Impact of anesthesia on micromagnetic stimulation (μMS) of the vagus nerve

To treat diseases associated with vagal nerve control of peripheral organs, it is necessary to selectively activate efferent and afferent fibers in the vagus. As a result of the nerve’s complex anatomy, fiber-specific activation proves challenging. Spatially selective neuromodulation using micromagnetic stimulation(μMS) is showing incredible promise. This neuromodulation technique uses microcoils(μcoils) to generate magnetic fields by powering them with a time-varying current. Following the principles of Faraday’s law of induction, a highly directional electric field is induced in the nerve from the magnetic field. In this study on rodent cervical vagus, a solenoidal μcoil was oriented at an angle to left and right branches of the nerve. The aim of this study was to measure changes in the mean arterial pressure (MAP) and heart rate (HR) following μMS of the vagus. The μcoils were powered by a single-cycle sinusoidal current varying in pulse widths(PW = 100, 500, and 1000 μsec) at a frequency of 20 Hz. Under the influence of isoflurane, μMS of the left vagus at 1000 μsec PW led to an average drop in MAP of 16.75 mmHg(n = 7). In contrast, μMS of the right vagus under isoflurane resulted in an average drop of 11.93 mmHg in the MAP(n = 7). Surprisingly, there were no changes in HR to either right or left vagal μMS suggesting the drop in MAP associated with vagus μMS was the result of stimulation of afferent, but not efferent fibers. In urethane anesthetized rats, no changes in either MAP or HR were observed upon μMS of the right or left vagus(n = 3). These findings suggest the choice of anesthesia plays a key role in determining the efficacy of μMS on the vagal nerve. Absence of HR modulation upon μMS could offer alternative treatment options using VNS with fewer heart-related side-effects.


Introduction
The vagus nerve is one of the most promising targets in autonomic neuromodulation.It is the 10th cranial nerve of the body and has nerve endings in almost all organs such as the heart, lungs, kidneys, pancreas, stomach etc.After the first report of vagus nerve stimulation (VNS) in the 1980s to stop seizures in dogs [1], it has found a wide variety of applications in the treatment of depression, obesity, diabetes, traumatic brain injury and many others [2][3][4][5][6][7][8][9][10][11][12].Hence, VNS devices and stimulation techniques have undergone several trials of research and development cycles in the recent past.Finally, in 1997 the first vagus nerve stimulation was approved by FDA for the treatment of epilepsy and depression [13].According to a report from 2018, over 100,000 patients have been undergone VNS [14,15].However, side effects from VNS are prominent and they show up in the form of bradycardia, bradypnea, indigestion, throat and tonsil pain etc [16].Recent investment into vagus nerve stimulation is sparked by this demand in the community on how to further reduce the side effects from VNS [17].Such side effects can be reduced significantly if specific fascicles of the vagus nerve are stimulated instead of the entire vagus nerve laying the grounds for spatially selective vagus nerve stimulation (sVNS) [18,19].
Of the several sVNS techniques developed in the past, special mention includes -fiber-selective stimulation [20,21], spatially selective stimulation [18,22], anodal block [23,24], kilohertz electrical stimulation (KES) block [25,26], reversible AC nerve conduction block [27] and neural titration [28,29].Such techniques have primarily focused on studying the impact of stimulation pulse [30], the electrode array geometry [31,32], and the stimulation protocol.Contrarily, there is promise in the application of micromagnetic neurostimulation (μMS) for selectively modulating neural activity in specific spatial regions [33][34][35][36].In this method, micrometer-sized coils, also known as microcoils (μcoils), are activated by a time-varying current, leading to the generation of a magnetic field.This magnetic field as per Faraday's law of induction induces an electric field that activates the neural tissue.Over electrical stimulation electrodes, the μcoils are advantageous because of the following reasons.First, these μcoils are not in any electrochemical contact with the neural tissues.This renders these μcoils much safer for chronic implantation due to much less biofouling effects [35,[37][38][39].Second, these μcoils offer spatially selective neural tissue activation.Saha et al has reported that, a solenoidal μcoil when driven by a sinusoidal current of 3 A, activates only 1.13 mm 3 out of 4.8 mm 3 neural tissue volume [36].sVNS could benefit significantly from these two unique properties of micromagnetic neurostimulation.The reported fiber-selective VNS has not been tested for any electrochemical safety issues [40].Therefore, micromagnetic VNS (μVNS) offers both-less biofouling after chronic implantation and spatially specific neural activation.
Jeong et al [41] has recently reported micromagnetic vagus nerve stimulation (μVNS) where the animals were anesthetized using isoflurane as the anesthesia.However, isoflurane has a depressive effect on the peripheral nervous system (PNS), particularly on the baroreflex [22,31].Thus, side effects from VNS that would normally occur in awakened animals could be absent or dampened in aimals anethesized with isoflurane.Therefore, isoflurane may reduce the efficacy of μVNS when compared to an awake animal.In this work we have used μcoils to study the impact of two different general anesthetics-isoflurane and urethane, and their impact on micromagnetic vagus nerve stimulation (μVNS).Unlike the previous report on μVNS [41], we used single cycle sinusoidal current of varying pulse widths (PW) at 3 A amplitude; each cycle repeated at 20 Hz.The orientation dependent spatially selective activation of the nerve has been discussed in zection 3.1.The impact on respiratory responses on the animals upon μVNS on both the left and the right cervical vagus under the influence of isoflurane has been reported in section 3.2 and the same under the influence of urethane has been reported in section 3.3.

The MagPen fabrication and elecrical circuit characterization
The efficacy of micromagnetic stimulation (μMS) on the vagus nerve was tested in vivo using the Magnetic Pen (MagPen) as the magnetic μcoil implant [33,34].The μcoil situated at the tip of the MagPen is of dimension 1 mm × 0.6 mm × 0.5 mm (μcoil model no.TDK Corporation MLG1005SR10JTD25).The green printed circuit board (PCB) of the MagPen is of length 3 cm.Due to orientation specificity of the μcoils in μMS (see section 3.1), the μcoil and has been prototyped in two orientations Type Horizontal (Type H or MagPen-Type H) and Type Vertical (Type V or MagPen-Type V) (see figure S1(a)).The complete details for the MagPen prototype have been provided in figure S1(b) (see Supplementary Information S1).Biocompatibility and anti-leakage current proofing for each of the MagPen prototypes has been ensured by providing a 10 μm thick Parylene-C coating at the tip of the MagPen.
An LCR meter (Model no.BK Precision 889B) has been used to study the electrical characteristics (or RLC characteristics) of the μcoil.To study the impact of μMS on the vagus nerve, the μcoil has been operated at the frequencies 1 kHz, 5 kHz and 10 kHz which corresponds to PWs of 1000 μsec, 500 μsec and 100 μsec respectively.Within this frequency range, the experimentally measured RLC characteristics of the μcoil have been discussed in Supplementary Information S1.Electrical components, the series resistance or DC resistance (R DC ) and the series inductance (L s ), have their individual importance's in successful μMS.
The μcoil when driven by a time-varying current (i(t)) generates a time-varying magnetic field (B(t)) as expressed by equations (1) and (2) respectively.
Where, f is the frequency at the which the μcoil is driving the current (pulse width (PW) for micromagnetic stimulus = f 1 ), A 1 is the amplitude of the current driving the μcoil, N and L are the number of turns and length of the μcoil respectively, μ r and μ 0 are the relative and vacuum permeability respectively.This magnetic field (B(t)) as per Faraday's law of induction induces the electric field (E ind ) on conducting mediums (here, biological conductor, the vagus nerve).
Where, l and S are the contour and the surface area of the neural tissue.Substituting equations (1) and (2) in equation (3) we get: Another form of equation ( 4) can be expressed as the electromotive force (emf or v(t)) induced in the nerve: Therefore, the induced electric field from the μcoil that activates neurons is a function of the series inductance (L s ) of the μcoil.
On the contrary, the R DC contributes to the heating on the neural tissues from the μcoil ) Numerical calculations and independent experiments conducted worldwide report that the temperature increase from these μcoils on neural tissues upon micromagnetic stimulation remains below 1 °C [39,41,42].

Animal preparation and surgery
Male Sprague Dawley rats (weight: 275-300 g: age 10-12 weeks) were purchased from Charles River Laboratories (Wilmington, MA) and housed in pairs in a temperature and light controlled room.Rats were allowed access to standard rat chow and distilled water ad libitum during this time.Rats were anesthetized with 5 v.% isoflurane (n = 7) and maintained at 2-3 v.% during surgical preparation.In the second part of experiments, rats (n = 3) were anesthetized using urethane of concentration/dose of 1.5 gm per 5 ml of saline that was infused at a rate of 40-50 μl/min.The femoral artery and vein were cannulated with PE50 tubing for continuous arterial pressure measurement and urethane perfusion throughout the experiment.A cervical midline incision was done to expose the left and right carotid arteries and vagus nerve.The vagus nerve was isolated inferior to the carotid bifurcation such that surrounding tissues were not going to be in contact with the μcoil.The baseline parameters for MAP and HR for the animal under two different general anesthetics are provided in table 1.The animal study was reviewed and approved by the Institutional Animal Care and Use Committee (IACUC), University of Minnesota, Twin Cities.Protocol no.2112-39640A.

The circuit powering the MagPen
The μcoil at the tip of the MagPen (see figure 1(b)) is being driven by a function generator (Model no.RIGOL DG1022Z) which generates single-cycle sinusoidal pulses (V in in figure 2(a)).This signal is then amplified by a class-D amplifier (Model no.Pyramid PB717X) (see figure 2(a)).The amplifier output forms the input to the μcoil.In this work, the amplitude and frequency of the sinusoidal current driving the μcoil has been kept constant at 3 A and 20 Hz respectively.Only the PW or duration of the sinusoidal current has been varied between 100 μsec, 500 μsec and 1000 μsec (see solidblack lines in figure 2(b)).This means that the current waveform driving the μcoil are single-cycle sinusoids of amplitude 3 A that have varying PW or duration; each sinusoid cycle is separated by 50 msec duration.In other words, each cycle is repeated at 20 Hz frequency.The μcoil driving waveform (amplitude, pulse width and frequency) have been adapted from a previous report by Stauss et al [43] on electrical stimulation of the left and right cervical vagus to observe successful hemodynamic and respiratory responses in rodents.
Figure 2(b) also shows the induced electric field waveform (see dotted-orange lines).Since, µ E t ( ) , i t d dt ( ) the shape for the induced electric waveform is the same as that of the current driving the μcoil but out of phase.As per our calculations, for a 3 A current driving the μcoil, the corresponding amplitude of the induced electric field on a neural tissue measured at a distance of 100 μm between the μcoil and the tissue will be 3.75 V m −1 .

Orientation specific and spatially selective μMS of the vagus nerve
Micromagnetic neurostimulation offers two interrelated features that is unique when compared to other neuromodulatory techniques.First, it offers μcoil orientation specific activation of neurons.Second, the alignment of the neural fibers with respect to the μcoil is also critical for successful activation.Combining these two features, arises the most important feature for μMS -its ability for spatially specific stimulation.Our in vitro experiments on rat hippocampal slices [33] and in vivo experiments on the rat sciatic nerve [42] provided a general overview on the correct alignment of the nerve fiber with respect to a particular orientation of the μcoil.In this work, we used a NEURON model to corroborate our prior experimental findings and project them to our experiments on the vagus nerve.The FEM modeling for both MagPen-Type H and MagPen-Type V were performed when the μcoil was driven by a 3 A sinusoidal current at 1 kHz frequency (see figure 3).The induced electric field was measured on a tissue of dimension 4 mm × 4 mm × 300 μm at a distance of 100 μm between the μcoil and the tissue (see figure 3).For MagPen-Type H μcoil orientation, when the nerve fiber is oriented along y-axis (see figure 3(a)), no activation of the neuron was observed.Keeping the μcoil orientation same, when the nerve fiber is oriented along x-axis (see figure 3(c)), successful activation of the neuron was observed.Similarly, for MagPen-Type V μcoil orientation, when the nerve fiber is oriented along both x-axis and y-axis (see figures 3(b) and (d)), no activation of the neuron was observed.
The interesting point to note is that the spatial distributions for induced electric field is constant for MagPen-Type H (see figures 3(a) and (b)) and  MagPen-Type V (see figures 3(c) and (d)), irrespective of the neuron orientation.Therefore, the key determining factor for successful activation of the neurons is the directionality of the induced electric field with respect to the neurons (marked by solid-red arrows in figure 3).The directionality of the induced electric field in figure 3(c) favors neuron activation.

Influence of isoflurane during μMS of the cervical vagus
The MagPen-Type H was held over the vagus nerve as per the orientation and directionality in figure 3(c).The experimental set-up for orienting the MagPen over the rodent vagus is shown in figures 1(a) and (b).The rodents under the primary influence of isoflurane as the anesthesia were provided with micromagnetic stimulation on both the left and the right cervical vagus.On applying μMS of varying pulse width to the left and right cervical vagus, changes in arterial blood pressure (BP) and heart rate (HR) were measured from the femoral artery.For the micromagnetic stimulus of 3 A amplitude, 100 μsec and 500 μsec duration repeated at 20 Hz frequency, no significant changes in the mean arterial pressure (MAP) or heart rate (HR) was observed when compared to baseline signal (see table S4).Baseline signal refers to the MAP and HR parameters in the rodent when no μMS was applied to the vagus (see the black bars in figure 4 and table 1).At the same amplitude and frequency of the micromagnetic stimulus but at 1000 μsec pulse width, stimulation of both the left and the right vagus resulted in decreased MAP when compared to baseline (see table S4).The left vagus showed an average drop in MAP of 16.75 mmHg upon μMS (n = 7) (see figure 4(a-i)).While the right vagus upon μMS showed an average drop of 11.93 mmHg in the MAP (n = 7) (see figure 4(b-i)).Surprisingly, we did not observe any significant changes in HR during micromagnetic stimulation of both left and right cervical vagus compared to the baseline signal (see figures 4(a-ii), (b-ii) and table S4).This suggests the drop in MAP associated with left vagus μMS was the result of stimulation of afferent, but not efferent fibers.

Influence of urethane during μMS of the cervical vagus
Isoflurane has a depressive effect on the peripheral nervous system (PNS), particularly on the baroreflex [22,31].Thus, side effects from VNS that would normally occur in awakened animals could be absent or dampened in animals anethesized with isoflurane.Therefore, isoflurane may reduce the efficacy of VNS when compared to an awake animal.Hence, we switched the anesthesia to urethane which has been reported to interfere less with peripheral nerve activity.
The left and the right cervical vagus of the rats were also treated with micromagnetic stimulus (see figures 1(a) and (b)) while the rats were under the influence of urethane as the primary anesthesia.Under the influence of urethane, the rats showed a lower baseline MAP and higher baseline HR signals (see black bars in figure 5 and table 1) when compared to the corresponding baseline signals under the influence of isoflurane (see black bars in figure 4).The micromagnetic stimulus was same-single cycle sinusoidal current of amplitude 3 A and pulse widths varying between 100 μsec, 500 μsec and 1000 μsec.Each cycle repeated at 20 Hz frequency.No significant changes in either MAP or HR of the rodents were observed under the influence of urethane as the primary anesthesia (see figures 5(a)-(b) and table S4).However, on the same animal when the supply of urethane was stopped and immediately isoflurane flow was initiated, the animal showed drops in MAP at 1000 μsec pulse width of the micromagnetic stimulus.This suggests the choice of anesthesia did play a critical role in investigating the efficacy of μMS on the rat cervical vagus.

Discussion
This report investigates the impact of micromagnetic stimulation (μMS) of the vagus nerve in vivo on a rodent model using two different general anesthetics.Ottaviani et al [49] reported that the structure and fuctions of the vagus nerve in mammals vary significantly.Therefore, any study on vagus nerve stimulation implant development is probably a proof-of-concept study as the structure and function of the human vagus nerve varies significantly from that of other mammals.Under the influence of isoflurane as the anesthesia, μMS of the left vagus nerve in rats showed significantly greater drops in MAP (16.2% drop in MAP upon micromagnetic stimulation of the left vagus) than when the right vagus nerve was stimulated (11.52% drop in MAP upon micromagnetic stimulation of the right vagus nerve).From this observation one can conclude that the left vagus in rats is more susceptible to μMS than the right (see section 3.2).However, this could also suggest that for our experimental set-up as in figures 1(a) and (b), the orientation of the μcoil was at an angle that activated fewer vagal nerve fibers in the right vagus.This directly highlights the importance of the angular alignment of these μcoils with respect to the tissue.This phenomena has been better explained in [50].
The fact that we were able to observe mean arterial pressure (MAP) changes upon micromagnetic stimulation of the left cervical vagus but no significant heart rate (HR) changes, was surprising (see section 3.2).This led us to suspect probably we were activating some other nerve or lymph node in and around vagus -for example, the phrenic nerve or the laryngeal nerve lymph node-due to the spatially selective nature of activation of the μcoils (discussed in section 3.1).Hence, for the last 3 animals (n = 3), we carefully isolated the vagus from the adjacent lymph nodes and nerve and applied micromagnetic stimulation to the vagus nerve.However, our experimental observations remained same as reported in section 3.2.The branches of the vagus extend beyond the cervical vagus into different organs-the heart, lungs, bronchi and oesophagus.However, our experiments were set to study respiratory effects on the animals upon micromagnetic vagus nerve stimulation (μVNS).The effect of μVNS on any other target organ must be investigated as future work.Prior to that, a better understanding of the vagal nerve anatomy in rodents is required to be made as an important step towards understanding selective vagus nerve stimulation (sVNS) [51,52].
In this work, we reported that during micromagnetic stimulation of the vagus, only 3 different pulse widths (PWs) were tested-100 μsec, 500 μsec, and 1000 μsec.Pulse widths below 100 μsec are too small to trigger any neural activation.Furthermore, at such high frequency of operation of the inductive μcoil (PW = 1000 μsec corresponds to f = 10 kHz), the μcoil adds increased impedance to the flow of current through it (see Supplementray Information S1).However, we did try to investigate the impact of micromagnetic stimuli for PWs greater than 1000 μsec.Under the influence of isoflurane as the anesthesia, at 2000 μsec PW for the micromagnetic stimuli at both the left and the right cervical vagus, the MAP dropped; whereas the HR remained same as the baseline (see Supplementary Information S3).However, PWs greater than 2000 μsec for the micromagnetic stimuli were avoided because of potential heat generation leading to burning of the μcoils, thereby causing loss of electrical connection [36].Similar kind of μcoil burning and loss of electrical connection has been reported by Jeong et al during their experiments on micromagnetic vagus nerve stimulation (μVNS) [41].

Conclusions
The efficacy of micromagnetic stimulation (μMS) of the left and the right cervical vagus in rodent model was investigated.In this in vivo study, μMS was delivered when the rat was under the influence of two anesthesiaisoflurane and urethane -independently.Under the influence of isoflurane, the left cervical vagus of the rat seemed to be more susceptible to μMS than the right cervical vagus.Upon μMS of the left vagus, the drop in mean arterial pressue (MAP) was 1.4 × than that observed during μMS of the right.In both cases, the μcoil was driven by a 3 A sinusoidal current at 1000 μsec pulsewidth at a frequency of 20 Hz and the rat was under the influence of isoflurane as the anesthesia.This drop in MAP, with no change in HR, is consistent with activation of afferent vagal fibers that regulate sympathetic control of the vasculature but not the heart.Assumming that the isoflurane anesthesia prevented the HR variability upon μMS of the cervical vagus, the anesthesia was switched to urethane.Under the influence of urethane no changes in either MAP or HR was observed.However, terminating the flow of urethane on the same animal and initiating isolfurane flow in the animal, the drop in MAP returned during μVNS.These findings suggest that anesthesia plays a critical role in the efficacy of μMS on the cervical vagus.Therefore, the absence of HR modulation upon μMS could offer alternative treatment options using VNS with fewer heart-related side-effects.

2. 5 .
NEURON modelingTo determine whether the induced electric field generated by the μcoil in MagPen could activate neurons, a simulation study was performed using the NEURON package (https://neuron.yale.edu/neuron/)[46,47].The model developed by Pashut et al[48] was modified.Their array of the induced electric field plot was replaced with the induced electric field generated from the FEMmodel of the μcoils on ANSYS-Maxwell (eddy current solver) (see section 2.4).Simulations of the time varying waveform at 3 A amplitude sinusoidal current at 2 kHz frequency through the MagPen were performed.The membrane potential was then measured at the soma and the volume of tissue activation around the μcoil of the MagPen was estimated (see section 3.1).

Figure 1 .
Figure 1.(a) Experimental set-up to demonstrate micromagnetic stimulation (μMS) of the vagus nerve.MagPen is held over the rat left cervical vagus nerve.(b) Zoomed in image of the μcoil in MagPen-Type H held at an angle over the rat cervical vagus.(c) DSI telemetry sensors over the femoral artery for mapping heart rate (HR), blood pressure (BP) and respiration rate (RR) changes during μMS.

Figure 2 .
Figure 2. (a) External connections to the μcoil for μMS includes an arbitrary function generator connected to a class-D n-channel MOSFET amplifier.(b) The single-cycle sinusoidal waveform of the current driving the μcoil (i(t)) represented in solid black lines.The corresponding waveform of the induced electric field from the μcoil (E(t)) represented in dotted orange lines.The amplitude of the current driving the μcoil, correspondingly the amplitude of the induced electric field (E(t)) and the frequency were kept constant at 3 A, 3.75 V/m and 20 Hz (= 50 msec) respectively.The pulse width (PW) of the micromagnetic stimulus has been varied between 100 μsec, 500 μsec and 1000 μsec in the experiments on the rodent cervical vagus nerve.

Figure 3 .
Figure 3. Different orientations of the μcoil in MagPen with respect to the vagus nerve showing their corresponding spatial distribution of the induced electric field (E TypeH and E TypeV ) calculated from FEM modeling.(a) Orientation of MagPen-Type H with vagal nerve fiber oriented along y axis showing no neural response.(b) Orientation of MagPen-Type V with vagal nerve fiber oriented along y axis showing no neural response.(c) Orientation of MagPen-Type H with vagal nerve fiber oriented along x axis showing neural response.(d) Orientation of MagPen-Type V with vagal nerve fiber oriented along x axis showing no neural response.

Figure 4 .
Figure 4. Impact of isoflurane during μMS of the left and right cervical vagus nerve.(a) μMS of the left cervical vagus.(i) Mean arterial pressure (MAP) and (ii) Heart rate (HR) variation with pulse width of the micromagnetic stimulus.(b) μMS of the right cervical vagus nerve.(i) Mean arterial pressure (MAP) and (ii) Heart rate (HR) variation with pulse width of the micromagnetic stimulus.The μcoils were being driven by single-cycle sinusoidal current of amplitude 3 A at a frequency of 20 Hz and varying PW (n = 7).

Figure 5 .
Figure 5. Impact of urethane during μMS of the left cervical vagus nerve.(a) Mean arterial pressure (MAP) and (b) Heart rate (HR) variation with pulse width (PW) of the micromagnetic stimuli.The μcoils were being driven by single-cycle sinusoidal current of amplitude 3 A at a frequency of 20 Hz and varying PW (n = 3).

Table 1 .
Baseline parameters for MAP and HR of the animals under 2 different types of general anesthetics.
[44]ent modeling (FEM) of μcoils FEM simulations on Ansys-Maxwell, eddy current solver[44](ANSYS, Canonsburg, PA, United States) were performed to understand the magnetic field and the induced electric field spatial distribution from the μcoils (see Supplementary Information S2).It generates the magnetic field and induced electric field spatial (MSI) at the University of Minnesota (8 cores of Intel Haswell E5-2680v3 CPU, 64 × 8 = 512 GB RAM and 1 Nvidia Tesla K20 GPU).The induced electric field values were then exported to be analyzed using a customized code written in MATLAB (The Mathworks, Inc., Natick, MA, USA).