Show Summary Details

Page of

PRINTED FROM the OXFORD RESEARCH ENCYCLOPEDIA, NEUROSCIENCE ( (c) Oxford University Press USA, 2016. All Rights Reserved. Personal use only; commercial use is strictly prohibited. Please see applicable Privacy Policy and Legal Notice (for details see Privacy Policy).

date: 24 June 2017

Plasticity of Stepping Rhythms in the Intact and Injured Mammalian Spinal Cord

Summary and Keywords

The spinal cord is a prime example of how the central nervous system has evolved to execute and retain movements adapted to the environment. This results from the evolution of inborn intrinsic spinal circuits modified continuously by repetitive interactions with the outside world, as well as by developing internal needs or goals. This article emphasizes the underlying neuroplastic spinal mechanisms through observations of normal animal adaptive locomotor behavior in different imposed conditions. It further explores the motor spinal capabilities after various types of lesions to the spinal cord and the potential mechanisms underlying the spinal changes occurring after these lesions, leading to recovery of function. Together, these observations strengthen the idea of the immense potential of the motor rehabilitation approach in humans with spinal cord injury since extrinsic interventions (training, pharmacology, and electrical stimulation) can modulate and optimize remnant motor functions after injury.

Keywords: locomotion, central pattern generator (CPG). spinal cord injury (SCI), locomotor training, kinematics, electromyography (EMG), spinal learning, sensory control of locomotion, descending pathways, neuropharmacology


The lyrics of the song “Do the Loco-Motion” offer a refreshing view of locomotion: “My little baby sister can do it with ease/ It's easier than learning your a b c's/ … Do it nice and easy now don't lose control/ A little bit of rhythm and a lot of soul.” Locomotion is “easy” because it stems from inborn neural circuitry on which learning and practice build during development. Locomotor movements must be expressed in several variants or formats to match our immediate or long-term intentions (soul) or external conditions such as walking on a dry or an unplowed sidewalk. Several structures of the central nervous system (CNS) are involved in such controls, and lesions of the CNS will perforce reduce the range of possibilities. However, through mechanisms of neuroplasticity, optimal remnant locomotor functions will be reexpressed.

This chapter seeks to identify some of the basic and efficient control mechanisms involved in locomotor control and their adaptation in normal conditions or after lesions. “Normal” patterns of locomotion are adapted to various voluntary demands (speed, direction, precision) and sensory afferents, as well as supraspinal commands from the cerebral cortex and brainstem, including neuromodulators; modify the basic spinal locomotor circuits in a coherent manner to meet these demands; and at the same time maintain the progression from A to B. After lesions, whether traumatic, neurodegenerative, or genetic, the balanced interactions of these control mechanisms are challenged, and therefore optimization of locomotion must rely on the reorganization of remnant neural structures.

Readers less familiar with this field may wish to consult previous reviews on locomotion generation and sensorimotor interactions during locomotion (Edgerton et al., 2004; Grillner, 1981; Kiehn, 2006; Rossignol, 1996; Rossignol et al., 2006; Rossignol & Frigon, 2011; Rossignol et al., 2014). The generic locomotor pattern involves an alternation between flexor and extensor muscles of one limb in two main phases of stepping (swing, stance), an interlimb coupling (out-of-phase or in-phase) between the two forelimbs or the two hindlimbs, as well as between fore- and hindlimbs. The basic defining features of this generic locomotor pattern must be flexible to accommodate body displacements in order to achieve various goals or meet external constraints. Thus, normal adaptive changes of swing, stance, joint excursions, and interlimb coupling must occur. This intrinsic flexibility implies that the locomotor circuitry itself has some degree of built-in functional plasticity in the normal mammalian spinal cord. The extent of this plasticity becomes even clearer after injury of the nervous system, namely, that of the spinal cord. In this short review, the term “plasticity” is taken to mean that some of the basic features of locomotion can change (acutely or chronically) to meet behavioral demands or to compensate for missing controls after lesions. These plastic changes not only involve alterations in muscle discharges and kinematics but also modifications in sensory afferent transmission, intrinsic spinal circuits, and supraspinal-spinal interactions. These changes can be anatomical (sprouting and/or regeneration of axons), physiological (phase and state-dependent interactions of sensory and supraspinal circuits), or biochemical (membrane and receptor properties).

This article first presents a brief overview of the adaptive capacities of stepping rhythms in mammals with an uninjured spinal cord when exposed to a variety of external conditions and, second, describes locomotor adaptation after various types of spinal lesions.

Adaptation of stepping patterns to behavioral demands in animals with intact spinal cord

The Normal pattern of forward walking

Single-limb kinematics

Plasticity of Stepping Rhythms in the Intact and Injured Mammalian Spinal CordClick to view larger

Figure 1. General methodology for the study of locomotion in cats.

A: Cats are placed on a motor-driven treadmill. Pairs of electromyographic EMG electrodes are led subcutaneously and are implanted chronically into various flexor (red) and extensor (blue) muscles in the hind- and forelimbs (only four are represented here). The multi-pin EMG connector linked to amplifiers is cemented to the skull. An intrathecal cannula with an inlet also cemented to the skull is led through the atlanto-occipital ligament down to the lumbar level. Reflective markers are placed at various joints on the limbs, and the angle measurements are taken in with the indicated orientations. For each video field (16.7 ms between fields when videorecorded at 60 frames/s), the coordinates of the reflective markers are obtained, and the hindlimb movement is reconstructed as indicated in the kinematic model (MTP = metatarsophalangeal joint). The digital (SMPTE) time code (upper right) is used to synchronize video images and EMG recordings. The complete spinal lesions are made at T13 unless otherwise specified, and the partial lesions at T10-T11. With complete lesion of the SC at T13, cats are positioned with the forelimbs standing on a platform fixed above the belt, while the hindquarters are free to walk on the moving treadmill. With partial lesion, the cat is free to walk with all four limbs on the treadmill as shown here.

B: From joint markers’ coordinates, the movements of the limbs during swing (red sticks) and stance (blue sticks) phases of each cycle are reconstructed. Note that to prevent overlap of the sticks when reconstructing the locomotor movement, each frame is displaced by an amount equal to the displacement of the foot along the horizontal axis. The colored arcs around each joint represent the orientation of the measurements illustrated in C.

C: Mean angular excursions of the hip, knee, ankle, and metatarsophalangeal (MTP) joints of 10 cycles of the left hindlimb at 0,8m/s. Angular values decrease in flexion and increase in extension for all joints. Note that some skin slippage of the knee marker accounts for the apparent angular decrease during stance rather than an increase as it should be represented for extension. The color code corresponds to the arc around each joint. Sub phases of the cycle are defined as Flexion (F), First Extension (E1) for swing and Second Extension (E2), and Third Extension (E3) for stance as proposed (Philippson, 1905). This figure is partly inspired by Rossignol and Frigon, 2011.

The step cycle is made up of different phases, all of which can be modified to some extent to meet particular demands. These various components of the cycle must first be described to understand where adaptation can occur. The kinematic and electromyographic (EMG) characteristics of the “generic” normal locomotor patterns in quadrupeds have been described before (Grillner, 1981; Halbertsma, 1983; Rossignol, 1996; Trank et al., 1996). This cycle corresponds, in the cat, to a leisurely walk at around 0.4 m/s. In brief, the stride of one limb is subdivided into two phases, swing and stance. During one complete step, however, when the proximal hip joint undergoes one flexion and one extension oscillation movement, bringing the limb forwards and backwards, the knee and ankle joints perform more complex movements. This is reflected in the currently accepted subdivision of the swing and stance periods in two subphases (Philippson, 1905). Figure 1 illustrates the basic methodology we have used to quantify locomotion in the cat on a treadmill, allowing us to measure kinematics synchronized to electromyographic activity during movement and also to inject drugs through an intrathecal cannula (similar set-ups have been done in rats (Alluin et al., 2015). At the onset of swing (Fig. 1B), all joints are initially flexed (F), as seen also in the stick representation and angular excursion graphs (Fig. 1C) below the stick figure. Then, while the hip continues its forward movement in flexion, the knee and ankle extend (upward deflection at around midswing) in preparation for foot landing and weight acceptance. This is termed the First Extension period (E1). At foot contact, the backward motion of the hip starts but the knee and ankle flex, as weight is partially borne by that limb. This period of flexion during the first part of stance is defined as the Second Extension period (E2) (note that in Fig. 1C, the extension of the knee that should be represented by an up-going displacement of the trace is partially hidden by a knee marker slippage). To complete the cycle, all joints extend (Third Extension phase or E3) during the terminal push-off phase. The distal metatarsophalangeal (MTP) joint, as the hip joint, also performs mainly one dorsiflexion and one plantarflexion. The plantarflexion starts toward the end of stance as other more proximal joints flex and ensure foot lift-off. This plantarflexion continues somewhat past lift-off, and then the MTP starts a dorsiflexion in prevision of foot landing and continues during the E1, E2, and E3 parts of the stance phase (Trank & Smith, 1996). The foot touches ground at a precise point in front of the hip, so that the combination of all joint angles is precise and repeatable at different speeds unless the animal is forced to place the foot on a rung or before or after an obstacle, as will be seen later in this article. Foot positioning at lift-off is more variable since presumably most of the adjustments can be made in the succeeding swing phase to ensure correct foot positioning at the next touch-down (Halbertsma, 1983). As will also be seen later, foot lift-off and foot touch-down can be translated backwards to accommodate a crouching position during walking (Trank et al., 1996) so that these kinematic events do not necessarily have a fixed relation to the hip joint since the overall length of the limb can change depending on posture.

Early work on decerebrate cat walking on a treadmill using electrical stimulation of the Mesencephalic Locomotor Region (MLR) (Shik et al., 1966b) has shown that, as speed increases, the stance phase decreases linearly, whereas the swing phase decreases proportionally less or remains relatively constant (Halbertsma, 1983; Kulagin & Shik, 1970; Shik et al., 1966a). This has been established as a standard of locomotion (Grillner, 1981). It is also of interest that the F and E1 phases in the hindlimbs are not necessarily equal in all joints so that F can increase and E1 decrease after a perturbation of the forelimbs, suggesting a somewhat independent control of flexion and extension neural circuits (Halbertsma, 1983) and thus a potential for plasticity with regard to foot positioning. This small variation in the onset of swing at all joints is well seen in the angular excursion diagram where the vertical line indicating the start of swing is a compromise for the four different joints.


Plasticity of Stepping Rhythms in the Intact and Injured Mammalian Spinal CordClick to view larger

Figure 2. Electromyography and duty cycles

A: The EMGs are recorded in both fore- and hindlimbs (only ankle extensors on both sides are represented to show alternation). The EMGs are full-wave rectified and averaged over 10-15 cycles. EMGs are normalized to the step cycle and are taken from different cats to best represent the characteristics of the bursts. The full vertical line indicates foot contact of the left hindlimb. Muscles are labeled as ipsilateral (i) or contralateral (c) to the camera (or often to the lesion side in hemisection experiments usually performed on the left). The muscles are: St, Semitendinosus (knee flexor/Hip extensor); Srt, Sartorius (knee extensor, hip flexor); VL, Vastus Lateralis (knee extensor); Gastrocnemius Medialis (GM, ankle extensor and knee flexors); Tibialis anterior (TA, ankle flexor). Extensor Digitorum Brevis (EDB) acting as a dorsiflexor of the digits; Brachialis, an elbow flexor, and Triceps Lateral (Tri), an elbow extensor.

This part of the figure was inspired by Rossignol and Frigon (2011.

B. Gait diagram derived from 19 step cycles in a cat walking at a steady speed of 0,4m/s. The average step cycle is repeated twice (0 to 1 and 1 to 2) so as to represent events at transitional phases.The order of the limbs follows that of Hildebrand (Hildebrand, 1980) and has the merit of clearly indicating the smooth sequence of Left Hindlimb (LH), Left Forelimb (LF), Right Hindlimb (RH), and Right Forelimb (RF). The period during which each limb is in contact with the belt (stance) is represented by a horizontal line, while the gaps in between the lines represent the swing phase. To simplify the illustration, a mean standard error of 0,025 (expressed as a phase value of the cycle) was indicated at each contact and lift because all the SDs fell within a range of 0,02 to 0,3. Note the strict alternation (0,5 phase) between LH and RH contact (similar out-of-phase coordination can be measured for the forelimbs). Note also that in both fore- and hindlimb pairs, some overlap indicates a dual support phase. Finally, there are several episodes of tripod walking with different combinations of fore- and hindlimb support.

Flexor and extensor muscles are associated with the various stride phases as seen in Figure 2. Flexor muscles generally show more complex patterns of discharge and, contrary to extensor muscles, may have different timing and profiles. For instance, the bifunctional muscle Sartorius (iSrt) discharges more or less throughout swing as a hip flexor but has a knee extensor (anterior part) and knee flexor (medial part) component and can thus discharge two bursts of activity during swing and stance (Pratt & Loeb, 1991).The knee flexor–hip extensor muscle Semitendinosus (St) discharges two distinct bursts, one at the very end of stance to initiate knee flexion and one at around foot contact. The ankle flexor Tibialis Anterior (iTA) discharges during the whole swing period. Extensor Digitorum Longus (EDL) also discharges during the whole swing phase, whereas Extensor Digitorum Brevis (EDB) discharges a burst just before foot contact and often a second burst during the stance phase (Fig. 2 and (Trank et al., 1996)). Extensor muscles discharge somewhat more uniformly, although their timing of onset, peak, and offset may vary. For instance the ankle extensors Gastrocnemius (iGM) burst starts before foot contact and peaks in the early stance phase, whereas the knee extensor, Vastus Lateralis (VL), is slowly recruited at around foot contact and reaches its peak activity in the later part of stance (see Fig. 2A and Halbertsma, 1983; Trank et al., 1996). Two forelimb muscles are included in Figure 21 (Bra, flexor and Tri, extensor around the elbow). A more thorough summary of EMG discharges during forward walking is given elsewhere (Rossignol, 1996; Trank & Smith, 1996; Widajewicz et al., 1994). Already, it can be understood that the locomotor pattern is not simply an alternation between flexors and extensors but rather is a complex pattern with detailed features of timing and EMG profiles. These characteristics can change with imposed demands on locomotor performance.

Interlimb coupling and gait patterns changes with speed

Gaits can be described by a foot-fall formula which illustrates the feet in contact with the ground at a given moment and speed, or more simply by a gait diagram (see Fig. 2B) which details the whole duration of stance and swing phases as horizontal bars for each limb and permits a rapid evaluation of their coupling (Hildebrand, 1976) at all times. In relation to our previous idealized description of a leisurely walk, the sequence of foot falls in a complete stride would be as follows: Left Hindfoot (LH), Left Forefoot (LF), Right Hindfoot (RH), and finally Right Forefoot (RF). The kinematic events between two homologous limbs (hindlimbs or forelimbs) are out-of-phase by 180 degrees. Hildebrand has described a variety of limb couplings in symmetrical gaits (such as trot, pace) in which homolateral or diagonal couplets are in ground contact and asymmetrical gaits such as in various forms of gallop (bound, half-bound) in which homologous limbs (fore or hindlimbs) are coupled in-phase (bound) or in which hindlimbs are in-phase while forelimbs maintain more or less an out-of-phase coupling. These gait changes can accommodate a variety of speeds. Other studies have also emphasized the flexibility in interlimb coupling, accounting for various gait patterns (English, 1979).

Interest has recently been raised in mouse locomotion in which gait patterns are clustered around defined patterns (e.g., trot, walk) when speed increases (Lemieux et al., 2016). These gait clusters may be subtended by specific neuronal circuits, as indicated by genetic ablations of key interneurons which can abolish certain gaits altogether, as will be described later in this article (Bellardita & Kiehn, 2015).

Walking in different conditions

Although one can describe a ”generic” locomotor pattern for simple undemanding (leisurely) forward treadmill locomotion in a quadruped, kinematic and EMG features can change within certain ranges to accommodate a variety of locomotor conditions encountered in daily life. Therefore, there is already a built-in provision for adaptive changes when the spinal cord is intact. Some of these special walking conditions and their associated kinematic and EMG changes are summarized here: walk with differential speeds on both sides mimicking turning; walk in a crouched position mimicking stalking; walk in a backward direction; walk on uncertain terrain, precision walk in which the feet have to be placed repetitively and precisely on restricted targets such as rungs on a ladder; and, finally, walk while accommodating incoming obstacles.

Walk with a differential speed

Between left and right sides

When animals turn, say to the left, the limbs on the right must cover a larger distance than the left limbs. To maintain a 1:1 rhythm, the interlimb asymmetry must be corrected by adjusting the subphases of the step cycle. Such asymmetries were studied in decerebrate cats, which can maintain a 1:1 coupling for a rather large speed differential until the animal then makes two steps on the fast side for one step on the slow side (2:1 rhythm) (Kulagin & Shik, 1970). This was studied in more detail in intact cats walking on a treadmill (Halbertsma, 1983). While keeping a 1:1 coupling, the fast limb makes a longer swing and a shorter stance, whereas the slow limb makes a shorter swing and a longer stance. In the slow limb, the touch-down often occurs more caudally than predicted, again indicating some flexibility in the time of touch-down. Similar statements can be made on the position of the foot at lift-off. It appears that events occurring at or around touch-down in one limb provide important signals for the contralateral homologous limb. The overall rhythm generated by all four limbs cannot necessarily be predicted by the speed of any one belt but appears as a compromise of complex interactions between the limbs (D’Angelo et al., 2014; Frigon et al., 2015; Halbertsma, 1983). When the differential speed increases, the fast limb can make two unequal steps, one during swing and the other during stance of the slow limb.

Between fore- and hindlimbs

Locomotor activity in fore- and hindlimbs can interact in both directions through propriospinal interactions, even though these might not be symmetrical (Juvin et al., 2012; Thibaudier & Hurteau, 2012). To study cervico-lumbar interactions during locomotion, studies were made on intact cats walking on a transverse treadmill in which both forelimbs walk at one speed and both hindlimbs at the same speed as the forelimbs or at higher or lower speeds (Thibaudier & Frigon, 2014; Thibaudier et al., 2013). When the hindlimbs walked faster than the forelimbs (up to a speed ratio of 3), forelimbs and hindlimbs could maintain a 1:1 coupling by altering their respective stance and swing phases or else adopt a 2:1 coupling in which the forelimbs performed two steps per hindlimb step cycle. However, when the forelimbs walked faster, the coupling with the hindlimbs was disrupted, and various coupling between fore and hindlimbs could occur. It should be recalled indeed that during locomotion, the fore-hindlimb neuronal coupling is somewhat flexible and probably accounts for the variety of observed gait patterns (Miller et al., 1975).Therefore, this suggests a predominant role of the movement generated at the cervical enlargement on the lumbosacral enlargement during locomotion. However, the overall tonic effects of the hindlimbs on the expression of forelimb locomotion is quite important, as suggested by experiments in fictive locomotion of the forelimbs whose vigor is enhanced by a spinalization at the low thoracic level (Saltiel & Rossignol, 2004).

Walking in a crouched position, on inclines, or backwards

Walking with different postures, such as crouching or walking downslope or upslope on a treadmill, forces the locomotor control system close to its limits and illustrates how much the parameters of the step cycle can be modified.

During crouched walking imposed by fixing a ceiling above the treadmill, cats have to modify the overall limb excursion as well as foot positioning to accommodate the reduction of hip height relative to foot contact. Basically, the overall shape of joint profiles is preserved, but the joint excursion amplitude is increased as well as the exact moment of change in movement direction of individual joints onset of hip flexion, of knee extension. At touch-down, the foot does not reach as far in front of the hip. Bursts of flexor muscles are longer, and some foot flexor muscles may show an enhanced discharge during stance (Trank et al., 1996). Significant changes in the kinematics and muscle discharge are also observed during upslope walking (Carlson-Kuhta et al., 1998) and downslope walking (Smith et al., 1998). Of interest, when walking upslope, double joint muscles such as St (hip extensor and knee flexor) can show an enhanced flexor discharge but also a prominent activity during the stance phase, which is even more prominent during gallop. During downslope walking (20 percent grade), there may even be co-activation of hip flexor iliopsoas (IP) and knee extensor VL, suggesting a high degree of flexibility in the recruitment of antagonist muscles to meet the behavioral demands. During backward walking, several changes occur in the discharge period and/or the discharge profile. For instance, St can discharge throughout swing, and specific muscles such as Flexor Hallucis Longus (FHL) may switch the timing of its single-burst discharge from the onset of swing to the onset of the stance phase (Trank and Smith, 1996).

These examples indicate that neural circuits controlling locomotion must possess enough intrinsic plasticity to accommodate a variety of behavioral demands such as are needed during crouching, up- or downslope or even backward walking. It was proposed (Grillner, 1981) that neural circuits generating locomotor patterns might be composed of spinal control units for flexor and extensor muscles acting at each joint. The possibility of coupling these units differently accounted for most of the changes seen in various behavioral demands (Smith et al., 1998). One can define various global synergies in which muscle discharges are grouped together and which can even include muscles of antagonistic functions. It has been considered that the CNS might have to deal with only a few synergies (spinal primitives) and that by adjusting the weight of some components in the synergy, a large number of movements can be achieved (Saltiel et al., 2001).

However, an exhaustive analysis of EMG discharge of the fore- and hindlimbs of the cat during treadmill walking has shown that discrete muscle discharges can be regrouped in sequential clusters (Krouchev et al., 2006), thus potentially allowing discrete changes in some elements of a given cluster to meet behavioral demands such as negotiating an obstacle (Drew et al., 2008b). This model fits much better the observations of supraspinal discharges from the motor cortex or the Red Nucleus, which are discrete and appropriate to change specifically some components of these synergies during execution of the movement.

Precision walking

Obviously, precisely positioning the foot in a constrained environment such as the rungs of a ladder or overcoming an obstacle requires significant adaptations of the locomotor pattern. Numerous devices have been used to study these adaptations, such as a circular ladder treadmill (Amos et al., 1987), obstacles on a treadmill (Drew, 1988), a fixed horizontal ladder or a walkway with obstacles (Beloozerova & Sirota, 1993), or a moving horizontal ladder (Rossignol et al., 2015). A number of these studies have implicated supraspinal structures (Armstrong, 1986), especially those using obstacles and the visuomotor strategies needed to overcome incoming obstacles (Drew, 1993; Drew et al., 2002; Widajewicz et al., 1994).

In line with studies indicating that locomotion results from the sequential discharge of synergists at given points of the step cycle, precision walking may be achieved by controlling only certain features of the clusters (sparse synergies) to achieve foot positioning (Krouchev & Drew, 2013). As we will discuss later, the precise timing of these events is most often perturbed after spinal lesions.

Taken together, the above descriptions clearly illustrate that a great deal of functional plasticity exists in the locomotor circuitry of the normal uninjured spinal cord to accommodate daily behavioral demands. It will now be seen that after various spinal lesions, neuroplastic changes in the CNS, namely, in the locomotor circuitry, allow quite impressive recovery of locomotion. and that a great deal of this adaptation occurs in the spinal cord together with other changes in remnant descending controls.

Plasticity of stepping patterns after spinal cord injury (SCI)

Plasticity of Stepping Rhythms in the Intact and Injured Mammalian Spinal CordClick to view larger

Figure 3. General scheme of locomotor control

At the center of the figure is the spinal cord with interneurons and motoneurons innervating flexor (F in red) and extensor (E in Blue) muscles. Groups of interneurons constitute a Central Pattern Generator (CPG), while other interneurons involved in transmitting sensory or descending inputs may be (or not) within the sphere of influence of the CPG. Descending pathways innervating flexor or extensor–related circuits course through the Dorsolateral Funiculus (DLF) or the Ventrolateral Funiculus (VLF) and can be rather specifically lesioned. DC stands for dorsal columns. Neurotransmitters released by descending pathways or local neurotransmitters that can change the properties of the spinal circuits are indicated. Sensory afferents from the skin or muscles (proprioceptive IA, IB, and II) impinge on interneurons and motoneurons. On the right side of the figure are illustrated only a few flexor and extensor muscles that can be selected in various flexor/extensor synergies involving different joints. The spinal section is meant to represent a rather rostral lumbar segment (L3), which are considered more active in rhythmogenesis. Adapted from Dancause, N., Nadeau, S., & Rossignol, S. (2015). Sensorimotor rehabilitation: at the crossroads of basic and clinical sciences. Amsterdam: Elsevier, with permission from Elsevier.

Adaptation to external demands depends partly on intrinsic spinal mechanisms and on their supraspinal modulation exerted through pathways coursing in various quadrants of white matter. It is therefore of interest to describe the effect of different types of spinal lesions on the expression of locomotor patterns, as well as the potential of external manipulations (such as training, drugs, and electrical stimulation) to modulate these patterns after spinal lesions.

The general framework illustrated in Figure 3 suggests several mechanisms that may come into play after an SCI. At the core is a spinal circuitry (labeled CPG) composed of interconnected interneurons generating a locomotor rhythmic activity and projecting to flexor and extensor motoneurons as well as to a fringe of several classes of interneurons that are part of reflex circuits or intrinsic spinal circuits. The excitability of these interneurons may thus be cyclically modulated by the CPG, even though they may not be part of the rhythmogenesis per se. On this spinal circuitry impinge several afferent fibers originating from the skin as well as from various muscle receptors (IA, II from muscle spindles and 1B from Golgi Tendon Organs) from the limbs (Rossignol et al., 2006). These spinal circuits are also modulated by major descending pathways interconnecting other spinal levels (propriospinal), brainstem (vestibulo, reticulospinal, rubrospinal pathways) as well as cortico or cortico-brainstem pathways. Finally, spinal circuits are controlled by neurochemically defined descending pathways whose cells of origin are in the brainstem and which synthesize, for instance, serotonin (5HT in the Raphe Nuclei) or Noradrenaline (Locus Coeruleus). These neurotransmitters can change the state of spinal circuits by inducing changes in membrane properties, and their release in the spinal cord can be almost completely abolished by spinal lesion. However, receptors of these neurotransmitters are still present on spinal cells and can be activated or inhibited by agonists or antagonists of the neurotransmitters.

The following paragraphs review complete or partial spinal lesions severing the axons of these structures. The lesions most often studied are produced surgically and are either complete or incomplete lesions produced by a lateral hemisection (unique or staggered) or a bilateral dorsal/ventral hemisection. Other lesions mimicking more closely human spinal cord injury use impact by weight drop or other controlled mechanical impactors (Basso et al., 2002; Young, 2002). A more recent approach consisting of genetically ablating spinal interneurons or afferents is also briefly discussed later.

Recovery of hindlimb locomotion after a complete spinal cord section

There is a long history relating to the recovery of hindlimb stepping after a complete spinal lesion, and it has been extensively reviewed (Grillner, 1981; Rossignol, 1996; Rossignol et al., 2006). A seminal observation remains that hindlimb locomotion can be observed in kittens, even when spinalized before they have learned to walk (Grillner, 1973), and that this pattern could be maintained and could evolve over several months (Bregman & Goldberger, 1983; Forssberg et al., 1980a, 1980b). Neonatally spinalized kittens may also maintain some form of transient and irregular quadrupedal walk during the first 5 months (Grillner, 1973; Howland et al., 1995). A key finding in cats has been that the spinal cord itself, after an acute spinalization and in absence of movement-related sensory feedback (curarization), can generate a detailed alternating pattern between flexor and extensor muscles of both hindlimbs, as recorded from specific muscle nerves (fictive locomotion) after injection of L-dopa, a noradrenalin precursor (Grillner & Zangger, 1979). Acutely spinalized adult cats can walk with the hindlimbs on a treadmill after injecting clonidine, an alpha-2 noradrenergic agonist (Forssberg & Grillner, 1973).

Behavioral observations after chronic SCI

Cats. In adult cats, spinalization at T13 initially induces complete hindlimb paralysis. Locomotor training on a treadmill for 2–3 weeks can improve several aspects of the locomotor pattern (Barbeau & Rossignol, 1987; Lovely et al., 1990), so that placing the hind paws on a moving belt suffices to trigger and maintain hindlimb locomotion with spontaneous partial weight bearing of the hindquarters. The cat can adjust its speed to that of the treadmill and the subphases of the step cycle (swing, stance) change with speed with very similar characteristics to normal animals. Dynamic equilibrium of the hindquarters can be maintained for a number of consecutive steps by simply holding the thorax or the tail. It was also shown that several of the EMG characteristics of hindlimb locomotion seen in the intact hindlimbs were preserved in the spinal cat (Bélanger et al., 1996; Edgerton et al., 1991). Even muscles such as St preserved its characteristic double burst.

In humans, the level of spinal cord injury determines the degree of locomotor functionality (Dietz et al., 1999). This leads to the question of the importance of postural thoracic muscles since several muscles span large parts of the trunk. The thoracolumbar junction appears to have a particular importance for rhythm generation (Antri et al., 2011; Cazalets, 2000; Kiehn, 2006; Kiehn & Kjaerulff, 1998; Kjaerulff & Kiehn, 1996), but the lumbosacral region still appears to be of paramount importance (Beliez et al., 2015). In a series of experiments using spinal lesions at different lumbosacral levels in the cat, it appears that rostral lumbar segments must be functional (rostral to L4) to observe an organized multisegmental muscle activation during the locomotor pattern (Langlet et al., 2005), which does not preclude the presence of alternating flexor-extensor activity in single segments (Grillner & Zangger, 1979).

Important postural deficits have been observed in chronic low spinal cats (Macpherson & Fung, 1999) and spinal rabbits (Lyalka et al., 2009), indicating that although transient balance can be achieved, the thoracic segments above the lesion level are mainly involved in the circuitry for balance control. Intrathecal injections of 5-HT agonists in severely lesioned rabbits showed a limited increase in reflex postural control probably, resulting mainly from an increase in spinal excitability (Lyalka et al., 2008).

Rodents. Neonatally spinalized rats (but not weanling rats) can develop a well-coordinated lumbosacral rhythmic pattern and also various forms of quadrupedal walking through some mechanical coupling between fore- and hindlimbs (Stelzner et al., 1975; Udoekwere et al., 2016). The spinal ability of rats to generate a spontaneous (i.e., without additional stimulation) locomotor pattern after complete spinalization as adults has been questionned (Gimenez y Ribotta et al., 2000; Majczynski & Slawinska, 2007), except when locomotor-enabling conditions are provided such as intrathecal drug administration, spinal electrostimulation, robotic weight support ,and locomotor training (Courtine et al., 2009; Gerasimenko et al., 2007; Ichiyama et al., 2008; Van den Brand et al., 2012). Experiments on P1/P2 rats show that weight-bearing locomotion of the hindquarters can be achieved, and cortical ablation of trunk-related cortex suggests that it participates in the regaining of weight support through activation of trunk musculature spanning the lesion level (Giszter et al., 2008). Even in adults, cortical lesions of the trunk areas affect the locomotor stability of the hindquarters. Intracortical microstimulation (ICMS) of the trunk cortex shows a significant extension of the cortical representation of the trunk after SCI (Oza & Giszter, 2014). Robotic trunk training in adult rats transected as neonates induced important changes in the cortex as shown with ICMS (Oza & Giszter, 2015). The importance of trunk muscles could relate to enhancement of spinal locomotion in adult rats trained with an upright posture (Courtine et al., 2009; Slawinska et al., 2012). Our recent work in completely spinalized rats has shown, however, that an intensive daily locomotor training in a horizontal posture using only perineal stimulation can induce well-coordinated speed-adapted locomotion of the hindlimbs after some weeks (Alluin et al., 2015). Similar findings were made in spinalized adult mice (Leblond et al., 2003).

Locomotor training and afferent inputs

The spinal circuitry responsible for locomotor generation must reorganize after removing all descending inputs, either propriospinal from cervical regions or from supraspinal centers. After a chronic SCI, some changes in motoneuron membrane properties and synaptic drive may account for increased spinal excitability, as shown in detailed studies (Hochman & McCrea, 1994a, 1994b, 1994c).

These intrinsic changes can be conditioned by locomotor training, which activates movement-related sensory feedback (Barbeau & Rossignol, 1987; de Leon et al., 1998; Edgerton et al., 1991; Lovely et al., 1986; Rossignol et al., 2006; Rossignol & Frigon, 2011). Locomotor training providing repetitive exercise adds further beneficial effects on locomotor performance. For instance, it was shown that the muscle force output and endurance could be enhanced by training (Edgerton et al., 2004). The importance played by sensory inputs in modifying the spinal circuitry has been emphasized and reviewed (Edgerton et al., 2004; Rossignol et al., 2006, 2011; Rossignol & Frigon, 2011). Two sensory inputs have been claimed to be important for participating in the initiation of the swing phase: one is the loss of force feedback in ankle extensors at the end of stance that would provide a trigger signal to initiate a swing (Pearson, 1995; Pearson et al., 1992), and the second is the position of the hip joint whose extension triggers a swing even when there is little or no tension in ankle extensors (Grillner & Rossignol, 1978).

Reflex studies in spinal cats indicated that stimulation of a circumscribed area on the dorsum of the foot (mechanical or electrical) could evoke hindlimb reflexes that were coherent with and emphasized the ongoing phase of the step cycle. This was coined as a phase-dependent reflex reversal in spinal cats (Forssberg, 1979; Forssberg et al., 1977) in which the same stimulus evoked either a flexion response or an extensor response, depending on whether the exact same stimulus was applied during the swing or stance phase. This was also confirmed later in intact cats, either in the forelimbs (Drew & Rossignol, 1987) or the hindlimbs (Frigon & Rossignol, 2008).

Cutaneous inputs are particularly important for the recovery of foot placement during spinal locomotion. Our studies indeed showed that whereas all the cutaneous inputs from both hindpaws can be removed without disturbing locomotion in the otherwise intact cat, the same cats after spinalization can perform bilateral rhythmic hindlimb stepping but are unable to touch ground on the plantar surface of the foot (Bouyer & Rossignol, 2003a, 2003b). It is of interest that locomotor training by itself might modulate the transmission in relevant sensory pathways (Côté & Gossard, 2004; Côté et al., 2003) so that training seems to modify not only the intrinsic spinal circuitry generating the stepping pattern but also reflex transmission through some specific sensory pathways important for enhancing aspects of stepping, such as an increase in weight support. The mechanisms through which specific cutaneous pathways and associated interneurons have been studied with genetic methods showed a very specific role such as in grasping (Bui et al., 2013; Panek et al., 2014). The importance of cutaneous information on the recovery of locomotion in completely spinalized rats has been emphasized by showing that the beneficial effect of training rats in the upright posture is mainly due to cutaneous receptors detecting the load on the limbs (Slawinska et al., 2012).

The beneficial effects of locomotor training after spinalization are probably achieved through the increased activation of afferent inputs (Edgerton et al., 2001; Rossignol et al., 2015). Although robotic devices can provide an indirect activation of afferents through imposed movements, attempts have been made to activate directly the spinal cord and afferents to provide an excitatory input to the spinal cord. In that vein, electrical stimulation of afferents in dorsal roots either through intrathecal stimulation by electrodes placed on the dura (Harkema et al., 2011a) or transcutaneous stimulation (Gerasimenko et al., 2015) or even electromagnetic magnetic stimulation (Gerasimenko et al., 2010) with or without vibration of muscles have been used. These studies are largely performed in normal humans in whom a remarkable involuntary activation of rhythmic locomotor pattern can be generated (as pioneered by Gurfinkel et al., 1998; Selionov et al., 2009) using only muscle vibration. In a recent paper (Gerasimenko et al., 2016), it was shown that transcutaneous spinal stimulation and plantar foot stimulation induced by a mechanical device can initiate a locomotor pattern that becomes more complete when the two are combined. This occurs in quiescent persons whose body weight is supported by a harness in a vertical position. These stimuli can combine and enhance voluntary locomotion. The interaction between such enhancing stimulation and voluntary movement is quite fascinating since it opens up the possibility of using such stimuli to potentiate remnant descending commands for voluntary movements (Angeli et al., 2014).

This is important since it has clinical implications for locomotor training in persons with SCI (see the literature review in Harkema et al., 2011b, 2012).

Neurochemistry after spinalization and locomotor training

Various neurochemical changes occur in the spinal cord after SCI. Even though these changes may not be specific to neurons involved in generating rhythmicity, their general nature makes it likely that several changes will occur in reflex pathways or transmission of descending commands on various interneurons.

Pharmacological stimulation with noradrenergic agonists (clonidine) was used to train adult spinal cats. The potent effect described before (Forssberg & Grillner, 1973) was used to facilitate daily locomotor training after daily clonidine injection and accelerated the recovery of hindlimb locomotion (Chau et al., 1998a, 1998b). In chronically spinalized adult cats trained to walk on a treadmill, spinal fictive locomotor patterns can also be observed without pharmacological stimulation, suggesting that the activity in the central circuitry can be revealed with chronic training alone (Pearson & Rossignol, 1991).

Studies in spinal cats (Giroux et al., 1999) have shown an increase in adrenergic and 5-HT receptors in the spinal cord. These might explain the major effects obtained with these neurotransmitters in either initiating or modulating the locomotor pattern (Rossignol et al., 2001). As mentioned before, changes in the constitutive activity of 5-HT receptors after SCI may lead to an increase of spinal excitability and participate in the recovery of rhythmicity (Murray et al., 2010).

It should also be considered that after SCI, an increase in GAD67, one of the GABA synthetic enzymes, is increased in both the dorsal and ventral horns (Tillakaratne et al., 2002), which could lead to increased inhibition. Step training reduced the levels of the enzyme in a task-specific manner (standing vs. walking) (Tillakaratne et al., 2002). The increase in inhibitory neurotransmitters seems paradoxical in view of a general increase in motoneuronal excitability that may lead to spasticity and an increase in interneuronal activity, such as reciprocal facilitation between antagonist motoneurons that may lead to co-contractions. This paradox was somewhat resolved by showing that the potassium-chloride co-transporter KCC2 is downregulated after spinal cord injury. Therefore, an increase in chloride concentration would lead to a reduced inhibitory effect (Boulenguez et al., 2010).

Other changes occur in the membrane properties of spinal neurones after SCI. For instance, Persistent Inward Currents (PICs), which regulate input-output functions of motoneurons, are altered after SCI since monoaminergic descending pathways normally modulating these currents may be mostly severed. A return of these PICs appears essential to recover spinal excitability and rhythmicity (Johnson & Heckman, 2014; Tazerart et al., 2008).

The underlying mechanisms of the beneficial effects of locomotor training may implicate several levels of plasticity: spinal circuitry, sensory afferents, descending inputs, and neurochemical changes. One current hypothesis suggests that the peripheral production during exercise of Brain Derived Neuro Factor (BDNF) and Neurotrophin-3 (NT-3) as measured through their respective mRNA could exert specifically an effect on the spinal circuitry as well as on muscles in the lesioned rat (Gomez-Pinilla et al., 2001). Since BDNF promotes neuronal survival and axonal sprouting of descending pathways (Jakeman et al., 1998; Kishino et al., 1997; Vavrek et al., 2006) and also regulates KCC2 levels in motoneurons (Boulenguez et al., 2010), it is likely that their effects will contribute to the benefit of locomotor training. In hemisected rats, BDNF decreased and exercise could normalize its level; the level of BDNF could be correlated with the intensity of training, as measured by the distance achieved on a running wheel by the rats (Ying et al., 2005). In complete spinal rats, bike training as well as step training increases BDNF as well as other neurotrophins and affects the recovery of H-reflex excitability (Cote et al., 2011). Another pertinent report (Petruska et al., 2007) has shown that step training in rats (especially neonatally transected) has profound effects on synaptic and cellular properties (spike amplitude, After Hyperpolarization Potential [AHP]) of ankle extensor motoneurons. Afferent inputs as well as Ventro Lateral Funiculus inputs (VLFs) were increased. It was suggested by the authors that the mechanisms leading to such plasticity might involve BDNF and NT-3.

Of great interest, cats spinalized at T13 were injected with NT-3/BDNF, producing fibroblasts, and were found to walk on the treadmill as well as cats that were trained on the treadmill (Boyce et al., 2007). This finding suggests the possibility that one of the effects of training might be through the production of these trophic factors. Intramuscular injections of adeno-associated virus engineered with NT-3 gene can indeed exert an effect on 1A reflex transmission to motoneurons and Excitatory Post Synaptic Potentials (EPSs) properties (Petruska et al., 2010), which might suggest a basic mechanism for the effectiveness of training through proprioceptive reflex pathways.

Spinal learning. One fascinating aspect emerging from work on the spinal cord is that “learning” can take place at the very last station before actual movements are generated. Edgerton’s group has emphasized that spinal animals can execute specific tasks such as standing or walking when these specific tasks are repeated in an intensive training regimen (Edgerton et al., 1997; Hodgson et al., 1994). Specific afferent inputs brought into use, as well as intrinsic neurochemical changes as described above, are most certainly involved in such use-dependent plasticity. This aspect has been pursued by others in a different context of spinal learning that may also implicate BDNF (Grau, 2014).

A most striking example of long-lasting induced spinal effects (Di Giorgio reported in (Chamberlain et al., 1963; Steinmetz & Patterson, 1985) reports that, in decerebrate rabbits, cerebellar lesions produced a postural asymmetry of the hindlimbs, which could be maintained after a complete spinalization if the spinalization was performed some minutes after the cerebellar lesion.

Work by Wolpaw (Thompson & Wolpaw, 2014) has shown in various animal species, including humans, that operant conditioning can be used to upregulate or downregulate a simple H-reflex, thus suggesting that specific operant conditioning could be used to alleviate symptoms which partly involve the H-Reflex such as spasticity after SCI. Therefore, this is another example of spinal plasticity in which transmission in a simple reflex can be changed on a long-term basis.

There is clearly a need to identify factors involved in spinal “learning” since these are probably crucial in determining how training regimens or pharmacological interventions could be used to improve functional recovery of spinal circuits after lesions. Although much more is still needed, a very basic model of spinal motor reflex learning hypothesizes that the 1a-motoneuron-Renshaw cell pathway may constitute a basic learning circuit in the spinal cord that can participate in “learning” or in improving motor tasks (Brownstone et al., 2015).

Recovery of hindlimb locomotion after incomplete SCI

The demonstration that stepping can be induced after a complete low thoracic spinalization by pharmacological stimulation acutely or by a combination of pharmacological stimulation and locomotor training in chronic spinal animals adds significantly to an observation made more than a century ago on the stepping behavior of spinal cats (Sherrington, 1910). The ability to generate a locomotor pattern has at times obscured the fact that spinal animals are not normal cats. The locomotor pattern itself may show a significant foot drag as well as a reduced weight support requiring external partial weight support. More complex postural synergies are also defective (Pratt et al., 1994). Most importantly, the absence of supraspinal signals needed for goal-directed locomotion is absent. However, given the fact that the spinal cord can do so much even in the absence of supraspinal interactions, and even movement-related sensory feedback, has led to numerous studies on the specific roles of these interactions, especially using partial spinal cord lesions.

Studies on partial lesions have been done on animals using weight drops or other impactors to study locomotor recovery (Young, 2002). Such lesions are useful to evaluate the recovery as a function of lesion size, and this has given rise to impairment scales such as the BBB scale (Basso et al., 1995). More specific lesions of dorsal/dorsolateral or ventral/ventrolateral white matter have been used to identify the roles that may be played by large areas of white matter largely corresponding to lateral and ventral motor control pathways (Kuypers, 1982).

Lesions of dorsolateral funiculi

In the cat, bilateral lesion of the dorsolateral funiculus at low thoracic levels (T12-13) is followed by recovery of locomotion with, however, some defects such as foot drag and prolongation of the stance phase. The persistence of such defects depends on the size of the lesion (Jiang & Drew, 1996; Rossignol et al., 1999). With reference to previous description of EMG timing, it was of interest that the knee flexor St was delayed relative to the hip flexor Sartorius. This could lead to a delay in the initiation of swing in other joints, as well as to a foot drag. This defect was, however, reestablished after a few weeks of training except for very large lesions. Similar findings about some disruption of hip and knee flexors were seen in complete spinal cats. This and an earlier onset of ankle flexors such as TA may explain the foot drag because the hip starts its flexion more or less as the ankle starts its own flexion, while the knee flexion is delayed. It is thus clear that lift-off is a critical turning point of the step cycle to which several control signals contribute: sensory afferents as described before (hip joint position at the end of stance, as well as the lowering of force feedback from ankle extensor) and defective supraspinal inputs leading to an abnormal synergy of muscle activation at this moment of the cycle. Eventually, the step cycle subcomponents will normalize to some extent (normal swing/stance relation to step cycle duration as a function of speed (Jiang & Drew, 1996) probably owing to spinal and sensory mechanisms evolving through training. The most amazing finding, however, is that cats with such large bilateral lesions of the dorsolateral funiculus containing the lateral corticospinal tract can walk voluntarily and quadrupedally, confirming that lesions of the motor cortex or the pyramids do not prevent voluntary locomotion (see reviews in Armstrong, 1986; Drew et al., 1996), although there are deficits in unrestrained locomotion probably due to several mechanisms related to Interlimb coordination (Gorska et al., 1993). Quite clearly, cortical structures are mainly implicated in controlling important synergies related to foot placement and especially in planning movements as animals are required to walk with a certain degree of precision (Drew et al., 2008a; Drew & Marigold, 2015; Krouchev & Drew, 2013).

Lesions of ventral funiculi

Such lesions are more difficult to perform experimentally since they require a ventrolateral approach through the vertebral pedicles at T11-13 (Brustein & Rossignol, 1998, 1999). Our work has shown that such large bilateral ventral lesions may lead to a severe walking postural incapacity to the point that cats could behave in the laboratory as completely spinal and dragged their hindquarters for several weeks. Histological analysis of retrograde labeling of brainstem cells showed that our lesion had destroyed vestibulospinal cells and a large number of reticulospinal cells. Of interest is that rubrospinal cells and corticospinal cells were of course less damaged because their axon course more dorsally.

Repeated locomotor training sessions providing as much as possible some weight support and equilibrium led to the recovery of quadruped locomotion not only on the treadmill but also to overground walking, with the ability to negotiate obstacles in the walking path (Brustein & Rossignol, 1998). The intralimb characteristics of the step pattern were preserved. When required to walk up- or downslope on the treadmill, the biomechanical demands were met mainly by amplitude changes in forelimb extensor muscles, which became, perforce, more propulsive than they normally are.

Intrathecal injection of clonidine, the alpha-2 noradrenergic agonist mentioned before, had a deleterious effect on voluntary locomotion (Brustein and Rossignol, 1999) probably through a reduction of postural tonus as alpha-2 agonists are recognized to have an antispastic effect. This raises the important issue that drug effects may differ completely, depending on the type of spinal lesion studied (Giroux et al., 1998).

On the contrary, serotonergic agonists had a beneficial effect on quadrupedal locomotion after ventral lesions, probably by exerting an excitatory effect on spinal excitability (Brustein and Rossignol, 1999), much as is the case in the completely spinal cat (Barbeau and Rossignol, 1990). Considering that serotoninergic agonists do not trigger locomotion after spinaliaation in cats, contrary to rats (Courtine et al., 2009; Gerasimenko et al., 2007; Slawinska et al., 2012, 2013, 2014a, 2014b; Van den Brand et al., 2012), we must suggest here that the effects of 5-HT agonists in cats with partial lesions is probably through a heightened spinal excitability (Rossignol et al., 2014) analogous to what has been observed in low spinal rats in which some 5-HT receptors become constitutive and increase cellular excitability (Murray et al., 2010).

Lateral spinal hemisection

Lateral spinal hemisection at the T10-T11 level (Barrière et al., 2008, 2010; Cohen-Adad et al., 2014; Helgren & Goldberger, 1993; Martinez et al., 2011, 2012, 2013; Rossignol et al., 2015) initially induces locomotor defects that are a combination of unilateral and dorsal lesions—that is, a combination of foot placement defects, increased stance length, and postural defects. However, regular treadmill training for 3 weeks removes most of these defects, and hindlimb locomotion becomes regular and symmetrical—a remarkable feat given the profound imbalance of descending inputs on both sides. As suggested before (Helgren & Goldberger, 1993), afferent signals during training are probably the result in great part of plastic changes in primary sensory afferents. But even more remarkable is the fact that when such trained hemisected cats are completely spinalized at T13, 3 weeks after the hemisection, thus removing all remnant descending inputs, spinal locomotion can reappear within 24 hours, whereas it usually takes 2-3 weeks to reappear when a single complete spinal lesion is performed without a preceding hemisection. It must then be concluded that in the interim period between the hemisection and the complete spinalization, plastic changes have occurred in the spinal cord circuitry, so that it is immediately available for action after the complete spinalization.

Other work has suggested that untrained hemisected cats (at least 3 weeks of inaction) can result in a persistent asymmetrical gait and that this asymmetry can also be observed after the complete spinalization. However, this spinal locomotor asymmetry can be corrected by a further locomotor training in the spinal state (Martinez et al., 2013). Trained hemisected cats that showed a symmetrical gait after 3 weeks of training had a much more symmetrical gait after complete spinalization, suggesting that the training during the interim period imprinted changes in the cord that could be carried over after spinalization. It was shown (Helgren & Goldberger, 1993) that after hemisection, primary afferents below the hemisection increase; this may correspond to an increase in reflex inputs to the cord and be related to the beneficial effects of locomotor training, which enhances inputs to the spinal cord during movements.

Cats that can normally step easily over an obstacle on their path (Drew & Marigold, 2015) have significant difficulties clearing obstacles after a unilateral hemisection, although they can succeed half the time (Doperalski et al., 2011)

Work on fictive locomotor patterns has shown that acutely hemisected cats could generate a much better rhythmicity on the side contralateral to the lesion. However, chronically trained hemisected cats showed a much better rhythmicity on the side ipsilateral to the lesion, suggesting that indeed chronic hemisection entrains changes in the locomotor circuitry that probably underlie the fast recovery of spinal locomotion after spinalization in chronic hemisected cats (Gossard et al., 2015). Similar findings were made with chronically isolated hemisected spinal cord (Kato, 1989).

Mechanisms of recovery after partial SCI

With reference to Figure 3, one can surmise that optimization of locomotor function will depend on several types of changes. The complete SCI model has indicated that major neuroplastic events occur in the spinal cord itself below the lesion. It was also emphasized in previous sections that, after SCI, afferent inputs originating from the limbs during movement will participate in the recovery of locomotion, especially when locomotor training enhances the movement-related sensory feedback. After SCI, other mechanisms should be considered: neuroanatomical/functional plastic changes in spinal neurons, reflex pathways, in descending or propriospinal pathways. The issue of sprouting/regeneration of descending pathways is very large and well beyond the scope of this review. However, some aspects will be discussed insofar as they lead to intraspinal changes.

Axons of commissural interneurons were shown to sprout through the midline after a midline incision (Fenrich & Rose, 2009, 2011). This kind of anatomical plasticity may be crucial to ensure that spinal interneurons above a spinal lesion may act as relays to bypass and carry information down the spinal cord after SCI. This was elegantly shown by the group of Schwab using anatomical and physiological techniques (Bareyre et al., 2004). In this study, a lesion of the corticospinal tract normally projecting down to the lumbar spinal cord was found to sprout on long propriospinal tract neurones. These propriospinal neurones also increased their terminal arborization on lumbar motoneurons. Thus, modifications in spinal propriospinal interneurons could act as efficient relays forming a detour to carry Cortico Spinal Tract (CST) inputs to the spinal cord. Given the importance of the sprouting of axons in the recovery after SCI, it has been important to try to antagonize growth inhibition molecules such as NOGO (Kempf & Schwab, 2013).

Propriospinal neurons also receive other inputs, namely, from the reticular formation. Thus, cortico-reticulo-propriospinal projections may provide a general and important distribution network for the recovery of locomotion. Several papers address the role that propriospinal pathways may play in the recovery of locomotion (Cowley et al., 2008, 2010; Rossignol et al., 2014; Zaporozhets et al., 2006). In mice, staggered spinal hemisections that remove long descending tracts is followed by a locomotor recovery which is attributed to propriospinal pathways bypassing the spinal lesions (Courtine et al., 2008). Of interest is the work on serotoninergic stimulation by quipazine through an intrathecal cannula of propriospinal pathways between dual staggered hemisections in in vivo rats (Cowley et al., 2015).

Emphasis has been put on the importance of the sprouting of axons from corticospinal pathways after SCI to recover voluntary movements. Of interest has been the demonstration that such sprouting is significant in primate models of spinal hemisection in Rhesus monkeys (Rosenzweig et al., 2010). Aside from the detour created by corticospinal projections to propriospinal pathways, it was shown that corticospinal axons could themselves sprout in the spinal cord and form other detours to reach spinal motoneurons involved in voluntary movements (Friedli et al., 2015). After hemisection in rats, electrical stimulation of the spared corticospinal tract or the motor cortex resulted in increased sprouting and better locomotor recovery (Carmel et al., 2010; Carmel & Martin, 2014).

Serotonergic descending pathways are also important after spinal injury. Sprouting of serotonergic terminals has been shown in cats after various lesions (Polistina et al., 1990). In rats, 5-HT fibers sprout after SCI, a process enhanced by NOGO-A antibody (Mullner et al., 2008). This has justified transplanting tissue (Diener & Bregman, 1998) or cells releasing 5-HT at the site of a spinal lesion or below the lesion (Ribotta et al., 2000; Slawinska et al., 2013; Yakovleff et al., 1995) to improve locomotion when grafting is performed weeks after the initial lesion.

Genetic lesions

Traumatic lesions of the spinal pathways allow us to describe the remnant functions within the spinal cord itself or the specific functions of some descending pathways as they relate to locomotion; one can observe changes in circuits or cellular properties of the remaining cells as discussed above. It has now proved possible to dissect parts of the locomotor circuitry using genetic approaches (Goulding, 2009; Grossmann et al., 2010; Kiehn, 2006), and this will undoubtedly become more and more important to determine the effects of specific cellular lesions on rhythm generation and adaptation to speed with various interlimb couplings. The genetic inactivation of several types of interneurons as a class of lesion is a broad topic outside the scope of this review. However, a recent observation points in directions that could be taken along this line. Specific ablation of ventral V0v interneurons abolished trot in the mutant mice. Ablation of V0v and V0d abolished all forms of coupling except bound. As mentioned in a previous section, limbed mammals change speed by adopting different gait patterns such as walk, trot, or gallop, which consist in specific clusters of interlimb couplings. The genetic ablation studies suggest that a modular organization underlies the recruitment of specific circuits to produce the various forms of gaits, which can switch from one to the other almost instantly (Bellardita and Kiehn, 2015; Grossmann et al., 2010). An optogenetic approach was used to demonstrate this modular organization (Hagglund et al., 2013). Optogenetic inactivation might be viewed as a new approach to induce specific spinal lesions and so is inactivation of neurons by diphtheria toxin.

Genetic ablation of muscle spindles as in Egr3 mutant mice has apparently little consequence for the basic locomotor capabilities, but the inactivation of muscle proprioceptive feedback strongly limits the recovery of locomotion after spinal hemisection, confirming the important role played by sensory information during recovery (Takeoka et al., 2014).

Human perspective

One of the main impetuses to study the basic mechanisms of plasticity after spinal lesions is to develop a framework that can be used to develop rehabilitation strategies to enable humans with SCI to recover walking. Some of the concepts developed in experimental animals have at times been questioned, namely, because a great deal of the animal work has been performed on quadrupeds and has generally neglected the postural constraints of human bipeds.

That being said, there are a few observations that must be recalled to link animal experiments and human work. “Old” and reliable observations have suggested that complete spinal transection of the thoracic cord in soldiers during the First World War (Lhermitte, 1919; Roussy & Lhermite, 1918) often resulted in autonomous rhythmic movements of the lower limbs and that such, often very pronounced, uncontrollable rhythmic movements of the limbs below the lesion could be triggered by gently removing the sheet covering the patient, that is, by a light cutaneous stimulation. Similar observations of reflex activity and rhythmic lower limb activity (stepping or cycling movements) were made at the end of the Second World War in completely spinalized persons (Kuhn, 1950). Other more recent evidence of some autonomous rhythmical spinal circuit has been provided (Bussel et al., 1996; Calancie, 2006; Calancie et al., 1994; Dimitrijevic et al., 1998; Nadeau et al., 2010). A recent paper by the Lacquaniti group reiterates the notion of locomotor primitives in humans constituting a sort of central pattern generator and details how the subcomponents of this human CPG can be altered to accommodate perturbations such as walking in uncertain terrain or pathologies such as in cerebellar ataxia (Martino et al., 2014, 2015). Of interest too is the observation that interlimb reflex pathways in humans can reappear several months after a SCI, indicating long-term plastic changes in the spinal cord (Calancie et al., 2002).

Based on concepts developed in animals, some researchers have been active in fostering the use of plastic changes in the spinal cord to promote locomotion in humans with SCI. This subject is outside the scope of this review, but a few papers offer some dramatic examples of such accomplishments (Alexeeva et al., 2011; Angeli et al., 2014; Barbeau et al., 2006; Harkema, 2008; Harkema et al., 2012; Lam et al., 2007).

Observing a child acquiring or selecting the basic skills of moving his or her body around to explore people and objects, one is struck by the gradual progression of movements from a basic alternate coordinated stepping when held upright to a basic frog-like crawl behavior on the ground allowing the child, with coordinated movements of all limbs, to go first backward by pushing forward and then going forward by backward rhythmic movements of the fore and lower limbs, with occasional raising of the buttock and being on all four before walking on two legs. This progression is rather stereotyped and signal basic spinal circuitry upon which already voluntary commands are imposed. It is through this basic movement alphabet that we all learn to move appropriately in normal conditions or optimally after lesions (Forssberg, 1999).

Concluding remarks

In this review, I have first outlined how various parameters (kinematics, EMGs) of locomotion can be modified to adapt to various walking conditions, thus suggesting an important functional plasticity of locomotor control mechanisms in the normal. Second, I have summarized some of the observations made after various types of spinal lesions. This has shown even more spinal plasticity since animals reexpress locomotion after all these lesions (complete spinal section, partial spinal sections). One of the major findings is that after spinal lesion, the spinal cord below the lesion is changed and that this changed spinal cord can further be modified by locomotor training, indicating again a basic spinal neuroplasticity on which various therapeutic approaches (training, pharmacology, and electrical stimulation) can be applied in the context of neurorehabilitation in persons with a SCI.


This review is the result of several interactions with colleagues and students over the years, and I thank them for having contributed to many of the insights reported here. I would like to thank the Department of Neuroscience, the Faculty of Medicine, and Université de Montréal as well as funding organizations (Canadian Institute of Health Research, Fonds de la Recherche en Santé du Québec, Wings for Life, Craig H. Neilsen Foundation, Christopher and Dana Reeve Foundation) for their support over the past 40 years. Thanks also to my wife Céline, my daughter Elsa, and my 18 months old grandson Émile for their enthusiastic support.


Alexeeva, N., Sames, C., Jacobs, P. L., Hobday, L., Distasio, M. M., Mitchell, S. A., et al. (2011). Comparison of training methods to improve walking in persons with chronic spinal cord injury: a randomized clinical trial. Journal of Spinal Cord Medicine, 34(4), 362–379.Find this resource:

Alluin, O., Delivet-Mongrain, H., & Rossignol, S. (2015). Inducing hindlimb locomotor recovery in adult rat after complete thoracic spinal cord section using repeated treadmill training with perineal stimulation only. Journal of Neurophysiology, 114(3), 1931–1946.Find this resource:

Amos, A., Armstrong, D. M., & Marple-Horvat, D. E. (1987). A ladder paradigm for studying skilled and adaptive locomotion in the cat. Journal of Neuroscience Methods, 20, 323–340.Find this resource:

Angeli, C. A., Edgerton, V. R., Gerasimenko, Y. P., & Harkema, S. J. (2014). Altering spinal cord excitability enables voluntary movements after chronic complete paralysis in humans. Brain, 137(Pt 5), 1394–1409.Find this resource:

Antri, M., Mellen, N., & Cazalets, J. R. (2011). Functional organization of locomotor interneurons in the ventral lumbar spinal cord of the newborn rat. PLoS One, 6(6), e20529.Find this resource:

Armstrong, D. M. (1986). Supraspinal contributions to the initiation and control of locomotion in the cat. Progress in Neurobiology, 26, 273–361.Find this resource:

Barbeau, H., Nadeau, S., & Garneau, C. (2006). Physical determinants, emerging concepts, and training approaches in gait of individuals with spinal cord injury. Journal of Neurotrauma, 23, 571–585.Find this resource:

Barbeau, H., & Rossignol, S. (1987). Recovery of locomotion after chronic spinalization in the adult cat. Brain Research, 412, 84–95.Find this resource:

Barbeau, H., & Rossignol, S. (1990). The effects of serotonergic drugs on the locomotor pattern and on cutaneous reflexes of the adult chronic spinal cat. Brain Research, 514, 55–67.Find this resource:

Bareyre, F. M., Kerschensteiner, M., Raineteau, O., Mettenleiter, T. C., Weinmann, O., & Schwab, M. E. (2004). The injured spinal cord spontaneously forms a new intraspinal circuit in adult rats. Nature Neuroscience, 7(3), 269–277.Find this resource:

Barrière, G., Frigon, A., Leblond, H., Provencher, J., & Rossignol, S. (2010). Dual spinal lesion paradigm in the cat: evolution of the kinematic locomotor pattern. Journal of Neurophysiology, 104(2), 1119–1133.Find this resource:

Barrière, G., Leblond H., Provencher, J., & Rossignol, S. (2008). Prominent role of the spinal central pattern generator in the recovery of locomotion after partial spinal cord injuries. Journal of Neuroscience, 28(15), 3976–3987.Find this resource:

Basso, D. M., Beattie, M. S., & Bresnahan, J. C. (1995). A sensitive and reliable locomotor rating scale for open field testing in rats. Journal of Neurotrauma, 12(1), 1–21.Find this resource:

Basso, D. M., Beattie, M. S., & Bresnahan, J. C. (2002). Descending systems contributing to locomotor recovery after mild or moderate spinal cord injury in rats: experimental evidence and a review of literature. Restorative Neurology and Neuroscience, 20(5), 189–218.Find this resource:

Bélanger, M., Drew, T., Provencher, J., & Rossignol, S. (1996). A comparison of treadmill locomotion in adult cats before and after spinal transection. Journal of Neurophysiology, 76(1), 471–491.Find this resource:

Beliez, L., Barriere, G., Bertrand, S. S., & Cazalets, J. R. (2015). Origin of thoracic spinal network activity during locomotor-like activity in the neonatal rat.” Journal of Neuroscience, 35(15), 6117–6130.Find this resource:

Bellardita, C., & Kiehn, O. (2015). Phenotypic characterization of speed-associated gait changes in mice reveals modular organization of locomotor networks. Current Biology, 25(11), 1426–1436.Find this resource:

Beloozerova, I. N., & Sirota, M. G. (1993). The role of the motor cortex in the control of accuracy of locomotor movements in the cat. Journal of Physiology, 461, 1–25.Find this resource:

Boulenguez, P., Liabeuf, S., Bos, R., Bras, H., Jean-Xavier, C., Brocard, C., et al. (2010). Down-regulation of the potassium-chloride cotransporter KCC2 contributes to spasticity after spinal cord injury. Nature Medicine, 16(3), 302–307.Find this resource:

Bouyer, L. J. G., & Rossignol, S. (2003a). Contribution of cutaneous inputs from the hindpaw to the control of locomotion: 1. Intact cats. Journal of Neurophysiology, 90(6), 3625–3639.Find this resource:

Bouyer, L. J. G., & Rossignol, S. (2003b). Contribution of cutaneous inputs from the hindpaw to the control of locomotion: 2. Spinal cats. Journal of Neurophysiology, 90(6), 3640–3653.Find this resource:

Boyce, V. S., Tumolo, M., Fischer, I., Murray, M., & Lemay, M. A. (2007). Neurotrophic factors promote and enhance locomotor recovery in untrained spinalized cats. Journal of Neurophysiology, 98(4), 1988–1996.Find this resource:

Bregman, B. S., & Goldberger, M. E. (1983). Infant lesion effect: I. Development of motor behavior following neonatal spinal cord damage in cats. Developmental Brain Research, 9, 103–117.Find this resource:

Brownstone, R. M., Bui, T. V., & Stifani, N. (2015). Spinal circuits for motor learning. Current Opinion in Neurobiology, 33, 166–173.Find this resource:

Brustein, E., & Rossignol, S. (1998). Recovery of locomotion after ventral and ventrolateral spinal lesions in the cat. I. Deficits and adaptive mechanisms. Journal of Neurophysiology, 80(3), 1245–1267.Find this resource:

Brustein, E., & Rossignol, S. (1999). Recovery of locomotion after ventral and ventrolateral spinal lesions in the cat. II. Effects of noradrenergic and serotoninergic drugs. Journal of Neurophysiology, 81(4), 1513–1530.Find this resource:

Bui, T. V., Akay, T., Loubani, O., Hnasko, T. S., Jessell, T. M., & Brownstone, R. M. (2013). Circuits for grasping: spinal dI3 interneurons mediate cutaneous control of motor behavior. Neuron, 78(1), 191–204.Find this resource:

Bussel, B., Roby-Brami, A., Neris, O. R., & Yakovleff, A. (1996). Evidence for a spinal stepping generator in man. Paraplegia, 34(2), 91–92.Find this resource:

Calancie, B. (2006). “Spinal myoclonus after spinal cord injury.” Journal of Spinal Cord Medicine, 29(4), 413–424.Find this resource:

Calancie, B., Molano, M. R., & Broton, J. G. (2002). Interlimb reflexes and synaptic plasticity become evident months after human spinal cord injury. Brain, 125(Pt 5), 1150–1161.Find this resource:

Calancie, B., Needham-Shropshire B., Jacobs, P., Willer, K., Zych, G., & Green, B. A. (1994). Involuntary stepping after chronic spinal cord injury. Evidence for a central rhythm generator for locomotion in man. Brain, 117 (Pt 5), 1143–1159.Find this resource:

Carlson-Kuhta, P., Trank, T. V., & Smith, J. L. (1998). Forms of forward quadrupedal locomotion. II. A comparison of posture, hindlimb kinematics, and motor patterns for upslope and level walking. Journal of Neurophysiology, 79(4), 1687–1701.Find this resource:

Carmel, J. B., Berrol, L. J., Brus-Ramer, M., & Martin, J. H. (2010). Chronic electrical stimulation of the intact corticospinal system after unilateral injury restores skilled locomotor control and promotes spinal axon outgrowth. Journal of Neuroscience, 30(32), 10918–10926.Find this resource:

Carmel, J. B., & Martin, J. H. (2014). Motor cortex electrical stimulation augments sprouting of the corticospinal tract and promotes recovery of motor function. Frontiers in Integrative Neuroscience, 8, 51.Find this resource:

Cazalets, J. R. (2000). Organization of the spinal locomotor network in neonatal rat. In Neurobiology of spinal cord injury (pp. 89–111). Totowa, NJ: Humana Press.Find this resource:

Chamberlain, T. J., Halick, P., & Gerard, R. W. (1963). Fixation of experience in the rat spinal cord. Journal of Neurophysiology, 26, 662–673.Find this resource:

Chau, C., Barbeau, H., & Rossignol, S. (1998a). Early locomotor training with clonidine in spinal cats. Journal of Neurophysiology, 79(1), 392–409.Find this resource:

Chau, C., Barbeau, H., & Rossignol, S. (1998b). Effects of intrathecal α‎1- and α‎2-noradrenergic agonists and norepinephrine on locomotion in chronic spinal cats. Journal of Neurophysiology, 79(6), 2941–2963.Find this resource:

Cohen-Adad, J., Martinez, M., Delivet-Mongrain, H., & Rossignol, S. (2014). Recovery of locomotion after partial spinal cord lesions in cats: assessment using behavioral, electrophysiological and imaging techniques. Acta Neurobiologiae Experimentalis (Wars), 74(2), 142–157.Find this resource:

Cote, M. P., Azzam, G. A., Lemay, M. A., Zhukareva, V., & Houle, J. D. (2011). Activity-dependent increase in neurotrophic factors is associated with an enhanced modulation of spinal reflexes after spinal cord injury. Journal of Neurotrauma, 28(2), 299–309.Find this resource:

Côté, M.-P., & Gossard, J.-P. (2004). Step-training dependent plasticity in spinal cutaneous pathways. Journal of Neuroscience, 24, 11317–11327.Find this resource:

Côté, M.-P., Menard, A., & Gossard, J.-P. (2003). Spinal cats on the treadmill: changes in load pathways. Journal of Neuroscience, 23(7), 2789–2796.Find this resource:

Courtine, G., Gerasimenko, Y., van den, B. R., Yew, A., Musienko, P., Zhong, H., et al. (2009). Transformation of nonfunctional spinal circuits into functional states after the loss of brain input. Nature Neuroscience, 12(10), 1333–1342.Find this resource:

Courtine, G., Song, B., Roy, R. R., Zhong, H., Herrmann, J. E., Ao, Y., et al. (2008). Recovery of supraspinal control of stepping via indirect propriospinal relay connections after spinal cord injury. Nature Medicine, 14, 69–74.Find this resource:

Cowley, K. C., MacNeil, B. J., Chopek, J. W., Sutherland, S., & Schmidt, B. J. (2015). Neurochemical excitation of thoracic propriospinal neurons improves hindlimb stepping in adult rats with spinal cord lesions. Experimental Neurology, 264, 174–187.Find this resource:

Cowley, K. C., Zaporozhets, E., & Schmidt, B. J. (2008). Propriospinal neurons are sufficient for bulbospinal transmission of the locomotor command signal in the neonatal rat spinal cord. Journal of Physiology, 586(6), 1623–1635.Find this resource:

Cowley, K. C., Zaporozhets, E., & Schmidt, B. J. (2010). Contribution of propriospinal neurons to the activation of locomotion. Annals of the New York Academy of Science, 1198, 42–53.Find this resource:

D’Angelo, G., Thibaudier, Y., Telonio, A., Hurteau, M. F., Kuczynski, V., Dambreville, C., et al. (2014). Modulation of phase durations, phase variations and temporal coordination of the four limbs during quadrupedal split-belt locomotion in intact adult cats. Journal of Neurophysiology, 112(8), 1825–1837.Find this resource:

de Leon, R. D., Hodgson, J. A., Roy, R. R., & Edgerton, V. R. (1998). Full weight-bearing hindlimb standing following stand training in the adult spinal cat. Journal of Neurophysiology, 80(1), 83–91.Find this resource:

Diener, P. S., & Bregman, B. S. (1998). Fetal spinal cord transplants support growth of supraspinal and segmental projections after cervical spinal cord hemisection in the neonatal rat. Journal of Neuroscience, 18(2), 779–793.Find this resource:

Dietz, V., Nakazawa, K., Wirz, M., & Erni, T. (1999). Level of spinal cord lesion determines locomotor activity in spinal man. Experimental Brain Research, 128(3), 405–409.Find this resource:

Dimitrijevic, M. R., Gerasimenko, Y., & Pinter, M. M. (1998). Evidence for a spinal central pattern generator in humans. Annals of the New York Academy of Science, 860, 360–376.Find this resource:

Doperalski, A. E., Tester, N. J., Jefferson, S. C., & Howland, D. R. (2011). Altered obstacle negotiation after low thoracic hemisection in the cat. Journal of Neurotrauma, 28(9), 1983–1993.Find this resource:

Drew, T. (1988). Motor cortical cell discharge during voluntary gait modification. Brain Research, 457, 181–187.Find this resource:

Drew, T. (1993). Motor cortical activity during voluntary gait modifications in the cat.I. Cells related to the forelimbs. Journal of Neurophysiology, 70, 179–199.Find this resource:

Drew, T., Andujar, J. E., Lajoie, K., & Yakovenko, S. (2008a). Cortical mechanisms involved in visuomotor coordination during precision walking. Brain Research Review, 57(1), 199–211.Find this resource:

Drew, T., Jiang, W., Kably, B., & Lavoie, S. (1996). Role of the motor cortex in the control of visually triggered gait modifications. Canadian Journal of Physiology and Pharmacology, 74(4), 426–442.Find this resource:

Drew, T., Jiang, W., & Widajewicz, W. (2002). Contributions of the motor cortex to the control of the hindlimbs during locomotion in the cat. Brain Research Review, 40(1–3), 178–191.Find this resource:

Drew, T., Kalaska, J., & Krouchev, N. (2008b) “Muscle synergies during locomotion in the cat: a model for motor cortex control. Journal of Physiology, 586(5), 1239–1245.Find this resource:

Drew, T., & Marigold, D. S. (2015). Taking the next step: cortical contributions to the control of locomotion. Current Opinion in Neurobiology, 33, 25–33.Find this resource:

Drew, T., & Rossignol, S. (1987). A kinematic and electromyographic study of cutaneous reflexes evoked from the forelimb of unrestrained walking cats. Journal of Neurophysiology, 57, 1160–1184.Find this resource:

Edgerton, V. R., de Guzman, C. P., Gregor, R. J., Roy, R. R., Hodgson, J. A., & Lovely, R. G. (1991). Trainability of the spinal cord to generate hindlimb stepping patterns in adult spinalized cats. In Neurobiological basis of human locomotion (pp. 411–423). Tokyo: Japan scientific Societies Press.Find this resource:

Edgerton, V. R., Leon, R. D., Harkema, S. J., Hodgson, J. A., London, N., Reinkensmeyer, et al. (2001). Retraining the injured spinal cord. Journal of Physiology, 533(Pt 1), 15–22.Find this resource:

Edgerton, V. R., Roy, R. R., de Leon, R. D., Tillakaratne, N., & Hodgson, J. A. (1997). Does motor learning occur in the spinal cord? The Neuroscientist, 3(5), 287–294.Find this resource:

Edgerton, V. R., Tillakaratne, N. J., Bigbee, A. J., de Leon, R. D., & Roy, R. R. (2004). Plasticity of the spinal neural circuitry after injury. Annual Review of Neuroscience, 27, 145–167.Find this resource:

English, A. W. (1979). Interlimb coordination during stepping in the cat: an electromyographic analysis. Journal of Neurophysiology, 42, 229–243.Find this resource:

Fenrich, K. K., & Rose, P. K. (2009). Spinal interneuron axons spontaneously regenerate after spinal cord injury in the adult feline. Journal of Neurosciemce, 29(39), 12145–12158.Find this resource:

Fenrich, K. K., & Rose, P. K. (2011). Axons with highly branched terminal regions successfully regenerate across spinal midline transections in the adult cat.” Journal of Comparative Neurology, 519(16), 3240–3258.Find this resource:

Forssberg, H. (1979). Stumbling corrective reaction: A phase-dependent compensatory reaction during locomotion. Journal of Neurophysiology, 42, 936–953.Find this resource:

Forssberg, H. (1999). Neural control of human motor development. Current Opinion in Neurobiology, 9(6), 676–682.Find this resource:

Forssberg, H., & Grillner, S. (1973). The locomotion of the acute spinal cat injected with clonidine i.v. Brain Research, 50, 184–186.Find this resource:

Forssberg, H., Grillner, S., & Halbertsma, J. (1980a). The locomotion of the low spinal cat. I. Coordination within a hindlimb. Acta Physiologica Scandinavica, 108, 269–281.Find this resource:

Forssberg, H., Grillner, S., Halbertsma, J., & Rossignol, S. (1980b). “The locomotion of the low spinal cat: II. Interlimb coordination. Acta Physiologica Scandinavica, 108, 283–295.Find this resource:

Forssberg, H., Grillner, S., & Rossignol, S. (1977). Phasic gain control of reflexes from the dorsum of the paw during spinal locomotion. Brain Research, 132, 121–139.Find this resource:

Friedli, L., Rosenzweig, E. S., Barraud, Q., Schubert, M., Dominici, N., Awai, L., et al. (2015). Pronounced species divergence in corticospinal tract reorganization and functional recovery after lateralized spinal cord injury favors primates. Science Translational Medicine, 7(302), 302ra134.Find this resource:

Frigon, A., & Rossignol, S. (2008). Adaptive changes of the locomotor pattern and cutaneous reflexes during locomotion studied in the same cats before and after spinalization. Journal of Physiology, 586(12), 2927–2945.Find this resource:

Frigon, A., Thibaudier, Y., & Hurteau, M. F. (2015). Modulation of forelimb and hindlimb muscle activity during quadrupedal tied-belt and split-belt locomotion in intact cats.” Neuroscience, 290, 266–278.Find this resource:

Gerasimenko, Y., Gad, P., Sayenko, D., McKinney, Z., Gorodnichev, R., Puhov, A., et al. (2016). Integration of sensory, spinal, and volitional descending inputs in regulation of human locomotion. Journal of Neurophysiology, 116(1), 98–105.Find this resource:

Gerasimenko, Y., Gorodnichev, R., Machueva, E., Pivovarova, E., Semyenov, D., Savochin, A., et al. (2010). Novel and direct access to the human locomotor spinal circuitry. Journal of Neuroscience, 30(10), 3700–3708.Find this resource:

Gerasimenko, Y., Gorodnichev, R., Moshonkina, T., Sayenko, D., Gad, P., & Reggie, E. V. (2015). Transcutaneous electrical spinal-cord stimulation in humans. Annals of Physical and Rehabilitational Medicine, 58(4), 225–231.Find this resource:

Gerasimenko, Y. P., Ichiyama, R. M., Lavrov, I. A., Courtine, G., Cai, L., Zhong, H., et al. (2007). Epidural spinal cord stimulation plus quipazine administration enable stepping in complete spinal adult rats. Journal of Neurophysiology, 98(5), 2525–2536.Find this resource:

Gimenez y Ribotta, M., Provencher, J., Feraboli-Lohnherr, D., Rossignol, S., Privat, A., & Orsal, D. (2000). Activation of locomotion in adult chronic spinal rats is achieved by transplantation of embryonic raphe cells reinnervating a precise lumbar level. Journal of Neuroscience, 20(13), 5144–5152.Find this resource:

Giroux, N., Brustein, E., Chau, C., Barbeau, H., Reader, T. A., & Rossignol, S. (1998). Differential effects of the noradrenergic agonist clonidine on the locomotion of intact, partially and completely spinalized adult cats. In Neuronal mechanisms for generating locomotor activity (pp. 517–520). New York: Annals of the New York Academy of Sciences.Find this resource:

Giroux, N., Rossignol, S., & Reader, T. A. (1999). Autoradiographic study of 1-, 2-Noradrenergic and Serotonin 1A receptors in the spinal cord of normal and chronically transected cats. Journal of Comparative Neurology, 406(3), 402–414.Find this resource:

Giszter, S., Davies, M. R., Ramakrishnan, A., Udoekwere, U. I., & Kargo, W. J. (2008). Trunk sensorimotor cortex is essential for autonomous weight-supported locomotion in adult rats spinalized as P1/P2 neonates. Journal of Neurophysiology, 100(2), 839–851.Find this resource:

Gomez-Pinilla, F., Ying, Z., Opazo, P., Roy, R. R., & Edgerton, V. R. (2001). Differential regulation by exercise of BDNF and NT-3 in rat spinal cord and skeletal muscle. European Journal of Neuroscience, 13(6), 1078–1084.Find this resource:

Gorska, T., Majczynski, H., Bem, T., & Zmyslowski, W. (1993). Hindlimb swing, stance and step relationships during unrestrained walking in cats with lateral funicular lesion. Acta Neurobiologiae Experimentalis, 53, 133–142.Find this resource:

Gossard, J. P., Delivet-Mongrain, H., Martinez, M., Kundu, A., Escalona, M., & Rossignol, S. (2015). Plastic changes in lumbar locomotor networks after a partial spinal cord injury in cats. Journal of Neuroscience, 35(25), 9446–9455.Find this resource:

Goulding, M. (2009). Circuits controlling vertebrate locomotion: moving in a new direction. Nature Reviews Neuroscience, 10(7), 507–518.Find this resource:

Grau, J. W. (2014). Learning from the spinal cord: How the study of spinal cord plasticity informs our view of learning. Neurobiology of Learning and Memory, 108, 155–171.Find this resource:

Grillner, S. (1973). Locomotion in the spinal cat.” In Control of posture and locomotion (pp. 515–535). Advances in Behavioral Biology, 7. New York: Plenum Press.Find this resource:

Grillner, S. (1981). Control of locomotion in bipeds, tetrapods, and fish. In Handbook of physiology (pp. 1179–1236). The nervous system II. Bethesda, MD: American Physiology Society.Find this resource:

Grillner, S., & Rossignol, S. (1978). On the initiation of the swing phase of locomotion in chronic spinal cats. Brain Research, 146, 269–277.Find this resource:

Grillner, S., & Zangger, P. (1979). On the central generation of locomotion in the low spinal cat. Experimental Brain Research, 34, 241–261.Find this resource:

Grossmann, K. S., Giraudin, A., Britz, O., Zhang, J., & Goulding, M. (2010). Genetic dissection of rhythmic motor networks in mice. Progress in Brain Research, 187, 19–37.Find this resource:

Gurfinkel, V. S., Levik, Y. S., Kazennikov, O. V., & Selionov, V. A. (1998). Locomotor-like movements evoked by leg muscle vibration in humans. European Journal of Neuroscience, 10(5), 1608–1612.Find this resource:

Hagglund, M., Dougherty, K. J., Borgius, L., Itohara, S., Iwasato, T., & Kiehn, O. (2013). Optogenetic dissection reveals multiple rhythmogenic modules underlying locomotion. Proceedings of the Natioanl Academy of Science U S A, 110(28), 11589–11594.Find this resource:

Halbertsma, J. M. (1983). The stride cycle of the cat: the modelling of locomotion by computerized analysis of automatic recordings. Acta Physiologica Scandinavica, Suppl., 521, 1–75.Find this resource:

Harkema, S., Behrman, A., & Barbeau, H. (2012). “Evidence-based therapy for recovery of function after spinal cord injury.” Handbook of Clinical Neurology, 109, 259–274.Find this resource:

Harkema, S., Gerasimenko, Y., Hodes, J., Burdick, J., Angeli, C., Chen, Y., et al. (2011a). Effect of epidural stimulation of the lumbosacral spinal cord on voluntary movement, standing, and assisted stepping after motor complete paraplegia: a case study. The Lancet, 377, 1938–1947.Find this resource:

Harkema, S. J. (2008). Plasticity of interneuronal networks of the functionally isolated human spinal cord. Brain Research Review, 57(1), 255–264.Find this resource:

Harkema, S. J., Behrman, A. L., & Barbeau, H. (2011b). Locomotor training: Principles and practice. New York: Oxford University Press.Find this resource:

Helgren, M. E., Goldberger, M. E. (1993). The recovery of postural reflexes and locomotion following low thoracic hemisection in adult cats involves compensation by undamaged primary afferent pathways. Experimental Neurology, 123, 17–34.Find this resource:

Hildebrand, M. (1976). Analysis of tetrapod gaits: general considerations and symmetrical gaits.” In Neural control of locomotion (pp. 203–236). New York: Plenum Press.Find this resource:

Hildebrand, M. (1980). The adaptive significance of tetrapod gait selection. Oxford Journals, 20(1), 255–267.Find this resource:

Hochman, S., & McCrea, D. A. (1994a). Effects of chronic spinalization on ankle extensor motoneurons. I. Composite monosynaptic Ia EPSPs in four motoneuron pools. Journal of Neurophysiology, 71, 1452–1467.Find this resource:

Hochman, S., & McCrea, D. A. (1994b). Effects of chronic spinalization on ankle extensor motoneurons. II. Motoneuron electrical properties. Journal of Neurophysiology, 71, 1468–1479.Find this resource:

Hochman, S., & McCrea, D. A. (1994c). Effects of chronic spinalization on ankle extensor motoneurons. III. Composite Ia EPSPs in motoneurons separated into motor unit types. Journal of Neurophysiology, 71, 1480–1490.Find this resource:

Hodgson, J. A., Roy, R. R., de Leon, R., Dobkin, B., & Edgerton, V. R. (1994). Can the mammalian lumbar spinal cord learn a motor task?. Medicine and Science in Sports and Exercise, 26(12), 1491–1497.Find this resource:

Howland, D. R., Bregman, B. S., Tessler, A., & Goldberger, M. E. (1995). Development of locomotor behavior in the spinal kitten. Experimental Neurology, 135(2), 108–122.Find this resource:

Ichiyama, R. M., Courtine, G., Gerasimenko, Y. P., Yang, G. J., Van den Brand, R., Lazrov, I. A., et al. (2008). Step Training reinforces specific spinal locomotor circuitry in adult spinal rats. Journal of Neuroscience, 28, 7370–7375.Find this resource:

Jakeman, L. B., Wei, P., Guan, Z., & Stokes, B. T. (1998). Brain-derived neurotrophic factor stimulates hindlimb stepping and sprouting of cholinergic fibers after spinal cord injury. Experimental Neurology, 154, 170–184.Find this resource:

Jiang, W., & Drew, T. (1996). Effects of bilateral lesions of the dorsolateral funiculi and dorsal columns at the level of the low thoracic spinal cord on the control of locomotion in the adult cat: I.Treadmill walking. Journal of Neurophysiology, 76(2), 849–866.Find this resource:

Johnson, M. D., & Heckman, C. J. (2014). Gain control mechanisms in spinal motoneurons. Frontiers in Neural Circuits, 8, 81.Find this resource:

Juvin, L., Le Gal, J. P., Simmers, J., & Morin, D. (2012) Cervicolumbar coordination in Mammalian quadrupedal locomotion: role of spinal thoracic circuitry and limb sensory inputs. Journal of Neuroscience, 32(3), 953–965.Find this resource:

Kato, M. (1989). Chronically isolated lumbar half spinal cord produced by hemisection and longitudinal myelotomy generates locomotor activities of the ipsilateral hindlimb of the cat. Neuroscience Letters, 98, 149–153.Find this resource:

Kempf, A., & Schwab, M. E. (2013). Nogo-A represses anatomical and synaptic plasticity in the central nervous system. Physiology (Bethesda), 28(3), 151–163.Find this resource:

Kiehn, O. (2006). Locomotor circuits in the mammalian spinal cord. Annual Review of Neuroscience, 29, 279–306.Find this resource:

Kiehn, O., & Kjaerulff, O. (1998). Distribution of central pattern generators for rhythmic motor outputs in the spinal cord of limbed vertebrates. Annals of the New York Academy of Science, 860, 110–129.Find this resource:

Kishino, A., Ishige, Y., Tatsuno, T., Nakayama, C., & Noguchi, H. (1997). BDNF prevents and reverses adult rat motor neuron degeneration and induces axonal outgrowth. Experimental Neurology, 144(2), 273–286.Find this resource:

Kjaerulff, O., & Kiehn, O. (1996). Distribution of networks generating and coordinating locomotor activity in the neonatal rat spinal cord in vitro: A lesion study. Journal of Neuroscience, 16, 5777–5794.Find this resource:

Krouchev, N., & Drew, T. (2013). Motor cortical regulation of sparse synergies provides a framework for the flexible control of precision walking. Frontiers in Computational Neuroscience, 7, 83.Find this resource:

Krouchev, N., Kalaska, J. F., & Drew, T. (2006). Sequential activation of muscle synergies during locomotion in the intact cat as revealed by cluster analysis and direct decomposition. Journal of Neurophysiology, 96, 1991–2010.Find this resource:

Kuhn, R. A. (1950). Functional capacity of the isolated human spinal cord. Brain73:1–51.Find this resource:

Kulagin, A. S., & Shik, M. L. (1970) Interaction of symmetrical limbs during controlled locomotion. Biophysics, 15, 171–178.Find this resource:

Kuypers, H. G. (1982). “A new look at the organization of the motor system.” Progress in Brain Research, 57, 381–403.Find this resource:

Lam, T., Eng, J. J., Wolfe, D. L., Hsieh, J. T. C., & Whittaker, M. (2007). A systematic review of the efficacy of gait rehabilitation strategies for spinal cord injury. Top Spinal Cord Injury Rehabilitation, 13, 32–57.Find this resource:

Langlet, C., Leblond, H., & Rossignol, S. (2005). The mid-lumbar segments are needed for the expression of locomotion in chronic spinal cats. Journal of Neurophysiology, 93(5), 2474–2488.Find this resource:

Leblond, H., L’Espérance, M., Orsal, D., & Rossignol, S. (2003). Treadmill locomotion in the intact and spinal mouse. Journal of Neuroscience, 23(36), 11411–11419.Find this resource:

Lemieux, M., Josset, N., Roussel, M., Couraud, S., & Bretzner, F. (2016). Speed-dependent modulation of the locomotor behavior in adult mice reveals attractor and transitional gaits. Frontiers in Neuroscience, 10, 42.Find this resource:

Lhermitte, J. (1919). La section totale de la Moelle Dorsale. Paris: Imprimerie Vve Tardy-Pigelet et Fils.Find this resource:

Lovely, R. G., Gregor, R. J., Roy, R. R., & Edgerton, V. R. (1986). Effects of training on the recovery of full-weight-bearing stepping in the adult spinal cat.” Experimental Neurology, 92, 421–435.Find this resource:

Lovely, R. G., Gregor, R. J., Roy, R. R., & Edgerton, V. R. (1990) Weight-bearing hindlimb stepping in treadmill-exercised adult spinal cat. Brain Research, 514, 206–218.Find this resource:

Lyalka, V. F., Musienko, P. E., Orlovsky, G. N., Grillner, S., & Deliagina, T. G. (2008). Effect of intrathecal administration of serotoninergic and noradrenergic drugs on postural performance in rabbits with spinal cord lesions. Journal of Neurophysiology, 100(2), 723–732.Find this resource:

Lyalka, V. F., Orlovsky, G. N., & Deliagina, T. G. (2009). Impairment of postural control in rabbits with extensive spinal lesions. Journal of Neurophysiology, 101(4), 1932–1940.Find this resource:

Macpherson, J. M., & Fung, J. (1999). Weight support and balance during perturbed stance in the chronic spinal cat. Journal of Neurophysiology, 82(6), 3066–3081.Find this resource:

Majczynski, H., & Slawinska, U. (2007). Locomotor recovery after thoracic spinal cord lesions in cats, rats and humans. Acta Neurobiologiae Experimentalis (Wars), 67(3), 235–257.Find this resource:

Martinez, M., Delivet-Mongrain, H., Leblond, H., & Rossignol, S. (2011). Recovery of hindlimb locomotion after incomplete spinal cord injury in the cat involves spontaneous compensatory changes within the spinal locomotor circuitry. Journal of Neurophysiology, 106(4), 1969–1984.Find this resource:

Martinez, M., Delivet-Mongrain, H., Leblond, H., & Rossignol, S. (2012). Incomplete spinal cord injury promotes durable functional changes within the spinal locomotor circuitry. Journal of Neurophysiology, 108(1), 124–134.Find this resource:

Martinez, M., Delivet-Mongrain, H., & Rossignol, S. (2013). Treadmill training promotes spinal changes leading to locomotor recovery after partial spinal cord injury in cats. Journal of Neurophysiology, 109(12), 2909–2922.Find this resource:

Martino, G., Ivanenko, Y. P., d’Avella, A., Serrao, M., Ranavolo, A., Draicchio, F., et al. (2015). Neuromuscular adjustments of gait associated with unstable conditions. Journal of Neurophysiology, 114(5), 2867–2882.Find this resource:

Martino, G., Ivanenko, Y. P., Serrao, M., Ranavolo, A., d’Avella, A., Draicchio, F., et al. (2014). Locomotor patterns in cerebellar ataxia. Journal of Neurophysiology, 112(11), 2810–2821.Find this resource:

Miller, S., Van der Burg, J., & Van der Meche, F. G. A. (1975). Coordination of movement of the hindlimbs and forelimbs in different forms of locomotion in normal and decerebrate cats. Brain Research, 91, 217–237.Find this resource:

Mullner, A., Gonzenbach, R. R., Weinmann, O., Schnell, L., Liebscher, T., et al. (2008). “Lamina-specific restoration of serotonergic projections after Nogo-A antibody treatment of spinal cord injury in rats.” European Journal of Neuroscience, 27(2), 326–333.Find this resource:

Murray, K. C., Nakae, A., Stephens, M. J., Rank, M., D’Amico, J., Harvey, P. J., et al. (2010). Recovery of motoneuron and locomotor function after spinal cord injury depends on constitutive activity in 5-HT(2C) receptors. Nature Medicine, 16(6), 694–700.Find this resource:

Nadeau, S., Jacquemin, G., Fournier, C., Lamarre, Y., & Rossignol, S. (2010). Spontaneous motor rhythms of the back and legs in a patient with a complete spinal cord transection. Neurorehabilitation and Neural Repair, 24(4), 377–383.Find this resource:

Oza, C. S., & Giszter, S. F. (2014). Plasticity and alterations of trunk motor cortex following spinal cord injury and non-stepping robot and treadmill training. Experimental Neurology, 256, 57–69.Find this resource:

Oza, C. S., & Giszter, S. F. (2015). Trunk robot rehabilitation training with active stepping reorganizes and enriches trunk motor cortex representations in spinal transected rats. Journal of Neuroscience, 35(18), 7174–7189.Find this resource:

Panek, I., Bui, T., Wright, A. T., & Brownstone, R. M. (2014). Cutaneous afferent regulation of motor function. Acta Neurobiologiae Experimentalis (Wars), 74(2), 158–171.Find this resource:

Pearson, K. G. (1995). Proprioceptive regulation of locomotion. Current Opinion in Neurobiology, 5(6), 786–791.Find this resource:

Pearson, K. G., Ramirez, J. M., & Jiang, W. (1992). “Entrainment of the locomotor rhythm by group Ib afferents from ankle extensor muscles in spinal cats.” Experimental Brain Research, 90, 557–566.Find this resource:

Pearson, K. G., & Rossignol, S. (1991). Fictive motor patterns in chronic spinal cats. Journal of Neurophysiology, 66, 1874–1887.Find this resource:

Petruska, J. C., Ichiyama, R. M., Jindrich, D. L., Crown, E. D., Tansey, K. E., Roy, R. R., et al. (2007). Changes in motoneuron properties and synaptic inputs related to step training after spinal cord transection in rats. Journal of Neuroscience, 27(16), 4460–4471.Find this resource:

Petruska, J. C., Kitay, B., Boyce, V. S., Kaspar, B. K., Pearse, D. D., Gage, F. H., et al. (2010). Intramuscular AAV delivery of NT-3 alters synaptic transmission to motoneurons in adult rats. European Journal of Neuroscience, 32(6), 997–1005.Find this resource:

Philippson, M. (1905). L’autonomie et la centralisation dans le système nerveux des animaux. Trav Lab Physiol Inst Solvay (Bruxelles), 7, 1–208.Find this resource:

Polistina, D. C., Murray, M., & Goldberger, M. E. (1990). Plasticity of dorsal root and descending serotoninergic projections after partial deafferentation of the adult rat spinal cord. Journal of Comparative Neurology, 299(3), 349–363.Find this resource:

Pratt, C. A., Fung, J., & Macpherson, J. M. (1994) Stance control in the chronic spinal cat. Journal of Neurophysiology, 71, 1981–1985.Find this resource:

Pratt, C. A., & Loeb, G. E. (1991). “Functionally complex muscles of the cat hindlimb. I. Patterns of activation across sartorius.” Experimental Brain Research, 85, 243–256.Find this resource:

Ribotta, M. G., Provencher, J., Feraboli-Lohnherr, D., Rossignol, S., Privat, A., & Orsal, D. (2000). Activation of locomotion in adult chronic spinal rats is achieved by transplantation of embryonic raphe cells reinnervating a precise lumbar level. Journal of Neuroscience, 20(13), 5144–5152.Find this resource:

Rosenzweig, E. S., Courtine, G., Jindrich, D. L., Brock, J. H., Ferguson, A. R., Strand, S. C., et al. (2010). Extensive spontaneous plasticity of corticospinal projections after primate spinal cord injury. Nature Neuroscience, 13(12), 1505–1510.Find this resource:

Rossignol, S. (1996). “Neural control of stereotypic limb movements.” In Handbook of Physiology. Section 12. Exercise: Regulation and Integration of Multiple Systems (pp. 173–216). New York: Oxford University Press.Find this resource:

Rossignol, S., Drew, T., Brustein, E., & Jiang, W. (1999). Locomotor performance and adaptation after partial or complete spinal cord lesions in the cat. Progress in Brain Research, 123(31), 349–365.Find this resource:

Rossignol, S., Dubuc, R., & Gossard, J. P. (2006). Dynamic sensorimotor interactions in locomotion. Physiological Reviews, 86(1), 89–154.Find this resource:

Rossignol, S., & Frigon, A. (2011). Recovery of locomotion after spinal cord injury: Some facts and mechanisms. Annual Review of Neuroscience, 34, 413–440.Find this resource:

Rossignol, S., Frigon, A., Barriere, G., Martinez, M., Barthelemy, D., Bouyer, L., et al. (2011). Spinal plasticity in the recovery of locomotion. Progress in Brain Research, 188, 229–241.Find this resource:

Rossignol, S., Giroux, N., Chau, C., Marcoux, J., Brustein, E., & Reader, T. A. (2001). Pharmacological aids to locomotor training after spinal injury in the cat. Journal of Physiology, 533(Pt 1), 65–74.Find this resource:

Rossignol, S., Martinez, M., Escalona, M., Kundu, A., Delivet-Mongrain, H., Alluin, O., et al. (2015). “The “beneficial” effects of locomotor training after various types of spinal lesions in cats and rats.” In Sensorimotor rehabilitation: At the crossroad of basic and clinical sciences (pp. 173–198). Oxford: Elsevier.Find this resource:

Rossignol, S., Schmidt, B. J., & Jordan, L. M. (2014). Spinal plasticity underlying the recovery of locomotion after injury. In Textbook of neural repair and rehabilitation (pp. 166–195). Cambridge, U.K.: Cambridge University Press.Find this resource:

Roussy, G., & Lhermite, J. (1918). Les blessures de la moelle et de la queue de cheval. Paris: Masson & Cie.Find this resource:

Saltiel, P., & Rossignol, S. (2004). Critical points in the forelimb fictive locomotor cycle and motor coordination: Evidence from the effects of tonic proprioceptive perturbations in the cat. Journal of Neurophysiology, 92(3), 1329–1341.Find this resource:

Saltiel, P., Wyler-Duda, K., d’Avella, A., Tresch, M. C., & Bizzi, E. (2001). Muscle synergies encoded within the spinal cord: evidence from focal intraspinal NMDA iontophoresis in the frog. Journal of Neurophysiology, 85(2), 605–619.Find this resource:

Selionov, V. A., Ivanenko, Y. P., Solopova, I. A., & Gurfinkel, V. S. (2009). Tonic central and sensory stimuli facilitate involuntary air-stepping in humans. Journal of Neurophysiology, 101(6), 2847–2858.Find this resource:

Sherrington, C. S. (1910). Flexion-reflex of the limb, crossed extension-reflex, and reflex stepping and standing. Journal of Physiology, 40, 28–121.Find this resource:

Shik, M. L., Orlovsky, G. N., & Severin, F. V. (1966a). Organization of locomotor synergism. Biophysics, 11, 1011–1019.Find this resource:

Shik, M. L., Severin, F. V., & Orlovsky, G. N. (1966b). Control of walking and running by means of electrical stimulation of the mid-brain. Biophysics, 11, 756–765.Find this resource:

Slawinska, U., Majczynski, H., Dai, Y., & Jordan, L. M. (2012). The upright posture improves plantar stepping and alters responses to serotonergic drugs in spinal rats. Journal of Physiology, 590(Pt 7), 1721–1736.Find this resource:

Slawinska, U., Miazga, K., Cabaj, A. M., Leszczynska, A., Majczynski, H., Nagy, J. I., et al. (2013). Grafting of fetal brainstem 5-HT neurons into the sublesional spinal cord of paraplegic rats restores coordinated hindlimb locomotion. Experimental Neurology, 247, 572–581.Find this resource:

Slawinska, U., Miazga, K., & Jordan, L. M. (2014a). 5-HT(2) and 5-HT(7) receptor agonists facilitate plantar stepping in chronic spinal rats through actions on different populations of spinal neurons. Frontiers in Neural Circuits, 8, 95.Find this resource:

Slawinska, U., Miazga, K., & Jordan, L. M. (2014b). The role of serotonin in the control of locomotor movements and strategies for restoring locomotion after spinal cord injury. Acta Neurobiologiae Experimentalis (Wars), 74(2), 172–187.Find this resource:

Smith, J. L., Carlson-Kuhta, P., & Trank, T. V. (1998). Forms of forward quadrupedal locomotion. III. A comparison of posture, hindlimb kinematics, and motor patterns for downslope and level walking. Journal of Neurophysiology, 79(4), 1702–1716.Find this resource:

Steinmetz, J. E., & Patterson, M. M. (1985). Fixation of spinal reflex alterations in spinal rats by sensory nerve stimulation. Behavioral Neuroscience, 99(1), 97–108.Find this resource:

Stelzner, D. J., Ershler, W. B., & Weber, E. D. (1975). Effects of spinal transection in neonatal and weanling rats: survival of functions. Experimental Neurology, 46(1), 156–177.Find this resource:

Takeoka, A., Vollenweider, I., Courtine, G., & Arber, S. (2014). Muscle spindle feedback directs locomotor recovery and circuit reorganization after spinal cord injury. Cell, 159(7), 1626–1639.Find this resource:

Tazerart, S., Vinay, L., & Brocard, F. (2008). The persistent sodium current generates pacemaker activities in the central pattern generator for locomotion and regulates the locomotor rhythm. Journal of Neuroscience, 28(34), 8577–8589.Find this resource:

Thibaudier, Y., & Frigon, A. (2014). Spatiotemporal control of interlimb coordination during transverse split-belt locomotion with 1:1 or 2:1 coupling patterns in intact adult cats. Journal of Neurophysiology, 112(8), 2006–2018.Find this resource:

Thibaudier, Y., & Hurteau, M. F. (2012). Sensory regulation of quadrupedal locomotion: a top-down or bottom-up control system? Journal of Neurophysiology, 108(3), 709–711.Find this resource:

Thibaudier, Y., Hurteau, M. F., Telonio, A., & Frigon, A. (2013). Coordination between the fore- and hindlimbs is bidirectional, asymmetrically organized, and flexible during quadrupedal locomotion in the intact adult cat. Neuroscience, 240, 13–26.Find this resource:

Thompson, A. K., & Wolpaw, J. R. (2014). Operant conditioning of spinal reflexes: from basic science to clinical therapy. Frontiers in Integrative Neuroscience, 8, 25.Find this resource:

Tillakaratne, N. J., de Leon, R. D., Hoang, T. X., Roy, R. R., Edgerton, V. R., & Tobin, A. J. (2002). Use-dependent modulation of inhibitory capacity in the feline lumbar spinal cord. Journal of Neuroscience, 22(8), 3130–3143.Find this resource:

Trank, T. V., Chen, C., & Smith, J. L. (1996). Forms of forward quadrupedal locomotion. I. A comparison of posture, hindlimb kinematics, and motor patterns for normal and crouched walking. Journal of Neurophysiology, 76(4), 2316–2326.Find this resource:

Trank, T. V., & Smith, J. L. (1996). Adaptive control for backward quadrupedal walking VI. metatarsophalangeal joint dynamics and motor patterns of digit muscles. Journal of Neurophysiology, 75(2), 678–679.Find this resource:

Udoekwere, U. I., Oza, C. S., & Giszter, S. F. (2016). Teaching adult rats spinalized as neonates to walk using trunk robotic rehabilitation: Elements of success, failure, and dependence.” Journal of Neuroscience, 36(32), 8341–8355.Find this resource:

Van den Brand, R., Heutschi, J., Barraud, Q., DiGiovanna, J., Bartholdi, K., Huerlimann, M., et al. (2012). Restoring voluntary control of locomotion after paralyzing spinal cord injury. Science, 336(6085), 1182–1185.Find this resource:

Vavrek, R., Girgis, J., Tetzlaff, W., Hiebert, G. W., & Fouad, K. (2006). BDNF promotes connections of corticospinal neurons onto spared descending interneurons in spinal cord injured rats. Brain, 129(Pt 6), 1534–1545.Find this resource:

Widajewicz, W., Kably, B., & Drew, T. (1994). Motor cortical activity during voluntary gait modifications in the cat. II. Cells related to the hindlimbs. Journal of Neurophysiology, 72, 2070–2089.Find this resource:

Yakovleff, A., Cabelguen, J. M., Orsal, D., Ribotta, M., Rajaofetra, N., Drian, M. J., et al. (1995). Fictive motor activities in adult chronic spinal rats transplanted with embryonic brainstem neurons. Experimental Brain Research, 106(1), 69–78.Find this resource:

Ying, Z., Roy, R. R., Edgerton, V. R., & Gomez-Pinilla, F. (2005). Exercise restores levels of neurotrophins and synaptic plasticity following spinal cord injury. Experimental Neurology, 193(2), 411–419.Find this resource:

Young, W. (2002). Spinal cord contusion models. Progress in Brain Research, 137, 231–255.Find this resource:

Zaporozhets, E., Cowley, K. C., & Schmidt, B. J. (2006). Propriospinal neurons contribute to bulbospinal transmission of the locomotor command signal in the neonatal rat spinal cord. Journal of Physiology, 572(Pt 2), 443–458.Find this resource:

Dancause, N., Nadeau, S., & Rossignol, S. (Eds.). (2015). Sensorimotor Rehabilitation: At the Crossroads of Basic and Clinical Sciences. Progress in Brain Research, vol. 218. Amsterdam: Elsevier, pp. 1–449.Find this resource: