1 Erwin B.
1 Department of Neurology and
2 Departments of Neuroscience and
Pharmacology,
montgomery@neurology.wisc.edu
Despite the efficacy of deep brain stimulation (DBS) for an expanding
range of neurological and psychiatric disorders little is known about its
mechanisms of action. We report results
from microelectrode recordings in the motor cortex (mCtx), sensory cortex
(sCtx), globus pallidus externa (GPe) and interna (GPi), and putamen (Pt) in
non-human primates during stimulation of the subthalamic nucleus (STN) in a
manner analogous to human DBS. We
stimulated with different frequencies of regular stimulation and with
pair-pulses to demonstrate resonance effects.
The latter are manifest as increase probability of neuronal discharge to
the second pulse as a function of the time interval from the first pulse. The results demonstrate the DBS activates
neurons in the basal ganglia-thalamic-cortical (BG-Th-Ctx) system. Responses consist of anti-dromic activation
in mCtx and GPe and short latency oligosynaptic activations. There is no difference in the direct effect
(first 8 ms) regardless of stim-ulation frequency (130 pulses per sec [pps],
100 pps or 50 pps). Paired-pulse stimulation demonstrates resonance effects at
short inter-stimulus pulse intervals suggestive of high frequency reentrant
oscillators and consistent with previous demonstration of multiple and high
simultaneous periodic activity in neuronal spike trains in the BG-Th-Ctx
system. The implications of these
findings for theories of physiology and pathophysiology are discussed.
2. METHODS
We completed a
series of experiments utilizing stimulation of the subthalamic nucleus (STN) of
non-human primates in a manner analogous to DBS in humans while conducting
microelectrode recordings of extra-celluar action potentials in motor cortex
(mCtx), sensory cortex (sCtx), putamen (Pt), caudate nucleus (Cn), globus
pallidus interna (GPi) and globus pallidus externa (GPe). We utilized continuous regular stimulation at
different frequencies and a paired-pulse regimen to study the dynamics of the
basal ganglia-thalamic-cortical (BG-Th-Ctx) system.
3. RESULTS
1.
DBS in the vicinity of the STN results in antidromic
activation of neurons in the mCtx (~2 ms) and GPe (~2ms). This is consistent with activation of axons
projecting to or in the vicinity of the STN.
2.
DBS in the vicinity of the STN results in short
latency oligosynaptic activation of neurons in the mCtx (~5 ms), sCtx (~4 ms),
GPi (~ 5ms), Pt (~5.5 ms), and GPe (~6 ms).
It is possible that these later responses in sCtx, GPe, GPi and Pt
represent propagation of the antidromic activations of mCtx to the sCtx and
through the BG-Th-Ctx system. Activation
of GPi could result from stimulation of the STN output neurons to GPi. This is strong evidence that STN DBS does not
inhibit the STN output. These
observations suggest that previous demonstrations of much slower conduction
times, particularly in the cortico-striatal pathways are in error 3.
3.
There is very little difference in the responses to
STN DBS in the first 8 ms to DBS at 130 pulses, 100, or 50 pps. This is strong evidence against the notion
that different DBS frequencies have different direct effects. Namely, high frequency DBS does not cause
inhibition while low frequency DBS causes activation.
4.
After the initial 8 ms response to the DBS pulse, the
neuronal activity appears to return to baseline. Thus, high frequency DBS may suppress baseline
activity by not allowing sufficient time subsequent to each DBS pulse for the
neuron to return to baseline activity.
In the case of disease, the high frequency stimulation prevents the
return to pathological neuronal activity. Alternatively, high frequency
stimulation may drive the BG -Th-Ctx system to regularity, thereby reducing the
misinformation content pathological system.
5.
Paired-pulse stimulation allows demonstration of
resonance effects (increased probability of neuronal discharge) when the second
stimulation of the pair is given when the effects of the first pulse have
traversed the system. The time period
between the pair of pulses associated with the resonance effect is the measure
of the time required to traverse the system or the fundamental frequency of the
reentrant oscillator that comprises the system.
All neurons recorded in the BG-Th-Ctx system demonstrated multiple and
high frequency resonance effects that could not be explained solely on membrane
dynamics associated with post-action potential conductance changes. Several neurons demonstrated harmonic effects
consistent with propagation through a reentrant oscillator. These observations are consistent with
demonstration that all neurons in the BG-Th-Ctx demonstrate multiple and high
frequency periodic activity in their neuronal spike train during baseline
activity 4.
4. DISCUSSION AND CONCLUSIONS
1. DBS does not
inhibit the target structure. This has
been confirmed by large body of evidence including microelectrode recordings in
animals and humans 5-7, regional cerebral blood
flow studies using PET 8, and neurotransmitter
studies 9, 10. A explanation based on mathematical modeling
has been provided 11.
2. Therapeutic STN
DBS increases activity in the GPi.
Consequently, increased activity in the GPi cannot be causally related
to the pathophysiology of Parkinson’s disease contrary to current theory 12, 13. Rather, the demonstrations of increased GPi
activity in animal models of parkinsonism and in human recordings is
epiphenomenal.
3. The effects of
STN DBS are propagated throughout the BG-Th-Ctx suggesting that the therapeutic
effects are a systems phenomena as opposed to a single (or few) structure. This hypothesis is consistent with the
observation DBS applied to a number of structures in BG-Th-Ctx is therapeutic
including the GPi 1, ventrolateral thalamus 14, mCtx 15, STN 1, and GPe 16. Thus, the system or network can be activated
via a number of input points.
4. If DBS exerts its
anti-pathophysiological (therapeutic) effect through the system, then could not
the pathophysiological mechanism be systems-based 17?
5. Alternative
theories of pathophysiology include: abnormalities in dynamic mechanisms
related to behavior not manifest in baseline discharge rate 18 and changes in the pattern
rather than rate 7 19.
6. Additional
alternatives include overwriting of misinformation to no by driving BG-Th-Ctx activity to high
frequency and regular activity 20. Preliminary studies in non-human primates
performing a behavioral task demonstrates that some neurons in the BG-Th-Ctx
system that normally modulate their discharge rates with behavior fail to do so
with high frequency STN DBS while continuing to modulate their activity with
lower frequency DBS. This theory
represents a significant departure from previous concepts of pathophysiology
characterized by one-dimensional push-pull systems of excitation and inhibition
12, 13. Pathophysiology may have more to do with
misinformation rather than loss of a specific neurotransmitter. An analogy would be a “run time” error in
computer programming in otherwise intact computer systems.
7. Demonstration
that the BG-Th-Ctx system can best be described as a large set of non-linear
reentrant oscillators loosely embedded in scale-free networks greatly expands
the range of normal and pathophysiological mechanisms and properties 17. The striking analogy to mathematical 21 and physical systems means
that mechanisms and properties for these systems may be translated into
neurophysiological mechanisms and properties.
This could be the basis for a new theoretical systems neuroscience and
engineering. Further, these implications
might apply to any brain system that can be characterized as reentrant
oscillators including the limbic system and the thalamo-cortical systems.
[1] Deep-Brain Stimulation for Parkinson's
Disease Study, G. Deep-brain stimulation of the subthalamic nucleus or the pars
interna of the globus pallidus in Parkinson's disease.
[2] Olanow, C. W. et al. A double-blind controlled trial of bilateral fetal nigral transplantation in Parkinson's disease. Annals of Neurology 54:403-414, 2003.
[3] Nambu, A., Yoshida, S. & Jinnai, K. Discharge patterns of pallidal neurons with input from various cortical areas during movement in the monkey. Brain Research 519:183-191, 1990.
[4] Gale JT, M. J. E.,
[5]
[6] Baker, K. B.,
[7] Hashimoto, T.; Elder, C. M.; Okun, M. S.; Patrick, S. K.; Vitek, J. L. Stimulation of the subthalamic nucleus changes the firing pattern of pallidal neurons. Journal of Neuroscience 23:1916-1923,2003
[8] Perlmutter, J. et al. Blood flow responses to deep brain stimulation of thalamus. Neurology 58:1388-94, 2002.
[9] Windels, F., Bruet, N., Poupard, A. & Savasta, M. High frequency stimulation of subthalamic nucleus increased extracellular contents of glutamate in substantia nigra and external pallidum 6th International Congress of Parkinson's disease and Movement Disorders. 212, 2000.
[10] Stefani, A. et al. Subthalamic Stimulation Activates Internal Pallidus: Evidence from cGMP Microdialysis in PD Patients. Annals of Neurology 57:448-452, 2005.
[11] McIntyre, C., C., Savasta, M., Walter Benjamin, L. & Vitek Jerrold, L. How Does Deep Brain Stimulation Work? Present Understanding and Future Questions. Journal of Clinical Neurophysiology 21:40-50, 2004.
[12] Albin, R. L., Young, A. & Penny, J. B. The functions anatomy of basal ganglia disorders. Trends in Neuroscience 12, 366-375 (1989).
[13] DeLong, M. R. Primate models of movement disorders of basal ganglia origin. Trends in Neuroscience 13:281-285, 1990.
[14] Koller, W. et al. High-frequency unilateral thalamic stimulation in the treatment of essential and parkinsonian tremor. Annals of Neurology 42:292-299, 1997.
[15] Canavero, S. et al. Therapeutic extradural cortical stimulation for movement disorders: a review
Neurological Research 25:118-122, 2003.
[16] Vitek, J. L., Hashimoto, T., Peoples, J., DeLong, M. R. & Bakay, A. E. Acute stimulation in the external segment of the globus pallidus improves Parkinsonian motor signs. Movement Disorders 19:907-915, 2004.
[17]
[18]
[19]
[20]
[21] Hoppensteadt, F. C. & Izhikevich, E. M. Synaptic organizations and dynamical properties of weakly connected neural oscillators. II. Learning phase information. Biological Cybernetics 75:129-135, 1996.
Acknowledgments
The research reported was supported in part by the American Parkinson Disease Association Advance Center for Research and the Roger Duvosin Fellowship of the American Parkinson Disease Association (EBM) and in part by NIH grant number 5P51 RR000167 to the National Primate Research Center, University of Wisconsin-Madison and NIH grant number 5T32HD07459 (JTG) to Department of Neuroscience, Georgetown University.