| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Zoologisches Institut Universität zu Köln, Cologne, Germany
Submitted 16 July 2004; accepted in final form 21 October 2004
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
; Orlovsky et al. 1999). For instance, in adult stick insects walking at moderate speeds, four legs are simultaneously in a stance phase forming a stepping pattern that corresponds to a tetrapod gait. With increasing speed, leg coordination shifts toward the stepping pattern of a tripod gait (Graham 1972, 1985; Wendler 1964).
This flexibility requires a sophisticated mechanism to coordinate the neural networks controlling leg movements. A basic component of those networks is an oscillatory neuronal circuit, a central pattern generator (CPG) that produces alternating activity in antagonistic motorneuron (MN) groups (Hooper and DiCaprio 2004; MacKay-Lyons 2002; Marder and Calabrese 1996; Pearson 1993, 2000). In stick insects, leg movements are generated by the activity of several CPGs, each of which controls the movement of individual leg joints (Bässler and Büschges 1998). The coordination among these CPGs required to produce coordinated joint movements in individual legs (intra-leg coordination) is believed to depend largely on afferent feedback arising within that leg (Akay et al. 2004; Bässler and Büschges 1998; Bucher et al. 2003).
With respect to inter-leg coordination, behavioral studies in which the position of individual legs of lobsters walking on a treadmill was perturbed showed that leg position influences the coordination of the remaining legs (Cruse and Müller 1986; Cruse et al. 1983). Similar behavioral experiments with stick insects (Bässler 1979; Cruse 1979, 1985; Cruse and Schwarze 1988; Dean and Wendler 1983; Graham 1979a, b) led to the proposition of six rules for coordination during walking (Cruse 1990; Cruse et al. 1998). Three of these rules are of particular importance for ipsilateral leg coordination: 1) an ongoing swing phase inhibits the start of the swing phase in the next rostral leg, 2) the start of a stance phase facilitates the start of the swing phase in the next rostral leg, and 3) the posterior movement of a leg during its stance phase increasingly facilitates the start of the swing phase in the next caudal leg. Simulations using computer and electro-mechanical models showed that these rules are sufficient to produce patterned leg movements similar to those observed in walking insects (Cruse et al. 1995, 1996, 1998). Maintenance of these rules depends on interactions between segmental ganglia, because cutting the ipsilateral connectives disrupts inter-leg coordination (Dean 1989).
However, the nature of these pathways, in particular the relative importance of intersegmental afferent signals versus central CPG interconnections, is still not fully understood. Studies in stick insects described interneurons that receive afferent input, code position, and movement of single leg joints or complete legs and project to neighboring ganglia (Brunn and Dean 1994; Büschges 1989). In locusts, similar identified intersegmental interneurons form synaptic connections with leg MNs and nonspiking interneurons in the next posterior ganglion (Laurent 1986, 1987; Laurent and Burrows 1988, 1989a, b). However, in all these studies, the preparations were so reduced that the animals no longer walked, and thus, although these intersegmental interneurons are a potential pathway for intersegmental coordination of leg movements, it is unknown if they are in fact used to coordinate walking movement (Burrows 1996).
We describe here an attempt to address this difficulty using a walking single leg preparation of the stick insect in which the body of the animal was fixed and all legs except a single front leg were amputated. This leg performed walking movements on a treadmill while MN activity was recorded from the deafferented and deefferented ipsilateral mesothoracic ganglion (the ganglion caudal to the ganglion that controls the walking leg). Thus this preparation provides a link between the behavioral studies performed in intact preparations and the electrophysiological data gathered from preparations, so reduced walking was impossible. Furthermore, because sensory input was limited to a single leg, the complexity of neural responses observed is much less than in a fully legged animal, which greatly facilitates studying the neural mechanisms underlying intersegmental coordination. This preparation thus provides "a reliable semi-intact preparation in which the function and physiology of central oscillators and sensory processes can be observed and manipulated, while the essential motor patterns are expressed" (Friesen and Chang 2001).
When the front leg performed walking movements, the activity of mesothoracic MN pools increased and produced rhythmic activity coupled to the front leg steps. In contrast, when the pro- and mesothoracic ganglia were isolated from the animal, mesothoracic MN activity was not coordinated with rhythmic activity of prothoracic leg MNs. In preparations in which a frontal leg sensory organ [the femoral chordotonal organ (fCO)] was left intact, fCO mechanical stimulation influenced mesothoracic MN activity. Taken together, these data suggest that sensory input from the frontal leg may underlie some of the intersegmental leg coordination observed in walking intact animals.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
In experiments using a treadmill, all legs except the right front leg were amputated, and the animals were fixed dorsal side up on a foam platform using dental cement (Protemp II, ESPE). The thorax was opened to allow access for recording from mesothoracic leg nerves. The gut was moved aside, connective tissue was carefully removed to expose the mesothoracic ganglion, and the cavity was filled with saline (composition according to Weidler and Diecke 1969). Recordings were made from right side leg nerves nl2 (protractor coxae), nl5 (retractor coxae), C1 (levator trochanteris), C2 (depressor trochanteris), nl3 (extensor tibiae; nomenclature according to Graham 1985 and Marquardt 1940) with monopolar hook electrodes (modified after Schmitz et al. 1991). All mesothoracic nerves were crushed or cut to exclude sensory input in the mesothoracic segment. In three experiments, activity of mesothoracic flexor tibiae MNs was recorded as EMG activity. For these recordings, the main leg nerve nCr (nervus cruris) was left intact, and two thin copper wires were inserted into the proximal femur. To exclude local sensory signals as much as possible, the leg was firmly glued to the platform, and the receptor apodeme of the fCO was cut because its afferents share nCr with flexor efferents. An EMG from the prothoracic flexor muscle was also obtained by inserting two thin copper wires into the proximal femur. A lightweight low-friction treadmill (Gabriel et al. 2003) was positioned under the animal's right front leg to allow the unrestricted leg to perform walking like movements. A DC motor attached to the treadmill served as a tachometer for treadmill velocity. Data were analyzed with respect to the start of the front leg stance phase, the latter being defined as the time during which the treadmill was accelerated. To describe a modulation of spiking rate of mesothoracic MNs, we examined an interval of ±0.5 s around the time of the start of front leg stance. This interval was divided in bins of 33 ms, and spikes were summed in each bin for all analyzed steps of an animal. We determined the two bins representing the highest or lowest spike count prior to and during the front leg stance phase. The percentage given in RESULTS describes the difference in spike count with respect to the bin representing the higher spike count, thereby indicating the average depth of modulation.
In experiments investigating the role of CPG interaction alone, a pair of ganglia (pro- and mesothoracic or meso- and metathoracic) was removed from the animal. The ganglia were placed in a Sylgard-lined dish with two separate compartments to allow independent superfusion of each ganglion. The interganglionic connective was carefully placed in a slit connecting the compartments and sealed with petroleum jelly. In some experiments, the separation of the two compartments was verified by applying a small amount of the dye Janusgreen B (Eastman Chemical, Rochester, NY) next to the petroleum jelly barrier. While one ganglion was superfused with normal saline, the other ganglion was superfused with saline containing 5 x 10–4 M pilocarpine (Sigma, Taufkirchen, Germany). Pilocarpine is a muscarinic agonist that activates arthropod CPGs and evokes rhythmic alternating activity in antagonistic MN pools (Büschges et al. 1995). In the isolated pro- and mesothoracic ganglia, extracellular recordings were made from leg nerves nl2 (protractor coxae) and nl5 (retractor coxae) of both ganglia using monopolar hook electrodes. In the isolated meso- and metathoracic ganglia, recordings were made of nerves C1 (levator trochanteris), C2 (depressor trochanteris), nl2 (protractor coxae), and nl3 (extensor tibiae).
In experiments where the fCO was stimulated, the front leg walking preparation was used, but the right front leg was cut 3 mm distal of the femur-tibia joint. The fCO receptor tendon was clamped in an electromechanical stimulator (Hofmann et al. 1985). Mechanical stimulation to the fCO simulated femur-tibia joint flexion between 40 and 120°. The abdomen of the animal was touched with a paintbrush that elicited short bursts of MN activity, often including alternating bursts from antagonistic MN pools. fCO stimulation was triggered to occur during these bursts. Nerve recordings were made from the stumps of mesothoracic nerves nl2, nl5 (protractor and retractor coxae), C1, C2 (levator and depressor trochanteris), and nl3 (extensor tibiae). Activity of flexor tibiae MNs was recorded from fine branches of nerve nCr, which innervate the flexor tibiae muscle. In addition to recordings in the mesothoracic segment, extensor tibiae MN activity was recorded in the prothoracic segment from nerve F2. This allowed monitoring local reflex behavior during fCO stimulation (Bässler 1988). For comparison, nerve activity in an interval (1 s duration) prior to fCO stimulation was integrated. Activity was also integrated in an equally long second interval after the stimulation and divided by the first number. This provided a value describing spiking activity after fCO stimulation normalized to spiking activity. The same method was applied to data from control experiments with disconnected stimulation device. Percent values given in RESULTS refer to the difference between experimental and control values.
All data were recorded using a MICRO 1401 A/D converter and SPIKE 2 data acquisition/analysis software (versions 3.13–4.12, Cambridge Electronic Design, Cambridge, UK). Data evaluation was done using custom written scripts within the SPIKE 2 software and plotted with Grapher 4.0 (Golden Software, Golden, CO). In the text and figures, N is the number of animals and n is the sample size.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
The animals walked with the right front leg on a treadmill while activity of mesothoracic MNs was recorded extracellularly. Tactile activation by touching the abdomen with a paintbrush typically elicited sequences of 8–15 consecutive steps lasting for 5–20 s. Spontaneous locomotor sequences also occasionally occurred. EMG recordings from the front leg flexor muscle monitored the stepping activity (Fig. 1). Shortly after the onset of flexor activity, the treadmill accelerated, indicated by a rising tachometer signal (1st dashed line). The onset of this rise was defined as the start of the front leg stance. Shortly before the end of the flexor burst, the leg lifted off the treadmill, and treadmill velocity decreased. The maximum treadmill velocity was defined as end of the stance phase (2nd dashed line). It is important to note that the decline in treadmill velocity after this peak does not give any information about leg movements, because during this time, the leg is lifted off the treadmill. The decline in treadmill velocity thus stems solely from the treadmill's mechanical properties. In the resting animal, mesothoracic slow MNs (small unit marked with asterisk) normally were spontaneously tonically active. When front leg walking began, activity of slow mesothoracic MNs increased and fast mesothoracic MNs (large unit) began to fire. In addition to this overall increase of firing rate, in most animals, MN activity was specifically modulated for each MN pool and correlated with front leg steps.
|
Protractor coxae MNs control the forward movement of the leg and are active mainly during swing phase, while retractor coxae MNs control the backward movement and are active mainly during stance phase. During walking activity of the front leg, mesothoracic protractor MN activity was coordinately modulated in 9 of 10 animals (Fig. 2A). To quantify this modulation, the start of front leg stance was taken as a reference, and the number of protractor MN spikes was plotted in a time window ±0.5 s around this time. This was done for all steps of a sequence, and the plots were aligned so that each row represents one step cycle (Fig. 2B). This raster plot shows some variability in the firing pattern during different steps, but during all 116 steps shown in the plot, mesothoracic protractor activity is reduced during front leg stance (average duration 0.29 s; indicated in Fig. 2, B and C, by vertical lines). A summary of mesothoracic protractor MN activity during front leg stance is shown in the histogram (Fig. 2C). The bin representing lowest spike activity during front leg stance contains 
65% fewer spikes than the one representing highest spike activity prior to the stance phase (arrows in Fig. 2C mark the bins compared). Eight other animals showed a similar modulation of protractor MN activity during front leg walking, with depth of modulation in the range of 15–90%.
|
25% higher during front leg stance, protractor activity is 45% lower. A similar modulation of mesothoracic retractor MN activity during front leg stance was observed in all four animals tested. Depressor and levator trochanteris motorneurons
Levator trochanteris MNs control the up movement of a leg and are active during the swing phase of a leg. The antagonistic depressor trochanteris MNs control the down movement during stance phase. We observed a front leg step correlated modulation of mesothoracic depressor MN activity in six of nine animals tested. In these six animals, depressor MN activity decreased in two animals (Fig. 3, A and C) and increased in four (Fig. 3D) during front leg stance. For a given animal, the modulation of depressor MNs with respect to front leg stance was similar during all steps, although the magnitude of the modulation varied from step to step (see raster plots). For instance, in Fig. 3C2, in steps 1–10, fast depressor MN activity was only weakly modulated, whereas in steps 20–60, it was almost completely abolished during front leg stance. The average depth of modulation ranged from 50 to 78% for slow depressor MNs and from 44 to 100% for fast depressor MNs.
|
Extensor and flexor tibiae MNs
The slow (SETi) and fast (FETi) extensor tibiae MNs and antagonistic flexor tibiae MNs control the movement of the femur-tibia ("knee") joint of the leg. Extensor MNs extend the tibia mainly during the swing phase of the leg, whereas flexor MNs flex it during the stance phase. In all four animals studied, FETi MNs showed a front leg step correlated modulation of activity. In three of four animals studied, SETi MNs showed such a modulation. Extensor MNs spike rate was lowest before front leg stance. It rapidly increased with the start of stance and reached its maximum at the end of the stance phase. SETi activity was modulated 34–50%, whereas FETi activity was modulated 64–100%.
It was difficult in this preparation to reliably identify the antagonistic flexor tibiae MNs in an extracellular recording because their axons share a nerve with those of tarsal MNs. We therefore used EMG recordings to study their activity during front leg walking. It was hence necessary to leave nervus cruris, which also carries afferents, intact. Through careful fixation of the middle leg stump and cutting of the fCO receptor strand, we minimized any local sensory input. In all three of these experiments, mesothoracic flexor MNs activity was modulated in phase with front leg stepping (Fig. 4C1). Mesothoracic flexor MN modulation was antagonistic to mesothoracic extensor MN modulation, with flexor MN spike rate being greatest before front leg stance and decreasing 90–100% during stance (Fig. 4C2).
|
Prothoracic CPG activity and mesothoracic motorneuron activity
The information used for intersegmental coordination of movements could originate from prothoracic CPGs as well as from sense organs associated with the front leg. To examine the influence of prothoracic CPG activity on mesothoracic MN output, we isolated the pro- and mesothoracic ganglia and recorded protractor and retractor coxae activity from nerve stumps. The prothoracic ganglion was superfused with saline containing 5 x 10–4 M pilocarpine to activate locomotor CPGs and produce alternating rhythmic motor activity (Büschges et al. 1995). Unlike in previous preparations of the stick insect, which revealed no strong intersegmental coordination (Büschges et al. 1995; Ryckebusch and Laurent 1994), the adjacent mesothoracic ganglion was isolated by a petroleum jelly barrier and not exposed to pilocarpine. Slow units of both pro- and mesothoracic protractor and retractor MNs were usually spontaneously active at a constant spike rate. Shortly after pilocarpine superfusion, spike rates increased in the pharmacological activated prothoracic ganglion and, to a smaller extent, in the adjacent mesothoracic ganglion. In the prothoracic ganglion, the tonic activity subsequently became rhythmic, and in 19% of the experiments, a stable, alternating rhythm was established. In these preparations mesothoracic protractor and retractor MNs continued in all cases to fire tonically (Fig. 5; protractor: N = 4, retractor: N = 3). To detect subtle spike rate modulation, the starts of prothoracic retractor bursts were taken as a reference point, and mesothoracic MN activity was plotted as histogram in a ±1 s interval around these times. No correlation between pro- and mesothoracic MN activity was found (Fig. 5, B and C).
|
In summary, we found no evidence for influences of an activated CPG on MN activity in adjacent ganglia in isolated preparations.
Influence of front leg femoral chordotonal organ stimulation on mesothoracic motorneuron activity
Because CPG activity in an adjacent ganglion did not seem to be sufficient to provide signals for intersegmental coordination of locomotor activity, we studied intersegmental influences of a sensory organ, the fCO. The fCO senses the position and movement of the femur-tibia joint and could thereby provide signals suitable for leg coordination during walking. Its physiology and its influence on networks in the segmental ganglion are well studied (e.g., Bässler and Büschges 1998; Hofmann et al. 1985). Information about the femur-tibia joint provided by the fCO is used for the coordination with other joints of the leg. In an "active" animal, fCO signals can evoke transitions between levator and depressor trochanteris MN pools (Hess and Büschges 1999). Bucher et al. (2003) presented evidence that position signals from the fCO have access to the CPGs controlling MNs of the coxa-trochanter joint.
We used a preparation similar to the front leg walking preparation, but the right front leg was fixed and cut distal to the femur-tibia joint. The other five legs were amputated to exclude uncontrolled sensory input. Tactile stimulation of the animal often evoked alternating bursts of antagonistic MN pools in the mesothoracic ganglion (Fig. 6, A1 and B1). The front leg's fCO was stimulated mechanically during such bursts to simulate femur-tibia joint flexion. When the stimulation occurred during a burst of mesothoracic protractor coxae MNs, it shortened the protractor burst and increased retractor coxae MN activity (Fig. 6A1). As the effect of the stimulus was often masked by the naturally varying protractor burst lengths, we compared the pooled data with data from a control situation in which the stimulus device was unplugged. Prothoracic fCO stimulation during a mesothoracic protractor MN burst reduced protractor MN activity by 35% and increased retractor MN activity by 61% (Fig. 6A2; protractor: N = 11/control: 9, n = 161/control: 148; retractor: N = 9/5, n = 112/59). A similar effect was observed when the stimulus was applied during extensor tibiae MN bursts (Fig. 6B1). Although extensor activity did not change significantly, activity of flexor tibiae MNs increased by 94% (Fig. 6B2; flexor: N = 8/4 n = 88/76; extensor: N = 7/6, n = 146/101, extensor data not shown). No influence of prothoracic fCO stimulation was found on mesothoracic depressor and levator trochanteris MNs (depressor: N = 7/5, n = 143/81; levator: N = 11/3, n = 200/59, data not shown).
|
; Bässler and Büschges 1998). The transmission of intersegmental information could thus depend on the state of the animal, reflected by the occurrence or absence of a resistance reflex. We therefore analyzed the data separately with respect to the occurrence (or absence) of a local resistance reflex in the prothoracic segment. No correlation between the occurrence (or absence) of a prothoracic resistance reflex and the strength of effects on mesothoracic MNs was detected. In summary, signals from the prothoracic fCO, indicating a flexion of the front leg, promote mesothoracic retractor coxae and flexor tibiae activity, while they repress protractor coxae activity.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
), were not observed. Since the mesothoracic ganglion was deafferented, these spike rate modulations reflect influences of intersegmental signals from the prothoracic ganglion on the mesothoracic ganglion. These data thus strongly suggest that these interganglionic signals help coordinate leg movements during walking.
|
Although the modulation seen in each mesothoracic MN pool was similar in all animals studied, step-to-step variability was observed in individual animals. For example, mesothoracic protractor MNs showed reduced activity during front leg stance in nine animals. The variation seen between steps included variations in underlying firing rate (e.g., Fig. 2B, steps 1–40 compared with 40–70) and in modulation depth (e.g., Fig. 3C2, steps 1–10 compared with 20–60). This variation was still observed with other reference points in the walking cycle (e.g., stance end, peak front leg flexor activity) and thus is not due to using the front leg stance as reference to analyze mesothoracic MN activity. Similarly, this step to step variation also remained when the data were re-analyzed with respect to phase within in the step cycle (i.e., dividing all burst durations and delays by step cycle period). The source of this variation is unknown but is comparable to that observed in animals walking in similarly regular environments, e.g., on treadmills (e.g., Fischer et al. 2001; Foth and Bässler 1985; Gabriel et al. 2003; Graham 1985; Graham and Wendler 1981). Interestingly, in Aplysia feeding under conditions in which environmental variation was also deliberately reduced as much as possible, wide variation in bite parameters was again observed (Horn et al. 2004). These authors ascribed this variation to an inherent variability in CPG activity and suggested that such inherent variability is functionally advantageous in dealing with a varying environment (in this case, different foodstuffs). Whether the variations observed in stick insect walking arises, at least in part, from an inherent variability in the system is unknown, but natural substrates do show wide variation, and it is thus possible that the variation we have observed in our preparations arises from inherent system properties. Regardless, despite this variability in individual step activity, the modulation in mesothoracic MN activity induced by front leg stepping was nonetheless present in all steps.
Comparison with the step pattern of an intact animal
Stick insect walking (Graham 1985) and searching (Dürr 2001) movements are generally described with respect to leg movement; relatively little data are available on MN activity. This distinction is important because of a significant difference between the timing of MN activity and leg movement (Watson and Ritzmann 1998). For example, the delay between the onset of flexor EMG activity and treadmill movement in our experiments was typically 100 ± 50 ms—similar to the delay of 126 ± 151 ms reported by Gabriel et al. (2003). Unfortunately, the magnitude of delays is not known for all MN types, and it is therefore difficult to predict MN activity from recordings of whole leg movement. Prior work that examined MN activity directly was performed in preparations in which protraction and retraction were prevented (Fischer et al. 2001; Schmidt et al. 2001), and thus these data also cannot be compared with those presented here. Furthermore, MN coordination seems to be strongly influenced by treadmill mechanical properties (friction, inertia) (Gabriel et al. 2003), and comparison of data using different treadmills is thus also problematic.
As a result of these difficulties and of the variety of step patterns that result in a tetrapod gait, we therefore chose to instead compare our data to an "idealized" tripod gate in which protractor, levator, and extensor MNs are active throughout the stance phase and a 180° phase shift exists between neighboring legs (Fig. 7B). A similar step pattern was observed under conditions of reduced proprioceptive feedback such as walking on a mercury surface ("gait I"; Graham and Cruse 1981). Nevertheless, it should be kept in mind that the tripod gait is only one of several step patterns shown by stick insects (Graham 1972, 1985; Wendler 1964). For an idealized tripod gait, it would be predicted that middle leg protractor, levator, and extensor MN activity should be greatest, and retractor, depressor, and flexor activity least, during front leg stance. These predictions are borne out for our extensor/flexor and levator data (Fig. 7B). With respect to the protractor/retractor MNs (although they do correctly fire in antagonism with each other), relative to front leg stance, their activity is almost in anti-phase to that predicted and would thus produce not forward, but backward, movement of the middle leg during its swing phase. With respect to the depressor MNs, in two of nine animals, the observed modulation was consistent with that predicted, but in the other four animals that showed modulation, the activity was again significantly out of phase with that predicted and would incorrectly depress the leg during its swing phase. Clearly, these protractor/retractor and (in the 4 animals) depressor MN activities are inconsistent with functional middle leg movement. This is true not only for a tripod gait, but also if a tetrapod gait would be chosen for comparison.
An explanation for the disparity is that, in the intact animal, local proprioceptive input as well as afferent signals from the contralateral front and the middle and hind legs results in appropriate phase shifts of protractor/retractor and depressor MN activities. The MN pools controlling the most proximal leg joint, protractor coxae, and retractor coxae MNs seem to be in-phase with those in the prothoracic segment. Such in-phase coordination has been described for front legs of decapod crustaceans, while the body weight is supported by only the back legs and the front legs are lifted above ground (Duysens et al. 2000). Similarly, leg stumps of rock lobsters show in-phase coordination after leg autotomy (Clarac 1981). Clarac (1981) suggested that peripheral proprioceptive inputs could shift this in-phase coordination toward the alternating pattern observed in the intact walking animal.
Amputation experiments with cockroaches led to the idea that "sensory input helps keep the legs synchronized with one another" (Delcomyn 1991a). In support of this hypothesis, local sensory signals [e.g., from the fCO (reviewed in Bässler and Büschges 1998) and the campaniform sensilla (Akay et al. 2004; Duysens et al. 2000; Ridgel et al. 2000)] are known to have an important function in the local control of leg movements. Further evidence of the importance of local sensory information in leg coordination comes from experiments by Graham (1985), in which he moved the coxal stump of an amputated middle leg while a stick insect walked on a double tread wheel. With the stump held still, retractor MNs of one leg showed weak rhythmic activity in a fixed phase relation to the activity of retractor MNs controlling other legs. When proximal sensory organs were stimulated by moving the stump, retractor MN activity was observed only if the movement occurred in a certain phase window with respect to the movements of the legs in front and behind it. This shows that local sensory input, which is not present in our single leg walking preparation, modifies motor output and contributes to the formation of the stepping pattern seen in a freely walking animal.
The experimental conditions (reduced sensory input) could induce searching in the mesothoracic segment. Searching movements are for example observed if the middle leg of a walking stick insect steps into a gap and thereby looses mechanical feedback (Dürr 2001). Searching movements are stereotypic, loop-like leg movements with simultaneous protraction, depression, and flexion (Dürr 2001). We observed roughly simultaneous activity of mesothoracic protractor, depressor (2 of 6 experiments), and flexor MNs (Fig. 7A). Although it is not possible to predict accurate leg movement from this MN activity, we do not want to exclude the possibility that the observed MN activity, in part, reflects searching movements.
Relevance of central coordinating mechanisms
MN activity that, in the absence of sensory input, is nonetheless relatively similar to that observed in intact animals (fictive motor patterns) can be induced by appropriate treatment in several multi-CPG systems (crayfish swimmeret system by application of the neuropeptide proctolin (Ikeda and Wiersma 1964; Mulloney et al. 1987), leech swimming by brief electrical stimulation of lateral nerve roots (Kristan and Calabrese 1976), and swimming in lamprey and tadpole by application of N-methyl-D-aspartate (NMDA) (Grillner and Wallén 2002; Skinner and Mulloney 1998). These data indicate that, although sensory signals play a role in modifying system output as necessary to maintain function, in several systems, the basic coordination among the different CPGs is to a large extent maintained by central mechanisms (Friesen and Chang 2001).
With respect specifically to walking, muscarinic agonists (pilocarpine or oxotremorine) induce "fictive locomotion" with slow cycle periods but basically correct leg coordination in crayfish (Chrachri and Clarac 1990). However, in stick insects, although pilocarpine application does induce alternating rhythmic activity of antagonistic motor pools, the activity of the MN pools controlling different leg joints is not correctly coordinated (Büschges 1995). In the work presented here, when CPG activity was pharmacologically induced in one ganglion of an isolated pair, no evidence that this activity influenced spontaneous activity of leg MNs of the neighboring ganglion was found. Similarly, previous work on ganglion chains in stick insects and locusts also found no evidence for a strong inter-CPG coupling when rhythmic activity was induced in the complete chain of ganglia, although a weak intersegmental coordinating influence could have been masked by strong local activity in these preparations (Büschges et al. 1995; Ryckebusch and Laurent 1994). It is, of course, possible that this lack of inter-joint and inter-leg coordination is due to incomplete activation of the neural networks underlying walking, and that in real walking, as observed in the swimmeret and leech, lamprey, and tadpole CPG systems, central connections mediate intersegmental coordination (Hill et al. 2003). Nonetheless, our data and the prior work mentioned above are also consistent with the alternative hypothesis that direct coupling of segmental CPGs is not involved in intersegmental coordination in stick insect walking.
It is important to note, however, that this lack of apparent central coordinating mediation is not present in all insects, because Johnston and Levine (2002) found coordinated, tripod-like activity of leg MNs in all three thoracic ganglia in pilocarpine-activated, isolated nerve cords in Manduca sexta. Although normal walking in the animals displays a more complex gait (Johnston and Levine 1996), these data nonetheless suggest that, in Manduca, central coupling at least partially underlies intersegmental coordination. This difference between Manduca and stick insects may be related to the different primary forms of locomotion the two animals use—Manduca mostly flies, whereas stick insects walk and climb. A more flexible control of walking, with a stronger influence of afferent feedback to better match walking movements with substrate variation, might therefore be advantageous for stick insects.
Relevance of sensory signals for intersegmental coordination
Early evidence for a role of sensory information in stick insect leg coordination came from experiments in which middle leg autotomy changed phase relations of the remaining legs (Wendler 1964). We show here an intersegmental effect of fCO afferents. This stimulus mimics flexion, which would occur at the start of the front leg stance phase during walking. The stimulus decreased mesothoracic protractor MN activity and increased mesothoracic retractor MN activity, which parallel the observations from the single leg walking preparation. Intersegmental fCO signals could therefore be at least partially responsible for the observed modulation of protractor and retractor MN activity. In contrast, prothroracic fCO stimulation caused an increase of flexor MN activity. This observation is contradictory to the decrease of flexor MN activity during front leg stance observed in the single leg walking preparation. Similarly, no evidence for intersegmental fCO modulation of mesothoracic extensor, levator, and depressor MN activity was found, although their activity was modified in the single leg preparation. These data thus suggest that, although fCO activity plays a role in intersegmental coordination, the combined activity of multiple front leg afferent pathways [e.g., signals from the fCO and campaniform sensilla (Akay et al. 2001) or hair fields (Wendler 1964)] are required for correct inter-leg coordination.
fCO stimulation can locally (i.e., in the same segment) induce either an assistance or a resistance reflex, with the assistance reflex being commonly associated with the system being in an "active" state (Bässler 1988). It was therefore possible that the effect of fCO stimulation on mesothoracic MNs would vary depending on whether the system was in this active state or not. We tested this by simultaneously observing the effect of fCO stimulation on prothoracic and mesothoracic MN activity, and found no correlation between the mesothoracic effects and whether an assistance or resistance reflex was induced in the prothoracic ganglion. These data thus suggest that the changes in local networks that underlie the "active" state do not gate or otherwise alter transmission in fCO intersegmental pathways.
In summary, we show that, in a preparation with a single walking front leg, intersegmental signals modulate MN activity in the adjacent deafferented segment. Experiments with completely isolated nervous systems indicate that these effects are unlikely to be due to central connections. The prothoracic fCO was shown to be an intersegmental input that could contribute to the observed MN modulation. These data suggest that inter-leg coordination in the stick insect is primarily due to intersegmental pathways whose activity is altered by afferent input.
|
GRANTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Present addresses: B. Ch. Ludwar, Department of Biological Sciences, Ohio University, Irvine Hall, Athens, OH 45701; M. L. Göritz, Neurobiology and Behavior, W159 Seeley G. Mudd Hall, Cornell University, Ithaca, NY 14853.
|
FOOTNOTES |
|---|
Address for reprint requests and other correspondence: B. Ch. Ludwar, Dept. of Biological Sciences, Ohio Univ., Irvine Hall, Athens, OH 45701 (E-mail: ludwar{at}ohio.edu)
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Akay T, Haehn S, Schmitz J, and Büschges A. Signals from load sensors underlie interjoint coordination during stepping movements of the stick insect. J Neurophysiol 92: 42–51, 2004.
Bässler U. Interactions of central and peripheral mechanisms during walking in first instar stick insects. Physiol Entomol 4: 193–199, 1979.
Bässler U. Functional principles of pattern generation for walking movements of stick insect forelegs: the role of the femoral chordotonalorgan afferences. J Exp Biol 136: 125–147, 1988.
Bässler U and Büschges A. Pattern generation for insect walking movements—multisensory control of a locomotor program. Brain Res Rev 27: 65–88, 1998.[CrossRef][Medline]
Brunn D and Dean J. Intersegmental and local interneurons in the metathorax of the stick insect Carausius morosus that monitor middle leg position. J Neurophysiol 72: 1208–1219, 1994.
Bucher D, Akay A, DiCaprio RA, and Büschges A. Interjoint coordination in the stick insect leg-control system: the role of positional signaling. J Neurophysiol 89: 1245–1255, 2003.
Burrows M. The Neurobiology of an Insect Brain. Oxford: Oxford University Press, 1996.
Büschges A. Processing of sensory input from the femoral chordotonal organ by spiking interneurons of stick insects. J. Exp Biol 144: 81–111, 1989.
Büschges A, Schmitz J, and Bässler U. Rhythmic patterns in the thoracic nerve cord of the stick insect induced by pilocarpine. J Exp Biol 198: 435–456, 1995.[Abstract]
Chrachri A and Clarac F. Fictive locomotion in the fourth thoracic ganglion of the crayfish, Procambarus clarii. J Neurosci 10: 707–719, 1990.
Clarac F. Decapod crustacean leg coordination during walking. In: Locomotion and Energetics in Arthropods, edited by Herreid CF II and Fourtner CR. New York: Plenum Press, 1981, p. 63ff.
Cruse H. The control of the anterior extreme position of the hindleg of a walking insect, Carausius morosus. Physiol Entomol 4: 121–124, 1979.
Cruse H. Coactivating influences between neighbouring legs in walking insects. J Exp Biol 114: 513–519, 1985.
Cruse H. What mechanisms coordinate leg movement in walking arthropods? Trends Neurosci 13: 15–21, 1990.[CrossRef][ISI][Medline]
Cruse H, Clarac F, and Chasserat C. The control of walking movements in the leg of the rock lobster. Biol Cybern 47: 87–94, 1983.
Cruse H and Müller U. Two coupling mechanisms which determine the coordination of ipsilateral legs in the walking crayfish. J Exp Biol 121: 349–369, 1986.
Cruse H and Schwarze W. Mechanisms of coupling between the ipsilateral legs of walking insect (Carausius morosus). J Exp Biol 138: 455–469, 1988.
Cruse H, Brunn D, Bartling Ch, Dean J, Dreifert M, Kindermann T, and Schmitz J. Walking: A complex behavior controlled by simple networks. Adapt Behav 3: 387–417, 1995.
Cruse H, Bartling Ch, Dean J, Kindermann T, Schmitz J, Schumm M, and Wagner H. Simplifying neural networks for controlling walking by exploiting physical properties. In: Artificial Neural Networks—ICANN 96, edited by von der Malsburg C, von Seelen W, Vorbrüggen JC, and Sendhoff B. New York: Springer, 1996, p. 433–438.
Cruse H, Kindermann T, Schumm M, Dean J, and Schmitz J. Walknet—a biological inspired network to control six-legged walking. Neural Netw 11: 1435–1447, 1998.[CrossRef][ISI][Medline]
Dean J. Leg coordination in the stick insect Carausius morosus: effects of cutting thoracic connectives. J Exp Biol 145: 103–131, 1989.
Dean J and Wendler G. Stick insect locomotion on a walking wheel: Interleg coordination of leg position. J Exp Biol 103: 75–94, 1983.
Delcomyn F. Pertubation of the motor system in freely walking cockroaches. I. Rear leg amputation and the timing of motor activity in leg muscles. J Exp Biol 156: 483–502, 1991a.
Delcomyn F. Pertubation of the motor system in freely walking cockroaches. II. The timing of motor activity in leg muscles after amputation of a middle leg. J Exp Biol 156: 503–517, 1991b.
Dürr V. Stereotypic leg searching movements in the stick insect: Kinematic analysis, behavioural context and simulation. J Exp Biol 204: 1589–1604, 2001.[Abstract]
Duysens J, Clarac F, and Cruse H. Load-regulating mechanisms in gait and posture: comparative aspects. Physiol Rev 80: 83–133, 2000.
Fischer H, Schmidt J, Haas R, and Büschges A. Pattern generation for walking and searching movements of a stick insect leg. I. Coordination of motor activity. J Neurophysiol 85: 341–353, 2001.
Foth E and Bässler U. Leg movements of stick insect walking wirh five legs on a treadmill and with one leg on a motor-driven belt. I General results and 1:1 coordination. Biol Cybern 51: 3131–318, 1985.
Friesen WO and Chang J. Sensory and central mechanisms control intersegmental coordination. Curr Opin Neurobiol 11: 678–683, 2001.[CrossRef][ISI][Medline]
Gabriel JP, Scharstein H, Schmidt J, and Büschges A. Control of flexor motorneuron activity during single leg walking of the stick insect on an electronically controlled treadwheel. J Neurobiol 56: 237–251, 2003.[CrossRef][ISI][Medline]
Graham D. A behavioural analysis of the temporal organisation of walking movements in the 1st instar and adult stick insect (Carausius morosus). J Comp Physiol 81: 23–52, 1972.[CrossRef]
Graham D. Effects of circum-oesophageal lesion in the behaviour of the stick insect Carausius morosus I. Cyclic behaviour patterns. Biol Cybern 32: 139–145, 1979a.[CrossRef][ISI]
Graham D. Effects of circum-oesophageal lesion in the behaviour of the stick insect Carausius morosus II.Changes in walking co-ordination. Biol Cybern 147–152, 1979b.
Graham D. Pattern and control of walking in insects. Adv Insect Physiol 18: 31–140, 1985.
Graham D and Cruse H. Coordinated walking of stick insects on a mercury surface. J Exp Biol 92: 229–241, 1981.
Graham D and Wendler G. Motor output to the protractor and retractor coxae muscles in stick insects walking on a treadwheel. Physiol Entomol 6: 161–174, 1981.
Grillner S and Wallén P. Cellular bases of a vertebrate locomotor system-steering, intersegmental and segmental co-ordination and sensory control. Brain Res Rev 40: 92–106, 2002.[CrossRef][Medline]
Hess D and Büschges A. Role of proprioceptive signals from an insect femur-tibia joint in patterning motorneuronal activity of an adjacent leg joint. J Neurophysiol 81: 1856–1865, 1999.
Hill AAV, Masino MA, and Calabrese RL. Intersegmental coordination of rhythmic motor patterns. J Neurophysiol 90: 531–538, 2003.
Hofmann T, Koch UT, and Bässler U. Physiology of the femoral chordotonal organ in the stick insect, Cuniculina impigra. J Exp Biol 114: 207–223, 1985.
Hooper SL and DiCaprio RA. Crustacean motor pattern generator networks. Neurosignals 13: 50–69, 2004.[CrossRef][ISI][Medline]
Horn CC, Zhurov Y, Orekhova IV, Proekt A, Kupfermann I, Weiss KR, and Brezina V. Cycle-to-cycle variability of neuromuscular activity in Aplysia feeding behavior. J Neurophysiol 92: 157–180, 2004.
Ikeda K and Wiersma CAG. Autogenic rhythmicity in the abdominal ganglia of the crayfish: the control of swimmeret movements. Comp Biochem Physiol 12: 107–115, 1964.[Medline]
Johnston RM and Levine RB. Locomotory behavior in the hawkmoth Manduca sexta: kinematic and elektromyographic analyses of the thoracic legs in larvae and adults. J Exp Biol 199: 759–774, 1996.[Abstract]
Johnston RM and Levine RB. Thoracic leg motorneurons in the isolated CNS of adult Manduca produce patterned activity in response to pilocarpine, which is distinct from that produced in larvae. Invert Neurosci 4: 175–192, 2002.[CrossRef][Medline]
Kristan WB and Calabrese RL. Rhytmic swimming activity in neurones of the isolated nerve cord of the leech. J Exp Biol 65: 643–668, 1976.
Laurent G. Thoracic intersegmental interneurones in the locust with mechanoreceptive inputs from a leg. J Comp Physiol A 159: 171–186, 1986.
Laurent G. The morphology of a population of thoracic intersegmental interneurones in the locust. J Comp Neurol 256: 412–429, 1987.[CrossRef][ISI][Medline]
Laurent G and Burrows M. A population of ascending intersegmental interneurones in the locust with mechanosensory inputs from a hind leg. J Comp Neurol 275: 1–12, 1988.[CrossRef][ISI][Medline]
Laurent G and Burrows M. Distribution of intersegmental inputs to nonspiking local interneurons and motor neurons in the locust. J Neurosci 9: 3019–3029, 1989a.[Abstract]
Laurent G and Burrows M. Intersegmental interneurons can control the gain of reflexes in adjacent segments of the locust by action on nonspiking local interneurons. J Neurosci 9: 3030–3039, 1989b.[Abstract]
MacKay-Lyons M. Central pattern generation of locomotion: a review of the evidence. Phys Ther 82: 69–83, 2002.
Marder E and Calabrese RL. Principles of rhythmic motor pattern generation. Physiol Rev 76: 687–717, 1996.
Marquardt F. Beiträge zur Anatomie der Muskulatur und der peripheren Nerven von Carausius (Dixippus) morosus. Zool Jahrb Abt Anat Ont Tiere 66: 63–128, 1940.
Mulloney B, Acevedo LD, and Bradbury AG. Modulation of the crayfish swimmeret rhythm by octopamine and the neuropeptide proctolin. J Neurophysiol 5
–0.2 s in C but not in E) is due to the different step cycle period in the 2 animals (0.4 s for the animal in A and C and 1.4 s for the animal in D and E). Thus in the 1st animal, the decrease in protractor activity that occurs between stance phases is captured within the analysis time window, whereas in the 2nd animal, this decrease in activity occurs at times more negative than –0.5 s.
