John A. Hodgson

Retraining the injured spinal cord.

The present review presents a series of concepts that may be useful in developing rehabilitative strategies to enhance recovery of posture and locomotion following spinal cord injury. First, the loss of supraspinal input results in a marked change in the functional efficacy of the remaining synapses and neurons of intraspinal and peripheral afferent (dorsal root ganglion) origin. Second, following a complete transection the lumbrosacral spinal cord can recover greater levels of motor performance if it has been exposed to the afferent and intraspinal activation patterns that are associated with standing and stepping. Third, the spinal cord can more readily reacquire the ability to stand and step following spinal cord transection with repetitive exposure to standing and stepping. Fourth, robotic assistive devices can be used to guide the kinematics of the limbs and thus expose the spinal cord to the new normal activity patterns associated with a particular motor task following spinal cord injury. In addition, such robotic assistive devices can provide immediate quantification of the limb kinematics. Fifth, the behavioural and physiological effects of spinal cord transection are reflected in adaptations in most, if not all, neurotransmitter systems in the lumbosacral spinal cord. Evidence is presented that both the GABAergic and glycinergic inhibitory systems are up‐regulated following complete spinal cord transection and that step training results in some aspects of these transmitter systems being down‐regulated towards control levels. These concepts and observations demonstrate that (a) the spinal cord can interpret complex afferent information and generate the appropriate motor task; and (b) motor ability can be defined to a large degree by training.

Electromyography of rat soleus, medial gastrocnemius, and tibialis anterior during hind limb suspension.

Hind limb suspension is a model frequently utilized to study muscle plasticity. One reason for its frequent use is that it is thought to mimic in many respects the conditions imposed on some muscles during spaceflight. Changes in muscle properties that follow hind limb suspension generally have been attributed to reductions in the recruitment of these muscles. To determine the validity of this assumption, the electromyographic (EMG) activity of three hind limb muscles, the soleus, a slow extensor, the medial gastrocnemius, a fast extensor, and the tibialis anterior, a fast flexor, was studied. The EMG was recorded in each rat for 25 min of each hour for 24 consecutive hours, 7 and 3 days prior, on the day of, and 3, 7, 14, 21, and 28 days after hind limb suspension. Control rats were treated similarly and their EMG recorded on corresponding days. Compared with presuspension, soleus activity was reduced significantly to 91% on the first day of suspension, but had recovered to 81% of its normal activity by the seventh day. Similarly, there was a significant reduction to 54% in activity of the medial gastrocnemius on the day of hind limb suspension which recovered to 98% of its presuspension values by day 7. In contrast, the tibialis anterior showed a significant increase in activity relative to presuspension values within 3 days of the initiation of suspension. These data indicate that hind limb suspension produced only a relatively short-term reduction in the activity of both the soleus and medial gastrocnemius and results in an increased activity in the tibialis anterior. Collaborative studies showed that significant alterations in muscle mass and metabolic and mechanical properties occurred and persisted in spite of the recovery of activation in the soleus and medial gastrocnemius. In addition, no alterations in mass and mechanical properties were evident in the tibialis anterior during a 4-week suspension even though the EMG increased after hind limb suspension. Together, these data indicate that the adaptations in muscle properties following hind limb suspension are not closely related to changes in the total amount of muscle EMG activity per day.

Mechanical output of the cat soleus during treadmill locomotion: In vivo vs in situ characteristics

To study the mechanical output of skeletal muscle, four adult cats were trained to run on a treadmill and then implanted under sterile conditions and anesthesia with a force transducer on the soleus tendon and EMG electrodes in the muscle belly. After a two-week recovery period, five consecutive step cycles were filmed at treadmill speeds of 0.8, 1.3 and 2.2 m s-1. Locomotion data in vivo included individual muscle force, length and velocity changes and EMG during each step cycle. Data for an average step cycle at each speed were compared to the force-velocity properties obtained on the same muscle under maximal nerve stimulation and isotonic loading conditions in situ. Results indicate that the force and power generated at a given velocity of shortening during late stance in vivo were greater at the higher speeds of locomotion than the force and power generated at the same shortening velocity in situ. Strain energy stored in the muscle-tendon unit during the yield phase in early stance is felt to be a major contributor to the muscles enhanced mechanical output during muscle shortening in late stance.

Can the mammalian lumbar spinal cord learn a motor task

Progress toward restoring locomotor function in low thoracic spinal transected cats and the application of similar techniques to patients with spinal cord injury is reviewed. Complete spinal cord transection (T12-T13) in adult cats results in an immediate loss of locomotor function in the hindlimbs. Limited locomotor function returns after several months in cats that have not received specific therapies designed to restore hindlimb stepping. Training transected cats to step on a treadmill for 30 min.d-1 and 5 d.wk-1 greatly improves their stepping ability. The most successful outcome was in cats where training began early, i.e., 1 wk after spinal transection. Cats trained to stand instead of stepping had great difficulty using the hindlimbs for locomotion. These effects were reversible over a 20-month period such that cats unable to step as a result of standing training could be trained to step and, conversely, locomotion in stepping-trained cats could be abolished by standing training. These results indicate that the spinal cord is capable of learning specific motor tasks. It has not been possible to elicit locomotion in patients with clinically complete spinal injuries, but appropriately coordinated EMG activity has been demonstrated in musculature of the legs during assisted locomotion on a treadmill.

Transmission of forces within mammalian skeletal muscles

Most models of in vivo musculoskeletal function fail to take into account the diversity of force trajectories defined by muscle fiber architecture. It has been shown for many muscles, across species, that muscle fibers commonly end within muscle fascicles without reaching a myotendinous junction, and that many of these fibers show a progressive decline in cross-sectional area along the length of the muscle. The significance of these anatomical observations is that the tapering would seem to preclude forces generated at the largest cross-sectional area of the fibers being transmitted to the sarcomeres toward the ends of the tapered fiber. If all of the forces are transmitted via the sarcomeres arranged in series, those few sarcomeres at the smaller ends of the fibers must tolerate the stress exerted by the more numerous sarcomeres arranged in parallel at the portions of the fiber with larger cross-sectional areas. A logical alternative would be for forces to be transmitted laterally along the length of a fiber to the cell membrane and the extracellular matrix. Such a structural arrangement would permit an alternative force transmission vector and minimize the necessity for a precise level of force to be generated along the entire length of a fiber. There are cytoarchitectural and biochemical data demonstrating the presence of a subcellular network which is appropriately located to transmit forces from the active intracellular contractile elements to the extracellular intramuscular connective tissues. However, to fully comprehend how forces are transmitted from individual cross bridges to the tendon, it will be necessary to understand the interactions of all of the components of the muscle tendon complex from the molecular to the multicellular level. It is insufficient to know the physiology of the individual components in a restricted experimental paradigm and assume that these conditions account for the functional characteristics in vivo. Thus, the challenge is to understand how the sarcomeres and all of the associated structures transmit the forces of the whole muscle to its attachments.

Mechanical and morphological properties of chronically inactive cat tibialis anterior motor units.

1. The lumbar spinal cord was functionally isolated in ten cats by cord transection at the junctions of segments T12‐T13 and L7‐S1 and cutting bilaterally all dorsal roots between the two transections. Two 24 h EMG recording sessions were used to verify that muscles in the lower limb were virtually electrically silent. The cats were maintained in excellent health for 6 months. 2. Six months after spinal cord isolation, an acute experiment was performed to isolate a single motor unit from the tibialis anterior of each hindlimb using ventral root splitting techniques. Each motor unit was characterized physiologically as either fast fatigable (FF, n = 11), fast fatigue resistant (FR, n = 4), fast intermediate (FI, n = 2), or slow (S, n = 1), and repetitively stimulated to deplete the motor unit of its glycogen. 3. Maximum tensions of the fast motor units were lower than mean maximum tensions of control, whereas the S motor unit remained within the range observed in controls. In general, the isometric contractile properties, as well as fatigability, were within the ranges for each of the motor unit types in control cats. The mean fibre cross‐sectional areas of the fibres within the FR and FF motor units were approximately 40 and 50% smaller than control, while the mean fibre size of the fibers within the S motor unit was similar to control. 4. Innervation ratios and specific tensions for all experimental motor units were within the ranges of those reported for tibialis anterior motor units in control cats. Thus, it appears that the decrease in maximum tension of the fast motor units was primarily related to a reduction in fibre size. 5. The spatial distribution of the fibres within fast motor units of a spinally isolated cat, as measured by interfibre distances of the motor unit fibres, was similar to that reported for control tibialis anterior motor units. 6. These data suggest that factors independent of activity play a prominent, if not dominant, role in maintaining the complement of motor unit types typical of adult cat muscles. In addition, normal innervation patterns appear to be maintained in the absence of activity.

Extensor- and flexor-like modulation within motor pools of the rat hindlimb during treadmill locomotion and swimming

EMG activity was recorded from the vastus lateralis (VL, knee extensor), rectus femoris (RF, hip flexor and knee extensor), tibialis anterior (TA, ankle flexor and digit extensor) and either the lateral or medial gastrocnemius (LG, MG, knee flexors and ankle extensors) muscles of 7 adult rats during treadmill locomotion and swimming. Most flexors and extensors are activated as a single burst but each is known to be modulated differently during locomotion. For example, the extensor EMG bursts are shortened and amplitude elevated as speed increases, whereas little change occurs in the EMG duration and amplitude in flexors. The RF and VL displayed a double burst of EMG activity per cycle during treadmill locomotion and a single burst during swimming. Kinematic and EMG analyses showed that during running, one of these EMG bursts occurred primarily during swing while the other burst occurred primarily during stance. Modulation of the burst occurring during swing approximated a flexor pattern, while the second burst was modulated like a typical extensor when running over a range of speeds and grades on a treadmill. These data suggest that motoneurons within a motor pool of a uniarticular (VL) as well as a biarticular (RF) muscle can be modulated by more than one cyclical input, probably of central origin, and that under some conditions several motor pools may share the same central commands.

Training effects on soleus of cats spinal cord transected (T12–13) as adults

Adult spinal cord transected (T12–13) cats were trained for 30 min/day, 5 days/week to either step on a treadmill (Stp‐T) or stand (Std‐T) for ∼5 months. Training ameliorated soleus atrophy and enhanced maximum force capability compared to nontrained (N‐T) spinal cats, with Stp‐T being significantly different from N‐T. Isometric twitch speed and maximum rate of shortening were unaffected by training; the soleus of all spinal groups was significantly faster than control. There was an elevation in myosin adenosine triphosphatase activity and a shift toward faster myosin heavy chain and fiber type compositions in N‐T and Std‐T, but not Stp‐T cats. Thus, rhythmical activity involving muscle length and force changes (stepping) was more effective than a similar amount of a more static activity (standing). This specificity related to the type of training should be considered when developing rehabilitative strategies following spinal cord injury.

Relationship between ankle muscle and joint kinetics during the stance phase of locomotion in the cat.

The purpose of this study was to examine the relationship between internal force production in selected skeletal muscles and the externally calculated joint moment during overground locomotion in the adult cat. Hindlimb segments were modelled as a linked system of rigid bodies and a generalized muscle moment (GMM), the sum over all active and passive tissues acting about the joint, was calculated using principles of inverse dynamics. Moments produced by individual muscles were calculated using tendon transducers implanted in freely moving cats and muscle moment arm information. Results indicated that the externally measured variables of peak ground reaction force and joint position were equally important to the determination of peak ankle GMM. Examination of peak moments revealed that increases in peak ankle GMM were met by increases in medial (MG) and lateral (LG) gastrocnemius output. Peak soleus (SOL) moments did not change significantly as a function of peak ankle GMM. The role of the plantaris (PLT) was less clear, with peak moments increasing significantly as a function of peak ankle GMM in one cat. All four ankle extensors were important to the attainment of peak ankle GMM early in stance. Subsequently, SOL and PLT contributed substantially to the ankle GMM throughout stance, LG moments declined to near zero, soon after peak ankle GMM; and MG moments demonstrated a substantial but more gradual decline. The relative contributions of these individual muscles to the ankle GMM were supported by their respective architecture, uniarticular versus multiarticular function, and physiological profiles.

Does Motor Learning Occur in the Spinal Cord

It is becoming clear that the plasticity of the sensory-motor networks of the adult mammalian lumbosacral spinal cord is much greater than and is more dependent on the specific patterns of use than has been previously assumed. Using a wide variety of experimental paradigms in which the lumbar spinal cord is isolated from the brain, it has been shown that the lumbosacral spinal cord can learn to execute stepping or standing more successfully if that specific task is practiced. It also appears that the sensory input associated with the motor task and/or the manner in which it is interpreted by the spinal cord are important components of the neural network plasticity. Early evidence suggests that several neurotransmitter systems in the spinal cord, to include glycinergic and GABAergic systems, adapt to repetitive use. These studies extend a growing body of evidence suggesting that memory and learning are widely distributed phenomena within the central nervous system. NEUROSCIENTIST 3:287–294, 1997