Activation of the Rostral Intralaminar Thalamus Drives Reinforcement through Striatal Dopamine Release

Abstract

Glutamatergic projections of the thalamic rostral intralaminar nuclei of the thalamus (rILN) innervate the dorsal striatum (DS) and are implicated in dopamine (DA)-dependent incubation of drug seeking. However, the mechanism by which rILN signaling modulates reward seeking and striatal DA release is unknown. We find that activation of rILN inputs to the DS drives cholinergic interneuron burst-firing behavior and DA D2 receptor-dependent post-burst pauses in cholinergic interneuron firing. In vivo, optogenetic activation of this pathway drives reinforcement in a DA D1 receptor-dependent manner, and chemogenetic suppression of the rILN reduces dopaminergic nigrostriatal terminal activity as measured by fiber photometry. Altogether, these data provide evidence that the rILN activates striatal cholinergic interneurons to enhance the pursuit of reward through local striatal DA release and introduce an additional level of complexity in our understanding of striatal DA signaling.

Keywords: Parkinson’s disease; addiction; associative thalamus; caudate; goal-directed behavior; movement; putamen.

Figure 1.. Thalamic rILN Activate Striatal Cholinergic Interneurons

(A) Schematic of experimental approach. (B) Representative expression of channelrhodopsin (ChR2)-eYFP at the rostral intralaminar nuclei of the thalamus (rILN) injection site. (C) Representative expression of ChR2-eYFP in rILN axon terminals (green) and cholinergic interneurons (red) in the dorsal striatum (DS). (D) Recorded tdTomato-expressing cholinergic interneuron. (E) Sample trace of a cell-attached cholinergic interneuron recording. (F) Left: NBQX and AP5 application reduced the average peak amplitude of post-synaptic currents evoked by optogenetic activation (oPSCs; one 2–4 ms pulse, blue bar) of ChR2-expressing rILN terminals (orange; n = 6 cells) versus control aCSF (gray; n = 4 cells). Right: representative oPSCs: control (black), after incubation in aCSF (gray), or NBQX and AP5 (orange). (G) Top: representative whole-cell current-clamp recording of burst firing in a cholinergic interneuron in response to optogenetic rILN terminal activation (15 pulses at 10 Hz), followed by a firing pause. Middle: representative raster plots of six consecutive responses. Bottom: corresponding histogram of average firing per second. (H) Time to first spike following burst firing was greater than the average interspike interval (ISI) during tonic firing (n = 11 cells). (I) Representative cholinergic interneuron recording following incubation in 5 μM sulpiride (top), with a raster plot of six consecutive responses (middle) and corresponding histogram (bottom). (J–M) Comparison of cholinergic interneuron burst-pause firing properties between control (n = 12 cells) and in sulpiride (n = 12 cells) of (J) baseline firing rate, (K) burst firing rate, (L) post-burst-pause length, and (M) ratio of post-burst ISI to mean baseline ISI. aCSF, artificial cerebral spinal fluid; cc, corpus callosum; fr, fasciculus retroflexus. Scale bars: 250 μm (B and C) and 40 μm (D). Unpaired t test (F); Wilcoxon signed-rank test (H); Mann-Whitney test (J–M): *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Individual values are represented in gray or orange, with mean ± SEM in black.

Figure 2.. rILN Afferent Activation Evokes DA Release in the DS

(A) Approach for fast-scan cyclic voltammetry (FSCV) recordings. (B) Left: optogenetic rILN terminal activation in the DS elicited a dopamine (DA) current with a corresponding cyclic voltammogram (top) and a plot of the background-subtracted current (bottom). Right: representative electrode calibration of the current to known DA concentrations (3–4 recordings per concentration). (C) DA release time course in control (black; n = 8 recordings) and following NBQX application (orange; n = 8 slices). (D) AP5 (orange) did not alter rILN-induced DA release (n = 8 slices). (E) Mecamylamine (orange) abolished DA release (n = 8 slices). (F) Quinpirole (orange) eliminated DA release (n = 7 slices). Unpaired t test (C–F): ****p < 0.0001. Data represent mean ± SEM.

Figure 3.. Chemogenetic Suppression of rILN Firing Decreases Dopaminergic Nigrostriatal Terminal Activity and Locomotion

(A) Approach for chemogenetic rILN terminal inhibition and synaptic transmission interrogation. (B) Left: recording from DS medium spiny neurons (MSNs) in striatal slices; application of clozapine-N-oxide (CNO; red) attenuated rILN-evoked oPSC amplitude (n = 6 cells) as compared to control aCSF (gray; n = 9 cells). Right: representative oPSC traces, before (black) and after aCSF (gray) or CNO (red) incubation. (C) Left: experimental strategy to chemogenetically suppress rILN activity and monitor dopaminergic nigrostriatal terminal activity. Middle: timeline of drug administration and behavioral testing. Right: representative expression of GCaMP6f (green) in the substantia nigra (SN). (D) Emission profile of GCaMP6f in the DS, with the peak GCaMP6f channel (white arrow) and off-peak channel (gray arrow). (E) Left: representative traces of the peak signal in hM4Di and mCherry-expressing mice moving freely in an open field following vehicle (Veh; black) or CNO (blue; 5 mg/kg) injection. Right: average peak signal was decreased in hM4Di-expressing mice (N = 16 mice), but not mCherry-expressing mice (N = 14 mice), by CNO. (F–H) Changes in movement between groups: (F) total distance traveled, (G) maximum velocity, and (H) percentage of time in motion. Scale bar: 250 μm. Unpaired t test (B); post hoc Holm-Šidàk test (E–H): *p < 0.05, **p < 0.01, ***p < 0.001. Individual values are represented in gray; data represent mean ± SEM.

Figure 4.. Optogenetic Activation of rILN Afferents in the DS Is Reinforcing

(A) Top left: approach for in vivo optogenetic rILN-terminal activation and ChR2-eYFP expression relative to fiber placement in DS. Right and bottom left: experimental timeline and procedure for a two-day optical intracranial self-stimulation paradigm (oICSS). (B) ChR2-eYFP-expressing mice pressed the light-paired lever (blue fill) more than the non-reinforced lever (black fill) in a light and DA D1 receptor-dependent manner. (C) eYFP-expressing mice did not press the light-paired lever more than the non-reinforced lever (N = 9 mice). (D) Left: strategy to ablate DS cholinergic interneurons at site of light delivery. Right: representative taCasp3-mediated ablation of cholinergic interneurons (red) and expression of ChR2-eYFP-expressing rILN terminals (green) in the DS. (E) mCherry-expressing mice (N = 7) pressed the light-paired lever (blue) more than the non-reinforced lever (black) on both test days. taCasp3-ablated mice (N = 6) only pressed the light-paired lever more on day 2. FR1, fixed-rate 1; SCH, SCH23390. Scale bars = 250 μm. Post hoc Holm-Šidàk test: *p < 0.05, **p < 0.01, ****p < 0.0001. Individual values are represented in gray; data represent mean ± SEM. See also Figure S2.