Joshua A. Goldberg

Striatal cholinergic interneurons (CINs) can drive local dopamine release via nicotinic acetylcholine receptors (nAChRs) expressed on dopaminergic axons, but their role in modulating serotonin (5-HT) signaling is poorly understood. Here, we show that synchronous activation of CINs directly triggers local 5-HT release in the dorsal striatum via nAChRs expressed on serotonergic axons. This CIN–5-HT coupling is not detectable in the ventral striatum, despite its substantially denser serotonergic innervation. The nAChR-dependent release not only increases 5-HT levels in the dorsal striatum, but also expands the spatial footprint of serotonergic signaling. In Sapap3^-/- mice, a model of obsessive-compulsive disorder (OCD)-like behavior, this mechanism is exaggerated due to a hypercholinergic state, selectively amplifying the nAChR-dependent component of monoamine release. These findings demonstrate a regionally confined form of acetylcholine–5-HT crosstalk in the striatum and identify CINs as regulators of 5-HT dynamics in both healthy and pathological states.

Striatal output is dynamically modulated by cholinergic interneurons (CINs), the primary source of acetylcholine in the striatum. CINs have been classically viewed as a random and homogeneous population, but recent evidence suggests heterogeneity in their anatomical and functional organization. Here, using systematic mapping and quantitative spatial analyses, we found that—contrary to current dogma—CINs exhibited striking enrichment and nonrandom clustering in the striosome compartment, particularly in the lateral striatum. Similar analyses carried out for parvalbumin- and somatostatin-expressing interneurons revealed that compartmental organization is interneuron specific. The strong “striosome preference” exhibited by CINs was confined within striosome borders, not extending to the surrounding matrix. We further found that striosome and matrix CINs differed in their expression levels of phospho-S6 ribosomal protein-Ser240/244 and choline acetyltransferase, suggesting functional differences, and clustered CINs differed from unclustered CINs in their intrinsic membrane properties. Finally, CINs expressing Lhx6, which defines a distinct γ-aminobutyric acid (GABA) coreleasing population, were notably absent from regions where highly clustered striosomal CINs appeared. Collectively, our findings uncover important dimensions of CIN organization, suggesting that modulation of regional and compartmental striatal output may depend upon the spatial–functional heterogeneity of CINs.

Objectives: Levodopa treatment, the gold standard therapy for Parkinson's disease, can lead to levodopa‐induced dyskinesias (LIDs). Recent studies in acute brain slices have implicated the neural dynamics of striatal cholinergic interneurons (CINs) in the pathophysiology of LID, reporting that their discharge becomes more bursty in dyskinetic mice. However, little is known about how parkinsonism and levodopa affect CINs in vivo. Methods: We used microendoscopic imaging in the dorsal striatum of the genetically encoded Ca ²⁺ indicator, GCaMP6f, expressed in cholinergic neurons. We imaged the collective, simultaneous activity of several molecularly identified CINs in the dorsal striatum, in control mice, as well as in mice rendered parkinsonian with 6‐hydroxydopamine (6‐OHDA). Results: We found that 6‐OHDA drove a levodopa‐insensitive increase in the baseline rate of activity of CINs and a reduction in their responsiveness to self‐initiated movement. Furthermore, levodopa treatment (even before the appearance of LIDs) drove an increase in the amplitude of the individual Ca ²⁺ events (perhaps indicative of more burst‐like discharge), which correlated with a further reduction in the number of CINs responsive to movement. Conclusions: 6‐OHDA treatment disrupts the robust encoding of movement by CINs. Levodopa disrupts the activity of CINs, possibly making them more bursty in vivo, even before the manifestation of LIDs. Thus, molecularly‐identified CINs are implicated in the pathophysiology of parkinsonism. Not only does levodopa fail to restore their normal dynamics, but it actually exacerbates their abnormal burstiness. © 2025 The Author(s). Movement Disorders published by Wiley Periodicals LLC on behalf of International Parkinson and Movement Disorder Society.

Striatal cholinergic interneurons (CINs) activate nicotinic acetylcholine receptors on dopamine axons to extend the range of dopamine release. Here we show that synchronous activation of CINs induces and extends the range of local serotonin release via a similar mechanism. This process is exaggerated in the hypercholinergic striatum of a mouse model of OCD-like behavior, implicating CINs as critical regulators of serotonin levels in the healthy and pathological striatum.

Striatal cholinergic interneurons (CINs) activate nicotinic acetylcholine receptors on dopamine axons to extend the range of dopamine release. Here we show that synchronous activation of CINs induces and extends the range of local serotonin release via a similar mechanism. This process is exaggerated in the hypercholinergic striatum of a mouse model of OCD-like behavior, implicating CINs as critical regulators of serotonin levels in the healthy and pathological striatum.

A hallmark of Parkinson's disease is the appearance of correlated oscillatory discharge throughout the cortico-basal ganglia (BG) circuits. In the primate globus pallidus (GP), where the discharge of GP neurons is normally uncorrelated, pairs of GP neurons exhibit oscillatory spike correlations with a broad distribution of pairwise phase delays in experimental parkinsonism. The transition to oscillatory correlations is thought to indicate the collapse of the normally segregated information channels traversing the BG. The large phase delays are thought to reflect pathological changes in synaptic connectivity in the BG. Here we study the structure and phase delays of spike correlations measured from neurons in the mouse external GP (GPe) subjected to identical 1–100 Hz sinusoidal drive but recorded in separate experiments. First, we found that spectral modes of a GPe neuron's empirical instantaneous phase response curve (iPRC) elucidate at what phases of the oscillatory drive the GPe neuron locks when it is entrained and the distribution of phases at which it spikes when it is not. Then, we show that in this case the pairwise spike cross-correlation equals the cross-correlation function of these spike phase distributions. Finally, we show that the distribution of GPe phase delays arises from the diversity of iPRCs and is broadened when the neurons become entrained. Modeling GPe networks with realistic intranuclear connectivity demonstrates that the connectivity decorrelates GPe neurons without affecting phase delays. Thus, common oscillatory input gives rise to GPe correlations whose structure and pairwise phase delays reflect their intrinsic properties captured by their iPRCs.

A hallmark of Parkinson’s disease is the appearance of correlated oscillatory discharge throughout the cortico-basal ganglia (BG) circuits. In the primate globus pallidus (GP), where the discharge of GP neurons is normally uncorrelated, pairs of GP neurons exhibit oscillatory spike correlations with a broad distribution of pairwise phase delays in experimental parkinsonism. The transition to oscillatory correlations is considered an indication of the collapse of the normally segregated information channels traversing the BG, and the large phase delays are thought to reflect pathological changes in synaptic connectivity in the BG. Here we study the structure and phase delays of correlations measured from independent neurons in the mouse external GP (GPe) subjected to an identical 1-100 Hz sinusoidal drive. We find that spectral modes of a GPe neuron’s empirical instantaneous phase response curve (iPRC), elucidate at what phases of the oscillatory drive the GPe neuron locks when it is entrained, and the distribution of phases at which it fires when it is not. We show, mathematically, that the spike cross-intensity function (CIF) is the cross-correlation function of the spike phase distributions of a neuronal pair. Moreover, the distribution of GPe CIF phase delays arises from the diversity of iPRCs, and is broadened when the neurons become entrained. Modeling GPe networks with realistic intranuclear connectivity demonstrates that the connectivity decorrelates GPe neurons without affecting CIF phase delays. Thus, common oscillatory drive gives rise to GPe correlations whose diverse structure and pairwise phase delays reflect their intrinsic properties as captured by their iPRCs. Author’s summary The external globus pallidus (GPe) is a hub in the basal ganglia, whose neurons impose a barrage of inhibitory synaptic currents on neurons of the subthalamic nucleus, substantia nigra and internal globus pallidus. GPe neurons normally fire independently, but in experimental parkinsonism, they become correlated in the frequency range associated with the pathological rhythms seen in human Parkinson’s disease, raising the possibility that they may be generators of the pathological oscillation. We drove individual pallidal neurons with an oscillatory input over a wide range of frequencies. Cross-correlations of these neurons reproduced many of the features seen in parkinsonism, suggesting that their correlated oscillations might derive from a shared input rather than internal interconnections.

Understanding dendritic excitability is essential for a complete and precise characterization of neurons ' inputoutput relationships. Theoretical and experimental work demonstrates that the electrotonic and nonlinear properties of dendrites can alter the amplitude (e.g., through amplification) and latency of synaptic inputs as viewed in the axosomatic region where spike timing is determined. The gold-standard technique to study dendritic excitability is using dual-patch recordings with a high-resistance electrode used to patch a piece of distal dendrite in addition to a somatic patch electrode. However, this approach is often impractical when distal dendrites are too fine to patch. Therefore, we developed a technique that utilizes the expression of Channelrhodopsin-2 (ChR2) to study dendritic excitability in acute brain slices through the combination of a somatic patch electrode and optogenetic activation. The protocol describes how to prepare acute slices from mice that express ChR2 in specific cell types, and how to use two modes of light stimulation: proximal (which activates the soma and proximal dendrites in a similar to 100 mu m diameter surrounding the soma) with the use of a high-magnification objective and full-field stimulation through a low-magnification objective (which activates the entire somato-dendritic field of the neuron). We use this technique in conjunction with various stimulation protocols to estimate model-based spectral components of dendritic filtering and the impact of dendrites on phase response curves, peri-stimulus time histograms, and entrainment of pacemaking neurons. This technique provides a novel use of optogenetics to study intrinsic dendritic excitability through the use of standard patch-clamp slice physiology.