Abstract
Ser54 of Gsα binds guanine nucleotide and Mg2+ as part of a conserved sequence motif in GTP binding proteins. Mutating the homologous residue in small and heterotrimeric G proteins generates dominant-negative proteins, but by protein-specific mechanisms. For αi/o, this results from persistent binding of α to βγ, whereas for small GTP binding proteins and αs this results from persistent binding to guanine nucleotide exchange factor or receptor. This work examined the role of βγ interactions in mediating the properties of the Ser54-like mutants of Gα subunits. Unexpectedly, WT–αs or N54-αs coexpressed with α1B-adrenergic receptor in human embryonic kidney 293 cells decreased receptor stimulation of IP3 production by a cAMP-independent mechanism, but WT-αs was more effective than the mutant. One explanation for this result would be that αs, like Ser47 αi/o, blocks receptor activation by sequestering βγ; implying that N54-αS has reduced affinity for βγ since it was less effective at blocking IP3 production. This possibility was more directly supported by the observation that WT-αs was more effective than the mutant in inhibiting βγ activation of phospholipase Cβ2. Further, in vitro synthesized N54-αs bound biotinylated-βγ with lower apparent affinity than did WT-αs. The Cys54 mutation also decreased βγ binding but less effectively than N54-αs. Substitution of the conserved Ser in αo with Cys or Asn increased βγ binding, with the Cys mutant being more effective. This suggests that Ser54 of αs is involved in coupling changes in nucleotide binding with altered subunit interactions, and has important implications for how receptors activate G proteins.
Introduction
Heterotrimeric G proteins mediate the effects of a vast array of extracellular signals on intracellular events. They consist of an α-subunit that reversibly binds a βγ-dimer (Gilman, 1987; Birnbaumer, 1990; Spiegel, 1992; Neer, 1995; Hildebrandt, 1997; Hamm, 1998). Activated G protein–coupled receptor (GPCR) catalyzes the exchange of GDP for GTP on Gα, leading to the release of α-GTP and free βγ from GPCR and on pathways to independently regulate effectors. Thus, G protein activation by GPCR involves a molecular mechanism coupling guanine nucleotide exchange to subunit dissociation (Gilman, 1987).
Ser54 of αs is a key residue involved in Mg2+ and nucleotide binding, and is a highly conserved residue among GTP binding proteins (Sprang, 1997; Hamm, 1998). The S54N mutant of αs (N54-αs) has a conditional dominant-negative phenotype (Hildebrandt et al., 1991; Cleator et al., 1999). It has intrinsic basal activity, increasing cAMP levels without agonist, but paradoxically blocks hormone stimulation of cAMP levels (Cleator et al., 1999). N54-αs does this by binding receptor nonproductively, preventing GPCR signaling to other G proteins (Cleator et al., 2004). The analogous N17-Ras mutant also has dominant-negative properties based on nonproductive interaction with guanine nucleotide exchange factor, its equivalent of GPCR (Farnsworth and Feig, 1991), and this is true of many small G proteins (Feig, 1999). Homologous mutations of other Gα subunits also often have dominant-negative properties similar to those of N54-αs. For example, studies of analogous S43C or S43N mutants of αt suggest that they are also dominant-negative proteins by directly binding GPCR, in this case rhodopsin (Natochin et al., 2006).
Mutation of the Ser54 homolog in Gα proteins does not always, however, result in proteins with identical characteristics or mechanisms of action. The site analogous to Ser54 of αs in αi2 is Ser48, and in αo is Ser47. Cys mutants of these sites also have a dominant-negative phenotype (Slepak et al., 1993, 1995), but the mechanism in this case is different from those of N17-Ras (Farnsworth and Feig, 1991), N54-αs (Cleator et al., 2004), and N43/C43-αt (Natochin et al., 2006; Ramachandran and Cerione, 2011). C47-αo and C48-αi2 prevent GPCR activation of G proteins by tightly binding βγ (Slepak et al., 1993, 1995). Receptors most efficiently recognize the α-GDPβγ complex (i.e., heterotrimer) (Fung, 1983; Yasuda et al., 1996). Thus, by diminishing the free βγ pool in cells, these α-mutants can decrease the ability of all cellular GPCRs to signal to downstream G proteins. Alternatively, in cases where βγ mediates the effects of a GPCR, such mutants can block downstream signaling by sequestering receptor-generated βγ.
Just recently, it was discovered that dominant-negative Gαi3 subunits are involved in auriculocondylar syndrome (ACS), a rare condition that impairs craniofacial development. ACS is caused by interference of the endothelin type A receptor (ETAR)/phospholipase Cβ (PLCβ) pathway that induces genes critical for craniofacial development (Marivin et al., 2016). This conclusion is supported by the finding that knockout of Gq or G11 in mice results in craniofacial features that resemble ACS (Offermanns et al., 1998; Dettlaff-Swiercz et al., 2005). In humans, mutations are found in genes encoding endothelin-1 (Gordon et al., 2013a), which is the ligand for the ETAR and the downstream effector PLCβ4 (Rieder et al., 2012; Gordon et al., 2013b), but not Gq or G11. Unexpectedly, mutations are found in Gαi3 and are all clustered around the nucleotide binding pocket of Gαi3. One of the mutations fully characterized was the Gαi3 S47R mutant that, remarkably, is homologous to the N54-αs, albeit an arginine (R) is substituted rather than an asparagine (N). The dominant-negative Gαi3 S47R mutant preferentially binds GDP and sequesters its receptor ETAR from activating Gq (Marivin et al., 2016), which is very similar to the way in which N54-αs sequesters the thyroid-stimulating hormone receptor (TSHR) and prevents Gq activation (Cleator et al., 2004).
Here, we looked for an explanation for why the mechanism of action of the Ser54-αs dominant negative is different from that of analogous αi2/o mutants. These studies indicate that the optimal mutations in each protein have opposite effects on their βγ-dimer interactions. These results have implications for the role of this site in translating Mg2+ and nucleotide binding into regulation of subunit interactions. N54-αs binds GTP with lower affinity secondary to altered Mg2+ binding and binds nonproductively to a receptor imparting a dominant-negative phenotype. These results suggest that N54-αs sequesters receptor in a βγ-free state, which challenges the current view of the sequence of events in GPCR activation of heterotrimeric G proteins.
Materials and Methods
Materials.
[3H]inositol, [3H]adenine, and [35S]methionine were from Amersham BioSciences UK Ltd. (Little Chalfont, UK).
Construction of Vectors.
cDNAs for αs, N54-αs, activated αs (αs*), and α1B-adrenergic receptor (α1B-AR) were as described previously (Cleator et al., 1999, 2004). Gβ and Gγ cDNAs were recently described (Dingus et al., 2005). A rat αo cDNA in plasmid Rc/cytomegalovirus from Dr. Randall Reed (Johns Hopkins University School of Medicine) was transferred into pcDNA3.1(+) as an EcoRI-XbaI fragment. The βARK-minigene in the pRK construct was obtained from Robert J. Lefkowitz, as described previously (Koch et al., 1994). The S47N and S47C mutants were generated from this construct using the QuikChange Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA). The sense primer for S47C-αo was ′′GGAGAATCAGGAAAATGCACCATTGTGAAGCAG, changing codon 47 from AGC (Ser) to TGC (Cys). The S47N mutant was made with similar primers changing codon 47 to AAC (Asn). The S54C mutant of αs was generated from the parent αs-pcDNA vector using the primer 5′′GCTTCACAATGGTGCATTTGCCAGACTCTCCAG, changing codon 54 from AGC to TGC.
Cell Transfections and Inositol Phosphate Determination.
Human embryonic kidney 293 (HEK293) cells were grown and transfected with LipofectAMINE (Thermo Fisher Scientific, Waltham, MA) as described previously (Cleator et al., 1999, 2004). Previous studies validated that WT–αs and N54-αs are expressed at comparable levels when cells are transfected with the same amount of cDNA (Cleator et al., 1999, 2004). IP3 production was measured using a [3H]inositol uptake assay (Cleator et al., 2004). Cells were labeled with 2 µCi/ml myo-[3H] inositol 24 hours prior to experiments. Data are presented as the percentage of total [3H]inositol recovered as [3H]inositol phosphates.
In Vitro Transcription/Translation and Biotinylated-βγ Binding Assay.
WT or mutant α was synthesized in a 50-μl solution containing 1 µg cDNA, 10 µCi [35S]Met, 1 mM cold Met, and 40 μl of TNT Quick Coupled Transcription/Translation System (Promega, Madison, WI), as described previously (Dingus et al., 2005). The biotinylated-βγ (bβγ) binding assay was as previously described (Dingus et al., 1994). Protein was translated at 30°C for 90 minutes (by which time there is little or no additional synthesis) and then frozen at −80°C until used in an experiment. Binding of αs or αo to bβγ was in a 200-µl solution containing 20 mM Hepes, pH 7.5, 1 mM EDTA, 1 mM dithiothreitol, 100 mM NaCl, 100 µM GDP, 0.5 µCi translation mixture, and 10 µl of UltraLink Neutravidin Beads (Thermo Fisher Scientific, Waltham, MA) with 0.1 µg of bβγ. Controls included [35S]-α incubated with beads lacking bβγ. Samples were incubated at 4°C for 2 hours on a rotary shaker. Beads were collected in a Picofuge microfuge for 10 seconds and labeled proteins in the pellet separated on 11% SDS-PAGE gels. Fixed, dried gels were exposed for 1–2 days on a Molecular Dynamics (Sunnyvale, CA) phosphoimaging screen and analyzed with a Molecular Dynamics Storm Imager. Data were expressed as relative densities from a single autoradiogram with all samples from an experiment run on a single gel.
Results
Effect of Wild-Type-αs or N54-αs on Signaling to PLCβ2 through the α1B-Adrenergic Receptor in HEK293 Cells.
Previously, we showed that N54-αs blocks Gs-coupled GPCRs from activating downstream G proteins (Cleator et al., 2004). Evidence for this included the observation that in COS-7 cells N54-αs blocked signaling of receptors coupled to Gs, such as the TSHR, vasoactive intestinal peptide (VIP) receptor, and β-adrenergic receptor, but not those coupled primarily to Gq, such as the α1B-AR. For the TSHR, which couples to both Gs and Gq (Allgeier et al., 1994), N54-αs prevented TSH stimulation of both cAMP levels through Gs, and IP3 levels through Gq (Cleator et al., 2004). Interestingly, in COS-7 cells, although N54-αs had essentially no effect on phenylephrine (PE) stimulation of IP3 levels through the α1B-AR, wild-type–αs actually suppressed agonist-stimulated IP3 levels (Cleator et al., 2004). This effect was even more prominent in HEK293 cells, where the expression of wild-type–αs suppressed PE-stimulated IP3 levels by as much as 60% (Fig. 1). In this case, N54-αs also decreased PE-stimulated IP3 levels, but, as in COS-7 cells, N54-αs was less effective than wild-type-αs (Fig. 1A). Importantly, there was a fundamental difference in the effect of N54-αs on signaling through the α1B-AR (Fig. 1A) and its effects on TSHR, VIP receptor, and β-adrenergic receptor that signal through Gs. For Gs-coupled receptors, N54-αs is more potent than wild-type-αs at blocking receptor signaling (Cleator et al., 2004), although here, N54-αs is less effective than wild-type-αs. Because there was a difference in the efficacy of wild-type–αs and N54-αs (Fig. 1B), and because this effect was opposite that for Gs-coupled receptors (Cleator et al., 2004), we wanted to understand the origin of this difference.
Activated Q213L-αs (αs*) was expressed with the α1B-AR in HEK293 cells to examine the possibility that increases in cAMP levels caused by wild-type–αs or N54-αs accounted for their ability to decrease α1B-AR–mediated stimulation of IP3 levels (Fig. 1B). Coexpression of αs*, which causes a greater cAMP increase than N54-αs (Cleator et al., 1999), actually increased PE stimulation of IP3 turnover (Fig. 1B), rather than inhibiting the response, as found with unactivated proteins. We therefore considered other possible mechanisms for an effect of αs on PE-stimulated IP3. One such mechanism would result from the expression of αs suppressing free βγ levels, as proposed for C47 mutants of the αi-related proteins (Slepak et al., 1993, 1995). Such an effect would be consistent with, for example, studies showing that the expression of αo (Yu et al., 1997) or αt (Lustig et al., 1993) blocks βγ-mediated effects in cells. This hypothesis also implied, however, that the N54-αs mutant has decreased affinity for βγ compared with wild-type-αs, since it was less effective in the suppression of PE stimulation of IP3.
Wild-Type or N54-αs Inhibition of β2γ2 Stimulation of PLCβ2.
Wild-type–αs or N54-αs inhibition of β2γ2 stimulation of PLCβ2 was used to study the βγ binding affinity of the proteins in vivo. The ability of αo to bind βγ, thus preventing βγ activation of PLCβ2, has been previously used to study in vivo binding of αo (Yu et al., 1997). β2γ2 coexpressed with PLCβ2 in HEK293 cells caused a 4-fold to 5-fold increase in phosphatidylinositol turnover (Fig. 2A), similar to previously reported results (Liu and Simon, 1996). At high levels of expression, wild-type–αs and N54-αs effectively inhibited β2γ2 stimulation of PLCβ2 (Fig. 2A).
Two possible mechanisms for inhibition effects of αs subunits on βγ stimulation of PLCβ2, one mediated by αs subunits complexing βγ and the other resulting from downregulation of PLCβ2 after its phosphorylation by protein kinase A (Liu and Simon, 1996). cAMP has been shown to downregulate PLCβ2 through phosphorylation mediated by protein kinase A (Liu and Simon, 1996). αs* could be used to explore the contributions of cAMP-mediated inhibition of PLCβ2, whereas the ability of the βARK minigene to bind/sequester βγ could be used to investigate the contributions of αs complexing βγ in inhibiting PLCβ2. Exploring the relative contributions of these two mechanisms, we found that αs* decreased βγ activation of PLCβ2 to a similar extent as the βARK minigene, which scavenges βγ (Koch et al., 1994). Neither αs* nor the βARK minigene was as effective as wild-type–αs or N54-αs. One possible reason for the greater effect of wild-type–αs or N54-αs to inhibit βγ-mediated PLCβ2 activation is that wild-type–αs and N54-αs have a dual effect of increasing cAMP and the ability to bind and sequester βγ. To nullify the inhibition of PLCβ2 by cAMP, αs* was coexpressed with wild-type–αs and N54-αs and β2γ2-mediated PLCβ2 stimulation was measured (Fig. 2B). This strategy was used to measure the binding of wild-type–αs or N54-αs to βγ while minimizing the confounding effects of cAMP. Wild-type–αs was clearly more potent in inhibiting β2γ2-mediated PLCβ2 stimulation compared with N54-αs, suggesting that N54-αs binds βγ with less affinity than wild-type–αs.
Binding of Wild-Type–αs or N54-αs to bβγ.
To directly examine βγ interactions, wild-type–αs and N54-αs were synthesized using an in vitro rabbit reticulocyte system and assayed for their ability to bind to bβγ immobilized on streptavidin beads (Dingus et al., 1994). Both proteins specifically bound bβγ (Fig. 3A). However, significantly less N54-αs than wild-type–αs bound to bβγ (Fig. 3B), closely paralleling the hypothesized reduced affinity of N54-αs seen in the cellular data above. Attempts to conduct a full dose-response curve to determine the affinity of the proteins for bβγ were hindered by the relatively low amounts of protein synthesized (data not shown). At low concentrations, however, particularly at levels below the equilibrium dissociation constant (KD) for binding, the amount of α bound is proportional to the affinity of the interaction. Thus, although saturation could not be reached with the low amounts of αs synthesized, the difference in binding observed is compatible with the affinity of N54-αs for bβγ being at least 3-fold lower than that of wild-type–αs. The argument for a 3-fold decrease in apparent affinity of the mutant for Gβγ is based upon the following: since we could not achieve saturation binding, our binding assays are likely at or below the KD of the binding interaction. Significantly below the KD for binding, the amount bound in a simple bimolecular interaction is directly proportional to the inverse of KD [i.e., bound = free × (Bmax/KD)], or is directly proportional to (KA) [i.e., bound = free × Bmax × KA]. Thus, if binding of the mutant is only 33% of the binding of wild type, and the two have the same Bmax, as designed in the experiment, then the KA of the mutant is 33% or less of that of mutant (i.e., the wild type has a 3-fold higher affinity). The difference is potentially greater than this because, to the degree the higher affinity interaction is saturated, the closer the weaker interaction approaches the same value as the higher affinity interaction, which then obscures the difference in affinities).
One potential complication of the results in Fig. 3B is the possibility that Ser54 mutants of αs are more labile because of decreased affinity for guanine nucleotides (Hildebrandt et al., 1991; Cleator et al., 1999) and increased thermal denaturation of the nucleotide-free protein. This was evaluated by testing the susceptibility of wild-type and N54 mutant protein to temperature-dependent denaturation prior to assaying bβγ binding at 4°C (Fig. 3C). Thus, aliquots of previously synthesized wild-type–αs and mutant N54-αs were then incubated at different temperatures for 30 minutes and then evaluated in the bβγ binding assay. The expectation of this experiment was that, if N54-αs is more temperature sensitive, less N54-αs would bind to bβγ in a subsequent binding assay because of a reduced functional concentration. Although both proteins were sensitive to denaturation at 39°C (essentially a positive control for the denaturation protocol), and the mutant somewhat more so, both proteins were stable to temperatures up to 30°C, indicating that differences in stability did not explain the differences in the apparent affinity of wild-type–αs and N54-αs (Fig. 3B).
Binding of S54-αS and S47-αo, Mutants to bβγ.
The decreased binding of N54-αs to Gβγ, compared with wild-type–αs, suggests differences in βγ binding properties of αs mutants compared with the analogous mutants of αo and αi (S47C), which have increased βγ binding (Slepak et al., 1993, 1995). A possible explanation for this could be that substitution of the conserved Ser by Cys results in a different phenotype than for substitution with Asn. To test this idea, Ser54 in αS was mutated to Cys (C54-αs), whereas Ser47 of αo was mutated to Asn or Cys (N47-αo and C47-αo, respectively) and βγ binding evaluated. C54-αS bound to bβγ less well than wild type, although this effect was clearly not as great as for N54-αs (Fig. 4A). The analogous mutations in αo, in contrast to results with αs, had increased affinity for βγ, with the C47-αo mutant having a greater increase than the N47-αo mutant (Fig. 4B). Thus, although the Asn substitution in αs is more effective than is the Cys substitution, and the Cys substitution is more effective in αo, in both cases the substitutions have phenotypic effects in the same direction.
Discussion
Ser54 of αs and Ser17 of Ras are part of the Mg2+ binding site and first conserved guanine nucleotide binding domain found in both small and heterotrimeric G proteins (Wittinghofer and Pai, 1991). N17-Ras is a dominant-negative protein with clear-cut properties that have made it a particularly important molecular tool for studying Ras signaling in cells (Feig, 1999). It binds nucleotides weakly, has increased preference for GDP, has high affinity for its upstream guanine nucleotide exchange factor, and does not activate downstream effectors (Feig, 1999). In contrast, the phenotype of N54-αs is more complex, which limits the use of N54-αs in cellular studies. N54-αs has a conditional dominant-negative phenotype whereby it has increased intrinsic basal cAMP activity, but paradoxically decreases hormone stimulation of cAMP (Hildebrandt et al., 1991; Cleator et al., 1999). As a dominant-negative phenotype, N54-αS also works upstream by binding the receptor nonproductively, preventing the activation of endogenous αS and, in the case of the TSHR, endogenous αq/11 as well (Cleator et al., 2004). Surprisingly, as shown here, N54-αs binds βγ with lower apparent affinity than does wild-type–αs. This is surprising because analogous αo and αi mutations (S47C) have increased apparent affinity for βγ (Slepak et al., 1993, 1995).
Evidence for decreased association of N54-αS with βγ includes the data presented here in HEK-293 cells, as well as previous data in COS-7 cells (Cleator et al., 2004), showing that wild-type–αs more effectively suppresses the stimulation of IP3 levels through the Gq-coupled α1B-AR than does N54-αs (Fig. 1). In contrast, N54-αs more effectively suppresses receptor increases in cAMP or IP3 mediated by receptors coupled to Gs in addition to or instead of Gq (Cleator et al., 2004). Second, wild-type–αs suppresses the stimulation of PLCβ2 more effectively than N54-αs by β2γ2 coexpressed in HEK293 cells (Fig. 2). Third, under similar conditions, less N54-αs binds to bβγ than wild-type–αs, which is consistent with a 3-fold or greater decrease in the affinity of N54-αs for βγ (Figs. 3 and 4). Finally, a decreased affinity for βγ explains, in part, some of the previously characterized properties of the N54-αs mutant. This relates particularly to the paradoxical increased basal activity of N54-αs despite the fact that it is not activated by receptors (Hildebrandt et al., 1991; Cleator et al., 1999). This can now be explained by its spontaneous activation after dissociation of βγ and by the fact that the GTP-bound protein does activate its downstream effector, adenylyl cyclase (Hildebrandt et al., 1991; Cleator et al., 1999).
Although dominant-negative GTP binding proteins are often the focus of cellular studies of signaling pathways (Feig, 1999; Barren and Artemyev, 2007), the molecular characterization of these proteins also offers opportunities to understand better the molecular mechanism of signaling through normal variants of these proteins (Wall et al., 1998). Thus, the results reported here may shed light on the mechanism by which guanine nucleotide exchange is coupled to subunit dissociation in the process of G protein activation by receptors, in that the N54-αs mutant may constitute an analog of a discrete functional state of the Gs protein. Historically, the properties of point mutations have been used to infer, for example, the concepts of a two-state model for GPCR activation (Lefkowitz et al., 1993), where point mutants of selective GPCRs generate constitutively active receptors (Samama et al., 1993). For G proteins, a discrete “state” is inferred from point mutants inhibiting their GTPase activity that trap the Gα subunit in its active state with GTP bound. In the case of N54-αs, the properties of the state of this protein are that it has increased guanine nucleotide exchange rates for both GDP and GTP (Hildebrandt et al., 1991; Cleator et al., 1999), resulting in an increased preference for GDP over GTP binding, decreased affinity for βγ (shown here), and an increased stable interaction with receptor accounting for its dominant-negative activity (Cleator et al., 2004). Although it is speculative that this would represent a discrete functional state of Gα, supporting evidence for this comes from the recent characterization of an αt-related mutant (Pereira and Cerione, 2005). In this case, mutation of an entirely different site in Switch III, R238E, generates a mutant with a phenotype strikingly similar to that of N54-αs. These similarities include increased nucleotide exchange rates and decreased affinity for guanine nucleotides; insensitivity to NaF, as previously shown for N54-αs (Hildebrandt et al., 1991; Cleator et al., 1999); decreased binding of βγ, a dominant-negative phenotype related to receptor sequestration (Cleator et al., 2004); and an inferred interaction of the βγ-free (or compromised) α-subunit as a stable complex with GPCR (Pereira and Cerione, 2005). Why would two unrelated mutations, in two different Gα isoforms, one a component of the Mg2+/guanine nucleotide binding site (αs Ser54) and the other a part of Switch III (αt Arg238), generate proteins with strikingly similar complexes of biochemical changes? One explanation would be that these residues participate in generating stable states of their respective proteins with specific functional properties related to their activation by receptors.
A long-standing idea about GPCR signaling is that G proteins are activated through a dual mechanism involving nucleotide exchange (GTP for GDP) and subunit dissociation. Figure 5A shows a classic interpretation of G protein activation whereby the receptor catalyzes nucleotide exchange prior to subunit dissociation. The idea that the N54 mutant identifies a discrete state of Gα with low affinity for Gβγ, high affinity for GPCR, and high affinity for GDP suggests an alternative activation sequence mediated by GPCR, one in which subunit dissociation precedes nucleotide exchange (Fig. 5B). In this model, the Gs-β2AR crystal (Dror et al., 2015) would correspond to State III (i.e., the nucleotide-free state). These ideas would also explain the curious observation that in the Gs–β2-adrenergic receptor crystal structure, Gβγ makes no contact with the receptor (Dror et al., 2015), even though Gβγ has long been thought to be required for GPCR activation of Gα (Fig. 5, legend).
Acknowledgments
We thank Alicia Arnold, Roneka Ravenell, and Brook White for expert technical assistance; Dr. Ravi Iyengar for the PLCβ2 and Q212L-αs cDNA; Dr. Diane Perez for the α1B-AR cDNA; Dr. Robert J. Lefkowitz for the βARK minigene; and Dr. Randell Reed for the αo cDNA.
Authorship Contributions
Participated in research design: Cleator, Wells, Dingus, Kurtz, and Hildebrandt.
Conducted experiments: Cleator, Dingus, and Hildebrandt.
Contributed new reagents or analytic tools: Cleator, Dingus, and Kurtz.
Performed data analysis: Cleator, Wells, Dingus, and Hildebrandt.
Wrote or contributed to the writing of the manuscript: Cleator and Hildebrandt.
Footnotes
- Received November 10, 2017.
- Accepted February 23, 2018.
↵1 Current affiliation: Vanderbilt University Medical Center, Division of Cardiovascular Medicine, Nashville, Tennessee.
J.H.C. and C.A.W. were supported by the Medical Scientist Training Program at Medical University of South Carolina. J.H.C. was funded by the Vanderbilt Clinical and Translational Research Award. In addition, this work was supported in part by National Institutes of Health Grants NS38534 and DK37219 (J.D.H.).
Abbreviations
- αs*
- activated αs
- α1B-AR
- α1B-adrenergic receptor
- ACS
- auriculocondylar syndrome
- bβγ
- biotinylated-βγ
- ETAR
- endothelin type A receptor
- HEK293
- human embryonic kidney 293
- GPCR
- G protein–coupled receptor
- IP3
- inositol triphosphate
- N54-αs
- S54N mutant of αs
- PE
- phenylephrine
- PLCβ
- phospholipase Cβ
- TSHR
- thyroid-stimulating hormone receptor
- VIP
- vasoactive intestinal peptide
- Copyright © 2018 by The American Society for Pharmacology and Experimental Therapeutics