Precisely how electricity changes nervous system function can be fully characterized, understood, and then leveraged to design electrical therapies. To understand the biophysics of electric-nervous tissue interactions, we have developed a research program structures along three “tiers” of characterization. The tiers are evidently not independent (the nervous system is not ‘divided’ along these lines), but they provide a rational basis for understanding electric field effects and they lend themselves to specific experimental questions.
This page is intended as a rather technical and historical overview of Quantiative Neurophysiology with Applied Electric Fields as developed in the CCNY Neural Engineering Lab based on innovative work by other groups.
These tiers also provide a basis for designing rational electrotherapies based on “Functional Targeting”. Our premise for functional targeting is that the same brain structures that are responsible for disease (and hence should be changed by stimulation) are also responsible for some normal functions (and so should be spared by stimulation). So while it is important to improve “spatial targeting” namely by guiding current only to specific structures – we propose that a concurrent scheme for functional targeting –namely changing the abnormal process while sparing the normal functions within a specific structure – is also necessary.
You can watch Dr. Bikson discuss Rational Electrotherapy and Functional Targeting at three tiers at the 2008 Neural Interface Conference here as well as follow recent presentations by our group (news). You can also find additional background information on the CCNY Neural Engineering Course page.
Functional Targeting of Electrical Stimulation
1st Tier: Single Cell
In this section we introduce our Single Cell tier (1st tier) for Rational Electrotherapy and Functional Targeting. This introduction is put in a broader historical and technical context, and links with our ongoing projects.
Transmembrane Potential: The starting point for understanding how electrical stimulation effects cells is considering how the electric fields generated by stimulation (e.g. the current flowing around the cells) leads to changes in transmembane potential. These changes in transmembane potential then determine what the outcome of stimulation is: understanding how transmembrane potential changes lead to functional outcomes is complex; none-the-less determining induced transmembrane potential is a necessary first step. There are other potential mechanisms by which stimulation might change neuronal function, such as joule heating, but transmembrane potential is largely the focus of rational electrotherapy. Electroporation results from large change in transmembrane potential so is a sub-phenomenon (e.g. endothelial cells). If transmembrane potential exceeds a threshold value, it may lead to action potential generation [link AP firiing] – but even small “sub-threshold” polarization may profoundly change function [link sub].
Biphasic Polarization and the ‘somatic doctrine’: It’s important to understand that there is no such thing as simply “all depolarizing” or “all hyper-polarizing” stimulation. As shown in the pictures below [add figure], for any stimulation and any cell, a current that cross into the membrane must also cross out – so there is always regions of depolarization and hyperpolarization. Sometimes people focus on just the soma compartment, in which case it is possible to discuss a “somatic depolarizing” of “somatic hyper-polarizing” field. The ‘somatic doctrine’ was useful to explain early animal results with CNS stimulation [link to early in vivo] but ignores the potentially profound effect of dendrite depolarization of cellular functions including synaptic integrations and plasticity [link synapse].
Single Neuron Stimulation Models: There is a significant history of using computational approaches to model the effects of electrical current on neuronal membrane polarization The consensus of these reports is that the magnitude of neuronal polarization is highly sensitive to detailed neuronal geometry, distributed (active) membrane properties, and the profile of the induced extracellular potentials (Ranck 1975; Yim 1986; Tranchina and Nicholson 1986; Vigmond 1997; Holt and Koch 1999; McIntyre and Grill, 1999; Bokil 2001; McIntyre and Grill, 2002; McIntyre 2004).
Computational neuroscientists have employed both analytical approaches using simplified cellular geometries such as spheres and numerical approaches which have recently evolved to include neurons with hundreds of compartments; each compartment loads with a range of voltage gated channels, each channel control by multiple gating kinetics. Whereas the former is limited by simplicity (mathematical tractability), the later can result in such complexity that results are phenomenological.
On our General approach to computational modeling: We believe strongly in the usefulness of models to make testable predictions, equally important is that models are precisely parameterized based on experimental data, including experiments where fully controlled (e.g. uniform [link]) is applied to reduced systems [link]. This emphasis has lead us to define and focus on specific metrics of electrical activation, such as “coupling constant” (see below) which facilitates a tight dialogue between specific models (with no more complexity then needed), quantitative predications, and experiments. We are determining and exploring similar metrics at the network level. This approach is of course not unique to our group, but we have developed an integration of expertise around our own efforts toward computational modeling in support of rational electrotherapy.
Early Animal Studies of Electrical Stimulation: In vivo animal experiments in the 60’s established qualitative relationships between electrical stimulation and CNS neuronal function including showing polarity dependent changes (see somatic doctrine below) and also changes that outlasted the duration of stimulation. However, in our view, significant steps forward in determining the quantitative relationship between electric fields and changes in neuronal function, largely initiated only with the use of isolated preparations: in these isolated preparations is was possible to 1) precisely and reproducibly control the applied currents (including uniform fields); and 2) directly link specific responses to electric fields.
Isolated Preparations: We believe that the majority of insight into how electrical stimulation affects CNS structures has come from experiments using isolated preparations: Indeed, our group continues to rely heavily on in vitro brain slices to test specific hypothesis about electrical modulation. As indicated above, there are two core features of isolated preparation necessary to draw quantitative insight on electric field effects. First, the electrical environment of the cells is precisely and fully controlled, so that we can always indicated and reproduce what we are applying. This is in contrast to in vivo or human stimulation, where even under the best controls, the induced electric fields will depend on subject specific anatomy. Second, isolated preparations allow direst quantification and isolation of stimulation effects. Through a range of electrophysiological (e.g. intracellular, extracellular), imaging (e.g. optical imaging), pharmacological, and other approaches, well worked out in isolated preparations, it is possible to determine in detail the specific pathways by which electric field modulated neurons – such detailed biophysical understanding is a necessary step toward rational electrotherapy. The limitation of isolated preparations will always be just that-the tissue is isolated for the animal, so things like behavior is evidently absent. But in our work, we aim to design experiments to take advantage of this isolation, allowing us to determine precise mechanisms of action, rather then make phenomenological observations.
“Sub-threshold” and “Supra-threshold” electrical stimulation: Sub-threshold (also call neuro-modulation) and supra-threshold fields are terms that have been applied widely, including by our group, to distinguish between a perceived fundamental division between how weak currents verse strong current affect brain function – and hence how distinct electrotherapies should be interpreted and designed. But, especially our work cortical neurons and network oscillations has made us increasingly wary of these popular terms.
These terms are generally interpreted at the single cell (1st tier) level. Supra-threshold stimulation, which is typically applied in (repetitive) pulses, results in action potential generation in the target neurons; in the simplest case in a one-pulse to one-action potential volley fashion. Sub-threshold stimulation nominally indicated stimulation that is applied to an isolated and quiescent neuron will not trigger action potential. [my video]. But neurons in the brain are hardly uncoupled and quiescent. For example during oscillation neurons are constantly near and crossing action potential threshold, in this case a very weak field can cause a neuron to fire; it is not straightforward to the say if this is a sub or supra threshold stimulation. Instead of “sub-threshold” we propose using “weak current” stimulation.
Uniform Fields and ‘Quasi-Uniform’ Assumption: A uniform field indicated an electric field that has only one direction and magnitude in a tissue region of interest; the latter emphasis is important because in reality there is not globally uniform field. Following the ‘quasi-uniform’ assumption, we have proposed that for many stimulation modalities, such as tDCS, the induced electric fields are largely uniform on the scale of each single neuron – for example in one cortical column. Though the electric field will vary across the brain, we can speak of the single electric field magnitude and direction in each region of interest. Though an abstraction, there are several advantages to the ‘quasi-uniform’ assumption in the understanding and design of electrotherapy. One advantage is the use of uniform field in isolated preparations.
Our work on modeling transcranial electrical stimulation is addressing how valid the quasi-uniform assumption is for specific electrotherapies like tDCS.
Adapting a technique applied by John Jefferys in 1981 (pubmed), in 1986, Steven Schiff and Bruce Gluckmanapplied uniform fields in the control of experimental seizures (pubmed), and approach that we and others followed.
The “Coupling Constant” Theory: The coupling constant, also called the ‘coupling strength’ or ‘polarization length’, is a key substrate of our efforts to quantify the effects of stimulation on neurons, relying on experimentally verified hypothesis. The concept is straightforward: We assume that for weak electric fields (stimulation intensities too weak to activation significant voltage gated membrane channels and below action potential threshold, that the membrane potential polarization at any given compartment is a linear of stimulation intensity. Moreover, for a uniform electric fields, the membrane potential polarization can be expressed as:
Equation Vtm = G * E
Weak current “Coupling Constant” Data: Coupling constant quantification for various neuron types have been done using isolated stimulation preparations. Some of the initial work quantifying uniform field coupling constants was done by Christopher Chan around 1986 and 1988, then in the lab of Charles Nicholson. Dr. Chan used very low frequency field and intracellular recording from multiple compartments in turtle cerebellum neurons. This pioneering work was then develoed into an analytical model.
Dr. Marom Bikson, working on the lab of John Jefferys determined coupling constants for rat CA1 neurons (PDF). John Jefferys group went on to determine these values for CA3 cells [pubmed]; values that were used our later CA3 network models (project). Working with Dr. Bikson, Dr. Radman then led our effort to determine coupling constants in morphological reconstructed cortical neurons (project).
Firing Thresholds: For supra-threshold stimulation the traditional focus at the 1st tier is “recruitment order”: or which cell fires first (as stimulation intensity is gradually increased). For example in DBS there has been significant modeling effort to model the first cellular responders. Dr. Radman characterized firing threshold in cortical neurons (project) – one important finding of his work is that neurons that are ‘down-stream’ from first activated neurons will also fire but indirectly due to synaptic input. Evidently in the brain, the synaptic connections are intact and wide-spread such that the concept of firing threshold could be considered in the context of the single neuron and all its afferents.
Small Polarization Big Effects I-Single Neurons: One of the fundamental challenges in rational electrotherapy with weak electric fields is understanding how relatively small polarization (e.g. 1 mV/mm field leading to a 0.1 mV somatic polarization) could have functional effects on CNS function and hence be used in therapy. In addition to quantification of coupling constants, this challenges if understanding how weak current are “transduced” the by CNS has been a focus of our research efforts at the 1st tier. For example, we showed how when responding to depolarizing ‘ramps’ (e.g. a gradual EPSP), the weak polarization induced in fields can be dramatically amplified through a change in firing time– this led us to propose that weak electric field that may be too weak to trigger action potentials or even change action potential threshold, may exert a powerful effect on CNS function through spike timing.
We have investigated additional ‘amplification’ mechanisms at the synapse and network levels, as well as through clinical collaborations.
Beyond Neurons: Glia and Endothelial Cells. In 2006, Dr. Bikson argued that any Rational Electrotherapyparadigm must consider non-neuronal cells in the brain. Indeed, this concept is hardly new, but essentially ignored in most studies of electrical stimulation and medical device design. Work on electrical stimulation on non-neuronal cells also links with our general studies of non-synaptic mechanisms in the CNS.
Glia represent the majority of cell in the CNS – the concept that they are just passive support cells is being increasingly outdated. As such, one might ask could electric fields modulate CNS function, in part, through direct action on glia (indirect modulation of glia following intense neuronal activation is more of a given). One of the original investigations in the role of glia in response to DC stimulation was Dr. Gardner-Medwin (who incidentally was John Jefferys’ PhD mentor).
Endothelial cells form the Blood-Brain Barrier (BBB), which tightly regulate the neuronal environment. Any electrical action on the BBB may thus profoundly impact brain function. We have been collaborating with Dr. John Tarbell (CCNY), and world expert on endothelial barriers, to directly evaluate this hypothesis (project).
Functional Targeting 2nd Tier: Synapses
Our group has developed the concept of “Functional Targeting” [link] at three “tiers” as a neurophysiological substrate for Rational Electrotherapy. Our 2nd tier is direct action of electricity on Synaptic Function. Synapses evidently underlie CNS function and changes in synaptic function (e.g. LTP, LTD) are though to underpin normal (e.g. learning) as well as abnormal (e.g. epilepsy) function. Understanding how applied current change synaptic function is therefore necessary for Rational Electrotherapy. We separately consider the acute (immediate and short lasting) and plastic (long lasting) changes in synaptic function. The latter is particularly relevant for therapies using non-invasive electrotherapy technologies where, because a subject cannot walk around forever with the stimulator, changes should outlast the treatment-period of stimulation.
Micro-Tetanic Stimulation LTP and LTD: The general concept that electrical stimulation can change synaptic function is well born out in the areas of experimental tetanic (high-frequency pulse) supra-threshold stimulation in animal LTP studies – these studies have elegantly characterized the pre- and post-synaptic molecular cascades leading to changes in synaptic efficacy. However, these studies have some key limitations in the context of Rational Electrotherapy namely 1) Electrical stimulation used by many neurophysiologist as an almost ‘mystical’ force that is little understood so applied in a poorly controlled (‘irrational’) manner – essentially the details of stimulation (e.g. monopolar or bipolar) are either ignored or considered irrelevant to outcome; 2) the nature of stimulation applied (e.g. using micro-electrodes) is often not translatable to a clinical setting, nor is that a consideration in these basic science studies.
Uniform currents on acute synaptic function: The use of uniform field in isolated preparations allows precise and reproducible characterization of how applied currents changes synaptic function. Perhaps the first systematic characterization along these lines was by Dr. Jefferys who illustrated that synaptic efficacy in the dentate gyrus (pubmed) could be modulated by weak fields. Dr. Marom Bikson worked with Dr. Jefferys to extend this analysis to the CA1 region of the hippocampus (pdf), some of the most important findings from this work were 1) that the different afferent bundles will be modulated in different ways by the same electric field (e.g. one bundle could be potentiated even as another was depressed); and 2) field were equally effective in modulating synaptic terminals and post-synaptic neurons – so that uniform fields applied perpendicular to the main post-synaptic neuron axis (and hence not significantly polarizing the post-synaptic neuron) but parallel to synaptic fibers (and hence polarizing pre-synaptic terminals) are effective in modulating synaptic function.
Uniform fields of plastic synaptic function: Having established acute effects on synaptic efficacy for electric fields, our attention has recently focused on plastic (lasting) changes following electrical stimulation. We are particularly interested in induction of plasticity by weak current stimulation, which is relatable to non-invasive electrotherapy technologies like tDCS – plastic changes induced by “blasting” the system with supra-threshold pulse trains are investigated elsewhere in the context of LTP. We have ongoing projects in this area focusing on the cortex, an important target for non-invasive stimulation, and the hippocampus, and important target for memory and disease treatment. We believe plastic changes my represent a new level of weak current amplification at our 2nd tier.
Small Currents Big Changes 2: Synapses. One concept we have been developing is that the effects of weak current, when applied over several minutes, is cumulative. Simply put the idea is that a particular current applied for only a short period, and then stopped, will not lead to any lasting changes in synaptic function. But a weak current applied for a long duration (e.g. minutes, hours) will interact with the CNS in such a manner that its effects are “amplified” through changes in synaptic plasticity.
On clinical translation area of this work is tDCS. Clinical studies with tDCS have in fact shows that DC stimulation needs to be applied for several minutes to have lasting effects. We are now directly investigating the mechanisms of action of this in cortical slices [link Asif work]. We already showed in hippocampal slices that application of even high-intensity DC current for short periods does not induce plastic changes (only acute changes) – a reasonable question is then what would happen is the field we applied and “left on” for several minutes.
We are also investigating the induction of plastic synaptic changes by long duration application of AC currents. This is relevant for clinical tACS (ref Paulus) but also in concerns about environmental exposure to AC fields (e.g. power-lines). Though there was an initial suggestion that short duration application of relatively weak AC currents could induce plastic changes (Bawin papers) – this result has not been reproduced (Jefferys paper). Preliminary data from our group suggest that weak AC fields induce cumulative changes in synaptic function over time-scales of 10’s of minutes [link Je Hi project].
Functional Targeting of Electrical Stimulation 3rd tier: Network
Our group has developed the concept of “Functional Targeting” at three “tier” as a neurophysiological substrate for Rational Electrotherapy. Our 3nd tier is actions of electricity that are only manifest at the network level, generally as modulation of coherent endogenous network activity such as oscillations.
This is a very interesting and active area of research (example project on gamma) for several reasons. Changes in network function and oscillations may be closely linked to several normal brain functions (e.g. slow-wave oscillations in sleep and memory consolidation). Moreover, was have observed phenomena under stimulation of coherent (oscillating networks) that are not predicted based solely on single neuron/synapse function, but rather seem to reflect the active and dynamic response of the coupled neuronal network.
Small Fields Big Changes 3: Networks: We have proposed several mechanisms by which new networks may functionally “amplify” the effects of field that appear too weak to modulate single neuron function, especially in the presence on normal brain noise.
Bridging Tiers: From Networks back to Synaptic Plasticity
Bridging Tiers: From Network back to Single Cell Responses
Learn more about Medical Device Development [link] at CCNY BME.