Neural control of locomotion: sensory control of the central pattern generator and its relation to treadmill training
Introduction
A common approach to investigate the neural basis of animal locomotion, is to make use of surgically reduced animal preparations. In this way the role of separate parts of the central nervous system (CNS) in rhythm generation can be determined. Studying such preparations makes it possible to answer important questions such as ‘is it possible to express stereotypic and locomotor related patterns in these isolated preparations i.e. in the absence of the eliminated neural structures?’ and ‘what is the relation of these patterns to those observed in the intact animal?’. First, the different types of reduced preparations, as used in this review, will be dealt with shortly.
Within the ‘decerebrated’ preparations, several classes can be distinguished. The classes used in this review are the ‘pre-mammillary or thalamic’ and ‘post-mammillary or mesencephalic’ preparation. The pre-mammillary and post-mammillary preparations are transected above and between the colliculi and the subthalamic nuclei in the brainstem, respectively. The pre-mammillary preparation exhibits spontaneous periods of locomotion in response to a moving treadmill. The post-mammillary preparation does not walk spontaneously. Stepping movements can be elicited by electrical stimulation of the mesencephalic region (MLR) of the brainstem 1, 2. Both types of the decerebrate preparations can be used to study the locomotor output in the absence of movement related feedback. During this so-called ‘fictive’ locomotion, feedback is completely eliminated through blocking of the movement. This can be achieved by either injection of neuromuscular relaxants [3]or transection of the efferent nerves at the ventral root or at muscle nerve level. In these preparations, locomotor output can be evoked only by electrical stimulation of the MLR region [4]or pharmacologically with substances mimicking the action of descending pathways (noradrenergic agonists and/or precursors: l-DOPA+nialamide or clonidine; [5]). In spinalized cats the connections between the neural networks at supraspinal and spinal level are interrupted by transection of the spinal cord. The level of transection can differ but it is mostly at thoracic level. Spinalized cats can restore locomotor function of the hind-limbs when trained on a treadmill. This stepping can be evoked or enhanced with the use of drugs. This will be treated in the later part of this review more extensively.
Before evaluating results of experiments on the various preparations one has to bear in mind some limits when these results are used to understand the neural system of the intact animal.
When recovery is allowed, the reduced preparation has the capability to reorganize. Let us consider first the spinalized animal. Even if the structures above and below the lesion are still in a functional preserved state, the capabilities of the system to function as one total system have changed drastically [6]. Hence, it is far too simple to consider the spinal cord below the lesion as being the same as in the intact animal, except for the absence of interaction with supraspinal levels. Basically this view does not take into account the potential of the spinal cord to reorganize (see also [7]on this point). Therefore, it is fair to argue that the locomotor output of a chronic spinal cat (i.e. spinalized cat which was trained regularly on a treadmill and recovers locomotor ability) may not be based on exactly the same circuitry as the locomotor output of the intact animal. It is, indeed, conceivable that some of the patterns observed in chronic spinal cats reflect the capacity of the nervous system to cope with an injured system.
Furthermore, the sensory related input to the CNS of walking spinal cats can differ considerably from normal. Although the basic locomotor pattern can be present in fictive locomotion (i.e. in complete absence of afferent input), it will be shown that the role of afferents is very important in shaping the rhythmic pattern, to control phase-transitions and to reinforce the ongoing activity [8]. Therefore, a rhythm generating structure without its normal afferent input can be very artificial and possibly therefore cannot entirely reproduce the motor output as seen in the intact cat.
Despite these restrictions it will be shown that the locomotor output of chronic spinal cats is nevertheless strikingly similar to the one seen in intact cats. According to Edgerton et al. 9, 10, there are two main features that are different in the locomotion of chronic adult spinal cats as compared to normal. First there is a delay in the initiation of the swing phase resulting in the paw dragging over the treadmill. Second, the force and the EMG of the fast extensor muscles decline prematurely at the end of the stance phase. Interestingly, enhanced loading of the hind-limb muscles (for example by pulling the tail downwards) can compensate for both deficits. Hence, this suggests that the main differences are related to changes in afferent input rather than to spinal ‘rewiring’.
Recently, it has become possible to regain some locomotor activity in patients suffering from a spinal cord injury (SCI) through an intense training on a treadmill. These improvements are thought to be a consequence of a (re)activation of neural circuits located at spinal level. These neural networks are collectively termed central pattern generators (CPGs). CPGs are thought to be responsible for the generation of the basic rhythmic muscle activity as seen during locomotion. The evidence for such a spinal CPG in both cats and humans is reviewed in part 1.
Since afferent input is important in refining the locomotor output, it is conceivable that the provision of adequate sensory input during such treadmill training is of the utmost importance to achieve a more optimal locomotor output of the spinal locomotor circuitry. Hence, it is essential to learn more about how such CPGs are controlled by sensory input produced during gait.
The next section will deal with the influence of afferents on the cats’ CPG. Next, the results of treadmill training for spinalized cats, monkeys and humans are treated. Finally, the treadmill training results and their relation to sensory inputs are discussed.
Section snippets
The influence of afferents on the cat CPG
In cats, it has long been known that the spinal CPG can be activated by a variety of input sources. Grillner and Zangger [11]showed that electrical stimulation of dorsal roots could effectively trigger the initiation of locomotion in spinal cats (similar experiments in mesencephalic cats were reported earlier by Budakova [12]). Furthermore, many authors have confirmed that spinal locomotion can be either induced or facilitated through ‘a-specific’ sensory stimulation (which usually consists of
Chronic spinalized cats
After transection of the spinal cord of adult cats and dogs at a thoracic level (typically Th 12) or at upper lumbar level and after recovery from spinal shock, the so-called ‘low spinal’ cats and dogs are able to generate rhythmic alternating movements of the hind-limbs. In the earliest experiments, this stepping was very limited because these animals could not fully support themselves 44, 45. However, Ten Cate 46, 47, 48provided additional support for these animals. The spinalized cat was
Afferent influence on human CPG during treadmill training
When BWS treadmill training was started in humans, it was suggested that locomotor related afferent input could be of importance for the enhancement of locomotor activity. Several studies mentioned, only at a qualitative level, the relationship between the locomotor related load and kinematic input and the observed locomotor output. In subjects with one completely paralyzed lower limb, a flexor-like stepping movement could be elicited in the paralyzed leg when the SCI person shifted the body
Alternative explanations for locomotor recovery on the treadmill
Although it is tempting to conclude that the results following treadmill training, as described above, can be ascribed to activation of CPGs, one first has to consider some alternative explanations.
Conclusions
As in cats, improvement of locomotor function is seen in both complete and incomplete SCI persons when intensively trained on a treadmill by using BWS. This suggests that spinal locomotor centres can be activated in paraplegic patients in a manner similar to that in the chronic spinal cat after training on a treadmill. The most important features are a better modulation pattern of EMG activity in muscles of the lower extremities and the increased ability to support body weight. The question
Acknowledgements
We want to thank B. Bussel for providing us with some very valuable references and for general guidance, D.A. Winter and V. Dietz for helpful discussions and comments. B. Van Wezel and S. Donker for reading this paper and for providing critical remarks. This work was supported by a grant from STW and Sint Maartenskliniek to HC and from Nato (twinning grant 910574) to JD.
References (81)
- et al.
Main characteristics of the hind-limb locomotor cycle in the decorticate cat with special reference to bifunctional muscles
Brain Res
(1980) - et al.
Locomotion evoked by brain stem stimulation: occurrence without phasic segmental afferent input
Brain Res
(1979) - et al.
Sensorimotor stimulation to improve locomotor recovery after spinal cord injury
TINS
(1997) - et al.
Co-ordination of motor pools controlling the ankle musculature in adult spinal cats during treadmill walking
Brain Res
(1991) Proprioceptive regulation of locomotion
Curr Opinion Neurobiol
(1995)Control of locomotion in the decerebrate cat
Prog Neurobiol
(1996)- et al.
Inhibition of flexor burst generation by loading ankle extensor muscles in walking cats
Brain Res
(1980) - et al.
On the initiation of the swing phase of locomotion in chronic spinal cats
Brain Res
(1978) - et al.
Recovery of locomotion after chronic spinalization in the adult cat
Brain Res
(1987) - et al.
Weight-bearing hind-limb stepping in treadmill-exercised adult spinal cat
Brain Res
(1990)
Consequences of spinal cord lesions upon motor function with special reference to locomotor activity
Prog Neurobiol
Locomotor activity in spinal man
Lancet
The effects of serotonergic drugs on the locomotor pattern and on cutaneous reflexes of the adult chronic spinal cat
Brain Res
Biceps femoris tendon jerk reflexes are enhanced at the end of the swing phase in humans
Brain Res
Modification of human spinal stretch reflexes: preliminary studies
Neurosci Lett
Control of walking and running by means of electrical stimulation of the mid-brain
Biophysics
Neurophysiology of locomotor automatism
Physiol Rev
The effects of DOPA on the spinal cord. 5. Reciprocal organization of pathways transmitting excitatory action to alpha motoneurones of flexors and extensors
Acta Physiol Scand
Model for the study of plasticity of the human nervous system: features of residual spinal cord motor activity resulting from established post-traumatic injury
Ciba Found Symp
Common principles of motor control in vertebrates and invertebrates
Ann Rec Neurosci
Trainability of the spinal cord to generate stepping patterns in adult spinalized cats
Recovery of full weight-supporting locomotion of the hind-limbs after complete thoracic spinalization of adult and neonatal rats
On the central generation of locomotion in the low spinal cat
Exp Brain Res
Stepping movements evoked by a rhythmic stimulation of a dorsal root in mesencephalic cat
Neurosci Behav Physiol
Can the mammalian lumbar spinal cord learn a motor task?
Med Sci Sports Exerc
Neural control of stereotypic limb movements
A physiological basis for the development of rehabilitative strategies for spinally injured patients
J Am Paraplegia Soc
Stimulation of the group I extensor afferents prolongs the stance phase in walking cats
Exp Brain Res
Proprioceptive input resets central locomotor rhythm in the spinal cat
Exp Brain Res
Reversal of the influence of group Ib afferents from plantaris on activity in medial gastrocnemius muscle during locomotor activity
J Neurophysiol
Transmission in a locomotor-related group Ib pathway from hind-limb extensor muscles in the cat
Exp Brain Res
Disynaptic group I excitation of synergist ankle extensor motoneurones during fictive locomotion
J Physiol
Entrainment of the locomotor rhythm by group Ib afferents from ankle extensor muscles in spinal cats
Exp Brain Res
Ankle extensor group I afferents excite extensors throughout the hind-limb during fictive locomotion in the cat
J Physiol
Function of segmental reflexes in the control of stepping in cockroaches and cats
Positive force feedback control of muscles
J Neurophysiol
Indications of positive feedback in the control of movement
J Neurophysiol
Reflexes induced by nerve stimulation in walking cats with implanted cuff electrodes
Exp Brain Res
Cited by (176)
Synergy-based knee angle estimation using kinematics of thigh
2021, Gait and PostureSingle channel EMG-based continuous terrain identification with simple classifier for lower limb prosthesis
2019, Biocybernetics and Biomedical EngineeringHuman Lower Limb Motion Intention Recognition for Exoskeletons: A Review
2023, IEEE Sensors Journal