Tonic GABAergic inhibition, via GABAA receptors containing αβƐ subunits, regulates excitability of ventral tegmental area dopamine neurons

Abstract The activity of midbrain dopamine neurons is strongly regulated by fast synaptic inhibitory γ‐Aminobutyric acid (GABA)ergic inputs. There is growing evidence in other brain regions that low concentrations of ambient GABA can persistently activate certain subtypes of GABAA receptor to generate a tonic current. However, evidence for a tonic GABAergic current in midbrain dopamine neurons is limited. To address this, we conducted whole‐cell recordings from ventral tegmental area (VTA) dopamine neurons in brain slices from mice. We found that application of GABAA receptor antagonists decreased the holding current, indicating the presence of a tonic GABAergic input. Global increases in GABA release, induced by either a nitric oxide donor or inhibition of GABA uptake, further increased this tonic current. Importantly, prolonged inhibition of the firing activity of local GABAergic neurons abolished the tonic current. A combination of pharmacology and immunohistochemistry experiments suggested that, unlike common examples of tonic inhibition, this current may be mediated by a relatively unusual combination of α4βƐ subunits. Lastly, we found that the tonic current reduced excitability in dopamine neurons suggesting a subtractive effect on firing activity.

There is growing evidence in other brain regions that low concentrations of ambient GABA can also persistently activate certain subtypes of GABA A receptor, located extrasynaptically, to generate a 'tonic' current (Brickley & Mody, 2012;Farrant & Nusser, 2005;Semyanov et al., 2004). Tonic activation of GABA A receptors can increase input conductance which is commonly hyperpolarizing and can lead to shunting of excitatory inputs, therefore playing an important role in regulating neuronal activity. Extrasynaptic GABA A receptors often differ from their synaptic counterparts by exhibiting activation to low concentrations of GABA as well as markedly slower rates of desensitization (Brickley et al., 1996;Bright et al., 2011). The properties of extrasynaptic GABA A receptors are conferred by receptor subunit composition, most commonly the inclusion of α4, α5 or α6, with β and δ subunits (Brickley et al., 2001;Caraiscos et al., 2004;Chandra et al., 2006). Furthermore, the properties of extrasynaptic GABA A receptors are often conferred by their subunit composition differing from their synaptic counterparts. For example, the inclusion of the δ subunit is commonly seen in extrasynaptic receptors, while postsynaptic GABA A receptors typically contain γ subunits (Essrich et al., 1998). Indeed, VTA GABA neurons exhibit a tonic current mediated by δ subunit containing receptors (Vashchinkina et al., 2012(Vashchinkina et al., , 2014. Of the remaining possible GABA A receptor subunits (i.e. ε, θ, π and ρ1-3), only Ɛ (Moragues et al., 2000(Moragues et al., , 2002Sinkkonen et al., 2000) and θ (Bonnert et al., 1999;Moragues et al., 2002;Sinkkonen et al., 2000) are expressed in the midbrain. Θ subunit is highly selective with subunit pairing to make functional GABA A receptors, and it can recruit γ subunits or Ɛ subunits (Bonnert et al., 1999;Ranna et al., 2006). In contrast, Ɛ subunits can form functional GABA A receptors with αβ subunits which generate tonic GABAergic current (Neelands et al., 1999). Although exogenous application of GABA can induce a tonic current in VTA dopamine neurons (Vashchinkina et al., 2014), it is not clear if dopamine neurons exhibit an endogenous extrasynaptic tonic GABAergic current. We, therefore, sought to examine this by conducting whole-cell recording from VTA dopamine neurons in acute brain slices from mice, along with immunohistochemical analysis of GABA A receptor subunits.

| Animal maintenance and breeding
Male Pitx3-GFP [B6.129P2-Pitx3 tm1MLi /Mmjax, MMRRC: 41479] (Zhao et al., 2004), TH-GFP [Tg(Th-EGFP)21-31 Koba, MGI: 3774290] (Matsushita et al., 2002;Sawamoto et al., 2001) heterozygous mice and C57BL/6J mice were used for experiments. All breeding and experimental procedures were conducted in accordance with the Animals (Scientific Procedures) Act of 1986 (UK) and approved by Imperial College London's Animal Welfare and Ethical Review Body. All mice were maintained in social groups of 2-4, where possible, with appropriate environmental enrichment (e.g., bedding and tunnels). Mice were kept in 12 hr light/dark cycle with a free access to standard chow and water ad libitum.

| Electrophysiology
Pitx3-GFP mice were used for all voltage-clamp recordings, and a mixture of Pitx3-GFP and C57BL/6J mice were used for current-clamp recordings. Acute brain slices were transferred to a slice recording chamber (Scientifica, UK) and were continuously perfused at a rate of 3-5 ml/min with fully oxygenated aCSF at 34°C. Neurons were visualized using infra-red differential interference contrast (IR-DIC) under an upright microscope (Olympus BXWI 51, Japan) equipped with a 40x objective (0.8 numerical aperture), an IR filter, DIC optics and a charge coupled device (CCD) video camera (Watec 902H). Dopamine neurons were identified in Pitx3-GFP mice by the presence of GFP (viewed under fluorescence illumination (Xcite120 unit; EXFO, UK) coupled to a GFP excitation filter) or in C57BL/6J mice by the presence of an Ih current in neurons located in the lateral parabrachial pigmented area (PBP) of the VTA (Ungless & Grace, 2012).
For voltage-clamp recording, neurons were clamped at −70 mV. α-amino-3-hydroxy-5 methyl-4-isoxazolepropionic acid (AMPA), N-methyld-aspartate (NMDA), and GABA B receptors were routinely blocked using NBQX (25 μM), D-AP5 (50 μM), and CGP-52432 (5 μM), respectively. Miniature inhibitory postsynaptic currents (mIPSCs) were recorded in the presence of 1 μM tetrodotoxin (TTX). To determine average holding currents, a continuous current recording was measured for 60 counts/1 s epoch, discarding those coincident with spontaneous IPSCs (sIPSCs) or mIPSC. A drug effect on the holding current was obtained after at least 10 min of drug application. Tonic GABA A currents (I tonic ) were defined as the difference between the average holding current (I hold ) in control (Baseline) versus that obtained during drug application and reported as ΔI hold . Frequency, amplitude, and decay time constants of mIPSCs were analysed off-line with the MiniAnalysis (Synaptosoft). Average IPSC waveforms were constructed using at least 100 IPSCs.
For current-clamp recordings, neurons were recorded in current-clamp conditions for a minimum of 1-2 min after breaking into the whole-cell mode before delivering depolarizing pulses. A series of hyperpolarization steps were initially applied to identify VTA dopamine neurons by the presence of an Ih current. For excitability experiments, neurons were clamped at −60 mV and step current injections of increasing amplitude from 0 to 60 pA (5 pA steps, 500 ms duration) were used to evoke action potentials. To examine the inputoutput relationship of the neurons, the injected current was plotted against the number of action potentials obtained.

| Data acquisition and analysis
Membrane currents were low-pass filtered at 5 kHz, digitized at 20 kHz, and analysed using pClamp 9.2 software (Molecular Devices). Access and input resistances were monitored throughout the experiments using a 5 mV voltage step. The access resistance was typically <20 MΩ, and results were discarded if it changed by more than 20%. Membrane capacitance (Cm) was measured under voltage clamp at −50 mV using a hyperpolarizing 10 mV, 250 ms step. Cm was measured from the change in membrane charge taken from the integrated capacity transients (pClamp, Molecular Devices). All potentials cited here have not been corrected for liquid junction potentials (9.2 mV, pClamp).

| Immunohistochemistry
Pitx3-GFP, TH-GFP and C57BL/6J mice (8-10 weeks old) were anaesthetized with pentobarbiturate (Euthatal), and transcardially perfused in an antibody-specific manner (see below for specific protocols). Brains were removed and post-fixed in the adequate fixative solution. Coronal sections were prepared with a cryostat at a thickness of 40 μm covering Bregma −3.0 to −3.65 mm and were collected as freefloating. Expression in the hippocampal area was used for the validation of successful immunostaining for each subunit (Drexel et al., 2015;Kralic et al., 2006). 2.5.1 | α1/α2/α4 subunits Mice were perfused via the ascending aorta with ice cold Tris-HCl buffer (pH7.4) for 2 min, immediately followed by 4% paraformaldehyde (PFA)/Tris buffer for 5 min. The brain was extracted immediately after decapitation and post-fixed for 1h. After rinsing with Tris buffer, the brain was stored at −80°C after 30% sucrose/Tris buffer cryoprotection until processing them further. Sections were briefly incubated in 0.3% hydrogen peroxide/Tris buffer for 5 min, then blocked in Tris buffer/0.2% Triton X-100 /20% normal donkey serum (NDS) for 30 min. Sections were then moved and incubated in primary antibody solution (Tris buffer/0.2% Triton X-100/2% NDS, Table 1) for 24 hr at 4°C. After several washes in Tris buffer, sections were incubated in secondary antibody solution (Tris buffer/0.2% Triton X-100/2% NDS, Table 2) for 1 hr at room temperature.

| α3/θ subunits
Mice were perfused and processed in the same manner as for α1/α2/α4 subunit staining and cryoprotected and stored for the same duration as previously described. Sections were briefly incubated in 0.3% hydrogen peroxide/Tris buffer for 5 min, and blocked in Tris buffer/5% skimmed milk /20% NDS for 30 min. Sections were then moved and incubated in primary antibody solution (Tris buffer/5% skimmed milk/2% NDS, Table 1) for 24 hr at 4°C. After several washes in Tris buffer, sections were incubated in secondary antibody solution (Tris buffer/5% skimmed milk/2% NDS, Table 2) for 1 hr at room temperature.

| ε subunit
Mice were perfused with ice cold oxygenated aCSF for 2 min, immediately followed by ice-cold 4% PFA/Tris-HCl buffer (pH7.4) for 5 min via the ascending aorta. The brain was extracted immediately after decapitation and blocks of midbrain were isolated and immersed in ice-cold fixative for 90 min. The block of brain was stored at −80°C after 30% sucrose/Tris buffer cryoprotection until processing them further. Sections were blocked in Tris buffer/0.2% Triton X-100/20% NDS for 30 min at RT. Sections were then moved and incubated in primary antibody solution (Tris buffer/0.2% Triton X-100/2% NDS, Table 1) for 48 hr at 4°C. After several washes in Tris buffer, sections were incubated in secondary antibody solution (Tris buffer/0.2% Triton X-100/2% NDS, Table 2) for 1 hr at room temperature.

| Microscopy
Images were captured with a SP5 (Leica, Germany) confocal microscope with 63 × 1.40 oil-immersion objective (1.4 NA) and the pinhole was set at 1 Airy unit. Acquisition of triple labelling was taken sequentially. All images were analysed and prepared with Fiji software using Ortho-slice extractor and cell counter. All pixels below the fixed threshold were removed first. All signals in the cytoplasm were omitted after consulting orthogonal views from the counting. For the quantitative analysis, 3-4 images were taken from the lateral PBP area from at least three brain slices spanning the anterior to posterior axis for each brain. The lateral PBP was identified by the position of the medial terminal nucleus (MT), the medial lemniscus (ml) and TH expression in the SNc. Example images were background-subtracted, and a low pass filer was applied. The images were shown using the maximal intensity Z projection mode (a stack of 11 confocal images spaced at 0.5 µm for × 2.5 digital zoom (× 100 total magnification), and six confocal images spaced at 0.3 µm for × 6 digital zoom (× 240 total magnification)).

| Statistical analyses
Data are presented as means ± SEM (standard error of the mean). n = sample number, N = animal number, Statistical comparisons were made using paired Student's t test or twoway ANOVA (with Šídák' post hoc tests), and performed using Prism (Graphpad).

GABAergic current
To explore a tonic GABAergic current in neuronal subtype specific manner, we used Pitx3-GFP mice which selectively express green fluorescent protein (GFP) in VTA dopamine neurons (Zhao et al., 2004). In order to magnify GABA A receptor-mediated chloride currents, we chose to use the internal pipette solution containing a high concentration of chloride and hold cell membrane at −70 mV to observe inward current. Ionotropic glutamate (AMPA/kainite and NMDA) and GABA B receptors were routinely blocked by adding NBQX (25 µM), D-AP5 (50 µM) and CGP-52432 (5 µM), respectively, to the extracellular solution. We first examined whether VTA dopamine neurons expressed a tonic GABAergic conductance by measuring the change in baseline holding current (I hold ) produced by blocking GABA A receptors with the non-competitive GABA A receptor antagonist, picrotoxin (PTX, 100 µM). We calculated the tonic GABA A receptor-mediated current (I tonic ) as the change in holding current from baseline (I hold ) after a 5 min application of picrotoxin. In addition to abolishing all spontaneous inhibitory postsynaptic currents (sIPSCs), we found that PTX significantly decreased I hold in VTA dopamine neurons (Figure 1a-c, t 7 = 4.395, p = 0.00632, n = 8 [N = 6]), indicative of a tonic GABAergic conductance. Application of bicuculline (BIC, 10 µM), a competitive GABA A receptor antagonist, also abolished sIPSCs and decreased I hold in VTA dopamine neurons (Figure 1b,c; t 7 = 2.502, p = 0.0409, n = 7 [N = 6]). Taken together, these results indicate that VTA dopamine neurons exhibit an endogenous background tonic GABA A receptor-mediated current.

| Tonic GABA A current in VTA dopamine neurons is sensitive to exogenous GABA concentration and maintained by VTA GABA firing neuron activity
The presence of a tonic conductance suggests that GABA must be present in the extracellular space at a sufficiently high concentration to cause persistent receptor activation and/or that these receptors open spontaneously. Identifying the source, or sources, of GABA is important for understanding how this tonic receptor activation is modulated. In some neuronal populations, such as in cerebellar granule cells, action potential-dependent vesicular release plays an important role (Brickley et al., 1996;Kaneda et al., 1995;Wall & Usowicz, 1997). However, some non-vesicular sources of GABA have also been suggested, including astrocytic release (Kozlov et al., 2006;Liu et al., 2000) and reversal of GABA transporters (Gaspary et al., 1998;Richerson and Wu, 2003). To explore this in VTA dopamine neurons, we firstly examined whether globally increasing GABA could affect this tonic current. Nitric oxide has been shown to increase GABA release at several synapses through potentiating presynaptic GABA release in either a Ca 2+ -dependent manner or by reversing GABA transporters (Li et al., 2002;Yu & Eldred, 2005). Moreover, the global increase in GABA release that occurs with nitric oxide signalling has been shown to selectively enhance the tonic GABA conductance in thalamic neurons (versus manipulations that increase local GABA release; Bright and Brickley, 2008). The nitric oxide donor, S-nitroso-N-acetylpenicillamine (SNAP), has been shown to potentiate IPSCs in dopamine neurons (Nugent et al., 2007) and could therefore be involved in modulation of the tonic current. We found that application of SNAP (400 µM) produced a dramatic increase in I hold (Figure 2a GABA transporters have well-documented effects on phasic inhibition, but they also have an important and dynamic influence on ambient GABA, and thus the amount of GABA available to activate extrasynaptic GABA A receptors that mediate tonic currents. For example, in dentate gyrus granule cells (Nusser & Mody, 2002) and hippocampal interneurons (Semyanov et al., 2003) the GABA uptake system can have a profound influence on the amount of tonic current. Astrocytic GABA transporters were also reported to have a possible role in regulating the extracellular GABA during excessive GABA release by the elevated network activity (Kersanté et al., 2013). This suggests that in acute brain slice preparations, GABA uptake may keep the ambient GABA concentration sufficiently low to prevent the activation of GABA A receptors or to allow activation of only a specific subset of receptors such as the high affinity, non-desensitizing receptors often involved in generating tonic currents. It is possible, therefore, that inhibition of GABA uptake could be used to selectively modulate the tonic current in VTA dopamine neurons. To test this, we used the specific GAT-1 GABA transporter blocker, NO711, to examine changes in I hold . We found that NO-711 (20 µM) increased I hold (Figure 2a We next wanted to examine the possibility that spontaneous activity of VTA GABA neurons may be the source of the tonic current in dopamine neurons. Extracellular GABA level decreases by 70% when action potentials were blocked for more than 20min (De Groote & Linthorst, 2007;Rowley et al., 1995). Here, we found that the prolonged incubation . These results suggest that the endogenous tonic current is a result of spontaneous firing activity of VTA GABA neurons. Together with results from NO-711 and SNAP application, it also suggests that tonic current requires a decent amount of GABA spillover, while the GABA transporter in both neurons and astrocytes are actively regulating the tonic current under basal condition.

| γ and δ subunit containing GABA A receptors do not contribute to the tonic current observed in VTA dopamine neurons
GABA A receptors that contain γ subunits are generally found in the postsynaptic GABA A receptor and importantly are relatively insensitive to Zn 2+ (Essrich et al., 1998;Smart et al., 1991;Stórustovu & Ebert, 2006). We found that Zn 2+ (20 µM) inhibited the tonic current (Figure 3a . These results suggest that the endogenous tonic current-mediating GABA A receptors do not contain the γ subunit and are located extrasynaptically, while postsynaptic GABA A receptors, which are responsible for mIPSCs, contain the γ subunit in VTA dopamine neurons. It has been shown that the δ subunit does not play a role in the tonic current induced by exogenous application of GABA in VTA dopamine neurons (or indeed in synaptic currents in these neurons) but does play a role in VTA GABA neurons (Vashchinkina et al., 2014). Consistent with this, even when we applied a relatively high concentration of THIP (4,5,6,7-te trahydroisoxazolo[5,4-c]pyridine-3-ol, 5 µM), a δ-preferring GABA A receptor agonist, we observed a very small change in I hold in VTA dopamine neurons (Figure 3e,f; t 4 = 3.003, p = 0.0398, n = 5 [N = 5]), suggesting that δ subunit containing receptors do not play a role in tonic inhibition in VTA dopamine neurons.

| Evidence that Ɛ subunit containing GABA A receptors are located extrasynaptically in VTA dopamine neurons
Electrophysiology results suggest a number of possibilities regarding the subunit composition of the receptors mediating this tonic current in VTA dopamine neurons. One possibility is that they contain only α and β subunits. However, THIP, which had only a small effect on dopamine neurons (Figure 3e,f), has been shown to act as an agonist at αβ receptors (Stórustovu & Ebert, 2006). Moreover, αβ receptors exhibit spontaneous opening, whereas we did not observe a tonic current in the absence of endogenously released GABA (i.e., in the presence of TTX).
Of the remaining possible GABA A receptor subunits that may be present, only ε and θ subunits are found to be expressed in midbrain. Previous reports suggested that both Ɛ subunits and θ subunits are expressed at high levels in the SNc, while not detected in the VTA (Bonnert et al., 1999;Moragues et al., 2000;Okada et al., 2004). However, immunostaining in the VTA can be challenging, and we therefore hypothesized that it may be possible to detect Ɛ and/or θ subunits in the VTA using methods optimized for receptor subunit immunostaining (Notter et al., 2014). Using this approach, however, we observed limited θ subunit expression in dopamine neurons in the VTA (Figure 4a(i,ii),d; % TH+/ Total subunit signals; 33.07% ± 2.70, n = 11 [N = 4]). In contrast, we observed that Ɛ subunits expression was colocalized with dopamine neurons in the VTA (Figure 4b,d; TH+/Total subunit signals; 60.02% ± 3.16, n = 13 [N = 4]). In addition, it was generally not colocalized with gephyrin, a postsynaptic marker of GABAergic synapses (Essrich et al., 1998;Fritschy et al., 2008;Kneussel & Loebrich, 2007) (Figure 4c; Co-localization/Gephyrin (TH+); 2.77% ± 0.73, n = 13 [N = 4]), suggesting that it may be located extrasynaptically, consistent with a role in tonic inhibition.

| Tonic GABA currents shunt depolarizing inputs in VTA dopamine neurons to reduce excitability
Finally, we wanted to examine the functional consequences of this tonic inhibition on dopamine neuron firing activity. Although in other neuronal types the functional consequences of tonic inhibition are not fully understood, one straightforward outcome is a persistent increase in the cell's input conductance. This affects the magnitude and duration of the voltage response to an injected current and increases the decrement of voltage with distance. Consequently, for a given excitatory input, the size and duration of the excitatory postsynaptic potential will be reduced, and the temporal and spatial window over which signal integration can occur will be narrowed, making it less likely that an action potential will be generated. Indeed, several studies have shown that blockage of tonic inhibition with GABA A antagonists decreases the current that is required to achieve a given firing rate | 1731 TOSSELL ET aL.
(i.e., input-output curves are shifted; Brickley et al., 1996;Hamann et al., 2002;Semyanov et al., 2003). It is also possible that shunting inhibition can change the slope of the input-output relationship and hence modulate neuronal gain; that is, it can alter the sensitivity of a neuron to changes in the excitatory input rate (Mitchell & Silver, 2003). Given that the tonic current in VTA dopamine neurons can be decreased by blocking GABA A receptors with PTX, we investigated how this manipulation of the tonic current would influence neuronal excitability. To do this, we used step current injections of increasing amplitude (5-60 pA in 5 pA steps, 500 ms duration, holding potential of −60 mV) to evoke action potentials under baseline conditions and in the presence of PTX in Pitx3-GFP or C57BL/6J mice (Figure 6a). VTA dopamine neurons were identified using GFP expression in Pitx3-GFP mice or by the presence of an Ih current in C57BL/6J mice. We found the removal of the tonic current significantly shifted the input-output relationship to the left (Figure 6b; main effect of group F 1,8 = 10.61, p = 0.0116, main effect of current step F 11,88 = 20.9, p = <0.0001, interaction: F 11,88 = 3.192, p = 0.0011; n = 9 [N = 9). Importantly, input resistance of recorded neurons was significantly increased after PTX application (Figure 6c; Baseline 551.3 ± 37.22 MΩ, PTX 782.3 ± 88.08 MΩ, t 9 = 3.401, p = 0.0093, n = 9 [N = 9]). PTX is a non-competitive blocker for GABA A receptor chloride channels and known to also act on GABAc receptors and glycine receptors. However, GABAc receptors are unlikely to be found on VTA dopamine neurons (Petri et al., 2002) and glycine receptors are reported to be insensitive to PTX in adult rats (Ye et al., 1998;Zheng & Johnson, 2001). Thus, the effect of PTX in this study is likely selective to GABA A receptors. This suggests that tonic activation of GABA A receptors can indeed influence the excitability of VTA dopamine neurons, and therefore may represent an important modulator of neuronal activity in vivo.

| DISCUSSION
GABAergic inhibition is essential for regulating the firing pattern of midbrain dopamine neurons and represents a significant proportion of the total synaptic inputs onto these neurons. Although fast phasic synaptic inhibition arising through the activation of postsynaptic GABA A receptors has been well-described, little is known about whether this strong GABAergic drive can also mediate a tonic conductance.
Here, we have shown that tonic GABA A receptor-mediated currents are also present in VTA dopamine neurons. Our results suggest that the source of this background tonic inhibition is spill over from spontaneously active local GABAergic neurons in the VTA. We also found that the GABA transporter plays an active role in maintaining this background tonic current, and it may be that it can modify its activity to compensate for moderate increases in ambient GABA. However, under conditions where GABA release is significantly increased in a global manner (i.e., when we applied a NO donor), the tonic current can be increased.
In addition, electrophysiological studies show that δ subunit specific agonists enhance tonic currents induced by exogenous GABA in VTA GABA neurons, but not in VTA dopamine neurons (Vashchinkina et al., 2014). Consistent with this, we found that even a relatively high concentration of THIP had a minimal effect on holding currents. Moreover, a δ subunit specific agonist significantly reduces sIPSCs in dopamine neurons (presumably as a consequence of inhibition of VTA GABA neurons; Vashchinkina et al., 2012). With respect to α5 subunits, in situ hybridization, single-cell PCR and immunostaining studies have found that it is not present in the VTA (Allen Mouse Brain Atlas; Lein et al., 2007) (Ciccarelli et al., 2012;Fritschy & Mohler, 1995;Guyon et al., 1999;Hörtnagl et al., 2013;Okada et al., 2004;Pirker et al., 2000). Although these are the GABA A receptor subunits classically associated with tonic inhibition, it is important to note, however, that there are other examples of neurons that exhibit tonic currents that do not involve these subunits. For instance basolateral nucleus principal cells of the amygdala exhibit α3 GABA A receptor-mediated tonic currents, and the epileptic hippocampal dentate granule cells were found to maintain tonic currents with α4γ2 subunit containing GABA A receptors (Brickley & Mody, 2012;Marowsky et al., 2012;Rajasekaran et al., 2010). In this case, what is the subunit composition of GABA A receptors mediating tonic inhibition in VTA dopamine neurons? Because we observed some sensitivity to zinc, we hypothesized that GABA A receptors mediating tonic inhibition do not contain γ subunits, consistent with them being located extrasynaptically. One possibility is that they contain only α and β subunits. However, the absence of much effect of THIP and the lack of spontaneous opening suggests that this is not the case (Stórustovu & Ebert, 2006). One interesting possibility is that they contain Ɛ subunits, which are relatively uncommon (e.g., Belujon et al., 2009). Indeed, αβƐ containing receptors exhibit insensitivity to THIP and intermediate sensitivity to Zn 2+ (Thompson et al., 1998(Thompson et al., , 2002Whiting et al., 1997). Our immunostaining results, using an approach optimized for detecting receptor subunit expression in the VTA, revealed extensive expression of Ɛ in the VTA, colocalized with dopamine neurons but not synaptic markers. Furthermore, we found that the α4 subunit in particular was expressed extrasynaptically in dopamine neurons. Therefore, based on our combined pharmacology and immunostaining results, taking into account the existing literature concerning possible subunit combinations and their properties, we propose that the tonic current in VTA dopamine neurons is mediated via extrasynaptic GABA A receptors composed of a relatively unusual α4βƐ subunit combination. This presents a useful working hypothesis that will need to be tested using further pharmacological, anatomical and electrophysiological approaches, ideally combined with subunit-and cell-type-specific genetic deletions. We also examined the functional consequences of tonic inhibition on dopamine neuron firing. We observed that by increasing tonic inhibition, dopamine neuron excitability was increased. Importantly, we observed these effects when neurons were depolarized from the same resting membrane potential and it is, therefore, not simply a consequence of the hyperpolarizing effect of the tonic current. The most straightforward explanation is that the tonic current increased cell conductance, thereby shunting depolarizing input and making firing less likely. It is tempting to speculate that some of the effects of GABA A receptor blockage on dopamine neuron firing in vivo may involve this tonic current (for example, increased burst firing; Paladini & Tepper, 1999). Moreover, in vivo under conditions of globally increased GABAergic synaptic inputs it may be that this tonic current is increased, thereby reducing excitability and shunting excitatory synaptic inputs. Consistent with this, dopamine neuron input resistant measurements in vivo (Grace & Bunney, 1983) are considerably lower than those taken in vitro (Johnson & North, 1992;Yung et al., 1991). Another possible explanation is that a balance of inhibitory and excitatory tonic currents itself are contributing to pacemaking activity. In dynamic clamp experiments, for instance, either an increase in tonic NMDA or a decrease in tonic GABA currents can produce a burst firing of midbrain dopamine neurons and balanced excitation and inhibition results in pacemaking firing (Canavier & Landry, 2006;Komendantov et al., 2004;Lobb et al., 2010). Furthermore, a high ratio of tonic GABA/NMDA leads to silence and a high ratio of tonic NMDA/GABA leads to bursting. Thus, the level of tonic GABAergic current level may regulate silencing and firing of dopamine neurons and this mechanism could contribute, for example, to pauses in dopamine neuron firing in response to reward omissions and aversive stimuli (Henny et al., 2012;Schultz, 1986;Ungless et al., 2004).
There is emerging evidence that tonic inhibition through GABA B receptors can regulate neuronal firing (Khatri et al., 2019;Wang et al., 2015). Wang et al. reported that GABA B receptor-mediated tonic inhibition has a direct GABA A receptor-independent regulatory role on the spontaneous activity of noradrenaline neurons in the locus coeruleus. On the other hand, tonically active GABA B receptors enhance extrasynaptic GABA A receptor-mediated tonic currents by increasing desensitization and trafficking of GABA A receptors in cerebellar granule cells (Khatri et al., 2019). These findings suggest that GABA B receptors also have a role for tonic inhibition and this can modulate the function of postsynaptic GABA A receptors. Since we demonstrated here that GABA A receptor mediating tonic inhibition in VTA dopamine neurons potentially have an unusual combination of subunits, it would be interesting to see how this rare type of GABA A receptor would interact with GABA B receptor on pacemaking mechanisms of midbrain dopamine neurons.
In conclusion, we reveal the existence of an endogenous tonic GABAergic current in VTA dopamine neurons, which is a result of firing activity in VTA GABA neurons, is potentially mediated by GABA A receptors containing a relatively unusual combination αβƐ subunits, and which can ultimately regulate VTA dopamine neuron firing activity.