POTENTIATION OF GPCR-SIGNALING VIA MEMBRANE TARGETING OF G PROTEIN a SUBUNITS
ABSTRACT
Different assay technologies are available that allow ligand occupancy of G protein coupled receptors to be converted into robust functional assay signals. Of particular interest are universal screening systems such that activation of any GPCR can be detected with a common assay end point. The promiscuous G protein Ga16 and chimeric G proteins are broadly used tools for setting up almost universal assay systems. Many efforts focused on making G proteins more promiscuous, however no attempts have been made to make promiscuos G proteins more sensitive by interfering with their cellular protein distribution. As a model system, we used a promiscuous G protein aq subunit, that lacks the highly conserved six amino acid N-terminal extension and bears four residues of ai sequence at its C-terminus replacing the corresponding aq sequence (referred to as D6qi4). When expressed in COS7 cells, D6qi4 undergoes palmitoylation at its N-terminus. Cell fractionation and immunoblotting analysis indicated localization in the particulate and cytosolic fraction. Interestingly, introduction of a consensus site for N-terminal myristoylation (the resulting mutant referred to as D6qi4myr) created a protein that was dually acylated and exclusively located in the particulate fraction. As a measure of G protein activation D6qi4 and D6qi4myr were coexpressed (in CHO cells) with a series of different Gi/o coupled receptors and ligand induced increases in intracellular Ca2+ release were determined with the FLIPRTM technology (Fluorescence plate imaging reader from Molecular Devices Corp.). All of the receptors interacted more efficiently with D6qi4myr as compared with D6qi4. It could be shown that increased functional responses of agonist activated GPCRs are due to the higher content of D6qi4myr in the plasma membrane. Our results indicate that manipulation of subcellular localization of G protein a subunits—moving them from the cytosol to the plasma membrane-potentiates signaling of agonist activated GPCRs. It is concluded that addition of myristoylation sites into otherwise exclusively palmitoylated G proteins is a new and sensitive approach and may be applicable when functional assays are expected to yield weak signals as is the case when screening extracts of tissues for biologically active GPCR ligands.
Key Words: G protein coupled receptor; G protein; Promiscuous; Ga16; Myristoylation; Signal transduction
INTRODUCTION
Heterotrimeric (abg) G proteins relay signals from activated seven transmembrane (7TM) receptors to intracellular effectors and ion channels, thereby mediating the actions of many hormones, neurotransmitters as well as vision, smell, and taste.[1,2] As 7TM receptors are plasma membrane proteins, the G proteins must become closely apposed to the GPCR to receive and transmit the signal into an intracellular response. However, none of the G protein subunits possess elements within their primary amino acid sequences that are compatible with them being transmembrane proteins. The way, G protein a subunits achieve plasma membrane localization is via amino-terminal myristoylation and/or palmitoylation, and association with the bg subunits.[3–6] Whereas myristoylation occurs cotranslationally and is stable throughout the lifetime of the a subunit, palmitoylation is a posttranslational and dynamic modification, the turnover of which is accelerated upon activation of the G protein.[7–11]
The Gai protein family (ai1, ai2, ai3, ao, az, and at) carries myristate at their amino-termini and all except at are also palmitoylated.[5,6,12,13] Gaq and Gas are exclusively palmitoylated proteins. Additional, but not yet characterized signals for membrane anchorage have been proposed for Ga subunits, but remain to be chemically characterized.[14,15] For Gaq and Ga11, subcellular distribution and the role of N-terminal palmitoylation has been studied in detail by several independent groups.[8,14,16–20] Cell fractionation and immunoblotting analysis indicates that Gaq is expressed in the membrane and the cytosolic fraction. To study the role of palmitoylation in membrane targeting of aq, cysteines 9 and 10, either individually or in combination, were replaced with serines: the C9S/C10S mutation prevents incorporation of [3H]palmitate and results in a substantially greater fraction of the expressed protein in the cytosol.[16,18,19] Whereas the role of palmitate for subcellular localization of Gaq has been studied in great detail,[8,14,16–19] and it is known that depalmitoylation correlates well with a movement of the a subunit from the membrane to the cytosol, it has not been addressed if cytosolic a subunits can be moved to the plasma membrane via inclusion of additional fatty acid anchors. Increased membrane expression of G protein a subunits may represent a useful tool to potentiate signals of activated GPCRs. To study the effect of introducing a myristoylation site into an otherwise exclusively palmitoylated G protein a subunit, we initially generated the following mutant Gaq construct, named D6qi4, as a model system (Fig. 1): (i) the extreme C-terminus (4 aa) of aq was replaced with the corresponding ai sequence. This modification has proven to channel Gi-linked receptors to the Gq-pathway thus allowing for stimulation of PLCb isoforms and intracellular calcium release;[21,22] (ii) the highly conserved, six amino acid extension, characteristic for the Gaq/11 class of G proteins has been deleted. It has been shown that the N-terminal extension is critical for constraining receptor coupling selectivity of Gaq: removal of the 6 aa extension of Gaq confers onto receptors that otherwise exclusively stimulate Gai or Gas the ability to stimulate PLCb and subsequent intracellular calcium release.[23,24] Thus, the mutant D6qi4 represents a G protein a subunit that combines two molecularly different principles of linking Gi-coupled receptors to the Gq-pathway. It will serve as a tool to study whether inclusion of myristate may be applied to move the protein from the cytosol to the plasma membrane and constitutes an attempt to generate an a subunit with expanded capability and sensitivity to interact with Gi-linked receptors.
Figure 1. Alignment of the N- and C-terminal amino acid sequences of selected mutant and wild type G protein a subunits. Gaps were introduced for optimum sequence alignment. D6qi4 denotes a mutant Gaq construct that lacks the highly conserved six amino acid extension, characteristic for Gaq/11 proteins and bears four residues of Gai sequence at the extreme C-terminus, replacing the corresponding Gaq sequence. Replacement of Gly for Ala at codon two of D6qi4 introduces a consensus sequence for N-terminal myristoylation with the resulting mutant being referred to as D6qi4myr.
MATERIALS AND METHODS
DNA Construction
The construction of D6q in pCDNA1 has been reported previously.[23,24] To generate D6qi4, a synthetic EcoRI-NsiI fragment containing the desired codon changes was used to replace the corresponding Gaq wt sequence. D6qi4myr was created by replacing ala at codon 2 of D6qi4 for gly in a synthetic BamHI-FspI fragment using oligonucleotide-based site directed mutagenesis, thus introducing a consensus sequence for the enzyme N-myristoyl transferase at the amino terminus of D6qi4. As already reported for D6q,[23,24] the initiator methionine codon was immediately preceded by the BamHI site of the pCDNA1 multiple cloning site (MCS). D6qi4 and D6qi4myr contain an internal hemagglutinin (HA) epitope tag (DVPDYA), which replaces wt q residues 125–130.[19] The sequence of both plasmids was verified by dideoxy sequencing in both directions.
Cell Culture
CHO-K1 cells were cultured in basal Iscove medium (Biochrom) supplemented with 10% fetal bovine serum (Biochrom), Penicillin–Streptomycin (10,000 IU/mL–10,000 mg/mL; GIBCO), Gentamicin (Roche), 2 mM L-Glutamin (Roche). Cells were cultivated at 37◦C in a humidified 5% CO2 incubator. COS7 cells were cultivated in Dulbecco’s modified Eagle’s medium supplemented with 10% FCS, Penicillin–Streptomycin (10,000 IU/mL–10,000 mg/mL) and 2 mM L-Glutamin under the same conditions.
Transfection and Fluorescence Imaging Plate Reader (FLIPRTM) Assay
For transient transfections 26105 CHO-K1 cells were seeded into 35 mm dishes. About 24 h later cells were transiently transfected at 50–80% confluency with the indicated receptor and G protein constructs (1 mg of plasmid DNA each) using the LipofectAMINE Reagent (GIBCO) for CHO cells according to the manufacturer’s instructions.
Sixteen to eighteen hours after transfection CHO cells were seeded into 96- well plates at a density of 80 cells per well and cultured for 18–24 additional h until used in the functional FLIPR assays. CHO cells were loaded with 95 mL of HBSS containing 20 mM Hepes, 2.5 mM probenecid, 4 mM fluorescent calcium indicator dye Fluo4 (Molecular Probes) and 1% fetal bovine serum for 1 h (37◦C, 5% CO2). Cells were washed three times with PBS containing 1 mM MgCl2, 1 mM EDTA and 2.5 mM probenecid in a Tecan cell washer. After the final wash, 100 mL residual volume remained on the cells in each 96 well. Ligands were aliquoted as 3X solutions into a 96-well plate prior to the assay. The fluorometric imaging plate reader (FLIPR, Molecular Devices) was programmed to transfer 50 mL from each well of the ligand microplate to each well of the cellplate and to record fluorescence during 3 min in one second intervals during the first minute and 3 second intervals during the last 2 min. Peak fluorescence counts from the 18 s to 37 s time points are used to determine agonist activity. The instrument software normalizes the fluorescent reading to give equivalent initial readings at time zero.
Cell Fractionation and Immunoblot Analysis
COS7 cells were transfected in 100 mm dishes with 8 mg of G protein plasmid DNA using the Fugene transfection reagent according to the supplied protocol. For 8 mg of plasmid DNA 12 mL of Fugene were used. 72 h after transfection, cells were washed with ice cold PBS and harvested in a total volume of 2 mL PBS containing 1 mM phenylmethysulfonyl fluoride (PMSF). Cells were pelleted (5 min, 3000 rpm, Hettich EBA 12 centrifuge) and either stored at —80◦C or immediately processed to membranes. Cell pellets were resuspended in an ice cold buffer containing 50 mM tris (pH 8), 2.5 mM MgCl2, 1 mM EDTA, 1 mM PMSF (buffer A) and ruptured with 20 strokes of a hand-held glass homogenizer followed by passage (10 times) through a 27-gauge needle. Nuclei were pelleted (3500 rpm, 5 min, 4◦C) and the postnuclear supernatant was then fractionated (200 g, 30 min, 4◦C) into membrane pellets and supernatants. The supernatant was removed carefully and pellets were resuspended in buffer A. Membrane proteins were quantified with the Bio-Rad protein assay kit using bovine serum albumin (BSA) as a standard. Samples (25 mg of membrane protein and normalized volumes of the supernatant fractions) were resolved by SDS-polyacrylamide gel electrophoresis (12%), transferred to nitrocellulose membranes, and probed with the 12CA5-peroxidase linked monoclonal antibody (1 : 1000). The ECL kit from Amersham was used for detection of antigen–antibody complexes.
[3H]myristate Labeling and Immunoprecipitation
About 48 h after transfections, COS7 cells were metabolically labeled for 2 h with 0.75 mCi/mL of [3H]myristic acid (Dupont NEN, 33 Ci/mmol) in 3 mL of serum free medium containing dialysed fetal calf serum (37◦C, 5% CO2). After labeling, cells were washed twice in PBS, resuspended in 50 mL of buffer B composed of 10 mM tris (pH 8), 150 mM NaCl, 1% SDS and incubated at 70◦C for 10 min. 200 mL of buffer C containing 10 mM tris (pH 8), 150 mM NaCl, 1% Triton X-100, 0.5% sodium deoxycholate, 2 mM EDTA, and 1 mM PMSF were added and samples were tumbled for 30 min (4◦C). Insoluble material and nuclei were pelleted and supernatants were mixed with an equal volume of buffer B and 20 mL of Sepharose 4B beads (Sigma) coupled to the 12CA5 monoclonal antibody (Roche Biosciences). Samples were tumbled for 2 h at 4◦C and then centrifuged gently to pellet the sepharose beads (2000 rpm, EBA12 microcentrifuge, Hettich). Pellets were washed twice with 500 mL of buffer C (diluted 1 : 5) and resuspended in 20 mL of Laemmli sample buffer (Bio-Rad) without reducing agents. Samples were heated for 10 min at 70◦C to gently pellet the beads and supernatants were used for SDS-PAGE (12%). 10% of the total sample volume was subjected to SDS-PAGE/Western blotting analysis, 90% was subjected to SDS-PAGE followed by fluorography. Films were typically developed after 4–6 weeks.
All G protein a subunits were detected with the 12CA5 monoclonal antibody coupled to peroxidase (Roche Biosciences). The antibody is directed against the HA-epitope tag present in all constructs. Samples (25 mg of membrane protein) prepared from transfected COS7 cells were separated by SDS-PAGE (12%), transferred onto nitrocellulose and probed with the 12CA5-POD antibody [1 : 1000 dilution in TBS-T (10 mM tris pH 8, 150 mM NaCl, 10% SDS, 0.1% Tween 20)]. Immunoreactive proteins were visualized with the enhanced chemiluminescence kit (Amersham).
Receptor Ligands
Rantes (recombinant, human) and Somatostatin 14 were from Bachem Biochemica GmbH, Dopamine was from Sigma and U-50488 from Research Biochemicals Inc. (RBI). The remaining chemicals were purchased from Sigma.
RESULTS
Replacement of the last 4–5 C-terminal amino acids of Gaq with the corresponding Gai sequence has been shown to confer onto Gi-linked receptors the ability to stimulate the Gq-pathway in assays measuring intracellular inositoltrisphosphate (IP3) generation or intracellular calcium mobiliza- tion.[21,22,25] Deletion of the highly conserved six amino acid extension, characteristic for Gaq/11 subunits renders these subunits promiscuous: an N-terminally truncated Gaq protein, lacking the first six amino acids—when coexpressed with Gi- and Gs-coupled receptors—confers onto these receptors the ability to stimulate phospholipase Cb and intracellular IP3 generation.[23,24] So far, no G protein a subunits have been created that combine lack of the N-terminal extension and modification of the extreme C-terminus. In an attempt to create a G protein a subunit with increased ability to link Gi-coupled receptors to the Gq-pathway, we initially generated the following mutant Gaq-construct: (i) the highly conserved N-terminal extension, characteristic for Gaq/11 proteins, was deleted and (ii) the last four amino acids of Gaq sequence were replaced by the corresponding Gai sequence. The mutant was named D6qi4 and is schematically depicted in Fig. 1. Initially, expression level and subcellular distribution of D6qi4 was compared with Gaq wt, using membrane preparations of transiently transfected COS7 cells. Figure 2 indicates that both a subunits are predominantly expressed in the particulate fraction and that the expression level of D6qi4 is significantly lower than that of Gaq wt. This finding is in agreement with published results reporting that deletion of the highly conserved N-terminal extension decreases the expression level of the mutant protein[19,23] as opposed to the C-terminal modifications that are reported to leave protein expression unchanged.[21,22,25,26]
Figure 2. Expression of Gaq wt and D6qi4 in COS7 cells. COS7 cells were transfected at 50% confluency in 100 mm dishes with vector DNA [pCDNA3.1(+)] or the indicated G protein a subunits (8 mg plasmid DNA/dish). 48 h after transfection cells were fractionated into membrane pellets (P) and supernatants (S) which were separated using SDS-PAGE (polyacrylamide gel electrophoresis) and immunoblotted using the anti-HA 12CA5-peroxidase liked monoclonal antibody as described under Materials and Methods. The fractionation pattern of Gaq wt and D6qi4 are similar: both a subunits are particulate and soluble with the predominant amount of protein being particulate. D6qi4 is expressed at lower levels as compared with Gaq wt. Similar results were obtained in two additional experiments.
Gaq proteins are known to be palmitoylated at their N-termini, with this modification playing a critical role: mutations that prevent palmitoylation markedly impair membrane association and signalling functions.[16,19,23] Whereas depalmitoylation of Gaq correlates with a movement of the a subunit from the membrane to the cytosol, no attempts have been made so far, to move a cytosolic G protein a subunit from the cytosol to the membrane. In our study, D6qi4 served as a model to explore the consequence of including a consensus sequence for the enzyme N-myristoyltransferase via site directed mutagenesis into the otherwise exclusively palmitoylated mutant Ga subunit. The myristoylated version of D6qi4 will be referred to as D6qi4myr throughout the paper. When comparing the expression levels of D6qi4 and D6qi4myr via immunoblot analysis, it is obvious that D6qi4myr is expressed at a higher level as compared with D6qi4 (Fig. 3A). Metabolic labeling with [3H]myristic acid indicated myristate incorporation by D6qi4myr as opposed to D6qi4 (Fig. 3B) whereas both a subunits incorporate [3H]palmitate (data not shown). In order to find out whether D6qi4myr is simply expressed at a higher level as compared with D6qi4 or whether the higher membrane expression may be due to moving the a subunit from the cytoplasm to the plasma membrane, we performed immunoblot analysis with the monoclonal 12CA5-anti HA antibody on transiently transfected COS7 cells after cell fractionation and compared the subcellular distribution pattern of D6qi4 and D6qi4myr, respectively. Figure 4 indicates that expression level of D6qi4 is lower and that this subunit is particulate and soluble. In contrast, D6qi4myr is exclusively particulate. To compare the functionality of both G protein a subunits, they were transiently coexpressed in CHO cells with a series of Gi/o coupled receptors (chemokine CCR5, kappa opioid KOR, somatostatin SSTR1 and dopamine D2) and agonist induced mobilization of intracellular Calcium was recorded with the FLIPRTM technology. Coexpression of these Gi/o-linked GPCRs with vector DNA and stimulation with the appropriate agonist ligands resulted in a negligible increase in intracellular calcium mobilization (Fig. 5). In contrast, all four receptors when coexpressed with D6qi4 gained the ability to productively mobilize intracellular calcium. Strikingly, coexpression with D6qi4myr resulted in a significant enhancement of agonist potency, shifting the agonist dose response curves to the left by a factor of 10–100 while significantly increasing efficacy of ligand–receptor interaction. In order to find out whether the increased potency and efficacy of ligand receptor interaction is due to the higher expression level of D6qi4myr as compared with D6qi4 or due to the mutation of Ala ! Gly at codon 2, the amount of plasmid DNA used for transfections of D6qi4myr was adjusted to obtain equivalent membrane expression as with D6qi4. Using this experimental set up, dose response curves for G protein coupled receptors coexpressed with D6qi4 and D6qi4myr and stimulated with the appropriate agonist ligand were practically superimposable: an example is given for the somatostatin SSTR1 receptor in Fig. 6.
Figure 3. Immunoblot analysis and incorporation of [3H]myristate of D6qi4 and D6qi4myr. COS7 cells were transfected with vector DNA as a control or the indicated G protein a subunits. B: 48 hours after transfection, cells were labelled with [3H]myristic acid (0.75 mCi/mL, 2 h) followed by immunoprecipitation with the monoclonal 12CA5 antibody. The immunoprecipitated proteins (90% of the samples) were separated on a SDS-12% polyacrylamide gel and analyzed by fluorography as described in the Experimental section. D6qi4 and D6qi4myr were the only immunoprecipitated proteins running at 42 kDa. 10% of the samples were subjected to SDS-PAGE and Westernblot analysis (A). The G protein a subunits were detected with the monoclonal anti-HA 12CA5-POD antibody. Results are indicative of a representative experiment, repeated at least twice.
Figure 4. Expression of D6qi4 and D6qi4myr in COS7 cells. COS7 cells were transfected at 50% confluency in 100 mm dishes with vector DNA [pCDNA3.1(+)] or the indicated G protein a subunits (8 mg plasmid DNA/dish). 72 h after transfection cells were fractionated into membrane pellets (P) and supernatants (S) which were separated using SDS-PAGE and immunoblotted using the anti-HA 12CA5-POD monoclonal antibody as described under Materials and Methods. Fraction S1 represents 90%, fraction S2 10% of the total soluble proteins. D6qi4 can be detected in the particulate and soluble fraction as opposed to D6qi4myr, which is exclusively particulate. Similar results were obtained in two independent experiments with different batches of membranes.
Figure 5. Functional interaction of selected Gi-coupled receptors with D6qi4 and D6qi4myr in a fluorescence plate imaging reader (FLIPRTM) assay. CHO cells were transiently cotransfected in 35 mm dishes with vector DNA, D6qi4 or D6qi4myr and one of the four receptors chemokine CCR5 (A), kappa opioid KOR (B), somatostatin SSTR1 (C) and dopamine D2 (D). Cells were loaded with the Calcium indicator dye Fluo4 and exposed to increasing concentrations of the respective agonists. Changes in fluorescence were recorded using the FLIPRTM technology (Fluorescence imaging plate reader, Molecular Devices corp.) Data are means SD of a representative experiment performed in triplicates. At least three independent experiments gave similar results.
Figure 6. Functional interaction of the somatostatin SSTR1 receptor with D6qi4 and D6qi4myr expressed at comparable levels in the membrane fraction. A: somatostatin-14 (SS-14) mediated intracellular Ca2+ release in CHO cells transiently transfected with the SSTR1 receptor and D6qi4 or D6qi4myr, respectively. Transfections were performed in 35 mm dishes using the following amounts of plasmid DNA: 1 mg of SSTR1 receptor + 1 mg of D6qi4 or vector; 1 mg of SSTR1 receptor + 0.125 mg D6qi4myr + 0.875 mg vector. Cells were loaded with the calcium indicator dye Fluo-4 and exposed to increasing concentrations of the respective agonists. Changes in fluorescence were recorded using the FLIPRTM technology and fluorescence change counts were normalized. Data are means ± SD of a representative experiment performed in duplicates. Two additional experiments gave similar results. B: original Ca2+ tracings of the experiment depicted in A. C: Immunoblot analysis of D6qi4 and D6qi4myr. CHO cells were transfected in 100 mm dishes with vector DNA (16 mg), D6qi4 (8 mg) + vector (8 mg), or D6qi4myr (1 mg) + vector (15 mg). D6qi4myr DNA amount for transfections was chosen to obtain equal membrane expression as compared with D6qi4. Plasma membranes were prepared 48 h after transfection and membrane proteins (40 mg each) separated via SDS-PAGE analysis. The G protein a subunits were detected with the monoclonal anti-HA 12CA5- POD antibody. Results are indicative of a representative experiment, repeated twice with different batches of membranes.
DISCUSSION
Several key regions within G protein a subunits governing receptor-G protein coupling specificity have been identified so far: (i) the extreme C-terminus,[21,22,25–29] (ii) the extreme N-terminus,[23,24] (iii) a region between a4 and a5 helices,[30] (iv) the a3/b5-loop,[31] and (v) a region within the loop linking the N-terminal a-helix to the b1 strand of the Ras-like domain.[32] Whereas all of these domains
have been evaluated independently for their specific interaction with GPCRs, relatively little is known whether these regions interact in an additive fashion.
Several independent groups reported successful usage of chimeric Gaqi proteins the for channeling Gi-coupled receptors to the Gq-pathway.[21,22,25,26] However, some receptors like the Gi-coupled SSTR1 receptor fails to activate Gaqi5.[26] In contrast, removal of the highly conserved N-terminal extension of Gaq (the result- ing mutant referred to as D6q), which is important for constraining receptor-G protein coupling selectivity of this particular a subunit, has not been widely used to link Gi-coupled receptors to the Gq pathway, probably due to the low magnitude of response upon coexpression with GPCRs. Nevertheless, the SSTR1 receptor upon coexpression with D6q, gained the ability to mobilize intracellular IP3 release.[23] In an attempt to create a G protein a subunit with increased sensitivity towards the Gi-coupled receptor class, we constructed the following mutant Gaq subunit: initially, the highly conserved six amino acid extension, characteristic for Gaq/11 proteins was deleted and the last four C-terminal residues of Gaq were replaced with the corresponding Gai sequence (the resulting mutant referred to as D6qi4, Fig. 1). As expected from published data, D6qi4 is expressed at a lower level as compared with Gaq wt (Fig. 2 and Refs.[19,23]). It is reasonable to assume that deletion of the N-terminal extension is responsible for the reduced expression level as various reports have shown that modifications at the extreme C-terminus of Gaq, comprising the last five amino acids, do not interfere with the protein expression level.[21,22,25,26] Nevertheless, due to its location in the particulate and soluble fraction (Fig. 2), D6qi4 constitutes an ideal model system to perform subcellular localization studies. It allows to test the hypothesis whether inclusion of a myristoylation site into the exclusively palmitoylated a subunit allows moving the protein from the cytosol to the plasma membrane. Interfering with the subcellular distribution pattern of a G protein a subunit could constitute an alternative approach of establishing a sensitive assay for transformation of extracellular signals at GPCRs into intracellular responses. Therefore, in a second step, a consensus sequence for the enzyme N-myristoyltransferase was introduced into D6qi4 via site directed mutagenesis replacing Gly for Ala at codon 2. We could show that introduction of a consensus sequence for N-terminal myristoyla- tion into D6qi4 resulted in a significantly increased level of membrane expression (Fig. 3) and that the higher membrane expression is due to a change of the subcellular distribution pattern of the protein: D6qi4 is particulate and soluble whereas D6qi4myr is exclusively particulate (Fig. 4). This study simply examined whether the expressed protein is soluble and/or particulate and it is clear that much of the particulate protein may not necessarily be associated with the plasma membrane. Indeed, it may be trapped in intracellular organelles like ER or Golgi apparatus and may be incorrectly folded or processed. We are aware that presence of a protein in the particulate fraction does not necessarily imply plasma membrane association, however the results of coexpression studies with a series of GPCRs (Fig. 5) clearly proves that the protein in the particulate fraction is functional: all four selected Gi-linked receptors coupled more productively to D6qi4myr as compared with D6qi4. As GPCRs are transmembrane proteins, G proteins must be closely apposed to the GPCR to allow G protein activation and signal transduction. If a G protein a subunit potently transduces signals of an activated GPCR, it must be located at the plasma membrane to fulfill this function. D6qi4myr is functionally superior to D6qi4 and it reasonable to assume that functional superiority is due to the higher plasma membrane expression.
Additional evidence that the particulate fraction of the G protein a subunit D6qi4myr is really plasmamembrane associated can be taken from the experiments diluting the D6qi4myr DNA to obtain equal expression as with D6qi4 (Fig. 6). When expressed at similar levels in the membrane fraction, D6qi4 and D6qi4myr are equipotent in transducing responses of agonist activated GPCRs. In contrast, D6qi4myr clearly achieves higher membrane expression when equal amounts of DNA are used for transfections correlating with increased functional responses of coexpressed GPCRs. Thus, being able to titrate the functional response of a plasmamembrane GPCR via the amount of coexpressed G a subunit is a clear indication for the a subunit being located in the plasmamembrane closely apposed to the GPCR. It should be noted that D6qi4myr potently coupled the somatostatin SSTR1 receptor to intracellular Ca2+ mobilization, a feature not inherent to Gaqi5 or Ga16.[26,30,33] Combining two mechanisms of switching Gi-linked receptors to the Gq-pathway plus high level of membrane expression via N-terminal myristoylation may have concomitantly achieved this. Given the fact, that most of the GPCRs that fail to activate Ga16 belong to the Gi-subfamily,[30] D6qi4myr may receive considerable attention in the future for linking Gi-coupled liganded or orphan receptors to the Gaq-PLCb-Ca2+-release pathway.
It should be mentioned that a previous study, utilizing transfected COS7 cells reported Gaq wt and an N-terminal deletion mutant, lacking the first six amino acids (D6q), to be located exclusively in the particulate fraction.[23] On the other hand, Edgerton et al.[16] have shown Gaq wt and D6q to be particulate and soluble in the same host cell line. In this study Gaq wt and the N-terminal deletion mutant D6qi4 were found to be particulate and soluble. A very recent study by Hughes et al.[18] utilizing HEK293 cells also reported Gaq wt to be particulate and soluble with the predominant fraction being plasma membrane associated. We have no explanation for the discrepancy between Ref. 23 and this study. However, it should be highlighted that moving G protein a subunits from the cytosol to the membrane may only be achieved provided that subcellular distribution of a given a subunit is particulate and soluble. In addition, the possibility exists that the choice of the mammalian expression system determines the relative subcellular localization pattern of respective G protein a subunits. Taken together, these observations suggest that moving G protein a subunits from the cytosol to the membrane to potentiate signaling of GPCRs may not be universally applicable and may depend on the type of mammalian cell line.
In summary, we reported here on the generation of a G protein a subunit with improved ability to recognize Gi-coupled receptors as exemplified by the SSTR1 receptor that does not productively interact with any a subunit available today.[26,33] In addition, we could demonstrate for the first time that increased membrane expression of G protein a subunits can be achieved via moving the protein from the soluble to the particulate fraction utilizing fatty acids as anchoring tools. Given the differences in the subcellular localization pattern of G protein a subunits in various mammalian cell lines, a prerequisite for successful application of this technique is knowledge of the individual subcellular localization pattern Zenidolol of any given a subunit in a respective mammalian cell line.