Spinal plasticity in the recovery of locomotion

Abstract

Locomotion is a very robust motor pattern which can be optimized after different types of lesions to the central and/or peripheral nervous system. This implies that several plastic mechanisms are at play to re-express locomotion after such lesions. Here, we review some of the key observations that helped identify some of these plastic mechanisms. At the core of this plasticity is the existence of a spinal central pattern generator (CPG) which is responsible for hindlimb locomotion as observed after a complete spinal cord section. However, normally, the CPG pattern is adapted by sensory inputs to take the environment into account and by supraspinal inputs in the context of goal-directed locomotion. We therefore also review some of the sensory and supraspinal mechanisms involved in the recovery of locomotion after partial spinal injury. We particularly stress a recent development using a dual spinal lesion paradigm in which a first partial spinal lesion is made which is then followed, some weeks later, by a complete spinalization. The results show that the spinal cord below the spinalization has been changed by the initial partial lesion suggesting that, in the recovery of locomotion after partial spinal lesion, plastic mechanisms within the spinal cord itself are very important.

Introduction

The study of locomotion offers several opportunities to investigate the various levels of controls by the central nervous system (CNS) of a robust and fundamental primitive motor act. I (S.R.) personally became fascinated by the subject of locomotion when I saw a film shown by Sten Grillner of a cat that had been previously completely spinalized at the low thoracic level as a kitten (Grillner, 1973). This cat could step with the hindlimbs on a small nature trail outside the lab while a research assistant held its tail to partially support the weight of its hindquarters and provide some lateral balance. Even more surprising, the spontaneous forward walking movements of the intact forelimbs were at times strong enough that the hindquarters would rise and the cat would make a few unaided steps with the hindlimbs before losing balance. Although there was a long history of the locomotor capabilities of spinal animals in many species (as well summarized by Grillner, 1981), this provided a model which could be investigated with modern tools of electrophysiology and which obviously provided a scientific framework that also had a great potential impact on spinal cord injured (SCI) patients. These observations were clearly showing that the spinal cord below a complete spinal section was capable of generating the basic pattern of locomotion with even some elaborate timing details. Therefore rehabilitation after SCI should strive to maintain or activate the sub-lesional spinal circuits. The concept of spinal generation of locomotion is robust (Delcomyn, 1980, Rossignol, 1995, Rossignol, 1996, Rossignol et al., 2000, Rossignol et al., 2002) and relevant even for other animal species such as the rat (Courtine et al., 2009, Gimenezyribotta et al., 2000), the mouse (Leblond et al., 2003), and humans (Bussel et al., 1988, Calancie, 2006, Dietz and Harkema, 2004, Gerasimenko et al., 2010, Harkema, 2008). The importance of spinal generation of locomotion was strongly revived more recently when we observed that, even after partial SCI (the most common lesion in humans), the recovery of hindlimb locomotion also depends to a great extent on changes that have occurred in the spinal circuits below the SCI (Barrière et al., 2008, Barrière et al., 2010, Rossignol et al., 2009). This observation is also of clinical importance since it emphasizes the possibility of profoundly modifying the spinal cord through rehabilitation strategies in humans after SCI. The aim of the present paper is to link various observations made over the years that lead to such conclusions.