Electrical or magnetic nerve stimulation enhance the BoNT/A-mediated muscle paralysis

Abstract

Background

Botulinum toxin type A (BoNT/A) is first-line treatment for muscle spasticity but the dose can be limited by side effects depending on the treated muscle mass and location. Stimulation of peripheral motor nerves upon BoNT/A injection into the innervated muscles may be a technique to increase the BoNT/A effect without increasing the dose. Our goal was to study the augmentation of the BoNT/A effect by transcutaneous electrical (TENS) or magnetic nerve stimulation (MS) in a mouse model.

Methods

The paralytic effect of BoNT/A with or without TENS was evaluated with the mouse digit abduction score (DAS) at 20, 30, and 40 U/kg. The paralytic effect of BoNT/A with or without peripheral MS was evaluated with the mouse DAS at 20 or 30 U/kg.

Results

After 20 U/kg BoNT/A treatment combined with TENS the DAS was higher and showed a longer duration of paralysis than without TENS. After 30 U/kg BoNT/A treatment combined with magnetic stimulation the DAS was higher than without magnetic stimulation.

Conclusion

We established an electrical and a magnetic nerve stimulation protocol combined with BoNT/A treatment in mice. Combination of i.m. BoNT/A injection with TENS or MS demonstrated an enhanced BoNT/A effect as compared to non-stimulated groups. Despite limitations of the magnetic stimulation method in the mouse our data revives the discussion of BoNT/A effect augmentation by stimulating the injected muscles’ efferent nerves during the BoNT/A uptake phase.

Introduction

The rationale for Botulinum neurotoxin type A (BoNT/A) treatment of spastic muscles which is a common symptom of the upper motor neuron syndrome [1–3] is the blockage of acetylcholine release from nerve terminals at the neuromuscular junction (NMJ) resulting in transient paralysis. Although the BoNT/A effect in clinical practice may last for about 12 weeks, repeated injections to maintain the treatment effect of BoNT/A are required which is sometimes perceived stressful by patients [4–9]. Depending on the treated muscle mass and location the required BoNT/A dose may also cause undesired effects [10–12]. For these cases it would be advantageous to have an adjuvant procedure at hand such as peripheral nerve stimulation during or immediately after the BoNT/A injection that increases the efficacy of the BoNT/A treatment. Several types of adjuvant treatment combined with BoNT/A injection such as muscle stretching, adhesive taping, therapeutic ultrasound, vibration, transcutaneous electrical nerve or muscle stimulation have been proposed for the management of spasticity in order to increase the efficacy and duration of the BoNT/A effect [13–15]. However, these adjuvant treatments have not been very successful, although there is some evidence for a positive effect of adhesive taping or transcutaneous electrical nerve or muscle stimulation [16–20].

In this study we focused on peripheral electric or magnetic nerve stimulation to enhance the BoNT/A effect following the hypothesis that additional peripheral nerve stimulation intensifies acetylcholine release from the neuromuscular terminal which in turn increases the probability of BoNT/A uptake into the release vesicles [21–23].

Electrical stimulation alone is used in neurorehabilitation for muscle strengthening and improving motor recovery [24]. High-frequency (100 Hz) transcutaneous electric nerve stimulation (TENS) can be beneficial in reducing post-stroke spasticity in the lower extremities of long-term stroke survivors as an adjuvant therapy. TENS lasting more than 30 min per session with an electrode placement along the nerve and an intensity that was twice of that of the sensory threshold has been recommended when performed for more than 2 weeks [25, 26]. However, there is few data on the optimal TENS parameters.

TENS has also been frequently applied as an adjunct to BoNT/A therapy in pediatric or adult patients [13], always based on the idea that the electrical nerve stimulation leads to an increased toxin uptake due to an increased release vesicle turnover. This should result in increased binding, internalization and translocation of BoNT/A into the presynaptic terminal at the neuromuscular junction [23]. Considering that BoNT/A uptake takes place within a few minutes after injection [7, 21, 22, 27, 28], electrical stimulation of the injected muscle should be performed immediately after BoNT/A injection. Supporting this hypothesis a reduction in latency of the onset of the BoNT/A paralytic effect after electrical stimulation of motor nerves was already found in the 1960s using an ex vivo rat phrenic nerve-diaphragm preparation [21, 22]. These authors as well as Simpson 1980 [22] found that the neuromuscular blocking effect of BoNT/A was enhanced after electrical stimulation.

In a recent review of clinical studies [23] on spasticity treatment using BoNT/A plus electrical nerve-muscle stimulation 6 reviewed studies found significant enhancing effects on clinical and neurophysiological outcomes after electrical stimulation in combination with BoNT/A injection compared to control conditions [17–20, 29, 30]. However, there was a large variability in the stimulation protocols as well as in outcome measures. Another limitation was the lack of data on the exact stimulation parameters. So, further experimental evidence for the most appropriate BoNT/A dose in combination with the appropriate electrical stimulation protocol is needed.

Magnetic stimulation for spasticity treatment has been applied as repetitive transcranial magnetic (rTMS) or repetitive peripheral nerve stimulation (rPMS) to reduce spasticity after stroke [15, 31]. rTMS was combined with BoNT/A treatment in a case report of one stroke patient, with BoNT/A treatment and occupational therapy in 40 post-stroke patients with spastic upper limb hemiparesis and with BoNT/A plus rehabilitation in 25 pediatric patients with spastic diplegia [32–34]. The objective with rTMS was to improve neuronal plasticity and functional connectivity in the CNS reorganization process and with rPMS an increased peripheral somatosensory feedback rather than an improvement of BoNT/A uptake at the neuromuscular junction. rPMS is a non-invasive and painless method which makes it more acceptable for patients than electrical stimulation. The magnetic field creates a secondary electrical field and induces a current in the peripheral nerves which downstream will cause neurotransmitter release at the nerve’s terminals and finally muscle contraction [26]. Thus, the magnetic nerve stimulation shares most physiological effects with the electrical stimulation. The electrical stimulation, however, does not only depolarize a nerve but also depolarizes a muscle directly [35, 36]. rPMS has been used as an adjuvant to manual muscle stretching to reduce wrist and finger flexor muscle spasticity [37] and as a stand-alone procedure to reduce upper limb spasticity in adolescents and bilateral spastic cerebral palsy in children [36–39].

Ten theta-bursts with 3 stimuli each at 50 Hz applied repeatedly, at a theta frequency of 5 Hz every 10 s, for a total duration of 200 s lead to a significant reduction of spasticity (determined by the Ashworth scale). In this study a commercial figure-eight shaped coil was placed over the muscle belly of the spastic biceps brachii and wrist and finger flexors as well as the rectus femoris, hamstrings, gastrocnemius and soleus muscles for magnetic stimulation [40]. A review of 8 randomized controlled trials and a meta-analysis of 6 trials involving 170 patients with post-stroke spasticity (> 6 months) evaluated with the Ashworth or modified Ashworth scale indicated that rPMS has the potential to reduce spasticity in the upper and lower extremities [15]. The stimulation parameters used in the trials varied ranging from 5 to 25 Hz in frequency and 3 to 30 min per session. However, these clinical studies did not allow to determine the optimal stimulation protocol. rPMS has so far not been applied to augment the BoNT/A effect.

The goal of our investigation was to perform a preclinical protocol for transcutaneous electrical stimulation (TENS or ES) in combination with BoNT/A treatment in mice in order to augment the effect of BoNT/A alone. The electrical stimulation scheme was based on a stimulation protocol for muscle training in a mouse model [41]. In analogy, the goal of repetitive peripheral magnetic stimulation (rPMS or MS) was to evaluate if this stimulation technique would booster the effect of BoNT/A and allow for lower BoNT/A dosages, fewer side effects and potentially lower costs.

Materials and methods

Animals

Adult male Balb/cAnNRj mice between 5 and 7 weeks old were purchased from Janvier Labs (Le Genest-Saint-Isle, France). Mice were kept on a 12 h light/dark cycle. All animals were maintained in an enriched environment under a constant temperature (22 ± 2 °C) and humidity (30–70%) with food and water available ad libitum. Animals were acclimatized for at least 7 days prior to experimentation.

Studies were performed in full compliance with the European Communities Council Directive 2010/63/EU. Experimental procedures were approved by the local authority for the use of animals for scientific purposes (reference 2347-13-2022-35-G, Regional Council of Brandenburg).

BoNT/A

Mice were assigned to each test group designated to receive the BoNT/A formulation or placebo with or without electrical or magnetic stimulation, respectively. BoNT/A groups were dosed with incobotulinumtoxinA (BoNT/A 150 kDa; Xeomin^®, Merz Pharmaceuticals GmbH, Germany) using ascending dosages of 20, 30, 40 U/kg. BoNT/A 100 U/vial was reconstituted using preservative-free saline (0.9% B.Braun Melsungen AG, Melsungen, Germany) and 0.2% human serum albumin (CSL Behring, Marburg, Germany). Signs of pain after the injection were not observed. The reference group was treated with an equal volume of human serum albumin (0.2%, v/v) in saline (0.9%).

Experimental protocol

Prior to BoNT/A administration mice were anesthetized with isoflurane and kept under general anesthesia throughout the experimental procedure. Once fully anesthetized in an induction chamber anesthesia was maintained with a nose cone (5% isoflurane in 100% medical oxygen, 1 L/min). Test and reference item were injected either in the right gastrocnemius muscle or in the right tibialis anterior (20 µl). For the i.m. injection the right hindlimb was shorn with an electric clipper (Aesculap AG, Tuttlingen, Germany) and the injection was performed with a 100 µl Hamilton syringe (Hamilton Company, Nevada, US) with a 30 gauge needle (B.Braun Melsungen AG, Melsungen, Germany).

Body weight was measured at various time points throughout the study. Animals that showed more than a 20% decrease from their body weight at treatment initiation, or signs of moribundity, were immediately euthanized and censored for the rest of the study.

In the first experiment investigating the influence of electrical stimulation (ES) on BoNT/A efficacy 7 groups of mice were included (n = 10 up to n = 18 mice per group). A single IM injection of BoNT/A was administered to the right gastrocnemius muscle of the corresponding animals at a dose of 20, 30 and 40 U/kg body weight. The placebo control group received a single IM injection of vehicle to the right m. gastrocnemius at a dose of 1 mL/kg body weight.

In the second experiment investigating the magnetic stimulation (MS) 5 groups were included. A single IM injection of BoNT/A was administered to the right tibialis anterior muscle of the corresponding animals at a dose equivalent to 20 and 30 U/kg body weight. The site of injection was chosen to coordinate the arrival of the neurotoxin at the neuromuscular junction in the tibialis anterior muscle with the magnetic stimulation of the respective motor nerve. The placebo control group received a single IM injection of vehicle to the right tibialis anterior muscle at a dose of 1 mL/kg body weight.

The test groups were stimulated within 2 min after BoNT/A injection either electrically or magnetically. The comparative groups received equivalent dosages of BoNT/A but were not stimulated. In addition, the placebo group was stimulated electrically or magnetically in order to investigate clinical signs of the stimulation per se. Animals were closely monitored daily for clinical signs throughout the experiment. At the end of the experiment mice were sacrificed via cervical dislocation in deep isoflurane anesthesia.

Electrical stimulation procedure

Mice were electrically stimulated with bipolar tweezers (Wirtschaftsgenossenschaft deutscher Tieraerzte eG, Garbsen, Germany) connected to a NeuroTrac Rehab stimulator (Tens4you.de, Hann. Muenden, Germany) in the middle of the right thigh. The tweezers were in contact with the skin and transcutaneously stimulated all femoral nerves passing underneath. The stimulation pattern was chosen according to previous work by Ambrosio et al. 2012 [41]: The pulse pattern was 150 µs at 9 mA current and 50 Hz frequency. One block consisted of 10 stimuli á 5 s followed by a pause of 10 s. Thus, one block had a duration of 150 s in total (10 × 5s on and 10 × 10s off). It was started and stopped by the experimenter manually. Two blocks of stimulation were carried out with a pause of 300 s in between. The full procedure from starting the first block to the beginning of the second took 8 min. During and after the procedure the mice were placed under an infrared lamp in order to recover body temperature after anesthesia before they were returned to their home cage.

Magnetic stimulation (MS)

Stimulation device

MS was delivered using a custom-built round air coil (TeslaPulseCoil25). The TeslaPulseCoil25 and the pulse generator TeslaPulsGen 1 were designed and built by preclinics Research Instruments GmbH (Potsdam, Germany).

A commercial device could not be identified although we tested a dedicated rodent research system by MagVenture GmbH (Cool-40 rat coil coupled to the stimulator unit MagPro X100 MO). It turned out that repetitive pulse trains at 2 Hz or higher at 80% stimulation intensity resulted in an overheating of the coil which prevented further stimulation. Maximum 3 repetitions including 3 pulses could be administered. After every stimulation the temperature increased > 20 °C and it was required to wait until the temperature was < 20 °C to be able to start another pulse train. Thus, a custom-built device was chosen for the experiment.

The coil dimensions of the TeslaPulseCoil25 were 14 mm for the inside diameter and 30 mm for the outside diameter. The coil was wound with two layers of copper wire (2.0 mm diameter) with 18 turns in total (2 × 9). The windings were fixed in wrapping fiberglass yarn around which is glued with epoxy resin. In order to fix the coil and isolate the electric terminals a POM (Polyoxymethylene) block with a hole for the coil and space for wiring was used. In addition, the coil terminals (14 mm diameter solid brass, suitable for high currents and discharging heat) were isolated electrically in another POM block each (Fig. 1A). Next to the blocks there was a small working table composed of a polycarbonate plate integrated (150 mm length x 150 mm width, 6 mm thick) in order to stabilize the apparatus and also hold the head and the upper part of the mouse body securely and comfortable while allowing the operator access to the spinal cord. The embedded coil and the table were positioned onto a box (280 mm length x 240 mm width, 150 mm height). The TMS coil was driven by the custom stimulator unit (Preclinics Research Instruments GmbH) (Fig. 1B). The whole device was built with short distances between capacitor bank, half-bridges and stimulation coil. The advantages are small parasitic resistance and inductance beside the stimulation coil. High current through the coil discharges the capacitor bank resulting in a cosine like waveform. The pulse power for constant time shape pulses was set up with the voltage on capacitor bank. The maximum voltage by design was 600 V with peak currents of up to 900 A.

Fig. 1.

Magnetic stimulation customized coil and stimulator setup. MS was delivered using a custom-built round air coil (TeslaPulseCoil25) coupled to the pulse generator TeslaPulsGen 1. A TeslaPulseCoil25; B TeslaPulseGen1 stimulator; C Top view: TeslaPulsGen 1 Stimulator and TeslaPulseCoil 25 with tape and marked stimulation location on working table

The magnetic stimulator is capable of producing rapid stimulating pulse trains. This instrument uses an external commercial power supply for charging a high pulse current capable capacitor bank. The stimulator unit has an external power supply with an adjustable voltage with a maximum voltage of 600 V DC and a max current of 3 A. The capacitor bank (2 × 180 µF, bipolar in parallel connection with a capability for driving peak currents about 1000 A) is then discharged via two high power half-bridges driven by commercial bridge drivers through the stimulating coil. The wave form is generated by a µController. Programmed is a rectangular waveform (25% +, 50% reverse, 25% +) with a duration of 120 µs. The time-varying magnetic field is generated by short-duration, large-amplitude current pulses delivered to the air coil located near the target tissue. This configuration generates a single cosine pulse. The half-bridge coil driver is comprised of a GD650HFX170P1S IGBT module (StarPower Semiconductor Ltd., China), 1.7 kV, 1.073 kA (pulse current at 25 °C for 1ms: 1300 A, 10 µs short current capability) as well as the half bridge driver: 2SC0106T2A1-12 (Power Integrations, San Jose, USA) with external power supply (15 V, 500 mA). There is also a half bridge driver control with a fixed gate time programmed (µC STM32F4xx, USB-power supply). The pulse train is started via an external foot pedal switch. The pulse form used here is bimorph with a duration of 120 µs and a frequency of 2 Hz for 15 s.

Procedure

In order to deliver the magnetic stimuli to the spinal cord of the mice at lumbar vertebrae L3-L5 a subject holder was constructed over the coil. The holder was marked with a cross to mark the position of the highest magnetic field and thus the highest secondary electrical field with the goal to administer the same stimulation intensity to every mouse at the same position of the spinal cord. The mice were placed on the back directly on the holder at the coil (Fig. 1C). The mice were thus attached to the coil in a dorsal position placing the spine at lumbar vertebrae L3-L5 at the maximum of the magnetic field induced to stimulate the spinal cord underneath.

Due to the small size of the mouse the magnetic impulse can only induce a potential difference in a small nerve volume which means that higher stimulation intensities are needed to exert the same effect as in larger mammals. The magnetic flux in a nerve as a conductive structure increases with the fiber diameter squared and decreases with the third power of the distance from the coil.

where Φ is the magnetic flux in a wire, B is magnetic field strength and the unit of B is Tesla (T), A is area of the coil (A=, θ is the angle between the magnetic field and the perpendicular to the coil.

where µ0 is the vacuum permeability, m is the magnetic moment of the coil, r is the distance from the coil.

The pulse frequency used was 2 Hz. One pulse train consisted of 30 stimuli followed by a pause of 50 s. The pulse train was started and stopped by the experimenter using a pedal coupled to the stimulator. The duration of one pulse train was 15 s. The pulse train was started 10 times (1 train: 15 s on, 50 s off). During pilot investigations visible muscle stimulation of mice started at 300 V. With 500 V a strong stimulation was produced. Thus, a voltage of 500 V was selected for this study.

DAS assay and body weight development in mice

The hind limb digit abduction reflex in the mouse was induced by grasping the animal lightly at the back fur of the torso and lifting it swiftly into the air. Animals were pre-screened for a normal digit abduction response (DAS 0) before the experiment and those showing abnormal digit abduction responses or hind paw deformities were excluded from the study. In this study no mouse was excluded. The digit abduction response of each mouse was scored using a modified five-point scale, from normal reflex/no inhibition (DAS 0) to full inhibition of the reflex (DAS 4) as described in female Swiss Webster mice [42]. The modification included 0.25 steps between the main scores leading to 17 steps as established by the experimenter. Mice were scored for digit abduction response at various intervals following BoNT/A injection over 25–30 days by the same experimenter. Body weight was measured before dosing as well as at the days of the digit abduction scoring to calculate the relative group mean body weight development over time.

Statistical analysis

In each experiment, the mean relative body weight changes were expressed as percentage of weight change at each time point compared to the initial body weight (day 0). Data were analyzed by two-way repeated-measures analysis of variance (RM ANOVA) followed by Bonferroni post-hoc analysis. The mean DAS calculated for each dose and time point was compared using two-way RM ANOVA to assess effects of stimulation combined with BoNT/A treatment on the functional outcome over time. Comparisons between groups were performed using Bonferroni post-hoc testing. In case of missing values, the mixed-effects model was applied followed by Bonferroni post-hoc analysis. Outliers were identified with the Dean-Dixon test.

An unpaired, one-tailed non-parametric Mann-Whitney-U test was used to compare two groups (e.g. mean peak DAS). Individual replicate area under the curve values were calculated to determine the mean AUC value for DAS (DAS_AUC) from DAS response versus time curves, respectively. Two groups were compared via unpaired, parametric t-test with Welch´s correction. Time (days) to return-to-DAS₂₅ were determined from the intersection of the duration offset lines with DAS = 1 on the appropriate graphs for each of the groups as the duration of action.

All data are presented as mean ± SEM. Statistical analyses were carried out using GraphPad Prism (version 9.5.1, GraphPad Software, San Diego, CA, USA). The threshold for the level of significance was set at α = 0.05.

Results

Electrical stimulation

Impact of BoNT/A and electrical stimulation on body weight

BoNT/A caused a dose-dependent transient decrease of body weight in the first 7 days after injection resulting in up to 2.0% mean loss in the 20 U/kg groups, up to 5.3% in the 30 U/kg and up to 8.5% in the 40 U/kg groups (BoNT/A-treated vs. placebo p < 0.0001 for 20, 30 and 40 U/kg). After this phase of decrease all dose groups gained weight at a similar rate (Fig. 2A-C). Electrical stimulation did not further decrease the body weight as compared to BoNT/A only treated groups, irrespective of the BoNT/A dose 20 U/kg, 30 or 40 U/kg.

Fig. 2.

Impact of electrical stimulation after i.m. injection of BoNT/A on body weight or digit abduction score. A-C Relative mean body weight after BoNT/A injection with and without electrical nerve stimulation (A, BoNT/A 20 U/kg; B, 30 U/kg; C, 40 U/kg). BoNT/A treated groups had an initial decrease in body weight followed by weight gain after day 7 (treated vs. placebo p < 0.0001 two-way RM ANOVA). Electrical stimulation did not further decrease the body weight as compared to BoNT/A only treated groups, irrespective of the BoNT/A dose. Mean ± SEM, n = 10–18. D The paralysis in BoNT/A 20 U/kg plus ES treated mice lasted 3.2 days longer until return to the DAS₂₅ baseline (corresponding to score 1). In addition, BoNT/A combined with ES caused an upwards shift of the DAS response curve over time (two-way RM ANOVA followed by Bonferroni´s multiple comparisons test until day 4 stim vs. non-stim **p < 0.01 on day 2). Mean ± SEM, n = 18. E Mean peak DAS dose response after BoNT/A injection or BoNT/A injection with electrical nerve stimulation. BoNT/A combined with ES increased toxin potency in 20 and 30 U/kg BoNT/A groups (stim vs. non-stim **p < 0.01; one-tailed Mann-Whitney-U test). ES combined with BoNT/A 40 U/kg did not further increase the DAS_max. Mean ± SEM, n = 10–18

Impact of BoNT/A and electrical stimulation on the digit abduction score

Toxin-induced local muscle paralysis following injection on BoNT/A into the m. gastrocnemius was assessed in the DAS assay. The assay is commonly used to measure the potency of BoNT/A products as well as the duration of action [42, 43] and was performed here to investigate if electrical stimulation has an impact on the paralytic effect of BoNT/A.

Mice injected with placebo displayed full digit abduction (DAS = 0) throughout the full observation period (data not shown) in contrast to all groups treated with BoNT/A. The treated groups showed a dose-dependently increased peak DAS from low to high dose groups (Table 1; Fig. 2E).

Table 1.

Digit abduction score (DAS) response observed on the injected leg after intramuscular administration of different BoNT/A doses with or without electric stimulation

Treatment	Dose/kg	No of animals	Mean Peak DAS	Duration of action (d)	DASAUC (DAS * d)
BoNT/A	20 U/kg	18	3.1 ± 0.1	16.1	36.9 ± 3.4
BoNT/A + ES	20 U/kg	18	3.7** ± 0.1	19.3	44.3 ± 2.6
BoNT/A	30 U/kg	10	3.4 ± 0.2	14.0	36.7 ± 5.1
BoNT/A + ES	30 U/kg	10	3.6 ± 0.1	16.8	42.6 ± 3.0
BoNT/A	40 U/kg	10	4.0 ± 0.0	23.3	56.1 ± 2.8
BoNT/A + ES	40 U/kg	10	4.0 ± 0.0	22.3	58.1 ± 1.8

Data are expressed as mean ± SEM (**p < 0.01 vs. non-stimulated group). Duration of action: Mean time to return to DAS₂₅; DAS_AUC : Total DAS response vs. time

Electrical stimulation combined with BoNT/A lead to an increase in BoNT/A effect in the 20 U/kg and 30 U/kg dose groups. The peak DAS after 20 U/kg BoNT/A was significantly increased over the unstimulated BoNT/A group on day 2 (Fig. 2D and E), after 30 U/kg BoNT/A group it was numerically but no longer significantly increased over the unstimulated BoNT/A group and after 40 U/kg BoNT/A it was not increased at all over the unstimulated BoNT/A group (Fig. 2E). The increase was 19% for 20 U/kg BoNT/A (mean peak DAS stim vs. non-stim 3.7 vs. 3.1; p < 0.01).

The higher peak DAS in the 20 U/kg dose group resulted in a 3.2 days longer duration of the mean time-to-return to DAS₂₅ (DAS = 1) and thus to a higher degree of muscle paralysis during that time in the stimulated group compared to BoNT/A injection only (Fig. 2D). In the 30 U/kg dose group, the duration of action until return to DAS₂₅ was < 3 days and in the 40 U/kg dose group < 2 days.

The area under the curve of the DAS over time curves (DAS_AUC) was slightly increased after ES in all BoNT/A dose groups. The highest increase was demonstrated by the group dosed with 20 U/kg by 20% (Table 1).

Magnetic stimulation

Impact of BoNT/A and magnetic stimulation on body weight

BoNT/A caused a dose-dependent transient decrease of body weight in the first 4 days after injection resulting in up to 2.5% mean loss in the 20 U/kg groups and up to 4.0% in the 30 U/kg groups (BoNT/A-treated vs. placebo p < 0.0001 for 20 and 30 U/kg). After this phase of decrease all dose groups gained weight at a similar rate (Fig. 3A-B). Magnetic stimulation did not further decrease the body weight as compared to BoNT/A only treated groups, irrespective of the BoNT/A dose of 20 U/kg or 30 U/kg. A placebo group without MS was not included because in a previous study the MS protocol did not cause weight loss per se (data not shown).

Fig. 3.

Impact of magnetic nerve stimulation after i.m. injection of BoNT/A on body weight and digit abduction score. A, B Relative mean body weight after BoNT/A injection with and without magnetic nerve stimulation (A, BoNT/A 20 U/kg; B, 30 U/kg). BoNT/A treated groups demonstrated an initial decrease in body weight followed by weight gain after day 4 (treated vs. placebo p < 0.0001 two-way RM ANOVA). Magnetic stimulation did not further decrease the body weight as compared to BoNT/A only treated groups, irrespective of the BoNT/A dose. Mean ± SEM, n = 10. C BoNT/A administered at 30 U/kg with or without magnetic stimulation caused initially the maximum DAS of 4. BoNT/A combined with magnetic stimulation caused an upwards shift of the DAS response curve from day 7 to day 25 (Mixed-effects model followed by Bonferroni’s multiple comparisons test stim vs. non-stim *p < 0.05 on day 23)

Impact of BoNT/A and magnetic stimulation on the digit abduction score

Mice treated with placebo and magnetic stimulation displayed unimpaired digit abduction (DAS = 0) throughout the full observation period. BoNT/A injection of 20 or 30 U/kg alone caused the maximum DAS score of 4.0 that can be achieved. BoNT/A combined with magnetic stimulation did not increase the DAS score in the 20 U/kg BoNT/A group (p > 0.05). The mean recovery time to DAS₂₅ and the DAS_AUC were unchanged by magnetic stimulation.

In the dose groups treated with 30 U/kg BoNT/A the combination with magnetic stimulation increased the DAS score from day 7 to day 25 reaching significance on day 23 (p < 0.05) (Fig. 3C). Consistently, the DAS_AUC of BoNT/A treatment plus magnetic stimulation was higher than with BoNT/A injection alone (see Table 2 for reference). The duration of action was slightly increased in the stimulated group of < 0.5 day (Fig. 3C).

Table 2.

Digit abduction score (DAS) response observed on the injected leg after intramuscular administration of different BoNT/A doses with or without magnetic stimulation

Treatment	Dose/kg	No of animals	Mean Peak DAS	Duration of action (d)	DASAUC (DAS * d)
BoNT/A	20 U/kg	10	4.0 ± 0.0	24.0	71.2 ± 1.7
BoNT/A + MS	20 U/kg	10	4.0 ± 0.0	23.9	68.9 ± 2.7
BoNT/A	30 U/kg	10	4.0 ± 0.0	24.5	77.5 ± 2.5
BoNT/A + MS	30 U/kg	10	4.0 ± 0.0	24.8	80.9 ± 1.7

Data expressed as mean ± SEM. Duration of action: Mean time to return to DAS₂₅; DAS_AUC: Total DAS response vs. time

Discussion

Botulinum neurotoxin type A (BoNT/A) blocks the release of acetylcholine from nerve terminals at the neuromuscular junction (NMJ) and causes reversible paralysis. This effect is used beneficially for the treatment of clinical conditions connected with involuntary muscle hyperactivity and contractions including spasticity [4, 7]. An increased uptake of BoNT/A into the presynaptic terminal of the neuromuscular junction could lead to increased efficacy and allow for reduced neurotoxin doses which is specifically relevant for pediatric patients because the dose per kg body weight up to 24 U/kg is close to the approved maximum dosing [44].

The effect of BoNT/A is nerve activity dependent and, subsequently, dependent on the neurotransmitter release [22]. This could be proven convincingly in ex vivo experiments [22] or in vitro in cell culture [27] which showed that BoNT/A endocytosis increases with synaptic activity. As transmitter release from motor terminals is increased by 4-aminopyridine even during tetanic stimulation in vitro [45] indicates that electrical or magnetic nerve stimulation in a spastic muscle may still increase the neurotransmitter release [28]. To our knowledge this concept has never been studied in vivo in an animal model but in several clinical studies [23, 46]. The results of these clinical studies were less consistent than that of the in vitro experiments, which may result from differing stimulation protocols and varying muscle activity in patients.

Thus, we tried to fill this gap with a well-controlled small animal study, which similar to patients exhibit varying muscle activity.

Electric stimulation

In this study the BoNT/A injection was immediately (< 2 min) followed by electrical peripheral stimulation of the corresponding motor nerves. It was evaluated in order to determine a stronger paralytic effect than after the injection of BoNT/A alone.

We found an enhanced DAS response after BoNT/A injection plus nerve stimulation vs. BoNT/A injection alone. The augmentation, however, decreased with increasing BoNT/A doses. The difference was statistically significant for 20 U/kg BoNT/A (p < 0.01) but no longer for 30 U/kg (p > 0.05) or 40 U/kg (p > 0.05). The finding indicates a ceiling effect of the DAS scoring reflecting that at a maximal BoNT/A dose for mice such as 0.8 LDU, the paralytic effect cannot be further enhanced by electrical nerve or muscle stimulation as it is already at a maximum level (score 4).

At the lowest dose 20 U/kg BoNT/A the augmentation by electrical nerve stimulation in DAS response resulted not only in a more pronounced but also in a 3.2 days longer duration of a muscle paralysis (DAS > 1). ES was performed within 2 min after the injection because BoNT/A uptake occurs within a few minutes after injection [7, 21, 22, 27, 28]. This is in line with clinical studies showing an improved effect using ES early after BoNT/A administration [23]. The observed longer effect duration of 3.2 days after BoNT/A injection combined with ES in mice corresponds to 19.9% increase in duration of effect compared to BoNT/A treatment alone. In a clinical setting in upper limb spasticity patients treated with BoNT/A ≤ 400 U this difference would translate into 17 days longer duration of effect (14.4 instead of 12 weeks [8]) if combined with electrical stimulation.

We also observed a dose-dependent transient decrease in body weight change after BoNT/A administration when compared to the placebo group. This phenomenon is in line with other observations after administering BoNT/A at 20–40 U/kg [47–49]. It has been associated with a decrease in locomotion compared to placebo-treatment [43]. Impaired mobility due to the local muscle relaxation reduces food and water intake during the first days after the injection [49]. In this study the weight loss by BoNT/A was not aggravated by concomitant electrical stimulation which is a bit surprising. It may be explained by the small effects that could be induced in mice and that were significant when assessing local paralysis but no longer when measuring a downstream behavioral effect like food intake.

Magnetic stimulation

Peripheral magnetic stimulation has so far not been considered as a booster for BoNT/A uptake although it is largely analogous to electrical stimulation. Several studies in which BoNT/A was administered together with repetitive TMS [32–34] aimed at improving neuronal plasticity rather than BoNT/A uptake. However, the stimulation protocol applied by Hirakawa [32] included a stimulation session immediately after BoNT/A injection which makes this work also relevant for the BoNT/A augmentation hypothesis. As this concept of augmented BoNT/A uptake has not been studied in vivo in an animal study we included this stimulation method as well. We think that peripheral magnetic nerve stimulation merits to be tested in patients with the aim to enhance neuronal BoNT/A uptake. The study protocol should include a concomitant application of BoNT/A and MS in order to provide the MS right at the time of BoNT/A uptake which takes place in the first 20 min after injection [22]. For this reason, the motor nerves were stimulated within 2 min after BoNT/A injection in this study. The appropriate timing is crucial for the success of this method.

In this study we investigated if 20 and 30 U/kg BoNT/A injection combined with magnetic stimulation would enhance the paralytic effect on the injected muscle. The highest dose BoNT/A 40 U/kg was omitted to avoid the ceiling effect observed for the combination of BoNT/A with electrical stimulation. The ceiling effect predominantly reflects the maximum paralysis induced by BoNT/A. In our paradigm 40 U/kg was the highest dose that could be injected intramuscularly into the gastrocnemius muscle of male BALB/cAnNRj mice without causing lethality. A stronger magnetic stimulation may have had a more pronounced effect but the technical effort would be huge and is, presumably, not required in larger species. As a consequence, we rather suggest performing follow-up studies at least in rats or even larger species.

In BoNT/A-treated groups with or without magnetic stimulation a dose-dependent transient decrease in body weight gain from day 1 to 4 after injection was observed which can be attributed to the partial hindlimb paralysis by the neurotoxin; the food intake behavior of the animals was initially impaired by the unilateral paralysis. Magnetic stimulation co-administered with BoNT/A injection had no additional effect on the relative mean body weight indicating that the magnetic stimulation procedure using two stimulation blocks of 10 pulses at 2 Hz did not impair the animals´ health.

A slightly increased DAS was found after magnetic stimulation in the higher dose groups (BoNT/A 30 U/kg) from day 7 to day 25 (p < 0.05 on day 23) indicating an enhanced potency of the neurotoxin by magnetic nerve stimulation. The peak DAS response was unchanged because all BoNT/A doses with or without magnetic stimulation reached the highest score of 4.0 reflecting the technical ceiling effect of the assay. The mean recovery time until DAS₂₅ was not increased after magnetic stimulation but there was a slight increase in the duration of action translating into an increased DAS_AUC with magnetic stimulation. The effect of BoNT/A 20 U/kg was not significantly enhanced by magnetic stimulation.

Compared to the electric stimulation higher DAS scores (DAS of 4 from day 1–4 in all groups) with a prolonged paralytic effect were observed after magnetic stimulation. This finding can be attributed to the direct administration of BoNT/A into the tibialis anterior muscle, the target muscle for DAS response. The test item is thus directly at the main site of effect without any diffusion to the target muscle which would lead to a later onset of the paralysis and a possible compound loss. Direct injection into the TA muscle was chosen because diffusion form the m. gastrocnemius to the m. tibialis anterior takes time, and we tried to coordinate the arrival of the neurotoxin at the neuromuscular junction in the tibialis anterior muscle with the magnetic stimulation of the respective motor nerve.

In the mouse paradigm a magnetic field strength of 2 Tesla (TeslaPulsGen 1 stimulator) or even 7 Tesla (MagPro X100 MO, MagVenture) was used in the pilot investigations which resulted in a rather small muscle contraction when stimulating peripheral nerves such as the sciatic nerve. This is, however, in line with the findings by Weissman et al., who calculated that the induced electrical field is decreasing in a linear manner dependent on the radius of a spheric electric conductor [50]. Due to the small size of the mouse the magnetic impulse can only induce a potential difference within a small volume which means that higher stimulation intensities are needed to exert the same effect as in larger mammals. In our pilot investigations we were unable to induce a consistent toe spreading by magnetic stimulation of the sciatic nerve, thus we decided to stimulate the spinal cord, in particular the cauda equina rather than the conus medullaris which, however, cannot be well discriminated in the magnetic field; the stimulation resulted in a strong and reproducible muscle response of the legs. As mentioned above, the reason behind may be the small diameter of the sciatic nerve fibers compared to the larger spinal cord fibers as the amplitude of the induced current in the peripheral nerve depends on its cross-sectional area (numerator in the equation) which means that the effect of magnetic stimulation is more pronounced in larger species with larger peripheral nerves.

The diameter of a myelinated nerve fiber in humans is 2–18 μm [51] for small and large fibers including more specifically diameters of 10–14 μm for alpha-motoneurons [52] or 10 μm in just the axon diameter for the peripheral posterior interosseous nerve fibers [53]. In contrast, the diameter of the majority (more than 70%) of myelinated fibers of the murine sciatic or tibial nerve is 2–4 μm [54]. The diameter of rat nerve fibers is 4 μm for an unmyelinated and 6 μm for a myelinated sciatic nerve fiber [55] and 3–6 μm for myelinated tibial nerve fibers in young rats [56] reflecting the differences in fiber diameters between species.

A higher stimulation frequency as used in our study or longer stimulation period could not be administered due to an overheating of the coil. We conclude that any observed DAS difference has to be considered as largely underestimated due to the selected small species and the technical limitations described above.

Our results indicate that magnetic stimulation in mice can augment the effect of BoNT/A by increasing the paralytic effect, as shown for the 30 U/kg dose, via an increased uptake of the toxin into the presynaptic terminals of the neuromuscular junction without impairing the animals´ health.

Limitations of this study are the species mouse due the small size as discussed above. It may also be discussed if a spastic muscle can be further driven at all to an increased uptake of BoNT/A by any sort of nerve stimulation because it is already more intensely innervated as compared to a relaxed muscle. Clathrin-dependent uptake and the early endosomal vesicle pathway into neurons both contribute to presynaptic BoNT/A uptake while the first is more common in neurons and activity-dependent and the second is also involved but not activity- dependent [57–59]. Overall, presynaptic neurotransmitter release is tightly regulated by autoreceptor feedback, enzymatic degradation and reuptake transporters which all together allow for an elevated but not unlimited neurotransmitter release [60] which means that even highly active neuromuscular terminals in spastic muscles are not unlimitedly releasing acetylcholine. Finally, muscle spasticity is not only mediated by increased acetylcholine release but also by increased afferent reflex tone which means that muscle innervation can still be short-term stimulated by electrical or magnetic stimulation.

Electrical and magnetic stimulation do not differ much from a physiological perspective. Magnetic stimulation purely induces nerve impulses whereas electrical stimulation also stimulations muscles directly [35, 36]. Magnetic stimulation is technically more challenging but on the other hand absolutely pain-free for the patients. Stimulators are available in most neurological departments for TMS applications, and a commercially available stimulator basically worked in our set-up but was just too weak for the mouse paradigm; for human or larger animals it presumably works flawlessly if using a suitable coil. The procedure can easily be adapted to peripheral nerve stimulation by adding an appropriate coil.

Conclusion

In this study an ES stimulation procedure or a magnetic stimulation protocol combined with BoNT/A treatment was established in mouse. BoNT/A treatment in combination with ES or MS demonstrated superior efficacy and duration of action compared to non-stimulated groups. An additive or synergistic effect can however only be observed if the treatment is not at its maximum dose because reaching the ceiling effect has to be avoided. Despite some limitations of the magnetic stimulation in mouse due to species differences and technical challenges our data provide valuable insights for the development of standardized magnetic stimulation protocols in order to boost the BoNT/A efficacy in clinical practice.

Abbreviations

AUC

Area under the curve

BoNT/A

Botulinum toxin type A

DAS

Digit abduction score

DAS₂₅

Digit abduction score 1, corresponding to 25% of max. paralysis score 4

ES

Electrical stimulation

IM

Intramuscular

MS

Magnetic stimulation

NMJ

Neuromuscular junction

rPMS

Repetitive reripheral nerve stimulation

TENS

Transcutaneous electric nerve stimulation

rTMS

Repetitive transcranial magnetic stimulation

TMS

Transcranial magnetic stimulation

Author contributions

S.H. performed the study, evaluated the data and drafted the manuscript, K.F. designed the study and edited the manuscript.

Funding

This work was funded by Merz Therapeutics GmbH.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author K. F. upon reasonable request.

Declarations

Ethics approval and consent to participate

Studies were performed in full compliance with the European Communities Council Directive 2010/63/EU. Experimental procedures were approved by the local authority for the use of animals for scientific purposes (reference 2347-13-2022-35-G, Regional Council of Brandenburg).

Consent for publication

All authors contributed to the article and approved the submitted version.

Competing interests

The authors declare no competing interests.