Abstract
Deep brain stimulation (DBS) effectively treats motor symptoms of advanced Parkinson's disease (PD), with the globus pallidus interna (GPi) commonly targeted. However, its therapeutic mechanisms remain unclear. We employed optogenetic stimulation in the entopeduncular nucleus (EP), the rat homolog of GPi, in a unilateral 6-hydroxydopamine lesioned female Sprague Dawley rat model of PD. We quantified behavioral effects of optogenetic EP DBS on motor symptoms and conducted single-unit recordings in EP and ventral lateral motor thalamus (VL) to examine changes in neural activity. High-frequency optogenetic EP DBS (75, 100, 130 Hz) reduced ipsilateral turning and corrected forelimb stepping, while low-frequency stimulation (5 and 20 Hz) had no effect. EP and VL neurons exhibited mixed response during stimulation, with both increased and decreased firing. The average firing rate of all recorded neurons in the EP and VL significantly increased at 130 Hz but not at other frequencies. Beta-band oscillatory activity was reduced in most EP neurons across high frequencies (75, 100, 130 Hz), while reductions in beta-band oscillations in VL occurred only at 130 Hz. These findings suggest that the neural firing rates within EP and VL circuits were differentially modulated by EP DBS; they may not fully explain the frequency-dependent behavioral effect. Instead, high-frequency optogenetic EP DBS at 130 Hz may ameliorate parkinsonian motor symptoms by reducing abnormal oscillatory activity in the EP–VL circuits. This study underscores the therapeutic potential of circuit-specific modulation in the pallidothalamic pathway using optogenetic EP DBS to alleviate motor deficits in a PD rat model.
Significance Statement
The contribution of entopeduncular nucleus (EP) local cells to the therapeutic effects of EP deep brain stimulation in Parkinson's disease (PD) has been unclear. Our study addressed this by employing local cell-specific optogenetic stimulation. Directly stimulating EP local neurons using optogenetics effectively reduced parkinsonian symptoms in the 6-hydroxydopamine–lesioned rat. These findings highlight the importance of precise circuit manipulation through optogenetic techniques within the pallidothalamic pathway, suggesting a promising approach for ameliorating motor deficits in PD.
Introduction
Deep brain stimulation (DBS) is a therapy for the motor symptoms of advanced Parkinson's disease (PD). Globus pallidus interna (GPi) and subthalamic nucleus (STN) are widely used DBS targets that have demonstrated clinical efficacy (Ghika et al., 1998; Perlmutter and Mink, 2006; Wichmann and Delong, 2006; Kringelbach et al., 2007; Vitek, 2008). Initially, studies suggested that STN DBS produced greater improvements in motor symptoms, including resting tremor and rigidity, medication reductions, and economic benefits compared with GPi DBS (Krack et al., 1998). However, subsequent randomized clinical trials revealed no significant differences in motor outcomes between STN and GPi DBS (Okun et al., 2003; Anderson et al., 2005; Okun and Foote, 2005; Weaver et al., 2012). Yet, reports of cognitive and neuropsychiatric adverse effects, as well as motor symptoms refractory to STN DBS, have prompted increased interest in GPi DBS. GPi DBS may offer advantages such as greater suppression of dyskinesia, increased flexibility in medication adjustment, and fewer cognitive adverse effects (Anderson et al., 2005; Okun and Foote, 2005; Okun et al., 2009; Williams et al., 2014). Despite these well-established benefits, the precise mechanisms underlying the therapeutic effects of GPi remain unclear. The limited understanding impedes the full development and optimization of GPi DBS and hinders rational selection between STN and GPi DBS for individual patients. Furthermore, understanding the mechanism of GPi DBS could contribute to understanding the effects of DBS in other targets and neurological disorders.
Our previous study demonstrated that selective activation of STN cell bodies in a rat model of PD effectively alleviated motor symptoms, with behavioral efficacy strongly correlating with changes in pathological beta oscillatory activity rather than neural firing rates (Yu et al., 2020). While it is known that GPi DBS contributes to the disruption of oscillatory activity and information flow in basal ganglia–thalamo–cortical circuits (Anderson et al., 2003; Liu et al., 2008; Agnesi et al., 2013; Muralidharan et al., 2017), whether similar mechanisms underlie the therapeutic effects of GPi DBS remains a fundamental question. Specifically, the selective modulation of specific neural circuit activity during therapeutic GPi DBS has not yet been thoroughly studied. Recent research indicated that optogenetic inhibition of GPi improved parkinsonian behaviors, while activation using channelrhodopsin-2 (ChR2) did not yield behavioral effects (Moon et al., 2018; Yoon et al., 2020). However, these findings remain uncertain due to the kinetic properties of ChR2, which cannot sustain high-frequency firing—a critical requirement for effective DBS (Boyden et al., 2005; Lin et al., 2009; Yizhar et al., 2011; Hight et al., 2015; Jun and Cardin, 2020). Addressing these gaps could enhance our understanding of GPi DBS, guide treatment decisions, and advance our comprehension of DBS mechanisms across diverse neural targets.
Here we employed optogenetic GPi DBS with Chronos to examine its effects on behavior and neural activity. Chronos is known for its faster kinetics and ability to faithfully follow high-frequency stimulation (Klapoetke et al., 2014; Hight et al., 2015; Jun and Cardin, 2020). Chronos was packaged into an adeno-associated virus serotype 5 (AAV5) and driven by a neuron-specific human synapsin I (hSyn1) promoter (Diester et al., 2011; Mastro et al., 2017). This promotor, active in both in excitatory and inhibitory neurons (Watakabe et al., 2015), generated local cell-specific expression in the GPi [entopeduncular nucleus (EP) in rat] in the 6-hydroxydopamine (6-OHDA)-lesioned rat model of PD. We evaluated the behavioral effects of optogenetic EP DBS on parkinsonian motor symptoms using circling behavior and forelimb adjusting steps. Additionally, single-unit recordings were conducted in the EP and ventral lateral motor thalamus (VL) during optogenetic EP stimulation, allowing for the quantification of its effects on neural activity. Our finding demonstrates that cell-specific high–frequency EP optogenetic DBS effectively alleviated pathological circling and improved abnormal stepping, accompanied by disruption in abnormal oscillatory activity within the EP and its downstream VL neural circuits.
Materials and Methods
All animal care and experimental procedures were approved by the Institutional Animal Care and Use Committee (IACUC, Protocol Number L0280) at Michigan Technological University.
Animals and surgery
Preparation
Sixteen female Sprague Dawley rats (∼10 weeks, weighing 250–300 g) were used for behavioral studies and electrophysiological recordings. Among them, six rats received injections of Chronos and had optical fibers implanted in the EP (AP, −2.3 mm from the bregma; ML, 2.8 mm), while four rats received ChR2 injections and had fibers implanted in the EP. Additionally, six rats were injected with Chronos for acute optogenetic stimulation and single-unit recording. All rats underwent unilateral injection of 6-OHDA into the medial forebrain bundle (MFB; AP, −2.0 mm; ML, 2.0 mm) to induce hemiparkinsonism. The surgery procedures were conducted under general anesthesia using isoflurane (induction at 4% in 4 L/min O2; maintained at 2% with 2 L/min O2). Dexamethasone (5.0 mg/kg) was subcutaneously administered to reduce potential for brain swelling during surgery, while body temperature was maintained at ∼ 37°C with a water heating blanket. Craniotomies were conducted over the unilateral EP and MFB based on a stereotactic atlas of the rat brain as previously described (Yu et al., 2020; Li et al., 2021).
Viral injection, dopamine depletion, and fiber implantation
The viral vectors rAAV5-Syn-Chronos-GFP (5.7 × 1012 vg/ml) and rAAV5-Syn-hChR2 (H134R)-EYFP (5.2 × 1012 vg/ml) were stereotyped and packaged by the University of North Carolina at Chapel Hill Vector Core. The virus (0.6 µl) was injected into the EP (DV −6.8 mm) at a rate of 0.1 µl/min through a 1 µl 32 gauge Hamilton syringe, with a waiting period of 5 min after every 0.2 µl injection and leaving the needle in place for an additional 10 min after the full injection. Subsequently, rats were lesioned to induce unilateral degeneration of dopaminergic neurons in the substantia nigra pars compacta (SNc) to produce hemiparkinsonism. The 6-OHDA (6 µl, 2.5 mg/ml in 0.02% ascorbic acid dissolved with saline, Sigma-Aldrich) was infused into MFB (DV, −7.5 mm) at a rate of 1 µl/min using a 10 µl Hamilton syringe, with a 5 min interval after every 2 µl infusion and leaving the needle in place for an additional 10 min after the full injection. Thirty minutes before lesioning, the rats were administered 5 mg/kg desipramine intraperitoneally (Sigma-Aldrich) to inhibit monoamine oxidase and protect noradrenergic neurons.
For rats with optogenetic stimulation, an optical fiber (200 μm core diameter, 1.25 mm OD ceramic zirconia ferrule, Precision Fiber Products) was inserted into the EP (DV, ∼−6.8 mm from the bregma) and cemented in place with dental acrylic. These rats were then single housed for a minimum of 5 weeks before starting behavioral paradigms to allow time for viral expression and recovery from the lesion and surgery. For electrophysiological recordings with optical stimulation, the surgical openings were sutured after 6-OHDA and virus injection. The rats were given a recovery period of at least 5 weeks before the acute recordings began.
Behavioral tests
In optogenetic experiments, rats underwent ∼5 weeks of viral expression and nigral neuron degeneration before being connected to a 473 nm DPSS laser (Shanghai Laser and Optics Century) via fiber optic cables and placed within the testing chamber. Given Chronos' faster off-kinetics and reliable tracking of high rates, a shorter pulse width of 1 ms was applied to assess behavioral effects. To maintain consistency with previous studies of optogenetic STN DBS using ChR2, 5 ms pulses for ChR2 stimulation was employed (Gradinaru et al., 2009; Yu et al., 2020). Fiber output was measured at 10 mW using a PM20A photodetector (Thorlabs).
Circling test
The methamphetamine-induced circling behavior is a well-established and widely used behavioral test for validating PD models and assessing DBS effects (Hefti et al., 1980; Schwarting and Huston, 1996; Meissner et al., 2002; Gradinaru et al., 2009; Cadet et al., 2010; McConnell et al., 2012; So et al., 2012; McConnell et al., 2016; So et al., 2017; Moon et al., 2018; Bjorklund and Dunnett, 2019; Swan et al., 2019; Yu et al., 2020; Yu et al., 2022; Li et al., 2024). The circling test procedure followed that outlined in the previous study (Yu et al., 2020; Li et al., 2024). Briefly, unilaterally lesioned rats were injected with a single dose of methamphetamine (1.875 mg/kg in 0.9% saline) 30 min before being placed in a cylinder to induce robust and sustained circling. The cylinder was positioned within a dark sound attenuating chamber equipped with an infrared camera to capture the rat’s behavior. To prevent cable twisting, we employed a rotating optical commutator (Doric Lenses).
The behavioral effects of optical DBS were assessed using both fixed 130 Hz stimulation and randomized blocks of five stimulation frequency, 5, 20, 75, 100, and130 Hz, with the presentation order randomized within each block. In fixed 130 Hz optical stimulation, pulses were administered in sets of six trials (each trial consisting of 10 s on and 20 s off) for a total duration of 3 min. Each behavioral recording session began with a 3 min control period, followed by a 3 min stimulation period, and concluded with a 3 min period without light, resulting in a total session duration of 9 min. In frequency randomized stimulation, pulse trains were delivered for 10 s at each frequency followed by a 20 s off interval. Each session began with a minimum of 2 min control period, followed by the stimulation period, and concluded with 2 min without light. Six sessions were recorded for each rat, covering both 130 Hz and the randomized frequencies. The repeated sessions were conducted at least 2 d apart. The circling behavior was captured on video and analyzed. The position of the nose and the base of the tail over time was precisely determined based on image moment calculation using behavioral analysis software (TopScan, Clever Sys). Angular velocity and distance traveled per minute (linear speed) were subsequently computed off-line from the tracking data using MATLAB. To assess the impact of stimulation, we determined the normalized angular velocity or linear speed for each trial by dividing the angular velocity or linear speed during the stimulation-on period by the average angular velocities or linear speeds during the prestimulation-off and poststimulation-off periods immediately before and after the stimulation-on period. Subsequently, normalized angular velocities or linear speed for each stimulation rate were calculated by averaging across all randomized blocks.
Adjusting steps test
Unilateral lesioned rats display distinct deficits in contralateral limb use, with impairments in forelimb adjusting steps serving as a validated measure of parkinsonian akinesia in rats (Schallert et al., 1978; Glajch et al., 2012). The test procedure was induced as described in our previous study (Yu et al., 2020). In brief, each rat was held with their hindlimbs elevated and moved backward at a steady rate traversing a 1 m glass corridor (Runway, Clever Sys) over a duration of 3–4 s. Video recordings were made and later analyzed off-line to determine the number of steps taken with the contralateral and ipsilateral forelimbs. Continuous optical DBS at 20 and 130 Hz DBS was administered, with control trials conducted in sessions without stimulation. Three trials were recorded in each session (n = 6 sessions). The behavioral effects of DBS were quantified by calculating the ratio of the number of steps taken with the contralateral forelimb to the number of steps taken with the ipsilateral forelimb. The index of stepping change was calculated using the following formula: index = (ratio of steps at stimulation − ratio of steps without stimulation) / (ratio of steps at stimulation + ratio of steps without stimulation). This formula allows us to assess the change in stepping behavior in relation to stimulation.
Neural recording
Electrophysiological recordings in the EP and VL were conducted on the ipsilateral side of lesioned rats under urethane anesthesia (1.2 g/kg, s.c.) following a recovery period of ∼5 weeks postviral injection in the EP. A midline incision of the scalp was performed, followed by gentle retraction of the soft tissue and the creation of a craniotomy over the EP (AP, −2.3 mm from the bregma; ML, 2.8 mm) and VL (AP, −2.0 mm from the bregma; ML, 2.0 mm) regions. The dura matter was carefully removed at each craniotomy location, and the brain surface was covered with saline for protection and to maintain hydration. Additionally, a bone screw was implanted at the back of the head to serve as electrical ground during the recordings. A custom-made optrode was used for simultaneous optical stimulation and unit recording in EP. The optrode consisted of an optical fiber (200 µm) attached to single tungsten microelectrode (0.5 MΩ, 75 µm diameter). The tip of the fiber was positioned ∼0.3–0.4 mm above the electrode tip and angled to ensure that the emitted light encompassed the electrode tip (Sparta et al., 2011). For single-unit recordings from the VL during optical stimulation in the EP, 16-channel electrode arrays were used. Each electrode/array was slowly advanced using a micromanipulator, with separate recording depths for each penetration spaced 300–400 μm apart.
Neuronal activity was recorded using a multichannel acquisition processor system (Bio-Signal Technologies). The neural signals were amplified, high-pass filtered at 300 Hz to isolate action potential-like activity, and digitally sampled at 40 kHz. The neuronal effects of optical DBS were assessed using both 130 Hz stimulation and varying frequencies at 5, 20, 75, and 100 Hz. Each stimulation frequency was applied in a train of 10 trials, with each trial consisting of 10 s stimulation followed by 20 s without stimulation. Each recording session began with a baseline recording period lasting at least 60 s, during which no stimulation was presented. The session concluded with another baseline recording period of at least 60 s.
Histology
After completing the experimental tests, rats were deeply anesthetized with urethane (1.8 g/kg, i.p.) and then perfused transcardially with 0.1 M phosphate-buffered saline (PBS), followed by 4% paraformaldehyde in 0.1 M PBS. The brains were postfixed in 4% paraformaldehyde overnight at 4°C and subsequently switched to 30% sucrose solution at 4°C for cryoprotection. Brain sections were cut into 25 or 40 µm sections in the coronal plane using a cryostat (HM525NX, Epredia) and processed for staining. Tyrosine hydroxylase (TH) immunohistochemistry was used to assess the extent of degeneration of dopaminergic neurons in the SNc and to locate electrode/fiber positions. Initially, brain sections were rinsed and blocked for 1 h in 10% goat serum. Subsequently, sections were incubated overnight at 4°C in a solution containing anti-TH antibody (AB152; 1:1,000, Vector Laboratories) diluted in 10% goat serum with 0.25% Triton X-100 with PBS. Afterward, sections were incubated with biotinylated goat anti-rabbit secondary antibody (BA-1000, 1 : 250, Vector Laboratories) in a solution containing 10% goat serum and 0.25% Triton X-100 in PBS for 1 h at room temperature. Following rinsing, sections were incubated in a VECTASTAIN Elite ABC Kit (Vector Laboratories) solution for 1 h and then visualized using 3,3′-diaminobenzidine solution. To visualize GFP positive cells, we directly mounted the neighboring sections with DAPI-Fluoromount-G (SouthernBiotech) after rinsing. For determining the cell-specific expression of Chronos-GFP, brain sections underwent immunostaining for glutamate or glutamic acid decarboxylase (GAD). Slices were incubated overnight at 4°C with primary anti-CaMKII antibody (1:300, Millipore Sigma, catalog #05-532) or primary anti-GAD antibody (1:500, Millipore Sigma, MAB5460) in a solution containing 10% goat serum and 0.25% Triton X-100. Secondary antibodies (Alexa Fluor 594 goat anti-mouse IgGI, A21125, 1:500, Life Technologies) were used to visualize glutamate or GABA. Following staining, all sections were mounted with DAPI-Fluoromount-G (SouthernBiotech). Image acquisition was performed using a Leica DM4 upright microscope. The localization of electrodes, fibers, and injections were compared with a rat brain atlas (Paxinos and Watson, 2005). To assess dopaminergic neuron loss in unilaterally 6-OHDA lesioned rats, we examined three consecutive TH-stained brain slices from each brain. We counted cells in the SNc on both the lesioned and unlesioned sides using the cellSens Standard software (Evident Scientific). We then compared the number of dopaminergic neurons in the lesioned hemisphere with those in the intact hemisphere.
Data analysis
The electrophysiological data were analyzed using custom-written MATLAB scripts in combination with the Plexon Offline Sorter spike sorting software (Plexon). Spikes were detected by applying a high-pass filter at 300 Hz, followed by the extraction of spike waveforms with a duration of 1,100 µs. Single-unit activity was sorted using principal component analysis, ensuring distinct separation from noise. Units with refractory period shorter than 1 ms were excluded from further analysis. Each trial's neural activity, within a window of −10–20 s aligned to stimulus onset, was binned into 200 ms intervals to construct peristimulus time histograms (PSTHs). The PSTHs were then normalized using z-scoring, where the mean baseline firing rate (measured from a baseline window from −10 to 0 s before trial initiation) was subtracted and the result was divided by the standard deviation. To classify units as either “excited” or “inhibited” units, we compared their activity during 10 s stimulation on with a baseline window from 10 s before the stimulus onset using paired t test (p < 0.05). Peak latency was computed representing the time from the stimulus onset to the first peak response within 100 ms. To explore temporal structure, we conducted spectral analysis of spike times using multitaper analysis, leveraging five Slepian data tapers on 10 s windows (Chronux version 2.12; Mitra and Pesaran, 1999). Spectral representation was presented on either a logarithmic scale or a normalized scale, determined by dividing the minimum power within prestimulation-off, stimulation, and poststimulation-off periods. Normalized spectral power within the beta band (12–30 Hz), low beta band (12–20 Hz), and high beta band (20–30 Hz) during the stimulation period were computed and compared with the average spectral power during the prestimulation off and poststimulation off periods using a paired t test (p < 0.05). Power spectrograms were generated within a window of −10–20 s aligned to the stimulus onset using 1 s sliding window and 0.1 s steps.
Statistical analysis
Statistical significance between conditions was determined using a one-way or two-way repeated measures analysis of variance (ANOVA). Data normality was accessed through the frequency distribution (histogram) and boxplot (Ghasemi and Zahediasl, 2012). Post hoc paired comparisons were conducted if the corresponding main effect or interaction reached significance at p < 0.05. To control for multiple comparisons, we performed pairwise post hoc tests with the Tukey's honestly significant different test with a significant cutoff p < 0.05. For single comparisons, either two-sample Mann–Whitney test or two-tailed paired t test was performed. All statistical tests were conducted using MATLAB or Prism (GraphPad). Results are reported as mean ± standard error (SE).
Results
Histological study
The unilaterally injection of 6-OHDA into MFB resulted in a unilateral loss of nigral dopaminergic cells indicated by a substantial decrease in TH immunoreactivity in the SNc of the lesioned hemisphere (Fig. 1A) across all rats (n = 16) participating in behavioral and electrophysiological studies. Coronal sections immunostained for TH confirmed that the unilateral 6-OHDA lesion produced over 90% loss of dopaminergic neurons in the SNc (92.4 ± 0.6%). For the assessment of optogenetic EP DBS behavioral effects, Chronos (n = 6) or ChR2 (n = 4) under the control of Syn promotor was injected, and optical fibers were implanted in the EP. The AAV viral vector injection labeled EP neurons and their projection terminals in the VL (Fig. 1B,C). Chronos-GFP was coexpressed with glutamate (Fig. 1D,F) and GAD67 (Fig. 1E,G) confirming successful and specific targeting of excitatory and inhibitory local neurons in the EP. The locations of implanted optical fiber (n = 10) were confirmed within the EP (Fig. 2A).
Anatomical verification of 6-OHDA lesion and Chronos-GFP expression in the EP and VL. A, Coronal section immunostained for TH confirms a unilateral 6-OHDA lesion, resulting in >90% loss of dopaminergic neurons in the SNc. B, Left, Representative TH-stained representative coronal section indicating the location of optical fiber in the EP. Right, Representative immunofluorescence-stained section showing neurons expressing Chronos-GFP following viral injection in the EP. C, Densely labeled terminals in the VL. D, E, Fluorescence images revealing EP neurons expressing Chronos-GFP and labeled for glutamate and GAD67. Arrows indicate representative Chronos-GPF–positive neurons that were also immunoreactive for glutamate or GAD67.
Modulation of methamphetamine-induced circling in rats with unilateral 6-OHDA lesions during optogenetic EP DBS. A, Fiber locations in the EP across all implanted rats used for circling and forelimb stepping analysis (Chronos, n = 6; ChR2, n = 4) on a coronal section. The diagram is based on a rat brain atlas (Paxinos and Watson, 2005). B, Trials in one representative stimulation session of a rat injected with Chronos show that high-frequency optogenetic stimulation in the EP at 130 Hz reduced angular velocity of methamphetamine-induced circling, while the linear speed of movement remained unchanged. C, Ipsilateral circling was markedly reduced in Chronos-injected rats indicated by a decrease in angular velocity of circling (p < 0.001; one-way repeated–measure ANOVA; n = 6). There was no notable reduction in ipsilateral circling in ChR2-injected rats (right, p = 0.792; n = 4) during 130 Hz stimulation. D, The mean changes of angular velocity during optical stimulation with Chronos and ChR2. The reduction of pathological circling during optogenetic activation with Chronos was greater than with ChR2 (p < 0.0001; Mann–Whitney test). E, Chronos-injected rats exhibited varied responses in movement speed (increased, decreased, or no change) during optogenetic stimulation (p = 0.911), while ChR2-injected rats showed no significant change (p = 0.933). F, The mean changes of linear speed during optogenetic stimulation. There was no significant difference in the change of linear speed between the two optical stimulation conditions (p = 0.129; Mann–Whitney test). ***p < 0.001. Data are presented as mean ± SE.
Improvement in pathological circling
Quantification of the effects of optogenetic EP DBS on methamphetamine-induced circling behavior revealed that EP DBS at 130 Hz in Chronos-injected animals effectively reduced pathological circling, indicated by a decrease in angular velocity during stimulation compared with the periods before and after light stimulation (Fig. 2B,C, left; F(2,641) = 25.15; p < 0.001; one-way repeated–measure ANOVA). Conversely, optogenetic EP DBS at 130 Hz in ChR2-injected animals did not alter pathological circling (Fig. 2C, right; F(2,413) = 0.23; p = 0.792) consistent with previous studies (Gradinaru et al., 2009; Yu et al., 2020). This is further confirmed by the significant reduction in pathological circling during optogenetic EP DBS using Chronos, which was markedly greater than that observed with optogenetic DBS using ChR2 (Fig. 2D; p < 0.0001; Mann–Whitney test). Notably, the reduction in circling with Chronos in the EP (median, 36.1) was less pronounced than the reduction achieved with optical STN DBS at 130 Hz (median, 55.1%; Yu et al., 2020; p = 0.001; Mann–Whitney test).
Additionally, both optogenetic stimulation in Chronos-injected and in ChR2-injected rats showed no changes in linear speed (Fig. 2E; Chronos, F(2,641) = 0.09; p = 0.911; ChR2, F(2,413) = 0.07; p = 0.933), with no significant differences between the two conditions (Fig. 2F; p = 0.129; Mann–Whitney test). Interestingly, there was also no difference in the percentage change in linear speed during optogenetic EP DBS (median, 6.06%) compared with STN DBS (median, 4.75%; Yu et al., 2020; p = 0.778; Mann–Whitney test). These results suggested that the decrease in angular velocity during DBS was not associated with a reduction in overall movement or freezing during stimulation consistent with previous findings in the STN DBS (Yu et al., 2020).
To assess the impact of different frequencies of optogenetic EP DBS on behavior, we employed five distinct stimulation frequencies. Optical DBS with Chronos demonstrated a reduction in ipsilateral circling at higher stimulation frequencies, evidenced by a decrease in angular velocity compared with the off conditions (Fig. 3A,B). A two-way ANOVA revealed significant main effects of the stimulation type (Chronos vs ChR2; F(1,1104) = 4.02; p = 0.045), no effects of stimulation frequency (F(4,1104) = 0.70; p = 0.591), and a significant interaction between these factors (F(4,1104) = 9.04; p < 0.001). Post hoc paired comparisons revealed that optical EP DBS in Chronos-injected rats at high frequencies (75, 100, 130 Hz; all p < 0.05) alleviated ipsilateral turning, while the low rate (5 and 20 Hz) DBS was not effective in reduction. No frequency-dependent effects were observed for optogenetic DBS in ChR2-expressing animals (all p > 0.05; Fig. 3B). Furthermore, the stimulation frequency showed no significant main effect on linear speed (Fig. 3C,D; two-way ANOVA; F(4,1104) = 0.22; p = 0.928). There was no difference between the two stimulation types (stimulation type, F(2,1104) = 1.21; p = 0.272), However, a significant interaction emerged between stimulation frequency and stimulation type (F(4,1104) = 3.11; p = 0.015), with high-frequency optogenetic DBS-induced varied changes in linear speed among Chronos-injected rats.
Effects of EP DBS frequencies on circling in the unilateral 6-OHDA–lesioned rats. A, Effects of DBS frequency on angular velocity of circling in a representative stimulation session. Top, Behavioral effect of optogenetic EP DBS in a Chronos-injected rat was dependent on stimulation frequency. EP DBS at 100 Hz (p = 0.039) and 130 Hz (p = 0.021) reduced ipsilateral turning, while low frequency at 5, 20, and 75 Hz had no effect (all p < 0.05). Bottom, optogenetic DBS in ChR2-injected rat had no discernible behavioral effect at high or low frequencies (all p < 0.05). B, Effects of DBS frequency on linear speed of movement in representative stimulation sessions. EP DBS did not change the linear speed of circling in both during Chronos-injected (top) and ChR2-injected (bottom) rats (all p > 0.05). C, Effects of DBS frequency on angular velocity of circling across all Chronos (n = 6) and ChR2 (n = 4) stimulated rats. Ipsilateral circling was reduced in Chronos injected at high-frequency stimulation (75, 100, 130 Hz; all p < 0.05, two-way ANOVA and post hoc comparison). Circling was not affected in ChR2-injected rats (all p > 0.05). D, Effects of DBS frequency on linear speed of movement across all Chronos (n = 6) and ChR2 (n = 4) stimulated rats. The stimulation frequency had no significant main effect on linear speed (two-way ANOVA; p = 0.928). There was no difference between the two stimulation types (stimulation type, p = 0.272), However, a significant interaction emerged between stimulation frequency and stimulation type (p = 0.015), with high-frequency optogenetic DBS-induced varied changes in linear speed among Chronos-injected rats. *p < 0.05. Data are presented as mean ± SE.
Improvement in forelimb adjusting stepping
We evaluated the effects of optogenetic EP DBS on forelimb akinesia using the adjusting step test. Optogenetic DBS at 130 Hz ameliorated the bias toward using the unimpaired forepaw in rats with Chronos expression, reflected by an increase in the ratio of contralateral to ipsilateral forelimb use. However, there were no observed changes in impaired forelimb stepping during optogenetic DBS at 20 Hz (Fig. 4A; F(2,284) = 56.79; p < 0.001; one-way ANOVA). Conversely, optogenetic DBS in ChR2-injected animals did not produce any improvement in the use of the impaired forepaw at either 130 or 20 Hz (Fig. 4B; F(2,215) = 0.27; p = 0.764). Furthermore, these findings were further supported by a significant increase in stepping during optogenetic EP DBS at 130 Hz using Chronos, which was notably greater than the stepping observed at 20 Hz (Fig. 4C; p < 0.0001; paired t test). No significant changes were seen at either frequency in ChR2-injected rats (p = 0.939). Notably, the increase in stepping at 130 Hz in the EP (median, 0.47) was less pronounced than the increase achieved with optical STN DBS at the same frequency (median, 0.57; Yu et al., 2020; p = 0.001; Mann–Whitney test).
Changes of forelimb adjusting steps in the unilateral 6-OHDA–lesioned rats during optogenetic EP DBS. Left, High-frequency optogenetic EP DBS at 130 Hz in Chronos-injected rats resulted in an increased ratio of contralateral-to-ipsilateral limb use, indicating an increment of impaired forelimb use. Low-frequency optogenetic EP DBS at 20 Hz showed no effect (p < 0.001; one-way repeated–measure ANOVA; n = 6). Right, Optical EP DBS in rats injected with the ChR2 did not produce significant behavioral effects (p = 0.764; n = 4). ***p < 0.001. Data are presented as mean ± SE.
Modulation of neural firing rates in EP and VL during 130 Hz EP stimulation
We recorded single-unit neural activity from EP (n = 71 units) and VL (n = 86) in six rats and confirmed the locations of recording electrode by postmortem histology (Figs. 5A, 6A). Optogenetic DBS at 130 Hz induced varied responses in EP neural activity (Fig. 5). Approximately 49% of EP single units showed an increased firing rate during light stimulation (“excited”; n = 35/71), rapidly escalating their firing rate at the stimulus onset and stabilizing at a higher rate throughout the stimulation. In contrast, 34% of EP neurons displayed a decrease in the firing rate (“inhibited”; n = 24/71), reducing their firing rates and stabilizing at lower rate for the duration of the stimulation (Fig. 5C,D). Of the EP neurons, 17% of neurons (n = 12) did not demonstrate significant changes in the firing rate during light stimulation. In comparison, VL neurons increased and decreased their firing rate during 130 Hz optogenetic EP DBS. Among VL neurons, 48% (n = 41/86) increased their firing, while 28% (n = 24/86) decreased firing in response to the light stimulation (Fig. 6). Interestingly, we observed a significant increase in the population average firing rate of all recorded neurons in the EP (n = 71; 23.49 ± 2.79) compared with the baseline firing rate of 19.75 ± 2.04 (paired t test; p = 0.037). Similarly, we found a significant increase in the average firing rate of all recorded neurons in the VL (n = 86; 21.38 ± 2.21) compared with the baseline firing rate (16.81 ± 1.75; p = 0.002).
Neural activity changes in the EP during 130 Hz optogenetic EP DBS. A, Representative TH-stained coronal sections illustrating the fiber/electrode tracks within the EP. B, Top, action potential waveforms of two representative single units. Middle, Raster plots display the activity of two representative neurons in response to 130 Hz optical stimulation in a Chronos-injected rat. Each row in the raster plot represents a single trial. Bottom, Poststimulus histograms (PSTHs) of representative neurons depict selective increases or decreases in firing rates upon activation of the light stimulus. C, Heatmap illustrating the neural firing rate sorted by responses to light stimulus, classified into excited, inhibited, and nonresponsive neurons. The X-axis indicates time from the stimulus onset, with each row representing the firing rate normalized by the maximum firing rate of each neuron D. Left, Z-score normalized population firing rate of excited, inhibited, and nonresponsive neurons. Right, The mean firing rate of all recorded neurons (excited, inhibited, and nonresponsive neurons) in the EP before and during stimulation (paired t test; p = 0.037). n = 6 rats. *p < 0.05. Data are presented as mean ± SE.
Neural activity changes in the VL during 130 Hz optogenetic EP DBS. A, Representative TH-stained coronal section shows the recording electrode tracks within the VL. B, Top, Action potential waveform of two representative excited and inhibited VL neurons. Middle, Raster plots illustrate the activity of representative neurons in response to EP optical stimulation in a Chronos-injected rat. Bottom, PSTHs of representative neurons represent their response dynamics. C, Heatmap depicting the neural firing rate sorted by responses to light stimulus for excited, inhibited, and nonresponsive neurons. The X-axis indicates time from the stimulus onset, with each row representing the firing rate normalized by the maximum firing rate of each neuron. D, Left, Z-score normalized population firing rate of excited, inhibited, and nonresponsive VL neurons. Right, The mean firing rate of all recorded neurons in the VL before and during stimulation (paired t test; p = 0.002). n = 6 rats. **p < 0.01. Data are presented as mean ± SE.
To further understand the information process between EP and VL, we quantified response latencies of units in EP and VL measured from the stimulus onset to the response peak for excited units and to the response trough for inhibited neurons (Fig. 7). Excited units in the EP responded rapidly to light stimulation with a mean latency of 19 ms (SE, ±2 ms; Fig. 7A). On the other hand, inhibited units decreased their firing rates at longer latencies of 31 ms (SE, ±6 ms; Fig. 7A; p = 0.059; Mann–Whitney test). This difference in latencies suggested that inhibitory responses in the EP may arise either from the activation of inhibitory neurons or the recruitment of inhibitory neurons downstream of light-activated excitatory neurons. Both excited and inhibited neurons in the VL exhibited longer mean latencies than EP excited neurons (excited, 27 ± 2 ms; p = 0.048; inhibited, 41 ± 8 ms; p = 0.16; Fig. 7B). The longer latencies of inhibited neurons in the VL suggested that the suppression of neural responses in the VL may result from network mechanisms involving local inhibitory recruitment and external inhibition from another input, such as substantia nigra pars reticulata (SNr).
Characteristics of EP and VL neural responses during optogenetic EP DBS at 130 Hz. A, Left, raster plots of representative excited and inhibited GPi neurons. Right, Peak response latencies of excited EP neurons and trough latencies of inhibited EP neurons measured from the PSTHs within 100 ms after the stimulus onset. B, Left, raster plots of representative excited and inhibited VL neurons. Right, Peak response latencies of VL neurons measured from the PSTHs within 100 ms after the stimulus onset. C, Correlations between changes in the firing rate and baseline firing rate in inhibited (left, n = 24) and excited (right, n = 35) EP neurons. D, Correlation between change in the firing rate and baseline firing rate in inhibited (n = 24) and excited (n = 41) VL neurons. n = 6 rats.
To understand further the suppression of neural activity in the EP and VL, we examined the baseline firing rates of all recording neurons and conducted a correlation analysis to examine the influence of the baseline firing rate on response during optogenetic DBS (Fig. 7C,D). There was no significant difference of baseline activity among three type of neurons in both EP (F(2,68) = 0.312; p = 0.733) and VL (F(2,83) = 1.00; p = 0.267). There was a significant negative correlation between the reduction of the firing rate and the baseline firing rate in EP inhibited neurons (r = −0.75; p < 0.0001; Pearson's linear correlation coefficient). A significant correlation was also found in EP excited neurons (r = 0.39; p = 0.020). Similarly, the reductions in the firing rate of inhibited neurons in VL were also inversely correlated with their baseline firing rates (r = −0.64; p = 0.0008). However, the increase in the firing rate of excited neurons in VL was not significantly dependent on their baseline firing rates (r = 0.23; p = 0.150).
Modulation of neural oscillatory activity in EP and VL during 130 Hz EP stimulation
Changes in neural oscillatory activity within basal ganglia circuits are closely linked with the parkinsonian state and can be modulated by electrical DBS. We employed spectral analysis of single-unit firing times to evaluate the effects of optogenetic DBS on oscillatory activity. Prior to stimulation, the power spectra of EP neurons exhibited a prominent peak in the beta band (12–30 Hz). This beta oscillatory activity was abolished during optical stimulation at 130 Hz but returned to baseline levels upon cessation of stimulation (Fig. 8A,C). Specifically, 48/71 EP neurons (68%) demonstrated reduced beta band oscillatory activity during 130 Hz optogenetic DBS (t(47) = 3.21; p = 0.002; paired t test). Similarly, in the VL, oscillatory activity decreased in 50/86 (58%) recorded neurons during optogenetic EP DBS and recovered after DBS cessation (Fig. 8B,D; t(49) = 4.68; p < 0.001; paired t test). Furthermore, we found that beta oscillatory activity significantly decreased across all recorded EP neurons (n = 71) during 130 Hz optical EP DBS (t(70) = 2.69; p = 0.009). Likewise, beta oscillatory activity was diminished in the VL population (n = 86; t(85) = 3.20; p = 0.002). Beta oscillatory activity can be further subdivided into low- (13–20 Hz) and high (21–30 Hz)-frequency components, with differential changes potentially impacting parkinsonian symptoms (Connolly et al., 2015). We observed a reduction in lower beta power in EP neurons during optogenetic EP DBS at 130 Hz (two-tailed t test; t(70) = 3.80; p < 0.001), while no changes were detected in high beta band (t(70) = 1.97; p = 0.053; Fig. 9A). Both low and high beta activity in VL were significantly suppressed during optical EP DBS (low beta, t(85) = 2.61; p = 0.011; high beta, t(85) = 2.93; p = 0.004; Fig. 9B). To further investigate the relationship between changes in the neural firing rate and oscillatory activity during EP DBS, we conducted two-way ANOVA of neuronal response type based on changes in firing rates (excited, inhibited, and nonresponsive) and changes in oscillatory power (prestimulation vs stimulation). However, we found no significant interactions between firing rate response and changes in oscillatory power in the EP (F(2,141) = 0.748; p = 0.478) and VL (F(2,171) = 0.74; p = 0.479) during optogenetic EP DBS.
Changes of neural oscillatory activity in the EP and VL during 130 Hz optogenetic STN DBS. A, Left, Representative time–frequency spectrogram shows effect of optogenetic EP DBS on oscillatory activity in a representative EP neuron. Right, Normalized power spectra of the representative single-unit spike times during prestimulation (PreStim), stimulation (Stim), and poststimulation (PostStim) periods. B, Left, Representative time–frequency spectrogram shows effect of optogenetic EP DBS on oscillatory activity in a representative VL neuron. Right, Normalized power spectra of the representative single-unit spike times during prestimulation, stimulation, and poststimulation periods. C, D, Population power spectra of EP (C) and VL (D) neurons. Optogenetic EP DBS at 130 Hz reduced the beta band oscillatory activity in the EP (p < 0.001, paired t test) and VL (p = 0.009). n = 6 rats. **p < 0.01; ***p < 0.001; data are presented as mean ± SE.
Changes in high and low beta band power of neurons in the EP and VL during EP optogenetic stimulation. A, Changes of beta band power of EP neurons at 13–20 Hz (low) and 21–30 Hz (high). There were significant changes in low beta band in EP neurons (paired t test, p < 0.001), but not in high beta band (p = 0.053). B, Changes of low and high beta band power in VL neurons. There were significant changes in both low beta (p = 0.011) and high beta (p = 0.043) bands in VL neurons. n = 6 rats. *p < 0.05; ***p < 0.001. Data are presented as mean ± SE.
Modulation of neuronal activity during different DBS frequency
To understand the neural effect of different frequencies in optogenetic EP DBS beyond 130 Hz, we recorded single-unit neural activity from EP (n = 29) and VL (n = 22) in five rats during stimulation at frequencies of 5, 20, 75, and 100 Hz. Population data analysis revealed a reduction in mean firing rates of EP neurons during 20 Hz stimulation (Fig. 10A; two-tailed paired t test; t(28) = 2.087; p = 0.046). However, there are no significant differences in mean firing rates between the baseline and stimulation in the EP at the other frequencies (5 Hz, t(20) = 1.545; p = 0.138; 75 Hz, t(28) = 0.900; p = 0.376; 100 Hz, t(28) = 0.844; p = 0.406). In contrast, we found no significant difference in mean firing rates between the baseline and stimulation in the VL at the tested frequencies (Fig. 10B; 5 Hz, t(21) = 1.644; p = 0.115; 20 Hz, t(21) = 1.324; p = 0.200; 75 Hz, t(21) = 1.073; p = 0.295; 100 Hz, t(21) = 0.700; p = 0.492). Furthermore, we computed the spectral power of single-unit firing to quantify changes in neural oscillatory activity across different frequencies. We observed a significant decrease in beta oscillatory activity among recorded EP neurons during 75 and 100 Hz optical EP DBS (Fig. 11A; 75 Hz, t(25) = 2.596; p = 0.015; 100 Hz, t(25) = 2.883; p = 0.008), but no at 5 and 20 Hz (5 Hz, t(19) = 0.042; p = 0.967; 20 Hz, t(25) = 1.356; p = 0.187). Conversely, beta oscillatory activity in the VL was not significantly affected by EP DBS at the tested frequencies (Fig. 11B; 5 Hz, t(21) = 1.764; p = 0.093; 20 Hz, t(21) = 0.627; p = 0.538;75 Hz, t(21) = 1.073; p = 0.295; 100 Hz, t(21) = 1.382; p = 0.182).
Changes in the neural firing rate during different frequencies of optogenetic EP DBS. A, Left, z-score normalized population firing rate of recorded EP neurons (n = 29) at stimulation frequencies of 5, 20, 75, and 100 Hz. Right, The mean firing rate of recorded EP neurons before and during stimulation (paired t test, 5 Hz, p = 0.138; 20 Hz, p = 0.046; 75 Hz, p = 0.376; 100 Hz, p = 0.406). B, Left, z-score normalized population firing rate of VL neurons (n = 22) at tested stimulation frequencies. Right, The mean firing rate of all recorded VL neurons before and during EP stimulation (paired t test, 5 Hz, p = 0.115; 20 Hz, p = 0.200; 75 Hz, p = 0.295; 100 Hz, p = 0.492). n = 5 rats. *p < 0.05. Data are presented as mean ± SE.
Changes in neural oscillatory activity during different frequencies of optogenetic EP DBS. A, Left, Population power spectra of EP oscillatory unit spike times during prestimulation (Pre) and stimulation (Stim) periods for each stimulation frequency. Right, Mean normalized beta band (13–30 Hz) power of EP neurons before and during stimulation. Optogenetic stimulation at 75 and 100 Hz suppressed beta band oscillations in the EP (75 Hz, p = 0.015; 100 Hz, p = 0.008), whereas no changes were observed at lower frequencies of 5 and 20 Hz (5 Hz, p = 0.967; 20 Hz, p = 0.187). B, Left, Population power spectra of VL oscillatory unit spike times during prestimulation (Pre) and stimulation (Stim) periods for each stimulation frequency. Right, Mean normalized beta band (13–30 Hz) power of VL neurons before and during stimulation. Optogenetic stimulation 5, 20, 75, and 100 Hz did not significantly affect beta band oscillations in the VL (5 Hz, p = 0.093; 20 Hz, p = 0.538;75 Hz, p = 0.295; 100 Hz, p = 0.182). n = 5 rats. *p < 0.05; **p < 0.01. Data are presented as mean ± SE.
Discussion
We assessed the effects of optogenetic EP DBS on motor behavior and neuronal activity in a unilateral 6-OHDA–lesioned rat model of PD using the ultrafast opsin Chronos (Hight et al., 2015). High-frequency EP DBS (75, 100, 130 Hz) improved circling bias and impaired forelimb use, while low-frequency stimulation (5 and 20 Hz) had no effect. Selective EP neuron activation caused mixed changes in EP and VL neural firing, with a significant firing rate increase observed only at 130 Hz. EP DBS suppressed beta oscillations in the EP (75, 100, 130 Hz) and VL (130 Hz). These findings suggest that reducing pathological oscillations in the VL at 130 Hz is critical, but may not fully explain the therapeutic effects. The inputs from M1 and the reticular thalamic nucleus in VL, along with GPi downstream circuits like the EP-pedunculopontine tegmental nucleus pathway, likely contribute to behavioral improvements. Furthermore, the larger size of the EP compared with the STN may necessitate higher frequencies for effective VL engagement. These factors may explain the lack of changes in VL neural activity at lower stimulation frequencies, such as 75 and 100 Hz. Notably, changes in spectral power and firing rates at 130 Hz were uncorrelated, suggesting that EP DBS primarily exerts therapeutic effects by modulating pathological beta oscillations rather than altering firing rates directly.
Studies have highlighted the efficacy of STN optogenetic DBS in alleviating PD motor symptoms (Yu et al., 2020; Li et al., 2024). Here we quantitatively compare the behavioral effects of EP DBS with those of STN DBS (Yu et al., 2020). Our results indicated EP DBS was less effective than STN DBS in improving deficits. One notable difference between the EP and STN is their size, with the EP being significantly larger. Consequently, the light density required for effective stimulation in the EP may contribute to the reduced therapeutic effect observed with EP stimulation. Methamphetamine induces a strong, unilateral motor response in PD animals, leading them to circle predominantly in one direction and mimicking the asymmetric motor deficits observed in PD patients (Djaldetti et al., 2006; Miller-Patterson et al., 2018; Voruz et al., 2023; Seuthe et al., 2024). We recognize that methamphetamine treatment can alter motor behavior and neuronal activity, complicating the interpretation of optogenetic DBS effects. To address this, we included a second behavior test—forelimb stepping—to evaluate outcomes in the absence of methamphetamine. This approach clarifies how DBS modulates symptoms under different conditions and helps isolate the effects of the simulation. Additionally, neural recording conducted in naive anesthetized lesioned rats eliminated methamphetamine effects on neural activity, elucidating the mechanisms underlying EP optical stimulation. Previous studies on optogenetic EP manipulation showed that low-frequency optical inhibition using halorhodopsin (NpHR) improved parkinsonian symptoms, while activation using ChR2 did not produce significant changes (Moon et al., 2018; Yoon et al., 2020). However, these conclusions remain uncertain due to the kinetic properties of the opsins. ChR2 cannot sustain high-frequency firing, limiting its effects, while the long recovery times of NpHR hinder repeated suppression, crucial for multiple trials (Gradinaru et al., 2009; Chow et al., 2010). Our study applied Chronos, capable of following high stimulation frequency, confirming the therapeutic effects of optical EP DBS. Further validation with faster inhibitory opsins like StGtACR2, known for its high sensitivity for blue light (Wiegert et al., 2017; Mahn et al., 2018; Yamanashi et al., 2019), should be explored in future studies. A limitation of this study is the absence of a GFP-only control group to rule out confounds like visual stimulation and laser-induced heating. Instead, ChR2 was used as prior studies showing it did not improve motor behavior, and the Chronos-specific effects suggest these confounds are unlikely (Gradinaru et al., 2009; Moon et al., 2018; Yu et al., 2020; Li et al., 2024). We acknowledge that the frequency ranges used differ from those in clinic DBS, limiting direct clinical relevance. While our results are informative, they are primarily relevant to preclinical models, and further research is necessary to translate these findings to human applications.
DBS modulates neural firing rates in the basal ganglia circuits associated with the parkinsonian state (Bar-Gad et al., 2004; Erez et al., 2009; Chiken and Nambu, 2016). Optogenetic STN DBS primarily targets hyperactive STN glutamatergic neurons, modulating the indirect pathway and restoring balance within the basal ganglia–thalamocortical circuitry. STN optical stimulation results in both activation and inhibition in the STN and its downstream SNr and external segment of the globus pallidus (GPe), with most SNr neurons inhibited and most GPe neurons activated. Conversely, GPi optical DBS at 130 Hz directly influences GPi neuron activity, potentially reducing excessive inhibition of thalamocortical projections and normalizing motor circuit function. GPi DBS may exert broader modulatory effects on both direct and indirect pathways, offering a more comprehensive approach to circuit modulation (Obeso et al., 2008).
Our study observed increased firing rates in the GPi and VL during 130 Hz GPi optical stimulation, despite GPi typically acting as an inhibitory influence on its targets, including VL. In a normal physiological context, elevated GPi activity would inhibit downstream targets, reducing firing rates. Several factors could explain this increase. First, optical stimulation may induce nonlinear responses in neural circuits; high-frequency stimulation could paradoxically excite GPi neurons or disinhibit downstream targets (Jensen and Durand, 2009). Second, feedback loops and compensatory mechanisms within the basal ganglia–thalamo–cortical circuit could amplify the VL firing rate in response to increased GPi activity. For instance, the GPi and SNr are interconnected via recurrent collaterals, forming a feedback loop within the basal ganglia circuitry (Galvan et al., 2015). This feedback loop coordinates activity between these two nuclei, contributing to the regulation of motor function. Additionally, prolonged optical stimulation may induce plasticity within the GPi→VL pathway, altering the balance of excitation and inhibition and leading to increased firing rates in both regions. While the GPi primarily projects inhibitory signals, it also has excitatory connections, such as glutamatergic excitatory projections to the thalamus (Kha et al., 2000; Barroso-Chinea et al., 2008). Thus, optical stimulation may activate these pathways, leading to increased firing rates in both regions. In summary, while the observed increase in firing rates in both the GPi and VL may appear unexpected based on traditional models of basal ganglia function, it underscores the complexity and dynamic nature of neural circuits.
In PD, abnormal oscillatory activity, particularly in beta (13–30 Hz) frequencies, is evident in the basal ganglia–thalamo–cortical circuits (Hammond et al., 2007). This pathological oscillatory activity is implicated in contributing to motor symptoms and fluctuations observed in PD patients, with increased beta oscillations linked to bradykinesia and rigidity. DBS has been demonstrated to modulate these oscillatory activities, providing relief from motor symptoms in PD patients (Raz et al., 2000; Goldberg et al., 2004; Hammond et al., 2007; Little and Brown, 2014). Consistent with these findings, our study revealed that optogenetic activation of the GPi at 130 Hz successfully suppressed pathological beta oscillatory activity in both the GPi and VL. Additionally, modifications in oscillatory activity were observed independently of changes in the firing rate, supporting the hypothesis that the therapeutic efficacy of DBS arises from its capacity to suppress pathological oscillatory activity. These findings align with previous research on optogenetic STN DBS, indicating that specific stimulation of GPi local neurons was adequate for manipulating neural firing patterns within basal ganglia circuits and ameliorating motor symptoms in PD rats.
Differential changes in the low (13–20 Hz) and high (21–30 Hz) beta bands may produce varied functional effects on parkinsonian symptoms (Priori et al., 2004; Toledo et al., 2014). These subfrequencies of beta activity shift with dopamine depletion or parkinsonian severity (Connolly et al., 2015), with low beta activity exhibiting greater responsiveness to dopaminergic treatment compared with the high beta component (Priori et al., 2004; Lopez-Azcarate et al., 2010). We examined the effects of GPi optogenetic DBS on both low and high beta oscillatory activity. In contrast to our previous findings with 130 Hz optical STN DBS (Yu et al., 2020), where high beta activity of STN neurons was significantly reduced while low beta activity remained unchanged, we found that low beta activity of GPi neurons was dramatically reduced, while high beta activity remained unchanged. Conversely, both low and high beta activity of VL neurons was dramatically reduced, similar to the effects observed in STN downstream SNr neurons.
Considering the distinct effects of GPI DBS and STN DBS on neural firing rates and beta oscillatory activity, these results may explain their differing impacts on motor symptoms of PD. GPi DBS provides greater suppression of dyskinesia and may also improve bradykinesia and rigidity, making it effective for managing dyskinesia and levodopa-induced complications. Conversely, STN DBS typically improves rigidity, tremor, and bradykinesia, with notable efficacy in tremor-dominant symptoms (Williams et al., 2014). Low and high beta oscillations in PD represent distinct frequency components with varying associations with motor symptoms, treatment responsiveness, and potential as therapeutic targets. Understanding their specific roles in PD pathology can guide the development of more targeted and effective treatments for managing motor symptoms in PD patients.
Data Availability
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Footnotes
This work was supported by the National Institute of Neurological Disorders and Stroke under Award Number R15NS115032-01A1 and R15NS133859 to Dr. Chunxiu Yu.
The authors declare no competing financial interests.
- Correspondence should be addressed to Chunxiuy Yu at chunxiuy{at}mtu.edu.