Available online at www.sciencedirect.com
ScienceDirect Ephaptic coupling to endogenous electric field activity: why bother? Costas A Anastassiou and Christof Koch There has been a revived interest in the impact of electric fields on neurons and networks. Here, we discuss recent advances in our understanding of how endogenous and externally imposed electric fields impact brain function at different spatial (from synapses to single neurons and neural networks) and temporal scales (from milliseconds to seconds). How such ephaptic effects are mediated and manifested in the brain remains a mystery. We argue that it is both possible (based on available technologies) and worthwhile to vigorously pursue such research as it has significant implications on our understanding of brain processing and for translational neuroscience. Addresses Allen Institute for Brain Science, Seattle, WA 98103, USA Corresponding author: Anastassiou, Costas A ([email protected])
Current Opinion in Neurobiology 2015, 31:95–103 This review comes from a themed issue on Brain rhythms and dynamic coordination Edited by Gyo¨rgy Buzsa´ki and Walter Freeman
http://dx.doi.org/10.1016/j.conb.2014.09.002 0959-4388/# 2014 Elsevier Ltd. All right reserved.
Introduction Charge transfer across the membrane of all structures in brain matter such as neurons, glial cells, etc. induces extracellular current sinks and sources that, in turn, give rise to an extracellular field, that is, a spatial gradient of the extracellular voltage (Ve) measured in comparison to a distant reference signal. On the other hand, the neural membrane is affected by the presence of extracellular fields. A crude example is the strong entrainment of the transmembrane voltage (Vm) and, eventually, of spiking, observed during and/or immediately after electric stimulation, for instance, in electric microstimulation studies (entraining one or a few neurons [1,2]) or deep brain stimulation (DBS) therapy (entraining neural circuits [3,4]). Defined as the difference between the intracellular (Vi) and extracellular voltage (Ve) at a given time t and location x, the membrane voltage Vm(x,t) = Vi(x,t) Ve(x,t) will be affected by changes in Ve(x,t), that is, the presence of extracellular fields. www.sciencedirect.com
Does the dynamic, endogenous electric field in the brain serve a functional role by altering the functioning of neurons? Can such spatiotemporal Ve-fluctuations influence the activity of neural functioning via so-called ephaptic coupling1? Such coupling would constitute a ‘feedback’ mechanism with electric fields altering the activity of the same neural elements that gave rise to them in the first place. Addressing whether ephaptic coupling plays a functional role is not straightforward. Since the extracellular electric field is always present in the living brain — caused by cells functioning — it is challenging to study such functioning in the absence of Ve-activity as this would imply a comatose or a deceased brain [8]. A more tractable question is: which neural components are sensitive to extracellular fields? In the first section, we employ a bottom-up approach, describing ephaptic effects in synapses (microscale), individual neurons (mesoscale) and neural networks (macroscale). In the second section, we discuss temporal aspects of ephaptic coupling. Importantly, while the primary focus of this overview is in ephaptic coupling via electric fields induced by endogenous activity, study of such effects inevitably involves experiments with externally imposed fields. While externally imposed fields lack many of the features of the endogenous field, such studies have greatly enhanced our understanding of ephaptic coupling and have had an impact in medicine (see ‘Applications and translational neuroscience’).
Synapses (microscale) For chemical synapses, a linear relationship is observed between synaptic current Isyn and Vm (with the important exception of the NMDA receptor). Once the reversal potential (Erev) has been accounted for, synaptic currents, to a first approximation, can be described as Isyn(t) = gsyn(t)(Vm(t) Erev) with gsyn being the synaptic conductance. Isyn depends on Vm, which, in turn, depends on Ve. Thus, changes in Ve will influence synaptic currents, though such changes are expected to be rather weak (though potentially important at the network level [9]). 1
The term ‘ephapse’ from the Greek word (i.e., the touch or junction) was coined by Arvanitaki [5] to describe electric field interactions occurring between juxtaposed axons. Over the years researchers have used alternative terms such as ‘electric field effects’ to describe the same phenomenon. To date, the most detailed review about ephaptic coupling remains the seminal review by Jefferys [6]. For a more recent review see Weiss and Faber [7]. Current Opinion in Neurobiology 2015, 31:95–103
96 Brain rhythms and dynamic coordination
A stronger effect may be exerted by the impact of the electric field on electrodiffusion of charged ions. Specifically, ionic fluxes within the synaptic cleft (from presynaptic to postsynaptic neuron) via electrodiffusion induce electric fields locally, within the synapse. Intrasynaptic electric fields, in turn, can alter the membrane potential of the presynaptic terminal leading to changes in the ionic flux. Such events alter, for example, Ca2+-dynamics in the vertebrate retina by depolarizing the presynaptic terminal and increasing neurotransmitter release [10]. Similar events have been hypothesized to occur in cortex [11,12]. Finally, to the extent that electric fields can alter spiking of neurons (see next section), correlations in spiking between two synaptically connected neurons can affect the synaptic strength of their connection by inducing plasticity [13] — this would constitute an indirect impact of ephaptic coupling on synaptic functioning (e.g., [14,15,16]).
Neurons (mesoscale) Extracellular fields along the cable-like structure of neural morphology can be thought of as effective intracellular current injections, leading to alterations in their membrane potential [17,18,19,20,21,22]. The specifics of such entrainment depend on the morphological and physiological features of the neuron as well as the characteristics of the extracellular field — passive cable theory dictates that the impact the field has on Vm depends on the cable-field alignment as well as on three length constants: the characteristic length of the external field, the cable length, and the electrotonic length [22]. The range of Vm induced by a spatially inhomogeneous extracellular field can be of the order of the extracellular spatial voltage oscillation (Figure 1a). For the induced Vm-amplitude to become of the order of the amplitude of Ve, two independent conditions must hold: the length of the cable must be larger than the characteristic length of the external field and the electrotonic length must be larger than the characteristic length of the external field. If the cable length is considerably smaller than the characteristic length of the field, the Ve-gradient along the cable becomes insignificant and Vm remains broadly unaltered. The same applies when the electrotonic length is much smaller than the characteristic field length: for a decreasing cable diameter (i.e., decreasing electrotonic length [23]), the impact of the spatially inhomogeneous extracellular Ve on Vm decreases until the amplitude of the Vm-deviation along the cable becomes zero. Thus, for thin fibers such as axons, the induced Vm will be very small. Ephaptic coupling of Vm to electric fields has an impact on the active conductances of a cell and, thus, to spiking (Figure 1b). Such effects have been studied by applying extracellular fields across brain slices and whole-cell patching individual neurons [24,25,26,27,28] with a focus on the entrainment of spiking to relatively weak and temporally slow (less than 60 Hz) fields emulating local field potential (LFP)-activity in the brain. Changes Current Opinion in Neurobiology 2015, 31:95–103
in somatic potential of 70 mV, that is, below the noise levels of neurons [29], can significantly impact spiking and, in particular, the spike phase (with reference to the oscillatory extracellular signal). Importantly, while most studies have focused on the impact of electric fields on excitatory neurons, the same effects impact subthreshold and spiking of inhibitory neurons [27]. The usefulness of such studies for understanding ephaptic coupling to endogenous fields is limited–chiefly, the cases emulated in slice oversimplify in vivo activity where neurons are continuously bombarded by hundreds of postsynaptic currents along their intricate morphology in the presence of a spatially inhomogeneous and temporally dynamic electric field (Figure 1c; compare to fields in Figure 1a,b). Such limitations are present both for fields induced across parallel plates positioned millimeters away from each other (e.g., [24,25,30]) as well as fields elicited via stimulation pipettes (e.g., [1,28]). To account for the impact of endogenous fields on single neurons, both the intracellular and extracellular voltage would not only need to be monitored along a single cell but also manipulated, and all this in the behaving animal. Such experiments are now becoming feasible. For example, Ozen and colleagues [31] performed simultaneous extracellular and intracellular recordings in vivo while imposing weak electric fields via transcranial electrodes and found that the impact of the stimulation on neurons did not only depend on electrode placement or stimulation details but also to the behavioral state of the animal. Notably, while the extracellular depth recordings were not from the vicinity (within, say, 50 mm) of the impaled (via a sharp electrode) neuron, such work can serve as a precursor for future ephaptic coupling studies, ideally, combining local field manipulation, for example, via light stimulation.
Networks (macroscale) Two approaches have been employed to address network-level effects of ephaptic coupling: first, large-scale electric stimulation of slices and entire brains of living animals and humans, and second, computer simulations. Electric stimulation of brain slices has been the primary method of characterizing biophysical aspects of electric field effects at the network level. For example, we know that altering the field strength (but not the frequency) imposed via parallel plates, slice oscillation frequency progressively shifts to that of the imposed field (or a subharmonic) [26]. Similarly, when imposing different stimulus frequencies, slice dynamics become progressively entrained [30] (Figure 2a,b). Importantly, parallel plate, whole-slice stimulation has shown that emergent properties of neural networks are more sensitive to fields than single neurons [24,26]. Entrainment of spiking to field strengths as low as 0.5 mV/mm [25,28] (Figure 1a,b) firmly suggests that ephaptic entrainment www.sciencedirect.com
Ephaptic coupling to brain fields Anastassiou and Koch 97
Figure 1
(b)
Ve (mV)
10 Hz 20 Hz 50 Hz 60 Hz 100 Hz
2.5
1
0.5
1
1s
I0 = 50 nA
I0 = 100 nA
I0 = 200 nA
90°
90°
90°
90°
0.04
1.0 180°
0
4
8
12
0°
10 Hz
20 Hz
50 Hz
0.04 180°
0°
0.04 0°
180°
0.04 0°
180°
270°
270°
270°
N=2669
N=2639
N=2806
N=2907
N=2783, p=2.8 10–3
N=2709, p=6.7 10–5
N=2856, p
ScienceDirect Ephaptic coupling to endogenous electric field activity: why bother? Costas A Anastassiou and Christof Koch There has been a revived interest in the impact of electric fields on neurons and networks. Here, we discuss recent advances in our understanding of how endogenous and externally imposed electric fields impact brain function at different spatial (from synapses to single neurons and neural networks) and temporal scales (from milliseconds to seconds). How such ephaptic effects are mediated and manifested in the brain remains a mystery. We argue that it is both possible (based on available technologies) and worthwhile to vigorously pursue such research as it has significant implications on our understanding of brain processing and for translational neuroscience. Addresses Allen Institute for Brain Science, Seattle, WA 98103, USA Corresponding author: Anastassiou, Costas A ([email protected])
Current Opinion in Neurobiology 2015, 31:95–103 This review comes from a themed issue on Brain rhythms and dynamic coordination Edited by Gyo¨rgy Buzsa´ki and Walter Freeman
http://dx.doi.org/10.1016/j.conb.2014.09.002 0959-4388/# 2014 Elsevier Ltd. All right reserved.
Introduction Charge transfer across the membrane of all structures in brain matter such as neurons, glial cells, etc. induces extracellular current sinks and sources that, in turn, give rise to an extracellular field, that is, a spatial gradient of the extracellular voltage (Ve) measured in comparison to a distant reference signal. On the other hand, the neural membrane is affected by the presence of extracellular fields. A crude example is the strong entrainment of the transmembrane voltage (Vm) and, eventually, of spiking, observed during and/or immediately after electric stimulation, for instance, in electric microstimulation studies (entraining one or a few neurons [1,2]) or deep brain stimulation (DBS) therapy (entraining neural circuits [3,4]). Defined as the difference between the intracellular (Vi) and extracellular voltage (Ve) at a given time t and location x, the membrane voltage Vm(x,t) = Vi(x,t) Ve(x,t) will be affected by changes in Ve(x,t), that is, the presence of extracellular fields. www.sciencedirect.com
Does the dynamic, endogenous electric field in the brain serve a functional role by altering the functioning of neurons? Can such spatiotemporal Ve-fluctuations influence the activity of neural functioning via so-called ephaptic coupling1? Such coupling would constitute a ‘feedback’ mechanism with electric fields altering the activity of the same neural elements that gave rise to them in the first place. Addressing whether ephaptic coupling plays a functional role is not straightforward. Since the extracellular electric field is always present in the living brain — caused by cells functioning — it is challenging to study such functioning in the absence of Ve-activity as this would imply a comatose or a deceased brain [8]. A more tractable question is: which neural components are sensitive to extracellular fields? In the first section, we employ a bottom-up approach, describing ephaptic effects in synapses (microscale), individual neurons (mesoscale) and neural networks (macroscale). In the second section, we discuss temporal aspects of ephaptic coupling. Importantly, while the primary focus of this overview is in ephaptic coupling via electric fields induced by endogenous activity, study of such effects inevitably involves experiments with externally imposed fields. While externally imposed fields lack many of the features of the endogenous field, such studies have greatly enhanced our understanding of ephaptic coupling and have had an impact in medicine (see ‘Applications and translational neuroscience’).
Synapses (microscale) For chemical synapses, a linear relationship is observed between synaptic current Isyn and Vm (with the important exception of the NMDA receptor). Once the reversal potential (Erev) has been accounted for, synaptic currents, to a first approximation, can be described as Isyn(t) = gsyn(t)(Vm(t) Erev) with gsyn being the synaptic conductance. Isyn depends on Vm, which, in turn, depends on Ve. Thus, changes in Ve will influence synaptic currents, though such changes are expected to be rather weak (though potentially important at the network level [9]). 1
The term ‘ephapse’ from the Greek word (i.e., the touch or junction) was coined by Arvanitaki [5] to describe electric field interactions occurring between juxtaposed axons. Over the years researchers have used alternative terms such as ‘electric field effects’ to describe the same phenomenon. To date, the most detailed review about ephaptic coupling remains the seminal review by Jefferys [6]. For a more recent review see Weiss and Faber [7]. Current Opinion in Neurobiology 2015, 31:95–103
96 Brain rhythms and dynamic coordination
A stronger effect may be exerted by the impact of the electric field on electrodiffusion of charged ions. Specifically, ionic fluxes within the synaptic cleft (from presynaptic to postsynaptic neuron) via electrodiffusion induce electric fields locally, within the synapse. Intrasynaptic electric fields, in turn, can alter the membrane potential of the presynaptic terminal leading to changes in the ionic flux. Such events alter, for example, Ca2+-dynamics in the vertebrate retina by depolarizing the presynaptic terminal and increasing neurotransmitter release [10]. Similar events have been hypothesized to occur in cortex [11,12]. Finally, to the extent that electric fields can alter spiking of neurons (see next section), correlations in spiking between two synaptically connected neurons can affect the synaptic strength of their connection by inducing plasticity [13] — this would constitute an indirect impact of ephaptic coupling on synaptic functioning (e.g., [14,15,16]).
Neurons (mesoscale) Extracellular fields along the cable-like structure of neural morphology can be thought of as effective intracellular current injections, leading to alterations in their membrane potential [17,18,19,20,21,22]. The specifics of such entrainment depend on the morphological and physiological features of the neuron as well as the characteristics of the extracellular field — passive cable theory dictates that the impact the field has on Vm depends on the cable-field alignment as well as on three length constants: the characteristic length of the external field, the cable length, and the electrotonic length [22]. The range of Vm induced by a spatially inhomogeneous extracellular field can be of the order of the extracellular spatial voltage oscillation (Figure 1a). For the induced Vm-amplitude to become of the order of the amplitude of Ve, two independent conditions must hold: the length of the cable must be larger than the characteristic length of the external field and the electrotonic length must be larger than the characteristic length of the external field. If the cable length is considerably smaller than the characteristic length of the field, the Ve-gradient along the cable becomes insignificant and Vm remains broadly unaltered. The same applies when the electrotonic length is much smaller than the characteristic field length: for a decreasing cable diameter (i.e., decreasing electrotonic length [23]), the impact of the spatially inhomogeneous extracellular Ve on Vm decreases until the amplitude of the Vm-deviation along the cable becomes zero. Thus, for thin fibers such as axons, the induced Vm will be very small. Ephaptic coupling of Vm to electric fields has an impact on the active conductances of a cell and, thus, to spiking (Figure 1b). Such effects have been studied by applying extracellular fields across brain slices and whole-cell patching individual neurons [24,25,26,27,28] with a focus on the entrainment of spiking to relatively weak and temporally slow (less than 60 Hz) fields emulating local field potential (LFP)-activity in the brain. Changes Current Opinion in Neurobiology 2015, 31:95–103
in somatic potential of 70 mV, that is, below the noise levels of neurons [29], can significantly impact spiking and, in particular, the spike phase (with reference to the oscillatory extracellular signal). Importantly, while most studies have focused on the impact of electric fields on excitatory neurons, the same effects impact subthreshold and spiking of inhibitory neurons [27]. The usefulness of such studies for understanding ephaptic coupling to endogenous fields is limited–chiefly, the cases emulated in slice oversimplify in vivo activity where neurons are continuously bombarded by hundreds of postsynaptic currents along their intricate morphology in the presence of a spatially inhomogeneous and temporally dynamic electric field (Figure 1c; compare to fields in Figure 1a,b). Such limitations are present both for fields induced across parallel plates positioned millimeters away from each other (e.g., [24,25,30]) as well as fields elicited via stimulation pipettes (e.g., [1,28]). To account for the impact of endogenous fields on single neurons, both the intracellular and extracellular voltage would not only need to be monitored along a single cell but also manipulated, and all this in the behaving animal. Such experiments are now becoming feasible. For example, Ozen and colleagues [31] performed simultaneous extracellular and intracellular recordings in vivo while imposing weak electric fields via transcranial electrodes and found that the impact of the stimulation on neurons did not only depend on electrode placement or stimulation details but also to the behavioral state of the animal. Notably, while the extracellular depth recordings were not from the vicinity (within, say, 50 mm) of the impaled (via a sharp electrode) neuron, such work can serve as a precursor for future ephaptic coupling studies, ideally, combining local field manipulation, for example, via light stimulation.
Networks (macroscale) Two approaches have been employed to address network-level effects of ephaptic coupling: first, large-scale electric stimulation of slices and entire brains of living animals and humans, and second, computer simulations. Electric stimulation of brain slices has been the primary method of characterizing biophysical aspects of electric field effects at the network level. For example, we know that altering the field strength (but not the frequency) imposed via parallel plates, slice oscillation frequency progressively shifts to that of the imposed field (or a subharmonic) [26]. Similarly, when imposing different stimulus frequencies, slice dynamics become progressively entrained [30] (Figure 2a,b). Importantly, parallel plate, whole-slice stimulation has shown that emergent properties of neural networks are more sensitive to fields than single neurons [24,26]. Entrainment of spiking to field strengths as low as 0.5 mV/mm [25,28] (Figure 1a,b) firmly suggests that ephaptic entrainment www.sciencedirect.com
Ephaptic coupling to brain fields Anastassiou and Koch 97
Figure 1
(b)
Ve (mV)
10 Hz 20 Hz 50 Hz 60 Hz 100 Hz
2.5
1
0.5
1
1s
I0 = 50 nA
I0 = 100 nA
I0 = 200 nA
90°
90°
90°
90°
0.04
1.0 180°
0
4
8
12
0°
10 Hz
20 Hz
50 Hz
0.04 180°
0°
0.04 0°
180°
0.04 0°
180°
270°
270°
270°
N=2669
N=2639
N=2806
N=2907
N=2783, p=2.8 10–3
N=2709, p=6.7 10–5
N=2856, p
