Strategies to promote peripheral nerve regeneration: electrical stimulation and/or exercise

Abstract

Enhancing the regeneration of axons is often considered to be a therapeutic target for improving functional recovery after peripheral nerve injury. In this review, the evidence for the efficacy of electrical stimulation (ES), daily exercise and their combination in promoting nerve regeneration after peripheral nerve injuries in both animal models and in human patients is explored. The rationale, effectiveness and molecular basis of ES and exercise in accelerating axon outgrowth are reviewed. In comparing the effects of ES and exercise in enhancing axon regeneration, increased neural activity, neurotrophins and androgens are considered to be common requirements. Similarly, there are sex-specific requirements for exercise to enhance axon regeneration in the periphery and for sustaining synaptic inputs onto injured motoneurons. ES promotes nerve regeneration after delayed nerve repair in humans and rats. The effectiveness of exercise is less clear. Although ES, but not exercise, results in a significant misdirection of regenerating motor axons to reinnervate different muscle targets, the loss of neuromuscular specificity encountered has only a very small impact on resulting functional recovery. Both ES and exercise are promising experimental treatments for peripheral nerve injury that seem to be ready to be translated to clinical use.

Keywords: electrical stimulation; exercise and nerve regeneration; exercise or electrical stimulation; nerve stimulation; peripheral nerve regeneration.

Figure 1. The conversion of neurons and Schwann cells to a growth mode

A. Motoneurons undergo chromatolytic changes including the movement of the nucleus to an eccentric position reflecting B. the increased gene expression of several growth associated genes including neurotrophic factors such as glial and brain derived neurotrophic factors and their receptors, cytoskeletal proteins including tubulin and (in A) the reduced expression of transmitter associated genes and neurofilament. C. Schwann cells in the denervated Schwann cells participate in Wallerian degeneration of the axon and myelin debris and upregulate neurotrophic factors and the p75 receptors for neurotrophic factors at the same time as downregulating myelination associated genes that include P₀.

Figure 2. Staggered nerve regeneration

A; Just proximal to the site of surgical repair, regenerating axons visualized with silver staining may turn back and even loop (as described earlier by Cajal (1928). B. Silver stained regenerating axons cross the surgical site (not shown on the left) into the distal nerve stump at different rates due to C. the staggering of the regenerating axons at the suture site as shown by GFP+ fluorescent axons from the proximal nerve stump (left) crossing into the distal nerve stump on the right. D. Femoral nerves that had crossed a site of nerve transection and surgical repair and had just entered the denervated distal nerve stump were backlabelled with a fluorescent dye to reveal the staggered axon regeneration where motoneurons send their axons into the distal nerve stump over a protected period of a month. E. Backlabelling of rat femoral motoneurons that regenerated their axons 25 mm into either appropriate motor or inappropriate sensory nerve branches with two different dyes demonstrated that the numbers of the motoneurons send their axons equally into both branches initially but, after two weeks progressively more motoneurons regenerate their axons and these axons enter into only the appropriate motor branch. Motoneurons that are colabeled are those that regenerated axons into both branches, the numbers of these remaining constant throughout the 8 to 10 period in which all the motoneurons regenerate axons.

Figure 3. Electrical stimulation (ES) accelerates axon outgrowth across a site of nerve transection and surgical repair

A. Immediately following the transection and surgical repair of the rat femoral nerve, the nerve proximal to the suture site, was electrically stimulated at 20Hz for one hour. B. All femoral motoneurons regenerated their axons a distance of 25 mm into motor and sensory nerve branches within three weeks after the repair surgery and the electrical stimulation. This accelerated growth compares with the 8–10 week period when the femoral nerve was subjected to sham stimulation shown in Fig. 2F. C. Confocal images of whole mounts of the cut common peroneal nerve of a thy-1-YFP-H transgenic mouse repaired with a nerve graft from a wild type littermate contain profiles of regenerating axons. The subset of ~35 motor and sensory axons in this nerve that express yellow fluorescent protein (YFP) in these mice are highly visible against the dark background of the graft. Each image is a reconstruction the entire repaired nerve made by stitching together single 10 μm thick optical sections through the nerve. There was an obvious accelerated outgrowth of axons across the suture site and into the distal nerve stump when the common peroneal nerve was subjected to 20Hz 1 hour ES as compared to sham ES (Unstimulated).

Figure 4. Electrical Stimulation (ES) accelerates target reinnervation

Electrical stimulation at a frequency of 20Hz for one hour accelerates the outgrowth of axons across the site of surgical repair of transected nerve stumps to result in accelerated target reinnervation.

Figure 5. Preferential and delayed expression of neurotrophic factors in the denervated distal stump

A. Figurative illustration of the time course of femoral neurons regenerating their axons into the cutaneous and muscle branches with all the axons entering into the appropriate pathway at times later than three weeks coincides with B. the delayed upregulation of neurotrophic factors in the dorsal (sensory) and ventral (motor) roots. Heparin growth factor (HGF), brain derived neurotrophic factor (BDNF), glial derived neurotrophic factor (GDNF), nerve growth factor (NGF), insulin growth factors 1 and 2 (IGF-1 and IGF-2) and pleiotrophin (PTN).

Figure 6. Electrical stimulation accelerates the upregulation of regeneration associated genes that are essential for the effect

A. Following a period of one hour 20Hz electrical stimulation (ES), upregulation of neurotrophic factors and their receptors followed by upregulation of cytoskeletal proteins is increased and accelerated in the neurons. B. ES increases the mean (± SE) axon outgrowth into distal nerve stumps after suture of common peroneal nerve in a thy-1-YFP-H transgenic mouse to a distal nerve stump from either a wild type litter mate, from NT4/5 ^−/− transgenic mice, or acellular grafts. Hence, it is the neuronal neurotrophic factors and not those of the Schwann cells that account for the accelerated axon outgrowth after ES.

Figure 7. Daily exercise is more effective that electrical stimulation in promoting axon regeneration

The cumulative distributions lengths of axon profiles that regenerate from the proximal nerve stump of the common peroneal nerve of a YFP-transgenic mouse into a common peroneal isografts from a wild type litter mate two weeks after the surgical repair of the cut nerve. The curve was shifted to the right when the CP proximal nerve stump was subjected to electrical stimulation for 1 hour at 20 Hz. This shift was much greater when the mice were subjected to daily treadmill exercise. In this example, male mice were subjected to continuous training in which the mice were placed on a level, motor-driven treadmill at a speed of 10 meters/min for an hour daily. Each symbol is the average of six mice and the size of the symbol is proportional to the SEM. The horizontal dashed line at the 50^th percentile shows the location of the median axon profile length for each group.

Figure 8. The efficacy of the specific exercise regimens for male and female mice in promoting nerve regeneration are mediated by androgen receptors

A. Confocal images of the common peroneal nerves in male and female thy-1-YFP transgenic mice which express yellow fluorescent protein in the axons of ~35 motor and sensory neurons, two weeks after either continuous or interval training. Continuous training (slow walking at 10 meters/min for one hour per day) was effective in promoting nerve regeneration in males but not females and interval training (four repetitions of short sprints at 20 meters/min for 2 minutes following by 5 minutes of rest) was effective in females and not males. B. Flutamide, an androgen receptor antagonist released from Silastic capsules implanted three days prior to onset of treadmill training or electrical stimulation (ES) for one hour at 20 Hz eliminated the effects of both ES and gender-appropriate exercise (*) in both males and females. Median regenerating axon profile lengths, expressed as mean fold changes (±95% confidence limits) relative to untreated controls (vertical dashed line).

Figure 9. Gender-specific exercise sustains synaptic contacts on axotomized motoneurons

A. In this 1μm thick confocal image of a single retrogradely labeled motoneuron (CTB-AF545), excitatory synapses containing the vesicular glutamate transporter 1 (VGLUT1), and arising mainly from primary afferent neurons, and inhibitory synapses containing glutamic acid decarboxylase 67 (GAD67) were identified using immunofluorescence, and used to measure the percentage of the perimeter of the neuron in contact with the different immunoreactive structures. B. The mean (±SEM) percent synaptic coverage by VGLUT1 (top) and GAD67 (bottom) is reduced on motoneurons after sciatic nerve transection (untrained) but is sustained in male and female rats when ‘continuous’ and ‘interval’ treadmill exercise, respectively, is initiated three days after transection and continued for two weeks.

Figure 10. Misdirection of regenerating motor nerves is increased after brief electrical stimulation

A. The relative caudo-rostral distribution of motoneurons with axons in the mouse common peroneal (CP) nerve is shown for mice in which the sciatic nerve is intact and for mice four weeks after sciatic transection and repair. The normal caudal-rostral distribution of CP motoneurons in the intact mice (Intact) was shifted to a more caudal position in the spinal cord after sciatic nerve repair in ‘Untreated’ mice. This shift was exacerbated when the transected sciatic nerve was subjected to ES at 20 Hz for one hour but not when the mice were subjected to daily exercise, continuous training (slow walking at 10 meters/minute for one hour per day) in male mice and interval training (four repetitions of short sprints at 20 meters/min for 2 minutes following by five minutes of rest) in female mice. B. Pie charts showing that 94% of the compound muscle action potentials recorded from soleus muscles were elicited by stimulation of the tibial nerve but that after sciatic nerve transection and surgical repair, this was reduced to 72%. Selective electrical stimulation of the CP nerve branch at 20Hz for one hour (ES) reduced this further to 27%. Hence a one hour period of selective ES enhanced regeneration of cut CP axons so that only 27% of the axons regenerated within the appropriate CP nerve pathway with the remaining being misdirected to antagonist muscles that they did not formerly supply. C. The normal alternation of rectified and integrated electromyographic (EMG) activity in the flexor tibialis anterior (TA) muscle and the extensor soleus (Sol) muscle was lost after reinnervation, irrespective of whether the CP nerve was subjected to selective ES. The arrows below the EMG traces indicate lift and fall of the foot during locomotion.

Figure 11. Brief low frequency electrical stimulation accelerates nerve regeneration and muscle reinnervation after delayed repair

A. The median nerve in humans was subjected to one hour electrical stimulation (ES) at 20Hz following carpal tunnel release surgery after chronic injury. In the patients in which B. the median nerve was not stimulated and the number of median nerves with intact target muscle connections (MUNE) was ~50% of normal numbers (shown by the dotted line), the release surgery resulted in a small but insignificant increase in MUNE whilst in. C. where the median nerve was subjected to ES, there was already a significant increase above the ~180 intact motor units with the numbers increasing to normal within 6–12 months. D. In rats, the common peroneal and tibial nerves were cut and ligated prior to their cross-suture after two months delayed nerve repair. E. The same ES paradigm as used in the patients resulted in a significant increase in the numbers of E. common peroneal (CP) motoneurons and F. common peroneal sensory neurons that regenerated their axons into the tibial nerve stump.