ABSTRACT - Compensation for load is necessary for posture and adaptive locomotion but may be limited during rapid movements. We are studying mechanisms underlying support of body load in cockroaches. Body weight is increased in freely moving animals by small magnets attached to the thorax and varied by a coil below the substrate. Our previous studies showed that 1) increases in body load in freely standing cockroaches can produce activation of leg receptors that detect forces (campaniform sensilla) and the trochanteral extensor muscle in all legs and 2) tonic increases in load enhance sensory and motor activities during the stance phase when animals walk slowly. We are currently examining the effects of 1) tonic load increases in rapid walking and 2) sudden load changes applied during stance. Animals are videotaped at high speed to measure kinematic parameters and body position. Sensory and motor activities are recorded neurographically. Animals with tonically increased load that are walking rapidly show substantial decreases and greater variability in height above the substrate than is seen in slow walking. Our data to date indicate that extensor firing frequencies are less adapted to load increases at higher walking rates. These effects are correlated with the rate of walking: cockroaches can change from poor (low and variable body height) to adequate compensation (increased height, decreased fluctuations) within single sequences if the walking rate is slowed. Animals also readily adapt to sudden load increases that are applied during slow walking and show transient decreases in body height and modulation of motoneuron firing frequencies. We are currently analyzing sensory and motor discharges that occur during perturbations at different walking rates. However, our results to date indicate that cockroaches can compensate for body load when walking slowly but do not adapt to load increases when running. These findings are consistent with the idea that running is a centrally determined program which is effective in escape but poorly modifiable by changes in load. Support Contributed By: NSF Grant IBN-0235997

FORCE CONTROL: Our project is studying how animals sense forces that act upon the legs and how this information is used to adapt posture and locomotion. When an animal is standing, the largest forces on the legs result from the effects of body weight. These forces are not constant but change when different postures are assumed. They also vary continuously in locomotion when legs are alternatively lifted and placed upon the substrate.  The nervous system is thought to compensate for these variations by adjusting activities of leg and body muscles using information from leg sense organs. The specific functions of force receptors in these compensatory reactions have not been determined and some studies suggest that adaptations to loading are reduced in rapid running. In the present experiments we have examined the effects of tonic loading of the body in walking and running and have also begun studying the effects of sudden load increases upon motor activity.

DETECTION OF BODY LOAD WHEN STANDING: Many studies support the idea that the nervous system can generate postural adjustments to changes in load using information from a variety of sensory modalities.    A number of types of receptors can detect loading of legs, including sense organs that directly encode forces, such as muscle tendons organs or receptors of the sole of the feet.  Other types of sense organs can signal the effects of forces on legs through changes in body position, muscle length or joint movements (muscle spindles, joint angle receptors).

SUPPORT OF BODY LOAD IN COCKROACHES: Our previous studies have shown that the trochanteral extensor muscle is strongly activated to compensate for increases in body load. The extensor also shows elevated activity in the stance phase of walking when body load is tonically increased by attaching weights to the thorax. The tibial campaniform sensilla, which detect forces in the legs, show similar sensitivites to changes in body load in standing and walking. In restrained preparations, activation of these receptors can excite the extensor motoneuron. Our working hypothesis is that these receptors can contribute to compensatory reactions by detecting increases in forces acting upon the legs before substantial changes in body position have occurred.

EFFECTS OF FEEDBACK ARE REDUCED IN RAPID WALKING AND RUNNING: Feedback from leg sense organs is essential in generating slow walking but effects of these inputs are thought to be reduced in rapid locomotion. In humans, electrical stimulation of leg nerves elicits H-reflexes through activation of muscle receptors. A number of studies have shown that these reflexes are significantly reduced in running.  In cockroaches, previous studies have shown that leg ablations that reduce sensory feedback can produce repeated leg placement and lifting (‘double stepping’) in slow walking. These effects are absent in running. However, the ability of insects to adapt to changes in body load during running has not been tested. 

MODEL OF FORCE DETECTION IN COCKROACH WALKING: We have studied load detection in insects, as sense organs that signal the effects of forces acting upon the limbs (campaniform sensilla) can be recorded in freely moving animals.  In posture, proximal sensilla of the front and hind legs can encode body weight.  In walking, proximal sensilla fire at stance onset and continue through midstance; distal sensilla fire short bursts at the end of stance when the leg is unloaded.  Our working hypothesis is that sensory discharges may result from different sources at different times in stance: firing early in stance can result from forces generated by contractions of muscles that press the leg as a lever against the substrate. Force feedback later in stance may reflect the use of the leg as a supportive strut when body weight is placed upon the limb.

CHANGING BODY LOAD WITH MAGNETS: An electrical coil is positioned below the arena and activated by currents that are generated by computer. Currents to the coil generate magnetic fields that attract or repel the magnet on the dorsum of the animal, which effectively varies the forces acting upon the legs. The forces generated by the coil are calibrated using magnets placed on a single axis force plate.  We also measured the effects of changing the distance between the magnet and coil (reflecting changes in the animal's height in the arena) by vertically raising and lowering the force plate and magnets using a micromanipulator while a series of ramp and hold waveforms were applied to the coil.  Raising the magnet decreased the forces generated by the coil and decrements in height resulted in force increases. However, these overall effects were small and resulted in approximately a 5% change in forces per 1 millimeter change in height. 

RAPID INCREASES IN BODY LOAD IN STANDING: We tested responses to sudden changes in load using waveforms that rose to a peak within 125 msec and returned back to the same level more slowly. The sequence below shows five perturbations applied to a standing animal. In each test, the stimulus onset is followed by a sharp increase in the firing frequency of the trochanteral extensor motoneuron in the middle leg. The perturbations also produced rapid and brief decreases in body height that reached minimum values following the peak of current application and slowly returned. The upper trace shows the rate of body movement (velocity) which reached negative peaks (downward movement) early in the perturbation and later positive values as the applied force declined and the body rose back to baseline. The histogram at right shows the average responses from 38 tests in 3 animals. The rate of motoneuron discharge did not follow the body position, which reached a minimum only slightly after the peak of current application. The maximum motoneuron firing, however, is nearly concurrent with the peak of downward velocity. These data support the idea that extensor firing is not strictly correlated with the amplitude of body height but is sensitive to velocity and the rate of force application. 

ANIMALS COMPENSATE FOR RAPID INCREASES IN LOAD WHEN WALKING SLOWLY: We have begun experiments testing the effects of sudden load increases in walking. Imposition of loads via the coil can produce decreases in body height and increases in firing frequency of the middle leg trochanteral extensor (Ds) during the stance phase. The sequence above shows extensor firing in a normal step and one in which current was applied as the animal walked over the coil. The onset of current application produces a rapid rise in extensor firing. The upper traces show the body position and velocity. The increase in body load produces a transient decrease in body height followed by a more gradual recovery. The middle traces show the patterns of leg lifting and placement for all the animal’s legs during this sequence. The perturbation produces an increase in the stance duration (arrows above trace) in the leg relative to that expected from the preceding step. The histogram at right shows the average motor frequency and body position during 12 tests from 3 animals. The extensor firing accelerates at short latency following the onset of the stimulus. As in standing animals, the change in body position is delayed relative to the current waveform and extensor firing reaches a maximum before the minimum peak of body displacement. The extensor discharge more closely approximates the movement velocity. The lower histogram plots the latency between the onset of current to the coil and the rise in extensor frequency to a level 1.25 times baseline. The latencies are equivalent in standing and walking. Thus, responses to sudden increases in body load are similar in posture and slow walking. 

PERTURBATIONS APPLIED DURING THE SWING PHASE DO NOT PRODUCE RAPID EXTENSOR ACTIVATION: Load increases applied during the swing phase of walking did not produce activation of extensor motoneurons at short latency but could have effects upon firing in the next stance phase. In addition, stimuli applied during very rapid walking could disrupt the pattern of extensor bursting. In the sequence below, the animal was walking rapidly and the load increase produced a fall (thorax touched) the substrate. Extensor firing was briefly inhibited at the peak movement and an additional burst was then initiated. Thus, compensation for rapid load increases may be less effective in rapid walking. 

ANIMALS WITH TONICALLY INCREASED LOAD DO NOT MAINTAIN BODY HEIGHT WHEN MOVING RAPIDLY: Animals with increased body weight that were carrying loads maintained their body height above the substrate when walking slowly. However, in many animals the body height was substantially lower when the walking rate increased. We studied this change by measuring the height of the base of the magnet to the substrate in still images taken from sequences when the middle leg was positioned under the magnet (early to mid stance). The histogram shows that the body height was significantly lower when the rate of walking was over 10 steps/second. The plot of body height vs step rate shows that the effect was not consistent as some animals rose and fell repeatedly in these sequences.  In many sequences, the effect was not all or none but animals increased their body height as the walking speed slowed. 

MOTONEURON ACTIVITY SHIFTS IN PHASE RELATIVE TO MOVEMENT DURING RAPID WALKING: Previous studies have shown that animals can use slow motoneurons in walking at moderately high rates of stepping  before fast motoneurons are recruited. In walking with no added weight at slow to moderate rates, the middle leg extensor was recruited immediately before foot placement and continued to fire through the stance phase. However, motoneuron activity shifted in phase during rapid walking and running. Firing was initiated earlier in swing and reached a peak relatively early in stance.   

EFFECTS OF TONIC LOAD INCREASES IN WALKING AND RUNNING: The effects of tonic increases in load were tested by adding magnets to the thorax to increase body weight by ~35%. The upper set of recordings are taken from an animal when it walked slowly first with minimum load (the default small magnet) and then increased load. The histogram at right shows the mean discharge of the trochanteral extensor in the step cycle. The effects of increases in load were specific: the discharge frequency was higher during stance and firing during the swing phase was not affected. The lower set of traces show similar recordings taken during rapid walking. The extensor fires earlier in the swing phase and bursts are terminated early in stance in both recordings. The histogram of extensor firing demonstrates that increased loading apparently produces a non-specific elevation in discharge rate throughout the burst. A plot of the mean change in firing of the extensor motoneuron after loading (below right) shows that compensation for increased body weight is reduced as the rate of stepping increases.

THE SHIFT IN PHASE OF MOTOR FIRING RELATIVE TO SENSORY FEEDBACK COULD CONTRIBUTE TO LESS EFFECTIVE ADAPTATION TO BODY LOAD DURING RAPID LOCOMOTION: To further characterize the timing of motor activities and sensory feedback, we obtained recordings in animals walking while held rigidly above an oiled glass surface. In slow walking, discharges of the slow extensor were similar to those seen in freely moving animals but firing often accelerated during a burst (due to increased resistance of the supported bar). In rapid walking, firing of the slow extensor shifted also in phase. In more rapid running the fast extensor was recruited and firing occurred most frequently early in stance. The lowest figure shows recordings of the tibial campaniform sensilla during rapid walking on oiled glass. Firing occurs only after foot placement, when forces developed by leg muscles act upon the substrate. Histograms of sensory discharges in slow and rapid walking show no shift in phase. Thus, discharge of the extensor muscle in rapid walking and running precedes both feedback from receptors monitoring movements and sense organs signaling force. This decrease in effectiveness could contribute to the lack of adaptation to body load during running.



1)  We have begun experiments in which body load is suddenly increased in animals during walking.  Increases in load applied in the stance phase produce rapid acceleration of firing of the trochanteral extensor motoneuron (Ds) of the middle leg. Comparison of responses of the same animals to perturbations applied in walking and standing show that the response latency of the initial acceleration is similar in both posture and locomotion. Load increases initiated during the swing phase do not produce rapid increases in extensor discharges although firing in the succeeding stance phase may be altered.  

2) Load increases also regularly produce transient decreases in body height. The extensor discharge more closely approximates the movement velocity than the body position, as is seen in tests of load increases in posture. Thus, the initial components of responses to rapid changes in body load are similar in standing and slow walking.

3) Animals readily compensate for tonic increases in body load when walking slowly. However, tonic loads produce significant decreases in body height when running. Experiments in which freely moving animals carried loads (~35% body weight) showed that, in slow walking, extensor firing is adjusted to load specifically after stance onset and firing during swing is largely unchanged. The magnitude of the compensatory discharge decreases as the rate of walking increases. Thus, the ability of the system to adjust motor outputs to load is apparently reduced in rapid walking or running.  

4) Activities of the slow extensor motoneuron (Ds) shift in phase relative to leg movements during rapid walking. In slow walking, extensor firing is initiated in swing and activity is sustained throughout the stance phase. In rapid walking, motoneuron bursts begin early in swing and end earlier in the stance phase. This phase shift could readily limit the ability of  the system to adjust extensor firing according to signals from sense organs that detect load after stance onset.

5) Future experiments will examine whether activities of tibial campaniform sensilla, which detect changes in body load in standing animals, are similarly activated by sudden load increases during walking. We will also examine whether compensation for sudden load increases is reduced when the effectiveness of sensory feedback is apparently decreased.



Copyright 2006 Joan C. Edwards School of Medicine. All rights reserved.