Corelease of acetylcholine and GABA from cholinergic forebrain neurons
Abstract
Neurotransmitter corelease is emerging as a common theme of central neuromodulatory systems. Though corelease of glutamate or GABA with acetylcholine has been reported within the cholinergic system, the full extent is unknown. To explore synaptic signaling of cholinergic forebrain neurons, we activated choline acetyltransferase expressing neurons using channelrhodopsin while recording post-synaptic currents (PSCs) in layer 1 interneurons. Surprisingly, we observed PSCs mediated by GABAA receptors in addition to nicotinic acetylcholine receptors. Based on PSC latency and pharmacological sensitivity, our results suggest monosynaptic release of both GABA and ACh. Anatomical analysis showed that forebrain cholinergic neurons express the GABA synthetic enzyme Gad2 and the vesicular GABA transporter (Slc32a1). We confirmed the direct release of GABA by knocking out Slc32a1 from cholinergic neurons. Our results identify GABA as an overlooked fast neurotransmitter utilized throughout the forebrain cholinergic system. GABA/ACh corelease may have major implications for modulation of cortical function by cholinergic neurons.
https://doi.org/10.7554/eLife.06412.001eLife digest
Neurons communicate with one another at junctions called synapses. When an electrical signal arrives at the presynaptic cell, it triggers the release of molecules called neurotransmitters into the synapse. These molecules then bind to receptor proteins on the postsynaptic cell, starting a chain of events that leads to the regeneration of the electrical signal in the second cell.
Broadly speaking, neurotransmitters are either excitatory, which means that they increase the electrical activity of the postsynaptic neurons, or they are inhibitory, meaning that they reduce postsynaptic activity. Initially, it was thought that neurons release only one type of neurotransmitter, but it is now known that this is not always the case. Many neurons within the spinal cord, for example, release two different inhibitory neurotransmitters, GABA and glycine, while some neurons in the midbrain release GABA and an excitatory neurotransmitter called glutamate.
Saunders, Granger, and Sabatini now provide the first direct evidence that cholinergic neurons in different regions of the forebrain also release two neurotransmitters. Collectively known as the ‘forebrain cholinergic system’, these cells are best known for producing the excitatory transmitter acetylcholine. However, Saunders et al. now show that this system also produces an enzyme that manufactures GABA, as well as a protein that pumps GABA into structures called vesicles, which are then released into the synapse.
Although this is not concrete evidence for the release of GABA, Saunders et al. also show—with a technique called optogenetics, which involves the use of light to control neuronal activity—that some of the neurons in this system can trigger inhibitory responses in postsynaptic cells. Moreover, these responses can be blocked using drugs that occupy GABA receptors, or by using genetic techniques to delete the GABA-pumping protein from cholinergic neurons.
Taken together, the results of these experiments strongly suggest that the cholinergic neurons throughout the forebrain—unlike, for example, the cholinergic neurons in the midbrain, the region of the brain that controls movement—possess the molecular machinery needed to produce and release GABA, in addition to acetylcholine. Given that the cholinergic system has a key role in cognition and is particularly susceptible to degeneration in Alzheimer's disease, the ability of these neurons to release GABA release could have widespread implications for the study and understanding of brain function.
https://doi.org/10.7554/eLife.06412.002Introduction
For many years, neurons were thought to release only a single fast neurotransmitter (Strata and Harvey, 1999). This assumption led to classifying neuronal subtypes based on released neurotransmitter, a convention which helped predict a neuron's circuit function. However, many neuronal subtypes that release multiple fast neurotransmitters have now been described (Hnasko and Edwards, 2012). In some cases, the coreleased neurotransmitters have similar post-synaptic effects, such as inhibition mediated by GABA and glycine from spinal interneurons (Jonas et al., 1998). In other instances, the effects of the two neurotransmitters may be different and synergistic. For example, coreleased GABA and glutamate are thought to control the balance of excitation and inhibition in the lateral habenula (Root et al., 2014; Shabel et al., 2014). In neuromodulatory systems, synaptic release of fast neurotransmitters along with slow neuromodulators has emerged as a common theme. In addition to the impact of dopamine, stimulation of dopaminergic terminals from the ventral tegmental area and substantia nigra compacta activates glutamate-mediated excitatory currents in the nucleus accumbens (Stuber et al., 2010; Tecuapetla et al., 2010) and GABA-mediated inhibitory currents in the striatum (Tritsch et al., 2012, 2014). Likewise, serotonergic neurons of the dorsal raphe can trigger glutamate-mediated currents in post-synaptic neurons of the ventral tegmentum and nucleus accumbens which contributes to the signaling of reward (Liu et al., 2014).
Several cholinergic neuron populations also release multiple neurotransmitters. Retinal starburst amacrine cells (SACs) differentially release GABA and acetylcholine (ACh) based on the pattern of light stimulation (Lee et al., 2010). In the central brain, selective activation of striatal cholinergic interneurons results in cholinergic and glutamatergic responses (Gras et al., 2008; Higley et al., 2011; Nelson et al., 2014). Similarly, the cholinergic projection from habenula excites interpeduncular neurons through glutamate and ACh (Ren et al., 2011). Some evidence suggests that the basal forebrain cholinergic (BFC) system, which provides the major source of ACh to cortex, may corelease GABA. The dorsal-most BFC neurons, which belong to the globus pallidus externus (GP), express molecular markers for GABA synthesis and vesicular packaging and trigger GABAA receptor currents in GP and cortex when activated (Tkatch et al., 1998; Saunders et al., 2015). We therefore asked whether GABA corelease was a general feature of forebrain cholinergic neurons.
To address this question, we selectively activated cholinergic fibers in the cortex, with the goal of identifying synaptic events triggered by endogenous release from forebrain cholinergic neurons. Recording from layer 1 interneurons, we observed not only the expected excitatory post-synaptic currents (EPSCs) mediated by nicotinic ACh receptors (nAChRs), but an unexpected inhibitory post-synaptic current (IPSC) mediated by GABAA receptors. IPSCs insensitive to nAChR antagonists had onset latencies slightly faster than the nicotinic EPSCs (nEPSCs), and could be directly evoked under pharmacological conditions in which action potentials were blocked, suggesting cholinergic neurons were directly releasing GABA in addition to ACh. Indeed, we found that cholinergic neurons throughout the forebrain commonly coexpressed the GABA synthetic enzyme GAD65 (Gad2), and the vesicular GABA transporter (Slc32a1), indicating that these neurons possess the necessary cell machinery for GABA transmission. Finally, we show that conditional deletion of Slc32a1 selectively in cholinergic neurons eliminates monosynaptic IPSCs while leaving nEPSCs intact, confirming the direct release of GABA from cholinergic terminals. These experiments suggest a previously overlooked capability of the cholinergic system to use GABA in synaptic signaling.
Results
To explore the effects that cholinergic neurons have on cortical circuitry, we used double transgenic mice to optogenetically activate neurons that express endogenous choline acetyltransferase (Chat). These mice carried a knock-in allele linking Cre recombinase expression to the Chat locus through an internal-ribosome entry site (Chat i-Cre) as well as a Cre-activated channelrhodopsin-EYFP fusion allele (Rosa26 lsl-ChR2-EYFP). Chat i-Cre; Rosa26 lsl-ChR2-EYFP mice expressed ChR2-EYFP throughout the forebrain, recapitulating known patterns of Chat expression in cortex (Ctx), striatum, globus pallidus externus (GP), and nucleus basalis (NB, Figure 1A). To test whether ChR2+ cells express endogenous Chat, we performed ChAT immunohistochemistry on sections of Chat i-Cre; Rosa26 lsl-ChR2-EYFP mice (Figure 1B). We focused on those Chat+ forebrain neurons positioned to innervate the cortex, including local Chat+ interneurons and the subcortical projections arising from the GP/NB. In both regions, ChR2+ neurons were immunopositive for ChAT, confirming our ability to selectively activate endogenous Chat+ inputs to cortex.

Optogenetic stimulation of cortical Chat+ fibers evokes fast, monosynaptic GABAA receptor-mediated currents.
(A) Low-magnification view of sagittal section from Chat i-Cre; Rosa26 lsl-ChR2-EYFP mouse forebrain. ChR2-EYFP (green) is expressed in the nucleus basalis (NB), globus pallidus externus (GP), striatum and cortex (Ctx). (B) Higher-magnification view of ChR2-EYFP expression combined with ChAT immunostaining (magenta) in frontal cortex (top) and nucleus basalis (bottom). Arrowheads indicate cells immunopositive for ChAT. (C) Example 2-photon stack from a layer 1 interneuron following whole-cell recording and dialysis with Alexa Fluor 594. (D) High-magnification view of ChR2-EYFP+ somata and fibers in layers 1 and 2/3 of frontal cortex. NeuN immunostain (magenta) highlights the distribution of neuronal somata across layers. Arrowhead indicates an example layer 1 interneuron surrounded by ChR2-EYFP fibers. (E) Example PSCs from voltage-clamp recordings of three different layer 1 interneurons in response to blue light stimulation (blue bar) of ChR2+ cholinergic fibers. Neurons were voltage-clamped at −70 mV (left) to isolate EPSCs or at 0 mV (middle and right) to isolate IPSCs. PSCs recorded in the presence of glutamate receptor antagonists CPP and NBQX are shown in black and after bath application of nAChR antagonists (MEC, MLA, and DHβE) in red. (F) Example light-evoked IPSCs from a layer 1 interneuron in a Chat i-Cre; Rosa26 lsl-ChR2-EYFP mouse voltage-clamped at 0 mV in the presence of CPP and NBQX (baseline) and following subsequent bath application of (from left to right) nAChR antagonists, TTX, 4AP, and SR95531. (G) Summary graph of IPSC peaks normalized to baseline (n = 5 cells from 4 mice). Asterisk, condition vs baseline p < 0.05, Mann–Whitney test). (H) Onset latencies for monosynaptic nEPSCs (n = 41 cells), monosynaptic IPSCs (n = 9 cells), and polysynaptic IPSCs (n = 19 cells from 9 mice). Mean (±sem) are shown in green. Asterisk, p < 0.05, Mann–Whitney test.
To identify the synaptic signaling mechanisms engaged by activation of the cortical cholinergic system, we targeted layer 1 interneurons for whole-cell voltage-clamp recordings in acute brain slices. Layer 1 is strongly innervated by ChAT+ cells of the basal forebrain across species (Mesulam, 1995; Mechawar et al., 2000) including in Chat i-Cre; Rosa26 lsl-ChR2-EYFP mice, where ChR2-EYFP is expressed in a dense plexus (Figure 1C,D). As expected, in a subset of interneurons (n = 41 of 58 cells from 9 mice), we observed robust excitatory postsynaptic currents (EPSCs) at −70 mV in response to brief pulses of blue light (2–7 ms, Figure 1E, left). These EPSCs were not blocked by NBQX and CPP, ruling out a glutamatergic source, but were blocked by the nicotinic ACh receptor (nAChR) antagonists DHβE, MLA, and MEC, confirming their cholinergic identity (nEPSCs). nEPSCs displayed a typical biphasic response, with a fast component, likely mediated by synaptic receptors containing the low-affinity α7 nAChR subunit, and a slow component, likely mediated by non-synaptic receptors expressing the high-affinity non-α7 subunits (Bennett et al., 2012).
In addition to the expected nEPSCs recorded at −70 mV, we also observed barrages of outward inhibitory postsynaptic currents (IPSCs) at 0 mV, indicative of signaling through GABA receptors (n = 28 of 58 cells, Figure 1E, center). One possible explanation for these IPSCs could be ACh-mediated feed-forward activation of local interneurons, resulting in disynaptic release of GABA. Indeed, when nAChR antagonists were applied, the delayed outward IPSCs disappeared. However, in a subset of recorded cells IPSCs remained (n = 9 of 58 cells, Figure 1F, right), suggesting that these PSCs were not dependent on nAChR signaling.
To test if nAChR-resistant IPSCs are caused by direct release of GABA from cholinergic fibers, we bath applied the voltage-gated sodium channel antagonist TTX, which blocked light-evoked IPSCs (Figure 1F,G). In the presence of TTX, IPSCs could be rescued by enhancing ChR2-mediated depolarization with the voltage-gated potassium channel blocker 4AP. Rescued IPSCs were subsequently blocked by the GABAA receptor antagonist SR95531 (n = 5 cells from 4 mice). Moreover, nAChR-resistant ‘direct’ IPSCs had faster average onsets than both nEPSCs and nAChR-sensitive ‘indirect’ IPSCs (nEPSCs, 4.0 ± 0.2 ms, n = 41 cells; direct IPSCs, 2.5 ± 0.2, n = 9 cells; indirect IPSCs, 11.8 ± 2.3, n = 19 cells from 9 mice, Figure 1H). These data suggest direct IPSCs are independent of nAChR signaling and mediated by GABAA receptors, consistent with monosynaptic release of GABA by cholinergic neurons. In support of this possibility, gene expression analyses have suggested some populations of Chat+ subcortical neurons contain the synthetic machinery for GABA (Kosaka et al., 1988; Tkatch et al., 1998).
GABA corelease has been observed in other neuromodulatory systems, namely from dopaminergic neurons of the substantia nigra (Tritsch et al., 2012). In those neurons, GABA is co-packaged with dopamine into vesicles by the transporter VMAT2 (Slc18a2), instead of by the typical vesicular transporter for GABA (VGAT, encoded by Slc32a1), which is necessary for packaging in most GABAergic neurons (Wojcik et al., 2006; Tong et al., 2008; Kozorovitskiy et al., 2012). To assess whether cholinergic neurons could use VGAT to package GABA into vesicles, we tested for Slc32a1/ChAT co-expression throughout the brain, including in cortex, GP, NB, diagonal band of broca (DBB)/medial septum (MS), and pedunculopontine nucleus (PPN, Figure 2A, top). We labeled cells expressing endogenous Slc32a1 using double-transgenic mice that link Cre recombinase expression to the Slc32a1 locus (Slc32a1i-Cre) and carry a zsGreen Cre reporter allele (Rosa26 lsl-zsGreen). Subsequent ChAT immunolabeling on sagittal and coronal sections from Slc32a1i-Cre; Rosa26 lsl-zsGreen mice was used to examine coexpression. Since zsGreen accumulates in somata and does not diffuse throughout the cytoplasm, this strategy allowed the clear identification of ChAT+ soma free from background fluorescence caused by Cre+ axons and dendrites. In cortex, GP, NB, and MS/DBB, nearly all of ChAT+ cell were Slc32a1+ (zsGreen+/ChAT+, from 3 mice: Ctx, 628/628; GP, 238/243; NB, 598/624; MS/DBB; 560/601, Figure 2A,B). In contrast, ChAT+ cells of the PPN very rarely expressed Slc32a1 (3/157, Figure 2B). These data suggest that in both cortical and subcortical forebrain regions, ChAT+ cells express the canonical molecular machinery to package GABA into vesicles. However, since ChAT+ neurons in the brainstem PPN do not express Slc32a1, this GABAergic marker is not a ubiquitous feature of the central cholinergic system.

Immunopositive ChAT cells in the forebrain express Slc32a1.
(A) Top, sagittal and coronal schematic views of a mouse brain showing cholinergic regions of interest. Red boxes indicate approximate locations for magnified regions below. Bottom, example single-plane image from a confocal stack from sections of a Slc32a1i-Cre; Rosa26 lsl-zsGreen mouse immunostained for ChAT (magenta) and reporting Cre expression (green). Ctx, cortex; GP, globus pallidus externus; NB, nucleus basalis; MS/DBB, medial septum/diagonal band of broca; PPN, pedunculopontine nucleus. Insets show magnified view of individual neurons indicated by the white arrowhead. (B) Quantification of colocalization between cells expressing zsGreen Cre reporter and ChAT immunostain by brain region (zsGreen+/ChAT+, from 3 mice: Ctx, 628/628; GP, 238/243; NB, 598/624; MS/DBB, 560/601; PPN, 3/157).
GABA is synthesized by one of two GABA synthetic enzymes, GAD65 or GAD67, encoded by the genes Gad2 or Gad1, respectively. GAD67 is expressed largely in cell bodies and is thought to be responsible for synthesizing GABA for general metabolic cell functions, whereas GAD65 expression is more prominent in axon terminals and is thought to mediate the majority of synthesis of synaptic GABA (Soghomonian and Martin, 1998). We examined if cholinergic neurons expressed GAD67 by immunostaining for ChAT in brain sections from knock-in mice where GFP replaces the first exon of the Gad1 gene (Gad1GFP). We detected only minor overlap between neurons that stained positive for ChAT and those that expressed GFP in the NB, GP, MS/DBB, or PPN, except for cortical ChAT+ interneurons, where we observed significant overlap (GFP+/ChAT + neurons: Ctx: 122/136; MS/DBB, 1/573; NB, 0/439; GP, 7/413; PPN, 12/246, Figure 3—figure supplement 1). This suggests that within the major subcortical cholinergic projections, GAD67-mediated GABA synthesis does not occur in cholinergic neurons. To test for coexpression of ChAT and GAD65, we immunostained sections of knock-in mice where Cre recombinase was targeted to the endogenous Gad2 gene (Gad2 i-Cre) and visualized Cre expression with the zsGreen Cre reporter. In contrast to Gad1, in cortex, GP, NB, and MS/DBB of Gad2 i-Cre; Rosa26 lsl-zsGreen mice, most ChAT+ neurons were Gad2+ (zsGreen+/ChAT+, from 4 mice: Ctx, 518/519; GP, 273/372; NB, 860/934; MS/DBB, 673/685, Figure 3A,B). In the brainstem, however, few of the ChAT+ cells were Gad2+ (PPN, 6/110, Figure 3A,B). This ChAT co-expression pattern for Gad2 is similar to Slc32a1, suggesting that forebrain cholinergic neurons possess the necessary cellular machinery to both synthesize and package synaptic GABA.

Immunopositive ChAT cells in the forebrain express Gad2.
(A) Top, sagittal and coronal schematic views of a mouse brain showing cholinergic regions of interest. Red boxes indicate approximate locations for magnified regions below. Bottom, example single-plane image from a confocal stack from sections of a Gad2 i-Cre; Rosa26 lsl-zsGreen mouse immunostained for ChAT (magenta) and reporting Cre expression (green). Ctx, cortex; GP, globus pallidus externus; NB, nucleus basalis; MS/DBB, medial septum/diagonal band of broca; PPN, pedunculopontine nucleus. Insets show magnified view of individual neurons indicated by the white arrowhead. (B) Quantification of colocalization between cells expressing zsGreen Cre reporter and ChAT immunostain by brain region (zsGreen+/ChAT+, from 4 mice: Ctx, 518/519; GP, 273/372; NB, 860/934;MS/DBB, 673/685; PPN, 6/110).
Though the fast onset and persistence of GABAergic IPSCs in the presence of nAChR antagonists and TTX/4AP argues strongly for direct release of GABA from cholinergic terminals, we wished to confirm this finding using conditional genetics. To determine if GABA release is indeed monosynaptic, we took advantage of the observation that cholinergic neurons express Slc32a1. If GABA is released directly from cholinergic neurons, then conditional knock-out of Slc32a1 selectively in these cells should abolish GABA release. We therefore bred triple transgenic mice which carried conditional (floxed) Slc32a1 alleles in addition to Chat i-Cre and Rosa26 lsl-ChR2-EYFP. These mice allowed us to compare the optogenetic stimulation of wild-type cholinergic neurons (Slc32a1+/+) to those lacking Slc32a1 (Slc32a1fl/fl) in acute brain slices. In voltage-clamp recordings from layer 1 interneurons, we observed fast onset IPSCs in 31% of cells recorded in Chat i-Cre; Slc32a1+/+ mice (n = 5 of 16 from 2 mice), but no direct IPSCs in Chat i-Cre; Slc32a1fl/fl mice (n = 0 of 34 cells from 4 mice, Figure 4A,B). In contrast, the proportion and average peak amplitude of nAChR responses remains unchanged between Slc32a1+/+ and Slc32a1fl/fl mice (Figure 4A,B). These data demonstrate that GABA but not ACh release from cholinergic neurons relies on cell autonomous Slc32a1, ruling out a disynaptic mechanism.

GABA release from cortical ChAT+ axons requires Slc32a1.
(A) Example light-evoked nEPSCs and IPSCs from four different layer 1 interneurons voltage-clamped at −70 or 0 mV from Chat i-Cre; Rosa26 lsl-ChR2-EYFP mice with wild-type cholinergic neurons (Slc32a1+/+) or conditional Slc32a1 knock-out (Slc32a1fl/fl). (B) Top, the number and proportion of layer 1 interneurons in which light-evoked nEPSCs or direct IPSCs were detected from Chat i-Cre; Rosa26 lsl-ChR2-EYFP mice with wild-type Slc32a1 alleles (Slc32a1+/+, from 2 mice) or following conditional Slc32a1 knock-out (Slc32a1fl/fl, from 4 mice). Bottom, PSC peaks for each condition. Means (±sem) are shown in green.
Discussion
Here, we provide evidence that the cortical cholinergic system is capable of GABAergic neurotransmission. In response to optogenetic stimulation of neurons expressing endogenous Chat, we observed PSCs mediated by both nAChRs and GABAA receptors in layer 1 interneurons. A subset of the evoked IPSCs appeared to be monosynaptic, based on latency and pharmacological analyses. In support of their GABAergic nature, the ChAT+ neurons which innervate cortex—local interneurons and subcortical projections arising from the GP/NB—express Gad2 and Slc32a1, the canonical molecular machinery for GABA synthesis and vesicular packaging. Indeed, conditional knock-out of Slc32a1 selectively in cholinergic neurons eliminates light-evoked monosynaptic IPSCs. These genetic results confirm that cholinergic neurons release GABA directly. Cholinergic GABA release is likely to be a feature of most, but not all, central cholinergic neurons: the ChAT+ neurons of the MS and DBB—which innervate the hippocampus—also express Gad2 and Slc32a1, whereas those of the midbrain PPN do not.
The mode of neurotransmitter corelease can vary across neuron classes. In some instances, both neurotransmitters are released from the same synaptic vesicles. This is the case for GABA/glutamate corelease onto neurons of the lateral habenula, where individual miniature EPSCs can be observed with dual GABA/glutamate components (Shabel et al., 2014). Copackaging in individual vesicles is also the case when the same vesicular transporter loads both neurotransmitters. For example, GABA is packaged in dopaminergic neurons by VMAT2, which also packages dopamine (Tritsch et al., 2012). Similarly, both GABA and glycine packaging in spinal interneurons rely on VGAT (Wojcik et al., 2006). However, corelease from separate pools of synaptic vesicles also occurs. In retinal SACs, release of GABA and ACh can be functionally separated through patterned light stimulation or pharmacology. In Chat+ GP axons in cortex, GABA and ACh appear to be released from distinct vesicular pools which can be located within the same or neighboring pre-synaptic terminals (Saunders et al., 2015). Though the precise mechanism by which each forebrain cholinergic population coreleases GABA and ACh remain unclear and further experiments are merited to explore how GABA/ACh cotransmission is regulated, the differences in the proportion of layer 1 interneurons showing ACh and GABA responses suggest some form of segregated release.
The cholinergic system's function in promoting attention, alertness, and learning has classically been attributed to acetylcholine and its action as a diffuse volume transmitter, affecting cortical activity at relatively slow time scales. This model is supported by anatomical evidence showing widespread distribution of cholinergic fibers through all cortical layers with significant separation between the sites of release and ACh receptors (Descarries and Mechawar, 2000; Mechawar et al., 2000). In addition, many in vitro pharmacological experiments have shown that ACh receptors can shape the signaling of other neurotransmitter systems, by altering properties of presynaptic release, synaptic plasticity, or the intrinsic excitability of targeted neurons (Picciotto et al., 2012). However, more recent work has focused on the participation of ACh in rapid, wired neurotransmission, acting at tightly apposed synapses (Sarter et al., 2009; Poorthuis et al., 2014; Sarter et al., 2014). Behaviorally relevant sensory cues can cause a fast, time-locked spike in ACh concentration, suggesting that ACh may mediate detection of that cue (Parikh et al., 2007). In addition, fast onset currents mediated by nAChRs can be recorded in cortical interneurons following optogenetic activation of cholinergic fibers (Arroyo et al., 2012; Bennett et al., 2012). Our finding that cholinergic neurons also elicit fast-onset synaptic GABAA responses lends further support to the notion that the cholinergic system can rapidly affect cortical computations by acting at classical synapses.
GABA corelease from cholinergic forebrain neurons may affect cortical function in several ways. First, at the circuit level, GABA release could act in a manner that reinforces the emerging concept that the cholinergic system disinhibits cortical firing. ACh release from basal forebrain neurons excites layer 1 and VIP+ interneurons, which in turn inhibit local interneurons that target principle neurons of the cortex (Letzkus et al., 2011; Pinto et al., 2013; Fu et al., 2014). Depending on the timing and targeted cell types, GABA corelease could conceivably enhance this effect by inhibiting local interneurons, thereby promoting cortical activity. Second, nAChRs can regulate both pre and post-synaptic GABAergic signaling. For example, in hippocampal interneurons, post-synaptic nAChRs are present in inhibitory synapses (Fabian-Fine et al., 2001) and when activated, reduce GABAA receptor-mediated IPSCs in a Ca2+ and PKC-dependent manner (Wanaverbecq et al., 2007; Zhang and Berg, 2007). Thus coreleased ACh and GABA could interact to modulate local synaptic signaling. Lastly, experiments poisoning or stimulating the basal forebrain cholinergic system have demonstrated that activity within this projection is necessary and sufficient for plasticity in sensory cortices (Kilgard and Merzenich, 1998; Weinberger, 2004; Ramanathan et al., 2009). While ACh alone can induce functional changes in cortical circuits (Metherate and Weinberger, 1990), GABA may also contribute to synaptic rewiring in vivo. Addressing these questions experimentally will benefit from future work to clarify the basic synaptic anatomy and biochemical regulation of ACh/GABA corelease. Given the presence of GABA signaling machinery throughout the distinct forebrain cholinergic systems, corelease likely has a significant and fundamental effect on brain activity and cognition.
Materials and methods
Mice
Cre recombinase was targeted to specific cell types using knock-in mice to drive Cre expression under endogenous gene-specific regulatory elements using an internal ribosome entry site. Cre knock-in mice for choline acetyltransferase (Chat) (Rossi et al., 2011) and vesicular GABA transporter (Slc32a1) (Tong et al., 2008) were provided by Brad Lowell (Beth Israel Deaconess Medical Center) and are available from the Jackson Labs (Bar Harbor, ME; Chat i-Cre, stock #006410; Slc32a1i-Cre, stock #016962). Gad2 i-Cre mice were purchased from Jackson Labs (stock #010802) (Taniguchi et al., 2011). Gad1GFP knock-in mice replace elements of Gad1 coding sequence with GFP (Tamamaki et al., 2003). We did not distinguish between mice hetero or homozygous for transgenic alleles except where indicated. All experimental manipulations were performed in accordance with a protocol (#03551) approved by the Harvard Standing Committee on Animal Care following guidelines described in the US National Institutes of Health Guide for the Care and Use of Laboratory Animals.
Fixed tissue preparation, immunohistochemistry, and imaging
View detailed protocolMice aged post-natal day 25–124 were deeply anesthetized with isoflurane and transcardially perfused with 4% paraformaldehyde (PFA) in 0.1 M sodium phosphate buffer (1× PBS). Brains were post-fixed for 1–3 days, washed in 1× PBS, and sectioned (40–50 μm) coronally or sagittally using a Vibratome (Leica Microsystems, Buffalo Grove, IL). For ChAT or NeuN immunohistochemistry, slices were incubated in a 1× PBS blocking solution containing 5% normal horse serum and 0.3% Triton X-100 for 1 hr at room temperature. Slices were then incubated overnight at 4°C in the same solution containing anti-choline acetyltransferase antibody (AB144P; 1:100; Millipore, Billerica, MA) or NeuN (MAB377; 1:100; Millipore). The next morning, sections were washed three times for five minutes in 1× PBS and then incubated for 1 hr at room temperature in the blocking solution containing donkey anti-goat Alexa 647 or Alexa 594 (for ChAT) or anti-mouse Alexa 647 (for NeuN) (1:500; Molecular Probes, Eugene, OR). Slices were then mounted on slides (Super Frost). After drying, slices were coverslipped with ProLong antifade mounting media containing DAPI (Molecular Probes) and imaged with an Olympus VS110 slide scanning microscope using the 10× objective. Confocal images (1–2 μm optical sections) were acquired with an Olympus FV1000 laser scanning confocal microscope (Harvard Neurobiology Imaging Facility) through a 60× objective. Colabeling quantification was carried on images obtained from the Olympus VS110 slide scanning microscope using ImageJ.
Slice preparation
Request a detailed protocolAcute brain slices were obtained from mice aged post-natal day 30–128 using standard techniques. Mice were anesthetized by isoflurane inhalation and perfused through the heart with ice-cold artificial cerebrospinal fluid (ACSF) containing (in mM) 125 NaCl, 2.5 KCl, 25 NaHCO3, 2 CaCl2, 1 MgCl2, 1.25 NaH2PO4, and 11 glucose (∼308 mOsm·kg−1). Cerebral hemispheres were removed, placed in ice-cold choline-based cutting solution (consisting of [in mM]: 110 choline chloride, 25 NaHCO3, 2.5 KCl, 7 MgCl2, 0.5 CaCl2, 1.25 NaH2PO4, 25 glucose, 11.6 ascorbic acid, and 3.1 pyruvic acid), blocked, and transferred into a slicing chamber containing ice-cold choline-based cutting solution. Sagittal slices (300–350 μm thick) were cut with a Leica VT1000s vibratome and transferred to a holding chamber containing ACSF at 34°C for 30 min and then subsequently at room temperature. Both cutting solution and ACSF were constantly bubbled with 95% O2/5% CO2.
Acute slice electrophysiology and two-photon imaging
Request a detailed protocolIndividual slices were transferred to a recording chamber mounted on a custom built two-photon laser scanning microscope (Olympus BX51WI) equipped for whole-cell patch-clamp recordings and optogenetic stimulation. Slices were continuously superfused (3.5–4.5 ml·min−1) with ACSF warmed to 32–34°C through a feedback-controlled heater (TC-324B; Warner Instruments). Cells were visualized through a water-immersion 60× objective using differential interference contrast (DIC) illumination. Epifluorescence illumination was used to identify those layer 1 interneurons surrounded by ChR2-EYFP processes. Patch pipettes (2–4 MΩ) pulled from borosilicate glass (G150F-3; Warner Instruments) were filled with a Cs+-based low Cl–internal solution containing (in mM) 135 CsMeSO3, 10 HEPES, 1 EGTA, 3.3 QX-314 (Cl–salt), 4 Mg-ATP, 0.3 Na-GTP, 8 Na2-Phosphocreatine (pH 7.3 adjusted with CsOH; 295 mOsm·kg−1) for voltage-clamp recordings. Series resistance (<25 MΩ) was measured with a 5-mV hyperpolarizing pulse in voltage-clamp and left uncompensated. Membrane potentials were corrected for a ∼7 mV liquid junction potential. In some cases after the recording was complete, cellular morphology was captured in a volume stack using 740 nm two-photon laser light (Coherent). To activate ChR2 in acute slices from ChAT i-Cre; Rosa26 lsl-ChR2-EYFP mice, 473 nm laser light (Optoengine) was focused onto the back aperture of the 60× water immersion objective to produce collimated whole-field illumination. Square pulses of laser light were delivered every 20 s and power (2–7 ms; 4.4 mW·mm−2) was quantified for each stimulation by measuring light diverted to a focal plane calibrated photodiode through a low-pass dichroic filter. Following bath application of TTX and 4AP, in some cases, light power or duration was increased slightly to recover currents (e.g., changing the duration from 2 to 4 ms).
Reagents
Drugs (all from Tocris, United Kingdom) were applied via bath perfusion: SR95531 (10 μΜ), tetrodotoxin (TTX; 1 μΜ), 4-aminopyridine (4AP; 500 μM), scopolamine (10 μΜ), 2,3-dihydroxy-6-nitro-7-sulfamoyl-benzo(f)quinoxaline (NBQX; 10 μM), R,S-3-(2-carboxypiperazin-4-yl)propyl-1-phosphonic acid (CPP; 10 μM), N,2,3,3-Tetramethylbicyclo[2.2.1]heptan-2-amine, (MEC; 10 μM), [1α,4(S),6β,14α,16β]-20-Ethyl-1,6,14,16-tetramethoxy-4-[[[2-(3-methyl-2,5-dioxo-1-pyrrolidinyl)benzoyl]oxy]methyl]aconitane-7,8-diol (MLA; 0.1 μM), (2S,13bS)-2-Methoxy-2,3,5,6,8,9,10,13-octahydro-1H,12H-benzo[i]pyrano[3,4-g]indolizin-12-one (DHβE; 10 μM). CPP and NBQX were combined to make a cocktail of antagonists to target ionotropic glutamate receptors, while MEC, MLA, and DHβE were combined to make a cocktail to antagonize nicotinic receptors.
Acute slice data acquisition and analysis
Request a detailed protocolMembrane currents and potentials were recorded using an Axoclamp 700B amplifier (Molecular Devices, Sunnyvale, CA) filtered at 3 kHz and digitized at 10 kHz using National Instruments acquisition boards and ScanImage (available at: scanimage.org) written in MATLAB (Mathworks, Natick, MA). Electrophysiology and imaging data were analyzed offline using Igor Pro (Wavemetrics, Lake Oswego, OR), ImageJ (NIH, Bethesda, MD) and GraphPad Prism (GraphPad Software, La Jolla, CA). In figures, voltage-clamp traces represent the average waveform of 3–6 acquisitions. Peak current amplitudes were calculated by averaging over a 1 ms window around the peak. For pharmacological analyses, 3–7 consecutive acquisitions (20 s inter-stimulus interval) were averaged following a 3-min wash-in period for NBQX and CPP or a 4-min wash-in period for MEC, MLA, and DHβE. For TTX and 4AP conditions, current averages were composed of the acquisitions following full block or first-recovery of ChR2 evoked currents, respectively. Data (reported in text and figures as mean ± sem) were compared statistically using the Mann–Whitney test. p values smaller than 0.05 were considered statistically significant.
References
-
Mechanisms generating dual-component nicotinic EPSCs in cortical interneuronsThe Journal of Neuroscience 32:17287–17296.https://doi.org/10.1523/JNEUROSCI.3565-12.2012
-
Ultrastructural distribution of the alpha7 nicotinic acetylcholine receptor subunit in rat hippocampusThe Journal of Neuroscience 21:7993–8003.
-
The vesicular glutamate transporter VGLUT3 synergizes striatal acetylcholine toneNature Neuroscience 11:292–300.https://doi.org/10.1038/nn2052
-
Neurotransmitter corelease: mechanism and physiological roleAnnual Review of Physiology 74:225–243.https://doi.org/10.1146/annurev-physiol-020911-153315
-
Cholinergic innervation in adult rat cerebral cortex: a quantitative immunocytochemical descriptionThe Journal of Comparative Neurology 428:305–318.https://doi.org/10.1002/1096-9861(20001211)428:23.0.CO;2-Y
-
Cholinergic pathways and the ascending reticular activating system of the human brainAnnals of the New York Academy of Sciences 757:169–179.https://doi.org/10.1111/j.1749-6632.1995.tb17472.x
-
Striatal cholinergic neurotransmission requires VGLUT3The Journal of Neuroscience 34:8772–8777.https://doi.org/10.1523/JNEUROSCI.0901-14.2014
-
Fast modulation of visual perception by basal forebrain cholinergic neuronsNature Neuroscience 16:1857–1863.https://doi.org/10.1038/nn.3552
-
Cholinergic circuit modulation through differential recruitment of neocortical interneuron types during behaviorThe Journal of Physiology 592:4155–4164.https://doi.org/10.1113/jphysiol.2014.273862
-
The basal forebrain cholinergic system is required specifically for behaviorally mediated cortical map plasticityThe Journal of Neuroscience 29:5992–6000.https://doi.org/10.1523/JNEUROSCI.0230-09.2009
-
Single rodent mesohabenular axons release glutamate and GABANature Neuroscience 17:1543–1551.https://doi.org/10.1038/nn.3823
-
Deterministic functions of cortical acetylcholineThe European Journal of Neuroscience 39:1912–1920.https://doi.org/10.1111/ejn.12515
-
Phasic acetylcholine release and the volume transmission hypothesis: time to move onNature Reviews Neuroscience 10:383–390.https://doi.org/10.1038/nrn2635
-
Two isoforms of glutamate decarboxylase: why?Trends in Pharmacological Sciences 19:500–505.https://doi.org/10.1016/S0165-6147(98)01270-X
-
Dopaminergic terminals in the nucleus accumbens but not the dorsal striatum corelease glutamateThe Journal of Neuroscience 30:8229–8233.https://doi.org/10.1523/JNEUROSCI.1754-10.2010
-
Green fluorescent protein expression and colocalization with calretinin, parvalbumin, and somatostatin in the GAD67-GFP knock-in mouseThe Journal of Comparative Neurology 467:60–79.https://doi.org/10.1002/cne.10905
-
Glutamatergic signaling by mesolimbic dopamine neurons in the nucleus accumbensThe Journal of Neuroscience 30:7105–7110.https://doi.org/10.1523/JNEUROSCI.0265-10.2010
-
Synaptic release of GABA by AgRP neurons is required for normal regulation of energy balanceNature Neuroscience 11:998–1000.https://doi.org/10.1038/nn.2167
-
Specific long-term memory traces in primary auditory cortexNature Reviews Neuroscience 5:279–290.https://doi.org/10.1038/nrn1366
Decision letter
-
Sacha B NelsonReviewing Editor; Brandeis University, United States
eLife posts the editorial decision letter and author response on a selection of the published articles (subject to the approval of the authors). An edited version of the letter sent to the authors after peer review is shown, indicating the substantive concerns or comments; minor concerns are not usually shown. Reviewers have the opportunity to discuss the decision before the letter is sent (see review process). Similarly, the author response typically shows only responses to the major concerns raised by the reviewers.
Thank you for sending your work entitled “Corelease of acetylcholine and GABA from cholinergic forebrain neurons” for consideration at eLife. Your article has been favorably evaluated by a Senior editor and three reviewers, one of whom, Sacha Nelson, is a member of our Board of Reviewing Editors.
The Reviewing editor and the other reviewers discussed their comments before we reached this decision, and the Reviewing editor has assembled the following comments to help you prepare a revised submission.
Because the reviews are concise and raise only minor concerns the Reviewing editor has included them in full here.
Reviewer 1:
This is a short straightforward report that uses mouse genetics, anatomical colocalization and physiology and pharmacology in brain slices to demonstrate that most forebrain cholinergic neurons are also GABAergic. Although there was one prior report hinting at this relationship based on marker expression, the results are largely surprising and change the way that we think about the function of cholinergic input to the cortex and other forebrain regions.
I have no substantive concerns except that it would be important to insure that there is not significant overlap with the other paper from this group currently in press.
There are two more minor suggestions for improvement:
1) It is slightly surprising that only 9/58 cells exhibited direct IPSCs (vs. 41/58 cholinergic EPSCs) while the overwhelming majority of cholinergic neurons examined expressed Vgat and Gad2. What if anything additional do the authors know about this? Did they try stimulating repeatedly to see if multiple action potentials may be required to release GABA at some synapses? Do they think there may be small IPSCs to which they are insensitive? Is it feasible that some of the presynaptic neurons do not express Vgat and Gad2 at the protein level? This latter point would be easy to test with immunohistochemistry.
If the authors have negative results that constrain the answers to these questions, describing them would be a useful addition. But at the very least they should speculate about the reason for the discrepancy.
2) What have the authors examined with respect to the identity of the L1 neurons that receive and don't receive GABAergic input? Are they indistinguishable in terms of firing properties, morphology etc.? Again, even negative information would be useful.
Reviewer #2:
The authors examine the putative corelease properties of cholinergic synapses in the cortex using a combination of transgenic and optogenetic approaches. Post-synaptic currents were recorded in layer 1 interneurons following the selective activation of ChAT+ inputs in acute slice. The authors present several lines of evidence which support the direct release of GABA from cholinergic terminals, including: 1) light-induced, nAChR-antagonist resistant IPSCs, 2) expression of the enzymatic machinery for GABA biosynthesis and packaging, and 3) elimination of putative monosynaptic IPSCs in conditional Vgat knockout mice. These findings may lead us to re-evaluate the role of the basal forebrain cholinergic system in modulating cortical excitability and thus, are of interest to a broad neuroscience community. The specific functional implications of ACh/GABA cotransmission in the cortex are somewhat unclear and warrant further study.
Minor comments:
1) It appears that nAChR-resistant IPSCs are observed only in a subpopulation (approximately 15%) of recorded cells (as described in the Results section). However, IHC data suggest that molecular machinery for the synthesis and packaging of GABA are co-expressed in nearly all cortical ChAT+ cells. How might this discrepancy be explained? Since the direct release of GABA occurs at few cholinergic synapses, can we expect that this phenomenon will have significant impact on cortical function?
2) Can the quality of images in Figures 2 and 3 be improved, perhaps by higher magnification photomicrographs, to more clearly represent the quantification in panels B?
3) The current onset data (see panel H in Figure 1 for reference) should be included in Figure 4 to demonstrate the specific loss of direct, monosynaptic IPSCs in Vgat-deficient mice?
4) Were ACh-mediated EPSCs prolonged in Vgat-deficient mice? It appears so based on the trace shown in Figure 4. What would be the explanation of this result?
Reviewer #3:
The authors report the corelease of GABA and ACh onto cortical layer 1 interneurons. This is a meticulous study employing several transgenic mice, immunohistochemistry, pharmacology, optogenics and electrophysiology. The resulting interpretation of corelease is convincing.
Minor comments:
The only minor question is whether individual neurons release both transmitters. This seems highly likely as all presynaptic neurons, i.e., those expressing ChR2, are Chat positive and many/most are also positive for Gad2 and Vgat (though why only ∼20% of the postsynaptic cells that have AChergic EPSCs also have GABAergic IPSCs is unclear; maybe the cited Nature paper explains this). Formally, however, an absolute demonstration of corelease would require the excitation of single presynaptic neuron or finding that minis have both components. This is a minor comment and certainly not necessary for this paper.
Also, the stated rationale for performing the experiment in Figure 4 (“the possibility remains that the direct IPSCs […] were due to nicotinic axon-axonic synapses onto GABAergic terminals”) seems to be incorrect. While multiple experiments leading to the same conclusion is always more satisfying that single experiments, it would seem that the nicotinic blocker experiment would also rule out the axo-axonic connections. Maybe a more reasonable introduction to Figure 4 would be something like: “Consistent with the pharmacology…” That said, the experiments in Figure 4 are elegant and convincing and should be included.
Do the authors have to use “impact” as a verb (in the Discussion)? Whatever happened to “affect”?
https://doi.org/10.7554/eLife.06412.008Author response
Because the reviews are concise and raise only minor concerns the Reviewing editor has included them in full here.
Reviewer 1:
This is a short straightforward report that uses mouse genetics, anatomical colocalization and physiology and pharmacology in brain slices to demonstrate that most forebrain cholinergic neurons are also GABAergic. Although there was one prior report hinting at this relationship based on marker expression, the results are largely surprising and change the way that we think about the function of cholinergic input to the cortex and other forebrain regions.
I have no substantive concerns except that it would be important to insure that there is not significant overlap with the other paper from this group currently in press.
We thank the reviewer for their support of our study. We too were surprised by the GABA corelease from cholinergic neurons and agree that this phenomenon will force a reconsideration of cholinergic modulation of forebrain structures.
Our paper in press at Nature is a detailed anatomical and functional description of a subcortical projection system to cortex. We discovered that the globus pallidus externus (GPe) sends a direct projection to cortex, thus altering the classical model of the basal ganglia. This projection is carried by two cell types, one of which is ChAT+ neurons of the GPe, which we demonstrate express Gad2 and Vgat and release GABA in addition to ACh. That finding is a relatively minor part of the paper. Furthermore, as referenced by the referee above, Gad2 and Vgat coexpression was known to occur in cells of that region.
Corelease of GABA and ACh from the GPe ChAT+ cells inspired us to look at the potential of corelease in other cholinergic centers throughout the brain, resulting in the study submitted to eLife. The results presented here do not significantly overlap with those of the in press study. Furthermore the eLife study includes in depth analysis such as the conditional knock out of Vgat from Chat+ neurons, which provides genetic evidence supporting that GABA is released directly from cholinergic neurons.
There are two more minor suggestions for improvement:
1) It is slightly surprising that only 9/58 cells exhibited direct IPSCs (vs. 41/58 cholinergic EPSCs) while the overwhelming majority of cholinergic neurons examined expressed Vgat and Gad2. What if anything additional do the authors know about this? Did they try stimulating repeatedly to see if multiple action potentials may be required to release GABA at some synapses? Do they think there may be small IPSCs to which they are insensitive? Is it feasible that some of the presynaptic neurons do not express Vgat and Gad2 at the protein level? This latter point would be easy to test with immunohistochemistry.
If the authors have negative results that constrain the answers to these questions, describing them would be a useful addition. But at the very least they should speculate about the reason for the discrepancy.
We thank the reviewer for raising this interesting point regarding differences in the occurrence of GABA and ACh driven synaptic currents. We have expanded the discussion of this point in the manuscript.
In a subset of experiments, we tried stronger single-pulse or pulse-trains of ChR2 activation, but these alternative stimulations did not reveal GABAergic currents. Of course we cannot rule out the possibility that our somatic recordings were not able to resolve small IPSCs or heavily filtered IPSCs resulting from the activation of synapses on distal dendrites.
Our guess is that there is high specificity of the cortical targets of GABA release from Chat+ neurons. Several possibilities could explain target specificity of GABA/ACh currents. If GABA and ACh are co-packaged the same vesicles, the recorded post-synaptic neuron could express either GABAAR or nAChR. Alternatively, differential GABA/ACh corelease could be achieved presynaptically either by: 1) different types of neurons within the cholinergic system (for example local ChAT+ interneuron vs. subcortical projections), or 2) through independent release sites from single axons. Precedence in the literature exists for the latter possibility, as ACh and GABA may be independently released from starburst amacrine cells of the retina (Lee et al. 2010. Neuron 68: 1159-72. doi: 10.1016/j.neuron.2010.11.031).
There could be additional regulation of release by activity, neuromodulators or behavioral state, including post-transcriptional regulation of Vgat and Gad2 mRNAs. Establishing the mechanisms of corelease is a priority of ongoing experiments in the lab.
2) What have the authors examined with respect to the identity of the L1 neurons that receive and don't receive GABAergic input? Are they indistinguishable in terms of firing properties, morphology etc.? Again, even negative information would be useful.
We have not examined the identity of L1 neurons that receive or do not receive GABA currents in detail. In many of our recordings, we capture the morphology of the neurons with 2P stacks. At the level of dendritic anatomy, there are not any striking differences, though we have not attempted a quantitative comparison of morphological properties. Biocytin-based morphological reconstructions would be necessary to test for differences in axonal anatomy, like the distinct flavors of L1 neurons that innervate the same or different cortical columns (Jiang et al. 2012. Nat Neuro. doi: 10.1038/nn.3305).
Our goal was to use the L1 neurons as a read out for both GABA and ACh mediated post-synaptic currents. Therefore we performed all of our analyses in voltage clamp with a Cs-based, low Cl- internal solution, discriminating currents by holding potential and pharmacology. We felt that these recording conditions offered the highest sensitivity, as necessary to detect potentially small EPSCs/IPSCs. However, these recording conditions precluded us from looking at differences in active membrane properties. We are now embarking on a large effort to discover the rules for selectivity of GABA and ACh neurotransmission using a large panel of GFP marker lines and post-hoc analysis. We expect this effort will take a long time to complete.
Reviewer #2:
The authors examine the putative corelease properties of cholinergic synapses in the cortex using a combination of transgenic and optogenetic approaches. Post-synaptic currents were recorded in layer 1 interneurons following the selective activation of ChAT+ inputs in acute slice. The authors present several lines of evidence which support the direct release of GABA from cholinergic terminals, including: 1) light-induced, nAChR-antagonist resistant IPSCs, 2) expression of the enzymatic machinery for GABA biosynthesis and packaging, and 3) elimination of putative monosynaptic IPSCs in conditional Vgat knockout mice. These findings may lead us to re-evaluate the role of the basal forebrain cholinergic system in modulating cortical excitability and thus, are of interest to a broad neuroscience community. The specific functional implications of ACh/GABA cotransmission in the cortex are somewhat unclear and warrant further study.
We thank the Reviewer for their comments. We agree that GABA corelease from ChAT+ cells of the forebrain forces an important reconsideration for how the cholinergic system regulates excitability and activity in neural circuits. We also agree that the study presented here is an initial description of the ACh/GABA corelease phenomenon; future work needs to be done to understand the functional impact of GABA release throughout the many cortical and subcortical cell types and networks which receive ACh/GABA synapses.
Minor comments:
1) It appears that nAChR-resistant IPSCs are observed only in a subpopulation (approximately 15%) of recorded cells (as described in the Results section). However, IHC data suggest that molecular machinery for the synthesis and packaging of GABA are co-expressed in nearly all cortical ChAT+ cells. How might this discrepancy be explained? Since the direct release of GABA occurs at few cholinergic synapses, can we expect that this phenomenon will have significant impact on cortical function?
Please see reviewer #1, comment 1, in which we speculate on several corelease mechanisms that could provide target specificity of GABA and Ach release.
The impact of corelease on cortical function is an active area of research in the lab, necessitating acute slice circuitry/anatomy work, in vivo optogenetics and behavioral work in the context of conditional mouse genetics to isolate GABA vs. ACh signaling in the cholinergic system. Furthermore, it requires careful comparison of the effects on cortical function of activity in cholinergic systems with and without GABA release. This is, of course, a major research endeavor.
2) Can the quality of images in Figures 2 and 3 be improved, perhaps by higher magnification photomicrographs, to more clearly represent the quantification in panels B?
We have now included high magnification insets for a clearer representation of the co-localization analysis. We left the low magnification images intact to illustrate the labeling patterns unique to each brain region and to illustrate the co-localization across multiple cells.
We have also added a figure supplement, showing that outside of the cortex, ChAT+ neurons do not coexpress GAD67, as described in our original submission.
3) The current onset data (see panel H in Figure 1 for reference) should be included in Figure 4 to demonstrate the specific loss of direct, monosynaptic IPSCs in Vgat-deficient mice?
The reviewer makes a good point. However, in our optogenetics experiments with the conditional Vgat allele, we focused on screening for direct GABA/ACh currents. Direct, monosynaptic currents were reliably evoked and quantified from ∼ 10 acquisitions. However the small number of acquisitions used for the conditional Vgat experiments were not conducive to detecting and quantifying indirect IPSCs, which are more unreliable and have less consistent onset times.
Nonetheless, we prepared a direct current onset latency figure for the reviewers (Author response image 1A). Since we did not detect any direct GABAergic IPSCs in the Vgatfl/fl condition, there is no latency data. For the Vgat+/+ condition, GABAergic IPSCs were reliable and fast, as in Figure 1H. Direct nEPSCs did not significantly differ in their mean (±sem) onset latencies across Vgat conditions (Vgatfl/fl: 5.1±0.5, n=12 cells; Vgat+/+: 4.2±0.7, n=7).
4) Were ACh-mediated EPSCs prolonged in Vgat-deficient mice? It appears so based on the trace shown in Figure 4. What would be the explanation of this result?
The kinetics of nEPSCs can vary widely due to different contributions of fast and slow components. The fast component is thought to be due to nAChRs containing α7-subunits and activated by point-to-point synaptic transmission. The slow component is thought to be due to non-α7 subunit containing nAChRs and volume transmission. In L1 interneurons, which have served as a model system for working out some of the properties of ionotropic ACh transmission, nEPSCs are quite variable (Bennett et al. 2012. J Neurosci. doi: 10.1523/JNEUROSCI.3565-12.2012; Arroyo et al. 2012. J Neurosci. doi: 10.1523/JNEUROSCI.0115-12.2012).
Based on the reviewers suggestion, we rechecked the kinetics of nEPSCs across the floxed Vgat allele conditions. nEPSCs also had highly variable fast and slow components in both conditional KO and wild-type mice. The slightly prolonged fast component of Vgatfl/fl is present in the normalized average nEPSC waveform (shown with sem, Author response image 1B). However, given the diversity in kinetics in both conditions, we are not able to determine if this difference is systematic or due sampling bias from small numbers. Many more recordings would be required to define systematic changes in nEPSC kinetics.
Reviewer #3:
The authors report the corelease of GABA and ACh onto cortical layer 1 interneurons. This is a meticulous study employing several transgenic mice, immunohistochemistry, pharmacology, optogenics and electrophysiology. The resulting interpretation of corelease is convincing.
We thank the reviewer for describing our study as “meticulous” and “convincing.”
Minor comments:
The only minor question is whether individual neurons release both transmitters. This seems highly likely as all presynaptic neurons, i.e., those expressing ChR2, are Chat positive and many/most are also positive for Gad2 and Vgat (though why only ∼20% of the postsynaptic cells that have AChergic EPSCs also have GABAergic IPSCs is unclear; maybe the cited Nature paper explains this). Formally, however, an absolute demonstration of corelease would require the excitation of single presynaptic neuron or finding that minis have both components. This is a minor comment and certainly not necessary for this paper.
Reviewer 3 makes an important point regarding the formal definition of corelease. We also appreciate the sentiment that resolving the mechanisms of corelease for multiple cell types within the cholinergic system should be reserved for future work. In regards to the corelease mechanisms, please see our response to referee #1, comment 1.
Also, the stated rationale for performing the experiment in Figure 4 (“the possibility remains that the direct IPSCs […] were due to nicotinic axon-axonic synapses onto GABAergic terminals”) seems to be incorrect. While multiple experiments leading to the same conclusion is always more satisfying that single experiments, it would seem that the nicotinic blocker experiment would also rule out the axo-axonic connections. Maybe a more reasonable introduction to Figure 4 would be something like: “Consistent with the pharmacology…” That said, the experiments in Figure 4 are elegant and convincing and should be included.
We thank the reviewer for pointing out this miscommunication of rationale. Our intention was to suggest that because of the known complexities in receptor makeup of brain nAChRs, and the resulting complexities in antagonist pharmacology, that our blocker experiments could be prone to cryptic, resistant nAChRs. Nevertheless, we agree with the reviewer and have changed the wording accordingly.
Do the authors have to use “impact” as a verb (in the Discussion)? Whatever happened to “affect”?
“Affect” is making a resurgence! We have changed the wording as per the reviewer’s suggestion.
https://doi.org/10.7554/eLife.06412.009Article and author information
Author details
Funding
National Institute of Neurological Disorders and Stroke (NINDS) (#NS072030)
- Adam J Granger
The funder had no role in study design, data collection and interpretation, or the decision to submit the work for publication.
Acknowledgements
The authors thank members of the Sabatini Lab for helpful discussions related to this project and manuscript. We thank the Neurobiology Department and the Neurobiology Imaging Facility for consultation and instrument availability that supported this work. This facility is supported in part by the Neural Imaging Center as part of an NINDS P30 Core Center grant #NS072030.
Ethics
Animal experimentation: All experimental manipulations were performed in accordance with a protocol (#03551) approved by the Harvard Standing Committee on Animal Care following guidelines described in the US National Institutes of Health Guide for the Care and Use of Laboratory Animals.
Reviewing Editor
- Sacha B Nelson, Brandeis University, United States
Publication history
- Received: January 9, 2015
- Accepted: February 26, 2015
- Accepted Manuscript published: February 27, 2015 (version 1)
- Version of Record published: March 24, 2015 (version 2)
Copyright
© 2015, Saunders et al.
This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.
Metrics
-
- 7,263
- Page views
-
- 1,724
- Downloads
-
- 133
- Citations
Article citation count generated by polling the highest count across the following sources: Scopus, Crossref, PubMed Central.
Download links
Downloads (link to download the article as PDF)
Open citations (links to open the citations from this article in various online reference manager services)
Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)
Further reading
-
- Computational and Systems Biology
- Neuroscience
Inhibition is crucial for brain function, regulating network activity by balancing excitation and implementing gain control. Recent evidence suggests that beyond simply inhibiting excitatory activity, inhibitory neurons can also shape circuit function through disinhibition. While disinhibitory circuit motifs have been implicated in cognitive processes including learning, attentional selection, and input gating, the role of disinhibition is largely unexplored in the study of decision-making. Here, we show that disinhibition provides a simple circuit motif for fast, dynamic control of network state and function. This dynamic control allows a disinhibition-based decision model to reproduce both value normalization and winner-take-all dynamics, the two central features of neurobiological decision-making captured in separate existing models with distinct circuit motifs. In addition, the disinhibition model exhibits flexible attractor dynamics consistent with different forms of persistent activity seen in working memory. Fitting the model to empirical data shows it captures well both the neurophysiological dynamics of value coding and psychometric choice behavior. Furthermore, the biological basis of disinhibition provides a simple mechanism for flexible top-down control of the network states, enabling the circuit to capture diverse task-dependent neural dynamics. These results suggest a biologically plausible unifying mechanism for decision-making and emphasize the importance of local disinhibition in neural processing.
-
- Medicine
- Neuroscience
The available treatments for depression have substantial limitations, including low response rates and substantial lag time before a response is achieved. We applied deep brain stimulation (DBS) to the lateral habenula (LHb) of two rat models of depression (Wistar Kyoto rats and lipopolysaccharide-treated rats) and observed an immediate (within seconds to minutes) alleviation of depressive-like symptoms with a high-response rate. Simultaneous functional MRI (fMRI) conducted on the same sets of depressive rats used in behavioral tests revealed DBS-induced activation of multiple regions in afferent and efferent circuitry of the LHb. The activation levels of brain regions connected to the medial LHb (M-LHb) were correlated with the extent of behavioral improvements. Rats with more medial stimulation sites in the LHb exhibited greater antidepressant effects than those with more lateral stimulation sites. These results indicated that the antidromic activation of the limbic system and orthodromic activation of the monoaminergic systems connected to the M-LHb played a critical role in the rapid antidepressant effects of LHb-DBS. This study indicates that M-LHb-DBS might act as a valuable, rapid-acting antidepressant therapeutic strategy for treatment-resistant depression and demonstrates the potential of using fMRI activation of specific brain regions as biomarkers to predict and evaluate antidepressant efficacy.