Download Deep Brain Stimulation Does Not Silence Neurons in Subthalamic

J Neurophysiol 103: 962–967, 2010.
First published December 2, 2009; doi:10.1152/jn.00363.2009.
Deep Brain Stimulation Does Not Silence Neurons in Subthalamic Nucleus
in Parkinson’s Patients
Jonathan D. Carlson,1 Daniel R. Cleary,1 Justin S. Cetas,1 Mary M. Heinricher,1,2 and Kim J. Burchiel1
1
Departments of Neurological Surgery and 2Behavioral Neuroscience, Oregon Health and Science University, Portland, Oregon
Submitted 24 April 2009; accepted in final form 28 November 2009
INTRODUCTION
The clinical efficacy of deep brain stimulation (DBS) in the
subthalamic nucleus (STN) for treating many of the symptoms
of Parkinson’s disease is now well documented (Deep Brain
Stimulation for Parkinson’s Disease Study Group 2001;
Weaver et al. 2009). What remains unclear is how local STN
neurons respond to DBS and how this effect is translated into
a clinical benefit (Kringelbach et al. 2007; Montgomery and
Gale 2008).
STN stimulation and the resultant amelioration of Parkinson’s disease symptoms have classically been explained by
postulating that the stimulation produces a functional lesion by
inhibiting surrounding neurons. This could be either through
direct inhibition or by activating inhibitory presynaptic afferent
terminals (Benabid et al. 2009; Filali et al. 2004). The functional lesion theory grew from the idea that overactivity in the
STN and its target, the globus pallidus pars interna, underlies
the motor symptoms of Parkinson’s disease (Albin et al. 1989;
Address for reprint requests and other correspondence: J. D. Carlson, Inland
Neurosurgery and Spine Associates, 105 West 8th Ave., Suite 200, Spokane,
WA 99204 (E-mail: [email protected]).
962
Delong 1990), and was supported by the clinical observation
that lesioning or otherwise inactivating the STN is effective in
treating Parkinson’s disease symptoms (Follett 2000; Levy et
al. 2001; Walter and Vitek 2004). Electrical stimulation was
thus inferred to mimic a lesion by suppressing output from the
STN.
The functional lesion hypothesis received support from
studies in humans and in a primate model of Parkinson’s
disease in which high-frequency stimulation in the STN was
seen to inhibit activity in surrounding cell bodies for periods of
up to several seconds (Filali et al. 2004; Meissner et al. 2005;
Welter et al. 2004). However, an important limitation of these
studies is that the stimulating current was delivered not through
the standard multicontact clinical DBS electrode, but using a
microelectrode or the guide tube as the active electrode. The
current density and distribution with either of the latter methods would be quite different from that resulting from stimulation through the low-impedance DBS electrode (Butson and
McIntyre 2006). The goal of this study was to determine how
STN neurons respond to stimulation delivered using the same
electrode used clinically to deliver DBS. To this end, we
monitored the activity of STN neurons using a microelectrode
adjacent to the DBS electrode during therapeutic stimulation in
patients with idiopathic Parkinson’s disease.
We show here that subsets of STN neurons recorded during
implantation of the DBS electrode display a brief (1–2 ms),
pulse-related inhibition or a change in firing pattern. However,
neurons in this nucleus are not silenced by the DBS. These
findings do not support the idea that DBS gives rise to a
functional lesion of the STN. They are instead more consistent
with an alternative hypothesis, which is that DBS functions
through white matter activation to effect changes in neuronal
activity throughout the basal ganglia-thalamocortical network
(Gradinaru et al. 2009; Grill et al. 2004; Hammond et al. 2008;
Li et al. 2007; Montgomery and Gale 2008).
METHODS
Patients
This research study was approved by the Oregon Health and
Science University Institutational Review Board (2947). Patients with
a planned DBS surgery to treat advanced, medically intractable
Parkinson’s disease were informed of the study protocol and consent
obtained. Patients were off all dopaminergic medications for ⱖ12 h
before surgery. A total of 33 neurons were studied in 11 patients: 1– 4
neurons in each individual.
Operative procedure
STN targets were selected using the FrameLink Stereotactic Linking System version 4.1.9 (Medtronic Navigation, Louisville, CO)
0022-3077/10 $8.00 Copyright © 2010 The American Physiological Society
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.6 on June 18, 2017
Carlson JD, Cleary DR, Cetas JS, Heinricher MM, Burchiel KJ.
Deep brain stimulation does not silence neurons in subthalamic
nucleus in Parkinson’s patients. J Neurophysiol 103: 962–967, 2010.
First published December 2, 2009; doi:10.1152/jn.00363.2009. Two
broad hypotheses have been advanced to explain the clinical efficacy
of deep brain stimulation (DBS) in the subthalamic nucleus (STN) for
treatment of Parkinson’s disease. One is that stimulation inactivates
STN neurons, producing a functional lesion. The other is that electrical stimulation activates the STN output, thus “jamming” pathological
activity in basal ganglia-corticothalamic circuits. Evidence consistent
with both concepts has been adduced from modeling and animal
studies, as well as from recordings in patients. However, the stimulation parameters used in many recording studies have not been well
matched to those used clinically. In this study, we recorded STN
activity in patients with Parkinson’s disease during stimulation delivered through a clinical DBS electrode using standard therapeutic
stimulus parameters. A microelectrode was used to record the firing of
a single STN neuron during DBS (3–5 V, 80 –200 Hz, 90- to 200-␮s
pulses; 33 neurons/11 patients). Firing rate was unchanged during the
stimulus trains, and the recorded neurons did not show prolonged (s)
changes in firing rate on termination of the stimulation. However, a
brief (⬃1 ms), short-latency (6 ms) postpulse inhibition was seen in
10 of 14 neurons analyzed. A subset of neurons displayed altered
firing patterns, with a predominant shift toward random firing. These
data do not support the idea that DBS inactivates the STN and are
instead more consistent with the hypothesis that this stimulation
provides a null signal to basal ganglia-corticothalamic circuitry that
has been altered as part of Parkinson’s disease.
STN NEURONS AND DBS STIMULATION
Microelectrode recording and DBS
Once a neuron was isolated, baseline activity was recorded for
between 30 and 120 s. Stimulation trials were conducted using the
hand-held DBS stimulator (Medtronic 3625 Test Stimulator) through
a temporary lead connected to the DBS electrode. Multiple stimulation trials were conducted at therapeutic stimulus voltages (3–5 V),
pulse widths (90 –200 ␮s), and frequencies (80 –200 Hz). Stimulation
trains lasted ⬃4 –10 s and were separated by 30 – 60 s. Bipolar
stimulation was used in therapeutic stimulation trials, with contact 0
the anode and contact 1 the cathode. In control trials, contact 3 was the
anode and contact 2 was the cathode. This latter contact pair was used
as a control because it was most distant from the microelectrode and
because the geometry of the DBS electrode and its trajectory dictated
that this pair would be outside of the STN. The impedance of the DBS
electrodes was ⬃1 kOhm, based on measurements in rat brain.
Estimated stimulation current thus ranged from 3 to 5 mA.
Analysis
The microelectrode recordings were digitized (15 kHz) and stored
(Alpha Omega). Data files were imported into Spike 2 for analysis
(CED, Cambridge, UK). Waveforms were sorted to extract activity of
a single neuron using automated template-matching followed by
cluster analysis. The spike sort during each DBS trial was verified on
a spike-by-spike basis to ensure accurate detection of the action
potential of interest. Of 58 neurons recorded, 33 were of sufficiently
high quality to insure a single cell was present throughout the
recording (range, 2- to 5-min duration). Cells that drifted in and out of
recordings as a result of brain shift or activity of the awake patient
were excluded. Examples of recorded waveforms as shown in Fig. 2.
With this recording system, the DBS stimulus artifact precluded
spike identification for a period of several milliseconds during and
immediately after each current pulse. Because the amplifier was
saturated for at least a portion of this time, we did not use an artifact
subtraction method (Hashimoto et al. 2002; Montgomery et al. 2005)
and instead opted to work around the stimulus artifact without altering
the analog signal. The duration of signal saturation was determined for
each recording. With a sufficiently short DBS pulse width (90 –100
␮s) and relatively low frequency (no more than 100 Hz), spikes were
visible for over one half of the interval between the DBS stimulus
pulses (Fig. 2, B–D). The ability to detect spikes between DBS
stimulus pulses within a train allowed us to derive firing rate during
the pulse train.
Four different analyses were performed. Firing rate in the 10 s
before onset of each stimulus train was compared with that in the 10 s
after the stimulus train. Spike histograms aligned with the onset and
the offset of the DBS pulse train were generated using the first and last
stimulus artifacts in the train as alignment points. Cell activity was
averaged across all trials and all cells to generate a total cumulative
pre- and posttrain histogram. The averaged firing rates in the 10 s
before and after DBS trains were compared with identify any residual
effect of DBS after the stimulation was terminated.
Second, firing rate during the stimulus train was determined in a
subset of trials in which pulse width was 90 –100 ␮s. Beginning and
end of the stimulus artifact were determined empirically for each
recording by an investigator blinded to cell firing rate, and firing rate
in the intervals between pulses within the train were determined.
Because stimulus trains varied in length (4 –10 s), the central 3 s of
each train were used for analysis. Initial exploration of the neuronal
firing data showed no effect of pulse train duration on firing rate or
pattern, and data from trials with different stimulus durations were
combined in this analysis.
For these trains, a peristimulus time histogram (0.5-ms bin width)
was also generated for each cell around individual pulses in the train.
Bins with ⬎2 SD change in firing rate relative to a prestimulus
baseline were considered significantly different from basal firing.
Finally, the firing pattern of each neuron was classified according to
the methods of Kaneoke and Vitek (1996). A density histogram
(MATLAB, MathWorks, Natick, MA) was calculated based on the
30 s before and immediately after the first stimulus train. The resulting
density histogram distribution was compared with the Poisson theoretical distribution for a randomly firing cell using a ␹2 test for
goodness-of-fit, with P ⬍ 0.05 considered significant. If firing was
significantly different from that expected for random activity, the
skew was evaluated and the cell classified as bursting (right skew) or
tonic (left skew).
RESULTS
FIG. 1. The physical relationship between the microelectrode and deep
brain stimulation (DBS) electrode, enhanced from an intraoperative X-ray and
overdrawn on the estimated outline of the subthalamic nucleus (STN). The
contact surface of the DBS electrode was positioned ⬃1.3 mm posterior to the
neuron being recorded with the microelectrode.
J Neurophysiol • VOL
Based on microelectrode recording depth from the dorsal
surface of the nucleus, cells were distributed throughout the
dorsal-ventral extent of the STN (Fig. 3). In general, cells were
located in the anterior portion of the STN because the record-
103 • FEBRUARY 2010 •
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.6 on June 18, 2017
from T1 and T2 (3-T scan) fused STN images (located 4 mm posterior
to the intercommisural point, 12 mm lateral from the midline, and 4
mm ventral to the intercommisural line). The Leksell stereotactic
frame was used to guide the electrodes to the surgical target and
bilateral burr holes were placed anterior to the coronal suture under
local anesthesia. Patients were awake during surgery without sedatives or narcotics.
A pair of microelectrodes (400 – 800 kOhm; FHC, Bowdoin, ME)
were stereotactically advanced (Alpha Omega Alpha computer-controlled microdrive, Alpha Omega, Alpharetta, GA) through the thalamus and zona incerta to the dorsal border of the STN. One microelectrode was directed toward the surgical target, the other was
centered 2.0 mm anteriorly at the same depth. The dorsal border of the
STN was neurophysiologically defined by a clear transition from
white matter into a field of multicellular neuronal activity with
increased background noise.
On encountering the dorsal surface of the STN with both microelectrodes, the central (posterior) microelectrode was removed and the
DBS electrode (Model 3387, Medtronic) was inserted to the same
depth as the microelectrode, centered on contact 0 (the most distal
contact). Separation between the microelectrode and the DBS electrode surface was estimated to be ⬃1.365 mm (based on the DBS
electrode radius of 0.635 mm). Electrode placement was visualized on
lateral X-rays (Fig. 1). The DBS and microelectrode pair were
advanced in micrometer increments until an STN neuron was identified.
963
964
CARLSON, CLEARY, CETAS, HEINRICHER, AND BURCHIEL
Sorted Spikes
B
Sorted Spikes
A
Analog Data
Analog Data
0.0
0
0.5
1.0
1.5
0
0
Time (s)
0.1
0.2
0.3
0.4
Time (s)
D
2 Hz
100 µs
5V
10 ms
Stimulus
Artifact
Analog Data
Spike
100 Hz
90 µs
5V
10 ms
200 Hz 5 V
100 µs
0
0.01
0.02
0.03
1 ms
Time (s)
ing microelectrode was 2 mm anterior to the standard STN
target.
STN neurons did not show prolonged changes in firing rate
after DBS stimulus trains
We determined firing rate in the 10 s immediately preceding
and immediately after the first stimulus train for all neurons.
For this sample, mean firing rate in the 10-s period immediately preceding stimulation trains was 19.1 ⫾ 2.7 (SD)
spikes/s. Comparison with an equivalent period immediately
after the end of the stimulus trains showed no change in firing
(19.0 ⫾ 1.9 spikes/s; Fig. 4; t-test for correlated means, P ⬎
0.05). Further analysis at a higher temporal resolution showed
that, even in the first 100 ms after stimulus train offset, firing
rate was not different from that before the stimulus (22.0 ⫾ 6.7
spikes/s). Stimulation through the DBS electrode thus did not
have a prolonged excitatory or inhibitory effect on the firing of
surrounding STN neurons.
Activity during DBS stimulus trains
Using short pulse widths (100 ␮s, 100 Hz) in a subpopulation of neurons (n ⫽ 14) allowed us to detect neuronal activity
between DBS stimulus pulses. (A longer pulse width or higher
frequency prevented detection of spikes during the train in
other neurons; see Fig. 2D for representative trace with 200 Hz
stimulation.) Although the stimulus artifact masked potential
neuronal activity for ⱕ4 ms after each pulse, spikes were
clearly detectable between 5 and 10 ms postpulse in all cases.
TN
IC
ro
(m m Top
m)
o
ce
F
0
1.0
2.0
Nu
2
mb
er o
4
fC
6
ells
8
3.0
4.0
5.0
SNr
FIG.
3. The depth of all cells studied based on microdrive measurements was
referenced to the dorsal surface of the STN determined from the microelectrode
recordings. The recorded neurons were distributed throughout the dorso-ventral
extent of the nucleus. Because of the physical arrangement of the microelectrode anterior to the DBS electrode, the cells were recorded from the anterior
portion of the nucleus. IC, internal capsule; SNr, substantia nigra pars reticularis; STN, subthalamic nucleus; ZI, zona incerta.
J Neurophysiol • VOL
After Stim
Before Stim
40
Firing Rate (Hz)
tan
Dis
STN
fS
ZI
30
20
10
0
-9
-7
-5
-3
-1
1
Time (s)
3
5
7
9
0.1 s bins
FIG. 4. Mean firing rate before and after DBS, averaged across all neurons
studied for all trials. Firing rate over the 10-s period after stimulus offset was
not significantly different from that before stimulus delivery.
103 • FEBRUARY 2010 •
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.6 on June 18, 2017
Sorted Spikes
C
FIG. 2. Detectability of action potential waveforms in STN. A: example of spike-sorting based on
waveform characteristics. B: example of spikes extracted from the analog signal during a DBS train.
C: expanded time base shows spikes between stimulus pulses in the train (same neuron as in B). D: the
stimulus artifact changed shape and duration as a
function of the stimulus frequency and pulse width.
DBS with pulse widths of 100 ␮s and ⬃100 Hz
allowed for visible spikes to be detected and sorted
during about one half of the time between DBS
pulses. Note the complete saturation of the amplifier
for several milliseconds during and after each DBS
pulse.
1
-7
-5
-3
-1
Stim
1
3
5
7
9
Time (s)
FIG.
5. Firing rate during DBS (Stim, bar below trace) was not significantly
different from that in the 10 s preceding or after stimulation. Firing during the
central 3 s of each stimulus train was used for analysis. Also, because the
stimulus artifact precluded detection of spikes for a period during each
stimulus duty cycle, firing rate during stimulation was determined using the
proportion of time that spikes were detectable. Mean ⫾ SD, 14 neurons, 1-s
bins.
STN neurons shifted to random firing after DBS
For 31 neurons, activity was classified as bursting, random,
or tonic (Kaneoke and Vitek 1996) based on a 30-s baseline
recording before the DBS. Five of 14 neurons classed as
bursting when first encountered changed to a random pattern of
activity after stimulation through the DBS electrode. Four of
six tonically firing neurons similarly changed to a random
pattern. Those neurons exhibiting a random pattern before
stimulation (n ⫽ 11) generally remained random, although one
neuron shifted to a tonic pattern. No neuron changed from a
random to bursting pattern after the DBS.
DISCUSSION
The principle finding of this study is that electrical stimulation of the STN in Parkinson’s patients using therapeutic
parameters and a DBS electrode does not result in a profound
or long-lasting inhibition of surrounding neurons. Although a
subset of neurons exhibited a brief, pulse-locked inhibition,
this change was not sufficient to alter overall firing rate
significantly, and firing rate immediately after the end of a
stimulus train was comparable to that exhibited before stimulation. However, approximately one half of the neurons originally classified as having bursting or tonic firing shifted to a
random pattern after delivery of the stimulus train.
Our observation that the overall firing rate of STN neurons
was unchanged after application of stimulation trains lasting up
to 10 s through the DBS electrode was surprising in light of
several previous studies of STN activity. Recording STN
discharge during surgery in humans, both Welter et al. (2004)
and Filali et al. (2004) reported suppression of activity for tens
J Neurophysiol • VOL
A
Active DBS Stimulation
3000
0
40
**
0
0
2
4
B
6
8
Time (ms)
Control DBS Stimulation
1000
0
30
0
0
2
4
6
8
Time (ms)
FIG. 6. Example of postpulse inhibition of an STN neuron. A: stimulation
delivered using contacts 1 and 0 (in STN) resulted in brief (1–2 ms),
short-latency (6 ms) inhibition of activity after each pulse in the train.
B: control stimulation using the most proximal contacts (3 cathode, 2 anode)
did not evoke a postpulse inhibition in this neuron. Stimulus: 100-␮s pulses at
100 Hz, 3–5 V. For both A and B, spike times in a 10-ms period aligned with
the beginning of the stimulus artifact (superimposed to show time during
which spikes were not detectable) were used to construct a pulse-aligned
histogram. The spike count in each bin was normalized by the total number of
pulses and bin size to give firing rate. Raster record above each histogram
shows spike times associated with individual pulses.
103 • FEBRUARY 2010 •
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.6 on June 18, 2017
The firing rate during the DBS train was estimated by taking
this duty cycle into account. With this approach, we saw no
change in overall firing rate during the stimulus train (Fig. 5)
compared with that before stimulus onset.
For these neurons, peristimulus time histograms were also
generated around individual pulses comprising the stimulus
trains. In 10 of 14 cells, a suppression of activity (⬎2 SD)
lasting ⬃1 ms with a latency of 6.4 ms from the pulse could be
detected (range 5– 8 ms; Fig. 6A). This time-locked postpulse
inhibition was not observed when stimulation was delivered
using the most distant contacts (3 anode, 2 cathode, 9 mm from
the recording microelectrode; Fig. 6B).
DBS Pulse
-9
DBS Pulse
0
Pulses
10
Firing Rate (Hz)
20
965
of milliseconds after the stimulus train in a significant number
of neurons. In nonhuman primates, Meissner et al. (2005) saw
a substantial decrease in overall firing rate that was accounted
for by a pause in firing lasting several milliseconds after each
stimulus pulse. In rats, Tai et al. (2003) observed inhibition of
adjacent neurons during the stimulus train. Local high-frequency stimulation is also reported to suppress action potentials of tonically firing neurons in STN recorded in vitro
(Garcia et al. 2005; Magarinos-Ascone et al. 2002).
The durations of the stimulus trains used by Welter et al.
(2004) and Filali et al. (2004) were comparable to those
delivered here (20 s and 500 ms, respectively, vs. ⱕ10 s in this
study). However, there is a critical difference between our
study and that earlier work. We used the DBS electrode itself
to stimulate during the recording session, whereas other investigators have used a microelectrode, single-contact macroelectrode or guide tube to deliver current. The electrical fields
generated and neuronal responses evoked by these different
Pulses
30
Firing Rate (Hz)
AVG Firing Rate (Hz)
STN NEURONS AND DBS STIMULATION
966
CARLSON, CLEARY, CETAS, HEINRICHER, AND BURCHIEL
J Neurophysiol • VOL
neurons sampled in the STN in Parkinson’s patients altered
their firing from irregular to bursting after stimulation through
a microelectrode or guide tube (Welter et al. 2004). Although
we did not specifically analyze oscillatory activity in our
sample, our findings are more consistent with reports of reduced abnormal oscillatory activity in the STN and globus
pallidus pars interna in parkinsonian nonhuman primates after
high-frequency stimulation (McCairn and Turner 2009; Meissner et al. 2005). Such oscillations have been linked to the
pathophysiology of Parkinson’s disease (Kühn et al. 2009).
Implications for mechanism through which DBS ameliorates
the motor symptoms of Parkinson’s disease
Overactivity in STN and its target, the globus pallidus pars
interna, has been theorized to underlie the motor symptoms in
Parkinson’s disease (Albin et al. 1989; Delong 1990). Together
with the clinical observation that lesioning or inactivation of
the STN is effective in treating Parkinson’s disease symptoms
(Follett 2000; Levy et al. 2001; Walter and Vitek 2004), this
idea was the basis for the classic hypothesis that DBS mimics
an ablative lesion by suppressing output from the STN. Our
observation that the spontaneous firing rate of STN neurons is
unchanged during and immediately after the DBS stimulus
train fails to support this view.
An alternative explanation for the efficacy of DBS in the
STN is that activation of STN efferents and/or surrounding
white matter results in a regular, high-frequency neuronal
signal that normalizes pathological activity or prevents transmission of such activity through the basal ganglia-thalamocortical network (Grill et al. 2004; Hammond et al. 2008; Montgomery and Gale 2008). Consistent with this view are modeling studies suggesting that therapeutic DBS depolarizes
myelinated fibers, without evoking or inhibiting discharges in
local cell bodies (McIntyre et al. 2004a,b; Miocinovic et al.
2006). Experimental support for the idea that DBS modifies the
STN outflow comes from the observation that therapeutic
stimulation in the STN in humans and animal models alters
activity in target structures including the substantia nigra pars
reticularis, pallidum, and motor thalamus (Galati et al. 2006;
Hahn et al. 2008; Hashimoto et al. 2003; Kita et al. 2005; Li et
al. 2007; Maltete et al. 2007; Maurice et al. 2003; Xu et al.
2008). We have no direct evidence from our recordings that
axons of STN neurons or adjacent white matter were activated
by the DBS in our patients. Nevertheless, the observation that
at least some neurons that originally exhibited bursting or tonic
firing shifted to a more random pattern after stimulation would
corroborate the concept that DBS suppresses abnormal oscillations in basal ganglia networks controlling motor output
(Meissner et al. 2005). Furthermore, such changes in firing
pattern would be consistent with a shift in the dynamics of the
basal ganglia-thalamocortical network as an integrated system.
In conclusion, the data we obtained in patients during
implantation of a DBS electrode do not provide support for the
classic view that DBS in the STN produces its therapeutic
effect by immediate inhibition of surrounding neurons. Our
findings are instead more consistent with the alternative proposal that DBS synchronously activates surrounding myelinated fibers, thus normalizing or “jamming” the pathological
activity in the basal ganglia-thalamocortical network.
103 • FEBRUARY 2010 •
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.6 on June 18, 2017
electrodes would presumably be quite different. The electrical
field generated by the clinical DBS electrode has been extensively modeled (Butson and McIntyre 2006; Miocinovic et al.
2006). As a rough approximation for comparison with guide
tube or microelectrode stimulation, one can consider that the
contact area of a standard DBS electrode is near 6 mm2
(diameter, 1.27 mm; length, 1.5 mm). With typical impedances
around 1 kOhm and stimulation voltages around 3– 4 V, this
contact area leads to static (DC) current densities of 0.5
mA/mm2. Most guide tubes used for macrostimulation have a
surface area in the range of 0.5–3 mm2 (diameter, 0.1– 0.7 mm;
length, 0.5–1.5 mm). Stimulation is often performed with
constant current sources ranging from 1 to 5 mA. These
parameters would result in significantly higher current densities, ranging from 3 to 10 mA/mm2. Estimation of the current
density surrounding a microelectrode tip is more challenging,
but currents in the range of 50 –500 ␮A are used, which places
the current density ⱖ1–2 orders of magnitude higher than that
obtained using the larger, lower-impedance DBS electrode.
Given that the particular neural element(s) activated is a
function of current density (Ranck 1975; Rattay and Aberham
1993), extrapolation from the effects of a microelectrode to a
DBS electrode stimulation may be erroneous. The strength of
this study is that cell activity was monitored during stimulation
using the clinical DBS electrode and typical therapeutic parameters.
Another factor that may be important is correction for
stimulus artifact, at least when considering firing during the
stimulus train. The overall duration of the artifact associated
with each pulse is related to the amplifier filter characteristics
and input impedance as well as the stimulus pulse width and
current density. The stimulus artifact in this study, as in earlier
studies, was quite prominent, and amplifiers were typically
saturated for 1–2 ms with each pulse in the train. This precludes knowledge of potential spikes during this time, even
with sophisticated stimulus artifact subtraction algorithms
(Hashimoto et al. 2002; Montgomery et al. 2005). We cannot
therefore rule out the possibility that pulse-locked spikes were
obscured or that the neuron under study was silenced for the
duration of the artifact. In the latter case, a decrease in firing
during the stimulus train of ⱕ40% could have gone undetected.
However, there was no suppression of activity immediately
after termination of the stimulus train. Moreover, our observation of continued activity during the DBS stimulus train is
consistent with local field potential recordings where beta-band
and higher-frequency oscillations continue to be elevated during and immediately after DBS (Bronte-Stewart et al. 2009;
Priori et al. 2006; Rossi et al. 2008).
The basis for the brief (⬃1 ms), short-latency (⬃6 ms),
pulse-locked pause in firing seen in a subset of neurons could
not be determined under our experimental conditions. One
possibility is that anterograde or retrograde activation of STN
efferents or fibers of passage sets up a pulse-locked reverberation through the basal-ganglia-thalamocortical network. Reports of time-locked, short-latency responses in the globus
pallidus (Hashimoto et al. 2003; Kita et al. 2005) and the
substantia nigra (Galati et al. 2006) after STN stimulation
would support such a mechanism.
We also observed a change in the firing of some bursting and
tonic neurons to a random pattern. This observation is in
contrast with a previous study in which almost a third of the
STN NEURONS AND DBS STIMULATION
REFERENCES
J Neurophysiol • VOL
Li S, Arbuthnott GW, Jutras MJ, Goldberg JA, Jaeger D. Resonant
antidromic cortical circuit activation as a consequence of high-frequency
subthalamic deep-brain stimulation. J Neurophysiol 98: 3525–3537, 2007.
Magarinos-Ascone C, Pazo JH, Macadar O, Buno W. High-frequency
stimulation of the subthalamic nucleus silences subthalamic neurons: a
possible cellular mechanism in Parkinson’s disease. Neuroscience 115:
1109 –1117, 2002.
Maltete D, Jodoin N, Karachi C, Houeto JL, Navarro S, Cornu P, Agid Y,
Welter ML. Subthalamic stimulation and neuronal activity in the substantia
nigra in Parkinson’s disease. J Neurophysiol 97: 4017– 4022, 2007.
Maurice N, Thierry AM, Glowinski J, Deniau JM. Spontaneous and evoked
activity of substantia nigra pars reticulata neurons during high-frequency
stimulation of the subthalamic nucleus. J Neurosci 23: 9929 –9936, 2003.
McCairn KW, Turner RS. Deep brain stimulation of the globus pallidus
internus in the Parkinsonian primate: local entrainment and suppression of
low-frequency oscillations. J Neurophysiol 101: 1941–1960, 2009.
McIntyre CC, Grill WM, Sherman DL, Thakor NV. Cellular effects of deep
brain stimulation: model-based analysis of activation and inhibition. J Neurophysiol 91: 1457–1469, 2004a.
McIntyre CC, Mori S, Sherman DL, Thakor NV, Vitek JL. Electric field
and stimulating influence generated by deep brain stimulation of the subthalamic nucleus. Clin Neurophysiol 115: 589 –595, 2004b.
Meissner W, Leblois A, Hansel D, Bioulac B, Gross CE, Benazzouz A,
Boraud T. Subthalamic high frequency stimulation resets subthalamic firing
and reduces abnormal oscillations. Brain 128: 2372–2382, 2005.
Miocinovic S, Parent M, Butson CR, Hahn PJ, Russo GS, Vitek JL,
McIntyre CC. Computational analysis of subthalamic nucleus and lenticular fasciculus activation during therapeutic deep brain stimulation. J Neurophysiol 96: 1569 –1580, 2006.
Montgomery EB, Gale JT. Mechanisms of action of deep brain stimulation
(DBS). Neurosci Biobehav Rev 32: 388 – 407, 2008.
Montgomery EB Jr, Gale JT, Huang H. Methods for isolating extracellular
action potentials and removing stimulus artifacts from microelectrode recordings of neurons requiring minimal operator intervention. J Neurosci
Methods 144: 107–125, 2005.
Priori A, Ardolino G, Marceglia S, Mrakic-Sposta S, Locatelli M, Tamma
F, Rossi L, Foffani G. Low-frequency subthalamic oscillations increase
after deep brain stimulation in Parkinson’s disease. Brain Res Bull 71:
149 –154, 2006.
Ranck JB Jr. Which elements are excited in electrical stimulation of mammalian central nervous system: a review. Brain Res 98: 417– 440, 1975.
Rattay F, Aberham M. Modeling axon membranes for functional electrical
stimulation. IEEE Trans Biomed Eng 40: 1201–1209, 1993.
Rossi L, Marceglia S, Foffani G, Cogiamanian F, Tamma F, Rampini P,
Barbieri S, Bracchi F, Priori A. Subthalamic local field potential oscillations during ongoing deep brain stimulation in Parkinson’s disease. Brain
Res Bull 76: 512–521, 2008.
Tai CH, Boraud T, Bezard E, Bioulac B, Gross C, Benazzouz A. Electrophysiological and metabolic evidence that high-frequency stimulation of the
subthalamic nucleus bridles neuronal activity in the subthalamic nucleus and
the substantia nigra reticulata. FASEB J 17: 1820 –1830, 2003.
Walter BL, Vitek JL. Surgical treatment for Parkinson’s disease. Lancet
Neurol 3: 719 –728, 2004.
Weaver FM, Follett K, Stern M, Hur K, Harris C, Marks WJ Jr, Rothlind
J, Sagher O, Reda D, Moy CS, Pahwa R, Burchiel K, Hogarth P, Lai
EC, Duda JE, Holloway K, Samii A, Horn S, Bronstein J, Stoner G,
Heemskerk J, Huang GD. Bilateral deep brain stimulation vs best medical
therapy for patients with advanced Parkinson disease: a randomized controlled trial. J Am Med Assoc 301: 63–73, 2009.
Welter ML, Houeto JL, Bonnet AM, Bejjani PB, Mesnage V, Dormont D,
Navarro S, Cornu P, Agid Y, Pidoux B. Effects of high-frequency
stimulation on subthalamic neuronal activity in parkinsonian patients. Arch
Neurol 61: 89 –96, 2004.
Xu W, Russo GS, Hashimoto T, Zhang J, Vitek JL. Subthalamic nucleus
stimulation modulates thalamic neuronal activity. J Neurosci 28: 11916 –
11924, 2008.
103 • FEBRUARY 2010 •
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.6 on June 18, 2017
Albin RL, Young A, Penny JB. The functional anatomy of basal ganglia
disorders. Trends Neurosci 12: 366 –375, 1989.
Benabid AL, Chabardes S, Mitrofanis J, Pollak P. Deep brain stimulation
of the subthalamic nucleus for the treatment of Parkinson’s disease. Lancet
Neurol 8: 67– 81, 2009.
Bronte-Stewart H, Barberini C, Koop MM, Hill BC, Henderson JM,
Wingeier B. The STN beta-band profile in Parkinson’s disease is stationary
and shows prolonged attenuation after deep brain stimulation. Exp Neurol
215: 20 –28, 2009.
Butson CR, McIntyre CC. Role of electrode design on the volume of tissue
activated during deep brain stimulation. J Neural Eng 3: 1– 8, 2006.
Deep Brain Stimulation for Parkinson’s Disease Study Group. Deep-brain
stimulation of the subthalamic nucleus or the pars interna of the globus
pallidus in Parkinson’s disease. N Engl J Med 345: 956 –963, 2001.
Delong MR. Primate models of movement disorders of basal ganglia origin.
Trends Neurosci 13: 281–285, 1990.
Filali M, Hutchison WD, Palter VN, Lozano AM, Dostrovsky JO. Stimulation-induced inhibition of neuronal firing in human subthalamic nucleus.
Exp Brain Res 156: 274 –281, 2004.
Follett KA. The surgical treatment of Parkinson’s disease. Annu Rev Med 51:
135–147, 2000.
Galati S, Mazzone P, Fedele E, Pisani A, Peppe A, Pierantozzi M, Brusa
L, Tropepi D, Moschella V, Raiteri M, Stanzione P, Bernardi G, Stefani
A. Biochemical and electrophysiological changes of substantia nigra pars
reticulata driven by subthalamic stimulation in patients with Parkinson’s
disease. Eur J Neurosci 23: 2923–2928, 2006.
Garcia L, D’Alessandro G, Bioulac B, Hammond C. High-frequency
stimulation in Parkinson’s disease: more or less? Trends Neurosci 28:
209 –216, 2005.
Gradinaru V, Mogri M, Thompson KR, Henderson JM, Deisseroth K.
Optical deconstruction of Parkinsonian neural circuitry. Science 324: 354 –
359, 2009.
Grill WM, Snyder AN, Miocinovic S. Deep brain stimulation creates an
informational lesion of the stimulated nucleus. Neuroreport 15: 1137–1140,
2004.
Hahn PJ, Russo GS, Hashimoto T, Miocinovic S, Xu W, McIntyre CC,
Vitek JL. Pallidal burst activity during therapeutic deep brain stimulation.
Exp Neurol 211: 243–251, 2008.
Hammond C, Ammari R, Bioulac B, Garcia L. Latest view on the mechanism of action of deep brain stimulation. Mov Disord 23: 2111–2121, 2008.
Hashimoto T, Elder CM, Okun MS, Patrick SK, Vitek JL. Stimulation of
the subthalamic nucleus changes the firing pattern of pallidal neurons.
J Neurosci 23: 1916 –1923, 2003.
Hashimoto T, Elder CM, Vitek JL. A template subtraction method for
stimulus artifact removal in high-frequency deep brain stimulation. J Neurosci Methods 113: 181–186, 2002.
Kaneoke Y, Vitek JL. Burst and oscillation as disparate neuronal properties.
J Neurosci Methods 68: 211–223, 1996.
Kita H, Tachibana Y, Nambu A, Chiken S. Balance of monosynaptic
excitatory and disynaptic inhibitory responses of the globus pallidus induced
after stimulation of the subthalamic nucleus in the monkey. J Neurosci 25:
8611– 8619, 2005.
Kringelbach ML, Jenkinson N, Owen SL, Aziz TZ. Translational principles
of deep brain stimulation. Nat Rev Neurosci 8: 623– 635, 2007.
Kühn AA, Tsui A, Aziz T, Ray N, Brücke C, Kupsch A, Schneider G-H,
Brown P. Pathological synchronisation in the subthalamic nucleus of
patients with Parkinson’s disease relates to both bradykinesia and rigidity.
Exp Neurol 215: 380 –387, 2009.
Levy R, Lang AE, Dostrovsky JO, Pahapill P, Romas J, Saint-Cyr J,
Hutchison WD, Lozano AM. Lidocaine and muscimol microinjections in
subthalamic nucleus reverse Parkinsonian symptoms. Brain 124: 2105–
2118, 2001.
967