Abstract
Ambient GABA in the brain tonically activates extrasynaptic GABAA receptors, and activity-dependent changes in ambient GABA concentration can also activate GABAB receptors. To investigate an interaction between postsynaptic GABAB and GABAA receptors, we recorded GABAA currents elicited by exogenous GABA (10 μm) from dentate gyrus granule cells (DGGCs) in adult rat hippocampal slices. The GABAB receptor agonist baclofen (20 μm) enhanced GABAA currents. This enhancement was blocked by the GABAB receptor antagonist CGP 55845 and intracellular solutions containing the GTP analog GDP-β-s, indicating that baclofen was acting on postsynaptic GABAB receptors. Modulation of GABAA currents by postsynaptic GABAB receptors was not observed in CA1 pyramidal cells or layer 2/3 cortical pyramidal neurons. Baclofen reduced the frequency of spontaneous inhibitory postsynaptic currents (sIPSCs) but did not alter sIPSC amplitude or kinetics. Thus, GABAA receptors activated at synapses were not modulated by postsynaptic GABAB receptors. In contrast, tonic GABA currents and currents activated by the GABAA receptor δ subunit-selective agonist THIP (10 μm) were potentiated by baclofen. Our data indicate that postsynaptic GABAB receptors enhance the function of extrasynaptic GABAA receptors, including δ subunit-containing receptors that mediate tonic inhibition in DGGCs. The modulation of GABAA receptor function by postsynaptic GABAB receptors is a newly identified mechanism that will influence the inhibitory tone of DGGCs when GABAB and GABAA receptors are both activated.
Introduction
The inhibitory neurotransmitter γ-aminobutyric acid (GABA) activates both ionotropic GABAA receptors and metabotropic GABAB receptors. GABAA receptors are Cl− ion channels that produce electrical signals when activated. GABAA receptors respond transiently to GABA released from synaptic vesicles and, in many areas of the brain including the hippocampus, high-affinity GABAA receptors at extrasynaptic sites are activated tonically by ambient GABA (Farrant and Nusser, 2005; Glykys and Mody, 2007). Activation of presynaptic and postsynaptic GABAB receptors stimulates intracellular G-protein signaling cascades that activate K+ channels, inhibit voltage-gated Ca2+ channels, and regulate cyclic AMP (cAMP) and protein kinase A (PKA) (Padgett and Slesinger, 2010). Because postsynaptic GABAB receptors are located at extrasynaptic sites away from GABA release sites, their activation is limited by GABA uptake and requires patterns of presynaptic activity that lead to GABA spillover and elevations of ambient GABA (Scanziani, 2000; Kulik et al., 2003). Under conditions of increased ambient GABA, such as occur with ischemia, epileptic seizures, or drugs that increase GABA concentration, coactivation of GABAA receptors and postsynaptic GABAB receptors will occur (Scanziani et al., 1991; During and Spencer, 1993; Wu et al., 2003; Allen et al., 2004).
In dentate gyrus granule cells (DGGCs), electron microscopy with immunogold labeling has identified GABAB receptors at perisynaptic sites on dendritic and somatic membranes (Kulik et al., 2003), a distribution pattern that has remarkable overlap with the distribution of extrasynaptic GABAA receptor subunits that mediate tonic inhibition in DGGCs (i.e., δ subunits) (Wei et al., 2003). The proximity of postsynaptic GABAB receptors to extrasynaptic GABAA receptors on DGGCs suggests that GABAA receptors will be exposed to intracellular signaling pathways activated by GABAB receptors. This potential interaction has likely been overlooked, because studies of GABAA receptors are routinely done in the presence of GABAB receptor antagonists.
We investigated the interaction between GABAB receptors and GABAA receptors in DGGCs. Our data show that activation of postsynaptic GABAB receptors enhances GABAA currents caused by exogenous GABA. This newly identified interaction was not present in CA1 pyramidal neurons or layer 2/3 cortical pyramidal neurons. In DGGCs, tonic GABA currents and currents mediated by δ subunit-containing receptors were also modulated by GABAB receptor activation. Our results indicate that extrasynaptic GABAA receptor function will be enhanced when postsynaptic GABAB receptors are activated, increasing the inhibitory tone of DGGCs.
Materials and Methods
Brain slice preparation.
Brain slices were prepared from 4–6 week old Sprague Dawley rats of both sexes. Rats were anesthetized with 4% isoflurane, decapitated, and the brain dissected free. Transverse hippocampal slices (300 μm) were prepared. These slices contained portions of temporal cortex that were used for experiments on cortical neurons. Slices were cut and stored in a solution containing (in mm): 125 NaCl, 3 KCl, 26 NaHCO3, 1.2 NaH2PO4, 0.5 CaCl2, 4 MgCl2, 20 dextrose, and 1 kynurenic acid. Slices were cut in ice-cold solution and stored at room temperature. Solutions were continuously gassed with 95% O2/5% CO2. Slices were allowed to recover for 1 h before recording. All animal use protocols were approved by the local Institutional Animal Care and Use Committee.
Electrophysiology.
Membrane currents were recorded using whole-cell patch clamp techniques. Neurons were visualized with an Axioskop 2 upright microscope with fixed stage (Carl Zeiss). Recordings were made using an Axopatch 200B amplifier, a Digidata 1200 series A-D converter, and pClamp 9 software (Molecular Devices). Data were acquired at 2 kHz and low-pass filtered at 1 kHz. Series resistance was compensated by 50–70% online. If series resistance exceeded 20 MΩ or changed by >20%, the experiment was discarded. Focal application of GABA or bicuculline was made by pressure ejection (Picospritzer II, General Valve) from a patch pipette containing (in mm): 150 NaCl, 3 KCl, 2 CaCl2, 2 MgCl2, 10 dextrose, and 10 HEPES with pH adjusted to 7.4 with NaOH. The pressure ejection pipette was positioned ∼20–30 μm from the soma. The recording chamber was continuously superfused at ∼2–2.5 ml/min with a bath solution containing (in mm): 134 NaCl, 3 KCl, 1.4 NaH2PO4, 24 NaHCO3, 10 dextrose, 2 MgCl2, 2 CaCl2, and 1 kynurenic acid (pH 7.35–7.4 when bubbled with 95% O2/5% CO2). Osmolarity was adjusted to 300–305 mOsm with H2O. Patch pipettes were pulled from borosilicate glass with filament (Sutter Instrument). Pipettes had resistances of 3–4 MΩ when filled with intracellular solution containing (in mm): 125 CsCl, 10 QX-314 chloride, 10 HEPES, and 1 EGTA (pH corrected to 7.25 with CsOH). Osmolarity was adjusted to 275–285 mOsm with H2O as needed. Data acquisition was started 4–6 min after establishing a whole-cell recording. Experiments were performed at room temperature (∼23°C). All chemicals were purchased from Sigma except SKF 89976a (Tocris-Cookson) and QX-314 (Alomone Labs). Baclofen was used at 20 μm (Dutar and Nicoll, 1988), and CGP55845 was used at 10 μm (Chen and Regehr, 2003).
Analysis.
Data analysis was performed with Clampfit (pClamp 10) and Origin (v6.1, Microcal Software) software. Tonic currents were measured as the change in holding current caused by the GABAA receptor antagonist bicuculline methiodide (40 μm), based on Gaussian fits to all-points current amplitude histograms constructed from 2–10 s of data with a bin width of 1 pA. Fits were performed using a Levenberg–Marquardt algorithm in Clampfit, and the holding current was taken as the center of the Gaussian curve. IPSCs were analyzed using template matching event detection. Data are presented as mean ± standard error of the mean (SEM), and all error bars represent SEM. Statistical analyses were performed using Microsoft Excel. A two-tailed, paired or homoscedastic Student's t test was used with a p value ≤ 0.05 considered as significant.
Results
GABAB receptor activation enhanced GABAA receptor currents in a cell type-specific manner
To investigate an interaction between metabotropic GABAB receptors and ionotropic GABAA receptors, we made focal applications of GABA (10 μm) to activate GABAA currents in DGGCs (Vm = −60 mV). Bath application of the GABAB receptor agonist baclofen (20 μm) increased GABA-evoked currents by 83 ± 18% (control, −358 ± 143 pA vs baclofen, −577 ± 184 pA, n = 5, p < 0.01) (Fig. 1A). Currents in the presence of baclofen were inhibited by bicuculline (20 μm) to a similar extent as under control conditions (control, 91 ± 2% inhibition, baclofen, 95 ± 1% inhibition, n = 3–4, p = 0.26), indicating that currents enhanced by baclofen were also mediated by GABAA receptors (data not shown). Modulation of GABAA currents by baclofen was blocked by the GABAB receptor antagonist CGP55845 (CGP, 10 μm). In the presence of CGP, GABAA currents were −521 ± 120 pA at baseline and −486 ± 138 pA after application of baclofen (n = 5, p = 0.38) (Fig. 1B). These data indicate that activation of GABAB receptors augments GABAA receptor function in DGGCs. CGP by itself reduced GABAA currents, demonstrating baseline activation of GABAB receptors (and modulation of GABAA receptors) during application of GABA. On average, CGP alone reduced GABAA currents from −309 ± 87 pA to −187 ± 62 pA (n = 5, p < 0.05) (Fig. 1C).
We tested the effects of baclofen on GABA-evoked currents in hippocampal CA1 pyramidal cells and cortical layer 2/3 pyramidal cells. In contrast to results from DGGCs, baclofen did not modulate GABAA currents in these other cell types (CA1: control −669 ± 173 pA vs baclofen −647 ± 163 pA, n = 4, p = 0.19; Layer 2/3: control −457 ± 103 pA vs baclofen −445 ± 96 pA, n = 6, p = 0.28) (Fig. 1D).
Modulation of GABAA current required postsynaptic G-protein activation
Bath-applied baclofen activates both presynaptic and postsynaptic GABAB receptors, raising the possibility of an indirect action (e.g., via modulation of GABA release). Because GABAB receptors are G-protein coupled, transduction of receptor activation requires guanine nucleotide exchange. This process involves dissociation of GDP from inactive G proteins and binding of GTP. Thus, postsynaptic GABAB receptor signaling can be prevented by intracellular GDP-β-S, a nonhydrolyzable GDP analog (Harayama et al., 1998; Lin and Dun, 1998). We recorded GABAA currents from DGGCs with intracellular solutions containing GDP-β-S (0.5 mm) and no GTP (Fig. 2A). Intracellular GDP-β-S prevented baclofen-induced potentiation of GABAA currents (control, −615 ± 141 pA vs baclofen, −532 ± 118 pA, n = 5, p = 0.12) (Fig. 2A). These results indicate that baclofen modulates GABAA currents via postsynaptic GABAB receptors and a G-protein signaling pathway.
Modulation of GABAA current was independent of synaptic vesicle release
Although changes in GABA release were not predicted to affect responses to exogenous GABA, we wished to exclude a contribution of vesicular GABA release to the effects of baclofen. To address this issue, we inhibited vesicular release of GABA by pretreating slices with the H+-ATPase inhibitor concanamycin A (ConA, 0.5 μm, 2 h). ConA treatment reduced the frequency of spontaneous IPSCs (sIPSCs) by 75% (n = 9, p < 0.05) (data not shown). However, ConA treatment did not affect the baclofen-induced potentiation of GABAA currents (control, −437 ± 157 pA vs baclofen, −974 ± 167 pA, n = 5, p < 0.01) (Fig. 2B).
Baclofen potentiated tonic GABAA receptor currents, but not synaptic GABAA receptor currents
The data presented above show that postsynaptic GABAB receptors can potentiate GABAA currents elicited by exogenous GABA. Distinct types of GABAA receptors with unique subunit compositions are transiently activated at synapses by vesicular GABA release or tonically activated by ambient GABA at extrasynaptic sites (Farrant and Nusser, 2005). To determine whether extrasynaptic GABAA receptors were subject to modulation by postsynaptic GABAB receptors, we measured tonic GABAA current as the change in holding current caused by focal application of bicuculline (40 μm). Similar to its effect on exogenous GABA currents, baclofen potentiated tonic currents (control, −7.5 ± 1 pA vs baclofen, −15.2 ± 0.1 pA, n = 5, p < 0.01) (Fig. 3A,B). Tonic currents were unaffected by baclofen in the presence of CGP (CGP, −5.5 ± 1.2 pA vs baclofen/CGP, −4.8 ± 0.9, p = 0.44, n = 5) (Fig. 3C). In contrast to GABA-evoked currents, CGP alone did not affect tonic currents (control, −13.6 ± 3.0 pA vs CGP, −15.3 ± 3.4, n = 4, p = 0.17) (Fig. 3C). This indicates that GABAB modulation of tonic currents is not basally active and requires periods of increased ambient GABA (such as periodic application of exogenous GABA, i.e., Fig. 1C). Following inhibition of GABA uptake by SKF 89976a (SKF, 30 μm) to increase ambient GABA and tonic current, blocking GABAB receptors with CGP significantly reduced tonic current (from −86.8 ± 17.9 pA to −45.9 ± 9.6 pA, n = 5, p < 0.01) (Fig. 3D,E). These data indicate that GABAB receptors modulate GABAA receptors that mediate tonic currents.
Tonic currents in DGGCs are primarily mediated by δ subunit-containing GABAA receptors (Glykys et al., 2008). To confirm that baclofen potentiates currents produced by δ subunit-containing GABAA receptors, we used the δ subunit-selective agonist THIP (10 μm). Baclofen increased currents evoked by THIP (control, −103 ± 19 pA vs baclofen, −167 ± 32 pA, n = 5, p < 0.05) (Fig. 4A). In the presence of CGP, baclofen had no effect on THIP-induced currents (CGP, −48 ± 13 pA vs baclofen/CGP, −46 ± 14 pA, p = 0.56, n = 5) (Fig. 4B). Similar to tonic currents, CGP alone did not affect THIP currents (control, −85.5 ± 18.9 pA vs CGP, − 88.0 ± 19.3 pA, n = 4, p = 0.50) (Fig. 4B). These results demonstrate that δ subunit-containing GABAA receptors are a target of postsynaptic GABAB receptors.
Measurements of sIPSCs showed that baclofen reduced their frequency to 52 ± 6% of control values (n = 5, p < 0.05) but did not significantly affect sIPSC amplitude, 10–90% rise times, or decay times (n = 5, p = 0.94, 0.44, and 0.30, respectively) (data not shown). CGP alone did not affect sIPSC frequency, amplitude, or decay times (p = 0.53, 0.52, and 0.33, respectively; n = 4). These data indicate that GABAB receptors reduce presynaptic release of GABA but do not alter the properties of GABAA receptors activated at synapses (Otis and Mody, 1992).
Discussion
Our results show for the first time that postsynaptic GABAB receptors can modulate GABAA receptor function. Specifically, our data indicate that postsynaptic GABAB receptors enhance the function of GABAA receptors that produce tonic currents in DGGCs (including δ subunit-containing receptors). Because postsynaptic GABAB receptors and the GABAA receptors they modulate are located extrasynaptically (Kulik et al., 2003; Wei et al., 2003), both types of receptors will experience similar levels of ambient GABA. Coactivation of these receptors during periods of increased ambient GABA associated with intense neural activity may represent a feedback mechanism to increase inhibitory tone of DGGCs (Scanziani et al., 1991; During and Spencer, 1993; Scanziani, 2000).
Mechanism of GABAB receptor modulation of GABAA currents
Several lines of evidence indicate that GABAA currents were modulated by postsynaptic GABAB receptors. The use of exogenous GABA to activate GABAA currents would minimize effects of altered GABA release due to presynaptic GABAB activation, and GABAB effects were seen after inhibition of vesicular GABA release with ConA. Altered GABA uptake caused by presynaptic GABAB receptor activation could potentially account for our observations. However, this is unlikely because GABAB effects were seen in the presence of the GAT1 antagonist SKF. Additionally, currents produced by the nontransported GABAA agonist THIP were potentiated by baclofen, indicating that GABAB receptors enhance GABAA currents independently of GABA uptake mechanisms (including GAT3). Finally, selectively inhibiting postsynaptic G proteins by including GDP-β-S in our pipette solutions blocked the effects of GABAB receptor activation, confirming a postsynaptic site of action.
Our data show that tonic currents caused by ambient GABA are modulated by postsynaptic GABAB receptors. Experiments using the δ subunit-selective agonist THIP (Mortensen et al., 2010; Meera et al., 2011) confirmed that receptors containing δ subunits were potentiated by GABAB receptors. GABAB receptor activation did not change sIPSC amplitude or kinetics but reduced sIPSC frequency, indicating that postsynaptic GABAB receptors did not modulate GABAA receptors activated at synapses. Because a large number of “synaptic-type” GABAA receptors (i.e., γ subunit containing) are located away from synapses in extrasynaptic membranes (near postsynaptic GABAB receptors) (Farrant and Nusser, 2005), our data do not exclude the possibility that GABAB receptors modulate these types of GABAA receptors in addition to δ subunit-containing receptors.
Signals mediated by postsynaptic GABAB receptors are transduced by the G proteins Gi/Go and Gβγ (Padgett and Slesinger, 2010). A downstream effect of Gβγ is K+ channel activation; however, this is unlikely to have contributed significantly to our results because control currents and currents in the presence of baclofen were both inhibited >90% by GABAA antagonists, and intracellular QX314 effectively blocks GABAB-activated K+ channels (Nathan et al., 1990; Andrade, 1991). Activation of Gi/Go inhibits adenylate cyclase, with subsequent reduction in cAMP levels and reduced activation of PKA. Our data show that inhibition of G-protein signaling prevented GABAB receptor modulation of GABAA currents. The final effect on GABAA receptors is not known, but the possibilities include increased single channel conductance, open probability, or surface expression. Future studies using different experimental methods are required to distinguish between these possibilities.
Cell type specificity of the postsynaptic GABAB–GABAA interaction
In DGGCs, activation of GABAB receptors substantially increased GABAA currents by 83% on average. This strong modulation was absent in CA1 pyramidal neurons and cortical layer 2/3 pyramidal neurons. One salient difference between DGGCs and CA1 pyramidal cells is that tonic inhibition in CA1 pyramidal cells is mediated by α5 subunit-containing GABAA receptors, whereas δ subunit-containing receptors are dominant in DGGCs (Caraiscos et al., 2004; Glykys et al., 2008). In layer 2/3 pyramidal neurons there is little, if any, functional expression of δ subunits (although δ subunit mRNA is present) (Yamada et al., 2007). Thus, we speculate that signals produced by postsynaptic GABAB receptors may preferentially affect δ subunits. Our results showed that baclofen did not affect sIPSC characteristics but potentiated tonic currents and THIP-evoked currents, consistent with a selective modulation of δ subunit-containing receptors by postsynaptic GABAB receptors. However, other factors in addition to differences in GABAA receptor subunit expression may also contribute to the cell type specificity of GABAB receptor effects.
Functional implications
Tonic inhibition affects neuronal excitability and network behavior (Chadderton et al., 2004; Glykys and Mody, 2006; Pavlov et al., 2009; Duguid et al., 2012), thereby influencing many physiologic and pathophysiologic processes, including synaptic plasticity and epileptic seizures (Maguire et al., 2005; Martin et al., 2010). The effect of tonic inhibition is generally inhibitory, but in some cell types (i.e., thalamocortical cells) increasing tonic inhibition is maladaptive and contributes to pathologic patterns of neuronal firing (Cope et al., 2009). Our results add to the ways that tonic inhibition can be modulated, including GABA uptake and release by transporters, drugs that increase GABA concentrations, voltage-dependent modulation of extrasynaptic GABAA receptors, and neurosteroids (Overstreet and Westbrook, 2001; Nusser and Mody, 2002; Stell et al., 2003; Wu et al., 2003; Pavlov et al., 2009; Ransom et al., 2010). The enhancement of tonic inhibition by postsynaptic GABAB receptors described here is predicted to reduce cellular excitability of DGGCs by shunting excitatory synaptic currents and raising action potential threshold (Stell et al., 2003; Holter et al., 2010; Arima-Yoshida et al., 2011; Gupta et al., 2012).
Footnotes
This work was supported by a Veterans Affairs (VA) Merit Review Award (W.J.S.), a VA Career Development Award (C.B.R.), and the VA Epilepsy Centers of Excellence (C.B.R., W.J.S., M.H.H.). The authors appreciate the expert technical assistance of Sue Usher and discussions with Professor Bertil Hille during the course of experiments.
- Correspondence should be addressed to Dr. Christopher B. Ransom, Department of Neurology, University of Washington, Epilepsy Center of Excellence, VA Puget Sound, 1600 South Columbian Way, Seattle, WA 98108. cbr5{at}uw.edu