Selective labeling and functional expression of ChR2 in specific cortical neurons projecting to striatal D1- vs. D2-SPNs.

(A) Schematic of SAD-ΔG-Rabies-ChR2-mCherry construct with the glycoprotein deleted and replaced with ChR2-mCherry. (B) Timeline of viral injections of Cre-dependent helper viruses and the modified rabies virus for slice and behavioral experiments. (C) Example of coronal brain section with rabies-eGFP injection in the dorsal medial striatum of D1-Cre (top) or A2a-Cre (bottom) mouse shows enriched eGFP expression in the MCC. Scale bar, 1 mm. Inset (right): Higher magnification of retrogradely-labeled striatal D1- or D2-SPN projecting neurons in the MCC expressing eGFP. Dotted lines demarcate cortical lamina. Scale bar, 200 µm. (D) Similar experiments of labeling striatal D1- vs. D2-SPN projecting neurons in M1 with rabies-eGFP. (E) Example of coronal brain section with rabies-ChR2-mCherry injection in the dorsal lateral striatum of A2a-Cre mouse. Scale bar, 1 mm. Inset (right): Higher magnification of retrogradely-labeled striatal D2-SPN projecting neurons in the M1 showed clear membrane expression of ChR2-mCherry. Scale bar, 200 µm. (F) Cartoon brain schematic of ChR2-mCherry expressing M1 neurons projecting to D1-SPNs (red) during whole-cell patch clamp recordings. (G) (left) 10x epifluorescent (red channel) of ChR2-mCherry positive neurons in M1. Scale bar, 250 µm. (middle) 40x image of a patched layer 5 pyramidal neuron under DIC optics. Scale bar, 50 µm. (right) Epifluorescent image (red channel) showing patched layer 5 pyramidal neuron somas expressing ChR2-mCherry signal. Red dotted line denotes patched neuron. Scale bar, 50 µm. (H) Current-voltage traces of a ChR2-mCherry positive layer 5 M1 neuron under current clamp responding to hyperpolarizing and depolarizing current injection steps. Scale bars, 200 ms, 25 mV. (I) Optogenetic stimulation (20 Hz) elicits robust action potentials with high fidelity in a ChR2-mCherry positive D1-SPN projecting M1 neuron in layer 5. Scale bars, 100 ms, 20 mV.

The excitatory inputs to striatal D1- vs. D2-SPNs are partially segregated with asymmetrical overlapping.

(A) Schematic for the possible organization of the excitatory inputs to striatal D1- vs. D2-SPNs from completely segregated (left), totally overlapping (middle), to partially mixed (right). The red and black filled circles indicate the individual neurons projecting to D1- vs. D2-SPNs, respectively. The half red and half black circles imply the neurons projecting to both. (B) (left) Schematic of rabies-ChR2 labeling of the inputs to D1-SPNs and whole-cell recordings of rabies-negative striatal D1-SPNs with local optogenetic stimulation. (right) Example of the average EPSC trace showing short latency response to paired pulses (50 ms ISI) stimulation (black), that is blocked by AMPAR and NMDAR antagonists (gray). All recordings were conducted in the presence of picrotoxin (PTX) to isolate excitatory transmission. Scale bar, 25 ms, 100 pA. Same conditions applied to all following recordings. (C) Whole-cell recording of rabies-negative striatal D2- SPNs with local optogenetic stimulation with rabies-ChR2 labeling of the inputs to D2-SPNs. (D) The likelihood of the inputs to D1-SPNs form a functional connection with nearby non-starter D1-SPNs, and the likelihood of the D2-SPN situation. Numbers above the bars denote number of cells that show functional connectivity within total recorded. Fisher’s exact test, P = 0.3137. (E-F) Whole-cell recording of rabies-negative striatal D2-SPNs with local optogenetic stimulation with rabies-ChR2 labeling of the inputs to D1- SPNs (E), and recording of rabies-negative D1-SPNs with stimulation of inputs to D2-SPNs (F). (G) The likelihood of the inputs to D1-SPNs form a functional connection with nearby non-starter D2-SPNs, and the likelihood of the inputs to D2-SPNs form a functional connection with nearby non-starter D1-SPNs. Fisher’s exact test, P = 0.0079. **, P < 0.01.

Different effects of optogenetic stimulation of D1- vs. D2-SPN projecting cortical neurons on locomotion and reinforcement learning.

(A) Schematic of largely segregated yet partially overlapping excitatory inputs to striatal D1- vs. D2-SPNs. (B) Schematic of dorsal medial striatum (DMS) injection of Cre-dependent AAV-ChR2 and optogenetic simulation in D1- or A2a-Cre mice. (C) (top) Example of locomotion path under control (black) and following 20Hz optogenetic stimulation (gray) of DMS D1-SPNs in open field. Scale bar, 5cm, same for below. (bottom) Stimulation of D1-SPNs in DMS facilitates locomotion (n = 5, unpaired two-tailed t-test, t = 3.386, P = 0.0046). **, P < 0.01. (D) 20Hz stimulation of D2-SPNs in DMS suppresses locomotion (n = 5, unpaired two-tailed t-test, t = 2.559, P = 0.0227). *, P < 0.05. (E) Schematic for dorsal lateral striatum (DLS) optogenetics. (F-G) 20Hz stimulation of D1-SPNs in DLS facilitates locomotion (F, n = 5, unpaired two-tailed t-test, t = 4.736, P = 0.0003), while stimulation of D2-SPNs in DLS does not significantly suppress locomotion in open field (G, n = 5, unpaired two-tailed t-test, t = 1.026, P = 0.3224). ***, P < 0.001. (H) Schematic of rabies-ChR2 labeling of the inputs to D1 or D2-SPNs and optogenetic stimulation in MCC. (I-J) 20Hz stimulation of MCC neurons projecting to D1-SPNs facilitates locomotion (I, n = 9, unpaired two-tailed t-test, t = 2.344, P = 0.0344), while stimulation of MCC neurons projecting to D2- SPNs does not alter locomotion (J, n = 10, unpaired two-tailed t-test, t = 1.214, P = 0.2447). *, P < 0.05. (K) Schematic of rabies-ChR2 labeling of the inputs to D1 or D2-SPNs and optogenetic stimulation in M1. (L-M) 20Hz stimulation of the M1 neurons projecting to D1-SPNs facilitates locomotion (L, n = 7, Unpaired two-tailed t-test, t = 3.276, P = 0.0055), while stimulation of the M1 neurons projecting to D2-SPNs does not significantly alter locomotion (M, n = 8, Unpaired two-tailed t-test, t = 0.5796, P = 0.5714). **, P < 0.01. (N) Schematic of a mouse performing intracranial self-stimulation (ICSS) behavior. (O-P) D1-SPN (red) but not D2-SPN stimulation (black) drives ICSS behavior in either DMS (O, D1, n = 6; D2, n = 5; Mann Whitney test, Day 9 D1 vs. A2a, P = 0.0130) or DLS (P, D1, n = 6; D2, n = 5; Mann Whitney test, Day 9 D1 vs. A2a, P = 0.0433). *, P < 0.05. (Q) Timeline of helper virus injections, rabies-ChR2 injections and optogenetic stimulation for ICSS behavior. (R-S) Optogenetic stimulation of the cortical neurons projecting to either D1- or D2-SPNs drive ICSS behavior in both MCC (R, n = 5 per group, no significant effect of genotype F(1,8) = 1.074, P = 0.3303) and M1 (S, n = 5 per group, no significant effect genotype F(1,8) = 2.767, P = 0.1348). n.s., not statistical significant.

Optogenetic stimulation of D1- vs. D2-SPN projecting cortical neurons differently modulates action sequence execution.

(A) Schematic of a mouse performing FR8 sequence. (B) Optogenetic stimulation (20Hz) of the D1-SPN projecting MCC neurons during FR8 sequence. Example lever pressing (black bar) in control (top) vs. stimulation (middle) trials aligned to the first press, where the blue transparent rectangle corresponds to the window of optogenetic stimulation (20Hz, 8s). The black and blue lines in the PETH (bottom) indicate the lever press rate for control and stimulation conditions, respectively, same for below. (C) Optogenetic stimulation (20Hz) of the D2-SPN projecting MCC neurons during FR8 sequence. (D) Average percent lever press rate change during optogenetic stimulation of D1- vs. D2-SPN projecting MCC neurons compared to control (MCC – D1, n = 8; MCC – D2, n = 7; Unpaired two-tailed t-test, t = 2.774, P = 0.0097). **, P < 0.01. (E) Timeline of helper virus injections, rabies-ChR2 injections and optogenetic stimulation during action sequence performance. (F-G) Optogenetic stimulation (20Hz) of the D1- (F) or D2-SPN (G) projecting M1 neurons during FR8 sequence. (H) Average percent lever press rate change during optogenetic stimulation of D1- vs. D2-SPN projecting M1 neurons compared to control (M1 – D1, n = 6; M1 – D2, n = 7; Unpaired two-tailed t-test, t = 0.7651, P = 0.4511).

Low-frequency (5 Hz) optogenetic stimulation of cortical neurons projecting to striatal D1- or D2-SPNs has little effect on locomotion activity.

(A) 5 Hz optogenetic stimulation elicits action potentials with high fidelity in a ChR2-mCherry positive M1 pyramidal neuron in layer 5 projecting to striatal D1-SPNs. Scale bars, 200 ms, 25 mV. (B-C) 5 Hz optogenetic stimulation on MCC neurons projecting to either D1- or D2-SPNs didn’t change the locomotion activity. MCC – D1, n = 9, unpaired two-tailed t-test, t = 0.1906, P = 0.8516. MCC – D2, n = 10, unpaired two-tailed t-test, t = 1.015, P = 0.3275. (D) 5 Hz optogenetic stimulation of M1 neurons projecting to D1-SPNs didn’t change the locomotion activity. n = 7, unpaired two-tailed t-test, t = 0.276, P = 0.7866. (E) 5 Hz optogenetic stimulation of M1 neurons projecting to D2-SPNs slightly increased locomotion activity. n = 8, unpaired two-tailed t-test, t = 2.48, P = 0.0265. *, P < 0.05.

The synaptic properties of projections from D1- or D2-SPN retrogradely-labeled cortical inputs to striatal D1- or D2-SPNs.

(A-D) The EPSC latency (A), amplitudes (B), paired pulse ratio (C) and variation (D) of whole-cell recordings of rabies-negative striatal D1- or D2-SPNs, with optogenetic stimulation of the terminals of D1- or D2-SPN retrogradely-labeled cortical neurons. n.s., P > 0.05, not statistical significant.

No effects of optogenetic stimulation of M1 on locomotion in mice with ChR2 expression in either D1- or D2-SPNs of DMS.

(A) Schematic of dorsal medial striatum (DMS) injection of Cre-dependent AAV-ChR2 in D1- and A2a-Cre mice with optogenetic simulation in M1. (B) 20Hz optogenetic stimulation of M1 in mice expressing ChR2 in striatal D1-SPNs didn’t change the locomotion activity. n = 5, unpaired two-tailed t-test, t = 0.1016, P = 0.9194. (C) 20Hz optogenetic stimulation of M1 in mice expressing ChR2 in striatal D2-SPNs didn’t alter the locomotion activity. n = 5, unpaired two-tailed t-test, t = 1.155, P = 0.2525.