点击显示 收起
Kimmel Cancer Center, Thomas Jefferson University, Philadelphia, Pennsylvania
Abstract
Regulator of G protein signaling (RGS) proteins modulate G protein signaling by acting as GTPase-activating proteins for G protein -subunits. RGS7 belongs to a subfamily of RGS proteins that exist as dimers with the G protein 5-subunit. In this report, we addressed the mechanisms of plasma membrane localization of 5RGS7. When expressed in human embryonic kidney 293 cells, 5RGS7 was found to be cytoplasmic and soluble. Expression of o promoted a strong redistribution of 5RGS7 to the plasma membrane. Expression of q, however, failed to affect the subcellular localization of 5RGS7. The constitutively active mutant oR179C, like wild-type o, strongly recruited 5RGS7 to plasma membranes; however, inactive oG204A, RGS-insensitive oG184S, and lipidation-deficient oG2A were all defective in the ability to promote plasma membrane localization of 5RGS7. In addition, palmitoylation of RGS7 was demonstrated, and palmitoylation required expression of o or oR179C. To examine potential palmitoylation sites of RGS7, several cysteines were substituted with serines. 5RGS7C133S failed to localize to plasma membranes when coexpressed with o, suggesting cysteine 133 of RGS7 as a putative palmitoylation site. Finally, deletion of amino acids 76 to 128 of RGS7, which includes part of the disheveled, EGL-10, pleckstrin (DEP) domain, prevented o-mediated plasma membrane recruitment of 5RGS7. These findings are the first to demonstrate G-regulated plasma membrane localization and palmitoylation of 5RGS7 and suggest that membrane targeting of 5RGS7 is a complex process requiring at least RGS domain-mediated interaction with o and RGS7 palmitoylation.
Heterotrimeric G proteins, composed of - and -subunits, function as a molecular switches, relaying extracellular stimuli to cytoplasmic signaling pathways. Nucleotide exchange of GTP for GDP on the -subunit sets off a signaling cascade, whereas hydrolysis of the -bound GTP turns off the signaling. A group of proteins called regulators of G protein signaling (RGS) accelerates this GTP hydrolysis and thus modulates the duration of the signal transduction. More than 20 subtypes of the RGS proteins have been identified and are commonly divided into six groups (Hollinger and Hepler, 2002).
RGS7 belongs to the R7 subfamily of the RGS proteins that contain a domain called the G-like (GGL) domain (Sondek and Siderovski, 2001; Witherow and Slepak, 2003). Through this unique domain, RGS7 interacts with the G protein 5-subunit, which deviates significantly from the other four -subunits (Snow et al., 1998). Native 5RGS7 complexes have been isolated from brain extracts (Witherow et al., 2000; Zhang and Simonds, 2000), and copurification experiments suggest that RGS7 always exists as a heterodimer with 5 (Witherow et al., 2000). One critical role for the interaction of RGS7 and 5 is to mutually stabilize each other. Efficient expression of RGS7 in COS-7 cells depends on coexpression of 5 and vice versa (Snow et al., 1999; Witherow et al., 2000), and, moreover, loss of 5 in a mouse knockout causes the complete loss of detectable RGS7 protein in retina and brain extracts (Chen et al., 2003).
Less well understood is the physiological role of 5RGS7, in particular which G protein -subunits interact with and are regulated by 5RGS7. In vitro GAP assays have demonstrated that 5RGS7 acts almost exclusively on o and not i or q (Posner et al., 1999; Hooks et al., 2003). In contrast, 5RGS7 seems to show less -subunit selectivity in various cell systems. For example, 5RGS7 was found to regulate Gi/o protein-coupled receptor activated K+ channels (Keren-Raifman et al., 2001) and to attenuate Ca2+ mobilization mediated by q (Shuey et al., 1998), suggesting interaction with both o and q in cells. A recent report demonstrated fluorescence resonance energy transfer (FRET) between cyan fluorescent protein-tagged q and yellow fluorescent protein-tagged RGS7 in transfected cells, indicating a direct protein-protein interaction between q and 5RGS7 (Witherow et al., 2003). It is interesting that no FRET was observed between o and 5RGS7 in those studies (Witherow et al., 2003). A difficulty in examining 5RGS7 interactions with -subunits has been the inability to detect stable 5RGS7 complexes using either purified proteins or employing coimmunoprecipitation techniques. This is in striking contrast to other RGS proteins, for which it is relatively easy to demonstrate interactions with particular activated -subunits.
To interact with G protein -subunits, RGS proteins are likely to be targeted to plasma membranes (PM). Previous work has demonstrated that 5RGS7 is detected in both cytosolic and membrane fractions of brain extracts and cultured cells (Rose et al., 2000; Witherow et al., 2000; Zhang et al., 2001), and a substantial amount of 5RGS7 was detected in a nuclear fraction (Zhang et al., 2001). Immunofluorescence microscopy of endogenous 5RGS7 in PC-12 cells or overexpressed 5RGS7 in PC-12 or HEK293 cells indicated a predominantly cytoplasmic distribution, with little or no 5RGS7 at the PM, along with some nuclear localization (Zhang et al., 2001; Rojkova et al., 2003). Thus, the molecular mechanisms of PM targeting 5RGS7 are poorly defined. For some RGS proteins, interaction with a PM-localized and activated -subunit promotes translocation of the RGS protein from the cytoplasm to PM. In addition, RGS proteins contain other membrane targeting signals, such as protein-lipid or protein-protein interaction domains or covalently bound lipids, that function to promote regulated or constitutive PM localization (Hollinger and Hepler, 2002). Palmitoylation has been demonstrated to occur on RGS7 when expressed in Sf9 insect cells (Rose et al., 2000), suggesting that this covalent modification could facilitate membrane binding of 5RGS7.
In this report, we examined whether G protein -subunits could promote PM localization and palmitoylation of 5RGS7. We demonstrate that expression of o, but not q, promotes a redistribution of 5RGS7 from the cytoplasm to PM. An analysis of o mutants suggests that the GTP-bound active form, rather than the GDP-bound form, of o preferentially induces PM targeting of 5RGS7. In addition, we demonstrate that o promotes palmitoylation of RGS7, and our results suggest that Cys 133 of RGS7 is a site of palmitoylation.
Materials and Methods
Cell Culture. HEK293 and COS-7 cells were obtained from the American Type Culture Collection (Manassas, VA) and grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum and maintained at 37°C in a 95% air, 5% CO2-humidified atmosphere.
Expression Vectors. pcDNA3.1 encoding o or triple hemagglutinin epitope-tagged human RGS7 (S1 or S2) was purchased from the Guthrie cDNA Resource Center (Sayre, PA). RGS7(S2) is full-length cDNA, and RGS7(S1) lacks amino acids from 76 to 128. A plasmid for myc-His-tagged 5 was provided by David P. Siderovski (University of North Carolina, Chapel Hill, NC). o mutants oG2A, (oC3S), oR179C, oG184S, and oG204A were created using the QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, CA), as were triple hemagglutinin epitope-tagged RGS7C120S, RGS7C133S, and RGS7C206S. The deletion mutants RGS7 17eC75 and RGS7 17eC112 were generated by sequential polymerase chain reaction amplification using pcDNA3.1-RGS7(S2) as a template, and then subcloned into pcDNA3.1 as a KpnI-XhoI fragment.
Transfection. Unless otherwise noted, cells were seeded 1 day before transfection. An indicated amount of DNA constructs was transfected into cells using FuGene 6 (Roche, Indianapolis, IN).
Immunofluorescence Microscopy. Cells were fixed with 3.7% formaldehyde in phosphate-buffered saline for 15 min and permeabilized by incubation in blocking buffer (2.5% nonfat milk and 1% Triton X-100 in Tris-buffered saline) for 20 min. Cells were then incubated with primary antibodies indicated in blocking buffer for 1 h. The cells were washed with blocking buffer and incubated in a 1:250 dilution of a goat anti-mouse or a goat anti-rabbit antibody conjugated with either Alexa 488 or Alexa 594 for 30 min. The coverslips were washed with 1% Triton X-100 in Tris-buffered saline, rinsed in distilled water, and mounted on glass slides with Prolong or Prolong Gold antifade reagent (Molecular Probes, Eugene, OR). Only cells displaying low to moderate levels of fluorescence were examined. Images were recorded with a Olympus BX60 microscope and Sony DKC-5000 digital camera or using an Olympus BX61 microscope and Hamamatsu ORCA-ER digital camera controlled by Slidebook v4.0 (Intelligent Imaging Innovations, Denver, CO). For deconvolved images, image stacks were deconvolved using a constrained iterative algorithm in Slidebook v4.0, and images of "x-y" planes through the middle of cells are presented. Images were transferred to Adobe Photoshop for digital processing.
Cell Fractionation Assay. Soluble and particulate fractions were isolated as described previously (Evanko et al., 2000; Takida and Wedegaertner, 2003). Densitometric quantitation of relative amounts in soluble versus particulate fractions was performed using a Kodak DC40 imaging system.
Palmitoylation Assay. Wild-type and mutants of RGS7 were transfected into COS-7 cells in conjunction with 5 in the absence or presence of o or oR179C. 36 h after transfection, the cells were metabolically labeled with [3H]palmitate for 3 h then lysed. RGS7 was immunoprecipitated using an anti-HA polyclonal antibody. The samples were separated by SDS-PAGE and transferred onto polyvinylidene difluoride membrane. The membrane was sprayed with EnHance (PerkinElmer Life and Analytical Sciences) and exposed to Hyperfilm MP (Amersham Biosciences) at -80°C for 24 to 60 days. After fluorography, the RGS7 protein was detected by immunoblotting using an anti-HA monoclonal antibody. COS-7 cells were used for palmitoylation rather than HEK293 cells because we routinely observe better palmitate labeling of proteins using COS-7 cells (Evanko et al., 2000).
Materials. The anti-o polyclonal antibody was provided by David R. Manning (University of Pennsylvania). Anti-HA polyclonal and anti-q polyclonal antibodies were from Santa Cruz Biotechnology (Santa Cruz, CA). 12CA5 and 9E10 monoclonal antibodies were from Roche and Covance (Berkeley, CA), respectively.
Results
o Specifically Promoted PM Targeting of 5RGS7. We examined effects of -subunit expression on localization of the 5RGS7 complex in HEK293 cells. Both 5 and RGS7 are enriched in brain with little or no expression in other tissues (Rose et al., 2000; Witherow and Slepak, 2003; Zhang et al., 2000), and o is lacking from HEK293 cells (Wang et al., 1999). Thus, this cell line facilitates studies independently of endogenous counterparts. 5 and RGS7 were expressed in HEK293 cells, and then localization of the complex was visualized using an anti-HA monoclonal antibody to detect HA-tagged RGS7. 5RGS7 displayed a diffuse distribution throughout the cytoplasm (Fig. 1A, a). Efficient expression of RGS7 requires coexpression of 5 (Snow et al., 1999; Witherow et al., 2000; Chen et al., 2003), and 5 and RGS7 have been shown to form tight complexes and to colocalize (Witherow et al., 2000; Zhang et al., 2001); thus, it is likely that localization of RGS7 is representative of the 5RGS7 complex, and we refer to RGS7 detection as 5RGS7 herein. Coexpression of o (Fig. 1A, d) led to strong PM localization of 5RGS7 (Fig. 1A, b). In contrast, q (Fig. 1A, e) did not promote 5RGS7's PM targeting (Fig. 1A, c), suggesting that o selectively induces PM localization 5RGS7. Cells were also stained with DAPI to define the location of nuclei (Fig. 1A, feCh).
To further examine subcellular localization of RGS7, we carried out a cell fractionation assay. Transfected cells were lysed in a hypotonic buffer, and the soluble and particulate fractions were separated by ultracentrifugation. Proteins in each fraction were analyzed by immunoblotting using 12CA5 anti-HA monoclonal antibody for RGS7. When expressed alone, 5RGS7 was mostly found in the soluble fraction (Fig. 1B, lanes 1 and 2), whereas coexpression of o resulted in a significant shift of 5RGS7 from the soluble to particulate fraction (Fig. 1B, lanes 3 and 4). In contrast, expression of q did not change the predominantly soluble distribution of 5RGS7 (Fig. 1B, lanes 5 and 6). The results were consistent with the observations of immunofluorescence microscopy, suggesting that o specifically promotes PM localization of 5RGS7 in HEK293 cells.
Effects of o Mutants on Subcellular Localization of 5RGS7. To investigate whether PM localization of 5RGS7 is preferentially mediated by inactive or active forms of o, we next tested the effects of two o mutants, oG204A and oR179C. oG204A is unable to undergo activating conformational changes and is thus considered to be locked in the inactive GDP-bound form. On the other hand, oR179C is constitutively active because of a greatly reduced ability to hydrolyze GTP. RGS7 and 5 were expressed with oR179C or oG204A and localization of the complex was examined as described above. Coexpression of oR179C promoted pronounced PM localization of 5RGS7 (Fig. 2A, aeCd), whereas, in the presence of oG204A, the 5RGS7 dimer predominantly exhibited cytoplasmic, dispersed distribution (Fig. 2A, eeCh). In cell fractionation, 5RGS7 was, as described above, mostly found in the soluble fraction (Fig. 2B, lanes 1 and 2) and coexpression of oR179C led to an almost complete shift of the RGS7 band from the soluble to the particulate fraction (Fig. 2B, lanes 3 and 4). On the other hand, oG204A promoted an increase in the amount of RGS7 in the particulate fraction, but the shift was much less compared with that induced by oR179C (Fig. 2B lanes 5 and 6). We observed that cells expressing transfected oG204A and 5RGS7 at high levels displayed some PM localization of 5RGS7 (data not shown), which probably accounts for the 5RGS7 band in the particulate fraction (Fig. 2B, lanes 5 and 6). Figure 2C shows that the expression level of RGS7 was not markedly changed upon coexpression of various o mutants. Taken together, our findings imply that activated, GTP-bound o preferentially mediates PM localization of 5RGS7.
The o-subunit is modified with fatty acids, and we examined whether this lipidation is required for o to promote PM recruitment of 5RGS7. A 14-carbon saturated myristate attaches to glycine at position 2, and a 16-carbon palmitate modifies cysteine at position 3 of o. oG2A is devoid of both myristate and palmitate because myristoylation is a prerequisite for palmitoylation. When the 5RGS7 dimer was expressed in conjunction with oG2A, 5RGS7 displayed virtually no PM localization (Fig. 2A, ieCl) and remained in the soluble fraction (Fig. 2B, lane 7 and 8), suggesting that lipid modification of o is required for it to mediate PM localization of 5RGS7.
Moreover, we examined 5RGS7's subcellular localization in the presence of an oG184S mutant. This mutant, termed RGS-insensitive, has been shown to have a reduced ability to interact with the RGS domain of a RGS protein (Lan et al., 1998). When oG184S was expressed together with 5RGS7, a majority of transfected cells displayed predominantly cytoplasmic distribution of 5RGS7 (Fig. 2A, meCp). In the fractionation assay, substantially less shift of 5RGS7 to the particulate fraction was induced by oG184S compared with wild-type o and oR179C (Fig. 2B, lanes 9 and 10). The presence of some increased 5RGS7 in the particulate fraction is probably due to some membrane localization when proteins are expressed at very high levels, because fractionation experiments use a population of cells and thus do not distinguish between cells expressing different levels of the proteins. As was the case when oG204A was expressed, when cells expressed oG184S proteins at high levels, some PM localization of 5RGS7 was observed (data not shown). The observations with oG184S suggest that interaction of o with the RGS domain is important in o-mediated PM localization of 5RGS7.
Next, we considered the possibility that the failure of q to promote PM localization of 5RGS7 (Fig. 1) was due to a requirement for the -subunit to be sufficiently activated. Thus, we tested the ability of the constitutively active mutant qR183C to affect localization of 5RGS7. As was the case with q, qR183C failed to promote PM localization of 5RGS7 (Fig. 2). When coexpressed with qR183C, 5RGS7 remained in the cytoplasm of cells (Fig. 2A, qeCt) and was found predominantly in the soluble fraction (Fig. 2B, lanes 11 and 12). These results confirm that q, in contrast to o,isnot effective at promoting PM localization of 5RGS7.
o Induced Palmitoylation of 5RGS7. It has been shown that some RGS proteins are modified with palmitate, a fatty acid known to serve as a membrane targeting signal, and a previous report demonstrated that RGS7 incorporated palmitate when expressed with 5 in Sf9 insect cells (Rose et al., 2000). We looked at whether RGS7 is palmitoylated in mammalian cells. RGS7 and 5 were transfected into COS-7 cells in the absence or presence of o or oR179C. Cells were metabolically labeled with [3H]palmitate and incorporation of radioactivity into RGS7 was analyzed as described under Materials and Methods. Control transfection of empty vector showed no nonspecific binding of the radioactivity (Fig. 3, top, lane 1). No incorporation of radiolabeled palmitate into RGS7 was seen without o expression (Fig. 3, top, lane 2). On the other hand, RGS7 incorporated radioactive palmitate in the presence of o or oR179C (Fig. 3, top, lanes 3 and 4). Expression of RGS7 was confirmed by Western blotting (Fig. 3, bottom).
The site(s) of palmitoylation on RGS7 have not been identified, but three cysteine residues that could serve as potential palmitoylation sites exist in the region between the DEP and GGL domains (Rose et al., 2000). We thus replaced each of those cysteines with serine to create the RGS7 mutants RGS7C120S, RGS7C133S, and RGS7C206S, and tested for o-promoted PM localization. The mutants were expressed in conjunction with 5 and o, and their localization was detected by using the 12CA5 antibody. RGS7C120S was not detected at the PM; however, its expression was extremely low (data not shown); thus, we could not conclusively demonstrate a defect in o-promoted PM localization. 5RGS7C133S expression was somewhat variable, but the typical expression level was approximately 25 to 50% of wild-type 5RGS7 (Fig. 4B, lanes 5eC7). Although expressed at a reduced level, RGS7C133S retained binding to 5 as determined by pull-down experiments using the hexahistidine tag on 5 (data not shown). We thus compared localization of 5RGS7C133S with wild-type 5RGS7. In immunofluorescence microscopy, 5RGS7C133S displayed a diffuse cytoplasmic distribution even when expressed with PM-localized o (Fig. 4A, a and c). When assayed by subcellular fractionation, 5RGS7C133S was partially recruited to the particulate fraction when coexpressed with o (Fig. 4B, lanes 1 and 2), although the portion of RGS7C133S in the particulate was significantly reduced compared with wild-type RGS7 when coexpressed with o (Fig. 1B, lanes 3 and 4). On the other hand, 5RGS7C206S was strongly recruited to PM (Fig. 4A, b and d) or the particulate fraction (Fig. 4B, lanes 3 and 4) by expression of o. Together, the data with the cysteine mutants indicate that Cys133 of RGS7 is critical for PM localization of 5RGS7 and suggest that Cys133 is a putative palmitoylation site.
Amino Acids 76 to 128 Are Crucial for PM Targeting. It has been shown that deletion of the DEP domain in RGS9 resulted in its mislocalization (Martemyanov et al., 2003b). To test a role of the DEP domain and its flanking region in RGS7 subcellular localization, we examined several RGS7 mutants that have deletions of all or portions of the DEP domain. The RGS7 DEP domain comprises amino acids 17 to 112 (Wong et al., 2000). RGS7 17eC75, a deletion of the N-terminal portion of the DEP domain, and RGS7 17eC112, a deletion of the entire DEP domain, were expressed together with 5 and o, but RGS7 17eC75 and RGS7 17eC112 protein was almost undetectable, as assessed by Western blotting (Fig. 5A, lanes 2 and 3). RGS7 76eC128, which lacks the C-terminal portion of the DEP domain and additional flanking residues, displayed a substantially greater level of expression, although slightly reduced compared with wild-type RGS7 (Fig. 5A, lanes 1 and 4). It is noteworthy that 5RGS7 76eC128 was refractory to o-promoted PM localization (Fig. 5B, a); it remained in the cytoplasm, implying that the deleted region is important in o-mediated PM localization of the complex.
Discussion
We demonstrated herein that 5RGS7 is mostly cytoplasmic and soluble when expressed in HEK293 cells, but coexpression of o promoted a strong redistribution of 5RGS7 to the PM. Expression of q, however, did not elicit a similar PM recruitment of 5RGS7. PM localization of 5RGS7 was promoted by constitutively active oR179C, but not the inactive mutant oG204A. Moreover, our results suggest that PM recruitment of 5RGS7 is mediated, at least in part, through interaction of its RGS domain with o, because the RGS-insensitive oG184S mutant showed a decreased ability to recruit 5RGS7 to the PM. In addition to PM recruitment of 5RGS7, we demonstrate that expression of o is required for detectable palmitoylation of RGS7 in COS-7 cells. Finally, mutational analysis of RGS7 indicates that Cys133 is a potential site of palmitoylation. These studies are thus the first to demonstrate regulated PM localization and palmitoylation of RGS7.
The preferred G protein -subunit target of 5RGS7 remains controversial. In GAP assays using purified proteins, 5RGS7 is highly selective for o and exhibits no GAP activity on q (Posner et al., 1999; Hooks et al., 2003). However, when expressed in cells, 5RGS7 can regulate o-, i-, and q-dependent signaling pathways (Shuey et al., 1998; Kovoor et al., 2000; Witherow et al., 2000; Keren-Raifman et al., 2001; Zhang et al., 2002; Witherow et al., 2003; Ghavami et al., 2004). In our studies, o but not q was able to promote PM localization of 5RGS7 in cells. Thus, our assays of PM recruitment are consistent with selectivity of 5RGS7 for o. We were surprised to find our studies in contrast to a recent report showing a FRET signal between wild-type q and 5RGS7 in cells but no FRET signal between o and 5RGS7 (Witherow et al., 2003). We observed no PM recruitment of 5RGS7 when wild-type q was expressed. An explanation for these seemingly contradictory results is not clear. It is interesting that stable complexes between purified 5RGS7 and o, either GDP- or -bound, cannot be detected (Posner et al., 1999), and coimmunoprecipitation and pull-down approaches from cell lysates have likewise failed to isolate o or q bound to 5RGS7 (data not shown) (Witherow et al., 2000). It seems as if the binding of 5RGS7 to o is relatively weak compared with other RGS/G pairs (Posner et al., 1999), and thus the PM recruitment of 5RGS7 by o may provide a surrogate method for monitoring association of 5RGS7 and o in cells. In addition, unidentified proteins probably influence the affinity and selectivity of 5RGS7 binding to -subunits. For example, the related complex, 5LRGS9eC1, fails to form a stable complex with t or o using purified proteins unless the t effector cGMP phosphodiesterase-subunit is also included (Martemyanov and Arshavsky, 2002; Martemyanov et al., 2003a). The identification of similar affinity adaptors, possibly o or q effectors, will shed light on 5RGS7/G specificity in vivo.
How does o mediate RGS7's PM localization Our results suggest a model in which the RGS domain of RGS7 interacts preferentially with active o. The inactive oG204A, a mutant that can serve a dominant-negative function because of an inability to undergo activating conformational changes, failed to induce strong PM localization of 5RGS7, whereas constitutively active mutant oR179C promoted strong PM recruitment of 5RGS7. It is interesting that, in our hands, wild-type o was just as effective as oR179C in recruiting 5RGS7 to the PM in transient transfection experiments. We suspect that wild-type o did so most probably because some fraction of the overexpressed protein was in fact active; it is not uncommon for overexpressed -subunits to show some ability to activate signaling pathways, even in the absence of receptor stimulation or an activating mutation. On the other hand, high amounts of overexpressed wild type o may simply overcome its lower affinity compared with activated o for RGS7. Active o probably interacts directly with 5RGS7's RGS domain to induce PM recruitment, and this proposal is supported by the failure of RGS-insensitive oG184S to promote strong PM localization of 5RGS7; however, the degree to which the G184S disrupts interaction with RGS7 has not been demonstrated (Lan et al., 1998) Our results are consistent with a number of other studies showing that certain activated -subunits can selectively recruit RGS domain-containing proteins to the PM (Druey et al., 1998; Heximer et al., 2001; Bhattacharyya and Wedegaertner, 2003; Day et al., 2003; Masuho et al., 2004). There have been suggestions, though no direct demonstrations, that complexes of 5 and R7 RGS family members, such as RGS7, can interact with inactive -subunits via 5 and thus form a novel G protein heterotrimer. Although we cannot rule out the possibility that PM recruitment of 5RGS7 is mediated by 5 binding to o, our results with o mutants are more consistent with the model that PM localization of 5RGS7 is mediated, at least in part, through RGS domain binding to active o.
However, the presence of the RGS domain of RGS7 seems not to be sufficient for o-induced PM localization of 5RGS7. For example, mutants of RGS7 containing an intact RGS domain but with partial deletion of the DEP domain, RGS7 76eC128, or the mutation C133S were deficient in o-mediated PM recruitment. We identified palmitoylation as a modification of RGS7, and our results suggest that palmitoylation serves as a membrane targeting signaling. A previous report demonstrated that RGS7 was palmitoylated when expressed in Sf9 cells, but we could not detect palmitoylation of RGS7 when expressed together with 5 in COS-7 cells. Coexpression of o or oR179C with 5RGS7, however, promoted RGS7 palmitoylation. Thus, our palmitoylation assays were consistent with observations of PM localization; both palmitoylation and PM localization of 5RGS7 were promoted by o. To address potential sites of RGS7 palmitoylation, several cysteines were mutated individually to serines and o-promoted PM localization was determined. 5RGS7C206S was recruited to the PM just like wild-type 5RGS7. In contrast, 5RGS7C133S displayed a clear defect in PM recruitment. These results are consistent with the idea that Cys133 is a site for palmitoylation of RGS7, although requirements for very long exposure times for palmitate labeling and variable expression of RGS7C133S precluded us from definitively demonstrating that it fails to incorporate palmitate. In addition, we cannot rule out that additional sites of palmitoylation, such as potentially Cys120, exist in RGS7.
Palmitoylation has been identified in several RGS proteins and in some it has been demonstrated to influence their GAP activity. Palmitoylation of RGS4 was shown to inhibit its GAP activity toward z (Tu et al., 1999), whereas palmitoylation of RGS16 increased GAP function (Osterhout et al., 2003). Rose et al. (2000) showed that, when purified from Sf9 cells, membrane-bound RGS7, which is palmitoylated, and cytosolic, nonpalmitoylated RGS7 equipotently stimulated the GTPase activity of o, suggesting that palmitoylation has no effect on RGS7's GAP activity. Thus, for RGS7, a primary function of palmitoylation is probably to facilitate PM targeting of 5RGS7.
Despite the recent breakthrough in identification of bona fide palmitoyltransferases in yeast (Lobo et al., 2002; Linder and Deschenes, 2003, 2004), molecular mechanisms underlying palmitoylation in mammalian cells are still unclear. We have found that mutants of the small GTPase Sar1, which are known to inhibit vesicle transport between ER and Golgi along the exocytic pathway, had no effect on o-mediated membrane localization of 5RGS7 (data not shown). Thus, our findings suggest that RGS7's PM targeting and, presumably, palmitoylation are independent of the conventional exocytic pathway. The o-subunit itself is modified with palmitate, and its palmitoylation has been shown to be Brefeldin A-insensitive (Gonzalo and Linder, 1998), thus suggesting that o does not require a functional Golgi, and by extension does not use the classic exocytic pathway, for palmitoylation and PM targeting. Furthermore, we showed that a lipidation defective oG2A mutant was unable to induce 5RGS7's PM localization (Fig. 2A, c), and oC3S, which is myristoylated but not palmitoylated, was also unable to recruit 5RGS7 to the PM (data not shown), suggesting that palmitoylation of o is required for PM targeting of the 5RGS7 complex. Whether the same palmitoyltransferase can catalyze attachment of palmitate to o and RGS7 remains to be seen.
DEP domains may play roles in subcellular targeting of proteins (Martemyanov et al., 2003b), although the mechanisms are not clear. RGS7 76eC128 was defective in o-promoted PM localization. This may indicate that the DEP domain plays a unique role in facilitating membrane targeting. On the other hand, the 76eC128 deletion may affect the ability of Cys133 to undergo palmitoylation. Our attempts to resolve this question by additional DEP domain deletions were thwarted by very poor expression of RGS7 17eC75 and RGS7 17eC112. These results suggest that deletion of the N-terminal portion of RGS7's DEP domain creates an unstable protein; moreover, these deletion experiments raise the possibility that amino acids 112eC128 could be a critical region for both stability and PM localization of RGS7. A recent report proposed that DEP domains influence subcellular targeting by interacting with SNARE or SNARE-like proteins, and R9AP, which interacts with the DEP domain of RGS9, seems to have a SNARE-like domain (Martemyanov et al., 2003b). In light of this proposal, it is particularly interesting that another recent report used two-hybrid studies to identify snapin, a SNARE complex protein, as a protein that interacts with the N terminus of RGS7 in a region that partially includes the DEP domain (Hunt et al., 2003).
In conclusion, this study demonstrated that o can specifically induce palmitoylation and PM localization of RGS7. To our knowledge, it is the first report to show regulated membrane targeting of RGS7 and to begin to characterize the molecular mechanisms involved.
References
Bhattacharyya R and Wedegaertner PB (2003) Characterization of G alpha 13-dependent plasma membrane recruitment of p115RhoGEF. Biochem J 371: 709-720.
Chen CK, Eversole-Cire P, Zhang H, Mancino V, Chen YJ, He W, Wensel TG, and Simon MI (2003) Instability of GGL domain-containing RGS proteins in mice lacking the G protein beta-subunit Gbeta5. Proc Natl Acad Sci USA 100: 6604-6609.
Day PW, Carman CV, Sterne-Marr R, Benovic JL, and Wedegaertner PB (2003) Differential interaction of GRK2 with members of the G alpha q family. Biochemistry 42: 9176-9184.
Druey KM, Sullivan BM, Brown D, Fischer ER, Watson N, Blumer KJ, Gerfen CR, Scheschonka A, and Kehrl JH (1998) Expression of GTPase-deficient Gi2 results in translocation of cytoplasmic RGS4 to the plasma membrane. J Biol Chem 273: 18405-18410.
Evanko DS, Thiyagarajan MM, and Wedegaertner PB (2000) Interaction with G is required for membrane targeting and palmitoylation of Gs and Gq. J Biol Chem 275: 1327-1336.
Ghavami A, Hunt RA, Olsen MA, Zhang J, Smith DL, Kalgaonkar S, Rahman Z, and Young KH (2004) Differential effects of regulator of G protein signaling (RGS) proteins on serotonin 5-HT1A, 5-HT2A and dopamine D2 receptor-mediated signaling and adenylyl cyclase activity. Cell Signal 16: 711-721.
Gonzalo S and Linder ME (1998) SNAP-25 palmitoylation and plasma membrane targeting require a functional secretory pathway. Mol Biol Cell 9: 585-597.
Heximer SP, Lim H, Bernard JL, and Blumer KJ (2001) Mechanisms governing subcellular localization and function of human RGS2. J Biol Chem 276: 14195-14203.
Hollinger S and Hepler JR (2002) Cellular regulation of RGS proteins: modulators and integrators of G protein signaling. Pharmacol Rev 54: 527-559.
Hooks SB, Waldo GL, Corbitt J, Bodor ET, Krumins AM, and Harden TK (2003) RGS6, RGS7, RGS9, and RGS11 stimulate GTPase activity of Gi family G-proteins with differential selectivity and maximal activity. J Biol Chem 278: 10087-10093.
Hunt RA, Edris W, Chanda PK, Nieuwenhuijsen B, and Young KH (2003) Snapin interacts with the N-terminus of regulator of G protein signaling 7. Biochem Biophys Res Commun 303: 594-599.
Keren-Raifman T, Bera AK, Zveig D, Peleg S, Witherow DS, Slepak VZ, and Dascal N (2001) Expression levels of RGS7 and RGS4 proteins determine the mode of regulation of the G protein-activated K+ channel and control regulation of RGS7 by G beta 5. FEBS Lett 492: 20-28.
Kovoor A, Chen CK, He W, Wensel TG, Simon MI, and Lester HA (2000) Coexpression of G5 enhances the function of two G subunit-like domain-containing regulators of G protein signaling proteins. J Biol Chem 275: 3397-3402.
Lan KL, Sarvazyan NA, Taussig R, Mackenzie RG, DiBello PR, Dohlman HG, and Neubig RR (1998) A point mutation in Go and Gi1 blocks interaction with regulator of G protein signaling proteins. J Biol Chem 273: 12794-12797.
Linder ME and Deschenes RJ (2003) New insights into the mechanisms of protein palmitoylation. Biochemistry 42: 4311-4320.
Linder ME and Deschenes RJ (2004) Model organisms lead the way to protein palmitoyltransferases. J Cell Sci 117: 521-526.
Lobo S, Greentree WK, Linder ME, and Deschenes RJ (2002) Identification of a Ras palmitoyltransferase in Saccharomyces cerevisiae. J Biol Chem 277: 41268-41273.
Martemyanov KA and Arshavsky VY (2002) Noncatalytic domains of RGS9eC1.G 5L play a decisive role in establishing its substrate specificity. J Biol Chem 277: 32843-32848.
Martemyanov KA, Hopp JA, and Arshavsky VY (2003a) Specificity of G protein-RGS protein recognition is regulated by affinity adapters. Neuron 38: 857-862.
Martemyanov KA, Lishko PV, Calero N, Keresztes G, Sokolov M, Strissel KJ, Leskov IB, Hopp JA, Kolesnikov AV, Chen CK, et al. (2003b) The DEP domain determines subcellular targeting of the GTPase activating protein RGS9 in vivo. J Neurosci 23: 10175-10181.
Masuho I, Itoh M, Itoh H, and Saitoh O (2004) The mechanism of membrane-translocation of regulator of G-protein signaling (RGS) 8 induced by G expression. J Neurochem 88: 161-168.
Osterhout JL, Waheed AA, Hiol A, Ward RJ, Davey PC, Nini L, Wang J, Milligan G, Jones TL, and Druey KM (2003) Palmitoylation regulates regulator of G-protein signaling (RGS) 16 function. II. Palmitoylation of a cysteine residue in the RGS box is critical for RGS16 GTPase accelerating activity and regulation of Gi-coupled signalling. J Biol Chem 278: 19309-19316.
Posner BA, Gilman AG, and Harris BA (1999) Regulators of G protein signaling 6 and 7. Purification of complexes with G5 and assessment of their effects on G protein-mediated signaling pathways. J Biol Chem 274: 31087-31093.
Rojkova AM, Woodard GE, Huang TC, Combs CA, Zhang JH, and Simonds WF (2003) G subunit-selective G protein 5 mutant defines regulators of G protein signaling protein binding requirement for nuclear localization. J Biol Chem 278: 12507-12512.
Rose JJ, Taylor JB, Shi J, Cockett MI, Jones PG, and Hepler JR (2000) RGS7 is palmitoylated and exists as biochemically distinct forms. J Neurochem 75: 2103-2112.
Shuey DJ, Betty M, Jones PG, Khawaja XZ, and Cockett MI (1998) RGS7 attenuates signal transduction through the Gq family of heterotrimeric G proteins in mammalian cells. J Neurochem 70: 1964-1972.
Snow BE, Betts L, Mangion J, Sondek J, and Siderovski DP (1999) Fidelity of G protein beta-subunit association by the G protein gamma-subunit-like domains of RGS6, RGS7, and RGS11. Proc Natl Acad Sci USA 96: 6489-6494.
Snow BE, Krumins AM, Brothers GM, Lee SF, Wall MA, Chung S, Mangion J, Arya S, Gilman AG, and Siderovski DP (1998) A G protein gamma subunit-like domain shared between RGS11 and other RGS proteins specifies binding to G5 subunits. Proc Natl Acad Sci USA 95: 13307-13312.
Sondek J and Siderovski DP (2001) Ggamma-like (GGL) domains: new frontiers in G-protein signaling and beta-propeller scaffolding. Biochem Pharmacol 61: 1329-1337.
Takida S and Wedegaertner PB (2003) Heterotrimer formation, together with iso-prenylation, is required for plasma membrane targeting of G. J Biol Chem 278: 17284-17290.
Tu Y, Popov S, Slaughter C, and Ross EM (1999) Palmitoylation of a conserved cysteine in the regulator of G protein signaling (RGS) domain modulates the GTPase-activating activity of RGS4 and RGS10. J Biol Chem 274: 38260-38267.
Wang Y, Windh RT, Chen CA, and Manning DR (1999) N-myristoylation and play roles beyond anchorage in the palmitoylation of the G protein o subunit. J Biol Chem 274: 37435-37442.
Witherow DS and Slepak VZ (2003) A novel kind of G protein heterodimer: the G 5-RGS complex. Recept Channels 9: 205-212.
Witherow DS, Tovey SC, Wang Q, Willars GB, and Slepak VZ (2003) G5.RGS7 inhibits Gq-mediated signaling via a direct protein-protein interaction. J Biol Chem 278: 21307-21313.
Witherow DS, Wang Q, Levay K, Cabrera JL, Chen J, Willars GB, and Slepak VZ (2000) Complexes of the G protein subunit G5 with the regulators of G protein signaling RGS7 and RGS9. Characterization in native tissues and in transfected cells. J Biol Chem 275: 24872-24880.
Wong HC, Mao J, Nguyen JT, Srinivas S, Zhang W, Liu B, Li L, Wu D, and Zheng J (2000) Structural basis of the recognition of the dishevelled DEP domain in the Wnt signaling pathway. Nat Struct Biol 7: 1178-1184.
Zhang JH, Barr VA, Mo Y, Rojkova AM, Liu S, and Simonds WF (2001) Nuclear localization of G protein 5 and regulator of G protein signaling 7 in neurons and brain. J Biol Chem 276: 10284-10289.
Zhang JH, Lai Z, and Simonds WF (2000) Differential expression of the G protein beta(5) gene: analysis of mouse brain, peripheral tissues and cultured cell lines. J Neurochem 75: 393-403.
Zhang JH and Simonds WF (2000) Copurification of brain G-protein beta5 with RGS6 and RGS7. J Neurosci 20: RC59.
Zhang Q, Pacheco MA, and Doupnik CA (2002) Gating properties of GIRK channels activated by Galpha(o)- and Galpha(i)-coupled muscarinic m2 receptors in Xenopus oocytes: the role of receptor precoupling in RGS modulation. J Physiol 545: 355-373.