Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 15 (1), 123-30

GABAergic Circuits Mediate the Reinforcement-Related Signals of Striatal Cholinergic Interneurons


GABAergic Circuits Mediate the Reinforcement-Related Signals of Striatal Cholinergic Interneurons

Daniel F English et al. Nat Neurosci.


Neostriatal cholinergic interneurons are believed to be important for reinforcement-mediated learning and response selection by signaling the occurrence and motivational value of behaviorally relevant stimuli through precisely timed multiphasic population responses. An important problem is to understand how these signals regulate the functioning of the neostriatum. Here we describe the synaptic organization of a previously unknown circuit that involves direct nicotinic excitation of several classes of GABAergic interneurons, including neuroptide Y-expressing neurogilaform neurons, and enables cholinergic interneurons to exert rapid inhibitory control of the activity of projection neurons. We also found that, in vivo, the dominant effect of an optogenetically reproduced pause-excitation population response of cholinergic interneurons was powerful and rapid inhibition of the firing of projection neurons that is coincident with synchronous cholinergic activation. These results reveal a previously unknown circuit mechanism that transmits reinforcement-related information of ChAT interneurons in the mouse neostriatal network.


Figure 1
Figure 1. Characterization of GABAergic IPSCs elicited in SPNs with optogenetic stimulation of ChAT interneurons
a. Confocal images of a ChR2-YFP expressing neuron (top left) immunostained for ChAT (middle; bottom: overlay). A larger field is shown at the right. b. Top: Optically elicited action potential in a ChAT interneuron. Bottom: A cell-attached recording of spontaneous activity and optically evoked action potentials (red arrows) of a ChAT interneuron. c. Optogenetically elicited IPSPs in an SPN (arrow) efficiently blocked firing induced by current injection. Bottom trace, corresponding IPSC, (Vhold=-80 mV). d. Optogenetically elicited IPSCs in 2 SPNs (blue traces) were blocked by bicuculline (left) or DHβE (right, red traces). e. Kinetic components of the compound IPSC. Left: 3 distinct components of the IPSC exhibiting different τdecay values. Right: Non-monotonic transition between the fIPSC and the sIPSC. Note the negative inflection following the transition (red arrowhead). Inset: Decomposition of the compound IPSC (black trace) into a fIPSC (blue trace) and a sIPSC (red trace). f. Independent trial-to-trial amplitude variance of the fIPSC and the sIPSC. Left: overlay of 4 responses exhibiting identical sIPSC but different fIPSC components (colored arrows point to fIPSC peaks). Shaded areas are averaging windows. Middle: variable sIPSC components. Right: relative sIPSC amplitudes plotted against corresponding relative peak fIPSC amplitudes (n=5). Red line is linear regression (p>0.2). g. Inhibition of GAT-1 selectively prolongs the sIPSC. Note that the fIPSC is unaffected (arrow and inset). Red arrow, transition point of the response components. In all figures blue bars represent optical stimuli.
Figure 2
Figure 2. Synaptic interactions of ChAT and NPY-NGF interneurons and SPNs
a. Top: Characteristic passive and active properties of ChAT and NPY-NGF interneurons. Bottom: Synaptic circuitry of a ChAT and a NPY-NGF interneuron and a SPN recorded simultaneously (same as in a, c-f). GABAergic (blue) and nicotinic (red) interactions are indicated. Circular arrow represents recurrent inhibition. b. Action potentials of the ChAT interneuron elicited in voltage clamp (bottom) elicited nEPSPs in the NPY-NGF interneuron (top; red, average) that were blocked by DHβE (blue). c. Recurrent GABAergic IPSCs in the same ChAT interneuron (spike subtracted). Note the short τdecay of the IPSC and block by bicuculline (blue trace). d. Train stimulation of the ChAT interneuron (middle) elicits nEPSPs in the NPY-NGF neuron (top traces, arrows). Note partial depression of the nEPSPs contrasted with the complete failure of the recurrent IPSC after the first stimulus (bottom traces, arrowheads). e. Spiking in the NPY-NGF interneuron (middle) elicits IPSCs in the SPN (bottom; blue, average) and the ChAT interneuron (top, blue, average). Note the slow time course of the IPSCs and the absence of a fIPSC in the SPN. Bicuculline block, green. f. Top: Inhibition of GAT-1 increases τdecay of the IPSC elicited in an SPN by an NPY-NGF interneuron (different neurons than in e-f). Bottom: Overlay of the same IPSCs (green traces) with optogenetic IPSCs recorded under the same pharmacological conditions (blue, same as in Fig. 1). Unitary responses were scaled in amplitude. Note the similarity of rise times and τdecay.
Figure 3
Figure 3. Optogenetic activation of ChAT interneurons elicits nEPSPs and GABAergic IPSPs and triggers action potential firing in NPY-NGF interneurons
a. Confocal image of ChAT interneurons expressing ChR2-mCherry (arrows) and 2 EGFP expressing NPY-NGF neurons (arrowheads) in a ChAT-Cre/NPY-GFP mouse. b. Top: Confocal images of a NPY-NGF interneuron (arrowhead) and a simultaneously recorded nearby SPN intracellularly labeled with Alexa 594 (arrow). Bottom: Photomicrographs of the neurons during recording. Inset: high magnification showing SPN dendritic spines (arrows, same neurons as in d-e). c. Top panel: Optical stimulation of ChAT interneurons elicited large amplitude depolarizations and action potentials in a NPY-NGF interneuron that were blocked by DHβE (blue trace). Simultaneously elicited optogenetic compound IPSC in the SPN (bottom green trace). Bottom panel: Single action potential in the NPY-NGF neuron (black trace) elicited a slow IPSC in the SPN (green trace). d. EPSP-IPSC sequence elicited with optical stimulation of ChAT interneurons in an NPY-NGF neuron (different neurons from b-e). Left: reversal of the IPSP (arrow). Right: The IPSP (green trace), the compound response (black trace) and the isolated nEPSP (blue trace) recorded at ∼-45 mV. Note the large amplitude and slow time course of the isolated IPSP. e. The compound response of another NPY-NGF interneuron (black trace) was gradually increased in amplitude by application of bicuculline (red trace), leading to action potential firing (top inset). The isolated nEPSP (red trace) was blocked by ∼95% by 200 nM DHβE (blue trace). The residual response (blue trace, bottom inset) was not DNQX sensitive (green trace).
Figure 4
Figure 4. FSIs do not mediate the inhibition of SPNs by ChAT interneurons
a. Paired recording from a synaptically connected FSI and SPN. Intracellularly injected current pulses in the interneuron elicit voltage responses that identify it as an FSI (bottom). Note the typical large amplitude IPSCs elicited in the SPN (arrows). b. Optical stimulus (blue bar) elicits a compound IPSC, including a large fIPSC in the SPN (blue trace, top) but fails to trigger action potentials or large depolarizing potentials in the FS interneuron (red trace, bottom). Compare to the responses elicited in NPY-NGF neurons in Fig. 3d, g.
Figure 5
Figure 5. Optogenetically reproduced pause-excitation population response of ChAT interneurons elicits powerful inhibition in SPNs in vitro
a. Photomicrograph of eNpHR3.0-YFP-expressing ChAT interneurons (arrow, arrowhead) labeled intracellularly with Alexa 594 (red). b. Top: Responses of an eNpHR3.0-YFP expressing ChAT interneuron to intracellular current (a, arrow). Whole cell (middle) and cell-attached (bottom) recordings demonstrated spontaneous activity and large amplitude optogenetic hyperpolarization (green bar) leading to rebound excitation (arrows). c. Synaptic responses of a SPN to a pause-excitation population response of ChAT interneurons. Top: Rebound excitation of the interneurons triggered coincident large IPSPs (arrow) that efficiently blocked action potential generation. Middle: Spike trains of ChAT interneurons recorded using cell-attached (asterisks) and extracellular (arrowheads) recording (second trace from top) and in current clamp (color traces). Inset: eNpHR3.0-YFP expressing ChAT interneurons and cell attached recording pipettes (red). Bottom: PSTH of ChAT interneurons demonstrating pronounced pause-excitation activity. d. Simultaneous current recordings (blue and red) from 2 SPNs showing IPSCs elicited by the rebound activation of ChAT interneurons (bottom) induced using eNpHR1.0-mCherry. Simultaneous voltage recording from a third SPN showed optically elicited spike delay (top, arrow). Bottom: Extracellular recording of a ChAT interneuron (spikes, vertical lines) showed optical inhibition and rebound firing. e. Top: Short and long latency inhibition (black trace, red and blue arrows) in a SPN elicited by ChAT interneurons were blocked by DHβE (red traces). Bottom: Cell attached recording demonstrated long latency spikes in a ChAT interneuron (arrows) coinciding with late inhibition. Note that the early inhibition is elicited by other ChAT interneurons.
Figure 6
Figure 6. Pause-excitation sequences of ChAT interneurons inhibit SPNs in vivo in freely moving mice
a. Left: Waveforms of distinct types of units included in analysis. Average unit waveform is shown in gray, population averages are in color. Overlay demonstrates feature differences between unit types. Right: Firing rates (mean±s.e.m) of the 3 type of units. SPNs exhibit significantly lower firing rates than other unit types (t-test, p<0.01). b. Examples of spike trains of putative SPNs and ChAT interneurons. Note that a bursting episode was selected for the SPN. c. Characteristics of the population response of ChAT interneurons elicited with optogenetic inhibition. Note the instantaneous inhibition of firing and the excitation phase that is similar to responses of putative ChAT interneurons in primates (bins, 30 ms). d. Inhibition of firing of SPNs (bottom 3 PSTHs) by pause-excitation activity pattern of ChAT interneurons (top). Lower 3 PSTHs show (respectively, from top to bottom) cumulative response of all SPNs, SPN responses following 200 ms optical inhibition and responses following 1000 ms inhibition. The population mean and mean-2SD firing rates are indicated by blue and red lines, respectively. Note that the end of the optical stimulus is closely followed by strong inhibition of firing in SPNs. Consecutive bins with firing rates below mean-2SD are indicated by bins colored in blue. Note that the optical inhibition itself does not elicit statistically significant firing rate changes in the SPNs. Horizontal bars denote periods of illumination (bins, 50 ms).

Similar articles

See all similar articles

Cited by 130 articles

See all "Cited by" articles


    1. Graybiel AM. Neurobiol Learn Mem. 1998;70:119–36. - PubMed
    1. Schultz W. J Neurophysiol. 1998;80:1–27. - PubMed
    1. Morris G, Arkadir D, Nevet A, Vaadia E, Bergman H. Neuron. 2004;43:133–43. - PubMed
    1. Joshua M, Adler A, Mitelman R, Vaadia E, Bergman H. J Neurosci. 2008;28:11673–84. - PMC - PubMed
    1. Hyland BI, Reynolds JN, Hay J, Perk CG, Miller R. Neuroscience. 2002;114:475–92. - PubMed

Publication types

MeSH terms