Characterization of the Gαs Regulator Cysteine String Protein*

Cysteine string protein (CSP) is an abundant regulated secretory vesicle protein that is composed of a string of cysteine residues, a linker domain, and an N-terminal J domain characteristic of the DnaJ/Hsp40 co-chaperone family. We have shown previously that CSP associates with heterotrimeric GTP-binding proteins (G proteins) and promotes G protein inhibition of N-type Ca2+ channels. To elucidate the mechanisms by which CSP modulates G protein signaling, we examined the effects of CSP1–198 (full-length), CSP1–112, and CSP1–82 on the kinetics of guanine nucleotide exchange and GTP hydrolysis. In this report, we demonstrate that CSP selectively interacts with Gαs and increases steady-state GTP hydrolysis. CSP1–198 modulation of Gαs was dependent on Hsc70 (70-kDa heat shock cognate protein) and SGT (small glutamine-rich tetratricopeptide repeat domain protein), whereas modulation by CSP1–112 was Hsc70-SGT-independent. CSP1–112 preferentially associated with the inactive GDP-bound conformation of Gαs. Consistent with the stimulation of GTP hydrolysis, CSP1–112 increased guanine nucleotide exchange of Gαs. The interaction of native Gαs and CSP was confirmed by coimmunoprecipitation and showed that Gαs associates with CSP. Furthermore, transient expression of CSP in HEK cells increased cellular cAMP levels in the presence of the β2 adrenergic agonist isoproterenol. Together, these results demonstrate that CSP modulates G protein function by preferentially targeting the inactive GDP-bound form of Gαs and promoting GDP/GTP exchange. Our results show that the guanine nucleotide exchange activity of full-length CSP is, in turn, regulated by Hsc70-SGT.

G protein signaling (e.g. guanine nucleotide exchange factors (GEFs), 1 guanine nucleotide dissociation inhibitors, and GTPase-activating proteins). Our previous work has identified cysteine string protein (CSP) as a novel modulator of G proteins (1)(2)(3). Although the functional parallels between CSP and established G protein modulators are evident, the molecular mechanism by which CSP regulates G proteins is not yet known.
CSPs are secretory vesicle proteins of 34 kDa that contain three conserved domains: a J domain, a linker domain, and a cysteine string region (Fig. 1A). The J domain is a 70-amino acid region of homology shared by DnaJ (a well characterized bacterial co-chaperone) and many otherwise unrelated eukaryotic proteins. The linker domain is a 30-amino acid domain conserved between human and Drosophila CSP. CSP, so called because it contains a cysteine-rich domain that in rats consists of a string of nine cysteine residues flanked on the C-terminal side by three additional cysteines, is localized on synaptic vesicles (4), zymogen granules (5), and chromaffin granules (6). Most of the cysteine residues are palmitoylated and are required for membrane attachment to the secretory vesicle (7,8). In contrast to the conserved J domain, linker domain, and cysteine string regions, the C terminus diverges among CSP isoforms (9). CSP plays a crucial role in neurotransmitter release; however, its function is controversial. Deletion of the CSP gene in Drosophila melanogaster severely impairs central and presynaptic transmission (10 -12). The small number of flies that survive to adulthood is characterized by spasmic jumping, intense shaking, temperature-sensitive paralysis, and premature death. Deletion of CSP in mice results in severe impairment of synaptic transmission and neurodegeneration with no survival beyond 3 months (13). Age-and temperaturedependent deterioration of CSP null mutants suggests a chaperone role for CSP in regulating conformation and the activity of synaptic protein(s). Indeed, CSP has been proposed to function as a trimeric complex with two molecular chaperones, Hsc70 (70-kDa heat shock cognate protein) (14 -16) and SGT (small glutamine-rich tetratricopeptide repeat domain protein) (17). However, the involvement of Hsc70 and SGT in the CSP modulation of G protein function has not yet been examined.
We have shown that CSP associates with G proteins and promotes G␤␥ inhibition of N-type Ca 2ϩ channels (1)(2)(3). N-type Ca 2ϩ channels are effectors of G␤␥ such that direct binding of G protein ␤␥ subunits to the Ca 2ϩ channel ␣ 1 subunit results in a stabilization of the deep closed states of the channel (18). The mechanism by which CSP modulates G protein function has not yet been fully evaluated. A striking feature of CSP is that it contains two G protein binding sites (Fig. 1). The region of CSP encoded by amino acids 83-112 associates with G␤ and/or G␣␤␥ (2). In contrast, an N-terminal site binds G␣ subunits but not G␤␥ subunits (1). These binding sites do not have homology with established modulators of G proteins. Each CSP binding site independently modulates G protein inhibition of N-type Ca 2ϩ channels, and we have shown that they do this by distinct but yet to be determined molecular mechanisms (2). Here we examine GTPase activity of G␣ s and G␣ i in the presence of CSP and CSP truncations encoding one or both G protein binding sites. Our findings demonstrate that CSP stimulates GDP/GTP exchange of G␣ s and that modulation of G proteins by CSP is, in turn, regulated by the molecular chaperones Hsc70-SGT.

MATERIALS AND METHODS
Preparation of Fusion Proteins-Glutathione S-transferase (GST) fusion proteins of full-length rat CSP1, Hsc70, SGT, CSP 1-82 , and CSP 1-112 were subcloned into pGEX-KG (19), expressed as GST fusion proteins in AB1899 or BL21 cells, and purified as described previously (2,5). G␣ s and G␣ il were expressed in Escherichia coli and purified as described previously (20).
Pull-down Assays-G␣ i1 and G␣ s (30 g) were incubated for 1 h at 25°C in 100 l of 20 mM Tris-HCl (pH 8.0) buffer containing 130 mM NaCl, 2 mM MgSO 4 , and either 10 M GDP or 10 M GDP, 30 M AlCl 3 , and 10 mM NaF or 10 M GTP␥S followed by the addition of 10 g of GST, GST-CSP 1-82 , or GST-CSP 1-112 and further incubation for 1 h at 25°C. The proteins were then mixed with glutathione-agarose beads (10-l bed volume) for 20 min at 25°C with constant shaking. The beads were spun down and washed four times with 1 ml of the same buffer containing 10 M GDP, or 10 M GDP, 30 M AlCl 3 , and 10 mM NaF, or 10 M GTP␥S. Bound proteins were eluted with SDS sample buffer and separated on 12% SDS-gels.
GTPase Assay-Steady-state GTPase reactions were performed using G␣ s or G␣ i1 subunits (0.3 M) in the presence or absence of various concentrations of GST-CSP 1-82 or GST-CSP 1-112 and 0.6 M G␤ 1 ␥ 1 (21). GTPase activity measurements were initiated by mixing proteins with 20 M [␥-32 P]GTP (5 Ci) in 100 l of 20 mM Tris-HCl (pH 8.0) buffer containing 130 mM NaCl and 10 mM MgSO 4 . Aliquots (20 l) were withdrawn at the indicated times and transferred to 100 l of 7% perchloric acid. Nucleotides were precipitated using 700 l of 10% (w/v) charcoal suspension in phosphate-buffered saline. 32  Aliquots of 20 l were withdrawn at the indicated times, mixed with 1 ml of the same buffer containing 1 mM GTP and passed through Whatman cellulose nitrate filters (0.45 m). The filters were then washed three times with 3 ml of the same buffer and counted in a liquid scintillation counter. The apparent rate constant (k app ) values for the binding reactions were calculated by fitting the data to the following equation: Preparation of Rat Brain Homogenate-Rat brains were homogenized in 20 mM Tris-HCl buffer (pH 7.4) containing 2 mM MgSO 4 , 1 mM phenylmethylsulfonyl fluoride, and EDTA-free inhibitor mixture (Roche Applied Science). The homogenate was centrifuged at 100,000 ϫ g for 1 h at 4°C to obtain soluble fractions (Fig. 6, S). The pellet was washed twice with the same buffer and solubilized in the same buffer containing 1% n-dodecyl-␤-D-maltoside (Calbiochem) for 60 min at 4°C. The resulting supernatant (100,000 ϫ g, 1 h, 4°C) constituted solubilized membrane fraction (Fig. 6, M). Protein concentrations were determined by a Bio-Rad protein assay using bovine serum albumin as the standard. All of the procedures were carried out in strict accordance with a protocol approved by the University of Calgary Animal Care Committee.
Immunoprecipitations-CSP polyclonal antibodies raised against the peptide sequence corresponding to amino acids 182-198 of rat CSP and the entire CSP recombinant protein were described previously (5). The membrane fraction solubilized by 1% n-dodecyl-␤-D-maltoside was precleared by incubation with GammaBind G-Sepharose (30-l bed volume) in 20 mM Tris-HCl (pH7.4) buffer containing 130 mM NaCl and 2 mM MgSO 4 for 30 min at 25°C. Immunoprecipitation was achieved by incubating the precleared membrane fraction with the indicated antibodies overnight at 4°C followed by GammaBind G-Sepharose (Amer-sham Biosciences) (20-l bed volume) for 30 min at 25°C. The mixtures were then centrifuged, and the pellets were washed three times with 200 l of incubation buffer containing 50 M GDP. Samples were washed and resuspended in 30 l of sample buffer prior to immunoblotting.
Immunoblotting-Proteins were transferred electrophoretically at 70 V for 45 min from polyacrylamide gels to 0.45 m of nitrocellulose in 20 mM Tris, 150 mM glycine, and 12% methanol. Transferred proteins were visualized with Ponceau S. Nitrocellulose membranes were blocked for nonspecific binding using 5% milk, 0.1% Tween 20, and phosphatebuffered saline (pH 7.3) (137 mM NaCl, 2.7 mM KCl, 4.3 mM Na 2 HPO 4 , 1.4 mM KH 2 PO 4 ) prior to a 2-h incubation with primary antibody. The membranes were washed three times in the above milk/Tween/phosphate-buffered saline solution and incubated with goat anti-rabbit or goat anti-mouse IgG-coupled horseradish peroxidase. Antigen was detected using chemiluminescent horseradish peroxidase substrate (ECL, Amersham Biosciences). Immunoreactive bands were visualized following exposure of the membranes to Amersham Hyperfilm-MP.
cAMP Assays-Human embryonic kidney tsa-201 (HEK) cells were seeded into 12-well plates and grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum. At 70% confluency, cells were transfected with pcDNA3/Rous sarcoma virus/His-hemagglutinin-␤2-adrenergic receptor (gift from M. Bouvier, University of Montreal), pMT2-CSP (1), or a combination of both using FuGENE 6 according to the manufacturer's directions. Forty-eight h post-transfection, cells were incubated for three min at 37°C with serum-free medium, 50 M isoproterenol, or 100 M forskolin. Medium was then aspirated, and cells were washed with phosphate-buffered saline. Cells were lysed with 280 l of 0.1 M HCl. Lysates were centrifuged at 600 ϫ g, and the supernatant was collected. cAMP levels of the supernatant (100 l) were quantified using a Direct cAMP Enzyme Immunoassay kit (Sigma) according to the manufacturer's instructions.

CSP Stimulates Steady-state Hydrolysis of GTP by G␣ s -To
begin to investigate the mechanism by which CSP modulates G proteins, we evaluated GTPase steady-state activity in the presence of purified recombinant full-length CSP (CSP  ). Fig. 1B shows that the GTP hydrolysis by G␣ s was not altered in the presence of CSP. In contrast, GTP hydrolysis was increased ϳ4-fold in the presence of the proposed trimeric chaperone complex CSP 1-198 -Hsc70 1-646 -SGT 1-314 . In the absence of CSP, Hsc70-SGT did not alter GTPase activity, indicating that CSP is required for the G␣ s modulation. Binary complexes of CSP 1-198 -Hsc70 1-646 and CSP 1-198 -SGT 1-314 both triggered an increase in steady-state GTPase activity but to a lesser extent than the trimeric complex. The addition of 50 M ATP to the assay required for Hsc70 activity did not alter GTP hydrolysis (data not shown). Neither the ATPase, Hsc70 (not shown) nor the proposed trimeric chaperone complex Hsc70-SGT-CSP was observed to stimulate GTP hydrolysis in the absence of G␣ s , which demonstrates that the chaperone complex does not hydrolyze GTP. Thus, stimulation of the GTPase activity of G␣ s by CSP is specific and requires Hsc70-SGT but cannot be attributed to a generalized folding/refolding chaperone-like action on G␣ s by Hsc70-SGT.
To further evaluate the modulation of G␣ by CSP, we utilized full-length CSP and purified recombinant truncations of CSP containing either one G protein binding site (CSP 1-82 ) or both G protein binding sites (CSP 1-112 ). Fig. 2 demonstrates that CSP 1-112 triggered a large increase in the rate of GTP hydrolysis by G␣ s but not G␣ i , whereas CSP 1-82 had a modest effect on G␣ s . SGT and Hsc70 did not further enhance the GTP hydrolysis by G␣ s in the presence of either CSP 1-112 or CSP 1-82 (data not shown). In agreement with the data shown in Fig. 1, CSP did not stimulate G␣ s above basal levels in the absence of Hsc70-SGT. The use of CSP 1-82 and CSP 1-112 truncations rather than full-length CSP eliminates any potential influence from the cysteine string region of full-length recombinant CSP in the GTPase assay, given that interactions can occur with this region that are not characteristic of native palmitoylated CSP (22). A dose dependence of the effect of CSP 1-112 on the GTPase activity of G␣ s revealed a relatively strong affinity between the two proteins (EC 50 415 nM) and a maximal stimulation of ϳ8-fold (Fig. 3). Taken together, these results indicate that CSP directly and specifically modulates the GTPase activity of G␣ s through a region encoded by amino acids 1-112, that this region is autoinhibited in full-length CSP, and that autoinhibition is removed by Hsc70-SGT.
Next we examined the effect of CSP 1-112 on GTPase activity of heterotrimeric G s . G␣ s was first preincubated with G␤ 1 ␥ 1 followed by the addition of CSP  . Consistent with the previous report (23), the addition of G␤ 1 ␥ 1 decreased the basal steady-state GTP hydrolysis by G␣ s (Fig. 3). The dose dependence for the effect of CSP 1-112 on heterotrimeric G s was shifted to the right (EC 50 740 nM), whereas the maximal effect was proportional to the stimulation of monomeric G␣ s (Fig. 3). This suggests that because of a partial overlap of the G␤␥ and CSP 1-112 binding sites on G␣ s , CSP is a more potent modulator of the monomeric G␣ subunit.
CSP Stimulates the Nucleotide Exchange of G␣ s -The rate of GTP hydrolysis by G protein ␣ subunits under steady-state conditions is believed to be limited by the rate of guaninenucleotide exchange due to slow release of GDP. Because CSP 1-112 was more effective in stimulating GTPase activity of G␣ s , the effects of GST-CSP 1-112 on guanine nucleotide exchange of G␣ s and G␣ i were tested using the GTP␥S binding assay. CSP 1-112 increased the initial rate of GTP␥S binding to G␣ s but not G␣ i (Fig. 4).
Conformational Selectivity of CSP-G proteins switch between two main conformations: inactive GDP-bound and active GTP-bound. To examine in detail the conformational selectivity of the G␣-CSP association, we investigated the binding of CSP 1-82 and CSP  GST fusion proteins to G␣ s GTP␥S, G␣ s GDPAlF 4 , and G␣ s GDP. The binding of GST fusion proteins was analyzed by coprecipitation of G␣ s with the CSP truncations using glutathione-agarose beads. The amounts of coprecipitated G␣ s were assessed with Coomassie Blue-stained gels. The full-length GST-CSP was not included in this assay as it co-migrates with G␣ s on SDS-PAGE. Fig. 5 demonstrates that both CSP 1-82 and CSP 1-112 bind tightly to G␣ s GDP but not to G␣ s GTP␥S. CSP 1-112 also bound G␣ s GDPAlF 4, reflective of the transition state between inactive and active states; however, CSP 1-82 did not interact with G␣ s GDPAlF 4 . This evidence demonstrates that CSP preferentially associates with inactive GDP-bound conformations of G␣ s subunits.
Association of Native G␣ s and CSP-Next, we evaluated the interaction of native G␣ s -CSP. To investigate the association of G␣ s -CSP in brain tissue, we solubilized rat brain in 1% n-dodecyl-␤-D-maltoside and performed immunoprecipitations. The immunoprecipitations were resolved by SDS-PAGE and were analyzed for the presence of G␣ s and Hsc70 by Western blotting. Fig. 6 demonstrates that anti-CSP polyclonal (peptide), anti-CSP polyclonal (protein), and anti-G␣ s (Santa Cruz) were capable of precipitating G␣ s . These results are consistent with our previous reports demonstrating the association of G␣ and recombinant CSP (1, 2) that have been confirmed by others (24). In contrast, G␣ s did not coprecipitate significantly with rabbit nonimmune serum, GammaBind G-Sepharose, or IgG 1 . Hsc70 was absent from the G␣ s -CSP complexes precipitated with anti-CSP (peptide) and anti-CSP (protein) but present in the G␣ s -CSP complexes precipitated with anti-G␣ s polyclonal. In agreement with these results, we have reported previously that the CSP/Hsc70 interaction is transient (1,14). A nonspecific band representing IgG heavy chains is observed in Fig. 6, lane 6. As expected, G␣ s was significantly enriched in the membrane-bound fraction, and Hsc70 was present in both the membrane and cytosol fractions.
CSP Stimulates cAMP Generation-To further evaluate the role of CSP in G protein signaling, we examined the ability of transiently expressed CSP to stimulate cAMP production. Human endothelial kidney cells were transfected with CSP, ␤ 2 adrenergic receptor, or CSP ϩ ␤ 2 adrenergic receptor. Stimulation of ␤ 2 adrenergic receptor-transfected cells with isoproterenol evoked a 10-fold increase in cellular cAMP levels. In cells expressing full-length CSP and the ␤ 2 adrenergic receptor, isoproterenol evoked a 27-fold increase in cAMP levels (Fig. 7). In the absence of isoproterenol, small increases in cAMP levels were observed in cells transfected with CSP compared with control cells; however, these increases were not significant. Likewise, only slight effects of CSP were observed on forskolininduced cAMP production, which does not require G␣ s ; however, these increases were not found to be significant. These data are consistent with the conclusions that CSP enhances G␣ s -mediated signaling by functioning as a GEF.
In summary, we have provided the first evidence that the DnaJ/Hsp40 family member and secretory vesicle protein CSP modulates the GDP/GTP exchange of G␣ s . Data presented here indicate that upon activation by Hsc70-SGT, CSP is a direct GEF for G␣ s . DISCUSSION We have found that CSP, a secretory vesicle protein, is able to significantly enhance the GTP hydrolysis of G␣ s by stimulating GDP/GTP exchange. CSP is selective for G␣ s and, as such, is the first GEF to be identified for G␣ s . The CSP-G␣ s complex was detectable by immunoprecipitation and affinity chromatography. In addition, expression of CSP in HEK cells increased isoproterenol-stimulated cAMP levels, demonstrating that CSP enhances signaling through G␣ s.
Data presented here show that the molecular chaperones, Hsc70 and SGT, serve as a molecular switch, turning on the GEF activity of CSP. Indeed, we have shown previously that CSP strongly activates the ATPase activity of Hsc70 via its J domain (14). It has also been shown that SGT interacts with the C terminus of both Hsc70 and CSP and increases the ATPase activity of Hsc70 (17). Consistent with our results showing that stimulation of G␣ s is greater for the trimeric CSP-Hsc70-SGT complex than the CSP-Hsc70 and CSP-SGT binary complexes (Fig. 1), the trimeric complex has been reported to have greater Hsc70 ATPase activity than the binary complexes (17). Based on our observation that deletion of the C terminus of CSP (amino acids 113-192) increases GEF activity and makes it independent of Hsc70 and SGT ( Fig. 2 and data not shown), we speculate that the C terminus of CSP serves as a regulatory domain that autoinhibits the GEF activity of CSP. C terminus autoinhibition of GEF activity is switched off in the presence of Hsc70 and SGT. We hypothesize that distinct regions of CSP serve as both the activator (J domain) and target (C terminus) for Hsc70 and SGT and that a conformational change in CSP unmasks the GEF activity (J domain and linker region) for G␣ s . Notably, four CSP isoforms have been identified in mammals with variable C termini but highly conserved J domains, linker regions, and cysteine string regions. Perhaps autoinhibition by amino acids 113-198 and removal of the inhibition by Hsc70-SGT is unique to CSP1 (the isoform examined in this study). Consistent with our domain analysis, recent reports utilizing HIT-T15 insulin-secreting cells (25) and Drosophila (26) as models propose independent function for the J domain, linker region, and C terminus of CSP.
Interestingly, several functional parallels between CSP and the structurally distinct protein, Ric-8A (synembryn), are evident. First, CSP and Ric-8A are both GEFs. CSP is a GEF for G␣ s but not G␣ i1 (Figs. 1-4), whereas Ric-8A is a GEF for G␣ i1 , G␣ q , and G␣ o but not G␣ s (27). Second, CSP and Ric-8A are both abundant neural proteins (5,28). Finally, deletion of either CSP or Ric-8 causes paralysis. In Drosophila, CSP null mutants that survive to adulthood (briefly) are characterized by paralytic uncoordinated sluggish movements, spasmic jumping, intense shaking, and temperature-sensitive paralysis (10). Deletion of CSP in mice causes no significant change in synaptic function during the first 2 weeks of postnatal life, but this is followed by severe impairment of synaptic transmission with no survival beyond 3 months (13). In Caenorhabditis elegans, genetic studies reveal that Ric-8 acts together with G␣ proteins to regulate neuronal transmitter release (29,30). Thus, it seems that CSP, like Ric-8, is responsible for integrating synaptic G protein signaling pathways. In addition, Ric-8 has been recently shown to play an essential role in the G protein receptor-1/2-and G␣-dependent asymmetric division of C. elegans embryos (31).
In contrast to Ric-8, CSP appears to be capable of activating heterotrimeric G␣ s (Fig. 3). In this respect, CSP is functionally similar to AGS1, a RAS-related G protein, which was shown to activate both heterotrimeric G proteins and free G␣ i /G␣ o subunits (32). Nonetheless, CSP is apparently a more potent modulator of monomeric G␣ s . In HEK cells, the activation of ␤ 2 adrenergic receptor by isoproterenol was required for CSP to produce a large additional increase in cAMP levels. Most likely, this effect was a result of CSP targeting and activating the pool of G␣ s GDP generated by the G protein-coupled receptor-dependent activation of G s and GTP-hydrolysis on G␣ s .
An important question concerning the function of CSP will be to determine the specific physiological pool of G␣ s that CSP (either membrane or intracellular) targets. When secretory vesicles are in close proximity to the plasma membrane, CSP may target and activate plasma membrane-bound G␣ s . An alternate and attractive hypothesis is that CSP modulates G proteins located on secretory vesicles. G proteins have been found on cholinergic synaptic vesicles from Discopyge ommata FIG. 7. CSP promotes G␣ s -mediated signaling. HEK cells were transfected with ␤ 2 adrenergic receptor (␤ 2 AR) cDNA, CSP  cDNA, or a combination of both. Forty-eight h later, cells were incubated for 3 min with serum-free medium (Basal), 50 M isoproterenol (␤ 2 AR ϩ I), or 100 M forskolin, lysed, and assayed for cAMP accumulation via a Direct cAMP Enzyme Immunoassay kit. Data represent the mean of duplicate determinations that is representative of four separate experiments. p values for unpaired t tests were 0.002 (***) and not significant (ns) as indicated. (33), rat synaptic vesicles (34), large dense core and small synaptic vesicles (35,36), insulin secretory granules (37), bovine chromaffin granules (38), rat pituitary granules (39), and rat parotid granules (40). It is not known how these G proteins become activated or what role these vesicle G proteins play in exocytosis. Because G protein-coupled receptors and effectors appear to be absent from the vesicles, CSP is an obvious candidate to be a regulator of vesicle G proteins. In fact, although it is clear that CSP plays a significant component in neurotransmission (10,13), its precise role is not yet established. Current reports support a role for CSP in (i) the regulation of exocytosis (41)(42)(43)(44)(45), (ii) the regulation of Ca 2ϩ transmembrane fluxes (46 -51), and (iii) the regulation of folding/refolding of synaptic proteins (13,14). Given that the contributions of vesicular and plasma membrane G␣ s in the various experimental models utilized are not known, it is possible that differences in G␣ s signaling cascades underlie some of the differences observed among these reports.
Two distinct G protein binding sites are found in CSP (2). Deletion mutants encoding site 1 (CSP 83-198 ) or site 2 (CSP 1-82 ) both increased the G protein inhibition of N-type Ca 2ϩ channels in transient expression systems (2), indicating that two separate actions occur within full-length CSP. In pull-down experiments, site 1 (CSP 83-112 ) associates with ␤␥ or ␣␤␥ subunits, whereas site 2 (CSP 1-82 ) binds only to the ␣ subunit. In vitro, the interaction of G proteins with site 1 is robust and stable, whereas association with site 2 is relatively weaker and sensitive to nucleotides and the purity of the G protein preparation (2). The guanine nucleotide exchange rates reported in this study are greatest for the CSP deletion mutants that encode both sites, indicating that both sites are important for G␣ s activation.
Because CSP has been shown to stimulate Hsc70 ATPase, it is possible that CSP activates Hsc70, which in turn modulates the conformation and activity of G␣ s . Two features of our data argue against this interpretation. First, CSP 1-112 stimulates G␣ s GTPase in the absence of Hsc70 ( Fig. 2A). Second, Hsc70-SGT in the absence of CSP does not independently increase GTPase steady-state activity (Fig. 1B). Thus, although Hsc70 is known to manifest its chaperone activity in several different contexts, here we show that Hsc70 modulates the GEF activity of CSP but does not directly modulate G␣ s .
In this study, we confirm that the secretory vesicle protein, CSP, modulates G protein-mediated signal transduction. Furthermore, we demonstrate that the mechanism by which CSP modulates G proteins involves the binding of the region encoded by residues 1-112 of CSP to the GDP-bound conformation of G␣ s and the stimulation of nucleotide exchange. GEF activity of full-length CSP but not CSP 1-112 requires activation by Hsc70-SGT, suggesting that the C terminus of CSP inhibits GEF activity. CSP stimulation of guanine nucleotide exchange is selective for G␣ s . In this regard, CSP serves as an ideal bridge between heterotrimeric GTP protein signaling, molecular chaperones, and secretory vesicles.