Short- and long-term dopamine depletion causes enhanced beta oscillations in the cortico-basal ganglia loop of parkinsonian rats

Experimental Neurology 286 (2016) 124–136

Contents lists available at ScienceDirect

Experimental Neurology journal homepage: www.elsevier.com/locate/yexnr

Research Paper

Short- and long-term dopamine depletion causes enhanced beta oscillations in the cortico-basal ganglia loop of parkinsonian rats Maximilian H. Beck, Jens K. Haumesser, Johanna Kühn, Jennifer Altschüler, Andrea A. Kühn, Christoph van Riesen ⁎ Charité University Medicine Berlin, Department of Neurology, Movement Disorder and Neuromodulation Unit, Berlin, Germany

a r t i c l e

i n f o

Article history: Received 17 June 2016 Received in revised form 3 October 2016 Accepted 10 October 2016 Available online 12 October 2016 Keywords: Parkinson's disease 6-OHDA Reserpine Beta oscillations Gamma oscillations Cortico-basal ganglia loop

a b s t r a c t Abnormally enhanced beta oscillations have been found in deep brain recordings from human Parkinson's disease (PD) patients and in animal models of PD. Recent correlative evidence suggests that beta oscillations are related to disease-specific symptoms such as akinesia and rigidity. However, this hypothesis has also been repeatedly questioned by studies showing no changes in beta power in animal models using an acute pharmacologic dopamine blockade. To further investigate the temporal dynamics of exaggerated beta synchrony in PD, we investigated the reserpine model, which is characterized by an acute and stable disruption of dopamine transmission, and compared it to the chronic progressive 6-hydroxydopamine (6-OHDA) model. Using simultaneous electrophysiological recordings in urethane anesthetized rats from the primary motor cortex, the subthalamic nucleus and the reticulate part of the substantia, we found evidence for enhanced beta oscillations in the basal ganglia of both animal models during the activated network state. In contrast to 6-OHDA, reserpine treated animals showed no involvement of primary motor cortex. Notably, beta coherence levels between primary motor cortex and basal ganglia nuclei were elevated in both models. Although both models exhibited elevated beta power and coherence levels, they differed substantially in respect to their mean peak frequency: while the 6OHDA peak was located in the low beta range (17 Hz), the reserpine peak was centered at higher beta frequencies (27 Hz). Our results further support the hypothesis of an important pathophysiological relation between enhanced beta activity and akinesia in parkinsonism. © 2016 Published by Elsevier Inc.

1. Introduction There is a growing amount of evidence indicating that pathologically enhanced beta oscillations (13–35 Hz) in the cortico-basal ganglia loop play a paramount role in the system level pathophysiology of Parkinson's disease (PD) and other parkinsonian disorders (Brown, 2007; Oswal et al., 2013; Stein and Bar-Gad, 2013). For this reason, beta oscillations are currently investigated as a potential biomarker for closed loop deep brain stimulation treatment of PD patients (Little et al., 2013a; Priori et al., 2013; Little et al., 2015). Enhanced oscillatory synchrony in the beta frequency band has been found in recordings of local field potentials and spiking neuron activity of the motor cortex and basal ganglia of humans and in different animal models of Parkinson's disease (Brown et al., 2001; Levy et al., 2002; Kuhn et al., 2004; Kuhn et al., 2005; Sharott et al., 2005; Leblois et al., 2007; Shimamoto et al., 2013). It has been hypothesized that beta oscillations are directly linked to disease symptoms (Eusebio and Brown, ⁎ Corresponding author at: Charité University Medicine Berlin, Chariteplatz 1, 10117 Berlin, Germany. E-mail address: [email protected] (C. van Riesen).

http://dx.doi.org/10.1016/j.expneurol.2016.10.005 0014-4886/© 2016 Published by Elsevier Inc.

2009; Timmermann and Fink, 2011). A key element underlying this theory is based on the fact that effective symptomatic treatments with levodopa and deep brain stimulation both reduce abnormal beta power (Doyle et al., 2005a; Kuhn et al., 2006a; Kuhn et al., 2009; Eusebio et al., 2011). In addition, this reduction of beta correlates with the alleviation of the parkinsonian motor symptoms akinesia and rigidity (Kuhn et al., 2006a; Weinberger et al., 2006; Ray et al., 2008; Kuhn et al., 2009). Furthermore, deep brain stimulation of the subthalamic nucleus (STN) at beta frequencies worsens motor symptoms in PD patients (Eusebio et al., 2008; Chen et al., 2011). Despite these and many other important findings advocating a central importance of abnormal beta oscillations, their role in the pathophysiology of PD has been put into question by several studies finding no direct correlation to disease severity in human patients and in animal models of PD (Weinberger et al., 2006; Leblois et al., 2007; Kuhn et al., 2009). Since beta oscillations in the basal ganglia of humans can only be studied in patients already in advanced disease states treated with deep brain stimulation, little is known so far about signaling in the cortico-basal ganglia loop early in the disease and in healthy human subjects. Furthermore, the direct relationship between reduced basal ganglia dopamine levels and the emergence of enhanced beta oscillations remains unclear. Experimental

M.H. Beck et al. / Experimental Neurology 286 (2016) 124–136

animal studies investigating this question by using an acute pharmacologic blockade of the dopamine receptors D1 and D2 showed seemingly ambiguous results: while some studies found evidence for increased beta band oscillations under D1-D2-receptor antagonists (Costa et al., 2006; Dejean et al., 2011), others showed no changes in this regard (Degos et al., 2005; Mallet et al., 2008b). Thus, it still remains unclear if beta oscillations are either directly linked to a lack of dopamine, or if they might be simply an epiphenomenon developing late in a state of chronic dopamine depletion. The aim of our study was to further clarify this question by investigating cortico-basal ganglia circuit abnormalities in two animal models of PD which differ substantially in the time course until their maximum phenotype is reached: We opted to investigate the reserpine model, which results in an acute, long-lasting dopamine depletion due to an irreversible blockade of the vesicular monoamine transporter (VMAT2) of monaminergic neurons (Leao et al., 2015), and compared it to the chronic progressive 6-hydroxydopamine MFB-model (medial forebrain bundle) where the dopaminergic deficits develop a few days after the lesion progressively over several weeks (Blandini and Armentero, 2012). Thus, we wanted to clarify the differential effects of short- and long-term dopamine depletion on signaling characteristics in the cortico-basal ganglia loop. The 6-OHDA model is well known to exhibit enhanced beta oscillations (Sharott et al., 2005; Mallet et al., 2008b; Delaville et al., 2015), while this has not been studied so far in the reserpine model. We performed simultaneous extracellular electrophysiological recordings from the primary motor cortex (ECoG, LFP), the subthalamic nucleus and the reticulate part of substantia nigra (both LFP and multi-unit activity) under urethane anesthesia and assessed the animals motor behavior and histology. 2. Methods 2.1. Animals and materials Experimental procedures were carried out on 27 male Wistar-rats (Harlan Winkelmann, Germany), and were conducted in accordance with the German Animal Welfare Act (last revised in 2014) and European regulations (2010/63/EU). Experiments were approved by the local animal welfare authority (LaGeSo, Berlin), and conformed to local department and international guidelines. Every effort was made to minimize the number of animals and to reduce the animal's harm caused by the experimental setting. Animals were kept on a 12 h light cycle in standard housing conditions and received food and water ad libitum. Unless indicated otherwise, materials were obtained from Sigma-Aldrich, Germany. All stereotactic coordinates were measured in relation to the bregma (Paxinos and Watson, 2006). 2.2. Experimental groups Three groups of animals (n = 9 each) were formed in this study: healthy control (285–363 g), reserpine-treated (305–377 g) and 6OHDA lesioned (304–355 g) animals. 2.3. Unilateral 6-hydroxydopamine lesions of dopaminergic neurons The neurotoxin 6-OHDA-hydrochloride was dissolved in NaCl 0.9% containing 0.02% ascorbic acid at a final concentration of 8 μg/μl of the free base. The solution was stored at −80 °C and defrosted directly before injection. Anesthesia was induced and maintained with a combination of fentanyl (5 μg/kg, s.c., Rotexmedica, Germany), medetomidine (150 μg/kg, s.c., Domitor ®, Provet AG, Germany) and midazolam (2 mg/kg, s.c., Hameln Pharma, Germany). Rats were placed in a stereotactic frame (David Kopf Instruments, CA, USA) with heads fixed with atraumatic ear bars. Ophthalmic ointment (Bepanthen™, Bayer, Germany) was applied to prevent corneal dehydration. Body temperature of 37 ± 0.5 °C was maintained throughout surgery using a self-adjusting

125

heating pad (CMA, Sweden). After incision of the skin, the head was aligned to the flat skull position using the rat alignment tool (David Kopf Instruments, CA, USA). Following a small craniotomy, a 33-gauge blunt-tip cannula fixed on a 10 μl Hamilton syringe (World Precision Instruments, FL, USA) was inserted into the left medial forebrain bundle (MFB, AP: − 2.6, ML: + 1.6, DV: − 8.4 mm) and 1 μl of 6-OHDA was injected at a pace of 0.125 μl/min via a precision syringe pump (Micro 4™, World Precision Instruments, FL, USA). After completion of the injection the cannula remained in place for 5 more minutes to avoid a reflux of the liquid. After wound closure, anesthesia was reversed using a combination of naloxone (120 μg/kg, s.c., B. Braun Melsungen AG, Germany), flumazenil (200 μg/kg, s.c., Inresa, Germany) and atipamezole (750 μg/kg, s.c., cp-pharma, Germany). On the first three days after surgery the animals received the analgesic carprofen (5 mg/kg, s.c., Pfizer, Germany) to minimize their distress. The efficacy of the 6-OHDA lesion was assessed behaviorally and post-mortem using immunochemistry for TH (both see below). The electrophysiological experiments were conducted 20–30 days following the injection of the neurotoxin, at a time when a maximal lesion of the dopaminergic system can be expected (Sharott et al., 2005; Blandini and Armentero, 2012). 2.4. Pharmacologic dopamine depletion with reserpine To achieve an acute and widespread dopamine depletion, known to last at least 48 h, (Leao et al., 2015), rats received systemic injections of reserpine (3 mg/kg, i.p., Sigma-Aldrich). Reserpine was dissolved in 0.9% NaCl and dimethyl-sulfoxide (4 mg/l). Efficiency of the pharmacologic dopamine depletion was assessed behaviorally (see below). Furthermore, body weight was monitored carefully, since it has been demonstrated before that a weight reduction exceeding 5% predicts widespread reserpine-induced amine depletion (Halaris and Freedman, 1975). All animals reached this criterion. Eighteen hours after the injection reserpine-treated rats underwent electrophysiological recordings. At this point in time, an almost total dopamine depletion (N95%) can be expected in the basal ganglia (Leao et al., 2015). 2.5. Behavioral testing Motor testing was performed in all animals prior to the electrophysiological recordings. Additionally, reserpine-treated and 6-OHDA-lesioned animals were evaluated in a baseline testing before reserpine application or 6-OHDA lesion. All behavioral examinations were videotaped (Canon Legria HF R506, Canon, Japan) for offline data analysis. Since the two examined animal models are known to show different motor phenotypes, hemiakinesia in case of 6-OHDA and bilateral akinesia in case of reserpine (Bezard and Przedborski, 2011; Leao et al., 2015), we conducted different tests to quantify the animal's motor performance. To assess the unilateral motor deficit in the 6-OHDA-lesioned animals we performed the limb use asymmetry test and the drag test (Meredith and Kang, 2006). For the limb asymmetry use test (cylinder test) rats were placed in a transparent acrylic glass cylinder (height: 45 cm, diameter: 20 cm) without prior habituation (Schallert et al., 2000). During the rat's vertical exploration of the walls, full contacts of the right and the left forepaw were counted. The rats were left in the cylinder until at least one paw had touched the walls a minimum of 15 times. For the drag test the rat's hind limbs were raised off the ground leaving the forelimbs in touch with the ground (Olsson et al., 1995). Then the animal was dragged backwards for the distance of one meter and the adjusting steps of each forepaw were counted separately. Results for both tests are presented as a ratio of right to left forelimb uses. The test was repeated until a minimum of 15 steps was made by at least one of the paws. Reserpine-treated animals were examined with the bar test (Sanberg et al., 1988). For this test the rat's forelimbs were placed on a small pedestal (height: 8 cm, width: 10 cm) and the time needed for removal of both forelimbs was measured. The test was also completed when a time limit of 300 s was exceeded. Control

126

M.H. Beck et al. / Experimental Neurology 286 (2016) 124–136

group animals underwent all three motor tests to provide comparability to both models. 2.6. In vivo electrophysiological recordings in urethane anesthesia We performed in vivo electrophysiological recordings on 6-OHDA and reserpine-treated animals as well as on controls. All animals in the PD groups showed profound motor impairment. At the time of recording 6-OHDA animals weighted between 329 and 384 g, reserpinetreated animals between 283 and 350 g and control animals between 285 and 363 g. Anesthesia was induced and maintained with urethane (1.3 g/kg, i.p.). The surgical approach was performed likewise to stereotactic 6-OHDA injections until craniotomy. Small craniotomies were drilled above the target regions leaving the dura mater intact. Two custom-made Ag/AgCl-spherical tip electrodes (made from 99.99% silver wire, Goodfellow, UK; diameter 200 μm, 8 kΩ impedance) were placed in the epidural space above the left primary motor cortex (M1, AP: +3.0 ML: +3.0) to record the electrocorticogram (ECoG). Two Ag/AgCl electrodes were placed over the ipsi- and contralateral cerebellum (AP: −9.0, ML: +3.0/−3.0) for referencing. All epidural electrodes were secured with acrylic dental cement (Technovit®, Heraeus-Kulzer, Germany). Two parylene insulated tungsten microelectrodes (Microprobes for Life Science, MD, USA; tip diameter 4 μm, tip separation 250 μm) were slowly inserted into the left subthalamic nucleus (STN; AP: −3.6, ML:

+ 2.5, DV: − 8.0 mm) and the left reticulate part of substantia nigra (SNr, AP: −4.8, ML: +2.5, DV: −8.0 mm) under stereotactic guidance and constant monitoring of multi-unit activity to secure successful targeting (Fig. 1). During the trajectory of the electrodes STN, SNr and neighboring structures can be distinguished by their characteristic multi-unit activity (Fig. 1A). The STN shows an irregular, bursty, high frequency firing pattern with a medium amplitude for 100-300 μm in the dorsoventral diameter, as neurons are arranged densely in this nucleus. Furthermore, the STN is clearly demarcated from its neighboring structures, the zona incerta (ZI) above and the cerebral peduncle below, as both show sparse neural activity. Another important landmark on the way to STN is the ventral posteriomedial thalamic nucleus with a highly characteristic firing pattern (sparse, bursty, high amplitude potentials) transitioning into the ZI about 1 mm above the STN. Compared to the STN, the SNr signal shows a higher amplitude and a more heterogeneous firing pattern during the trajectory. The SNr also shows a greater dorsoventral diameter (~ 500 μm). Correct electrode placement was ensured histologically (see below). Due to the impedance (1–1.5 MΩ) of the tungsten electrodes we were able to record local field potentials (LFP) as well as multi-unit activity from the basal ganglia target regions. STN and SNr electrodes were referenced to the contralateral cerebellum, while the epidural ECoG electrodes were referenced to the ipsilateral cerebellar electrode. After successful implantation and electrode placement the animal was

Fig. 1. Representative traces of multi-unit activity and local field potentials under urethane anesthesia. A. Characteristic multi-unit recordings from the ventral posteromedial thalamic nucleus (VPM), the zona incerta (ZI), the subthalamic nucleus (STN) and the substantia nigra, reticular part (SNr). Beside the stereotactic coordinates, the highly specific multi-unit activity (MUA) of the four different structures seen during the positioning of the recording electrodes was a key information for target structure identification during the electrophysiological recordings. VPM-MUA demonstrated rare, high-amplitude potentials, which ceased about 1 mm above the STN. Usually the proximate ZI was characterized by sparse or absent spiking activity. Subthalamic MUA showed an irregular, homogeneous, bursty, high-frequency firing pattern with a medium amplitude, while SNr spiking demonstrated higher amplitudes and a frequent, more regular firing pattern. B Brain states during urethane anesthesia in the electrocorticogram (ECoG) of the primary motor cortex (M1) in a 6-OHDA lesioned animal. In urethane anesthesia two brain states can be observed: the slow wave activity state (SWA) is dominated by a 0.05–1 Hz oscillation and a higher amplitude signal. In the activated state (AS), the ECoG is characterized by a faster rhythm with smaller amplitude. The topmost illustration displays 2500 s of M1-ECoG, note the temporal alterations in amplitude in urethane anesthesia. The two dashed grey lines point to representative ECoG sections of AS and SWA with a higher temporal resolution. The bottommost illustration is a time frequency plot of the 0 to 5 Hz relative power of the uppermost 2500 s ECoG record. Warmer colors indicate higher relative power values. Reductions in amplitude are accompanied by a reduction of the relative 0–5 Hz power. The red line in both illustrations marks the same point of time. C Two 300 s ECoG time frequency plots of a 6-OHDA-lesioned animal: one during activated state and the other during slow wave activity. Again warmer colors indicate higher relative power values. Beside the reduction of the 0–5 Hz power during the activated state, an increase in power in the beta frequency band (13–35 Hz) can be observed in the activated state. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

M.H. Beck et al. / Experimental Neurology 286 (2016) 124–136

positioned in a faraday cage and electrophysiological recordings were started using a programmable neural data acquisition system (Omniplex, Plexon, TX, USA). The recorded wide band signal was bandpass filtered (0.05–8000 Hz), amplified (1750 ×) and sampled at 40 kHz. For later offline data analysis the acquisition system immediately performed further bandpass-filtering to split the signal into mainly LFP (0.05–250 Hz, down sampled to 1 kHz) and multi-unit activity (300–8000 Hz) data channels. During recordings electrocorticogram, respiration rate and body temperature were monitored closely to ensure the animal's wellbeing. 2.7. Histology and immunochemistry At the end of the recordings animals were sacrificed using a lethal dose of anesthetics. Rats were transcardially perfused with 250 ml of 0.1 M phosphate buffered saline (PBS, pH = 7.4), followed by 250 ml of 4% paraformaldehyde in 0.1 M PBS at 4 °C. Brains were removed and postfixed in the fixative for 24 h. For cryoprotection, brains were immersed in sucrose solutions of ascending concentrations (10–30%) and afterwards cryopreserved at −80 °C until sectioning. Coronal sections (40 μm) were cut at −20 °C using a cryotome (Leica, Germany). Sections dedicated to verification of electrode placement were collected on glass slides and were then Nissl stained (cresyl violet). Striatal and mesencephalic serial sections were processed for tyrosine hydroxylase immunohistochemistry (TH) in free floating sections using a standard protocol (Steiner et al., 2008) with a primary anti-TH mouse antibody (1:10,000, Sigma (T1299), Germany) and a secondary anti-mouse biotin-antibody (1:200, Vector BA-9200, Vector Laboratories, CA, USA). After staining sections were mounted on glass slides for further analysis. The correct placement of electrodes was verified by microscopic examination of the target and neighboring structures on the Nissl-stained sections. Only recordings from correctly placed electrodes in structurally intact target areas were further analyzed. TH-positive cells of the substantia nigra pars compacta were quantified using a Stereo Investigator System (MicroBrightField Bioscience, VT, USA) on a DMRE microscope (Leica, Germany) as previously described (Harnack et al., 2008). Cells were counted systematically on every 7th section and the total cell count was estimated using the built-in optical dissector method of the Stereo-Investigator system. Results from stereological counts are expressed as a ratio of the lesioned to the intact control hemisphere. Densitometric analysis of striatal TH content was conducted using the MCID analysis system (Northern Light R95 Precision Illuminator, MCID, UK; Cool Snap EZ Camera, Roper Scientific, Germany). We assessed three representative coronal sections of the striatum per animal. The sections were taken from the anterior (AP: + 1.70), middle (AP +0.20) and posterior (AP −0.50) striatum. We focussed our measurements on the dorsolateral striatum, because it is known to be involved in the motor loop of the basal ganglia. Here, ten measuring fields were randomly selected for each hemisphere. Values of assumed background-staining noise defined as the average optical density of the cortex were subtracted from the average striatal optical density of each hemisphere. We estimated the average cortical optical density by randomly assigning ten measuring fields in the ipsi- and contralateral cortex. The results are displayed as normalized relative density (in %) in relation to the non-lesioned hemisphere.

127

dominated by slow 0.05–2 Hz LFP oscillations, while in the activated state, the LFP is characterized by decreased power of these slow oscillations and a smaller amplitude. A separate analysis for the slow wave activity and activated network state was performed. Unless stated otherwise, 50s epochs of robust SWA and AS were selected for further data analysis. All selected data segments were checked for 50 or 100 Hz artifacts and we applied appropriate filters. 2.9. Power and coherence calculation Power spectral densities of the LFP for the three recording sites (STN, SNr, ECoG) were calculated by the Fast Fourier Transformation using the Hann window function (block size of 1024) with a resulting ~1 Hz resolution. The computed power spectrograms were normalized by the average 50-100 Hz power. To visualize power changes over time we constructed time-evolving power spectra from continuous LFP sections using non-overlapping windows with a resulting 1 Hz resolution. The similarity in frequency content of the LFP waveforms from the different recording sites was estimated by calculating means of coherence. In brief, coherence is a measure of linear correlation between two signals and ranges from 0 for totally incoherent waveforms to 1 for a total dependency of the both signals. Coherence is defined as the normalized cross-spectrum (Nolte et al., 2004) and can be calculated through the Fast Fourier Transformation by the following equation: 2 X   csdab ð f Þ  X cohð f Þ ¼ X psdb ð f Þ psda ð f Þ In this equation coherence (coh) is calculated from the cross spectral density (csd) between the two waveforms (a and b) normalized by the power spectral density (psd) of each waveform for a given frequency (f). We computed coherence between the motor cortex and the recorded basal ganglia nuclei (M1-STN, M1-SNr) and also between STN and SNr. Power and coherence values were averaged across the following frequency ranges: the beta (13–35 Hz) and gamma (36–80 Hz) band first defined by Hans Berger (Berger, 1938). Since there is evidence for a functional role of subdivisions of the traditional frequency bands in the beta and gamma spectra (Priori et al., 2004; Lopez-Azcarate et al., 2010; Hohlefeld et al., 2013; Little et al., 2013b; Toledo et al., 2014), we separately evaluated those frequency bands: in case of the beta band in a low beta (13–20 Hz) and a high beta spectrum (21–35 Hz) and for the gamma spectrum in a low gamma (36–50 Hz) and a high gamma (51–80 Hz) band. Significant peaks in power spectra or in coherence were identified by the following criteria (Delaville et al., 2015): 1) A local peak maximum had to be greater than the surrounding twelve 1 Hz bins (six before and after). 2) The peak had to have the highest amplitude in the frequency range of interest. 3) The first derivative of the waveform had to be positive before and negative after the peak maximum. 4) The second derivative of the waveform had to be negative at the peak, indicating a concave shape. We evaluated peaks for all sub-bands. Differences in power, coherence and peak frequency values between the three different groups were contrasted.

2.8. Data analysis

2.10. Spike sorting and spike-triggered waveform analysis

LFP and spike data was analyzed using build-in Spike2 analysis software functions (Cambridge Electronic, Design, UK) and custom-written scripts for Spike2 and Matlab (MathWorks, MA, USA). Data sections free of visually detectable artifacts were selected for further analysis. For STN and SNr data, we chose the electrode that showed robust local multiunit activity. As previously described two different altering cortical activation states can be observed under urethane anesthesia (Steriade, 2000; Mallet et al., 2008b). The Slow Wave Activity (SWA) state is

Apart from increased beta amplitudes in basal ganglia LFP's, phase locking between local basal ganglia spiking and LFPs have been described in human PD patients (Kuhn et al., 2005; Weinberger et al., 2006) and animal models of PD (Mallet et al., 2008b; Delaville et al., 2015). We intended to investigate if altered spike-field interactions can be found both in acute and chronic states of dopamine depletion by calculating spike-triggered waveform averages (STWAs). Spike identification and sorting was performed offline after the recordings using

128

M.H. Beck et al. / Experimental Neurology 286 (2016) 124–136

the Offline Sorter Software (Plexon, TX, USA) and its principal component analysis function. The threshold was set N 2 standard deviation above the background noise. We controlled our data for spikes with an interspike interval of b1 ms. We opted to use multi-unit spike activity instead of sorted single unit activity, since it has been shown previously that single unit activity does not provide additional benefits over multi-unit activity, while lowering the sensitivity for functional spike-LFP interactions (Fries et al., 2001). To assess the temporal interaction between spiking activity in the basal ganglia and LFPs from the same structure or the motor cortex we calculated spike-triggered waveform averages (STWAs) (Delaville et al., 2015). 100 s epochs LFP of activated brain state were bandpass-filtered to the frequency range of interest (13–20 Hz, 21–35 Hz or 36– 50 Hz) using a finite impulse response (FIR) filter (Spike2). Bandpassfiltered LFP waveforms were averaged ±0.25 s around the occurrence of a spike. To quantify the extent of phase-locking of multi-unit activity with LFP oscillations we used the peak-to-trough amplitude at or around a spike (zero point) as a measure. As a control condition each interspike interval was then shuffled 20 times and each shuffled version was used to calculate another STWA (Delaville et al., 2015). The peakto-trough amplitudes of the shuffled STWAs were normally distributed. We assumed a significant phase-locking between the multi-unit activity and the LFP oscillation if the unshuffled peak-to-trough amplitude exceeded the mean + 3 SDs of the mean of the shuffled condition (Delaville et al., 2015). To evaluate the extent of spike-field phaselocking in the predefined frequency ranges and differences between the three groups, we compared the number of significant phase locked animals and the mean of the ratio of the unshuffled to shuffled peak-totrough amplitudes. 2.11. Statistical analysis Statistics were calculated using Prism GraphPad (GraphPad Software, CA, USA). Normal distribution was tested using the D'Agostino & Pearson omnibus normality test. For data sets containing three groups we evaluated statistical significance using one-way ANOVA and post hoc analysis testing in case of normal distribution, for data sets that did not show normal distribution we performed the Kruskal-Wallis test. Posthoc tests were only calculated, if the initial ANOVA or Kruskal-Wallis test showed significant results. To compare two single groups, we used the unpaired t-test for normally distributed data and the Mann-Whitney test for not normally distributed data sets. The significance level of all test was defined to be alpha b 0.05 (* b 0.05; ** b 0.01; *** b 0.001; **** b 0.0001). Unless stated otherwise, group data are expressed as the mean ± SEM. p6OH-Res: refers to posthoc analysis of 6-OHDA vs reserpine-treated animals. p6OH-Con: refers to posthoc analysis of 6-OHDA vs control animals. pRes-Con: refers to posthoc analysis of reserpine-treated animals vs control animals. 3. Results 3.1. Effects of 6-OHDA and reserpine treatment on motor activity and histology The primary aim of our study was to assess oscillatory activity in the cortico-basal ganglia loop in two different animal models of Parkinson's disease that differ substantially in the time course until their maximal phenotype is reached. We compared the acute reserpine model (Res), that is known to reach its maximal phenotype within a few hours (Heeringa and Abercrombie, 1995; Oe et al., 2010) and the chronic progressive 6-OHDA model (6-OH) of PD with untreated controls (Con). All reserpine and 6-OHDA animals showed profound motor impairment and underwent further electrophysiological characterization. Severe bilateral motor dysfunction was proven in the reserpine model 18 h after the injection of the toxin using the bar test (Sanberg et al.,

1988). Reserpine-treated rats showed a high degree of akinesia, that could not be seen in control rats (Supplementary Fig. 1A, Supplementary Table 5). In the unilateral 6-OHDA model the cylinder test and the drag test (Olsson et al., 1995; Schallert et al., 2000) were employed 20–30 days after the injection of 6-OHDA into the medial forebrain bundle. Both tests documented almost maximal hemi-akinesia with significantly reduced use of the right forepaw vs. the left compared to controls (Supplementary Fig. 1A, Supplementary Table 5). An almost maximal dopaminergic lesion was further confirmed in all 6-OHDA-lesioned rats by post-mortem histological analysis of remaining dopaminergic fibers in the striatum (STR) and cells in the substantia nigra pars compacta (Supplementary Fig. 1B/C, Supplementary Table 4), which was not possible in the reserpine model, since it does not lead to a change in these measures (Leao et al., 2015). Correct electrode placement at intact recording sites could be verified in all animals by lightmicroscopic examination of Nissl-stained coronal sections.

3.2. Both models show increased beta oscillatory power in the basal ganglia but differ in peak frequency and involvement of the motor cortex We recorded simultaneously from the primary motor cortex (M1; ECoG; LFPs), the subthalamic nucleus and the reticulate part of substantia nigra (STN, SNr; LFPs and MUA) under urethane anesthesia. As previously described two different altering cortical activation states can be observed under urethane anesthesia. (Steriade, 2000; Mallet et al., 2008b). The Slow Wave Activity (SWA) state is dominated by slow 0.05–2 Hz LFP oscillations, which resemble the cortical activity patterns during natural sleep. In the activated state, the LFP is characterized by decreased power of these slow oscillations and a smaller amplitude. The activated state (AS) is therefore comparable to brain activity of awake and alert individuals. Although those brain states may differ from neural activity seen in the unanesthetized brain, the urethaneanesthetized animal still serves as well-established model for characterizing basal ganglia activity (Magill et al., 2001; Mallet et al., 2008b) and functional connectivity within and between basal ganglia and the cerebral cortex (Magill et al., 2006). A separate analysis for the slow wave activity and activated network state was performed (see Fig. 1 for representative traces of LFP and MUA, for further details see methods). As has been previously reported in the literature on the 6-OHDA model (Mallet et al., 2008a; Mallet et al., 2008b), we could not detect any significant changes in SWA-LFPs in spectral power between the examined models and controls. Apart from the characteristic 1 Hz peak, no other distinct peaks were observable in the beta or any other frequency range in the power spectra (data not shown). In view of the lack of any detectable peaks and changes in power, we focused our further analysis on the activated network activity state. Detailed results for the activated network state (power, coherence and spike-triggered waveform averages) are displayed in the Supplementary Tables 1–3. In agreement with previously published data from recordings in urethane-anesthetized animals, the AS-LFPs from the primary motor cortex of 6-OHDA-lesioned animals showed a distinct peak in the beta frequency range (Figs. 1B and 2A) (Sharott et al., 2005; Mallet et al., 2008b). Neither in the M1-ECoG from reserpine-treated nor in control animals, a peak in the beta range could be detected. In the statistical analysis for mean average beta power as well as for low and high beta power, 6-OHDA-lesioned animals showed highly significantly enhanced power levels in comparison with reserpine-treated rats and elevated whole beta levels compared to controls (Fig. 2A). The average beta peak frequency in 6-OHDA-lesioned animals was 18.6 ± 1.3 Hz (Fig. 4A). Although the majority of animals showed a single frequency peak in the low beta range (6/9), some animals had peaks that also extended to high beta (2/9; Fig. 4). The statistical analysis of the other frequency bands revealed significantly lower gamma levels (36–80 Hz) in reserpine-treated rats compared to controls and 6-OHDA-lesioned animals (whole gamma power: 6-OH 5.3 ± 0.2 vs Res 3.9 ± 0.4 vs Con 5.3 ±

M.H. Beck et al. / Experimental Neurology 286 (2016) 124–136

M1

Beta [13-35 Hz] High Beta

Low Gamma

200

150

Control

50

0 10

20

30

40

50

60

70

80 [Hz]

Control

Control

Low Gamma

75

6-OHDA Reserpine

Low Beta [13-20 Hz] Normalized Power

Beta [13-35 Hz] High Beta

6-OHDA Reserpine

Low Gamma [36-50 Hz]

6-OHDA Reserpine

Normalized Power

Low Beta

Control

6-OHDA Reserpine

Control

6-OHDA Reserpine

50

Low Gamma [36-50 Hz]

High Beta [21-35 Hz]

25

0 20

30

40

50

60

70

80 [Hz]

Control

SNr

C

6-OHDA Reserpine

Control

Beta [13-35 Hz]

Low Beta

High Beta

Normalized Power

100 Low Gamma

75

50 40 30 20 10

6-OHDA Reserpine

Low Beta [13-20 Hz] Normalized Power

10

Normalized Power

Normalized Power 0

80 60 40 20 0

Control 6-OHDA Reserpine

Control 6-OHDA Reserpine

50 Normalized Power

High Beta [21-35 Hz]

25

0 10

20

30

40

50

60

70

80 [Hz]

40

Low Gamma [36-50 Hz] Normalized Power

Normalized Power [a. u.]

Control

STN 100

Normalized Power [a. u.]

6-OHDA Reserpine

Normalized Power

100

B

Low Beta [13-20 Hz]

High Beta [21-35 aHz] Normalized Power

Normalized Power [a. u.]

Low Beta

Normalized Power

250

Normalized Power

A

129

30 20 10 0

20 15 10 5 0

Control 6-OHDA Reserpine

Control 6-OHDA Reserpine

Fig. 2. Normalized power spectra of M1 (A), STN (B) and SNr LFPs (C). Frequency-power plots for the 7–80 Hz range on the left side are displaying group-averaged normalized power values for control (black), 6-OHDA (red) and reserpine (blue). The low beta (13–20 Hz), high beta (21–35 Hz) and low gamma (36–50 Hz) frequency band are shaded in different grey tones. Bar plots displaying results for average beta (13–35 Hz), low beta, high beta and low gamma normalized power are corresponding to the frequency-power plots on the figure's left side. Results for controls are shown in black, for 6-OHDA in dark grey and for reserpine in bright grey. Stars are indicating significant group differences and the level of significance. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

0.3 a.u.; p = 0.0072: p6OH-Res b 0.05; p6OH-Con N 0.05; pRes-Con b 0.05). No distinct gamma peaks for 6-OHDA or controls were observed. In the recorded basal ganglia targets, STN and SNr, we observed prominent beta oscillatory power increases in both PD models. Notably, the prevalent beta peak frequency was significantly lower in 6-OHDA compared to reserpine-treated animals, both for the STN as well as for

the SNr (Fig. 4). Although the spectral power peak in the reserpine model was always identified to be in the beta band, the elevation in the power spectrum extended also into the low gamma range, which was never the case for 6-OHDA treated animals (Fig. 2B/C). While there was a relatively uniform distribution of peaks in low beta in the 6-OHDA model (STN: 16.8 ± 0.9 Hz; SNr: 15.4 ± 0.4 Hz; Fig. 4), the

130

M.H. Beck et al. / Experimental Neurology 286 (2016) 124–136

peaks of reserpine-treated rats tended to be much more focused in the high beta frequency range (STN: 26.5 ± 2.2 Hz; SNr: 27.9 ± 2.3 Hz; Fig. 4). There was no significant difference between beta peak frequencies in STN or SNr within the models. The statistical analysis of subthalamic and nigral average beta power showed for 6-OHDA-treated rats a significant elevation compared to control animals, but not to reserpine-treated animals (Fig. 2B/C). Due to the observed difference in beta peak frequency, we proceeded with the statistical analysis of the low and high beta range to further delineate the changes observed in the examined models. As expected, we found a significant low beta increase in the 6-OHDA-lesioned animals compared to the reserpine group and controls for both basal ganglia nuclei (Fig. 2B/C). In the STN the high beta frequency band was significantly elevated in both models over controls, while there was no difference between the two models (Fig. 2B). In SNr high beta power was significantly higher in reserpine than in controls, while 6-OHDA showed a trend for an elevation that did not reach statistical significance (Fig. 2C). Again, there was no difference between the animal models of PD in this measure (Fig. 2B/C). Due to the spread of the spectral power elevation from the beta into the low gamma band in reserpine, low gamma power (36–50 Hz) was also significantly increased. Note that there were no distinct gamma peaks in any group. In the STN reserpine and 6OHDA treated animals displayed elevated low gamma levels compared to controls, while in the SNr only reserpine-treated animals had a

significant low gamma increase (Fig. 2B/C). In summary, only chronically dopamine depleted 6-OHDA-lesioned animals showed a significant beta power increase in motor cortex, while no change could be detected in the reserpine group. In the basal ganglia, there was a significant increase in beta activity in both PD models with substantially different peak frequencies. 3.3. Cortical-subcortical beta coherence is elevated in both models but differs in peak frequency To further characterize cortical-subcortical and basal ganglia LFP interactions, we proceeded with the analysis of coherences between the recorded structures. We found substantially increased cortico-subthalamic and cortico-nigral coherences in 6-OHDA-lesioned as well as in reserpine-treated animals, but not in controls (Fig. 3A/B). While peak frequencies of spectral elevations were in the beta frequency band in both animal models of PD, the coherence extended substantially into the low gamma range (Fig. 4B). Comparable to our findings in the power analysis, 6-OHDA-lesioned animals had their mean maximal peak in low beta, whereas reserpine-treated animals exhibited a significantly higher mean peak frequency (Fig. 4B). The statistical analysis of the spectral frequency bands revealed significantly enhanced corticonigral synchrony in the beta range in the 6-OHDA and reserpine model over controls but no differences between the PD models (Fig.

Fig. 3. Coherence spectra between ECoG (M1) and LFPs in STN (A) and SNr (B) and between STN and SNr (C). Frequency-coherence plots for the 7–80 Hz range on the left side are displaying group-averaged coherence values for control (black), 6-OHDA (red) and reserpine (blue). Bar plots displaying results for average low beta (13–20 Hz), high beta (21– 35 Hz) and low gamma (36–50 Hz) coherence are corresponding to the frequency-coherence plots on the figure's left side. Results for controls are shown in black, for 6-OHDA in dark grey and for reserpine in bright grey. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

M.H. Beck et al. / Experimental Neurology 286 (2016) 124–136

A

Peak Frequency - Power

Power Peak Frequency [Hz]

M1

Coherence Peak Frequency [Hz]

B

STN

SNr

***

***

131

gamma range. A statistic trend (p = 0.097) was observed for elevated high beta coherence levels in the reserpine treated group. STN-SNr coherence levels were generally elevated compared to cortico-basal ganglia coherence. For further details, see Figs. 3C and 4B. In short, there was a broad beta-low gamma cortico-nigral coherence increase in both animal models of PD which could not be seen in the control group. In case of the cortico-subthalamic coherence, the beta-low gamma elevation was only significant for the reserpine-treated group. There were no group differences in case of the subthalamo-nigral coherence. 3.4. Only 6-OHDA-lesioned animals show significant spike-LFP entrainment in the beta frequency range

Peak Frequency - Coherence M1-STN

M1-SNr

STN-SNr

**

**

**

Fig. 4. Beta peak frequencies in power (A) and coherence (B). Grey squares are representing 6-OHDA-lesioned animals, grey circles reserpine-treated animals. A. Power peak frequencies in M1, STN and SNr. Note that reserpine-treated animals did not show power peaks in the beta range in M1. B. Peak frequencies for cortico-subthalamic (M1STN), cortico-nigral (M1-SNr) and subthalamo-nigral (STN-SNr) coherence. 6-OHDA peaks were significantly more centered in the low beta and reserpine peaks more in the high beta range. Statistical differences between 6-OHDA and reserpine peak frequencies within the same structure were compared, stars are indicating the significance level. Data are means ± SD.

5B). The cortico-subthalamic beta coherence increase was only significant for reserpine compared to controls (Fig. 3A). For 6-OHDA we found a clear trend in the low beta range, where the maximum of the 6-OHDA coherence peak was located (p = 0.0836; Fig. 3A), while the elevated reserpine cortico-subthalamic beta coherence was significantly more shifted in the high beta range (Fig. 3A). Similar results were seen in cortico-nigral coherence. The 6-OHDA beta peak centered in the low beta range, while reserpine beta coherence was centered in high beta (Fig. 3B). Due to the spread of the coherence elevation from beta into the low gamma band, we found elevated cortico-subthalamic low gamma coherence levels for reserpine and increased cortico-nigral low gamma coherence for 6-OHDA or reserpine over controls (Fig. 3A/B). No distinct peaks were found in the low gamma spectrum. STN-SNr coherence was also calculated, but no significant group differences were found for the low beta, high beta and low

To gain deeper insight into alterations on a local cell level and the interaction between spiking and LFPs, we calculated spike triggered waveform averages (STWA) within the STN or SNr and between motor cortex and STN or SNr (Fig. 5 and Supplementary Fig. 2). As aforementioned, alterations in power and coherence mainly occurred in the low and high beta and in the low gamma band. Therefore, we focused our STWA analysis on these frequency bands. In both PD models, STN multi-unit activity was entrained to low beta oscillations in the majority of animals (6-OH: 7/9, Res: 6/9), whereas this was not the case in untreated controls. Although a similar number of animals showed significant spike-field phase-locking in the low beta range, only 6-OHDA-lesioned animals demonstrated a significantly elevated mean ratio of the unshuffled-to-shuffled STWA peak-to-trough amplitudes compared to controls (Fig. 5A). Subthalamic high beta STWAs of all three groups exhibited significant phase locking in a majority of animals (6-OH 8/9 vs Res 6/9 vs Con 6/9). Unshuffled-to-shuffled high beta STWA peak-to-trough amplitudes were also elevated in all conditions, but 6-OHDA amplitudes differed significantly from controls and reserpine-treated animals (Fig. 5B). Within the SNr we found no substantial elevation in the mean ratio of unshuffled-to-shuffled low or high beta STWA peak-to-trough amplitudes for any group and only a few animals showed significant phase-locking (Fig. 5D/E). We also investigated the extent to which STN and SNr multi-unit activity was phase locked to low or high beta M1-ECoG activity (Supplementary Fig. 2). In the examined groups we found STN multi-unit activity significantly entrained to M1-LFPs in a minority of animals (Supplementary Fig. 2). Although there was a marked increase in the mean ratio of unshuffled-to-shuffled STWA peak-to-trough amplitudes for 6-OHDA, it did not reach statistical significance due to a high variation of the degree of phase-locking (p = 0.0967; Supplementary Fig. 2). There was no evidence for an entrainment of SNr multi-unit activity with the beta-LFP of the primary motor cortex. Only one animal in the 6-OHDA group showed significantly enhanced low beta STWA peak-to-trough amplitudes, and four of the 6-OHDA animals displayed significant high beta spike-LFP coupling (Supplementary Fig. 2). In conclusion, we found significant spike-field interactions in the low- and high beta frequency only within the STN in the 6-OHDA group. The number of animals with phase-locked low beta-STWAs was significantly elevated in the reserpine group for the STN, but did not reach statistical significance in the mean ratio of unshuffled-toshuffled STWA peak-to-trough amplitudes. 3.5. Low gamma spike-LFP interactions are high in all groups Since the spectral power and coherence analysis had also indicated a role of enhanced low gamma LFP activity in the two animal models of PD, we also analyzed spike-LFP interactions in this frequency band. In the STN, the analysis showed significantly phase locked multi-unit activity in 8/9 in control animals and 7/9 each for the 6-OHDA and the reserpine group. The mean ratio of unshuffled-to-shuffled STWA peak-totrough amplitudes was generally high, but similar between the experimental groups (Fig. 5C). In the SNr we observed no differences in the

132

6/9

6/9

Control

Uns huffled to Shuffled STWA Amplitude Ratio

Amplitude [mV]

[s]

Reserpine

F

7/9 6-OHDA

7/9 Reserpine

Uns huffled to Shuffled STWA Amplitude Ratio

8/9 6-OHDA

1/9 6-O HDA

5/9 Control

0/9 Reserpine

3/9 6-OHDA

2/9 Reserpine

SNr STWA [36-50 Hz] Uns huffled to Shuffled STWA Amplitude Ratio

Amplitude [mV]

Uns huffled to Shuffled STWA Amplitude Ratio

6/9

2/9 Cont rol

SNr STWA [21-35 Hz ]

Amplitude [mV]

E

Control

8/9

[s]

Reserpine

STN STWA [36-50 Hz]

[s]

Uns huffled to Shuffled STWA Amplitude Ratio

Amplitude [mV]

7/9 6-O HDA

[s]

7/9 Control

7/9 6-OHDA

5/9 Reserpine

Fig. 5. A–F: Effects of dopamine depletion on phase-locking of basal ganglia spiking to simultaneously recorded local beta and gamma LFPs from the STN and SNr. Spike triggered waveform averages (STWA) within STN (A–C) and SNr (D–E) are plotted for each condition, displaying group-averaged values for controls (black), 6-OHDA (red) and reserpine (blue). LFPs were band-pass filtered at 13–20 Hz (A/D), 21–35 Hz (B/E) or at 36–50 Hz (C/F). Bar plots are showing the corresponding average ratio of the unshuffled to shuffled STWA peak-to-trough amplitudes. Results for controls are encoded in black, for 6-OHDA in dark grey and for reserpine in bright grey. Within the bar plots the corresponding number of significantly phase locked animals is indicated in white letters. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

M.H. Beck et al. / Experimental Neurology 286 (2016) 124–136

C

0/9 Cont rol

STN STWA [21-35 Hz ]

[s]

SNr STWA [13-20 Hz]

Amplitude [mV]

Uns huffled to Shuffled STWA Amplitude Ratio

[s]

B

D

STN STWA [13-20 Hz]

Amplitude [mV]

A

M.H. Beck et al. / Experimental Neurology 286 (2016) 124–136

number of animals with spike-low gamma entrainment. The statistical analysis of the mean coupling index also showed elevated amplitudes, but again no significant group differences (Fig. 5F). M1-low gammaLFP activity was neither coupled to STN nor to SNr spiking activity (for details see Supplementary Fig. 2). 4. Discussion 4.1. Importance of excessive beta band oscillations in the pathophysiology of akinesia We found beta oscillations to be elevated in the basal ganglia nuclei STN and SNr of the acute reserpine and the chronic 6-OHDA model of PD. To our knowledge, we are the first to show enhanced beta oscillations are an integral part of the pathophysiology of the reserpine model. This finding provides further support for the hypothesis that abnormally enhanced beta band oscillations are linked to akinesia in PD and other disease conditions accompanied by slow movements, such as drug induced parkinsonism. This conception of a unifying pathophysiology of akinesia is supported by published data demonstrating a drastic increase of beta band synchrony in dystonia patients treated with drugs that are known to provoke parkinsonism (Kuhn et al., 2008). Previously published experimental studies using an acute pharmacologic block of the dopamine receptors D1 and D2, showed seemingly ambiguous results, thus casting doubt on a common pathophysiologic mechanism (Costa et al., 2006; Mallet et al., 2008b; Degos et al., 2009; Dejean et al., 2011). While some studies found evidence for increased beta band oscillations under a D1-D2-receptor blockade (Costa et al., 2006; Dejean et al., 2011), others showed no changes in this regard (Degos et al., 2005; Mallet et al., 2008b). The reason for this difference remains unknown. It can only be speculated that the reported differences can be attributed to different anesthesia regimes used in these studies. Apart from the increased beta amplitudes in basal ganglia LFPs, it has been reported that subthalamic spiking activity is significantly phase locked to beta oscillations in parkinsonian patients (Kuhn et al., 2005; Weinberger et al., 2006). Similar findings are known from the 6-OHDA model of PD (Mallet et al., 2008b; Delaville et al., 2015). Our 6-OHDA data is in line with these reports and shows enhanced phase locking between subthalamic spiking and LFPs in the beta range. We observed similar, but less pronounced results in our acute model. In both models the majority of animals displayed significant low- and high beta phase locking, albeit the 6-OHDA animals showed markedly greater amplitudes of spike triggered waveform averages. Interestingly, the reserpine animals did not display enhanced power or exaggerated cortico-subthalamic coherence levels in the low beta frequency range. Since we found STN was the only target structure showing strong Spike-LFP interactions in the beta frequency band, it can be assumed the STN might be of primary importance for the generation and consolidation of excessive beta oscillatory synchrony. 4.2. The role of the motor cortex for the generation of beta band synchrony Our study revealed a difference in the involvement of the basal ganglia and particularly of motor cortex in acute reserpine induced parkinsonism in comparison to the chronic 6-OHDA PD model. Interestingly, in the reserpine model, the primary motor cortex did not itself demonstrate increased beta power, but displayed enhanced coherence in the beta range with both basal ganglia nuclei. In contrast, the 6-OHDA model exhibited self-contained cortical beta power oscillations as well as exaggerated cortico-nigral beta coherence levels. We also noted a statistical trend for altered cortico-subthalamic beta coherence. Although, this observation is in line with the currently held theory of Parkinson's disease being a network disorder (Hammond et al., 2007; Oswal et al., 2013), it calls into question, if all parts of the loop are equally involved in the generation of exaggerated beta oscillation. In fact, there are a

133

number of theories where beta oscillations are primarily generated in the cortico-basal ganglia loop. Besides the external globus pallidusSTN subnetwork (Plenz and Kital, 1999; Bevan et al., 2002; Mallet et al., 2008a; Holgado et al., 2010) and the striatum (McCarthy et al., 2011), the cortex has also been hypothesized to be a possible generator of pathological beta oscillations (Magill et al., 2001; Li et al., 2014). Theoretically, the cortex seems to be very well suited to generate excessive beta oscillations in PD, because it is known to generate physiological levels of beta oscillations (Alegre et al., 2002; Pfurtscheller et al., 2003; Doyle et al., 2005b) and because of its multiple direct and indirect projections to the basal ganglia. However, our ECoG power data does not support a central role of the cortex for the generation of excessive beta synchrony in parkinsonism, as we cannot demonstrate any changes in beta oscillatory power in our acute model of parkinsonism in the motor cortex. The emergence of autonomous beta power increases in the cortex of the 6-OHDA lesioned animals might be related to adaptive neuroplastic changes that are triggered by beta oscillations generated in the basal ganglia. Although we did not observe an increase in cortical beta power, cortico-subthalamic and cortico-nigral beta coherence levels were markedly elevated in the reserpine (and the 6-OHDA) model. Enhanced cortical-subcortical beta coherence might trigger or contribute to the formation of excessive basal ganglia beta synchrony. Since coherence is viewed as an expression of neuronal communication (Fries, 2005), it can be speculated that pathological synchronization at beta frequencies disturbs information flow between motor cortex and basal ganglia nuclei and thereby generates akinesia. Previously published data further supports this hypothesis as elevated beta coherence levels can be found in patients suffering from PD (Williams et al., 2002; Fogelson et al., 2006; Litvak et al., 2011). Furthermore, it has been reported that dopaminergic medication causes a reduction in cortical-subthalamic beta coherence in patients and also in the 6-OHDA model of PD (Williams et al., 2002; Sharott et al., 2005; Hirschmann et al., 2013). Thus, it can be assumed that increased cortico-basal ganglia beta coherence might plays an important role in the formation of exaggerated beta synchrony. 4.3. Significance of excessive high and low beta synchrony Another noteworthy finding of our current study was the striking difference in the predominant beta peak frequency in power and coherence elevations that we found in the STN and SNr in the two animal models of PD. While we found a low beta increase in 6-OHDA-lesioned animals, there was a pronounced elevation in high beta power and coherence in the reserpine group. This result is especially interesting, as there is evidence for a diverging importance of excessive high and low beta synchrony in the pathophysiology of PD (Priori et al., 2004; Lopez-Azcarate et al., 2010; Little et al., 2013b; Toledo et al., 2014). It has been demonstrated that low beta is more responsive to antiparkinsonian medication than are high beta band oscillations and that it correlates better with clinical motor scores (Priori et al., 2004; Lopez-Azcarate et al., 2010; Hohlefeld et al., 2014). Our data does not refute the importance of different oscillating beta networks in PD, but shows that excessive synchrony in high beta seems to be sufficient for the generation of a high degree of akinesia. The explanation for the difference in peak synchrony might be directly related to the temporal dynamics of the evolution of beta in the two models. In a study on the MPTP primate model of PD, it has been shown that the oscillatory beta band peak frequency might shift into lower frequency ranges as a function of disease severity or disease duration (Connolly et al., 2015). Since all parkinsonian animals in our current study were severely akinetic, regardless of the PD model, our data may favor disease duration to be of critical importance for beta peak frequency although this has not been shown for human data so far. Hypothetically, long-lasting dopamine depletion leads to neuroplastic network adaptions that result in a reduction in beta peak frequency over time. A confounding factor in this regard could be attributed to the fact that we compared a unilateral

134

M.H. Beck et al. / Experimental Neurology 286 (2016) 124–136

with a bilateral model of PD. However, it is not very probable that this difference is the cause for the difference in dominating beta peak frequency, since studies on the bilateral MPTP non-human primate model of PD have demonstrated exaggerated low beta band power increases (Leblois et al., 2007; Moran et al., 2012). Interindividual variance of beta peak frequencies has been reported across PD patients (8–35 Hz), while the activity of individual subjects centers usually around a single main frequency (Stein and Bar-Gad, 2013). Animal models of PD also display considerable differences. In nonhuman primates treated with MPTP the dominating beta frequency is shown to be focused in the low beta range (10–15 Hz; (Leblois et al., 2007, Moran et al., 2012)), while peaks in the whole beta spectrum have been reported for the 6-OHDA model in different experimental settings (12–40 Hz; (Magill et al., 2001, Sharott et al., 2005, Avila et al., 2010, Grieb et al., 2014, Delaville et al., 2015, Javor-Duray et al., 2015)). An important factor in this regard is the influence of anesthesia on beta peak frequency. In general, a high beta peak can be found in awake parkinsonian animals (Avila et al., 2010; Grieb et al., 2014; Javor-Duray et al., 2015), while a low beta peak dominates in anesthetized animals (Magill et al., 2001; Sharott et al., 2005). A direct comparison of awake and urethane anesthetized animals, showed that the anesthesia lowers the beta peak frequency substantially (Brazhnik et al., 2014). 4.4. Limitations All electrophysiological recordings were conducted under urethane anesthesia, which might lower the sensitivity for relevant interactions between motor cortex and basal ganglia. However, it has been repeatedly proven that recordings from the cortico-basal ganglia loop under urethane anesthesia provide findings comparable to data from awake animals and humans (Magill et al., 2006; Mallet et al., 2008b; Brazhnik et al., 2014). In addition, the well-defined brain states that can be found herein, also allow to delimitate highly consistent intervals of local field potential and multi-unit activity recordings, thereby increasing intra- and interindividual comparability. Also, movement related artifacts are dramatically minimized. Nevertheless, future investigations on network changes after short-term and long-term dopamine depletion should be performed in awake animals, as it is possible to directly correlate motor behavior to oscillatory synchrony in the basal ganglia. The conclusion that we can draw from our STN-SNr coherence calculations are inconclusive because basal coherence levels were found to be already high. A likely explanation for this limitation can be found in the anatomical proximity of STN and SNr, which inevitably leads to effects of volume conduction (Kajikawa and Schroeder, 2011). However, a similar trend for enhanced high beta oscillations in reserpine treated rats could still be seen here as in the other reported combinations for cortico-basal ganglia coherency. Another inevitable limitation is the deviating pathophysiology of both of the examined animal models to the one seen in Parkinson's disease. The course of the disease even of our chronic PD model is much shorter than what can be found in human PD patients. Furthermore, pathological characteristics of PD such as Lewy bodies cannot be found in both of the models. However, our animal models share the main hallmark of the disease, i.e. a profound dopaminergic deficit, with human PD patients. Since the loss of dopamine is the major determinant of motor symptoms in PD, and since the dopaminergic deficit seems to be crucial for the development of beta oscillations in PD, we are confident to provide valuable information on the pathophysiology of PD. Differing biochemical characteristics of the acute and the chronic model used in our study should also be noted. The used 6-OHDA-toxin injection into the MFB leads to a chronic degeneration of the A9 and A10 dopaminergic neurons of the substantia nigra pars compacta and the ventral tegmental area developing after only a few days progressively for about 3 weeks (Schwarting and Huston, 1996). The result is an almost complete dopaminergic deafferentation of the striatum, the

nucleus accumbens, the olfactory tubercle and the medial frontal cortex (Schwarting and Huston, 1996). The effect of reserpine on the other hand is mediated by an acute irreversible inhibition of the vesicular monoamine transporter (VMAT2) that leads to subtotal global monoamine depletion, including dopamine, noradrenaline and serotonin (Leao et al., 2015). We cannot exclude that the additional depletion of serotonin and noradrenaline has an effect on LFP synchronization in the cortico-basal ganglia loop. However, the dependence of excessive beta oscillation on a loss of dopamine has been proven repeatedly (Brown et al., 2001; Kuhn et al., 2006b; Mallet et al., 2008b) and it seems very unlikely that our observed phenotype is not mainly dependent on dopamine depletion. In addition, recent evidence suggests that a reduction of serotonin and noradrenaline transmission on top of a chronic 6-OHDA lesion results only in minor changes of firing characteristics in the globus pallidus and in the SNr networks (Delaville et al., 2012). LFP characteristics have not been explored in this regard, yet. For this reason, we think that our data provides further insights on pathophysiological mechanisms in PD. 5. Conclusions In the present study we demonstrate that pathologically enhanced beta oscillatory power and coherence levels are a hallmark of the cortico-basal ganglia network changes in the reserpine and 6-OHDA animal model of Parkinson's disease. Thus, we add evidence in support of a theory on the existence of a common pathophysiology of akinesia in states of low dopamine, independent of the underlying disease pathomechanism. Furthermore, our data questions the role of the primary motor cortex for the initial generation of enhanced beta oscillations in the cortico-basal ganglia network. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.expneurol.2016.10.005. Contributions CvR designed the study. MHB, JKH, JK performed behavioral, electrophysiological and histologic experiments. CvR, MHB and AAK wrote the paper. Funding This study was funded by Deutsche Forschungsgemeinschaft KFO 247 (grant number GE 2629/1-1). Conflict of interest Andrea Kühn: consultancies: Medtronic, St Jude Medical, Boston Scientific. Advisory Boards: Boston Scientific, Medtronic. The authors declare that they have no conflict of interest. Acknowledgements We would like to thank Judith R. Walters and her laboratory for the great support with the STWA calculation and Constance Scharff for the help with antibody staining. We also thank the Deutsche Forschungsgemeinschaft (DFG), KFO 247, for funding our study. References Alegre, M., Labarga, A., Gurtubay, I.G., Iriarte, J., Malanda, A., Artieda, J., 2002. Beta electroencephalograph changes during passive movements: sensory afferences contribute to beta event-related desynchronization in humans. Neurosci. Lett. 331, 29–32. Avila, I., Parr-Brownlie, L.C., Brazhnik, E., Castaneda, E., Bergstrom, D.A., Walters, J.R., 2010. Beta frequency synchronization in basal ganglia output during rest and walk in a hemiparkinsonian rat. Exp. Neurol. 221, 307–319. Berger H (1938) Das Elektrenkephalogramm des Menschen. In: Nova Acta Leopoldina, vol. Bd. 6 (1938/39), pp S.173–309.

M.H. Beck et al. / Experimental Neurology 286 (2016) 124–136 Bevan, M.D., Magill, P.J., Terman, D., Bolam, J.P., Wilson, C.J., 2002. Move to the rhythm: oscillations in the subthalamic nucleus-external globus pallidus network. Trends Neurosci. 25, 525–531. Bezard, E., Przedborski, S., 2011. A tale on animal models of Parkinson's disease. Mov. Disord. 26, 993–1002. Blandini, F., Armentero, M.T., 2012. Animal models of Parkinson's disease. FEBS J. 279, 1156–1166. Brazhnik, E., Novikov, N., McCoy, A.J., Cruz, A.V., Walters, J.R., 2014. Functional correlates of exaggerated oscillatory activity in basal ganglia output in hemiparkinsonian rats. Exp. Neurol. 261, 563–577. Brown, P., 2007. Abnormal oscillatory synchronisation in the motor system leads to impaired movement. Curr. Opin. Neurobiol. 17, 656–664. Brown, P., Oliviero, A., Mazzone, P., Insola, A., Tonali, P., Di Lazzaro, V., 2001. Dopamine dependency of oscillations between subthalamic nucleus and pallidum in Parkinson's disease. J. Neurosci. 21, 1033–1038. Chen, C.C., Lin, W.Y., Chan, H.L., Hsu, Y.T., Tu, P.H., Lee, S.T., Chiou, S.M., Tsai, C.H., Lu, C.S., Brown, P., 2011. Stimulation of the subthalamic region at 20 Hz slows the development of grip force in Parkinson's disease. Exp. Neurol. 231, 91–96. Connolly, A.T., Jensen, A.L., Bello, E.M., Netoff, T.I., Baker, K.B., Johnson, M.D., Vitek, J.L., 2015. Modulations in oscillatory frequency and coupling in globus pallidus with increasing parkinsonian severity. J. Neurosci. 35, 6231–6240. Costa, R.M., Lin, S.C., Sotnikova, T.D., Cyr, M., Gainetdinov, R.R., Caron, M.G., Nicolelis, M.A., 2006. Rapid alterations in corticostriatal ensemble coordination during acute dopamine-dependent motor dysfunction. Neuron 52, 359–369. Degos, B., Deniau, J.M., Thierry, A.M., Glowinski, J., Pezard, L., Maurice, N., 2005. Neuroleptic-induced catalepsy: electrophysiological mechanisms of functional recovery induced by high-frequency stimulation of the subthalamic nucleus. J. Neurosci. 25, 7687–7696. Degos, B., Deniau, J.M., Chavez, M., Maurice, N., 2009. Chronic but not acute dopaminergic transmission interruption promotes a progressive increase in cortical beta frequency synchronization: relationships to vigilance state and akinesia. Cereb. Cortex 19, 1616–1630. Dejean, C., Arbuthnott, G., Wickens, J.R., Le Moine, C., Boraud, T., Hyland, B.I., 2011. Power fluctuations in beta and gamma frequencies in rat globus pallidus: association with specific phases of slow oscillations and differential modulation by dopamine D1 and D2 receptors. J. Neurosci. 31, 6098–6107. Delaville, C., Navailles, S., Benazzouz, A., 2012. Effects of noradrenaline and serotonin depletions on the neuronal activity of globus pallidus and substantia nigra pars reticulata in experimental parkinsonism. Neuroscience 202, 424–433. Delaville, C., McCoy, A.J., Gerber, C.M., Cruz, A.V., Walters, J.R., 2015. Subthalamic nucleus activity in the awake hemiparkinsonian rat: relationships with motor and cognitive networks. J. Neurosci. 35, 6918–6930. Doyle, L.M., Kuhn, A.A., Hariz, M., Kupsch, A., Schneider, G.H., Brown, P., 2005a. Levodopainduced modulation of subthalamic beta oscillations during self-paced movements in patients with Parkinson's disease. Eur. J. Neurosci. 21, 1403–1412. Doyle, L.M., Yarrow, K., Brown, P., 2005b. Lateralization of event-related beta desynchronization in the EEG during pre-cued reaction time tasks. Clin. Neurophysiol. 116, 1879–1888. Eusebio, A., Brown, P., 2009. Synchronisation in the beta frequency-band—the bad boy of parkinsonism or an innocent bystander? Exp. Neurol. 217, 1–3. Eusebio, A., Chen, C.C., Lu, C.S., Lee, S.T., Tsai, C.H., Limousin, P., Hariz, M., Brown, P., 2008. Effects of low-frequency stimulation of the subthalamic nucleus on movement in Parkinson's disease. Exp. Neurol. 209, 125–130. Eusebio, A., Thevathasan, W., Doyle Gaynor, L., Pogosyan, A., Bye, E., Foltynie, T., Zrinzo, L., Ashkan, K., Aziz, T., Brown, P., 2011. Deep brain stimulation can suppress pathological synchronisation in parkinsonian patients. J. Neurol. Neurosurg. Psychiatry. Fogelson, N., Williams, D., Tijssen, M., van Bruggen, G., Speelman, H., Brown, P., 2006. Different functional loops between cerebral cortex and the subthalmic area in Parkinson's disease. Cereb. Cortex 16, 64–75. Fries, P., 2005. A mechanism for cognitive dynamics: neuronal communication through neuronal coherence. Trends Cogn. Sci. 9, 474–480. Fries, P., Reynolds, J.H., Rorie, A.E., Desimone, R., 2001. Modulation of oscillatory neuronal synchronization by selective visual attention. Science 291, 1560–1563. Grieb, B., Engler, G., Sharott, A., von Nicolai, C., Streichert, T., Papageorgiou, I., Schulte, A., Westphal, M., Lamszus, K., Engel, A.K., Moll, C.K., Hamel, W., 2014. High-frequency stimulation of the subthalamic nucleus counteracts cortical expression of major histocompatibility complex genes in a rat model of Parkinson's disease. PLoS One 9, e91663. Halaris, A.E., Freedman, D.X., 1975. Loss of body weight as a predictor of reserpine-induced amine depletion. Eur. J. Pharmacol. 32, 93–101. Hammond, C., Bergman, H., Brown, P., 2007. Pathological synchronization in Parkinson's disease: networks, models and treatments. Trends Neurosci. 30, 357–364. Harnack, D., Meissner, W., Jira, J.A., Winter, C., Morgenstern, R., Kupsch, A., 2008. Placebocontrolled chronic high-frequency stimulation of the subthalamic nucleus preserves dopaminergic nigral neurons in a rat model of progressive Parkinsonism. Exp. Neurol. 210, 257–260. Heeringa, M.J., Abercrombie, E.D., 1995. Biochemistry of somatodendritic dopamine release in substantia nigra: an in vivo comparison with striatal dopamine release. J. Neurochem. 65, 192–200. Hirschmann, J., Ozkurt, T.E., Butz, M., Homburger, M., Elben, S., Hartmann, C.J., Vesper, J., Wojtecki, L., Schnitzler, A., 2013. Differential modulation of STN-cortical and cortico-muscular coherence by movement and levodopa in Parkinson's disease. Neuroimage 68, 203–213. Hohlefeld, F.U., Huchzermeyer, C., Huebl, J., Schneider, G.H., Nolte, G., Brucke, C., Schonecker, T., Kuhn, A.A., Curio, G., Nikulin, V.V., 2013. Functional and effective connectivity in subthalamic local field potential recordings of patients with Parkinson's disease. Neuroscience 250, 320–332.

135

Hohlefeld, F.U., Huchzermeyer, C., Huebl, J., Schneider, G.H., Brucke, C., Schonecker, T., Kuhn, A.A., Curio, G., Nikulin, V.V., 2014. Interhemispheric functional interactions between the subthalamic nuclei of patients with Parkinson's disease. Eur. J. Neurosci. 40, 3273–3283. Holgado, A.J., Terry, J.R., Bogacz, R., 2010. Conditions for the generation of beta oscillations in the subthalamic nucleus-globus pallidus network. J. Neurosci. 30, 12340–12352. Javor-Duray, B.N., Vinck, M., van der Roest, M., Mulder, A.B., Stam, C.J., Berendse, H.W., Voorn, P., 2015. Early-onset cortico-cortical synchronization in the hemiparkinsonian rat model. J. Neurophysiol. 113, 925–936. Kajikawa, Y., Schroeder, C.E., 2011. How local is the local field potential? Neuron 72, 847–858. Kuhn, A.A., Williams, D., Kupsch, A., Limousin, P., Hariz, M., Schneider, G.H., Yarrow, K., Brown, P., 2004. Event-related beta desynchronization in human subthalamic nucleus correlates with motor performance. Brain 127, 735–746. Kuhn, A.A., Trottenberg, T., Kivi, A., Kupsch, A., Schneider, G.H., Brown, P., 2005. The relationship between local field potential and neuronal discharge in the subthalamic nucleus of patients with Parkinson's disease. Exp. Neurol. 194, 212–220. Kuhn, A.A., Doyle, L., Pogosyan, A., Yarrow, K., Kupsch, A., Schneider, G.H., Hariz, M.I., Trottenberg, T., Brown, P., 2006a. Modulation of beta oscillations in the subthalamic area during motor imagery in Parkinson's disease. Brain 129, 695–706. Kuhn, A.A., Kupsch, A., Schneider, G.H., Brown, P., 2006b. Reduction in subthalamic 8– 35 Hz oscillatory activity correlates with clinical improvement in Parkinson's disease. Eur. J. Neurosci. 23, 1956–1960. Kuhn, A.A., Brucke, C., Schneider, G.H., Trottenberg, T., Kivi, A., Kupsch, A., Capelle, H.H., Krauss, J.K., Brown, P., 2008. Increased beta activity in dystonia patients after drug-induced dopamine deficiency. Exp. Neurol. 214, 140–143. Kuhn, A.A., Tsui, A., Aziz, T., Ray, N., Brucke, C., Kupsch, A., Schneider, G.H., Brown, P., 2009. Pathological synchronisation in the subthalamic nucleus of patients with Parkinson's disease relates to both bradykinesia and rigidity. Exp. Neurol. 215, 380–387. Leao, A.H., Sarmento-Silva, A.J., Santos, J.R., Ribeiro, A.M., Silva, R.H., 2015. Molecular, neurochemical, and behavioral hallmarks of reserpine as a model for Parkinson's disease: new perspectives to a long-standing model. Brain Pathol. 25, 377–390. Leblois, A., Meissner, W., Bioulac, B., Gross, C.E., Hansel, D., Boraud, T., 2007. Late emergence of synchronized oscillatory activity in the pallidum during progressive parkinsonism. Eur. J. Neurosci. 26, 1701–1713. Levy, R., Ashby, P., Hutchison, W.D., Lang, A.E., Lozano, A.M., Dostrovsky, J.O., 2002. Dependence of subthalamic nucleus oscillations on movement and dopamine in Parkinson's disease. Brain 125, 1196–1209. Li, Q., Qian, Z.M., Arbuthnott, G.W., Ke, Y., Yung, W.H., 2014. Cortical effects of deep brain stimulation: implications for pathogenesis and treatment of Parkinson disease. JAMA Neurol. 71, 100–103. Little, S., Pogosyan, A., Neal, S., Zavala, B., Zrinzo, L., Hariz, M., Foltynie, T., Limousin, P., Ashkan, K., Fitzgerald, J., Green, A.L., Aziz, T.Z., Brown, P., 2013a. Adaptive deep brain stimulation in advanced Parkinson disease. Ann. Neurol. Little, S., Tan, H., Anzak, A., Pogosyan, A., Kuhn, A., Brown, P., 2013b. Bilateral functional connectivity of the basal ganglia in patients with Parkinson's disease and its modulation by dopaminergic treatment. PLoS One 8, e82762. Little, S., Beudel, M., Zrinzo, L., Foltynie, T., Limousin, P., Hariz, M., Neal, S., Cheeran, B., Cagnan, H., Gratwicke, J., Aziz, T.Z., Pogosyan, A., Brown, P., 2015. Bilateral adaptive deep brain stimulation is effective in Parkinson's disease. J. Neurol. Neurosurg. Psychiatry. Litvak, V., Jha, A., Eusebio, A., Oostenveld, R., Foltynie, T., Limousin, P., Zrinzo, L., Hariz, M.I., Friston, K., Brown, P., 2011. Resting oscillatory cortico-subthalamic connectivity in patients with Parkinson's disease. Brain 134, 359–374. Lopez-Azcarate, J., Tainta, M., Rodriguez-Oroz, M.C., Valencia, M., Gonzalez, R., Guridi, J., Iriarte, J., Obeso, J.A., Artieda, J., Alegre, M., 2010. Coupling between beta and high-frequency activity in the human subthalamic nucleus may be a pathophysiological mechanism in Parkinson's disease. J. Neurosci. 30, 6667–6677. Magill, P.J., Bolam, J.P., Bevan, M.D., 2001. Dopamine regulates the impact of the cerebral cortex on the subthalamic nucleus-globus pallidus network. Neuroscience 106, 313–330. Magill, P.J., Pogosyan, A., Sharott, A., Csicsvari, J., Bolam, J.P., Brown, P., 2006. Changes in functional connectivity within the rat striatopallidal axis during global brain activation in vivo. J. Neurosci. 26, 6318–6329. Mallet, N., Pogosyan, A., Marton, L.F., Bolam, J.P., Brown, P., Magill, P.J., 2008a. Parkinsonian beta oscillations in the external globus pallidus and their relationship with subthalamic nucleus activity. J. Neurosci. 28, 14245–14258. Mallet, N., Pogosyan, A., Sharott, A., Csicsvari, J., Bolam, J.P., Brown, P., Magill, P.J., 2008b. Disrupted dopamine transmission and the emergence of exaggerated beta oscillations in subthalamic nucleus and cerebral cortex. J. Neurosci. 28, 4795–4806. McCarthy, M.M., Moore-Kochlacs, C., Gu, X., Boyden, E.S., Han, X., Kopell, N., 2011. Striatal origin of the pathologic beta oscillations in Parkinson's disease. Proc. Natl. Acad. Sci. U. S. A. 108, 11620–11625. Meredith, G.E., Kang, U.J., 2006. Behavioral models of Parkinson's disease in rodents: a new look at an old problem. Mov. Disord. 21, 1595–1606. Moran, A., Stein, E., Tischler, H., Bar-Gad, I., 2012. Decoupling neuronal oscillations during subthalamic nucleus stimulation in the parkinsonian primate. Neurobiol. Dis. 45, 583–590. Nolte, G., Bai, O., Wheaton, L., Mari, Z., Vorbach, S., Hallett, M., 2004. Identifying true brain interaction from EEG data using the imaginary part of coherency. Clin. Neurophysiol. 115, 2292–2307. Oe, T., Tsukamoto, M., Nagakura, Y., 2010. Reserpine causes biphasic nociceptive sensitivity alteration in conjunction with brain biogenic amine tones in rats. Neuroscience 169, 1860–1871. Olsson, M., Nikkhah, G., Bentlage, C., Bjorklund, A., 1995. Forelimb akinesia in the rat Parkinson model: differential effects of dopamine agonists and nigral transplants as assessed by a new stepping test. J. Neurosci. 15, 3863–3875.

136

M.H. Beck et al. / Experimental Neurology 286 (2016) 124–136

Oswal, A., Brown, P., Litvak, V., 2013. Synchronized neural oscillations and the pathophysiology of Parkinson's disease. Curr. Opin. Neurol. 26, 662–670. Paxinos, G., Watson, C., 2006. The Rat Brain in Stereotactic Coordinates. Academic Press, Sydney. Pfurtscheller, G., Woertz, M., Supp, G., Lopes da Silva, F.H., 2003. Early onset of post-movement beta electroencephalogram synchronization in the supplementary motor area during self-paced finger movement in man. Neurosci. Lett. 339, 111–114. Plenz, D., Kital, S.T., 1999. A basal ganglia pacemaker formed by the subthalamic nucleus and external globus pallidus. Nature 400, 677–682. Priori, A., Foffani, G., Pesenti, A., Tamma, F., Bianchi, A.M., Pellegrini, M., Locatelli, M., Moxon, K.A., Villani, R.M., 2004. Rhythm-specific pharmacological modulation of subthalamic activity in Parkinson's disease. Exp. Neurol. 189, 369–379. Priori, A., Foffani, G., Rossi, L., Marceglia, S., 2013. Adaptive deep brain stimulation (aDBS) controlled by local field potential oscillations. Exp. Neurol. 245, 77–86. Ray, N.J., Jenkinson, N., Wang, S., Holland, P., Brittain, J.S., Joint, C., Stein, J.F., Aziz, T., 2008. Local field potential beta activity in the subthalamic nucleus of patients with Parkinson's disease is associated with improvements in bradykinesia after dopamine and deep brain stimulation. Exp. Neurol. 213, 108–113. Sanberg, P.R., Bunsey, M.D., Giordano, M., Norman, A.B., 1988. The catalepsy test: its ups and downs. Behav. Neurosci. 102, 748–759. Schallert, T., Fleming, S.M., Leasure, J.L., Tillerson, J.L., Bland, S.T., 2000. CNS plasticity and assessment of forelimb sensorimotor outcome in unilateral rat models of stroke, cortical ablation, parkinsonism and spinal cord injury. Neuropharmacology 39, 777–787. Schwarting, R.K., Huston, J.P., 1996. Unilateral 6-hydroxydopamine lesions of mesostriatal dopamine neurons and their physiological sequelae. Prog. Neurobiol. 49, 215–266. Sharott, A., Magill, P.J., Harnack, D., Kupsch, A., Meissner, W., Brown, P., 2005. Dopamine depletion increases the power and coherence of beta-oscillations in the cerebral cortex and subthalamic nucleus of the awake rat. Eur. J. Neurosci. 21, 1413–1422.

Shimamoto, S.A., Ryapolova-Webb, E.S., Ostrem, J.L., Galifianakis, N.B., Miller, K.J., Starr, P.A., 2013. Subthalamic nucleus neurons are synchronized to primary motor cortex local field potentials in Parkinson's disease. J. Neurosci. 33, 7220–7233. Stein, E., Bar-Gad, I., 2013. Beta oscillations in the cortico-basal ganglia loop during parkinsonism. Exp. Neurol. 245, 52–59. Steiner, B., Winter, C., Blumensath, S., Paul, G., Harnack, D., Nikkhah, G., Kupsch, A., 2008. Survival and functional recovery of transplanted human dopaminergic neurons into hemiparkinsonian rats depend on the cannula size of the implantation instrument. J. Neurosci. Methods 169, 128–134. Steriade, M., 2000. Corticothalamic resonance, states of vigilance and mentation. Neuroscience 101, 243–276. Timmermann, L., Fink, G.R., 2011. Pathological network activity in Parkinson's disease: from neural activity and connectivity to causality? Brain 134, 332–334. Toledo, J.B., Lopez-Azcarate, J., Garcia-Garcia, D., Guridi, J., Valencia, M., Artieda, J., Obeso, J., Alegre, M., Rodriguez-Oroz, M., 2014. High beta activity in the subthalamic nucleus and freezing of gait in Parkinson's disease. Neurobiol. Dis. 64, 60–65. Weinberger, M., Mahant, N., Hutchison, W.D., Lozano, A.M., Moro, E., Hodaie, M., Lang, A.E., Dostrovsky, J.O., 2006. Beta oscillatory activity in the subthalamic nucleus and its relation to dopaminergic response in Parkinson's disease. J. Neurophysiol. 96, 3248–3256. Williams, D., Tijssen, M., Van Bruggen, G., Bosch, A., Insola, A., Di Lazzaro, V., Mazzone, P., Oliviero, A., Quartarone, A., Speelman, H., Brown, P., 2002. Dopamine-dependent changes in the functional connectivity between basal ganglia and cerebral cortex in humans. Brain 125, 1558–1569.