Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

A family of heterotrimeric G proteins transduces chemical and sensory signals across the plasma membrane by sequential interactions with receptor and effectors such as second messenger-generating enzymes or ion channels. Because of the wide array of cellular processes that are mediated by G proteins, the study of G protein function and regulation has provided a cornerstone for research in signal transduction.

Recently, numerous studies have suggested an important function for G proteins at cellular locations other than the plasma membrane. Certain G proteins have been detected at intracellular membranes, such as the Golgi complex (Kehlenbach et al., 1994; Denker et al., 1996; Ugur and Jones, 2000) or light vesicle fractions (Drmota et al., 1999), whereas others associate with cytoskeletal structures, such as microtubules and microfilaments (Wilson et al., 1994; Ibarrondo and Gullin, 1995; Roychowdhury and Rasenick, 1997; Roychowdhury et al., 1999). The mechanisms that govern the cellular destinations of G protein and the relative proportions of the G protein that traffic to subcellular compartment remain unresolved.

Previous studies, using cell fractionation and immunocytochemical assays, suggested that activation by  adrenergic agonist, cholera toxin or direct binding of a hydrolysis-resistant GTP analog resulted in G_s translocation from plasma membrane to cytoplasm (Rasenick et al., 1984; Ransas et al., 1989; Levis and Bourne, 1992). G_q/11 has also been reported to transiently translocate to plasma membrane of adrenal glomerulosa cells in response to stimulation by angiotensin II for 1 to 5 min (Cote et al., 1997). However, in other studies, a relocation of activated G_s was not observed (Jones et al., 1997; Huang et al., 1999). The mechanism by which G_s is released from the membrane is not yet known, but there are suggestions that activated G_s is depalmitoylated and released from the membrane (Wedegaertner and Bourne, 1994). This continues to be a subject of some controversy, however, because at least one report demonstrates that G_s remains associated with the plasma membrane, regardless of palmitoylation state (Huang et al., 1999).

The use of GFP in the study of cellular signaling allows not only the observation of G protein trafficking, but it also provides the opportunity to study the functional dynamics of G proteins in real time. At present, most GFP fusion proteins are constructed by fusing GFP to either the amino or carboxyl terminus of the protein of interest (Barak et al., 1997; Hack et al., 2000; Kallal and Benovic, 2000). For G protein  subunits, this is problematic, because the NH₂ region is important for association with G protein  and  subunits and the COOH terminal is required for interaction with receptor. Recently, functional G-GFP fusion proteins were obtained by inserting GFP into an internal loop of G_q and of G₂ from Dictyostelium discoideum (Hughes et al., 2001; Janetopoulos et al., 2001). Thus, a biologically active G_s-GFP fusion protein might need to incorporate GFP at some interior position in G_s.

This report demonstrates that incorporation of GFP into the internal sequence of G_s results in a biologically active protein that seems to function identically to native G_s in the binding of GTP and activation of adenylyl cyclase. This G_s-GFP displays a heterogeneous distribution on the plasma membrane of the cells in which it is expressed and translocation from plasma membrane to cytoplasm after activation. This study raises the possibility that G_s might interact with effector molecules and cytoskeletal elements at multiple cellular sites.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Construction of G_s-GFP Fusion Protein Expression Vectors. Full-length cDNAs encoding G_s were excised from the pcDNA1 vector by digesting with SamI and XbaI restriction enzymes. The full-length EGFP cDNA obtained by PCR from the pEGFP-N2 using appropriate primers (sense, 5'-GGAATTCATGGTGAGCAAGGGCGAGGAACTG-3'; antisense, 5'-GCTCTAGACGACTTGTACAGCTCGT-3') and adding restriction sites to its cDNA (EcoRI at initiation codon and XbaI at end of cDNA). To insert the EGFP within the sequence of G_s, the first fragment of G_s cDNA (from 1 to 71 amino acids) was amplified by PCR with restriction sites for KpnI at initiation codon and EcoRI at end of the fragment and cloned into pcDNA3 vector. Modified EGFP cDNA was ligated to the first fragment of G_s by cloning it into pcDNA3 vector at EcoRI and XbaI restriction sites. The second fragment of G_s cDNA (from 82 to 394 amino acids) was also obtained using PCR with appropriate primers. The sense primer contained sequence overlapping with the 3' end of EGFP (5'-GAGCTGTACAAGTCGTCTAGAAACAGCGATGGTGAGAA-3'). The anti-sense primer contained an additional XbaI restriction site (5'-CGCTCTAGAGAACATCTAAGCAAG-3'). The second fragment of G_s cDNA was recombined to 3' end of EGFP using PCR strategy. Finally, the full-length G_s-GFP was cloned into pcDNA3 at KpnI and XbaI restriction sites. All DNA manipulations, including ligations, bacterial transformation, and plasmid purification, were carried out using standard procedures. To produce Gs-NGFP and G_s-CGFP, full-length G_s cDNA was subcloned into the pEGFP-C3 vector and pEGFP-N2 vector, respectively. Platinum Taq DNA polymerase (Invitrogen, Carlsbad, CA) was used in PCR. All PCR fragments were identified by DNA sequencing.

Adherent Cell Culture and Transient Transfection. COS-1 cells and HEK 293 cells in DMEM (Invitrogen) containing 10% fetal bovine serum, 1% antibiotic (penicillin and streptomycin), and PC12 cells in DMEM medium containing 5% fetal bovine serum, 5% horse serum, and antibiotic were cultured at 37°C with 5% CO₂. The cells in the 12-well culture plates or 60-mm tissue culture dishes were transfected using GenePORTER Transfection reagent (Gene Therapy Systems, Inc. San Diego, CA) according to the manufacturer's instruction. Briefly, 60 to 80% confluent cells washed once with Opti-MEM I reduced serum medium (Invitrogen). In each well of 12-well culture plates, 0.5 ml of mixture of plasmid DNA (5 µg) and GenePORTER Transfection reagent (10 µl) in Opti-MEM I medium was applied. In 60-mm tissue culture dish, 2.5 ml of mixture of plasmid DNA (15 µg) and GenePORTER Transfection reagent (50 µl) in Opti-MEM I medium was added. Six hours after incubating, cells were cultivated in complete medium as described above. All experiments were performed in transiently transfected cells.

Microscopy. Cells were observed 24 h after transfection using fluorescence microscopy or confocal microscopy. Before observation, the medium in 12-well culture dishes was changed to serum-free DMEM containing 20 mM HEPES and placed immediately on the microscope stage. Cells were maintained at 37°C during the entire period of observation.

Subcellular Distribution of G_s-GFP Fusion Protein in COS-1 Cells. COS-1 cells were harvested and resuspended in HEPES-sucrose buffer (15 mM HEPES, 0.5 mM sucrose, 1 mM DTT) containing Complete protease inhibitor tablets (Roche Molecular Biochemicals, Indianapolis, IN). Cell suspensions were homogenized with a Teflon/glass homogenizer. After a low-speed centrifugation to remove unbroken cells and a nuclear pellet, samples were centrifuged for 15 min at 200,000g at 4°C in a centrifuge (TL100; Beckman Coulter, Fullerton, CA). The pellets (particulate fractionation) and the supernatants (soluble fractionation) were separated by 12% SDS-PAGE and transferred to Immobilon-P transfer membranes (Millipore, Bedford, MA). Blots were incubated for 1 h in Tris-buffered saline/Tween 20 (10 mM Tris-HCl, pH 8.0, 159 mM NaCl, and 0.1% Tween 20) containing 5% powdered skim milk and 1% bovine serum albumin. After three washes with Tris-buffered saline/Tween 20, membranes were incubated for 2 h with the primary antibody and for 1 h with horseradish peroxidase-conjugated goat anti-rabbit IgG. Proteins were detected using the enhanced chemiluminescence detection kit (Amersham Biosciences, Piscataway, NJ).

Photoaffinity Labeling. [³²P]-p³-1,4-Azidoanilido-p¹-5'-GTP ([³²P]AAGTP) was synthesized as described previously (Rasenick et al., 1994). Membranes (100 µg of protein) from cells transfected with the indicated construct were incubated with [³²P]AAGTP for 3 min in binding buffer (20 mM HEPES, 1 mM EDTA, 2 mM MgCl₂, 20 mM NaCl, 2 mM -mercaptoethanol) plus 0.2 µM GDP for 3 min and followed by 50 µM isoproterenol for each sample. After incubation for 10 min at 37°C, the samples were UV-irradiated on ice for 30 s with a 9-W Mineralight (Black Light Eastern Corp, Westbury, NY) at a distance of 3 cm. The samples were washed with binding buffer by centrifugation three times, and the pellets were resuspended in 20 µl of binding buffer containing 1 mM DTT. Proteins were resolved by SDS-PAGE and visualized by autoradiography of dried gels. Densitometric analysis was carried out. Statistical significance was determined by paired Student's t test.

Measurement of cAMP Accumulation in cyc-S49 Lymphoma Cells. cAMP levels were determined by labeling cyc cells with [³H]adenine and measuring [³H]cAMP formation from [³H]ATP (Marsh et al., 1998). G_s-GFP was introduced into cyc- cells (2 × 10⁷ cells in 0.8 ml of 20 mM HEPES-buffered minimal essential medium) by electroporation at room temperature using a BTX Electro Cell Manipulator 600 (capacitance setting, 1600 µF; voltage setting, 250 V; resistance setting, R4 = 72 ; BTX Inc. San Diego, CA). After electroporation, the cells were added to 5.0 ml of DMEM containing 10% heat-inactivated horse serum in 25-ml tissue culture flask. Approximately 50% of cells survived electroporation, and about 4% of the cells (revealed by fluorescence microscopy) were expressing G_s-GFP. At 24 h after electroporation, the cells were labeled with 14 µCi/ml of [³H]adenine. Twenty-four hours after the [³H]adenine addition, cAMP accumulation was measured. The cells first were washed in assay medium (20 mM HEPES-buffered DMEM) and then were suspended in 5 ml of the same medium with 1 mM 3-isobutyl-1-methylxanthine; and transferred to 1.5-ml microcentrifuge tubes (0.8 ml for each) with and without 0.1 mM isoproterenol. Tubes were incubated at 37°C for 20 min and reactions were terminated by addition of 0.2 ml of 25% cold trichloroacetic acid plus 1 mM concentrations of ATP and cAMP. Nucleotides were separated on ion exchange columns (Salomon et al., 1974). cAMP accumulation was expressed as [³H]cAMP / ([³H]ATP + [³H] cAMP) × 1000. Statistical significance was determined by one-way analysis of variance.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Construction and Expression of G_s-GFP Fusion Proteins. G_s exists as short and long splice variants, G_ss and G_sL, as well as a high-molecular-mass G splice variant, G_sXL (Kehlenbach et al., 1994). Compared with G_ss, G_sL contains an additional 15 amino acids inserted at position 72 of the polypeptide chain, and there is an exchange of glutamate for aspartate at position 71. Although there has been some indication that subtle differences between short G_s and long G_s exist (Seifert et al., 1998; Bourova et al., 1999), the general function of the two forms is similar. Levis and Bourne (1992) modified the long G_s form at a site (residues 77-81) within the 15-amino-acid insert to confer upon it recognition by antibody directed against a well-defined peptide of the influenza hemagglutinin (HA). Addition of a HA epitope did not alter the ability of wild-type G_s to mediate hormonal stimulation of adenylyl cyclase or attach to cell membrane. In the study, a G_s-GFP fusion protein was constructed by replacing the residues (72-81) within G_sL with the GFP sequence (Fig. 1). Thus, GFP was inserted in linker 1 between the helical and GTPase domains.

View larger version (43K): [in this window] [in a new window]   Fig. 1.   Construction of the Gs-GFP fusion protein. A, Schematic of Gs-GFP. Gs-GFP defines the fusion protein in which GFP was inserted within the NH2-terminal domain of the long form of Gs. B, model of Gs-GFP. GFP is depicted as an insert into the Gs structure described by Sunahara et al. (1997).

View larger version (46K): [in this window] [in a new window]   Fig. 2.   Expression of Gs-GFP in COS-1 cells. COS-1 cells were lysed 24 h after transiently transfection with Gs-GFP construct. Thirty micrograms of protein was loaded, separated by SDS-PAGE, and Gs-GFP was detected with polyclonal antibody against Gs (B) or a monoclonal antibody against GFP (A), as indicated. Lane 1 represents a lysate from cell transfected with Gs-GFP, lane 2 is the lysate from nontransfected cells.

View larger version (69K): [in this window] [in a new window]   Fig. 3.   Gs-GFP is associated with the plasma membrane in COS-1, PC12, and HEK 293 cells. Twenty-four hours after transfection, cells were observed by confocal microscopy at 37°C. A computer-generated cross-section of a typical cell is displayed on the top (x-z plane) and on the right (y-z plane). Each image shown is representative of at least 20 cells observed and subjected to a z-scan analysis with similar results.

View larger version (35K): [in this window] [in a new window]   Fig. 4.   Subcellular distribution of Gs-GFP in COS-1 cells. A, particulate and soluble fractions were isolated from cells transfected with Gs-GFP constructs as described under Materials and Methods. Twenty micrograms of protein was loaded, separated by SDS-PAGE gel, and detected with a polyclonal antibody against the C-terminal peptide of Gs. Lanes 1 and 2 represent the soluble portion from the cells transfected with GFP or with Gs-GFP construct, respectively. Lanes 3 and 4 indicate the particulate fraction from cells transfected with GFP or with Gs-GFP construct, respectively. B, densitometric analysis of Gs distribution after cell fractionation by sucrose density gradient sedimentation. Fractions 1 to 5 correspond to the 5% sucrose layer, fractions 6 to 10 are the 15% sucrose layer, fractions 11 to 15 (lanes 11-15) are the 30% sucrose layer, and fractions 16 to 20 (lanes 16-20) are 40% sucrose layer. The insert represents the immunoblot of fractions 11 to 20.

GTP Binding Properties of G_s-GFP Fusion Protein. Based on models derived from the crystal structure of the -subunit of the retinal G-protein transducin, the sequence of the 15-amino-acid insert localized in G_sL serves as a linker between the GTPase domain and the -helical domain (Noel et al., 1993; Sunahara et al., 1997). The guanine nucleotide-binding site is embedded between these two domains. Thus, an expansion of this linker sequence might be expected to diminish the ability of G_s to exchange GDP for GTP in response to agonist. To test this, COS-1 cells were cotransfected with G_s-GFP and -adrenergic receptor. Membranes from those cells were then incubated with the photoaffinity GTP analog [³²P]AAGTP in the presence or absence of isoproterenol. G_s from control cells bound [³²P]AAGTP and this binding was increased significantly in the presence of agonist. Transfected cells showed agonist-induced [³²P]AAGTP binding in both native G_s and G_s-GFP (Fig. 5). Thus, it seems that G_s-GFP fusion protein is coupled to the -adrenergic receptor and binds GTP in response to agonist. Note that in COS-1 cells, about 30% of cells expressed the G_s-GFP construct. In the cells expressing the construct, the content of G_s-GFP is about twice that of the endogenous G_s. Thus, in a given cell population, endogenous G_s exceeds G_s-GFP by a ratio of 3:2. [³²P]AAGTP labeling is consistent with this ratio.

View larger version (19K): [in this window] [in a new window]   Fig. 5.   Gs-GFP binding to AAGTP. COS-1 cells were cotransfected with cDNA encoding Gs-GFP (1 µg) and  adrenergic receptor (4 µg). A, cell membranes were prepared 24 h after transfection and were incubated with [32P]AAGTP in the presence and absence of isoproterenol (as indicated). Proteins were resolved by SDS-PAGE and autoradiography. The result shown is from one of four similar experiments. B, densitometric analysis of Gs-GFP binding to AAGTP. Densitometric analysis of four independent experiments were carried out and displayed in densitometric units. Shown is the mean ± S.E., n = 4; **, significant difference in incorporation compared with control (no isoproterenol) (p < 0.01).

Activation of Adenylyl Cyclase by G_s-GFP Fusion Protein in cyc-S49 Lymphoma Cells. To further test the ability of G_s-GFP to couple to -adrenergic receptor and activate adenylyl cyclase, we measured receptor-dependent accumulation of cAMP after transient transfection of cyc- cells, which lack endogenous G_s. Basal cAMP levels are identical in these cells transfected with G_s-GFP or the GFP vector alone. Isoproterenol did not stimulate adenylyl cyclase in cells expressing the GFP construct, but isoproterenol elicited a significant increase in cAMP in cells expressing G_s-GFP (Fig. 6).

View larger version (16K): [in this window] [in a new window]   Fig. 6.   Gs-GFP increases cAMP accumulation in cycS49 lymphoma cells after isoproterenol treatment. Cyc cells were transfected with GFP or Gs-GFP as indicated; cAMP levels were measured in the absence () and presence () of 0.1 mM isoproterenol. All values represent the mean ± S.E. of five independent experiments. **, significant difference from cells without isoproterenol (p < 0.01).

Internalization of Cholera Toxin Activated G_s-GFP. G proteins are bound at the inner face of the plasma membrane, where they are positioned strategically to interact with membrane-spanning receptors and appropriate effectors. The G_s subunit possesses the guanine nucleotide-binding site, intrinsic GTP hydrolytic activity, and an ADP-ribose acceptor site. Cholera toxin activates G_s by ADP-ribosylating arg201 of Gs and inactivating the intrinsic GTPase. During real-time observation, it was seen that after 20-min treatment with cholera toxin (3 µg/ml), marked internalization of G_s-GFP protein occurred. At 1 h, most of G_s-GFP protein was no longer associated with plasma membrane and was found in the cytoplasm (Fig. 7A). A computer-generated cross-section of cells further indicates that almost complete G_s-GFP internalization occurred at 1 h after cholera toxin treatment (Fig. 7B). Control cells did not show observable internalization of G_s-GFP during this time period. Quantitative image analysis for seven random cells under treated with cholera toxin established that the effect was statistically significant (Fig. 7C).

View larger version (88K): [in this window] [in a new window]   Fig. 7.   Cholera toxin treatment translocates Gs-GFP fusion protein in living PC12 cells. A, twenty-four hours after transfection with GS-GFP, media was replaced as described under Materials and Methods and living cells were viewed by confocal microscopy at 37°C. Cells were initially imaged (0 min), cholera toxin (3 µg/ml) was added, and cells were observed for 1 h. Bar, 10 µm. B, computer-generated cross-section of the whole cell after completion of the hour is displayed on the top (x-z plane) and on the right (y-z plane). C, quantitative image analysis of cells before and during treatment with cholera toxin. Mean gray value were determined and used to quantify the fluorescence intensity within the cytoplasm in cells. **, P < 0.01.

Isoproterenol Activation Rapidly Dissociates G_s-GFP from the Plasma Membrane. In COS-1 cells, activation of G_s is generally caused by activation of the -adrenergic receptor. The COS-1 cells 24 h after transfection with G_s-GFP construct were treated with 50 µM isoproterenol. This resulted in a rapid, partial internalization of G_s-GFP protein. Within 2 min after addition of isoproterenol, the significant release of G_s from plasma membrane was seen, but it only occurred at some regions of the membrane. It is noteworthy that cell shape changes slightly in response to isoproterenol [see Fig. 8 and videos 1 and 2 (http://www.molpharm.aspetjournals.org)], but this does not account for displacement of G_s-GFP protein from the membrane. Although minimal changes may occur in focus during the 2-min period, control cells did not show any significant changes of G_s-GFP protein on the plasma membrane (Fig. 8A and video 1). Figure 8B presents confocal images and results similar to these obtained with the digital fluorescent microscopy in Fig. 8A. It is noteworthy, however, that in some regions of plasma membrane, G_s-GFP fusion protein fluorescence was increased, suggesting that "naive" G_s was inserted into the plasma membrane. During the internalization process, clusters of G_s-GFP protein formed subjacent to the G_s-GFP "rich" regions of the cell (Fig. 8A and video 2).

View larger version (67K): [in this window] [in a new window]   Fig. 8.   Isoproterenol-stimulated rapid internalization of Gs-GFP fusion protein in living COS-1 cells. Cells were transfected with Gs-GFP construct and observed 24 h later at 37°C with confocal or digital fluorescent microscopy. A, cells were treated with (top) or without (bottom) isoproterenol (50 µM), and digital fluorescent images were captured every 5 s. Video 1 shows assembled images at 5-s intervals for the control cells and video 2 shows this for the isoproterenol treated cells. The videos are available at http://molpharm.aspetjournals.org. Arrows indicate areas of major release of membrane-bound Gs-GFP. Clusters of Gs-GFP protein form subjacent to the plasma membrane (indicated by open arrowhead). B, isoproterenol-induced Gs-GFP release from plasma membrane visualized by confocal microscopy. Arrows display regions where clusters of Gs-GFP were released from plasma membrane due to activation of  receptors by isoproterenol. The arrowheads indicate the sites at which clusters of Gs-GFP were inserted during the response to isoproterenol. Bar, 10 µm. These results are typical of 40 of 58 cells observed during the course of 15 experiments. Eighteen cells did not show Gs-GFP internalization in response to isoproterenol. C, COS-1 cells transfected with Gs-GFP and subcultured into 12-well culture plates were treated with isoproterenol for the indicated times. Cells were rapidly harvested and the content of Gs-GFP in soluble fraction was examined by immunoblot analysis.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

This report indicates that GFP can be inserted into the sequence of G_s, without substantially altering the normal function of that G protein. G_s-GFP exhibits properties associated with G_s, suggesting that insertion of GFP at position 71 was transparent with regard to function. We have successfully tracked the behavior of G_s in living cells and investigated the effect of activation by receptor and cholera toxin on the localization of G_s. This study confirms previous studies that indicated translocation of activated G_s (Ransas et al., 1989; Levis and Bourne, 1992; Wedegaertner et al., 1996) and provides insights into the cellular location and behavior of G_s in cells.

Images from unstimulated living cells indicate that although most G_s-GFP are associated with the plasma membrane, some of this protein is in the cytosol. Cell fractionation studies were consistent with this (Fig. 4A). Similar findings were made with a G_q-GFP construct (Hughes et al., 2001). Immunolocalization studies had suggested that G_s might be highly concentrated within such limited membrane regions (Aoki et al., 1992; Aronin and DiFiglia, 1992). Our results provide evidence that distribution of G_s-GFP on the plasma membrane is not uniform (see Fig. 2). The functional significance of this clustering remains to be determined.

Previous studies, using cell fractionation or observations of fixed cells, have suggested that activated G_s may be released from plasma membranes (Rasenick et al., 1984; Ransas et al., 1989; Levis and Bourne, 1992). In this study, cholera toxin-activated G_s-GFP was released from plasma membrane during the course of activation and were gradually distributed through the cytosol. This result coordinates well with observations of HA tagged G_s in HEK 293 cells (Wedegaertner and Bourne, 1994; Wedegaertner et al., 1996). It provides the first real-time evidence from living cells for translocation of activated G_s from plasma membrane into cytoplasm.

The -adrenergic receptor signals primarily through G_s and agonist binding to this receptor causes activation of G_s. We tested receptor-mediated activation of G_s-GFP fusion protein. Within 2 min after addition of isoproterenol, G_s-GFP was translocated from the plasma membrane to cytoplasm. The extent of this release was maximal by 1 min and did not increase significantly after that point. The observation is basically consistent with immunohistochemistry and cell fractionation studies, suggesting that the significant redistribution of activated G_s occurred 5 min after isoproterenol treatment (Levis and Bourne, 1992). It is noteworthy, however, that in some regions of the plasma membrane in COS-1 cells, the fluorescence density was actually increased in the presence of isoproterenol (Fig. 8B). Levis and Bourne (1992) suggested that epitope-tagged G_s, released from the membrane after -receptor activation and might be returned to membrane fraction. Although data in this study might be consistent with this, it is also possible that the G_s-GFP recruited to the membrane during the course of isoproterenol treatment was not identical to the G_s-GFP that was released.

The activated G_s translocation from plasma membrane into cytoplasm might be explained by rapid depalmitoylation of _s, as the activation of _s mediated by both receptor and other mechanisms (e.g., cholera toxin), causes depalmitoylation of _s (Wedegaertner and Bourne, 1994). Moreover, some studies have shown that both mutational activation and removal of the palmitoylation sites reduce association of G subunits with plasma membrane (Levis and Bourne, 1992; Wedegaertner and Bourne, 1994; Wedegaertner et al., 1996; Hughes et al., 2001). Nevertheless, one recent report demonstrated that G_s remained associated with the plasma membrane, regardless of palmitoylation state (Huang et al., 1999). Adding GFP to the NH₂ terminal of G_s, a modification likely to block palmitoylation, yielded a protein that did not associate with the membrane.

It is possible that G_s is unique with regard to agonist-induced membrane dissociation. Experiments with G_q labeled with AAGTP did not reveal that this G protein was released into the cytosol upon activation (Popova et al., 1997; Popova and Rasenick, 2000) and agonist activation did not evoke relocalization of GFP-modified G_q (Hughes et al., 2001). It is noteworthy in this regard that G_s may have an association with specialized membrane domains that is not seen with G_q proteins (Oh and Schnitzer, 2001).

The role and fate of the G_s released from the plasma membrane are unknown at this time. G released from the membrane in response to agonist might interact with cytoskeletal components (Cote et al., 1997). G binds to tubulin and has been shown to regulate microtubule polymerization and microtubule dynamics (Roychowdhury et al., 1999). G_s seems to activate the intrinsic GTPase of tubulin and, in doing so, removes the "GTP cap" that confers microtubule stability. This results in rapid microtubule depolymerization. Changes in cell shape subsequent to agonist activation have been demonstrated (Popova and Rasenick, 2000; Witt-Enderby et al., 2000) and such changes may be particularly relevant in the central nervous system, where it seems that cellular stimulation may result in a rapid reorganization of dendritic spines (Maletic-Savatic et al., 1999). Slight shape changes of COS-1 cells in response to -adrenergic agonist were seen in this study. Thus, it is possible that G_s, released from the membrane subsequent to agonist activation, could influence rapid changes in synaptic shape.

In summary, this report reveals that insertion of GFP into G_s yields a functional protein that can be used to track the behavior of this G protein in living cells. G_s-GFP displays a heterogeneous distribution on the plasma membrane of the cells in which it is expressed and translocates from plasma membrane to cytoplasm after activation. It is suggested that the release of G_s from plasma membranes may play a specific role in the process of cellular signal transduction.