hyaluronan, hyaluronan synthase 2 (Has2); the urokinase/tPA receptor (uPAR-1); and the p-glycoprotein multidrug resistance gene (Mdr1a), although not directly linked to apoptosis, are reported to either mediate TNF- actions or to be up-regulated by TNF- (Lee et al., 1993; Piguet et al., 1999; Ijuin et al., 2001; Theron et al., 2003). A shared link between many of the remaining genes and apoptotic pathways or TNF- signaling is not readily apparent.

Our findings suggest that the downstream actions of Gq and G14 may converge on signaling pathways used by TNF- or that they may stimulate VSMC to synthesize or release TNF- to exert autocrine actions. Indeed, TNF- has been shown to stimulate apoptosis in other VSMC by direct and autocrine mechanisms (Boyle et al., 2003). However, we found that activation of TNF- receptors alone with high concentration of cytokine was not sufficient to stimulate death of our aortic smooth muscle cells, nor did TNF- act synergistically with ionophore to sensitize VSM cell death to lower levels of intracellular calcium (data not shown). Altho