Receptor-dependent influence of R7 RGS proteins on neuronal GIRK channel signaling dynamics

Abstract

Most neurons are influenced by multiple neuromodulatory inputs that converge on common effectors. Mechanisms that route these signals are key to selective neuromodulation but are poorly understood. G protein-gated inwardly rectifying K⁺ (GIRK or Kir3) channels mediate postsynaptic inhibition evoked by G protein-coupled receptors (GPCRs) that signal via inhibitory G proteins. GIRK-dependent signaling is modulated by Regulator of G protein Signaling proteins RGS6 and RGS7, but their selectivity for distinct GPCR-GIRK signaling pathways in defined neurons is unclear. We compared how RGS6 and RGS7 impact GIRK channel regulation by the GABA_B receptor (GABA_BR), 5HT_1A receptor (5HT_1AR), and A₁ adenosine receptor (A₁R) in hippocampal neurons. Our data show that RGS6 and RGS7 make non-redundant contributions to GABA_BR- and 5HT_1AR-GIRK signaling and compartmentalization and suggest that GPCR-G protein preferences and the substrate bias of RGS proteins, as well as receptor-dependent differences in Gα_o engagement and effector access, shape GPCR-GIRK signaling dynamics in hippocampal neurons.

INTRODUCTION

Many neurotransmitters and related drugs inhibit synaptic transmission by targeting G protein-coupled receptors (GPCRs) that signal via G proteins to regulate the activity of downstream effectors (Davies and Tomas, 2023; Jiang et al., 2022; Kumar and Puthenveedu, 2022). Dysregulation of inhibitory G protein-dependent signaling underlies many neurological, psychiatric, and neurodegenerative disorders (Wong et al., 2023). GPCRs evoke canonical biological responses by coupling to a limited array of heterotrimeric G proteins (Voss and Muller, 2022), which then regulate the activity of a relatively small pool of effectors (Davies and Tomas, 2023; Jiang et al., 2022; Kumar and Puthenveedu, 2022). However, regulatory factors and mechanisms that shape inhibitory GPCR-effector signaling dynamics and establish the functional compartmentalization of signaling pathways are poorly understood in endogenous settings, particularly in neurons.

In the brain, inhibitory GPCRs signal via the G_i/o subfamily of Gα subunits that includes 3 distinct Gα_i isoforms (Gα_i1, Gα_i2, Gα_i3), 2 alternative splice variants of Gα_o (Gα_oA, Gα_oB), and Gα_z (Voss and Muller, 2022). These Gα subunits are functionally intertwined with Gβγ complexes (Senarath et al., 2018). Regulator of G protein Signaling (RGS) proteins regulate G protein-dependent signaling pathways by enhancing the intrinsic GTP hydrolysis rate of Gα, accelerating the process of inactive heterotrimer (Gαβγ) re-formation (Sprang, 2016). RGS proteins differ in terms of their Gα substrate specificities and structural elements that confer distinct functionality (Abramow-Newerly et al., 2006; Willars, 2006). Members of the R7 sub-family of RGS proteins (RGS6, RGS7, RGS9, and RGS11) contain a “Gγ-like” (GGL) domain that recruits Gβ5, the atypical/non-signaling member of the Gβ subunit family (Anderson et al., 2009; Slepak, 2009). R7 RGS/Gβ5 complexes exhibit strong substrate selectivity for Gα_o (Hooks et al., 2003; Lan et al., 2000; Posner et al., 1999).

The embedding of Gβ5 within R7 RGS protein complexes prompted our early interest in the impact of these complexes on signaling involving a prototypical Gβγ-regulated effector – the G protein-gated inwardly rectifying K⁺ (GIRK or Kir3) channel (Xie et al., 2010). GIRK channels are found predominantly in the somato-dendritic compartment of glutamatergic pyramidal neurons (Koyrakh et al., 2005; Kulik et al., 2006) and GABAergic interneurons (Booker et al., 2013) in the rodent HPC. GIRK channels mediate the G protein-dependent, postsynaptic inhibitory effect evoked by activation of many GPCRs (Luo et al., 2022), including the GABA_B receptor (GABA_BR) (Koyrakh et al., 2005; Luscher et al., 1997). GIRK channels exhibit strong co-clustering with GABA_BR around glutamatergic synapses on dendritic spines of HPC pyramidal neurons (Kulik et al., 2006), where they are well-positioned to mediate the postsynaptic inhibitory effect of GABA_BR on glutamatergic neurotransmission (Otmakhova and Lisman, 2004). Interestingly, GIRK channels and GABA_BR exhibit less co-clustering in dendritic shafts (Kulik et al., 2006). RGS6 (Stewart et al., 2014) and RGS7 (Ostrovskaya et al., 2018) are also found in the somato-dendritic compartment of hippocampal (HPC) neurons. Ultrastructural analysis in the mouse HPC CA1 region revealed that RGS7 is found primarily in the plasma membrane of dendritic shafts and spines (Ostrovskaya et al., 2018).

Previously, we reported that ablation of Gβ5 slowed the deactivation of GABA_BR-induced GIRK currents and increased GIRK channel sensitivity to GABA_BR activation in cultured hippocampal (HPC) neurons, and prolonged synaptic GABA_BR-GIRK currents in mouse HPC CA1 pyramidal neurons (Xie et al., 2010). These effects were recapitulated by ablation of RGS7 (but not RGS6) (Ostrovskaya et al., 2014; Ostrovskaya et al., 2018), suggesting that RGS7/Gβ5 complexes negatively regulate GABA_BR-GIRK signaling in HPC neurons. GIRK channels also mediate the somato-dendritic inhibitory influence of other GPCRs in HPC neurons, including 5HT_1A (5HT_1AR) and A₁ adenosine (A₁R) receptors (Luscher et al., 1997). The amplitude and kinetics of GIRK channel responses in HPC neurons differs in a GPCR-dependent manner (Leaney, 2003). Whether and how RGS7 contributes to differences in GPCR-GIRK signaling in HPC neurons is unknown. In addition, while RGS6 ablation did not impact GABA_BR-GIRK signaling in HPC neurons (Ostrovskaya et al., 2014), RGS6 modulates the GABA_BR-dependent regulation of voltage-gated Ca²⁺ channels (Gao et al., 2020) and the 5HT_1AR-dependent regulation of cAMP dynamics in these neurons (Stewart et al., 2014), and it negatively regulates A₁R- and GABA_BR-GIRK signaling in other cell types (Anderson et al., 2020; DeBaker et al., 2023; Maity et al., 2012).

Here, we sought to understand how R7 RGS proteins shape the strength, sensitivity, and kinetics of GABA_BR-, A₁R-, and 5HT_1AR-GIRK signaling in mouse HPC neurons. Using mutant mouse lines and electrophysiological approaches, we found that RGS6 and RGS7 exert an array of non-redundant and receptor-dependent influences on GPCR-GIRK signaling dynamics.

MATERIALS AND METHODS

Animals.

All animal experiments were approved by the University of Minnesota Institutional Animal Care and Use Committee. Male and female C57BL/6J mice were purchased from The Jackson Laboratory (#000664; Bar Harbor, ME) and used to establish an in-house breeding colony. Generation of Girk2^−/− (Signorini et al., 1997), Rgs6^−/− (Posokhova et al., 2010), Rgs7^−/− (Cao et al., 2012), and Gα_o^fl/fl (Ang et al., 2016) mice has been described previously. Genotypes were determined using validated PCR-based genotyping protocols. All mice were group-housed with no more than 4 males or 5 females in a single cage. Mice were maintained on a 14:10 h light/dark cycle and were provided ad libitum access to food and water.

Reagents.

Baclofen and all cell culture reagents were purchased from Sigma-Aldrich (Saint Louis, MO), and serotonin hydrochloride and adenosine were purchased from Tocris Bioscience (Bristol, UK). All drugs were stored, handled, and resuspended according to provider specifications. ML297 was a generous gift from Dr. C. David Weaver and was dissolved in DMSO on the day of the experiment. High-titer (>1 × 10¹² genocopies/mL) AAV8-CaMKIIα-GFP, as well as AAV8-CaMKIIα-Cre(mCherry) and control vector (AAV8-CaMKIIα-ΔCre(mCherry), were generated and purified by the University of Minnesota Viral Vector Cloning Core.

DNA constructs.

GABA_BR1 (NM_001470), GABA_BR2 (NM_005458), A₁R (AY136746), 5HT_1A serotonin receptor (AF498978), Gα_oB (AH002708), Gα_z (J03260), Gα₁₁ (AF493900), Gα₁₂ (NM_007353), Gα₁₃ (NM_006572), Gα₁₄ (NM_004297), Gα₁₅ (AF493904), Gαs long isoform (Gα_sL) (NM_000516), Gα_olf (AF493893), and Gβ_5S (NM_006578) in pcDNA3.1(+) were purchased from cDNA Resource Center (www.cDNA.org). The pCMV5 plasmids encoding rat Gα_oA, rat Gα_i1, rat Gα_i2, rat Gα_i3, human Gα_q, and bovine Gα_s short isoform (Gα_sS) were gifts from Dr. Hiroshi Itoh. Venus 156–239-Gβ₁ (amino acids 156–239 of Venus fused to a GGSGGG linker at the N terminus of Gβ₁ without the first methionine (NM_002074)) and Venus 1–155-Gγ₂ (amino acids 1–155 of Venus fused to a GGSGGG linker at the N terminus of Gγ₂ (NM_053064)) were gifts from Dr. Nevin A. Lambert. Flag-tagged Ric-8A (NM_053194) in pcDNA3.1 was a gift from Dr. Jean-Pierre Montmayeur. Flag-tagged Ric-8B (NM_183172 with one missense mutation (A1586G)) in pcDNA3.1 was a gift from Dr. Bettina Malnic. The masGRK3ct-Nluc-HA constructs were constructed by introducing HA tag at the C-terminus of masGRK3ct-Nluc, as described (Masuho et al., 2015b; Masuho et al., 2021).

Primary cell culture.

Primary HPC neuron cultures were prepared from neonatal (P0–2) C57BL/6J pups, as described(Wydeven et al., 2012). Briefly, hippocampi were extracted and placed into ice-cold modified Hank’s Balanced Salt Solution (HBSS, Ca²⁺ and Mg²⁺ free, with 1 mM HEPES) containing 20% FBS, rinsed twice with FBS-free HBSS, digested for 20 min at 37°C with occasional inversion using papain (2.5% v/v) and DNase I (0.1% v/v) in digestion solution (137 mM NaCl, 0.5 mM KCl, 0.7 mM Na₂PO₄, 2.5 mM HEPES, pH 7.2). Tissue was then mechanically dissociated by pipetting in Neurobasal A-based plating medium (with 1x B27, 1x Glutamax, 1x antibiotic/antimycotic, and 0.05% DNase I). Cells were pelleted by centrifugation (2500 rpm for 10 min at room temperature). Cells were diluted with plating media accordingly and plated onto poly-L-Lysine (0.0005%) pre-coated 8-mm glass coverslips in 48-well plates for single-cell electrophysiological analysis. Before experimentation, cultures were maintained in a humidified 5% CO₂ incubator at 37°C for 11–14 d, and half of the medium was replaced with fresh growth medium (Neurobasal A with 1x B27, Glutamax, and antibiotic/antimycotic) every 3 d. Viral infections were performed on the day of the first (AAV8-CaMKIIα-Cre(mCherry) or ΔCre control) or third (AAV8-CaMKIIα-GFP) media change.

Electrophysiology.

Whole-cell patch-clamp recordings in cultured HPC neurons were performed as described (Wydeven et al., 2014). In brief, coverslips with neurons were transferred to a chamber containing a low K⁺ bath solution (130 mM NaCl, 5.4 mM KCl, 1 mM CaCl₂, 1 mM MgCl₂, 5.5 mM D-Glucose, 5 mM HEPES, pH 7.4). GFP-positive neurons with pyramidal morphology and capacitance values of 100–200 pF were targeted for analysis. Fire-polished borosilicate patch pipettes (4–6 MΩ) were filled with K-Gluconate internal solution (140 mM K-Gluconate, 2 mM MgCl₂, 1.1 mM EGTA, 5 mM HEPES, 2 mM Na₂-ATP, 0.3 mM Na-GTP, and 5 mM phosphocreatine, pH 7.2). After achieving whole-cell access, neurons were held at −70 mV; liquid-junction potential was not corrected. Whole-cell currents were measured in a high-K⁺ bath solution (120 mM NaCl, 25 mM KCl, 1 mM CaCl₂, 1 mM MgCl₂, 5.5 mM D-Glucose, 5 mM HEPES, pH 7.4). GPCR agonists and ML297 were diluted in the high-K⁺ bath solution and perfused directly onto the neuron using the ValveLink 8.2 rapid perfusion system (AutoMate Scientific; Berkeley, CA). Whole-cell currents were acquired with an Axopatch-200B amplifier and pCLAMP v.8.2 software (Molecular Devices, LLC; San Jose, CA). All currents were low-pass filtered at 2 kHz, digitized at 10 kHz with a Digidata 1322A (Molecular Devices), and stored on a computer hard disk for subsequent analysis. Data for each separate study were obtained from 3–4 different cultures per group. Only experiments in which the access resistance (Ra) was stable (change pre- and post-perfusion <20%) and low (<15 MΩ) were included in the analysis. Current density (pA/pF) was calculated as the ratio of steady-state current amplitude to cell capacitance. Activation (baseline to peak upon agonist addition) and deactivation (steady-state to baseline upon agonist removal) time constants were extracted from appropriate regions of stable current traces and were fit with a 1-term Boltzmann equation. The constants were included in the analysis when a good curve fit (R² ≥ 0.99) was obtained. Acute desensitization of the GIRK channel response was expressed as a percentage (desensitization %) and calculated as follows: 100 × [(peak current - steady-state current)/peak current].

Bioluminescence resonance energy transfer (BRET) assay.

GPCR-G protein preferences were evaluated using a fast-kinetic BRET assay, as described (Anderson et al., 2020; Masuho et al., 2015a). In brief, HEK293T/17 cells were grown in Dulbecco’s Modified Eagle Media supplemented with 10% fetal bovine serum, minimum Eagle’s medium nonessential amino acids, 1 mM sodium pyruvate, and antibiotics (100 units/mL penicillin and 100 mg/mL streptomycin) at 37°C in a humidified incubator containing 5% CO₂. Culture dishes (3.5 cm) were incubated for 10 min at 37°C with 1 mL of 10 mg/mL growth factor-reduced Matrigel (BD Biosciences, Franklin Lakes, NJ) in culture medium. For transfection, cells were seeded into 3.5-cm dishes at a density of 2 × 10⁶ cells/dish. After 2 h, expression constructs (total 5 μg/dish) were added to the cells using PLUS (5 μg/dish) and Lipofectamine LTX (6 μL/dish) reagents. The GPCR (GABA_BR1 (1), GABA_BR2 (1), A₁R (1), and 5HT_1AR receptor (1)), Gα (Ga_oA (2), Gα_oB (1), Gα_i1 (1), Gα_i2 (2), Gα_i3 (1.5), Gα_z (1.5), Gα_q (2), Gα₁₁ (2), Gα₁₄ (4), Gα₁₅ (2), Gα_s-short (6), Gα_s-long (4), Gα_olf (6), Gα₁₂ (3), or Gα₁₃ (4)), Venus 156–239-Gβ₁ (1), Venus 1–155-Gγ₂ (1), and BRET donor (masGRK3ct-Nluc-HA (1)) were transfected; the number in parentheses indicates the ratio of transfected DNA (ratio 1 = 0.21 μg). Gα_14/15 and Gα_olf were transfected with Ric-8A (1) and Ric-8B (1), respectively. A construct carrying catalytic subunit of pertussis toxin (PTX-S1) were transfected with Gα_z, Gα_s subfamily, Gα_q subfamily, and Gα_12/13 subfamily to inhibit the possible coupling of endogenous Gα_i/o to GPCRs.

An empty vector (pcDNA3.1(+)) was used to normalize the amount of transfected DNA. After transfection (16–24 h), HEK293T/17 cells were washed once with BRET buffer (Dulbecco’s Phosphate-Buffered Saline containing 0.5 mM MgCl₂ and 0.1% glucose) and detached by gentle pipetting over the monolayer. Cells were harvested with centrifugation at 500xg for 5 min and resuspended in BRET buffer. Approximately 50,000 to 100,000 cells per well were distributed in 96-well flat-bottomed white microplates (Greiner Bio-One, Kremsmünster, Austria). The Nluc substrate furimazine was purchased from Promega (Madison, WI) and used according to manufacturer specifications. BRET measurements were made using a microplate reader (POLARstar Omega; BMG Labtech, Offenburg, Germany) equipped with two emission photomultiplier tubes, allowing for detection of two emissions simultaneously with a highest possible resolution of 20 ms per data point. All measurements were performed at room temperature. The BRET signal was determined by calculating the ratio of the light emitted by the Venus-Gβ₁γ₂ (535 nm with a 30-nm band path width) over the light emitted by the masGRK3ctNluc-HA (475 nm with a 30-nm band path width). The average baseline value (basal BRET ratio) recorded prior to agonist stimulation was subtracted from the experimental BRET signal values and obtained DBRET ratio. The time constants (tau) of the activation and deactivation phases were obtained by fitting a single exponential curve to the traces with Clampfit 10.3. BRET measurements of Venus-Gβ₁γ₂ and masGRK3ct-Nluc-HA were performed to measure G protein activation by GABA_BR, A₁R, and 5HT_1AR, as described(Anderson et al., 2009).

Statistical analyses.

Data are presented throughout as mean ± SEM. Outlier identification tests (ROUT, Q=1%) were run for all datasets to remove statistical outliers prior to analysis. Statistical analyses were conducted with Prism v10.0.2 (GraphPad Software; Boston, MA) using unpaired t-test (with or without Welch’s correction), Mann-Whitney test (when one or more groups of data did not pass the normality test), or 2-way ANOVA and post hoc comparisons (Šídák’s multiple comparisons or uncorrected Fisher’s LSD), as appropriate. In all cases, differences were considered significant if P<0.05.

RESULTS

R7 RGS proteins make non-redundant contributions to GPCR-GIRK signaling in HPC neurons

We generated primary HPC cultures from neonatal mice to study GPCR-GIRK signaling in neurons. Cultures were infected with AAV8-CaMKIIα-GFP to facilitate targeted analysis of excitatory neurons (Fig. 1A). The GABA_BR agonist baclofen evoked whole-cell inward currents in all cultured HPC neurons from control (Girk2^+/+) mice, but not in neurons from Girk2^−/− mice or in control neurons in the presence of the GABA_BR-selective antagonist CGP54626 (Fig. 1B,C). Similarly, serotonin and adenosine reliably evoked inward currents in neurons from control but not Girk2^−/− mice, or in control neurons in the presence of the 5HT_1AR-selective antagonist NAD299 or the A₁R-selective antagonist DPCPX, respectively (Fig. 1B,C). Thus, baclofen, serotonin, and adenosine activate GIRK channels in HPC neurons via GABA_BR, 5HT_1AR, and A₁R, respectively.

Figure 1. GIRK channels mediate GABA_BR-, 5HT_1AR-, and A₁R-induced currents in cultured mouse HPC neurons.

A. GFP expression in a primary mouse HPC culture, 4-d after infection with AAV8-CaMKIIα-GFP; scale: 100 μm. GFP-labeled neurons with pyramidal morphology were targeted for electrophysiological analysis.

B. Whole-cell currents evoked by baclofen (100 μM), serotonin (10 μM), and adenosine (10 μM) in GFP-labeled HPC neurons, measured in a high-K⁺ bath solution at a holding potential of −70 mV. Inward currents were reliably evoked in HPC neurons from control (Girk2^+/+; black) but not Girk2^−/− (gray) mice or in Girk2^+/+ neurons treated with selective antagonists (purple) for GABA_BR (CGP54626; 1 μM), 5HT_1AR (NAD299; 0.01 μM), or A₁R (DPCPX; 0.1 μM); scale: 500 pA/5 s.

C. Summary of agonist-induced steady-state current densities (pA/pF). Baclofen/CGP54626− vs baclofen/CGP54626+ (unpaired t-test with Welch’s correction: t_3.029=16.53, ***P=0.0005; n=4/group); baclofen/Girk2^+/+ vs baclofen/Girk2^−/− (unpaired t-test with Welch’s correction: t_3.032=16.42, ***P=0.0005; n=4/group); serotonin/NAD299− vs serotonin/NAD299+ (unpaired t-test with Welch’s correction: t_4.022=9.060, ***P=0.0008; n=5/group); serotonin/Girk2^+/+ vs serotonin/Girk2^−/− (unpaired t-test with Welch’s correction: t_4.066=8.805, ***P=0.0009; n=5/group); adenosine/DPCPX− vs adenosine/DPCPX+ (unpaired t-test: t₆=39.79, ****P<0.0001; n=4/group); adenosine/Girk2^+/+ vs adenosine/Girk2^−/− (unpaired t-test with Welch’s correction: t_3.136=41.58, ****P<0.0001; n=4/group).

Data are represented as mean ± SEM.

The catalytic activity of RGS proteins accelerates the rate of Gα inactivation (Fig. 2A). Accordingly, we predicted that R7 RGS protein ablation would prolong G protein inactivation for GPCRs that signaled via Gα_o G proteins, resulting in a slowing of current deactivation (Fig. 2B,C). Indeed, this influence of R7 RGS proteins on GABA_BR-GIRK signaling in cultured HPC neurons was reported previously, along with a slowing of activation and a leftward shift in the baclofen concentration-response relationship (Ostrovskaya et al., 2014; Xie et al., 2010) (Fig. 2B,C). The latter findings may be due to the predicted increased ratio of active to inactive G protein in the absence of RGS7 and positive cooperativity associated with the Gβγ-dependent activation of GIRK channels (Wang et al., 2016). Some studies have shown increased amplitude of GPCR-GIRK currents in the absence of R7 RGS proteins (Anderson et al., 2018; Anderson et al., 2020; DeBaker et al., 2023); this influence of R7 RGS proteins may indicate that the level of activated G protein, rather than GIRK channel, limits the magnitude of the GPCR-GIRK response.

Figure 2. R7 RGS protein influence on GPCR-GIRK signaling.

A. R7 RGS proteins (RGS) enhance the intrinsic GTP hydrolysis rate of Gα_o, accelerating the process of inactive heterotrimer (Gα_oβγ) re-formation (deactivation). GIRK channels are activated by direct binding of Gβγ subunits following G_i/o protein activation.

B. In the absence of R7 RGS influence, the rate of Gα_o GTP hydrolysis will be reduced, leading to slowed deactivation of GIRK currents evoked by activation of GPCRs that couple to Gα_o. In some systems, R7 RGS ablation has been shown to prolong activation and enhance GPCR-GIRK current amplitude, likely due to a reduction in inactive G protein and Gβγ accumulation, respectively.

C. Predicted and known consequences of R7 RGS protein ablation (dashed line) on GPCR-GIRK signaling, including prolonged activation and deactivation, enhanced GIRK current amplitude, and increased coupling efficiency (leftward shift in the concentration-response relationship).

With these precedents and predictions in mind, we examined GIRK-dependent signaling in neurons from Rgs7^−/− mice. Using the direct GIRK channel activator ML297 (Kaufmann et al., 2013; Wydeven et al., 2014), we first established that loss of RGS7 did not alter the level of GIRK channel activity in HPC neurons (Supp. Fig. 1). RGS7 ablation did, however, impact GIRK-dependent signaling dynamics in a GPCR-dependent manner that deviated in some cases from our predictions. Consistent with our prior report (Ostrovskaya et al., 2014), loss of RGS7 did not impact the amplitude (Fig. 3A,B), desensitization (Fig. 3A,C), or activation rate (Fig. 3A,D) for GIRK channel activation by a saturating concentration of baclofen, but it did prolong current deactivation (Fig. 3A,E). In contrast, loss of RGS7 slowed the activation for 5HT_1AR-induced GIRK channel responses (Fig. 3F,I), while leaving other parameters intact (Fig. 3G,H,J). All measures of GIRK-dependent signaling associated with saturating concentration of adenosine were normal in HPC neurons from Rgs7^−/− mice (Fig. 3K–O).

Figure 3. Impact of RGS7 ablation on GABA_BR-, 5HT_1AR-, and A₁R-GIRK signaling and coupling sensitivity.

A. Representative currents evoked by baclofen (100 μM) in excitatory (GFP-labeled) HPC neurons from Rgs7^+/+ (black) and Rgs7^−/− (green) mice; scale: 500 pA/5 s.

B. Steady-state baclofen-induced current density in neurons from Rgs7^+/+ (n=11) and Rgs7^−/− (n=10) mice (unpaired t-test: t₁₉=0.1469, P=0.8847).

C. Desensitization of baclofen-induced currents in neurons from Rgs7^+/+ (n=11) and Rgs7^−/− (n=10) mice (Mann-Whitney test: U_112,119=46, P=0.5573).

D. Activation rates of baclofen-induced currents in neurons from Rgs7^+/+ (n=11) and Rgs7^−/− (n=8) mice (unpaired t-test: t₁₇=1.176, P=0.2559).

E. Deactivation rates of baclofen-induced currents in neurons from Rgs7^+/+ (n=11) and Rgs7^−/− (n=11) mice (Mann-Whitney test: U_73,180=7, ***P=0.0001).

F. Representative currents evoked by serotonin (10 μM) in excitatory (GFP-labeled) HPC neurons from Rgs7^+/+ (black) and Rgs7^−/− (green) mice; scale: 500 pA/5 s.

G. Steady-state serotonin-induced current density in neurons from Rgs7^+/+ (n=12) and Rgs7^−/− (n=12) mice (Mann-Whitney test: U_130.5,169.5=52.50, P=0.2716).

H. Desensitization of serotonin-induced currents in neurons from Rgs7^+/+ (n=12) and Rgs7^−/− (n=12) mice (Mann-Whitney test: U_174,126=48, P=0.1733).

I. Activation rates of serotonin-induced currents in neurons from Rgs7^+/+ (n=12) and Rgs7^−/− (n=12) mice (unpaired t-test: t₂₂=3.663, **P=0.0014).

J. Deactivation rates of serotonin-induced currents in neurons from Rgs7^+/+ (n=12) and Rgs7^−/− (n=12) mice (unpaired t-test: t₂₂=1.029, P=0.3146).

K. Representative currents evoked by adenosine (10 μM) in excitatory (GFP-labeled) HPC neurons from Rgs7^+/+ (black) and Rgs7^−/− (green) mice; scale: 500 pA/5 s.

L. Steady-state adenosine-induced current density in neurons from Rgs7^+/+ (n=10) and Rgs7^−/− (n=10) mice (unpaired t-test: t₁₈=0.07027, P=0.9448).

M. Desensitization of adenosine-induced currents in neurons from Rgs7^+/+ (n=10) and Rgs7^−/− (n=10) mice (unpaired t-test: t₁₈=0.8939, P=0.3832).

N. Activation rates of adenosine-induced currents in neurons from Rgs7^+/+ (n=10) and Rgs7^−/− (n=8) mice (unpaired t-test: t₁₆=0.6995, P=0.4943).

O. Deactivation rates of adenosine-induced currents in neurons from Rgs7^+/+ (n=10) and Rgs7^−/− (n=10) neurons (unpaired t-test: t₁₈=0.2905, P=0.7748).

P. Representative concentration-response experiments for baclofen-induced currents in excitatory (GFP-labeled) HPC neurons from Rgs7^+/+ (black) and Rgs7^−/− (green) mice; scale: 500 pA/5 s.

Q. Concentration-response curves for baclofen-induced GIRK currents in neurons from Rgs7^+/+ and Rgs7^−/− mice.

R. Normalized concentration-response curves for baclofen-induced GIRK currents in neurons from Rgs7^+/+ and Rgs7^−/− mice.

S. Log(EC₅₀) for baclofen-induced GIRK currents in neurons from Rgs7^+/+ (n=11) and Rgs7^−/− (n=12) mice (unpaired t-test: t₂₁=2.271, *P=0.0338).

T. Representative concentration-response experiments for serotonin-induced currents in excitatory (GFP-labeled) HPC neurons from Rgs7^+/+ (black) and Rgs7^−/− (green) mice; scale: 500 pA/5s.

U. Concentration-response curves for serotonin-induced GIRK currents in neurons from Rgs7^+/+ and Rgs7^−/− mice.

V. Normalized concentration-response curves for 5HT-induced GIRK currents in neurons from Rgs7^+/+ and Rgs7^−/− mice.

W. Log(EC₅₀) for serotonin-induced GIRK currents in neurons from Rgs7^+/+ (n=11) and Rgs7^−/− (n=10) mice (unpaired t-test with Welch’s correction: t_12.92=2.533, *P=0.0251).

X. Representative concentration-response experiments for adenosine-induced currents in excitatory (GFP-labeled) HPC neurons from Rgs7^+/+ (black) and Rgs7^−/− (green) mice; scale: 500 pA,5 s.

Y. Concentration-response curves for adenosine-induced GIRK currents in neurons from Rgs7^+/+ and Rgs7^−/− mice.

Z. Normalized concentration-response curves for adenosine-induced GIRK currents in neurons from Rgs7^+/+ and Rgs7^−/− mice.

AA. Log(EC₅₀) for adenosine-induced GIRK currents in neurons from Rgs7^+/+ (n=10) and Rgs7^−/− (n=11) mice (unpaired t-test: t₁₉=1.527, P=0.1432).

Data are represented as mean ± SEM.

We also evaluated the coupling efficiency (EC₅₀) of the 3 inhibitory GPCRs to GIRK channels in HPC neurons from Rgs7^−/− mice. Consistent with our published report (Ostrovskaya et al., 2014), loss of RGS7 increased the sensitivity of GIRK channels to GABA_BR activation (Fig. 3P–S). In contrast, RGS7 ablation decreased GIRK channel sensitivity to 5HT_1AR activation (Fig. 3T–W), and it had no impact on A₁R-GIRK signaling sensitivity (Fig. 3X–AA). Thus, RGS7 impacts GPCR-GIRK coupling efficiency in HPC neurons in distinct fashion, exerting a negative influence on GABA_BR-GIRK signaling, positive influence on 5HT_1AR-GIRK signaling, and no influence on A₁R-GIRK signaling.

Also expressed in the HPC (Ostrovskaya et al., 2014; Stewart et al., 2014), RGS6 has been reported to modulate GABA_BR regulation of voltage-gated Ca²⁺ channels (Gao et al., 2020) and 5HT_1AR-dependent regulation of cAMP signaling (Stewart et al., 2014). Thus, we also examined GPCR-GIRK signaling in HPC neurons from Rgs6^−/− mice (Fig. 4). RGS6 ablation did not alter the level of GIRK channel activity, as assessed by measuring ML297-induced GIRK currents (Supp. Fig. 1B). Similarly, loss of RGS6 did not impact any measure related to GABA_BR-GIRK (Fig. 4A–E,P–S) or A₁R-GIRK signaling (Fig. 4K–O,X–AA). While most measures of 5HT_1AR-GIRK signaling were also unaffected by RGS6 ablation (Fig. 4F–I,T–W), we did observe a slowed deactivation of the 5HT_1AR-induced GIRK current (Fig. 4F,J).

Figure 4. Impact of RGS6 ablation on GABA_BR-, 5HT_1AR-, and A₁R-GIRK signaling and coupling sensitivity.

A. Representative currents evoked by baclofen (100 μM) in excitatory (GFP-labeled) HPC neurons from Rgs6^+/+ (black) and Rgs6^−/− (blue) mice; scale: 500 pA/5 s.

B. Steady-state baclofen-induced current density in neurons from Rgs6^+/+ (n=13) and Rgs6^−/− (n=18) mice (unpaired t-test: t₂₉=0.8133, P=0.4227).

C. Desensitization of baclofen currents in neurons from Rgs6^+/+ (n=13) and Rgs6^−/− (n=18) mice (Mann-Whitney test: U_207,289=116, P=0.9843).

D. Activation rates of baclofen-induced current in neurons from Rgs6^+/+ (n=13) and Rgs6^−/− (n=13) mice (Mann-Whitney test: U_161,190=70, P=0.4793).

E. Deactivation rates of baclofen-induced current in neurons from Rgs6^+/+ (n=8) and Rgs6^−/− (n=14) mice (unpaired t-test: t₂₀=1.064, P=0.3000).

F. Representative currents evoked by serotonin (10 μM) in excitatory (GFP-labeled) HPC neurons from Rgs6^+/+ (black) and Rgs6^−/− (blue) mice; scale: 500 pA/5 s.

G. Steady-state serotonin-induced current density in neurons from Rgs6^+/+ (n=12) and Rgs6^−/− (n=17) mice (unpaired t-test: t₂₇=0.1052, P=0.9170).

H. Desensitization of serotonin-induced currents in neurons from Rgs6^+/+ (n=12) and Rgs6^−/− (n=17) mice (unpaired t-test: t₂₇=0.8806, P=0.3863).

I. Activation rates of serotonin-induced currents in neurons from Rgs6^+/+ (n=13) and Rgs6^−/− (n=15) mice (Mann-Whitney test: U_159,247=68, P=0.1846).

J. Deactivation rates of serotonin-induced currents in neurons from Rgs6^+/+ (n=13) and Rgs6^−/− (n=13) mice (unpaired t-test with Welch’s correction: t_17.45=2.345, *P=0.0311).

K. Representative currents evoked by adenosine (10 μM) in excitatory (GFP-labeled) HPC neurons from Rgs6^+/+ (black) and Rgs6^−/− (blue) mice; scale: 500 pA/5 s.

L. Steady-state adenosine-induced current density in neurons from Rgs6^+/+ (n=14) and Rgs6^−/− (n=13) mice (unpaired t-test: t₂₅=0.4678, P=0.6440).

M. Desensitization of adenosine-induced currents in neurons from Rgs6^+/+ (n=14) and Rgs6^−/− (n=13) mice (unpaired t-test: t₂₅=1.378, P=0.1805).

N. Activation rates of adenosine-induced currents in neurons from Rgs6^+/+ (n=12) and Rgs6^−/− (n=13) mice (Mann-Whitney test: U_181,144=53, P=0.1861).

O. Deactivation rates of adenosine-induced currents in neurons from Rgs6^+/+ (n=14) and Rgs6^−/− (n=14) mice (unpaired t-test: t₂₆=0.4754, P=0.6385).

P. Representative concentration-response experiments for baclofen-induced currents in excitatory (GFP-labeled) HPC neurons from Rgs6^+/+ (black) and Rgs6^−/− (blue) mice; scale: 500 pA/5 s.

Q. Concentration-response curves for baclofen-induced GIRK currents in neurons from Rgs6^+/+ and Rgs6^−/− mice.

R. Normalized concentration-response curves for baclofen-induced GIRK currents in neurons from Rgs6^+/+ and Rgs6^−/− mice.

S. Log(EC₅₀) for baclofen-induced GIRK currents in neurons from Rgs6^+/+ (n=7) and Rgs6^−/− (n=8) mice (unpaired t-test: t₁₃=0.5009, P=0.6248).

T. Representative concentration-response experiments for serotonin-induced currents in excitatory (GFP-labeled) HPC neurons from Rgs6^+/+ (black) and Rgs6^−/− (blue) mice; scale: 500 pA/5 s.

U. Concentration-response curves for serotonin-induced GIRK currents in neurons from Rgs6^+/+ and Rgs6^−/− mice.

V. Normalized concentration-response curves for serotonin-induced GIRK currents in neurons from Rgs6^+/+ and Rgs6^−/− mice.

W. Log(EC₅₀) for serotonin-induced GIRK currents in neurons from Rgs6^+/+ (n=6) and Rgs6^−/− (n=6) mice (unpaired t-test: t₁₀=0.8880, P=0.3954).

X. Representative concentration-response experiments for adenosine-induced currents in excitatory (GFP-labeled) HPC neurons from Rgs6^+/+ (black) and Rgs6^−/− (blue) mice; scale: 500 pA/5 s.

Y. Concentration-response curves for adenosine-induced GIRK currents in neurons from Rgs6^+/+ and Rgs6^−/− mice.

Z. Normalized concentration-response curves for adenosine-induced GIRK currents in neurons from Rgs6^+/+ and Rgs6^−/− mice.

AA. Log(EC₅₀) for adenosine-induced GIRK currents in neurons from Rgs6^+/+ (n=6) and Rgs6^−/− (n=6) mice (Mann-Whitney test: U_42,36=15, P=0.6991).

Data are represented as mean ± SEM.

GABA_BR, 5HT_1AR, and A₁R exhibit distinct coupling preferences for inhibitory G proteins

As R7 RGS/Gβ5 complexes exhibit a strong substrate preference for Gα_o (Hooks et al., 2003; Lan et al., 2000; Posner et al., 1999), we suspected that inhibitory G protein coupling preferences for GABA_BR, 5HT_1AR, and A₁R might contribute to the selective impact of RGS7 ablation on GPCR-GIRK signaling. Overnight pre-treatment of HPC cultures with pertussis toxin (PTX) eliminated currents evoked by baclofen, serotonin, and adenosine (Supp. Fig. 2), confirming that GABA_BR, 5HT_1AR, and A₁R all couple to GIRK channels via one or more members of the Gα_i/o subfamily of G proteins.

We next probed the Gα coupling preferences for GABA_BR, 5HT_1AR, and A₁R using an established bioluminescence resonance energy transfer (BRET) assay (Fig. 5A). By quantifying the maximal amplitudes (Fig. 5B,D,F) of agonist-evoked BRET responses mediated by specific Gα isoforms, we demonstrated that GABA_BR, 5HT_1AR, and A₁R can all signal effectively via the five PTX-sensitive Gα_i/o isoforms (Gα_oA, Gα_oB, Gα_i1, Gα_i2, Gα_i3). Since the amplitudes of BRET responses may be influenced by differences in the intrinsic extent of heterotrimer dissociation (Digby et al., 2008), we relied on the activation rates to compare Gα coupling preferences for the receptors, as these rates reflect the catalytic activity of GPCRs (Masuho et al., 2013). By this measure, 5HT_1AR (Fig. 5E) and to a lesser degree GABA_BR (Fig. 5C) exhibited a preference for Gα_o over Gα_i isoforms, while A₁R did not discriminate between Gα_o and Gα_i isoforms (Fig. 5G).

Figure 5. G protein coupling preferences of GABA_BR, 5HT_1AR, and A₁R.

A. Schematic representation of the BRET assay for real-time optical analysis of G protein activity. Agonist-induced activation of a GPCR leads to the dissociation of Gα-GTP and Venus-Gβγ subunits. Venus-Gβγ can the interact with the Gβγ effector mimetic masGRK3ct-Nluc-HA to produce the BRET signal.

B,C. G protein coupling summary diagrams for GABA_BR. Maximum amplitudes (B, red) and activation rate constants (C, blue) for 15 different G proteins were normalized to the largest value and plotted in the wheel diagrams. Line thickness represents the SEM of three technical replicates performed independently.

D,E. G protein coupling summary diagrams for 5HT_1AR. Maximum amplitudes (D) and activation rate constants (E) summary of G protein coupling for 5HT_1AR.

F,G. G protein coupling summary diagrams for A₁R. Maximum amplitudes (F) and activation rate constants (G) summary of G protein coupling for A₁R.

Data are represented as mean ± SEM.

GABA_BR, 5HT_1AR, and A₁R use Gα_o to varying degrees to regulate GIRK channels in HPC neurons

We next used a viral Cre approach to assess the extent to which GABA_BR, 5HT_1AR, and A₁R engage with Gα_o to activate GIRK channels in native neurons. HPC cultures from Gα_o^fl/fl mice were treated with AAV8-CaMKIIα-Cre(mCherry) or control vector (AAV8-CaMKIIα-ΔCre(mCherry)) and agonist-evoked GIRK currents were measured in infected neurons (Fig. 6A–O). ML297-induced currents did not differ between viral treatment groups, indicating that Gα_o ablation did not impact GIRK channel expression or activity (Supp. Fig. 3). Consistent with the relative preferences for Gα_o over Gα_i exhibited by 5HT_1AR (strong), GABA_BR (modest), and A₁R (no preference) in the reconstitution system, 5HT_1AR-GIRK signaling was most profoundly impacted by the loss of Gα_o (Fig. 6A,F,K). Steady-state 5HT_1AR-GIRK current amplitude was decreased (Fig. 6G) and desensitization was absent (Fig. 6H), a finding likely attributable to a prominent slowing of current activation (Fig. 6I). Deactivation was also prolonged (Fig. 6J). Gα_o ablation had a comparable effect on GABA_BR-GIRK signaling, albeit to a lesser degree (Fig. 6A–E). Loss of Gα_o correlated with a borderline significant reduction in A₁R-GIRK current amplitude and prolonged deactivation (Fig. 6K–O).

Figure 6. Impact of Gα_o ablation on GABA_BR-, 5HT_1AR-, and A₁R-GIRK signaling and coupling sensitivity.

A. Representative currents evoked by baclofen (100 μM) in excitatory (mCherry-labeled) HPC neurons from Gα_o^fl/fl mice infected with either AAV-CaMKIIα-Cre(mCherry) (cre, orange) or AAV-CaMKIIα-ΔCre(mCherry) control (Δcre, black) vector; scale: 500 pA/5 s.

B. Steady-state baclofen-induced current density in neurons treated with cre (n=7) and Δcre (n=7) vectors (unpaired t-test: t₁₂=2.479, *P=0.0290).

C. Desensitization of baclofen-induced currents in neurons treated with cre (n=7) and Δcre (n=7) vectors (unpaired t-test: t₁₂=2.333, *P=0.0379).

D. Activation rates of baclofen-induced currents in neurons treated with cre (n=6) and Δcre (n=6) vectors (unpaired t-test: t₁₀=6.447, ****P<0.0001).

E. Deactivation rates of baclofen-induced currents in neurons treated with cre (n=6) and Δcre (n=7) vectors (unpaired t-test: t₁₁=10.50, ****P<0.0001).

F. Representative currents evoked by serotonin (10 μM) in excitatory (mCherry-labeled) HPC neurons from Gα_o^fl/fl mice infected with either AAV-CaMKIIα-Cre(mCherry) (cre, orange) or AAV-CaMKIIα-ΔCre(mCherry) control (Δcre, black) vector; scale: 500 pA/5 s.

G. Steady-state serotonin-induced current density in neurons treated with cre (n=8) and Δcre (n=11) vectors (unpaired t-test: t₁₇=6.869, ****P<0.0001).

H. Desensitization of serotonin-induced currents in neurons treated with cre (n=8) and Δcre (n=11) vectors (unpaired t-test: t₁₇=8.717, ****P<0.0001).

I. Activation rates of serotonin-induced currents in neurons infected with cre (n=8) and Δcre (n=11) vectors (unpaired t-test: t₁₇=12.56, ****P<0.0001).

J. Deactivation rates of serotonin-induced currents in neurons treated with cre (n=8) and Δcre (n=11) vectors (Mann-Whitney test: U_68,122=2, ***P=0.0001).

K. Representative currents evoked by adenosine (10 μM) in excitatory (mCherry-labeled) HPC neurons from Gα_o^fl/fl mice infected with either AAV-CaMKIIα-Cre(mCherry) (cre, orange) or AAV-CaMKIIα-ΔCre(mCherry) control (Δcre, black) vector; scale: 500 pA/5 s.

L. Steady-state adenosine-induced current density in neurons treated with cre (n=10) and Δcre (n=10) vectors (unpaired t-test: t₁₈=2.077, P=0.0524).

M. Desensitization of adenosine-induced currents in neurons treated with cre (n=9) and Δcre (n=10) vectors (unpaired t-test: t₁₇=0.6014, P=0.5555).

N. Activation rates of adenosine-induced currents in neurons infected with cre (n=8) and Δcre (n=10) vectors (unpaired t-test: t₁₆=0.4176, P=0.6818).

O. Deactivation rates of adenosine-induced currents in neurons treated with cre (n=8) and Δcre (n=9) vectors (unpaired t-test: t₁₅=5.820, ****P<0.0001).

P. Representative concentration-response experiments for baclofen-induced currents in excitatory (mCherry-labeled) HPC neurons from Gα_o^fl/fl with either AAV-CaMKIIα-Cre(mCherry) (cre, orange) or AAV-CaMKIIα-ΔCre(mCherry) control (Δcre, black) vector; scale: 500 pA/5 s.

Q. Concentration-response curves for baclofen-induced GIRK currents in neurons treated with cre or Δcre vector.

R. Normalized concentration-response curves for baclofen-induced GIRK currents in neurons treated with cre or Δcre vector.

S. Log(EC₅₀) for baclofen-induced GIRK currents in neurons treated with cre (n=8) or Δcre (n=7) vector (Mann-Whitney test: U_54,66=26, P=0.8665).

T. Representative concentration-response experiments for serotonin-induced currents in excitatory (mCherry-labeled) HPC neurons from Gα_o^fl/fl with either AAV-CaMKIIα-Cre(mCherry) (cre, orange) or AAV-CaMKIIα-ΔCre(mCherry) control (Δcre, black) vector; scale: 500 pA/5 s.

U. Concentration-response curves for serotonin-induced GIRK currents in neurons treated cre or Δcre vector.

V. Normalized concentration-response curves for serotonin-induced GIRK currents in neurons treated with cre or Δcre vector.

W. Log(EC₅₀) for serotonin-induced GIRK currents in neurons treated with cre (n=7) or Δcre (n=9) vector (unpaired t-test: t₁₄=3.307, **P=0.0052).

X. Representative concentration-response experiments for adenosine-induced currents in excitatory (mCherry-labeled) HPC neurons from Gα_o^fl/fl with either AAV-CaMKIIα-Cre(mCherry) (cre, orange) or AAV-CaMKIIα-ΔCre(mCherry) control (Δcre, black) vector; scale: 500 pA/5 s.

Y. Concentration-response curves for adenosine-induced GIRK currents in neurons treated with cre or Δcre vector.

Z. Normalized concentration-response curves for adenosine-induced GIRK currents in neurons treated with cre or Δcre vector.

AA. Log(EC₅₀) for adenosine-induced GIRK currents in neurons treated with cre (n=9) or Δcre (n=8) vector (Mann-Whitney test: U_105,48=3, ***P=0.0006).

Data are represented as mean ± SEM.

GIRK channel sensitivity to agonist-induced activation was also impacted by Gα_o ablation in a GPCR-dependent manner (Fig. 6P–AA). Baclofen-induced current amplitudes were smaller in the absence of Gα_o (Fig. 6P,Q), but GABA_BR-GIRK coupling sensitivity was unaltered (Fig. 6R,S). Loss of Gα_o decreased the sensitivity of GIRK channels to serotonin (Fig. 6T–W). In contrast, GIRK channels were more sensitive to adenosine in the absence of Gα_o (Fig. 6X–AA).

GABA_BR, 5HT_1AR, and A₁R access a common pool of GIRK channels in HPC neurons

The differential impact of RGS7 and Gα_o ablation on GPCR-GIRK signaling in HPC neurons could indicate that GABA_BR, 5HT_1AR, and A₁R couple to unique subsets of GIRK channels in HPC neurons. To test this premise, we asked whether GIRK channel activation evoked by one GPCR occluded responses to the others. Neither adenosine (Fig. 7A,B) nor serotonin (Fig. 7C,D) induced an additive current during a maximal baclofen-induced GIRK current response. In contrast, baclofen augmented currents during maximal responses evoked by adenosine (Fig. 7A,B) or serotonin (Fig. 7C,D). Additive currents evoked by serotonin during a maximal adenosine-induced response, or by adenosine during a maximal serotonin-induced response, were smaller than maximal currents evoked by either agonist on its own (Fig. 7E,F). In addition, concurrent application of adenosine and serotonin evoked a smaller GIRK current response than the maximal baclofen-induced current, and a maximum baclofen-induced current occluded the response to co-applied adenosine and serotonin almost completely (Fig. 7G). Collectively, these data suggest that GABA_BR has the greatest access to a largely shared GIRK channel pool in HPC neurons, with 5HT_1AR and A₁R coupling primarily to overlapping but distinct subsets of GIRK channels within this pool (Fig. 7H).

Figure 7. Occlusion of GPCR-GIRK signaling in HPC neurons.

A. A representative trace showing a maximal response to baclofen (100 μM) followed application of adenosine (10 μM) in an HPC neuron from a C57BL/6J mouse (left), and a trace showing a maximal response to adenosine followed by application of baclofen (right); scale: 500 pA/5 s.

B. Current densities evoked by application of baclofen (b, 100 μM) or adenosine (a, 10 μM), additive current densities induced by co-application of adenosine (Δa) or baclofen (Δb), and total current densities measured after adenosine (b+a) or baclofen (a+b) co-application. Main effects of drug (F_1,44=49.31, P<0.0001) and sequence of application (F_1,44=103.5, P<0.0001) were detected by 2-way ANOVA (****P<0.0001 within drug, ^####P<0.0001 within application sequence; uncorrected Fisher’s LSD), but no difference was found in the sum current densities (t₂₂=0.2758, P=0.7853; n=12/group). Numbers within the bars are mean values.

C. A representative trace showing a maximal response to baclofen (100 μM) followed application of serotonin (10 μM) in an HPC neuron from a C57BL/6J mouse (left), and a trace showing a maximal response to serotonin followed by application of baclofen (right); scale: 500 pA/5 s.

D. Current densities evoked by application of baclofen (b, 100 μM) or serotonin (s, 10 μM), additive current densities induced by co-application of serotonin (Δs) or baclofen (Δb) application, and total current densities measured after serotonin (b+s) or baclofen (s+b) co-application. Main effects of drug (F_1,48=63.96, P<0.0001) and sequence of application (F_1,44=127.4, P<0.0001) were detected by 2-way ANOVA (****P<0.0001 within drug, ^####P<0.0001 within application sequence; uncorrected Fisher’s LSD), but no difference was found in the sum current densities (t₂₄=0.3388, P=0.7377; n=13/group). Numbers within the bars are mean values.

E. A representative trace showing a maximal response to adenosine (10 μM) followed by application of serotonin (10 μM) in an HPC neuron from a C57BL/6J mouse (left), and a trace showing a maximal response to serotonin followed by application of adenosine; scale: 500 pA/5 s.

F. Current densities evoked by application of adenosine (a, 10 μM) or serotonin (s, 10 μM), additive current densities induced by co-application of serotonin (Δs) or adenosine (Δa), and total current densities measured after serotonin (a+s) or adenosine (s+a) co-application. There was a main effect of application sequence (F_1,56=50.00, P<0.0001) but not drug (F_1,56=0.06591, P=0.7983) was detected by 2-way ANOVA (^####P<0.0001 within application sequence; uncorrected Fisher’s LSD); no difference was found in the sum current densities (t₂₈=0.05743, P=0.9546; n=15/group). Numbers within the bars are mean values.

G. A representative trace showing that a maximal response to baclofen (100 μM) occludes the response to co-applied serotonin (10 μM) and adenosine (10 μM) (left), but the maximal response observed following co-application of serotonin and adenosine does not fully occlude the response to baclofen; scale: 500 pA/5 s.

H. Venn diagram representing access of GABA_BR, A₁R, and 5HT_1AR to GIRK channels in HPC neurons, derived from mean currents evoked by agonists applied alone and in combination.

Data are represented as mean ± SEM.

Lastly, we examined occlusion of GPCR-GIRK signaling pathways in HPC neurons from Rgs6^−/− and Rgs7^−/− mice. Loss of RGS7 did not impact additive currents (Fig. 8A–D), indicating that this R7 RGS protein does not regulate GPCR access to the GIRK channel pool in HPC neurons. While the loss of RGS6 did not alter occlusion between GABA_BR- and A₁R-GIRK signaling pathways (Fig. 8E), or between the GABA_BR- and 5HT_1AR-GIRK signaling pathways (Fig. 8F), it did increase the occlusion between the 5HT_1AR- and A₁R-GIRK signaling pathways (Fig. 8G). Thus, RGS6 contributes to the functional compartmentalization of these 2 signaling pathways in HPC neurons, likely via its regulatory influence on the more Gα_o-dependent 5HT_1AR-GIRK signaling pathway (Fig. 8H).

Figure 8. Impact of RGS7 and RGS6 ablation on occlusion of GPCR-GIRK signaling.

A. Additive current densities induced by application of adenosine (Δa, 10 μM) or baclofen (Δb, 100 μM) during concurrent application of a saturating concentration of baclofen (100 μM) or adenosine (10 μM), respectively, in neurons from Rgs7^+/+ (n=12/group) and Rgs7^−/− (n=10/group) mice. Only a main effect of drug (F_1,40=57.11, P<0.0001) was detected by 2-way ANOVA (****P<0.0001 within genotype, Šídák’s multiple comparisons).

B. Additive current densities induced by application of serotonin (Δs, 10 μM) or baclofen (Δb, 100 μM) during concurrent application of a saturating concentration of baclofen (100 μM) or serotonin (10 μM), respectively, in neurons from Rgs7^+/+ (n=13/group) and Rgs7^−/− (n=10/group) mice. Only a main effect of drug (F_1,42=95.45, P<0.0001) was detected by 2-way ANOVA (****P<0.0001 within genotype, Šídák’s multiple comparisons).

C. Additive current densities induced by application of serotonin (Δs, 10 μM) or adenosine (Δa, 10 μM) during concurrent application of a saturating concentration of adenosine (10 μM) or serotonin (10 μM), respectively, in neurons from Rgs7^+/+ (n=15/group) and Rgs7^−/− (n=10/group) mice. No main effects of drug or genotype, or drug × genotype interaction, were detected by 2-way ANOVA.

D. Venn diagram representing access of GABA_BR, A₁R, and 5HT_1AR to GIRK channels in HPC neurons from Rgs7^−/− mice.

E. Additive current densities induced by application of adenosine (Δa, 10 μM) or baclofen (Δb, 100 μM) during concurrent application of a saturating concentration of baclofen (100 μM) or adenosine (10 μM), respectively, in neurons from Rgs6^+/+ (n=6/group) and Rgs6^−/− (n=8/group) mice. Only a main effect of drug (F_1,24=18.30, P=0.0003) was detected by 2-way ANOVA (*P<0.05 and **P<0.01 within genotype, Šídák’s multiple comparisons).

F. Additive current densities induced by application of serotonin (Δs, 10 μM) or baclofen (Δb, 100 μM) during concurrent application of a saturating concentration of baclofen (100 μM) or serotonin (10 μM), respectively, in neurons from Rgs6^+/+ (n=6/group) and Rgs6^−/− (n=7/group) mice. Only a main effect of drug (F_1,22=29.81, P<0.0001) was detected by 2-way ANOVA (*P<0.05 and ***P<0.001 within genotype, Šídák’s multiple comparisons).

G. Additive current densities induced by application of serotonin (Δs, 10 μM) or adenosine (Δa, 10 μM) during concurrent application of a saturating concentration of adenosine (10 μM) or serotonin (10 μM), respectively, in neurons from Rgs6^+/+ (n=8/group) and Rgs6^−/− (n=7/6/group) mice. Only a main effect of genotype (F_1,25=27.43, P<0.0001) was detected by 2-way ANOVA (*P<0.05 and ***P<0.001 within drug, Šídák’s multiple comparisons).

H. Venn diagram representing access of GABA_BR, A₁R, and 5HT_1AR to GIRK channels in HPC neurons from Rgs6^−/− mice.

Data are represented as mean ± SEM.

DISCUSSION

In this study, we evaluated how R7 RGS proteins (RGS6 and RGS7) shape signaling involving inhibitory GPCRs that engage pertussis toxin-sensitive G proteins to regulate a common effector, the GIRK channel, in HPC neurons. We had expected that R7 RGS protein ablation would increase the duration (prolong the deactivation) of GPCR-GIRK signaling by extending the phase of G protein activation. Other predictions based on the catalytic activity of R7 RGS proteins included a potential slowed activation, increase in current density, and increased sensitivity of the GIRK channel to GPCR activation. What we found is that R7 RGS proteins impact multiple facets of GPCR-GIRK signaling in a receptor-dependent manner, and sometimes in ways that are not obviously linked to a catalytic action on active G proteins. Our data show that RGS7 accelerates deactivation and dampens the sensitivity of GABA_BR-GIRK signaling, while accelerating activation and enhancing the sensitivity of 5HT_1AR-GIRK signaling. In contrast, RGS6 accelerates the deactivation of 5HT_1AR-GIRK signaling and contributes to the functional compartmentalization of this signaling pathway. Neither R7 RGS protein exerts a discernable impact on A₁R-GIRK signaling.

RGS6 negatively regulates GABA_BR-GIRK signaling in VTA dopamine neurons (DeBaker et al., 2023) and cerebellar granule cells (Maity et al., 2012), indicating that it can regulate GABA_BR-GIRK signaling under the right circumstances. The lack of impact of RGS6 ablation on GABA_BR-GIRK signaling in HPC neurons is not likely explained by compensatory up-regulation of RGS7 in Rgs6^−/− cultures (Ostrovskaya et al., 2014). Furthermore, the impact of RGS7 ablation on GABA_BR-GIRK signaling aligns with that seen in HPC neurons from Gβ5^−/− mice (Ostrovskaya et al., 2014; Xie et al., 2010), which exhibit a profound reduction in all R7 RGS proteins (Anderson et al., 2009). Thus, RGS7 appears to provide the main if not only R7 RGS influence on GABA_BR-GIRK signaling in HPC neurons. It is possible, however, that the influence of RGS6 is augmented under certain conditions. Indeed, RGS6 expression is enhanced in adult-born HPC neurons by voluntary running and this blunts the GABA_BR-induced suppression of voltage-gated Ca²⁺ channels (Gao et al., 2020).

Among the inhibitory G protein isoforms, Gα_o is the preferred substrate for the catalytic (GTPase) activity of R7 RGS proteins (Hooks et al., 2003; Lan et al., 2000; Posner et al., 1999). Our data suggest that overlapping but distinct G protein coupling preferences of GABA_BR, 5HT_1AR, and A₁R explain in part the receptor-dependent influence of R7 RGS proteins on GPCR-GIRK signaling in HPC neurons. Using a BRET-based GPCR-G protein coupling assay (Masuho et al., 2015a), we found that 5HT_1AR and to a lesser degree GABA_BR preferred signaling via Gα_o over Gα_i, while A₁R did not discriminate between Gα_o and Gα_i isoforms. To examine the utilization of Gα_o by these receptors in a more natural context, we measured the impact of Gα_o ablation on GPCR-GIRK signaling in HPC neurons. Consistent with the BRET data, Gα_o ablation in HPC neurons profoundly disrupted 5HT_1AR-GIRK signaling and to a lesser degree GABA_BR-GIRK signaling, while exerting only a modest impact on A₁R-GIRK signaling. Therefore, among these receptors and in these neurons, 5HT_1AR relies most heavily on Gα_o for signaling to GIRK channels.

Our occlusion experiments show that GABA_BR, 5HT_1AR, and A₁R all regulate GIRK channel activity in HPC neurons but engage this somato-dendritic effector pool to different extents. GABA_BR exhibits the greatest access to this effector pool and can occlude almost completely the GIRK channel response to concurrent A₁R or 5HT_1AR stimulation. This suggests that A₁R- and 5HT_1AR cannot mobilize enough active G protein to evoke a maximal GIRK channel response. This could reflect relatively lower levels of functional A₁R or 5HT_1AR, overlapping but distinct G protein coupling preferences of the GPCRs, different rates of G protein activation by the receptors, and/or differences in the relative proximities of the receptors to GIRK channels. Ultrastructural analyses examining GIRK channel colocalization with 5HT_1AR or A₁R in HPC pyramidal neurons have not been reported, but both receptors are found in the somato-dendritic compartment (Rebola et al., 2003; Riad et al., 2000). As noted previously, GIRK channels and GABA_BR exhibit strong co-clustering in the perisynaptic space on dendritic spines of HPC pyramidal neurons, and more segregation in dendritic shafts (Kulik et al., 2006). While the whole-cell voltage-clamp recording approach used in the present study sheds light on regulatory mechanisms relevant to GPCR-GIRK signaling in the somato-dendritic compartment, mechanisms that impact GPCR-GIRK signaling dynamics may differ substantially at the synapse.

Our occlusion data further show that neither RGS6 nor RGS7 limits the maximal GPCR-GIRK response in HPC neurons. Moreover, neither R7 RGS protein promotes preferred access of GABA_BR to the GIRK channel pool, nor limits the access of 5HT_1AR or A₁R to GIRK channels. Loss of RGS6 did, however, increase occlusion between the 5HT_1AR- and A₁R-GIRK signaling pathways. This suggests that RGS6 promotes the functional compartmentalization of 5HT_1AR- and A₁R-GIRK signaling in HPC neurons, likely via its influence on the more Gα_o-dependent 5HT_1AR-GIRK signaling pathway.

The catalytic activity of R7 RGS proteins accelerates the rate of Gα_o inactivation. Provided the rate of agonist dissociation from the GPCR is faster than the intrinsic GTPase activity of Gα_o, R7 RGS proteins should accelerate deactivation of an agonist-induced, Gα_o-dependent GIRK channel response. The prolonged deactivation of GABA_BR-, 5HT_1AR- and A₁R-GIRK responses in neurons lacking Gα_o indicate that agonist dissociation rate is sufficiently fast for all 3 GPCRs to allow detection of an R7 RGS protein influence on this kinetic parameter, if present. Indeed, RGS7 ablation prolonged deactivation of the GABA_BR-GIRK response, and RGS6 ablation prolonged deactivation of the 5HT_1AR-GIRK response. Interestingly, RGS6 also negatively regulates the 5HT_1AR-induced suppression of cAMP production in HPC and cortical neurons (Stewart et al., 2014). As such, RGS6 may be a dedicated regulator of 5HT_1AR-dependent signaling in HPC neurons.

The lack of impact of either RGS6 or RGS7 ablation on A₁R-GIRK deactivation could reflect a redundant influence of these R7 RGS proteins, combined with the relatively modest recruitment of Gα_o by A₁R in HPC neurons. It is also possible that another RGS protein regulates A₁R-GIRK signaling in HPC neurons. In this regard, it is noteworthy that RGS4 is expressed in HPC pyramidal neurons (Cembrowski et al., 2016; Gold et al., 1997), can accelerate the GTPase activity of Gα_i/o and Gα_q (Berman et al., 1996; Huang et al., 1997; Saugstad et al., 1998), and modulates GIRK-dependent signaling in expression systems (Fowler et al., 2007; Ulens et al., 2000).

The catalytic activity of R7 RGS proteins should also increase the level of inactive Gα_o-GDP/Gβγ heterotrimer available for activation. In a signaling pathway involving collision coupling (Berlin et al., 2020; Kahanovitch et al., 2017), this influence should accelerate the activation of an agonist-induced, Gα_o-dependent GIRK channel response. Consistent with this premise, Gα_o ablation prolonged the activation of both the maximal GABA_BR-GIRK and 5HT_1AR-GIRK response, the latter in dramatic fashion. RGS7 ablation similarly prolonged the activation of the maximal 5HT_1AR-GIRK response. Although RGS7 ablation did not slow the deactivation of the maximal GABA_BR-GIRK response, loss of RGS7 did correlate with prolonged activation of GIRK current evoked by low concentrations of baclofen (Ostrovskaya et al., 2014). While we do not understand why this influence RGS7 is evident with maximal stimulation of the 5HT_1AR-GIRK signaling pathway in HPC neurons, and not with maximal stimulation of GABA_BR-GIRK signaling, the influence of an R7 RGS protein on the activation rate of a GPCR-GIRK response may be conspicuous when that response is limited by the amount of activated G protein generated by the receptor.

This framework may also help explain why RGS7 ablation enhances the sensitivity of GIRK channels to GABA_BR activation and reduces the sensitivity of GIRK channels to 5HT_1AR activation. Reduced sensitivity of GIRK channels to 5HT_1AR activation was also seen in neurons lacking Gα_o. When the GIRK channel response is limited by the amount of activated G protein mobilized by the receptor, as we envision is the case for the 5HT_1AR-GIRK signaling pathway, then the positive influence of R7 RGS proteins on production of inactive Gα_o-GDP/Gβγ is a critical determinant of both activation rate and response amplitude. Removing that influence slows activation and decreases response amplitude. When another factor(s) limits the response, as we suspect is the case for the GABA_BR-GIRK signaling pathway, then the RGS7 influence of GIRK current deactivation is more prominent. This influence should suppress the amplitude of sub-maximal GABA_BR-GIRK responses, decreasing the apparent sensitivity of the GIRK channel to GABA_BR activation.

An intriguing outcome of this study is that RGS7 ablation slowed activation, while RGS6 ablation prolonged deactivation, of the 5HT_1AR-GIRK response. This finding could suggest that RGS6 (deactivation) and RGS7 (activation) collaborate in non-redundant fashion to accelerate the kinetics of 5HT_1AR-GIRK signaling in HPC neurons. However, as the core catalytic influence of an R7 RGS protein should accelerate both the activation and deactivation of a Gα_o-dependent GIRK channel response, it is difficult to envision how RGS6 and RGS7 could make separable contributions to these discrete kinetic elements of the 5HT_1AR-GIRK response. This finding may suggest that 5HT_1AR-GIRK signaling pathway is regulated by RGS6 in a subset of HPC neurons, and by RGS7 in other HPC neurons. Alternatively, within pool of GIRK channels accessible to 5HT_1AR in a HPC neuron, there may be subsets regulated primarily by RGS6 or RGS7.

The findings in this study support a conceptual framework wherein the dynamics of GPCR-GIRK signaling in neurons are determined by GPCR coupling preferences for inhibitory Gα protein isoforms, the Gα_o substrate bias of R7 RGS proteins, and the extent of GPCR access to a largely shared pool of GIRK channels. Our findings further suggest that the relative abundance of specific G protein isoforms in proximity to each GPCR will shape GPCR-GIRK signaling dynamics. Similarly, the roles of other proteins implicated in GPCR-GIRK signaling dynamics such as GPCR kinase 2/3 (GRK2/3) (Raveh et al., 2010), Gα inhibitory interacting protein GINIP (Park et al., 2023), and potassium channel tetramerization domain-containing proteins (KCTDs) (Fritzius et al., 2024; Fritzius et al., 2017; Zheng et al., 2019; Zuo et al., 2019) warrant further investigation, as these factors may contribute in important and receptor-specific ways to neuronal GPCR-GIRK signaling. This conceptual framework may help explain prior reports of neuroadaptations involving GIRK-dependent signaling linked to one but not another inhibitory GPCR. For example, A₁R-GIRK but not GABA_BR-GIRK signaling was enhanced in HPC neurons following NMDA receptor activation (Chung et al., 2009), and chronic morphine treatment enhanced 5HT_1AR-GIRK but not GABA_BR-GIRK signaling in HPC cultures (Nassirpour et al., 2010). The GPCR specificity of the neuroadaptations involving GIRK-dependent signaling could result from changes in the level or function of pathway-specific regulators, including RGS proteins.

Previously, we highlighted the influence of the GABA_BR-RGS7-GIRK signaling axis on critical HPC-dependent processes (Ostrovskaya et al., 2014). RGS7 ablation diminished long-term depression and depotentiation of long-term potentiation in CA1 pyramidal neurons, which was consistent with the diminished performance of Rgs7^−/−mice in tests of fear learning, as well as spatial learning and memory (Ostrovskaya et al., 2014). Thus, RGS7 in the HPC may be a promising target for treatment of conditions involving cognitive dysfunction. Moreover, therapeutic interventions targeting RGS7 or other RGS proteins may afford fine-tuning of specific GPCR-effector signaling pathways in a manner that avoids undesirable side effects, which may prove particularly beneficial in pathological settings involving dysfunction of only specific GPCR-effector pathways.

HIGHLIGHTS.

• GIRK channels are a common effector in G protein signaling pathways in neurons

• Our understanding of GPCR-GIRK signaling mechanisms in neurons is incomplete

• RGS6 and RGS7 regulate neuronal GIRK channels in a GPCR-dependent manner

• Inhibitory GPCRs engage Gα_o to varying extents to activate GIRK channels in neurons

• G protein use, RGS substrate bias, and effector access shape GPCR-GIRK signaling

Abbreviations

AAV

adeno-associated virus

A₁R

adenosine receptor type 1

G protein

heterotrimeric GTP-binding protein

Gα

G protein alpha subunit

Gβγ

G protein beta-gamma subunit

GABA

gamma aminobutyric acid

GABA_BR

GABA receptor type B

GFP

green fluorescent protein

GIRK channel

G protein-gated inwardly rectifying K⁺ channel

GPCR

G protein-coupled receptor

HPC

hippocampus

Kir3

inwardly rectifying K⁺ channel sub-family 3

5HT

serotonin

5HT_1AR

5HT receptor type 1A

RGS

Regulator of G protein Signaling

R7 RGS

R7 sub-family of RGS proteins

Data Availability

Data will be made available on request.