Zill Lab NSF Research

RESEARCH PROJECT: DETECTING LOAD IN PEDAL EXTREMITIES: STRUCTURE AND RESPONSES OF SENSE ORGANS THAT ENCODE FORCES IN THE TARSI OF COCKROACHES

B.R. Keller, D. Neff, E.R. Duke, A.S. Amer, S.N. Zill*, Dept. Anat. and Pathol., J.C. Edwards Sch. Med., Marshall Univ., Huntington, WV, USA

ABSTRACT:- Our research is characterizing how forces are detected in the legs of cockroaches. Recent studies in a variety of animals have shown that sensory inputs from the feet can provide important information in the control of posture and walking and may also aid in adapting motor responses to variations in the substrate. To characterize information provided by sense organs of the feet and compare it with responses of receptors detecting forces in the proximal leg, we are recording sensory activities in the tarsus by wires placed in the first tarsal segment. Units are confirmed as sensory by recordings taken simultaneously in the tibia, which show 1:1 spiking at a delay due to centrifugal conduction. Forces are applied to the tarsus in restrained preparations using a probe mounted to a piezoelectric crystal. Recordings regularly show a single unit with a particularly large spike that responds to forced extension of the joint between the fourth and fifth tarsal segments (Ta45). The frequency of afferent firing encodes both the rate and amplitude of applied force. Control experiments strongly suggest that this activity is derived from a campaniform sensillum located on the distal condyle of the fourth tarsal segment (Ta4): ablation of the tarsus distal to the joint does not eliminate the sensory discharge while ablation of the cuticle of Ta4 eliminates all responses of equivalent amplitude. In addition, small indentations of the tarsus in the region of the cuticular cap of the sensillum produce action potentials of amplitude equivalent to that seen in bending tests. We have also studied the structure of the distal tarsus by histological section and confocal microscopy and characterized movements of the joint by high speed video imaging. These studies suggest that the Ta45 joint serves as a fulcrum for movements of the distal tarsus. However, forces are transmitted through the joint only when the claws and arolium are engaged with the substrate by the action of the pretarsal depressor (retractor unguis) muscle. We are planning to study responses of tarsal sense organs in freely standing animals during tests in which body load is changed by attaching a magnet to the dorsum of the animal and applying currents to a coil below the arena. Preliminary tests show that decreases in body load can produce firing of a large unit when animals can grasp the substrate and that this discharge is eliminated by ablation of cuticle at the Ta45 joint. Sensory inputs from the tarsi may, therefore, provide information about the forces exerted by the leg within the frame of reference of the foot and potentially aid in adapting motor responses to properties of the substrate. Support Contributed by: NSF Grant IBN-0235997

Introduction - How are forces detected and regulated in posture and locomotion? Many recent studies have emphasized the importance of sensory inputs from the feet in postural control. In the present study, we have sought to characterize sensory receptors that can encode forces acting on the tarsi of cockroaches. We have found that a campaniform sensillum located near an intrinsic tarsal joint has an exceptionally large extracellular action potential that can be readily identified in sensory recordings. This sensillum responds to forces that are imposed upon the tarsus and has response properties similar to those seen in the tibial campaniform sensilla, receptors that detect forces in the proximal leg. In addition, the tarsal campaniform sensillum also fires vigorously to cuticular strains produced by contractions of the pretarsal depressor (retractor unguis) muscle. We have also begun studies to characterize the responses of the receptor in freely moving animals.

Multimodal Integration in Load Compensation - Many studies support the idea that the nervous system uses multimodal integration of sensory inputs to control and adapt standing and walking. A number of types of receptors in the legs can detect perturbations of posture including sense organs, such as Golgi tendons organs, that directly encode forces. Other types of sense organs, such as muscle spindles, detect changes in muscle length and can signal the effects of forces on the legs. However, in some perturbations, such as substrate rotation in standing humans, muscles are activated that are antagonists to the muscles that are stretched. These studies have suggested that receptors of the feet may detect ground reaction forces directly and provide inputs that allow for directional tuning of compensatory responses.

Adaptation of posture and walking to properties of substrate - Sense organs of the feet may also play a role in the adaptation of posture and walking to the specific properties of the substrate. In stick insects, forces generated by a leg during walking depend upon the mechanical properties (stiffness) of the substrate. Much larger forces are produced if the substrate is stiff, rather than soft. Insects can also vary the effective friction with the substrate by actively grasping the surface with the claws. The effects of this mechanism on motor outputs and compensatory reactions have not been systematically studied.

Force Detection - We have studied load detection in insects. In cockroaches, sense organs that signal the effects of forces acting upon the leg (campaniform sensilla) can be recorded in freely moving animals. We have recorded activities of the tibial group of campaniform sensilla during perturbations that change body load. These studies have shown that, in upright posture, individual receptors discharge to either load increases (proximal tibial receptors) or load decreases (distal sensilla). These experiments also suggest that the sensilla can provide information over a wide range of variations in body load. In addition, the tibial sensilla show strong responses to contractions of leg muscles. However, it has not been clear how forces produced by muscle contractions influence responses to load. We have also not systematically tested how the effects of the tibial sensilla are integrated with inputs from force receptors of other leg segments.

Anatomy and movements of the tarsus - The cockroach tarsus is composed of five segments (tarsomeres numbered Ta1-5) and the pretarsus (claws and arolium). The tarsus has no intrinsic muscles but the claws are actively engaged with the substrate by the pretarsal depressor (retractor unguis) muscle. This muscle takes origin in the proximal leg and inserts on the pretarsus. Because the retractor tendon (apodeme) traverses all tarsal segments, it produces depression at each tarsal joint and brings the segments closer together (more tightly packed). Most of the intrinsic tarsal joints permit only limited movements but the joint between the fourth and fifth tarsal segments (Ta45) is highly mobile. The fifth tarsal segment simply abuts against the condyle ('backstop') of Ta4, which serves as a fulcrum, allowing the pretarsus to be engaged at a number of angles. While depression of Ta45 is actively generated, elevation is produced by joint elasticity. Previous studies have shown that a resilin ligament at the joint is stretched by joint depression and lifts the distal tarsus when released.

The tarsi of insects contain a large number of sensory receptors but a limited number of proprioceptive sense organs. A large chordotonal organ is found in the distal tarsus and encodes the position and movement of the claws. We have confirmed the early account of Pringle and found that campaniform sensilla are located on the distal end of each of the tarsal segments adjacent to the joint condyles. In whole mount preparations, the cap of the campaniform sensillum at the end of Ta4 appears to be linked to the joint membrane and backstop. Measurements in wholemount preparations show that the cap size of the Ta4 sensillum is comparable to that found in the largest proximal and distal tibial campaniform sensilla.

Sensory activities were recorded by pairs of 50 micron silver wires implanted into the first tarsal segment and tibia. The use of two sets of recordings allowed us to insure that activity was sensory, as it occurred first in the tarsus and was followed 1:1 by spikes in the tibial recording after a delay due to centripetal conduction. In studies on isolated legs, we were also able to record the activities of the Ta45 campaniform sensillum extracellularly through pins placed through the tibia adjacent to branches of the main leg nerve. Forces were applied to the tarsus in restrained preparations using a probe mounted to a piezoelectric crystal. Recordings regularly showed a single unit with a particularly large spike that responded to forced extension of the joint between the fourth and fifth tarsal segments (Ta45). Activities of sensory units of smaller amplitude that respond to the mechanical stimulus were also present in most recordings. Control experiments strongly suggested that the activity of the largest spike was derived from a campaniform sensillum located on the distal condyle of the fourth tarsal segment (Ta4): ablation of the tarsus distal to the joint did not eliminate the sensory discharge while ablation of the cuticle of Ta4 eliminated responses of equivalent amplitude. In addition, small indentations of the tarsus in the region of the cuticular cap of the sensillum produced action potentials of amplitude equivalent to that seen in bending tests. In some tests, the activity of the smaller units persisted after ablation at the fourth tarsal segment but was eliminated by similar ablation of the distal cuticle of the third tarsal segment at the location of the cuticular caps of the homologous group of campaniform sensillum.

The tarsal campaniform sensilla typically showed little tonic activity at rest. Forces applied to the fifth segment using rapid half sine waveforms produced intense bursts at very short latencies. The histogram at right plots the mean response frequency for 184 tests using half sine wave functions in 3 animals. Firing was initiated immediately following the onset of force application and typically ceased or decreased to a very low level when force levels declined. In contrast, previous studies showed that the proximal tibial sensilla were tonically active and fired at higher frequencies in more prolonged bursts to imposed forces.

Sensitivities to the rate of force application were studied by application of forces using ramp and hold functions of varying rates of rise and decline. All recordings showed a strong sensitivity to the rate of force application. Log-log plots of these tests showed that the dynamic sensitivities were comparable to that previously found in the tibial campaniform sensilla. Similar results were also obtained in tests in which forces were applied using sine wave functions.

We also characterized the sensitivity of the sensillum to the magnitude of forces by applying ramp and hold stimuli that had similar rates of rise but varying amplitudes. Receptors typically fired only phasically to low levels of force but discharged tonically to higher force amplitudes. Plots of the discharge during the hold phase show that the tonic firing encodes the magnitude of force and that range of response frequencies were similar to that obtained from the distal tibial sensilla.

We have begun testing responses of the tarsal campaniform sensilla to forces generated by the resisted contractions of retractor unguis in restrained animals by having the pretarsus grasp a force probe or by placing the probe against the fifth tarsal segment. Myographic activity of the retractor muscle is recorded in the femur. The tarsal sensilla show vigorous discharges when contractions of the retractor muscle are resisted. The anatomy of the leg suggests that muscle contractions produce strains in Ta4 as the apodeme pulls against the backstop and draws the segments into a closely packed position.

We have also initiated experiments to record the activities of the sensilla in freely standing animals. Preliminary tests of responses to changes in body load were performed using a small magnet on the dorsum of the animal and a coil placed below the arena. The tarsal campaniform sensillum did not discharge to increased loads but fired when the animal was lifted from the substrate. Animals resisted these forces by grasping the wire mesh on the bottom of the arena. Preliminary tests suggest that these discharges result from contractions of the retractor unguis muscle, as they can also be elicited when forces act to pull the animal from the side of the arena when it is grasping the wall of the cage. Control experiments confirmed the identity of the sensory discharge: activity was eliminated by cap ablation. Tests using half sine wave functions could elicit bursts of firing but at longer latencies than seen in restrained preparations. Sensilla also showed strong sensitivity to the rate of force application when single postures were held through repeated perturbations but bursts of activity were elicited at longer latencies.

1- This study has characterized responses of campaniform sensilla that can encode forces in the tarsi of cockroaches. The design of the tarsus allows for flexible coupling to the substrate via the claws, which can be engaged by contraction of the retractor unguis muscle. Extracellular recordings from sensory nerves in the first tarsal segment show a sensory unit with a large spike that does not discharge at rest. However, the receptor responds with intense phasico-tonic discharges to forces applied to the fifth tarsal segment. Tests using half sine or sine waveforms show that firing does not occur to force decreases. Control experiments confirm that the activity is derived from a single large campaniform sensillum at the Ta45 joint.

2- Tests applied as ramp and hold functions show that the phasic discharge can encode the rate of force application. The tonic frequency also can reflect the level of force. The receptor also discharges to resisted contractions of the retractor unguis (depressor pretarsus) muscle. These characteristics are similar to those obtained from the Group 6 tibial sensilla.

3- Preliminary tests using recordings in freely standing animals show that the receptor fires when forces are applied that lift the animal away from the substrate. These discharges occur at long latencies and apparently depend upon the engagement of the claws with the substrate. Our working hypothesis is that the receptor encodes the net forces transmitted through the leg when the claws engage the substrate. This hypothesis is compatible with a model of action and force generation by the retractor unguis muscle. We are planning to test this hypothesis in studies of freely standing animals and in tests in which forces are applied directly to the muscle apodeme.

POSTER FROM SOCIETY FOR NEUROSCIENCE MEETING