Electrophysiologic assessment of sciatic nerve regeneration in the rat: Surrounding limb muscles feature strongly in recordings from the gastrocnemius muscle

Abstract

Striking inconsistencies between the results of morphometric and electrophysiologic examinations of the regenerating nerve were observed in a previous study featuring the bridging of a 14 mm gap in the rat sciatic nerve.

To shed light on this dichotomy, seven further rats were subjected to permanent sciatic nerve transection and assessed electrophysiologically, histologically and by retrograde axonal tracing at various postoperative intervals (1 h to 8 weeks).

The results of the histological examinations and retrograde tracing revealed that in spite of the fact that compound muscle action potentials could be recorded in the gastrocnemius muscle, no reinnervation of the gastrocnemius muscle, either physiological or aberrant, had actually taken place. Furthermore, it was established that the electrical activity recorded in the gastrocnemius muscle after stimulation of the proximal or distal stump is generated by surrounding hind limb muscles unaffected by denervation. These are stimulated either directly, or indirectly due to spreading of the impulse.

It is therefore strongly recommended that caution should be exercised when interpreting recordings from the gastrocnemius muscle after stimulation of a regenerating sciatic nerve in laboratory rodents.

Introduction

Weak correlations between functional and structural evaluations in sciatic nerve regeneration studies in rats have frequently been mentioned and described (Dellon and Mackinnon, 1989, Hadlock et al., 1999, Howard et al., 2000, Kanaya et al., 1996, Nichols et al., 2005, Shenaq et al., 1989); in this context it has been emphasised that abundant axonal regrowth and adequate impulse conduction do not necessarily trigger a return of function (De Medinaceli, 1995).

Recently, the reverse situation, i.e. almost full return of sensory function without any histological evidence of axonal regeneration in the corresponding nerve segments was observed in a study carried out to assess the ability of biocompatible collagen tubes to sustain neural regeneration across a 14 mm gap in the rat sciatic nerve. Rats experiencing no axonal regeneration and also no return of motor function, as determined by footprint analysis, regained close to full sensation in the foot of the operated hind limb. Retrograde tracers were able to prove the saphenous nerve as the primary source of sensory reinnervation of these animals (Rupp et al., 2007b). The same study also ascertained striking discrepancies between the morphometric examinations and the electrophysiologic assessments of the regeneration sciatic nerve, the latter of which were carried out to complement the functional investigations (data not shown). After 8 weeks of regeneration the motor nerve conduction velocities (NCVs) recorded in the operated hind limbs of all rats were statistically indistinguishable, even though morphometric and electron microscopic examinations of transverse sections of the regenerating sciatic nerve at mid-lesion level revealed that whilst some rats experienced abundant regeneration, others showed none at all. Further investigation of the distal stump and the regenerating tibial nerve 0.5 cm distal to its submersion into the gastrocnemius muscle showed, that only those animals exhibiting large amounts of countable myelinated fibres at mid-lesion level also possessed easily distinguishable and countable myelinated fibres at the more distal levels of examination (unpublished observations). No correlation could be found between the number of myelinated fibres and the NCVs for the operated hind limb of the individual animals.

Furthermore, the compound muscle action potentials (CMAPs), which were recorded from the gastrocnemius muscle (GM) after direct stimulation of the sciatic nerve proximal and distal to the lesion site, exhibited a similar morphology in all rats. These CMAPs were used to calculate the NCVs, since no CMAPs could be recorded in the interosseus muscles of the operated hind limbs in any of the rats.

Two different aetiologies can be put forward as the most probable reason for the large discrepancy between regained impulse conduction and nerve fibre counts observed.

The first is aberrant innervation of the GM by a different nerve (Gassel, 1964), such as one of the proximal branches of the sciatic nerve. The second possibility implies technical difficulties, such as accidentally recording CMAPs from other muscles of the hind limb, a problem also referred to as “cross-talk” (Kuiken et al., 2003).

To determine which of the two aetiologies was responsible for the inconsistencies observed, seven further rats were subjected to extraction of a 14 mm segment from the sciatic nerve with no repair of the defect. At different time-points they were assessed electrophysiologically, histologically and by using retrograde axonal tracers in order to determine the source of innervation of the GM or, alternatively, the source of electrical activity recorded in the GM.