Ulrich Bässler

Pattern generation for stick insect walking movements : multisensory control of a locomotor program

8. Discussion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 8.1. Distinction between central, peripheral and coordinating influences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84 8.2. Generation of reflex reversal by a distributed network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 8.3. Decisions in modular systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85

The walking-(and searching-) pattern generator of stick insects, a modular system composed of reflex chains and endogenous oscillators

Stick insects (Cuniculina impigra) possessing only one front leg with restrained coxa performed searching movements or “walked” on a treadband. The movements are described. Ablation, surgical manipulation or experimental stimulation of different sense organs (femoral chordotonal organ, campaniform sensilla on trochanter and femur basis, proprioceptors at the coxatrochanter joint) were performed, and the resulting changes in motor output were recorded. These experiments demonstrate that the walking- and searching-pattern generators cannot be separated, at least not for the movements investigated. This walking- and searching-pattern generator consists of central modules, each of which produces irregular alternation of the activity of motor neurones of antagonistic muscles of a single joint, and of “reflex loops”. At least some of these reflex loops are only present in the active animal. They are responsible either for the control of a single joint or for the coordination of the movements of separate joints. The performance of these reflexes does not only depend on the state of activity of the animal; some of them additionally seem to depend on the context signalled by other sense organs.

Afferent control of walking movements in the stick insectCuniculina impigra

Summary1.On a motor-driven treadband decerebrateCuniculina usually walked regularly, but only forwards. Animals having only the two forelegs showed the same behavior. Decerebrate animals with just the two hindlegs walked only backwards.2.In the duration of the electrical activity of the depressor and levator muscles of the trochanter and the extensor and flexor muscles of the tibia, there was no significant difference between the forward walking forelegs of these animals and the active walk of the forelegs of an intact insect.

Interruption of searching movements of partly restrained front legs of stick insects, a model situation for the start of a stance phase?

Stick insects (Cuniculina impigra) possessing only one foreleg with restrained coxa performed searching movements. A force transducer was introduced as an obstacle into the plane of movement of femur or tibia. Depending on where it was introduced and whether it was touched for the first time during an upward or a downward movement, different kinds of behaviour of the leg were released. For these different movements, the forces applied to the obstacle and the electrical activity of the depressor, levator, retractor and protractor muscles are described. In addition the alterations occurring after ablation of several sense organs including the trochanteral campaniform sensilla are mentioned. The described movements were similar to the corresponding behaviours during walking at the end of swing phase and the beginning of stance phase. Therefore there is some probability that results obtained by this experimental paradigm can also be applied to the swing-stance transition.

Leg movements of stick insects walking with five legs on a treadwheel and with one leg on a motor-driven belt

Five legs of a fixed stick insect walked on a double treadwheel. The left hindleg (L3) walked on a motor-driven belt. When the belt was slower than the wheels L3 made less steps than the other legs and when the belt was faster than the wheels it made more steps than the other legs. In the case of slowlier stepping of the “belt-leg”, the motor neurons of the retractor coxae muscle of this leg showed a high activity when the leg was pulled backwards by the belt. This activity was modulated in the step rhythm of the “wheel-legs”. When all legs showed the same stepping frequency (1:1-coordination) the protraction duration of L3 was almost independent of step-period, as well as the lag between onset of protraction of L3 and that of L2. In some cases only L3 could be made to step on the belt even when all other legs did not walk.

Sensory influences on the coordination of two leg joints during searching movements of stick insects

Animals (Cuniculina impigra) possessing only one foreleg with restrained coxa perform very stereotyped searching movements during which the movements of the femur-tibia and coxa-trochanter joints are well coordinated. After ablation of either hairfield BF1 (measuring the position of the coxa-trochanter joint) or the apodeme of the femoral chordotonal organ (measuring the position of the femur-tibia joint) each joint can still be moved but the coordination changes and becomes very labile. The consequences for the ideas about the construction principles of the pattern generator for searching movements are discussed.

Distributed processing on the basis of parallel and antagonistic pathways simulation of the femur-tibia control system in the stick insect.

In inactive stick insects, sensory information from the femoral chordotonal organ (fCO) about position and movement of the femur-tibia joint is transferred via local nonspiking interneurons onto extensor and flexor tibiae motoneurons. Information is processed by the interaction of antagonistic parallel pathways at two levels: (1) at the input side of the nonspiking interneurons and (2) at the input side of the motoneurons. We tested by a combination of physiological experiments and computer simulation whether the known network topology and the properties of its elements are sufficient to explain the generation of the motor output in response to passive joint movements, that is resistance reflexes. In reinvestigating the quantitative characteristics of interneuronal pathways we identified 10 distinct types of nonspiking interneurons. Synaptic inputs from fCO afferents onto these interneurons are direct excitatory and indirect inhibitory. These connections were investigated with respect to position and velocity signals from the fCO. The results were introduced in the network simulation. The motor output of the simulation has the same characteristics as the real system, even when particular types of interneurons were removed in the simulation and the real system.

Effects of crossing the receptor apodeme of the femoral chordotonal organ on walking, jumping and singing in locusts and grasshoppers

SummaryInSchistocerca gregaria andOmocestus viridulus ♂ the receptor apodeme of the femoral chordotonal organ was moved from its natural origin on the tibia, dorsal to the axis of rotation of the joint, and fixed to the apodeme of the flexor tibiae muscle. The femoral chordotonal organ then signaled the opposite of a real movement of the tibia. During walking middle or hind legs manipulated in this way were often held up with fully extended, immobile femur-tibia-joints for many steps of the other legs. Animals with operated hind legs jumped well, but more rarely. The number and duration of the songs were smaller in operated animals, but there were only a few weak influences on the way the legs were moved during singing. The type of motor programme in use determined the response to a particular afference (programme-dependent reaction).

Information processing in the femur-tibia control loop of stick insects

The complicated response characteristics of the identified nonspiking interneuron type E4 upon elongation stimuli to the femoral chordotonal organ (fCO) can be obtained by a computer simulation using the neuronal network simulator BioSim, if the following assumptions were introduced: (1) The interneurons receive direct excitatory input from position- and velocity-sensitive fCO afferents but also, in parallel delayed inhibition from the same velocity-sensitive afferents. (2) Position-sensitive afferents in part show adaptation with a rather long time-constant. A subsequent experimental analysis demonstrated that all these assumptions fit the reality: (1) Interneurons of type E4 receive direct excitatory input from fCO afferents. (2) Interneurons of type E4 are affected by velocity dependent delayed inhibitory inputs from the fCO. (3) The fCO does contain adapting position-sensitive sensory neurons, which have not been described before. The described principle of the information processing is also able to generate the response in interneurons of type E6 with less steep amplitude-velocity characteristic due to a different weighting of the direct excitation and delayed inhibition.

Flexibility of a Proprioceptive Feedback System Results from its “Parliamentary” (Distributed) Organization

For a long time neurobiology of vertebrate and invertebrate locomotor and rhythm generating systems has focused on the neural basis of fairly stereotype aspects of behaviors and motor patterns. However, since about 15 years task-dependent flexibility and adaptivity in the production of motor programs has merged increasingly into focus. Consequently an increasing number of investigations has started to enlighten the mechanisms of this flexibility (for review see Pearson 1993). Along with these investigations theories of neuronal architectures for adaptive behavior emerged (Morton & Chiel 1994). Three different neuronal architectures have been distinguished that can be involved in the generation of flexibility: 1. For distinct types of motor programs there exist specific neuronal networks called “dedicated circuits”. In such networks all pathways involved support the performance of motor responses generated. Evidence for this kind of information processing derived from initial findings on the generation of postural reflexes by direct connections between afferent neurons and motoneurons in the cat by Eccles, Lloyd and coworkers (for review see Schomburg 1990). From these findings and subsequent investigations on other vertebrate and invertebrate sensory-motor systems the notion emerged that many if not all reflexes may be generated by dedicated circuits (e.g., Burrows 1992).