The signaling proteins GPR158 and RGS7 modulate excitability of L2/3 pyramidal neurons and control A-type potassium channel in the prelimbic cortex

Induction of depressive-like behaviors by stress is accompanied by selective reduction in intrinsic excitability of L2/3 prelimbic neurons

We began our studies by examining the effects of stress on the intrinsic excitability of L2/3 neurons in the prelimbic area of the mPFC. Following physical restraint stress (PRS), the depression-like behavior of mice was evaluated in a forced swim test (FST), and whole-cell patch-clamp recordings were obtained from the L2/3 pyramidal neurons (Fig. 1, A and B). The cell identity was confirmed by action potential (AP) properties and input–output relationship (38, 39). Analysis of the behavioral data indicated that mice subjected to PRS displayed significantly increased immobility during the FST compared with control mice (t(6) = 6.97, p = 0.0004), demonstrating a depressant-like effect (Fig. 1C). Whole-cell recordings revealed a significant (t(22) = 2.17, p = 0.041) hyperpolarization of the resting membrane potential (RMP) in mice subjected to PRS relative to control subjects (Fig. 1D). Interestingly, the excitability of L2/3 pyramidal neurons was also significantly reduced in PRS-exposed mice, as evidenced by fewer action potentials in response to depolarizing current steps compared with control neurons, when the cells were held at their resting potentials (Fig. 1, E and F). A repeated-measures analysis of variance (ANOVA) revealed significant main effects of group (F(1,22) = 5.55, p = 0.028) and current intensity (F(1.4, 30.5) = 223, p < 0.001; Greenhouse–Geisser-corrected). There was also a significant interaction of group with current density (F(1.4, 30.5) = 3.88, p = 0.045; Greenhouse–Geisser-corrected). Post hoc analysis using t test revealed a significant group effect when depolarizing current steps were ≥200 pA (all values, p < 0.05). The observed modulation of intrinsic excitability by stress was specific to L2/3 neurons (Fig. 1G), as the excitability of layer 5 (L5) neurons obtained from stressed animals was comparable with those from control unstressed mice (F(1,14) = 0.57, p = 0.46 for main effect of group and F(2, 112) = 1.6, p = 0.22 for interaction; repeated-measures ANOVA). Taken together, these data suggest that chronic stress induces specific reduction in excitability of L2/3 pyramidal neurons in prelimbic cortex in parallel with development of depressive-like behaviors.

Stress-induced changes in excitability of L2/3 neurons of mPFC are mediated by GPR158–RGS7 complex

In search for the molecular mechanisms underlying changes in intrinsic excitability, we focused on probing the involvement of a GPR158–RGS7 complex, which is prominently enriched in L2/3 neurons as knockout of either GPR158 or RGS7 results in a marked antidepressant phenotype and resilience to stress (35). To exclude the impact of changes in RMP, in the following experiments all cells were held at −70 mV unless noted otherwise. Indeed, elimination of GPR158 or RGS7 significantly increased the intrinsic excitability of layer L2/3 mPFC neurons (Fig. 2). Specifically, Gpr158^−/− neurons fired more spikes as compared with littermate control Gpr158^+/+ neurons (Fig. 2A). A repeated-measures ANOVA revealed significant main effects of group (F(1,23) = 8.77, p = 0.007) and current intensity (F(1.4, 31.7) = 540, p < 0.001; Greenhouse–Geisser-corrected). There was also a significant interaction of group by current intensity (F(1.4,31.7) = 7.6, p = 0.005; Greenhouse–Geisser-corrected). Post hoc analysis revealed a significant group effect when depolarizing current steps were ≥200 pA (all values, p < 0.01; t test). In addition to enhanced excitability, knockout of GPR158 also induced a significant depolarization of RMP (t(27) = 2.42, p = 0.022), increase in input resistance (t(27) = 2.25, p = 0.033), decrease in rheobase current (t(24) = 2.43, p = 0.023), and decrease in delay to the first spike evoked by depolarizing current injection (t(23) = 2.99, p = 0.007; Table 1). The effect of RGS7 ablation was quantitatively similar to that of GPR158 loss. Layer 2/3 mPFC neurons of Rgs7^−/− also fired more action potentials relative to Rgs7^+/+ neurons in response to depolarizing current injections, whereas chronic stress did not significantly change the excitability of Rgs7^−/− neurons (Fig. 2B). A repeated-measures ANOVA revealed significant main effects of group (F(2,40) = 4.43, p = 0.018) and current intensity (F(1.5,61.8) = 580.0, p < 0.001; Greenhouse–Geisser-corrected), but not interaction of group by current intensity (F(3.1,61.8) = 2.5, p = 0.066; Greenhouse–Geisser-corrected). Post hoc analysis revealed that the excitability of Rgs7^+/+ neurons was significantly different from Rgs7^−/− neurons either with (p = 0.024) or without (p = 0.007) chronic stress. Furthermore, ablation of RGS7 phenocopied the effects seen upon GPR158 loss, including depolarization of RMP (t(27) = 2.41, p = 0.028), increase in input resistance (t(27) = 13.9, p = 0.002), decrease in rheobase current (t(17) = 2.33, p = 0.033), and decrease in the delay to first spike evoked by depolarizing current injection (t(17) = 2.38, p = 0.029; Table 1). To test molecular specificity of this effect, we next probed the involvement of R7BP that also prominently regulates RGS7 but does not have an impact on depression-related behaviors (35). We found that elimination of R7BP did not change the intrinsic excitability of L2/3 neurons (Fig. 2C). Furthermore, the observed changes following Gpr158 or Rgs7 deletion were specific to L2/3 neurons, as no genotype differences were observed in excitability of L5 neurons, with or without stress (Fig. 2, D, E, and F, and Table 2). Together, these data suggest that the GPR158–RGS7 complex specifically modulates intrinsic excitability of layer 2/3 neurons in PFC in response to stress.

GPR158–RGS7 complex exerts its effects on excitability via cAMP

The GPR158–RGS7 complex has been shown to modulate signaling via the second messenger cAMP (37). Thus, we next tested whether the effects of GPR158 and RGS7 on excitability involve changes in cAMP. We found that in WT neurons elevating cAMP by adding the nonhydrolyzable analog S_p-cAMP into the internal recording solution significantly increased intrinsic excitability of layer 2/3 mPFC neurons, mimicking the effect of GPR158 or RGS7 ablation (Fig. 3, A and B; repeated-measures ANOVA; F(1,28) = 11.18, p = 0.002). Importantly, elimination of GPR158 completely prevented the effects of S_p-cAMP, which was not observed in Gpr158^−/− neurons (Fig. 3, C and D; repeated-measures ANOVA; F(1,6) = 0.13, p = 0.73). Similarly, S_p-cAMP failed to alter the excitability in neurons lacking RGS7 (Fig. 3, E and F; repeated-measures ANOVA; F(1,13) = 1.2, p = 0.29). These data indicate that GPR158 and RGS7 modulate intrinsic excitability of L2/3 prelimbic neurons through regulating intracellular cAMP.

GPR158 and RGS7 modulate A-type potassium channels

The decrease in RMP amid increase in input resistance upon deletion of GPR158 and RGS7 suggests the likely closing of inhibitory ion channels in these neurons as a mechanism for increasing excitability. To establish the identity of such channels, we investigated barium-sensitive potassium current, which includes both inward and outward currents and greatly affects membrane properties and excitability (40–42). We found that deletion of GPR158 or RGS7 did not significantly change inward currents when the membrane potential was stepped to a range from −90 to −120 mV (Fig. 4, A and B). Instead, the outward current was significantly inhibited in the range from −70 to −50 mV (main effects of voltage when tested with repeated-measures ANOVA, F(1,18) = 9.27, p = 0.007 and F(1,15) = 7.10, p = 0.018 for Gpr158 and Rgs7, respectively). This difference was also significant when we normalized the barium-sensitive current to a fixed potential at −80 mV (Fig. 4C; t(18) = 3.5, p = 0.002 and t(15) = 3.0, p = 0.009 for Gpr158 and Rgs7, respectively). These observations suggest the involvement of A-type potassium channels, whose genetic deletion or pharmacological inhibition has been shown to produce effects on depolarizing RMP, input resistant, and firing frequency as well as delaying the first spike in response to depolarizing current injection (43–45), which are similar to changes seen in Gpr158^−/− and Rgs7^−/− neurons. To test this hypothesis, we directly measured I_A in brain slices using an established protocol (Fig. 5A) (46). Indeed, we observed that loss of GPR158 or RGS7 (Fig. 5, B and C) significantly reduced I_A (t(16) = 2.41, p = 0.029 for GPR158 and t(25) = 2.89, p = 0.009 for RGS7. These data support the involvement of A-type potassium channels in modulating membrane properties and intrinsic excitability of L2/3 pyramidal neurons caused by inactivation of GPR158–RGS7 complex.

GPR158 directly modulates Kv.4.2 A-type potassium channel

Kv4.2 potassium subunit is the major contributor of A-type currents (47). Interestingly, proteomic profiling has identified GPR158 to be present among a large number of proteins associated with Kv4.2 potassium channel (48, 49). Thus, we next turned to a transfected HEK293 cell system to specifically test the interaction of GPR158–RGS7 with Kv4.2 and study its functional consequences. Indeed, we found that GPR158 effectively coimmunoprecipitated with Kv4.2 when coexpressed in HEK293 cells (Fig. 6A), validating their physical binding. In support of our model, the Kv4.2 channel is prominently modulated by cAMP via PKA-dependent phosphorylation of Thr-38 and Ser-552, which inhibits its plasma membrane trafficking (50, 51). To investigate the impact of GPR158 on Kv4.2 gating and the relevance of cAMP in this process, we coexpressed GPR158 with WT Kv4.2 or a Kv4.2AA double mutant (Kv4.2 T38A/S552A) that abolishes PKA regulation at the two sites. Cells were further cotransfected with the construct encoding mCherry to identify the cells that express the desired constructs (Fig. 6B). Whole-cell patch-clamp recordings revealed that transfecting Kv4.2 resulted in a rapidly activating and inactivating K⁺ current typical of I_A, which was absent in untransfected cells (Fig. 6C). Mutation of Kv4.2 channels or coexpression of GPR158 might not affect RMP because the current used to hold the cells at −50 mV was comparable between groups (Fig. 6D, left; F(3,36) = 1.46, p = 0.24). Furthermore, coexpression of GPR158 did not affect the peak amplitude of the current carried by either WT or mutant Kv4.2 channel (Fig. 6D, right; F(3,36) = 0.84, p = 0.48; one-way ANOVA), suggesting the lack of the channel modulation at the baseline when cAMP levels are low (47).

We next studied the effect of GPR158 on Kv4.2 upon elevating cAMP concentration by forskolin treatment, which activates the cAMP-producing enzyme adenylyl cyclase (Fig. 6E). Consistent with previously noted inhibitory effects of cAMP, forskolin quickly inhibited the peak amplitude of Kv4.2 current in all groups of cells (main effect of time, F(3.1,102.7) = 36.4, p < 0.001; repeated-measures ANOVA, Greenhouse–Geisser-corrected). There was also a significant effect of group (F(3,33) = 7.59, p = 0.001) and group by time interaction (F(9.3,102.7) = 3.60, p = 0.001). This effect was further underscored by the analysis of the maximal effect size at the effect plateau (Fig. 6F); one-way ANOVA (F(3,33) = 5.57, p = 0.003). The follow-up post hoc tests confirmed the role of cAMP in this effect: the Kv4.2AA mutant was insensitive to cAMP modulation, and forskolin was significantly less effective in inhibiting its currents (p < 0.05). Similarly, introduction of GPR158 also made Kv4.2 less sensitive to inhibition by forskolin (p < 0.05). Finally, addition of GPR158 to Kv4.2AA did not further alter its forskolin inhibition, indicating an occlusion of the effect (p < 0.05). It is notable that forskolin also induced a slight reduction of the currents in cells expressing Kv4.2AA constructs. This is likely due to the phosphorylation of the pore-forming subunit mediated by the extracellular signal-regulated kinase/mitogen-activated protein kinase signaling cascade (52, 53). These observations suggest that the GPR158 complexes counteract the inhibitory action of cAMP on A-type conductances via phosphorylation of the PKA-dependent sites on Kv4.2 channels (Fig. 6G).

Animal models

All studies were carried out in accordance with the National Institute of Health guidelines and were granted formal approval by the Institutional Animal Care and Use Committee of The Scripps Research Institute. The generation of Gpr158^−/− (36, 37), Rgs7^−/− (76), and R7bp^−/− (77) mice was described previously. All mice were maintained on C57/Bl6 background and were 2–5 months old during the experiments. All animals used for comparing genotypes were littermates derived from heterozygous breeding pairs. Mice were housed in groups on a 12-h light–dark cycle with food and water available ad libitum. Both male and female mice were used in this study.

Genetic constructs

Cloning of full-length mouse GPR158 into the pcDNA3.1/V5-His-TOPO was described previously (78). The plasmid encoding full-length mouse Kv4.2 (GenBank^TM accession number NM_019697; myc-DDK-tagged in C terminus) for mammalian expression was purchased (Origene). Kv4.2 bearing mutations to express Ala at the two PKA target sites Thr-38 and Ser-552 (Kv4.2AA) was generated using the In-Fusion HD cloning system (Clontech) in pcDNA3.1. Each construct was verified by DNA sequencing.

Behavior and physical restraint stress

PRS was performed in plastic tubes (30-mm diameter × 115-mm length) with holes for ventilation. Stressed mice were restrained horizontally in tubes for 2 h for 14 days; nonstressed mice were left undisturbed in their home cages. On day 15, all mice underwent an FST that was conducted using a vertical glass cylinder filled with water (25 °C). The mice spent 6 min in the water, and immobility was scored from 2 to 6 min by an independent researcher. A mouse was regarded as immobile when floating motionless or making only those movements necessary to keep its head above the water. Brain slices were prepared within 1 h following the FST, and whole-cell patch-clamp recordings were obtained from pyramidal cells in L2/3 of prelimbic area.

Slice preparation and whole-cell recordings from brain slices

Coronal slices (300 μm) containing prefrontal cortex (anteroposterior +1.5–2.2) were cut in ice-cold aCSF (124 mm NaCl, 2.8 mm KCl, 1.25 mm NaH₂PO₄, 2 mm CaCl₂, 1.25 mm MgSO₄, 26 mm NaHCO₃, 10 mm glucose, pH 7.5, bubbled with 95% O₂, 5% CO₂) using a vibrating tissue slicer (VT1200, Leica). Slices were transferred to a holding chamber (79) where they remained in oxygenated aCSF at 32–35 °C until use. During recording, slices were transferred to a submerged recording chamber where they were continuously perfused at 2 ml/min with oxygenated aCSF and maintained at 32–36 °C using an inline temperature controller. Neurons were visualized with IR differential video interference microscopy. Whole-cell recordings were obtained with standard-wall borosilicate glass pipettes (2–5 megaohms) filled with the following solution: 110 mm K-gluconate, 20 mm KCl, 10 mm di-Tris-P-creatine, 10 mm HEPES, 2 mm MgCl₂, 2 mm Na₂ATP, 0.3 mm Na₂GTP with a pH of 7.3 and osmolarity of 290 mosmol. Prelimbic L2/3 neurons were recognized by their large somata, prominent apical dendrite, and firing properties. To verify identity, some cells were injected with 0.1% biocytin through the recording electrode and visualized using a streptavidin Alexa Flour 488 reaction as described previously (39). One neuron was studied in each slice. Only neurons with an RMP more negative than −60 mV, an input resistance (R_N) >50 megaohms, an AP amplitude >45 mV relative to threshold, and an access resistance less than 20 megaohms were included.

RMP was obtained immediately by switching to current clamp after whole-cell mode was established. The membrane capacitance was estimated by using the auto C-slow function in Patchmaster. Intrinsic properties of mPFC neurons were recorded 3 min later (to allow the diffusion of internal solution) according to the following protocols. 1) V–I relationships were obtained from a series of 500-ms current injections (range −300 to 50 pA) and plotting the plateau voltage deflection against current amplitude. Neuronal R_N was determined from the slope of the linear fit portion of the V–I plot where the voltage sweeps did not exhibit sags or active conductance (80). 2) AP properties, including the minimum current necessary to elicit an AP (I_threshold), were studied with an ascending series of 500-ms depolarizing pulses to elicit one single spike. 3) Neuronal excitability was assessed by counting the number of spikes evoked in response to a series of 1-s depolarizing steps (range, 100–600 pA at 100-pA increments or 50–450 pA at 50-pA increases with a 20-s intertrial interval).

To record barium-sensitive current, cells were held at −50 mV and V–I relationships were obtained from a serious of voltage steps (from −50 to −120 mV) before and 10 min after bath application of 100 μm barium chloride. Barium-sensitive currents were calculated by subtraction of current responses to voltage steps in the presence and absence of barium. A-type current was isolated using a voltage protocol described previously (46). Briefly, the cells were first held at −40 mV to inactivate most of the outward current. After stabilization, a step was made to + 40 mV for 500 ms and then returned to −40 mV for 5 s followed by a prepulse to −100 mV (150-ms duration) to allow preferential recovery of the A-type current. This prepulse duration is long enough to recover most of the A current but is too short to recover the majority of the delayed rectifier currents (81, 82). A-type current was defined as the difference between the peak current during the test potential (+40 mV) after the prepulse and current during the test pulse without prepulse (relevant to the current used to hold the cells at −40 mV).

Cell culture and whole-cell recording from transfected HEK293 cells

Cell culture was carried out using standard procedures. HEK293T/17 cells were cultured at 37 °C in 5% CO₂ in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, minimum essential medium nonessential amino acids, 1 mm sodium pyruvate, and 2 mm GlutaMAX. 80,000 cells were plated on 12-mm poly-d-lysine–coated glass coverslips and transfected using Lipofectamine LTX (Invitrogen) and Plus reagent (Invitrogen) with a 1:1 ratio of GPR158 and Kv4.2 expression constructs and used 16 h later. Transfected cells were identified by expression of the fluorescent protein mCherry. Coverslips were bathed in recording solution for electrophysiological recording containing the following: 141 mm NaCl, 4.7 mm KCl, 1.2 mm MgCl₂, 1.8 mm CaCl₂, 10 mm glucose, 10 mm HEPES, pH 7.4, with NaOH. Whole-cell recordings were performed on fluorescently labeled cells at room temperature with glass pipettes filled with internal solution containing 125 mm KCl, 4 mm MgCl₂, 10 mm HEPES, 10 mm EGTA, 5 mm Mg-ATP, pH 7.2, with KOH. Pipette resistance was between 2 and 5 megaohms, and series resistance ranged between 2 and 10. The cells were held at −50 mV when the whole-cell configuration was obtained. The capacitance and series resistance were compensated using the auto function of Patchmaster software before recording the currents. Under voltage clamp, Kv4.2 channel currents were evoked by a brief prepulse to −80 mV followed by pulses of depolarization (Fig. 6C, left). Kv4.2 currents exhibited rapid activation and inactivation typical of A-type potassium currents, whereas endogenous Kv4.2 currents in HEK cells were relatively negligible; therefore, they were not quantified (Fig. 6C, right). To study the effect of forskolin on the peak amplitude of Kv4.2 channel current, a test pulse (+40 mV, 500 ms) was given every 10 s, and the peak current of each pulse was measured. After a baseline recording, forskolin (100 μm) was directly applied to the cell with an SF-77B rapid perfusion system (Warner Instruments, Inc., Hamden, CT). Percent inhibition of the peak Kv4.2 current by forskolin was expressed as (1 − I_test/I_baseline) × 100% where I_test is the peak current of each test pulse and I_baseline is the baseline amplitude (average of four pulses before forskolin application). All chemicals were purchased from Sigma. All data were collected with a HEKA EPC10 amplifier system (HEKA Instruments, Holliston, MA) and transferred to a personal computer using an ITC-16 digital-to-analog converter (HEKA Instruments). The signals were filtered at 2.9 kHz and digitized at 10 kHz using Patchmaster software (HEKA Instruments). Data were analyzed offline using Patchmaster. Voltages were not corrected for the liquid–liquid junction potential.

Immunoprecipitation

Two million HEK293T/17 cells/well were plated in a 6-well culture plate and transfected using Lipofectamine LTX and Plus reagent with a 1:1 ratio of GPR158 and Kv4.2myc plasmids. Cells were harvested 24 h later and lysed in 500 μl of ice-cold immunoprecipitation buffer (300 mm NaCl, 50 mm Tris-HCl, pH 7.4, 1% Triton X-100, complete protease inhibitor mixture) by sonication. Lysates were cleared by centrifugation at 14,000 × g for 15 min. 50 μl of supernatant were saved as input fraction, and the remaining 450 μl were incubated with 20 μl of Protein G beads (GE Healthcare) and 2 μg of antibodies on a rocker at 4 °C for 1 h. After three washes with immunoprecipitation buffer, proteins were eluted with 40 μl of 2× SDS sample buffer. 10 μl of each sample were analyzed by SDS-PAGE followed by Western blotting using horseradish peroxidase–conjugated secondary antibodies and an ECL West Pico (Thermo Scientific) detection system. Both lysates and immunoprecipitated fractions were run on the same gel followed by taking different exposures reported on different panels to accommodate for differences in protein content and resulting immunoreactivity.

Data analysis and statistics

A minimum of three animals were used in each group throughout the study. For statistical analyses, data were analyzed from individual cells. Statistical analysis was performed using GraphPad Prism (Prism 6.0, GraphPad, San Diego, CA) and SPSS Statistics 25 software (IBM). Student's t test was used to compare means between two groups, and one-way or two-way ANOVA followed by Bonferroni post hoc test was used to determine significant differences among multiple groups. Repeated-measures ANOVA was used when a dependent variable was measured multiple times such as intrinsic excitability, barium-sensitive current, and forskolin-mediated Kv4.2 current inhibition. Statistical tests were performed two-sided. Differences were considered significant if p was <0.05. All data are expressed as means ± S.E.