Abstract
The neuromodulatory effects of GABA on pyramidal neurons are mediated by GABAB receptors (GABABRs) that signal via a conserved G-protein-coupled pathway. Two prominent effectors regulated by GABABRs include G-protein inwardly rectifying K+ (GIRK) and P/Q/N type voltage-gated Ca2+ (CaV2) ion channels that control excitability and synaptic output of these neurons, respectively. Regulator of G-protein signaling 7 (RGS7) has been shown to control GABAB effects, yet the specificity of its impacts on effector channels and underlying molecular mechanisms is poorly understood. In this study, we show that hippocampal RGS7 forms two distinct complexes with alternative subunit configuration bound to either membrane protein R7BP (RGS7 binding protein) or orphan receptor GPR158. Quantitative biochemical experiments show that both complexes account for targeting nearly the entire pool of RGS7 to the plasma membrane. We analyzed the effect of genetic elimination in mice of both sexes and overexpression of various components of RGS7 complex by patch-clamp electrophysiology in cultured neurons and brain slices. We report that RGS7 prominently regulates GABABR signaling to CaV2, in addition to its known involvement in modulating GIRK. Strikingly, only complexes containing R7BP, but not GPR158, accelerated the kinetics of both GIRK and CaV2 modulation by GABABRs. In contrast, GPR158 overexpression exerted the opposite effect and inhibited RGS7-assisted temporal modulation of GIRK and CaV2 by GABA. Collectively, our data reveal mechanisms by which distinctly composed macromolecular complexes modulate the activity of key ion channels that mediate the inhibitory effects of GABA on hippocampal CA1 pyramidal neurons.
SIGNIFICANCE STATEMENT This study identifies the contributions of distinct macromolecular complexes containing a major G-protein regulator to controlling key ion channel function in hippocampal neurons with implications for understanding molecular mechanisms underlying synaptic plasticity, learning, and memory.
Introduction
The hippocampus plays a crucial role in learning, memory, and spatial navigation by processing the incoming signals from cortex through the trisynaptic circuit composed of sequentially connected neurons in dentate gyrus, CA3, and CA1 regions (Stepan et al., 2015). Critical to this process is the inhibitory influence of GABA imposed by a variety of interneurons on the CA1 pyramidal neurons that provide main output from hippocampus (Klausberger and Somogyi, 2008; Pelkey et al., 2017). Many of the GABA effects on CA1 pyramidal neurons are mediated by GABAB receptors (GABABRs) that belong to G-protein-coupled receptor (GPCR) superfamily and signal via heterotrimeric Gi/o proteins (Padgett and Slesinger, 2010).
The GABABRs mediate their inhibitory effects by activating several signaling pathways most prominently including direct modulation of ion channels by G-protein βγ subunits. In the dendrites, Gβγ, liberated from Gαi/o by GABABRs, binds and opens G-protein-gated inwardly rectifying K+ (GIRK/Kir3) channels, producing slow IPSCs (sIPSCs) and the ensuing hyperpolarization and decrease in excitability (Lüscher and Slesinger, 2010; Dascal and Kahanovitch, 2015). In the axons, Gβγ, released by the GABABRs, binds and inhibits CaV2 voltage-gated channels (N and P/Q types), suppressing neurotransmitter release and thereby inhibiting the synaptic output of CA1 neurons (Zamponi and Currie, 2013). The inhibitory signaling by GABABRs via GIRK and CaV2 is important for hippocampal synaptic plasticity and memory formation (Davies et al., 1991; Wagner and Alger, 1995; Schuler et al., 2001) and its dysfunctions are thought to contribute to a variety of neuropsychiatric conditions including epilepsy, Down syndrome, and motor and cognitive impairments (Schuler et al., 2001; Alonso et al., 2008; Cramer et al., 2010; Zamponi et al., 2010; Victoria et al., 2016)
A critical role in controlling the strength and timing of GABABR signaling to GIRK in CA1 neurons belongs to the RGS7/Gβ5 protein complex (Xie et al., 2010; Zhou et al., 2012; Ostrovskaya et al., 2014). Being a constitutive dimer, RGS7/Gβ5 functions as a GTPase activating protein (GAP) that accelerates G-protein inactivation (Anderson et al., 2009) to limit Gβγ-mediated GIRK activation and facilitate current deactivation upon termination of the GABABR signaling. Accordingly, elimination of either RGS7 (Ostrovskaya et al., 2014) or Gβ5 (Xie et al., 2010) in mice profoundly slows GIRK channel deactivation kinetics and sensitizes GIRK for the inhibitory effect of GABA. Intriguingly, recent studies revealed that, in the brain, RGS7 forms macromolecular complex with two other auxiliary subunits: RGS7 binding protein (R7BP) (Drenan et al., 2005; Martemyanov et al., 2005) and orphan receptor GPR158 (Orlandi et al., 2012). Biochemical studies show that both proteins stimulate RGS7 activity in catalyzing Gαi/o deactivation (Drenan et al., 2006; Masuho et al., 2013; Orlandi et al., 2015). Furthermore, R7BP and GPR158 also promote membrane localization of RGS7/Gβ5 in transfected cells and, in the brain, neurons (Drenan et al., 2005; Song et al., 2006; Anderson et al., 2007a; Orlandi et al., 2012). GPR158 was also documented to influence RGS7 abundance in the brain (Orlandi et al., 2015). Interestingly, interaction of RGS7 with R7BP and GPR158 is mutually exclusive (Orlandi et al., 2012), indicating that RGS7 exists in two distinct alternative configurations at the plasma membrane and raising a provocative possibility of functionally distinct complexes involved in regulation of GABABR signaling. Indeed, evidence supports R7BP involvement in GABABR function revealing distinct effects on GIRK currents (Zhou et al., 2012; Ostrovskaya et al., 2014). However, the role of GPR158 in this process is completely unknown, as is the contribution of any components of the RGS7 complex to CaV2 regulation.
In this study, we identify CaV2 as an effector ion channel regulated by RGS7 and use molecular and genetic approaches to investigate the role of alternative configurations of RGS7 complexes in controlling GABABR signaling to CaV2 compared with its impacts on GIRK. We revealed an unexpected mechanism whereby the ability of RGS7 to regulate GIRK and CaV2 is promoted by R7BP but inhibited by GPR158.
Materials and Methods
Animals.
All studies were performed in accordance with National Institutes of Health (NIH) guidelines and were granted formal approval by the Institutional Animal Care and Use Committee of the Scripps Research Institute. The generation of Gβ5−/− (Chen et al., 2003), RGS7−/− (Cao et al., 2012), R7BP−/− (Anderson et al., 2007b), and GPR158−/− (Orlandi et al., 2015) mice were described earlier. All animals used for comparing genotypes were littermates derived from heterozygous breeding pairs. Double knock-out (DKO) mice were generated by breeding R7BP and GPR158 KO mice and then crossing the R7BP KO/GPR158 heterozygous HET parents. Mice were housed in groups on a 12 h light/dark cycle with food and water available ad libitum. Males and females (2–5 months of age) were used for all experiments.
Antibodies, Western blotting, and recombinant proteins.
Lysates were prepared by homogenizing hippocampal tissue from age-matched littermates by sonication in lysis buffer containing 300 mm NaCl, 50 mm Tris-HCl, pH 7.4, and 1% Triton X-100 and complete protease inhibitor mixture (Roche Applied Science), incubated on a rocker for 30 min at 4°C, and cleared by centrifugation at 14,000 × g for 15 min. The supernatant was saved and the protein concentration was obtained using the 660 nm Protein Assay (Thermo Fisher Scientific). Samples were diluted in 4× SDS sample buffer and analyzed by SDS-PAGE. Signals were captured on film, scanned by densitometer, and band intensities were determined using ImageJ software. Rabbit anti-R7BP (TRS) and rabbit Gβ5 (ADTG) were generous gifts from Dr. William Simonds [National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK)–NIH]. Rabbit anti-GIRK1 antibodies were a generous gift from Dr. Kevin Wickman (University of Minnesota, Minneapolis, MN). Rabbit anti-Gβ1 was a generous gift from Dr. Barry Willardson (Brigham Young University, Provo, UT). Rabbit antibodies against the intracellular C terminus of mouse GPR158 (GPR158CT) and N terminus of RGS7 (RGS7NT) were described previously (Orlandi et al., 2015). The following antibodies were used: anti-GAPDH (Millipore), anti-Gαo (Cell Signaling Technology), anti-GIRK2 (Alomone Laboratories), anti-GABABR2 (Neuromab), and anti-GFP (Roche Applied Science). Rabbit anti-RGS7 (7RC1) antibodies used for immunogold electron microscopy were a kind gift from Dr. William Simonds (NIDDK/NIH).
Recombinant RGS7 was coexpressed with Gβ5 in Sf9 insect cells via baculovirus-mediated delivery and the recombinant complexes were purified by HisTALON Superflow Cartridge (Clontech Laboratories) chromatography using His-tag present at the N termini of RGS7 proteins as described previously (Martemyanov et al., 2005).
Subcellular fractionation.
For subcellular fractionation experiments, tissues were homogenized in ice-cold lysis buffer containing 150 mM NaCl, 50 mM Tris-HCl, pH 7.4, 1 mM EDTA, 2.5 mM MgCl2, and complete protease inhibitor mixture (Roche Applied Science) by sonication. Lysates were adjusted to the same protein concentration with lysis buffer and equal amounts were subjected to ultracentrifugation (200,000 × g for 30 min/4°C). The supernatant was recovered and designated as cytosolic fraction. The pellet was washed with the lysis buffer and re-sedimented by centrifugation (200,000 × g for 30 min/4°C). The pellet was then resuspended in detergent buffer containing 300 mm NaCl, 50 mm Tris-HCl, pH 7.4, 1% Triton X-100, and complete protease inhibitor mixture, incubated on a rocker for 30 min/4°C, and cleared by centrifugation at 14,000 × g for 15 min. The supernatant was saved and designated as the membrane fraction.
Recombinant helper-dependent adenovirus for GPR158 overexpression.
GPR158 was cloned into the high-level neuronal transgene expression cassette pUNISHER (Montesinos et al., 2011) for rapid and long term in vivo neuronal expression in the CNS as described previously. HdAd was stored at −80°C in storage buffer containing 10 mm HEPES, 250 mm sucrose, and 1 mm MgCl2, pH 7.4). Viral particles per milliliter were calculated as follows: viral particles/ml = (A260) × (dilution factor) × (1.1 × 1012) × (36)/(size of the vector in kb) viral titer: HdAd 23E4 Pun GPR158 syn EGFP 7.11 × 1012 vp/ml.
In situ hybridization.
Expression of Gpr158 and R7bp mRNAs was evaluated with the ViewRNA 2-plex In Situ Hybridization Assay (Affymetrix) using the following probe sets: Gpr158 (NM_001004761; catalog #VB1-11518), R7bp (NM_029879; catalog #VB6-16884). A probe against the E. coli gene DapB (NC_000913; catalog #VF1-10272) was used as a specificity control as recommended by the manufacturer. Briefly, mouse brains were embedded in optimal cutting temperature medium, flash frozen in liquid nitrogen, cut in 14 μm coronal sections, and rapidly fixed in 4% paraformaldehyde for 10 min. Sections were then washed and incubated for 2 h at room temperature in prehybridization mixture containing 50% deionized formamide, 5× SSC, 5× Denhardt's solution, 250 μg/ml yeast tRNA, and 500 μg/ml sonicated salmon sperm DNA, followed by overnight incubation at 40°C with the manufacturer's hybridization solution containing TYPE 1 and TYPE 6 QuantiGene ViewRNA probe sets diluted 1:100. Sections were then processed according to manufacturer's instructions. To identify the soma of the cells, each section was counterstained with NeuroTrace 435/455 Blue Fluorescent Nissl Stain (1:100, Invitrogen) and mounted using Fluoromont-G (Southern Biotech). Images were acquired at The Light Microscopy Facility of the Max Planck Florida Institute using an LSM 880 Zeiss confocal microscope. Image acquisition and processing were performed using ZEN 2011 software (Carl Zeiss) and setting the fluorescence intensity in nonsaturating conditions.
Immunogold electron microscopy.
Immunohistochemical reactions were performed using the preembedding immunogold method as described previously (Lujan et al., 1996). Briefly, after blocking with 10% serum for 1 h at room temperature, free-floating sections were incubated for 48 h with anti-RGS7 antibodies (1–2 mg/ml). Sections were washed and incubated for 3 h with goat anti-rabbit IgG coupled to 1.4 nm gold (Nanoprobes) at a 1:100 dilution. Sections were washed, postfixed in 1% glutaraldehyde, and processed for silver enhancement of the gold particles with an HQ Silver kit (Nanoprobes). The reacted sections were treated with osmium tetroxide (1% in 0.1 m PB), block stained with uranyl acetate, dehydrated in graded series of ethanol, and flat embedded on glass slides in Durcupan (Fluka) resin. Regions of interest were cut at 70–90 nm on an ultramicrotome (Reichert Ultracut E; Leica). Staining was performed on drops of 1% aqueous uranyl acetate followed by Reynolds's lead citrate. Ultrastructural analyses were performed on a Jeol-1010 electron microscope.
To establish the relative the abundance of RGS7 immunoreactivity along the plasma membrane of pyramidal cells, we used 60 μm coronal slices processed for preembedding immunogold immunohistochemistry. The procedure was similar to that used previously (Lujan et al., 1996). Briefly, for each of three animals from different postnatal ages and adult, three samples of tissue were obtained for preparation of embedding blocks (totaling nine blocks for each age). To minimize false negatives, electron microscopic serial ultrathin sections were cut close to the surface of each block because immunoreactivity decreased with depth. We estimated the quality of immunolabeling by always selecting areas with optimal gold labeling at approximately the same distance from the cutting surface. Randomly selected areas were then photographed from the selected ultrathin sections and printed with a final magnification of 45,000×. Quantification of immunogold labeling was performed in reference areas totaling ∼1800 μm2 for each age. Immunoparticles identified in each reference area and present in different subcellular compartments (dendritic spines, dendritic shafts, and somata) were counted. We measured the radial distance of each immunoparticle to the plasma membrane, being 0 for those just located in the plasma membrane. The data are expressed as thepercentage of immunoparticles along the radial distance from the plasma membrane expressed in nanometers.
Hippocampal cultures.
Primary cultures of hippocampal neurons were prepared using a modified version of a previously published protocol (Xie et al., 2010). Briefly, hippocampi were extracted from neonatal (P1–P3) pups and placed into an ice-cold HBSS/FBS solution (Sigma-Aldrich) containing 4.2 mm NaHCO3, 1 mm HEPES, and 20% FBS. The tissue was washed twice with 20% FBS and then three times with HBSS. Hippocampi were digested at room temperature for 5 min with 10 mg/ml trypsin type XI (Sigma-Aldrich) in a solution containing the following (in mm): 137 NaCl, 5 KCl, 7 Na2HPO4, and 25 HEPES, pH 7.2. The tissue was washed yhree times with 20% FBS and HBSS and then hippocampi were mechanically dissociated in HBSS supplemented with 12 mm MgSO4 using Pasteur pipettes of decreasing diameter. The neurons were pelleted by centrifugation (600 × g for 10 min at 4°C) and plated onto 8 mm glass coverslips pretreated with Matrigel (BD Biosciences) in a 48-well plate. Neurons were allowed to adhere for 30 min before adding 0.3 ml of prewarmed culture medium consisting of Neurobasal A (Life Technologies), 2 mm GlutaMAX-I (Life Technologies), 2% B-27 supplement, and 5% FBS. After 4–12 h, the culture medium was completely replaced with the same medium without FBS. Neurons were incubated at 37°C/5% CO2 and half of the medium was replaced with fresh medium every 2–4 d of culture. Neurons were cultured for 10–14 d before experiments.
Somatodendritic GIRK current recordings.
For GIRK currents, coverslips containing neurons at DIV 10–13 were transferred to a chamber containing a low-K+ bath solution containing the following (in mm): 145 NaCl, 4 KCl, 1.8 CaCl2, 1 MgCl2, 5.5 d-glucose, and 5 HEPES, pH 7.4 with NaOH. Borosilicate patch pipettes (2.5–5 MΩ) were filled with the following (in mm): 130 KCl, 10 NaCl, 1 EGTA, 0.5 MgCl2, 10 HEPES, pH 7.25 with KOH, 2 Na2ATP, 5 phosphocreatine, 0.3 GTP. Baclofen (R-(+)-b-(aminomethyl)-4-chlorobenzenepropanoic acid hydrochloride) was purchased from Sigma-Aldrich. Baclofen-induced currents were measured at room temperature using a high-K+ bath solution containing the following (in mm): 120 NaCl, 25 KCl, 1.8 CaCl2, 1 MgCl2, 5.5 d-glucose, and 5 HEPES, pH 7.4 with NaOH. For Ba+2 current recordings, the external solution contained the following (in mm): 138 NaCl, 4 KCl, 2.5 BaCl2, 1 MgCl2, 10 d-glucose, and 10 HEPES, pH 7.4 with NaOH. The internal solution contained the following (in mm): 100 CsCl, 20 TEA-Cl, 10 EGTA, 10 HEPES, 0.5 CaCl2, 1 MgCl2, 3 MgATP, and 0.3 Na3GTP, pH 7.25 with KOH.
The solution (±baclofen) was applied directly to the soma and proximal dendrites with an SF-77B rapid perfusion system (Warner Instruments). The holding potential was −80 mV. Membrane potentials and whole-cell currents were measured in large neurons (>75 pF) with hardware (Axopatch-700B amplifier, Digidata 1440A) and software (pCLAMP v. 10.3) from Molecular Devices. All currents were low-pass filtered at 2 kHz, sampled at 5 (GIRK) or 50 (Ba+2 current) kHz, and stored on computer hard disk for subsequent analysis. For GIRK, activation rates were extracted from a standard exponential fit of the current trace corresponding to the onset of drug effect and the peak evoked current and deactivation rates were extracted from an exponential fit of the trace corresponding to the return of current to baseline following removal of drug (Clampfit version 10.3 software). Current desensitization was defined as percentage change in steady-state current from the maximal baclofen-evoked response amplitude during 10 s of continuous drug application. Only experiments where access resistances (Ras) were stable and low (<20 MΩ) were included in the analysis. In experiments with Ba2+ currents, Ras were compensated at 50–90% rate.
Hippocampal slices.
Mice were euthanized under isoflurane anesthesia and brains were rapidly removed and placed in ice-cold artificial CSF (aCSF) supplemented with kynurenic acid 10 mm containing the following (in mm): 124 NaCl, 3 KCl, 24 NaHCO3, 1.25 NaH2PO4, 1 MgSO4, and 10 d-glucose equilibrated with 95% O2 and 5% CO2. The tissue was cut in 300-μm-thick sections with a vibrating microtome (Leica VT1200S). The slices were warmed to 35°C for 25–45 min in aCSF supplemented with 2 mm CaCl2 and equilibrated with 95% O2 and 5% CO2. Then slices were maintained in gassed aCSF at room temperature until being transferred to submerged-type recording chambers of volume ∼1.5 ml. The slices were constantly superfused (1–2 ml/min) with warmed (30–31°C) and gassed aCSF. All measurements were performed by an experimenter blinded to genotype.
Patch-clamp recordings in slices.
CA1 neurons were visually identified in the hippocampal transverse slices of 300 μm thickness using the Scientifica SliceScope system. Glass microelectrodes with an open tip resistance of 2.5–5 MΩ were used. The internal solution contained the following (in mm): 120 K-gluconate, 20 KCl, 10 K-HEPES, 0.2 EGTA, 2 MgCl2, 0.3 Na3GTP, and 4 Na2ATP, pH 7.3 with KOH. Cells with series resistance >20 MΩ or resting membrane potentials > −55 mV were excluded from analysis. Liquid junction potential was −14 mV. Kynurenic acid 2 mm was added to aCSF to block glutamatergic transmission.
Data analysis.
Statistical analyses were performed using Prism (GraphPad Software). Data are presented throughout as the mean ± SEM. Student's t test, one-way or two-way ANOVA, followed by Bonferroni's or Tukey's post hoc tests were used as appropriate. The minimal level of significance was set at p < 0.05.
Results
R7BP and GPR158 each control a significant fraction of RGS7 in the hippocampus
We began by studying the expression of GPR158 and R7BP subunits in the mouse hippocampus. Using sensitive in situ hybridization at a single-cell resolution, we found coexpression of R7bp and Gpr158 mRNAs in the majority of hippocampal neurons across all regions including CA1 pyramidal neurons (Fig. 1A,B). A negative control probe did not show any fluorescent labeling, indicating the specificity of the detection (Fig. 1C). To confirm the complex formation between RGS7 and its membrane anchoring subunits R7BP and GPR158 in hippocampal tissue, we performed coimmunoprecipitation experiments. Using this approach, we readily detected complexes of RGS7 with both R7BP (Fig. 1D) and GPR158 (Fig. 1E). The binding specificity was confirmed by using hippocampal lysates obtained from KO animals in which coimmunoprecipitation was not observed (Fig. 1D,E). To determine the relative contribution of R7BP and GPR158 to controlling RGS7, we compared the expression of RGS7 directly in hippocampi of R7BP and GPR158 KOs (Fig. 2A,B). We further generated R7BP and GPR158 DKO line to evaluate possible redundancy and interdependence of the respective adaptor subunits. The DKO mice were viable and did not show overt issues with development. Consistent with previous observations (Orlandi et al., 2015), deletion of GPR158 significantly reduced RGS7 content (to 66 ± 7% of WT levels). The effect of R7BP ablation was much smaller (to 86 ± 3%) and did not reach our criteria for significance (Fig. 2A,B). Interestingly, concurrent deletion of both R7BP and GPR158 resulted in a stronger reduction in RGS7 levels to 50 ± 9% relative to WT (Fig. 2A,B). We did not detect any significant differences in the expression of other components of the RGS7 complex and GABAB signaling pathway: Gαo, Gβ5, GABABR2, GIRK1, and GIRK2 (Fig. 2A,B), indicating selective effects of adaptor ablation on RGS7 complex stability.
We next examined the relative contributions of R7BP and GPR158 to membrane targeting of RGS7 by performing subcellular fractionation of hippocampal tissues from individual KOs compared with DKO. We found that both R7BP and GPR158 each significantly contributed to membrane association of RGS7 as judged by the reduction in RGS7 membrane content in respective KOs (Fig. 2C). Quantification of absolute content of RGS7 on the membrane from Western blotting data calibrated against recombinant RGS7 protein standards spiked into RGS7 KO samples revealed that the effect of GPR158 ablation was greater than that of R7BP ablation: reducing content of membrane RGS7 to 83 ± 2% and 50 ± 4%, respectively (Fig. 2D,E). Interestingly, only barely detectable 9 ± 4% of RGS7 remained on the membranes from DKO hippocampi (Fig. 2C–E). These results suggest that GPR158 and R7BP together account for the vast majority of membrane targeting of RGS7 in hippocampal neurons, with GPR158 providing a greater contribution.
We further confirmed these observations while examining precise subcellular distribution of RGS7 in hippocampal pyramidal neurons using an electron microscopy immunogold labeling technique (Fig. 3). Consistent with a previous report (Fajardo-Serrano et al., 2013), we found RGS7 to be positioned mostly on the plasma membrane across major neuronal compartments: dendritic shafts and spines as well as axonal terminals (Fig. 3A–C, G–I). In contrast, in DKO, most of RGS7 immunolabeling was confined to intracellular sites, in good quantitative agreement with our biochemical fractionation data (Fig. 3D–I).
RGS7 is the sole contributor to regulation of sIPSC kinetics by Gβ5-containing GAP complexes
To begin comparing the contributions of auxiliary subunits to RGS7-mediated regulation of GABAB-GIRK signaling, we first defined the role of RGS7 in controlling inhibitory signaling in CA1 pyramidal neurons. We previously found that KO of Gβ5 subunit, which eliminates the expression of four RGS complexes containing RGS6, RGS7, RGS9, and RGS11 dramatically slows down the recovery phase of IPSCs (slow IPSCs, sIPSCs) (Xie et al., 2010), a major form of GABAB-mediated inhibitory synaptic inputs onto CA1 neurons (Lüscher et al., 1997). To determine how much of this effect is mediated by RGS7, we studied sIPSC kinetics evoked in hippocampal slices comparing RGS7 KO side by side with WT littermates and Gβ5 KO mice.
Stimulation of interneuron projections in stratum lacunosum moleculare (Fig. 4A) elicited outward currents of similar amplitudes in CA1 neurons of Gβ5 KO, RGS7 KO and WT slices (Fig. 4B,C). We confirmed that the sIPSC currents were largely mediated by GIRK because treatment with tertiapin Q abolished these synaptically evoked events (data not shown). We observed a drastic slowing of both activation and decay kinetics of sIPSC in RGS7 KO. The response reached the peak significantly later in RGS7 KO compared with WT, reflected in an increase in time-to-peak (Fig. 4B,D). Furthermore, there was an ∼5-fold increase in decay constant in slices lacking RGS7 (Fig. 4B,E). Importantly, the RGS7 KO phenotype was quantitatively indistinguishable from Gβ5 KO when results were analyzed in parallel (Fig. 4B–E). These results reveal that RGS7 is the sole physiological regulator of GABAB-GIRK signaling that drives sIPSC among the members of Gβ5-containing R7 family RGS complexes.
R7BP, but not GPR158, regulates sIPSC kinetics
Given the indispensable and large contribution of RGS7 to the regulation of sIPSC kinetics, we next compared the effects of genetic loss of its adaptor subunits, R7BP and GPR158. We found that sIPSCs in R7BP KO slices showed significantly decelerated kinetics of both activation as evidenced by increased peak latency and deactivation as reflected by the increased deactivation time constant (Fig. 5A,C,D). This effect, however, was smaller than that seen in Gβ5 KO or RGS7 KO, suggesting that R7BP is responsible for regulating only a fraction of RGS7 actions in this process. Strikingly, we observed no differences in sIPSC kinetics between WT and GPR158 KO neurons. We also did not find any differences in the response amplitudes across all genotypes. Because compensation by R7BP may mask a possible effect of GPR158, we further analyzed the responses in the DKO mice compared with R7BP KO littermates (Fig. 5B–E). We did not detect appreciable differences in any response parameters between R7BP KO and DKO, suggesting that GPR158 does not shape sIPSCs even in the absence of regulatory R7BP influence.
GPR158 does not regulate GABABR-GIRK kinetics
We have previously demonstrated that RGS7 and R7BP regulate the kinetics of GIRK channel deactivation in primary hippocampal pyramidal neurons (Ostrovskaya et al., 2014). Therefore, we next studied the effects of GPR158 ablation on GIRK-mediated responses driven by GABABR activation also in this system, which offers greater sensitivity and mechanistic precision. We further compared GIRK properties measured in neurons of GPR158 KO mice with those from R7BP KO and DKO evaluated in parallel. A saturating concentration of GABAB agonist baclofen evoked GIRK currents with slowed deactivation in R7BP KO. However, the response recorded in GPR158 KO neurons was indistinguishable from WT (Fig. 6A,B). Furthermore, GIRK current deactivation kinetics in DKO neurons was slower than in WT, but did not differ from that in R7BP KO (Fig. 6A,B). Other parameters of baclofen-evoked currents, amplitudes, activation, and desensitization rates, were similar in all genotypes (Fig. 6C–E). These findings suggest that GPR158 does not contribute to GABAB-GIRK signaling in hippocampal pyramidal neurons.
Recovery of Ca2+ channels from GABABR-mediated inhibition is facilitated by RGS7/R7BP complexes
Because the regulation of N/P/Q types of Ca2+ channels constitutes the second major branch of GABAB signaling, we next probed the contribution of RGS7 complex to this process, which has not been defined before. We used primary hippocampal neuron system to record voltage-gated Ba2+ currents in RGS7 KO, R7BP KO, GPR158 KO, and WT. Depolarizing voltage steps from holding potential of −70 mV elicited family of currents characteristic of the Ca2+ channels, with no difference in current density and voltage dependence between the genotypes (Fig. 7A,B). Next, we assessed the magnitude and speed of baclofen-mediated block onset and relief upon rapid agonist application and removal. To increase the frequency of steps from −70 to −10 mV, we shortened the pulse duration to 12 ms and recorded the sweeps every 2 s (Fig. 7C), an approach reported previously (Greif et al., 2000). The amplitude of the current was taken 10–12 ms after the onset of the voltage step to −10 mV and was within 90–100% of the peak current estimated compared with currents recorded for the I–V relationship. Application of 100 μm baclofen caused a pronounced decrease in current amplitude with no significant differences between genotypes (64.8 ± 3.4% in WT vs 64.2 ± 3.3% in RGS7 KO, 63.3 ± 4.1% in R7BP KO, 69.7 ± 3.8% in GPR158 KO) (Fig. 7C,D). The densities of the current portions blocked by baclofen were also not different among the genotypes (Fig. 7E). These results show that RGS7, R7BP, and GPR158 do not affect the CaV abundance on the membrane and the magnitude of its inhibition by baclofen. To study the kinetic aspects of channel modulation by GABABRs, we used a fast perfusion system to apply and wash out baclofen (Fig. 7F). Although the timing of the baclofen-mediated inhibition onset could not be accurately quantified due to its fast speed, we observed that RGS7 KO neurons showed a significant delay in the current return to control levels after baclofen washout (Fig. 7F–H). The half-times (T1/2) of recovery were 2.4 ± 0.2 and 5.0 ± 0.8 s in WT versus RGS7 KO, respectively (Fig. 7H). Having established that RGS7 controls the timing of CaV inhibition by GABAB, we next evaluated the contributions of its subunits GPR158 and R7BP to this process in neurons obtained from GPR158 KO and R7BP KO mice. We found a significant delay in recovery of CaV currents from GABAB-mediated inhibition in R7BP KO (T1/2 of 3.8 ± 0.4 s) compared with WT neurons (Fig. 7I–K). This effect was smaller compared with RGS7 deletion, indicating that, as with GIRK regulation, R7BP controls only a fraction of RGS7 activity on the channel. In contrast, we observed no difference between WT and GPR158 KO (Fig. 7K), indicating that GPR158 does not affect the timing of GABAB-mediated regulation of CaV channels.
GPR158 opposes the function of R7BP in accelerating GABABR-GIRK kinetics
The lack of GPR158 effect on modulation of GABAB signaling to CaV and GIRK channels suggests that RGS7 complexed with it may be excluded from participating in this process. To test this possibility directly, we studied the effect of GPR158 overexpression on GABAB signaling through GIRK and CaV channels. Infecting primary neurons with adenovirus carrying GPR158 (Fig. 8A) resulted in elevation of GPR158 protein levels detectable by both Western blotting and immunohistochemistry (Fig. 8B,C). The concurrent EGFP expression encoded by the same vector was used to identify positively transduced neurons. In control experiments, cultures were infected with adenovirus carrying EGFP only (Fig. 8B,C). Electrophysiological recordings of GIRK currents from fluorescent neurons infected with AV-GPR158 showed a significant increase in deactivation rate of the response compared with those infected with empty AV-EGFP virus (Fig. 8D,E). Accordingly, quantification revealed a prominent increase in τ deactivation from 1226 ± 104 ms in control neurons to 2240 ± 202 ms in neurons overexpressing GPR158 (Fig. 8E).
Similar observations were made when studying modulation of CaV currents, in which we measured the effect of GPR158 overexpression on the timing of relief from baclofen inhibition (Fig. 8F–J). Again, we detected no differences in the density of total IBa currents across voltage steps (Fig. 8F) or the fraction blocked by baclofen (Fig. 8G), indicating unaffected expression and targeting of CaV to the plasma membrane. In contrast, we observed significantly slower current recovery in neurons infected with AV-GPR158, in which T1/2 was 3.1 ± 0.1 s compared with 2.3 ± 0.2 s in neurons infected with AV-EGFP control virus (Fig. 8H–J). Together, these findings indicate that GPR158 overexpression negatively affects the ability of RGS7 to regulate the kinetics of GABAB signaling to GIRK and CaV ion channels without affecting CaV expression and localization.
Discussion
The key observation reported in this study is that the macromolecular composition of major neuronal G-protein regulator, the RGS7 complex, determines its ability to influence downstream effectors (Fig. 9). Specifically, our findings indicate that R7BP, but not GPR158, modulates RGS7 effect on GABABR regulation of ion channels, GIRK, and CaV2. Moreover, shifting the balance toward the formation of RGS7-GPR158 complexes results in the loss of the RGS7 influence on these signaling pathways. Along with this central message, novel findings reported in this study include: (1) RGS7 is the sole member of R7 family that regulates GABABR signaling to GIRKs and sIPSCs in the hippocampal CA1 pyramidal neurons, (2) RGS7 also substantially affects the modulation of the CaV2 current by GABABRs, and (3) GABABR-mediated effects on ion channels are selectively mediated by R7BP-RGS7, but not GPR158-RGS7 complexes.
Previous studies have shown that regulation of GIRK channels by GABAB is substantially affected by members of the R7 family of RGS proteins (RGS6, RGS7, RGS9, and RGS11) that exist as constitutive complexes with the central scaffolding subunit Gβ5 (Xie et al., 2010; Maity et al., 2012; Ostrovskaya et al., 2014). KO of Gβ5 that eliminates all R7 RGS proteins drastically increases the sensitivity of GIRK modulation by GABAB and slows down its deactivation kinetics. In hippocampal neurons, KO of RGS7, but not RGS6, resulted in a similar phenotype (Ostrovskaya et al., 2014), whereas the contributions of other R7 RGS proteins remained unclear. By direct side-by-side comparison of Gβ5 and RGS7 KOs, measuring GIRK kinetics in primary neurons and GIRK-mediated sIPSCs in brain slices, we documented complete phenotypic equivalency of Gβ5 and RGS7 elimination, thus establishing RGS7 as the sole R7 RGS involved in the regulation of GABAB-GIRK signaling. We next addressed the contribution of membrane targeting of RGS7 complex to its effects. Because RGS7 is a soluble protein, its ability to regulate membrane delimited GABAB-GIRK was thought to require the action of membrane-targeting adaptors. Accordingly, elimination of its membrane anchor R7BP resulted in similar phenotype seen upon RGS7 elimination, although to a significantly lesser extent (Zhou et al., 2012; Ostrovskaya et al., 2014). Interestingly, biochemical analysis showed that elimination of R7BP brought about only a moderate decrease (∼25%) in membrane content of RGS7 (Jayaraman et al., 2009; Panicker et al., 2010), prompting a hypothesis that another membrane adaptor compensates for R7BP loss. Indeed, such protein was subsequently identified to be GPR158 and its ablation alone eliminated a greater fraction (∼50%) of RGS7 on the membrane (Orlandi et al., 2012, 2015), making it an ideal candidate for a missing redundant subunit functionally compensating in GABAB-GIRK regulation when R7BP is lost. While formally testing this hypothesis, the current study resulted in a surprising conclusion: not only does GPR158 not functionally compensate for R7BP, it counteracts the effects of R7BP on GIRK. Quantitative biochemical experiments revealed that deletion of both R7BP and GPR158 synergistically removed the majority of RGS7 from the plasma membrane, indicating that GPR158 and R7BP compete for RGS7 binding and control distinct pools of it. We further confirmed that RGS7 indeed forms complexes with both GPR158 and R7BP by direct immunoprecipitation experiments. We think that these observations suggest that RGS7 may exist in two alternative configurations on the plasma membrane and, although the complex with R7BP plays a permissive role in ion channel regulation, GPR158 prevents the action of RGS7 on GIRK. This inhibition could be explained by an occlusion mechanism in which GPR158 association simply makes RGS7 unavailable for binding to R7BP that normally facilitates GIRK regulation (Fig. 9). Alternatively, we cannot rule out that GPR158 instead selectively targets RGS7 to other intracellular effectors.
Our observations further suggest that a very small fraction of RGS7 remaining on the plasma membrane without R7BP and GPR158 is capable of effective GIRK regulation. The degree of this regulation is revealed by comparing RGS7 KO with R7BP/GPR158 DKO, which suggested that ∼10% of “anchor-free” RGS7 is responsible for the majority of GIRK control exceeding the fraction regulated by RGS7-R7BP complex (revealed by comparing WT and R7BP KO). We think that the disproportionately large contribution of small RGS fraction remaining on the membrane in the absence of R7BP and GPR158 is likely explained by direct association of RGS7/Gβ5 with the GIRK (Xie et al., 2010), placing it in the immediate molecular vicinity of the channel and thus enhancing regulation of its target (Fig. 9).
Another major finding of this work is in the first-time implication of the RGS7 complex in the regulation of voltage-gated calcium channels of the P/Q and N types, CaV2.1/CaV2.2. The GABABRs on cell bodies and presynaptic terminals also control the extent of their synaptic signaling via inhibition of the P/Q/N CaV2 Ca2+ channels and neurotransmitter release by Gβγ liberated from PTX-sensitive Gαi/o subunits (Holz et al., 1986; Doze et al., 1995; Dolphin, 2003). The role of RGS complexes in this process is less clear, with only few RGS proteins across the entire nervous system implicated in regulation of CaV2-mediated calcium influx. For example, RGS2 and RGS3 disinhibit CaV2 channels and increase transmitter release (Han et al., 2006; Toro-Castillo et al., 2007), whereas RGS4 and RGS12 accelerate the time course of desensitization of norepinephrine-mediated current inhibition (Diversé-Pierluissi et al., 1999; Schiff et al., 2000). In this report, we reveal RGS7 to be a key player in regulation of CaV function in hippocampal CA1 neurons. By accelerating the relief of CaV2 current blockade upon termination of GABABR activation, RGS7 acts to lessen the effects of GABABR-mediated inhibition while increasing its temporal resolution. Because dynamic regulation of CaV2 by GABABRs is crucial for fine tuning synaptic signaling, this mechanism likely contributes to plasticity and learning (Xu et al., 2007; Jung et al., 2016; Nanou et al., 2016).
Strikingly, we found that, as with GIRK, regulation of CaV2 by RGS7 is facilitated by R7BP and inhibited by GPR158. Because both GIRK and CaV2 are controlled by GABABRs via the same mechanism, this outcome suggests that exclusion of RGS7 from regulation of Gβγ signaling to ion channels may be a general mode of GPR158 action. The exact mechanisms of this inhibitory influence of GPR158 are unclear at this point and will require further investigation, yet some speculations may be warranted. Because GPR158 can modulate G-protein signaling initiated by traditional GPCRs (e.g., μ-opioid receptor) when tested in a reconstituted system (Orlandi et al., 2012), it is possible that GPR158 may be involved in setting the selectivity of RGS7 actions toward certain GPCRs and/or effectors. Consistent with this idea, studies on the highly homologous orphan receptor GPR179, which likewise associates with RGS7, indicate that this protein complex imparts regulation of mGluR6 signaling to ion channel TRPM1 in retina ON-bipolar neurons (Ray et al., 2014). Remarkably, this regulation requires an assembly of an elaborate macromolecular complex that includes many components of the signaling cascade (Orlandi et al., 2013; Sarria et al., 2016). Identification of the receptors and effectors controlled by GPR158-RGS7, as well as possible additional elements required for such control, will undoubtedly be an exciting research direction to pursue. Alternatively, it appears possible that GPR158 recruits RGS7 for its own intrinsic needs as a signaling GPCR. A recent study has suggested that GPR158 may signal via heterotrimeric G-proteins when activated with its proposed endogenous ligand osteocalcin (Khrimian et al., 2017). RGS proteins have been noted to associate with several canonical GPCRs, which is thought to modify their signaling properties (Abramow-Newerly et al., 2006; Anderson et al., 2009). In this context, RGS7 may thus act on G-proteins activated by GPR158 and thereby be excluded from regulation of G-proteins activated by other GPCRs such as GABAB.
Overall, our findings support an emerging concept that R7 RGS proteins serve as an integral part of macromolecular signaling assemblies consisting of GPCRs, auxiliary subunits, and ion channels. We extend this model by illustrating that the composition of these complexes can be specifically tailored and that these changes in organization allow bidirectional functional tuning endowing GPCR cascades with the plasticity and high spatiotemporal precision needed for coordination of synaptic signaling.
Footnotes
This work was supported by the National Institutes of Health (Grants DA021743, DA026405, and MH105482 to K.A.M.), the Spanish Ministry of Education and Science (Grant BFU-2015-63769-R to R.L.), and the European Union (HBP Project Reference 720270 to R.L.). We thank Dr. Said Kourrich for technical advice and discussions and Mrs. Natalia Martemyanova for help with animal breeding and genotyping.
The authors declare no competing financial interests.
- Correspondence should be addressed to Kirill A. Martemyanov, Department of Neuroscience, The Scripps Research Institute, 130 Scripps Way #3C2, Jupiter, FL 33458. kirill{at}scripps.edu