Calcium imaging fiber photometry reveals that STN glutamatergic neurons activate during spontaneous exploratory movement.

A, Parasagittal section showing the optical fiber tract reaching STN and GCaMP7f fluorescence expressed in glutamatergic neurons around the fiber ending. The section was aligned with the Allen brain atlas. ZI, zona incerta; SNr, substantia nigra pars reticulata; STN, subthalamic nucleus. B, Cross-correlation between movement and STN F/Fo for the overall (black traces), rotational (red) and translational (cyan) components (upper panel). Per session (dots) and mean±SEM (rectangle) linear fit (correlation, r) between overall movement and STN F/Fo, including the rotational and translational components (lower panel). The lighter dots show the linear fits after scrambling one of the variables (lower panel, shuffled). C, F/Fo calcium imaging time extracted around detected spontaneous movements. Time zero represents the peak of the movement. The upper traces show F/Fo mean±SEM of all movement peaks (black), those that had no detected peaks 3 sec prior (red), and peaks taken at a fixed interval >5 sec (cyan). The lower traces show the corresponding movement speed for the selected peaks. All traces in the paper are mean±SEM.

STN glutamatergic neurons code the direction of spontaneous contraversive exploratory turning movements.

A, F/Fo calcium imaging, overall movement, rotational movement, and angle of turning direction for detected movements classified by the turning direction (ipsiversive and contraversive; red and cyan) versus the side of the recording (implanted optical fiber). At time zero, the animals spontaneously turn their head in the indicated direction. The columns show all turns (left), those that included no turn peaks 3 sec prior (middle), and peaks selected at a fixed interval >5 sec (right). Note that the speed of the movements was similar in both directions. B, Population measures (area of traces 3 sec around the detected peaks) of F/Fo and movement (overall, rotational, and translational) for the different classified peaks. Asterisks denote significant differences (p<0.05) between ipsiversive and contraversive movements.

A subgroup of STN glutamatergic neurons code contraversive movements.

A, Parasagittal section showing a miniscope GRIN lens tract reaching STN and GCaMP7f fluorescence expressed in STN glutamatergic neurons. The section was aligned with the Allen brain atlas. The red inset shows a FOV of imaged cells during a recording session. ZI, zona incerta; STN, subthalamic nucleus. B, Classification of STN glutamatergic neurons during spontaneous movement onsets with k-means reveals three classes (mean±SEM). The top traces show F/F calcium imaging, and the bottom traces show the movement onset. Class 1 neurons activated weakly during movement onset. Class 2 neurons were inhibited while Class 3 neurons activated sharply during movement onset. C, Classification of STN glutamatergic neurons during spontaneous turning movements with k-means reveals three classes (mean±SEM). The top traces show F/F calcium imaging, and the bottom traces show angle of turning direction for detected movements separated by class. The left panels show the activation difference (bias) between contraversive-ipsiversive directions used to classify the cells. The middle and right panels show the corresponding contraversive and ipsiversive movements. Class 1 neurons did not activate during turns and did not code turn direction. Class 2 neurons showed stronger activation in the ipsiversive direction. Class 3 neurons activated more strongly than Class 2 in the contraversive direction. D, Population comparison of F/F Peak amplitude bias (difference between contravesive-ipsiversive direction) for the three classes of neurons. Asterisks denote significant differences (p<0.05) between both directions. The k-means clusters of the three cell classes from C are shown on the top panel. E, Unilateral optogenetic STN excitation drives an ipsiversive head orienting bias. Light patterns (2 sec at different frequencies) are delivered randomly as mice explore an arena. The top left panel shows head orienting angle change from the onset of the light. The bottom left panel shows the movement speed. Note the strong ipsiversive bias during 40-66 Hz STN excitation. The right panels show population data comparing Opsin mice expressing ChR2 in STN neurons with No Opsin mice. Asterisks denote significant differences (p<0.05) between Opsin and No Opsin mice.

STN glutamatergic neurons discharge to auditory tones in association with movement.

A, Example F/Fo calcium imaging and movement traces (mean±SEM) evoked from STN neurons by auditory tones (1 sec) of different saliency. The tones vary in frequency (4-16 kHz) and SPL (low and high dB). B, Area of F/Fo, overall movement, and movement components (rotational and translational) measured during a time window (0-2 s) after tone onset.

STN glutamatergic neuron activation in the context of signaled active avoidance.

A, Arrangement of the shuttle box used during signaled avoidance tasks. B, Behavioral performance during the four different avoidance procedures (AA1-4) showing the percentage of active avoids (black circles), avoidance latency (orange triangles), and ITCs (cyan bars). C, F/Fo and overall movement traces from CS onset for AA1, AA2 and AA3 (CS1 and CS2) procedures per trials classified as avoids (left) or escapes (right) of CS-evoked orienting responses measured by tracking overall head speed. D, Same as C from response occurrence. E, Population measures of F/Fo and speed for avoids and escapes during AA1, AA2 and AA3 (CS1). Asterisks denote significant differences (p<0.05) between avoids and escapes.

STN glutamatergic neurons track the avoidance and escape movement.

A, F/Fo and overall movement traces from CS onset (left) and response occurrence (right) for avoids during the AA4 procedure, which include three CSs that signal avoidance intervals of different durations. B, Population measures (-3 to 3 sec area, mean ± SEM) from response occurrence. Asterisks denote significant differences vs other stimuli. C, F/Fo and overall movement traces from US onset for escapes during the unsignaled US procedure, which includes the US, or each of its components delivered alone (foot-shock and white noise). D, Population measures (0 to 5 sec area, mean ± SEM) for the data in C. Asterisks denote significant differences (p<0.05) between a condition and the other conditions.

Different classes of STN glutamatergic neurons during signaled active avoidance.

A, k-means clustering of the speed time series from CS onset (gray panel) revealed three distinct classes of signaled active avoidance responses. Class 1 avoids start early after CS onset while Class 2 and Class 3 avoids start later reflecting delayed responding. The top panel shows the F/F activation of all the recorded STN neurons during each class of avoids. Class 1 avoids were associated with little STN neuron activation followed by Class 2 avoids, while Class 3 avoids were associated with strong STN activation.B, k-means clustering of the F/F time series within each Class of avoids in A revealed three classes (types: a-c) of neurons. Type a (1-3a) neurons showed little STN activation. Type b (1-3b) neurons activated during all avoids classes but showed inhibition at CS onset prior to Class 1 avoid. Type c (1-3c) showed stronger activation than the other types but activated most sharply during Class 3 avoids, which have the longest response delays. C,D, The same data shown in A and B are displayed from avoidance response occurrence (instead of CS onset). E, Population comparison of F/F parameters for the classes of avoids and neuron types shown in A-D. Asterisks denote significant differences (p<0.05) between the noted avoid classes or neuron types. The panels show baseline (prior to CS onset; left panel), orienting and avoidance responses from CS onset (middle panels) or from avoid occurrence (right panel). F, Same as E for the corresponding movement (speed) parameters (baseline corrected areas of speed traces).

Optogenetic activation of ChR2-expressing STN glutamatergic neurons drives signaled active avoidance and can substitute for the natural CS.

A, Parasagittal section showing the optical fiber tract reaching STN and ChR2 fluorescence expressed in glutamatergic neurons. The section was aligned with the Allen brain atlas. B, Effects of blue light patterns on signaled active avoidance. Blue circles show the effect of CS+Light trials (filled blue circles) compared to control CS trials (open blue circle) and catch NoCS trials (open blue squares). Note the shortening of the latency for higher frequencies including Cont light. The cyan triangles show the results for equivalent Light trials, where the CS is excluded, and the light serves as the CS. Light trials at low frequencies are ineffective at driving avoids. C, Traces of the overall speed during AA1 for CS trials and CS+Light trials for different light patterns. The trials are aligned by CS onset (left and middle panels), which reveals the orienting response evoked by the CS followed by the avoidance action, or from avoid occurrence (right panel), which reveals the peak speed as the mice cross between compartments. D, Population measures of peak speed and time to peak speed for CS and CS+Light trials from CS onset, for the orienting response and the avoidance response peaks from CS onset (left and middle panels). The right panel compares the peak speed measured from the avoid occurrence. CS+Light trials data are combined according to the light pattern into low frequency (2-20 Hz) or high frequency (>20 Hz and Cont). Asterisks denote significant differences (p<0.05) between CS trials and CS+Light trials. E, and F, Same as C and D but compares CS trials and Light trials. This tests the ability of the light alone to serve as CS in the absence of the natural tone.

Optogenetic activation of ChR2-expressing STN glutamatergic neurons drives crossings in naïve mice, suppresses the development of caution, and is incompatible with passive avoidance.

A, Effects of patterns of blue light applied in the STN of naïve mice where the light does not predict the US (NoCS+Light trials, blue circles). This was followed by the addition of the US in regular Light trials (red triangles). Note that STN excitation in naïve mice drives a high rate of crossings. B, Development of caution between AA1 and AA2 (by punishing ITCs) and reflected in the avoidance latency is suppressed by STN excitation in STN-ChR2 mice compared to No Opsin controls. The blue circles show AA1 sessions followed by red squares showing AA2 sessions in the same mice for STN-ChR2 and No Opsin mice. Note the abolishment of ITCs during AA2 and the shift of the latency longer only in No Opsin mice. Asterisks denote significant differences (p<0.05) between AA1 (blue) and AA2 (red). C, Traces of overall speed during AA1 and AA2 for the data shown in B. Note the characteristic changes in No Opsin mice are blunted by STN excitation in STN-ChR2 mice. This includes the rightward shift of the speed from CS onset, and the larger peak speed from response occurrence during AA2 compared to AA1. D, STN excitation blocks signaled passive avoidance. In AA3, CS1 signals the animal to actively avoid, while CS2 signals the animal to passively avoid by not crossing. In addition, we tested CS2+Light trials and NoCS+Light trials (40-60 Hz light). STN excitation virtually abolished passive avoids to CS2 so that mice crossed despite being punished. Moreover, mice also crossed during NoCS trials when the light was delivered. E, Traces of overall speed during AA3 for the data in D. F, Population measurements of peak speed for the data in E. Asterisks denote significant differences (p<0.05) between CS2 and the other conditions.

Optogenetic STN excitation is not aversive.

A, Signaled active avoidance procedures where the normal aversive US is substituted for STN optogenetic excitation that drives fast escape responses. Mice were subjected to AA1, followed by AA2 and then AA3 using 3 different CS’s. During AA1, CS1 predicts STN stimulation, while CS2 and CS3 predict nil. During AA2, ITCs are punished with STN stimulation. During AA3, mice must passively avoid by not responding when CS2 is presented, so active avoids driven by CS2 are errors. The bars show side by side the results for Opsin (black, red, cyan) and No Opsin mice (gray) in the same procedures. Asterisks denote significant differences (p<0.05) between Opsin vs No Opsin mice. B, The mice in A are subjected to the same procedures using the normal aversive US. Asterisks denote significant differences (p<0.05) between Opsin vs No Opsin mice.

Optogenetic inhibition of STN glutamatergic neurons impairs signaled avoidance.

A, Effect of Cont green light delivered at different powers on AA1 (green circles), AA2 (red circles) and AA3 (right panel) in mice expressing eArch3.0 in STN glutamatergic neurons. Note the strong abolishment of active avoidance responses in CS+Light trials for AA1, AA2, and AA3-CS1. In contrast, passive avoids during AA3-CS2 were not impaired. B, Traces of overall movement (speed) during AA1, AA2 and AA3 for CS trials and CS+Light trials combined for different light powers. The trials are aligned by CS onset, which reveals the orienting response evoked by the CS followed by the ensuing avoid action. C, Population data of peak speed from CS onset for orienting and avoidance responses during AA1, AA2, and AA3. Asterisks denote significant differences (p<0.05) between CS vs CS+Light.

Lesions of STN glutamatergic neurons impair signaled avoidance learning and performance.

A, Coronal Neurotrace (green) stained section of a Vglut2-Cre mouse injected with a Cre-dependent AAV-dtA in the STN to kill glutamatergic neurons. We counted the number of cells in the STN in controls and lesion mice. There was a significant reduction (p<0.05) in the number of STN neurons in the lesion mice. B, Behavioral performance during learning of AA1, followed by AA2 and AA3 procedures showing the percentage of active avoids (upper), avoid latency (middle), and ITCs (lower) for control and lesion mice. The AA3 procedure shows CS1 and CS2 trials for the same sessions. Active avoids during AA3-CS2 trials are errors, as the mice must passively avoid during CS2. Lesion mice were significantly impaired compared to control mice. C, Movement (speed) from CS onset (left) and from avoid occurrence (right) during AA1 and AA2 procedures for control and lesion mice. The lesion had significant effects on movement measured from CS onset or avoid occurrence. D, Same as C for AA3. E, Population measures of orienting, avoidance, and escape responses from CS onset (left) and from response occurrence (right) for overall movement. Asterisks denote significant differences (p<0.05) between Control and Lesion. F, Bilateral electrolytic lesions targeting the STN. G, Effect of bilateral electrolytic STN lesions on behavioral performance in a repeated measures design. The plot shows the percentage of active avoids (filled black circles), avoid latency (open orange squares), and ITCs (cyan bars). Mice were trained in AA1 prior to the lesion and then placed back in AA1, followed by AA2 and AA3. The lesion decreased the percentage of active avoids compared to AA1. During AA2, mice learned to suppress their ITCs. During AA3, lesion mice were impaired in active avoids during CS1 but passively avoided during CS2.

Optogenetic activation of STN glutamatergic pathways to the midbrain, not to GPe, mediate the effects of excitation within STN.

A, Schematic of optical fiber locations for the midbrain targeting midbrain tegmentum (SNr and mRt combined) and GPe groups to target fibers originating in STN. B, Effects of blue light patterns (CS+Light) applied in the Midbrain (filled triangles) or in the GPe (filled red circles) of STN-ChR2 mice on signaled active avoidance. A group of No Opsin mice are also included (filled back squares). Both optogenetic groups activated in the Midbrain and GPe received low and higher power light, but this is shown separately for the midbrain group as there was no effect of the light in the GPe group. Note that the effects in the Midbrain group occur for high frequency trains 20-100 Hz but not Cont light, which is consistent with the effective sustained activation of fibers with pulsed trains, not with Cont light. C, Development of cautious responding from AA1 to AA2 (induced by punishing ITCs) is reflected in increased avoidance latencies and is suppressed by excitation of STN fibers in the midbrain, but not by excitation of STN fibers in the GPe or in No Opsin controls. Note the abolishment of ITCs during AA2 across groups, and the latency shift toward longer response times observed only in GPe and No Opsin mice. D, Signaled passive avoidance is blocked when STN fibers to midbrain are excited but not those to GPe or in No Opsin mice. In AA3, CS1 signals the animal to actively avoid (open black squares), while CS2 signals the animal to passively avoid by not crossing (open red circles). Application of CS2+Light (filled red circles) and NoCS+Light (filled gray triangles) trials (40-60 Hz light) in Midbrain but not in GPe virtually abolished passive avoids to CS2 and induced crossings in response to the light alone. E, Ability of STN fiber excitation in the Midbrain or GPe to serve as a CS assessed by comparing CS and Light trials. Only excitation of STN fibers in the midbrain was sufficient to drive active avoidance responses, whereas stimulation of STN fibers in the GPe was ineffective and comparable to responses in No Opsin controls.

Optogenetic inhibition of STN glutamatergic fibers in Midbrain impairs signaled avoidance.

A, Effect of Cont green light delivered at different powers (Lo or Hi) on AA1, AA2 and AA3 in mice expressing eArch3.0 in STN glutamatergic neurons. Note the strong abolishment of active avoidance responses in CS+Light trials for AA1, AA2, and AA3-CS1. In contrast, passive avoids during AA3-CS2 were not impaired. The light had no effect in No Opsin mice (filled gray squares). B, Traces of overall movement (speed) during AA1, AA2 and AA3 for CS trials and CS+Light trials combined for different light powers. The trials are aligned by CS onset, which reveals the orienting response evoked by the CS followed by the ensuing avoid action. C, Population data of peak speed from CS onset for orienting and avoidance responses during AA1, AA2, and AA3. Asterisks denote significant differences (p<0.05) between CS vs CS+Light.