| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
The Journal of Neurophysiology Vol. 78 No. 6 December 1997, pp. 3133-3143
Copyright ©1997 by the American Physiological Society
Laboratoire de Neurobiologie et Mouvements, Centre National de la Recherche Scientifique, 13402 Marseille Cedex 20, France
 |
ABSTRACT |
|---|
 Abstract
 Introduction
Methods
Results
Discussion
References
|
|---|
Le Ray, Didier, François Clarac, and Daniel Cattaert. Functional analysis of the sensory motor pathway of resistance reflex in crayfish. I. Multisensory coding and motor neuron monosynaptic responses. J. Neurophysiol. 78: 3133-3143, 1997. An in vitro preparation of the fifth thoracic ganglion of the crayfish was used to study in detail the negative feedback loop involved in the control of passive movements of the leg. Release-sensitive primary afferents of from the coxo-basipodite chordotonal organ (CBCO), a proprioceptor whose strand is released by upward movement of the leg, monosynaptically connect to depressor motor neurons (Dep MNs). Extracellular identification of sensory units from the CBCO neurogram allowed us to determine the global coding of a sine-wave movement, imposed from the most released position of the CBCO strand. Intracellular recordings from sensory terminals (CBTs) and ramp movement stimulations applied to the CBCO strand allowed us to characterize two groups of release-sensitive CBCO fibers. The first group, divided into two subgroups (phasic and phaso-tonic), is characterized by discontinuous firing patterns: phasic CBTs fired exclusively during release movements; phaso-tonic CBTs displayed both a phasic firing and a tonic discharge during the more released plateaus. The second group was continuously firing whatever the movement, with higher frequencies during the release phase of the movement stimulation. All CBTs displayed a marked sensitivity for release movements while only the phaso-tonic ones showed a clear sensitivity to maintained positions. Surprisingly, no pure tonic sensory fibers were encountered. Systematic intracellular recordings from all resistant Dep MNs, performed in high divalent cation saline, allowed us to describe two shapes of monosynaptic resistance reflex responses. A phasic response was characterized by bursts of excitatory postsynaptic potentials (EPSPs) occurring exclusively during CBCO strand release movements. A phaso-tonic response was characterized by a progressive depolarization occurring all along the release phase of the stimulation: during maintained released positions, the amplitude of the sustained depolarization was position dependent; in addition, each release movement produced a phasic burst of EPSPs in the MN. The parallel study of the Dep MN properties failed to point out any correlation between the type of reflex response recorded from the MN and the MN intrinsic properties, which would indicate that the type of MN response is entirely determined by the afferent messages it receives.
The stretch reflex is a very widespread phenomenon described in all groups of vertebrates and invertebrates (Granit 1955 Preparation
Results are based on >130 intracellular recordings from CBTs and Dep MNs performed on adult male and female crayfish, Pacifastacus leniusculus and Procambarus clarkii. Animals were maintained in aquaria at 18°C and fed once a week.
Stimulations/recordings
Extracellular recordings and nerve stimulations were performed using platinum electrodes contacting the nerves, isolated from the bath with petroleum jelly (Vaseline), and directed to a 4-channel differential AC amplifier (A-M Systems). Single and paired intracellular recordings from CBTs and MNs (Fig. 1B) were made with thin-walled glass microelectrodes filled with a potassium chloride solution (3 M) and having a 25- to 30-M
Salines
A normal saline (in mM: 195 NaCl, 5 KCl, 13 CaCl2, and 2 MgCl2) was used to perform the sensory coding analysis. Because strong central synaptic inputs masked the monosynaptic sensory inputs from the CBCO, the Dep MN responses were analyzed with saline in which divalent cation concentrations were increased (in mM: 34 CaCl2 and 6.4 MgCl2, with the sodium concentration reduced accordingly) to raise the activation threshold of all central neurons; a Vaseline wall was used to restrict the perfusion of this high divalent cation saline exclusively on thoracic ganglia, in order not to affect the CBCO sensory inputs (Fig. 8). Saline solutions were buffered with 3 mM N-2-hydroxyethylpiperazine-N
Analysis
Signals were analyzed using the CED programs SPIKE2 and SIGAVG. The "Wave Marker" SPIKE2 program was used to classify the different profiles of extracellular spikes recorded from the CBCO nerve. Circular statistics (Batchelet 1981
As shown in Fig. 1, C and D, the mechanical stimulation of the CBCO strand is able to produce a reflex activity in the in vitro preparation. This is characterized by the discharge of the sensory units that can be recorded extracellularly in the CBCO neurogram and intracellularly in the CBTs (Fig. 1C). This sensory firing induces, in the postsynaptic Dep MN, a depolarization of its membrane potential (Fig. 1D) because of the summation of each compound excitatory postsynaptic potential (EPSP).
Analysis of the CBCO sensory coding of the leg movements
Two complementary studies have been performed to describe the coding of the CBCO sensory information. First, from the neurogram, an extracellular approach allowed us to characterize in a general manner the different CBCO units. Because of the limitations of this method, we carried out a second analysis, the intracellular characterization of the CBCO fibers. To facilitate our work, we focused the intracellular study on the release-sensitive afferents, impaled in their terminals.
EXTRACELLULAR STUDY FROM THE CBCO NEUROGRAM.
The extracellular study has been performed from a sinusoidal movement stimulation of the CBCO strand. In the angular range of the imposed sinusoidal movement, and depending on the preparation, about one-half of the CBCO units were activated as shown on Figs. 2C and 3. In our conditions, release-sensitive afferents were predominantly recruited (11/20 ± 2 release-sensitive and 7/20 ± 3 stretch-sensitive CBCO fibers, mean ± SE). The table of Fig. 2C gives the number of the sensory afferents activated by sinusoidal movements imposed to 10 CBCOs. A total number of 195 extracellular spikes has been studied to determine their sensitivity to CBCO strand movement; 55.5% of the recorded units were release-sensitive fibers (n = 108) against 28% of stretch-sensitive units (n = 55). In both cases, most of the CBCO units coded one direction of the imposed movement, although their discharge frequently persisted during a part of the opposite movement. In all experiments, some CBCO fibers (16.5%, n = 32) were classified in the "Nonspecific" group because their response was not clearly related to either phase of the imposed movement (R
PHASIC CBCO UNITS.
CBCO units that responded exclusively during the release movements of the CBCO strand (see example in Fig. 4A), and were silent during plateaus (see detail in inset), were classified as phasic units. The instantaneous firing frequency of this kind of sensory terminal varied from 5 ± 3 to 75 ± 8 Hz during release movements, and mean frequency varied from 10 ± 5 to 50 ± 10 Hz. As shown by the mean frequency histogram of Fig. 4A (bottom), there was no clear relationship between the mean firing frequency and the angular sector in which movement was applied. Moreover, the pattern of discharge was very variable from one cycle to the next. This explains the apparent discrepancy between the raw data (Fig. 4A, top) and the mean frequency (histogram in Fig. 4A). When low-velocity (0.25 mm/s) ramp stimulation was applied (Fig. 4B), the mean frequency of every phasic CBTs was dramatically reduced to about one-third of that obtained with high-velocity (1.25 mm/s) ramp stimulation.
PHASO-TONIC CBCO UNITS.
Phaso-tonic fibers (Fig. 5) were characterized by both a phasic response during release movements and a response to position. In general, the tonic response developed from a particular release position (different for each unit) and increased with the degree of release up to the maximum release position. As shown on Fig. 5, A and B, phaso-tonic CBCO units can be separated into two groups depending on the hysteresis of position coding: the first one (CBT 1, Fig. 5A) was characterized by spike emission for the most released positions, whatever the direction of the movement applied between maintained positions. However, for a given position, the discharge was larger after a release movement than after a stretch movement. The second group (CBT 2, Fig. 5B) stopped firing during the stretching phase; even for most release position, a stretch movement induced an inhibition of tonic firing that persisted during the two seconds of maintained position. Nevertheless, firing patterns of both types (CBT 1 and 2) were similar during the CBCO release (see detail in inset): the high-frequency phasic firing was followed by a lower frequency firing during the plateaus. Histograms of both mean frequencies and regressions showed that phasic (white histogram) and tonic (dark histogram) firings were increasing with the release of the CBCO strand. It was noticeable that in both cases the phaso-tonic units fired with a higher frequency during movements than during plateaus (up to 48 ± 7 Hz during ramps against up to 12 ± 5 Hz during plateaus). The velocity sensitivity of the phaso-tonic CBTs was exclusively limited to the phasic part of their response as shown by frequency histograms on Fig. 5C. As is the case for phasic CBCO units, when the velocity of the ramp was increased from 0.25 to 1.25 mm/s (i.e., multiplied by 5) the firing frequency of the movement-sensitive component of the phaso-tonic fiber was multiplied by 3 (white histogram). By contrast, the firing frequency of the tonic component appeared to be nonsignificantly affected by the change in velocity (dark histogram).
CONTINUOUS DISCHARGE CBCO UNITS.
Never described in detail in previous reports, this kind of fiber was characterized by continuous firing (at least 10 ± 4 Hz) that was modulated by movements imposed to the CBCO strand. Nevertheless, release-sensitive units showed a specificity for the release phase of the mechanical stimulation (where firing frequency could raise to a mean value of 65 ± 15 Hz during ramps). As shown on Fig. 7A, continuous discharge fibers fired with a higher frequency during the CBCO release but were still spiking during the stretch of the sensory organ. In this type of unit, movement sensitivity was also evident. The mean frequency histograms (Fig. 7A) show a marked difference between the movement response (white histogram) and the position response (dark histogram), as was the case for phaso-tonic units, during the CBCO release. The movement coding discharge displayed a clear-cut position sensitivity: the response to the same ramp movement in the most released position reached 65 Hz, whereas in more stretched position it was <25 Hz. However, the position-related movement sensitivity was not linear but hyperbolic. In contrast, the activity during maintained positions did not seem to be correlated to the position itself. Nevertheless, this activity was greater during the release than the stretch phase of the imposed movement. During the CBCO stretch, there was no clear difference between both types of firing (ramps or plateaus). Although these fibers were activated by release movements, they did not code the velocity, neither for the movement-related firing (Fig. 7B, white histogram) nor for the maintained position activity (Fig. 7B, dark histogram).
Dep MN monosynaptic reflex responses
To analyze the MN monosynaptic response, we performed intracellular recordings from the Dep MNs in high-Ca2+ and high-Mg2+ concentration saline. As shown in Fig. 8 (left), in a normal saline the Dep MN monosynaptic reflex response induced by CBCO mechanical stimulation was strongly masked by all the central inputs that reach the Dep MN. Perfusing the divalent cation-enriched saline exclusively onto the ganglion allowed us to reduce dramatically the central activities without affecting the sensory afferent information (Fig. 8, right). In these conditions, the Dep MN monosynaptic responses were unmasked, and the functional relationship between the type of response of the Dep MNs and their intrinsic properties could be analyzed.
DEP MN MONOSYNAPTIC REFLEXES: THE DIFFERENT RESISTANCE RESPONSES.
The reflex responses displayed by the Dep MNs are variable (Le Ray and Cattaert 1997
TONIC/PHASIC PROPERTIES OF MN AND MONOSYNAPTIC RESPONSE TO MOVEMENT.
Because it has often been claimed that phasic MN have larger diameters and higher conduction velocities, we used these parameters to classify the Dep MNs. Figure 10A shows the conduction delays of 59 Dep MNs recorded from 12 experiments. The different MNs are grouped according to the monosynaptic reflex response they displayed, i.e., assistance, phasic resistance, phaso-tonic resistance, or no response. In the four cases, no correlation appeared between conduction delays (which were measured as the time required for spikes to travel from the intracellular recording sites to the extracellular en passant electrode on the depressor motor nerve) and the kind of monosynaptic response. The dispersion observed in the delays in each group is quite large: from 2.35 to 10 ms for the no responding Dep MNs (
ACTIVE MEMBRANE PROPERTIES OF MN AND MONOSYNAPTIC RESPONSE TO MOVEMENT.
The Dep MN V-I curves were performed in a 5-mV range around the membrane potential to estimate the role of the membrane properties in the reflex response, because CBT-related EPSPs generally may reach up to 5 mV in MNs. V-I curves, resulting from injection of depolarizing and hyperpolarizing current pulses, revealed four types of electrical behavior in high divalent cation saline (Fig. 10B). Some Dep MNs (35%) displayed almost a linear response to current injection. The majority (51%) presented a negative rectification for negative current pulses ( In this paper, we have characterized the properties of both elements involved in the "stretch reflex" in crustacean walking legs: the CBCO sensory afferents and the connected MNs. We will discuss successively the functional aspects of the present results, the types of sensory coding and MN responses.
Crayfish thoracic ganglion in vitro preparation
The crayfish thoracic walking system is dissected out in such a way as to eliminate most of the sensory inputs that can excite the MNs. Only the CBCO is kept intact in the Petri dish with its sensory nerve that innervates the fifth thoracic ganglion. Perfusing the ganglion with high-Ca2+ and high-Mg2+ saline raised the threshold for spiking and thus strongly limits polysynaptic pathways. This allowed us to study much more clearly the monosynaptic connections between the CBCO sensory afferents and the MNs, which were emphasized by this procedure (Fig. 8). In all experiments, we were very careful to have very stable conditions and reproducible responses before performing the recordings (~2 h after the onset of the dissection). The CBCO was dissected out and arranged in the Petri dish in such a position that the puller imposed movements to the strand in exactly the same way as if it were in the intact leg. The CBCO strand length was measured in the fully released position, to apply stretches that were one-third of the original length. All stimulations were applied in this angular sector that corresponds to the angular range covered during the locomotor movements of the leg: the CB joint allows movements within the angular sector from the maximum of levation when the leg contacts the thorax, to the maximum of depression when the leg contacts the ventral carapace. Accidentally the CB joint can reach these extreme positions, for example when the animal is upside down. Nevertheless, movements involved in locomotion or in posture are restricted to ~60°, between the horizontal plane and the complete levation. During levation the CBCO strand is completely relaxed, and during depression it is slightly stretched; we focused our study on this functional part of CB joint movements.
Sensory unit coding
In the past, the crustacean walking leg chordotonal organs have been classified into two types: the first type presents a symmetrical discharge, and comprises the pro-dactylopodite (PD), the carpo-propodite (CP1), the mero-carpopodite (MC1), and the CB chordotonal organs, i.e., they have fairly the same number of stretch-sensitive and release-sensitive sensory units. The second type only possesses release-sensitive sensory fibers and comprises the CP2, the MC2 and the basi-ischiopodite chordotonal organs (Mill 1976 Monosynaptic motor neuronal responses
Recently, different studies on the crustacean MNs have demonstrated that, in the thorax, the abdomen and the stomatogastric ganglia, motor neuronal pools are quite complex and not homogenous. Their intrinsic properties and their connections have been described to be responsible for this heterogeneity (Harris-Warrick et al. 1992
INTRODUCTION
Abstract
Introduction
Methods
Results
Discussion
References
; Pringle 1961). At first considered rigid and stereotyped, the concept of the reflex as being capable of extensive modulation has evolved in parallel with the data accumulated on the central pattern generator (Barnes and Gladden 1985; Clarac 1991; Pearson 1993).
; Whitear 1962; Wiersma 1959). They are composed of an elastic strand in which 10-100 bipolar cells are inserted. Mill (1976) defined four different classes of sensory afferents in the dactyl chordotonal organ. Two are sensitive to changes of the chordotonal organ strand length; they operate in opposite direction, one being stretch sensitive, the other release sensitive, whatever the position. The two others code the strand length (stretched or released states of the strand) and are mainly activated for extreme angular positions and minimally activated for midramp positions. Bush (1962, 1965b) demonstrated that chordotonal organs are able to induce a stretch reflex, called a resistance reflex, which consists in the activation of the passively stretched muscle. The stretch reflex is not limited to its proper joint but can spread out on other joints or even on other legs (Clarac 1977; Vedel and Clarac 1979). Intensity and gain of this reflex appear to be very variable and may lead to complete reversal. For example, in crayfish, no resistance reflexes seem to exist during treadmill walking, while they could be generated during unintended movements (Barnes 1977). In stick insects (Bässler 1976), in crabs (Di Caprio and Clarac 1981), and in crayfish (Skorupski and Sillar 1986), it has been demonstrated that the same mechanoreceptor is able, depending on "the central state" of the CNS, to change a resistance reflex into an assistance reflex. Such variability has been similarly found in mammals (Pearson 1993), including humans (Duysens and Tax 1994).
; Burrows 1975). The variability of the reflex has often been associated with interneurons that are inserted within the reflex pathway, between sensory terminals and motor neurons (MNs) (Burrows 1992; Büschges and Schmitz 1991; Büschges and Wolf 1995; Le Ray and Cattaert 1997). However, in the crayfish, monosynaptic connections between the coxo-basal chordotonal organ (CBCO) and the levator (Lev) and depressor (Dep) MNs that command the coxo-basal joint have been demonstrated to be responsible for efficient resistance reflexes (El Manira et al. 1991).
) but keep intact the direct CBCO afferent-motor neuronal pathways.
). Also, we have restricted mechanical movements imposed to the CBCO strand to its most released range, where this reflex is the most efficient. In this paper, we have classified the different release-sensitive sensory afferents and the different postsynaptic Dep MNs according to a qualitative and quantitative analysis of their responses and properties. In the following paper, using paired intracellular recordings, the wiring of the sensory-motor connections will be established, and the role of specific synaptic parameters controlling MN response will be presented.
METHODS
Abstract
Introduction
Methods
Results
Discussion
References
View larger version (28K):
[in a new window]
FIG. 1.
Diagram of the experimental arrangement used in the analysis of sensory-motor activities. A: in vitro preparation of the crayfish walking system consisted of the last 3 thoracic ganglia (G3, G4, G5), and the nerves that innervate the 2 proximal joints of the left 5th leg. We kept also intact the coxo-basipodite chordotonal organ (CBCO), a proprioceptor that encodes the vertical movements of the leg. A mechanical puller allowed us to reproduce physiological stimulations to the CBCO strand, mimicking the vertical movements of the leg. Large dots indicate the location of en passant electrodes onto the depressor nerve (Dep n) and the CBCO sensory nerve (CB n). B: intracellular microelectrodes within the neuropile were used to record from a sensory terminal (CBT) and a Dep motor neuron (Dep MN). C: sine-wave movements imposed to the CBCO strand induced sensory activity recorded from both the CBCO nerve and a CBT (resting potential: 
70 mV). D: the same movement elicited a resistance reflex recorded from a Dep MN (resting potential: 68 mV).
resistance. The signals were amplified by an Axoclamp 2B amplifier (Axon Instruments). Intracellular current pulses and nerve stimulations were controlled by an eight-channel digital stimulator (A.M.P.I.). All signals were monitored on an eight-channel oscilloscope and a four-channel digital oscilloscope (Yokogawa DL 1200), stored on D.A.T. tapes (BioLogic digital tape recorder), and digitized on a PC-based computer through an A/D interface (from Cambridge Electronic Device, CED 1401PLUS). Intracellular and extracellular recordings were digitized at 5-10 kHz and written to disk.
View larger version (24K):
[in a new window]
FIG. 4.
Phasic release-sensitive CBTs. A: phasic CBTs were characterized by spike bursts occurring only during release ramps (1.25 mm/s). As shown in the inset, their firing stopped since the plateaus started. The mean firing frequency calculated over 9 stimulation cycles for each ramp is presented below and demonstrates that the position of the leg did not affect the firing frequency of this phasic CBT (resting potential: 77 mV). B: mean firing frequencies from all individual release ramps over the 9 cycles were calculated for 0.25 and 1.25 mm/s ramp velocities. The phasic coding of this CBT was highly velocity dependent, because a 5-fold ramp velocity increase leaded to a 3-fold firing mean frequency increase (P < 0.001). Error bars: SE. Mvt, imposed ramp movement.
-2-ethanesulfonic acid (HEPES) and pH adjusted at 7.7 at 15°C.
View larger version (34K):
[in a new window]
FIG. 8.
Effect of high-Ca2+, high-Mg2+ saline on reflex activity. In normal saline (left), the Dep MN (intracellular recording, resting potential: 70 mV) is subjected to a lot of central inputs that mask the purely monosynaptic influences. By perfusing high-Ca2+, high-Mg2+ saline (right), the central activity threshold is increased, which dramatically reduces the inputs that reach Dep MNs and allows us to unmask the CBCO direct excitation. a Lev and P Lev, extracellular recordings from the anterior and posterior levator nerves. Mvt, imposed ramp movement.
) were used to study the sensory coding of the cyclic imposed movement. Each sensory unit was identified over several movement cycles. Each cycle is defined from the starting of the release phase (chosen as zero) to the next release phase (chosen as 360°). Each spike is figured by a unit vector 
i, with a phase angle according to its occurrence in the stimulus cycle. The resulting mean vector 
is the vectorial sum of all unit vectors, divided by the number (n) of spikes (Fig. 2A)
with Rx and Ry representing the x and y coordinates of the resulting vector
.
View larger version (30K):
[in a new window]
FIG. 2.
Extracellular study; methodology. A: procedure used to analyze the different sensory units extracellularly recorded from the CBCO nerve. In this illustration, 3 units were identified by their shape and amplitude. The activity of each unit was then represented using circular statistics. Within the stimulation cycle, each occurrence of the unit (unit 1 in the example) was represented by a unit vector ( i). The vectorial sum of these unit vectors was calculated over several cycles and divided by the number n of occurrences; the resulting mean vector () indicated the coding characteristics (angle 
and vector length R are polar coordinates) of the unit (see text for details). B: bimodal distribution of resulting vector lengths, calculated from all analyzed CBCO units. This distribution was used to distinguish between specific (R > 0.3) and nonspecific (R 
0.3) CBCO units. C: with the use of the calculated and R values, CBCO units were classified in 10 different experiments. In the angular sector we studied, most of the release-sensitive fibers were activated. Stretch-sensitive units were generally, and nonspecific units were always activated less frequently.
) has a value between 0 and 1, and is a measure of the dispersion of the spikes within a stimulus cycle. A value of zero means that the spikes are evenly distributed; a value of one would indicate that all spikes occurred at the same phase. The frequency distribution of over all experiments (Fig. 2B) is bimodal, the separation between the two peaks being 0.3. Therefore units with value below 0.3 were considered asnonspecific units (Fig. 2C), whereas the others were classified as release or stretch-sensitive units.
RESULTS
Abstract
Introduction
Methods
Results
Discussion
References
0.3; see METHODS). Figure 3 presents a typical experiment in which different types of CBCO units were characterized from the CBCO neurogram, using the "Wave Marker" Spike2 program. The different identified extracellular spike templates are presented with the corresponding event histograms, which represent the occurrence of each extracellular spike in the stimulation cycle (bin size: 3 ms). In this experiment, 14 distinct release-coding and 5 stretch-sensitive CBCO units could be differentiated. Nevertheless, three sensory units did not code preferentially one direction of the imposed movement (R 0.3) All the release-sensitive fibers responded during the whole range of the release movement. Therefore it appears that the CBCO coding is distributed, i.e., sensory fibers do not code an extremely specific range of the movement. The phase () of the mean of each release-sensitive unit varied between 92.3 and 152.5°, which indicates some differences in the coding of the various sensory units. Moreover, although some interunit differences exist (event histogram peaks occurred between 65 and 125°), most of the unitsfired when the velocity of the imposed movement was maximum (90°).
View larger version (35K):
[in a new window]
FIG. 3.
Characterization of the CBCO units recruited by sinusoidal stimulation. During sine-wave stimulation applied to the CBCO strand, the activity of 19 distinct sensory units was analyzed. For each unit, an event distribution histogram was calculated over 12 cycles. The movement cycle is drawn at the bottom of each column (scale bar: 1 s). On the right of each histogram, the corresponding unit profile is shown (scale bar: 2.5 ms). Below each histogram, the calculated and R values are reported.
View larger version (33K):
[in a new window]
FIG. 5.
Phaso-tonic release-sensitive CBTs. Phaso-tonic CBTs fire phasically during release ramps and develop a tonic discharge for the most released position of the CBCO strand. A: 1st type of phaso-tonic CBTs continued to produce spikes even after the beginning of the CBCO stretch (resting potential of CBT1: 73 mV). B: 2nd type stopped once the stretch started (resting potential of CBT2: 70 mV). A and B: mean firing frequency histograms (same disposition as in Fig. 4A): both phasic (white bars) and tonic (filled bars) codings were dependent on the position of the leg. Regression lines are drawn for each histogram during the release phase (correlation coefficient are in A: 0.99 for the phasic component, 0.96 for the tonic component; in B: 0.97 for the phasic component, 0.93 for the tonic component). C: same calculation as in Fig. 4B, for both phasic and tonic components. Only the phasic pattern (white histogram) was velocity dependent (P < 0.001), whereas the tonic discharge (filled histogram) was not significantly affected by the ramp speed. Error bars: SE. Mvt, imposed ramp movement.
) plotted as a function of time, after the mechanical stimulation was stopped in the most released position (time 0), for one phaso-tonic fiber. The basal instantaneous frequency value was obtained after only 6 s in the released position.
View larger version (34K):
[in a new window]
FIG. 6.
Phaso-tonic CBTs show fast adaptation. A: when movement is stopped in a released position, the firing frequency of the CBT (raw data) quickly reduces to minimum value (resting potential of CBT: 70 mV). B: plot of the instantaneous frequency () and of the mean frequency (white bars), corresponding to the recording in A. The basal spiking frequency (~2.5 Hz) is reached within 5 s after the maintained released position. Error bars: SE. Mvt, imposed ramp movement.
View larger version (44K):
[in a new window]
FIG. 7.
Continuously firing release-sensitive CBTs. A: during the application of a ramp stimulation to the CBCO strand, some CBTs were continuously active, but fired many more spikes during the release phase of the stimulation. As is illustrated by the mean frequency histogram (same disposition as in Fig. 4A), the continuously firing CBT shows a clear movement sensitivity (white bars) and a much weaker purely tonic activity (filled bars). The line represents the hyperbolic correlation between the phasic component mean frequency and the position of the leg (r = 0.96). During the CBCO stretch, only spontaneous firing persisted (resting potential of CBT: 70 mV). B: same disposition as in Fig. 5C. Neither the phasic component (left) nor the tonic discharge (right) were velocity dependent. Mvt, imposed ramp movement.
): there exist one Dep MN involved in assistance reflex (i.e., activated during stretch of the CBCO), and eight Dep MNs involved in the resistance reflex (i.e., activated by CBCO release). If the response of the assistance Dep MN is stereotyped, i.e., EPSP bursts occurring exclusively during stretching ramps of the CBCO stimulation, the eight resistance Dep MNs are able to display responses of different shape. Ramp movements (with the same plateau duration, 2 s) imposed to the CBCO strand were able to produce in the resistance Dep MNs two different types of resistance responses (Fig. 9). Some Dep MNs showed a reflex response characterized by bursts of EPSPs (~1-4 mV amplitude) during the releasing ramps (Fig. 9A). These bursts of EPSPs were almost constant from the first to the last releasing movement of a stimulation cycle. This kind of response, where EPSP bursts stopped at the beginning of the plateaus (see detail in inset of Fig. 9A), was defined as a "phasic response." The second type of reflex response (Fig. 9B) was characterized by the same phasic bursts of EPSPs, which summated on a general depolarization of the MN that occurred during the whole release of the CBCO. In this case, the amplitude of the EPSP burst weakly increased from the first to the last releasing ramp (1 ± 0.4 mV during the 1st ramp, and 3 ± 0.3 mV during the last). In parallel, the Dep MN depolarization was increasing from the beginning until the end of the CBCO release (from 0 to 1.7 ± 1 mV). This kind of response (see detail in Fig. 9B, insert) defined a "phaso-tonic" Dep MN response. As shown in Fig. 9C, although the phasic bursts of EPSPs swiftly disappeared, the tonic depolarization of the phaso-tonic Dep MNs persisted (with an almost constant value) when the stimulation was stopped in a released position. The Dep MN completely repolarized only when the CBCO stretch started.
View larger version (25K):
[in a new window]
FIG. 9.
Monosynaptic resistance Dep MN responses. Resistance Dep MNs displayed 2 types of monosynaptic response to CBCO release. A: some only responded with phasic bursts of excitatory postsynaptic potentials (EPSPs) occurring during the release ramps (detail in inset; resting potential of Dep MN: 71 mV). B: others displayed an increasing tonic depolarization of their membrane potential on which the phasic EPSP bursts summated (detail in inset; resting potential of Dep MN: 70 mV). C: the tonic component of the resistance response persisted when the CBCO was kept released, and disappeared since the 1st stretching movement (resting potential of Dep MN: 70 mV). Mvt, imposed ramp movement.
); from 2.5 to 8.75 ms for the phaso-tonic resistance MNs (
); from 2.19 to 8 ms for the phasic resistance MNs (
); and from 3.44 to 5.39 ms for the assistance Dep MN (*). This dispersion of conduction delay values for the unique assistance Dep MN would indicate some variability from one animal to the other. Nevertheless, in a single experiment where all Dep MNs were recorded, the distribution of conduction delays was similar and did not show any relationship with the reflex response displayed by the different Dep MNs.
View larger version (20K):
[in a new window]
FIG. 10.
Dep MN properties and reflex responses. A: column scatter of conduction velocities of Dep MNs with respect to their monosynaptic reflex responses. B: 4 types of voltage-current curves observed in Dep MNs: a linear behavior, a double negative rectification, or a simple negative rectification for hyperpolarizing ( 
V1) or depolarizing (V2) currents (see details in the text). C: plotting of V2 against V1 for 37 Dep MNs. The distribution is not homogeneous, but the Dep MNs are grouped in 5 clusters. Only the assistance responding Dep MN showed particular active electrical properties (cluster 5).
V1) and a weak positive rectification for positive current pulses (V2). A smaller number of Dep MNs (8.5%) showed a double negative rectification (V1 and V2). The last type of electrical behavior, a positive rectification V1 and a negative rectification V2, was detected in few cases (only 5.5% of the impaled Dep MNs). V1 and V2 values were calculated as the differences between the linear regression (around 0) and the real values for 5 and +5 nA injected current amplitudes.
V2 as a function of V1 for each MN (Fig. 10C). This diagram revealed clusters of Dep MNs that would indicate distinct behavioral groups of MNs. However, each cluster contained phasic () and phaso-tonic () resistance Dep MNs, except for cluster 5 (positive V1 and negative V2), which contained exclusively the assistance Dep MN (*). Nonresponding Dep MNs () were also found in clusters 1 and 2 (moderate negative V1), but not in clusters 3 and 4 (larger negative V1).
DISCUSSION
Abstract
Introduction
Methods
Results
Discussion
References
; Mill and Lowe 1973). Other chordotonal organs are not disposed at a joint: the cuticular stress detectors (CSD1 and CSD2), located near the CB joint, are linked to small windows of soft cuticle and respond to mechanical stretching and releasing of their strand when the cuticle is stimulated (Marchand et al. 1995). In 1991, El Manira et al. described the CBCO: it is composed of 40 cells, 1/2 of which are sensitive to the stretch of the strand, whereas the others are sensitive to its release. In this study the analysis of the sensory coding was performed using two types of mechanical stimulation imposed to the CBCO strand.
,b; Mendelson 1963; Mill 1976) and, from more sophisticated tools, give new insights concerning chordotonal organ fine sensory coding. In the past, most of the studies were performed from extracellular recordings (Bush 1965a,b; Clarac 1970), the analysis being based on the amplitude of the action potentials, a technique that makes discrimination of units of similar size uncertain. Our extracellular analysis, based on the extracellular action-potential shape recognition, thoroughly studied the discharge of almost all the sensory units excited during movements imposed over a physiological range. Sensory terminal intracellular recordings allowed us to identify specifically different types of sensory units. They can be divided into three subtypes that are the phasic, the phaso-tonic and the continuously firing sensory fibers. This classification is mainly based on the position sensitivity because all of them are strongly sensitive to movement. The phasic units only responded to movement, their response being unchanged whatever the angular sector where it was measured. By contrast, in both phaso-tonic and continuously firing units, the response to movement was considerably affected by the angular sector where it was measured. Therefore, among movement-sensitive responses (dynamic responses), we can distinguish the position-unaffected dynamic (phasic units) from the position-modulated dynamic responses (phaso-tonic and continuously firing units). In addition, the dynamic response may be velocity sensitive (phasic and phaso-tonic units) or not (continuously firing units). Surprisingly, we did not find purely position-coding units as are usually described (Bush 1965a; Clarac 1970; Matheson 1990), for all units coding positions were also activated during movements (phaso-tonic CBTs). Whatever the imposed movement (ramp or sine wave), it appeared that each sensory fiber is activated in a wide angular sector. We never recorded any fiber coding specifically for a small joint angular sector.
; Le Ray and Cattaert 1997; Leibrock et al. 1996; Skorupski et al. 1992; Wine 1984). In the same MN pool, it seems that not all the MNs have the same roles. Skorupski et al. (1992) demonstrated that there were two groups of MNs in the promotor pool: one involved in the resistance reflex and the second in the assistant pathway. More recently, Le Ray and Cattaert (1997) showed in the Dep MN pool the existence of one assistance-specialized Dep MN. In this study, we classified the Dep MNs according to their monosynaptic reflex response and their intrinsic properties. We did not take into account the electrotonic coupling that has been described as being very weak between Dep MNs (Chrachri and Clarac 1989). Classically MNs have been classified into tonic and phasic depending on the type of muscle fibers they innervate and their activity (Kennedy and Takeda 1965a,b). Tonic MNs display continuous discharge, whereas phasic MNs are activated only during rapid movements (Kennedy and Davis 1977). Generally, phasic MNs have cell bodies and axon diameters larger than tonic ones, and they convey spikes more rapidly. Functionally, the resistance reflex is acting during posture and should involve preferentially postural MNs, i.e., tonic MNs. In the present work, we did not attempt to classify Dep MNs according to the muscle fibers they innervate or their activity in motor behavior. However, we did attempt to classify Dep MNs according to their conduction velocity (Fig. 10A) and their firing pattern. Surprisingly, this study demonstrates that this classification did not match the different types of reflex responses, because resistance reflex Dep MNs were found in the whole range of the conduction delays (Fig. 10A). It cannot, however, be excluded that conduction delays may not be related to the tonic or phasic behavior of a given MN. Concerning the three MNs that are not directly connected by the CBCO sensory afferents, they were also found in the whole range of Dep MN conduction delays (Fig. 10A). Nevertheless, this group may also be heterogeneous, some of them being excited through one or several interneurons and others not. Moreover the properties of interneurons involved in polysynaptic pathways may be very different from one MN to the other. Such an organization has already been demonstrated in the case of the insect locomotor system (Burrows and Pflüger 1988; Sauer et al. 1995) and in the case of the crayfish reflex pathways (Clarac et al. 1991), and might be very important in reflex modulation.
| |
ACKNOWLEDGEMENTS |
|---|
We thank Dr. L. Vinay for helpful comments on the manuscript.
This work was supported by Groupement d'intérêt scientifique (sciences de la cognition) contract (CNA10) and the Centre National de la Recherche Scientifique.
| |
FOOTNOTES |
|---|
Address for reprint requests: D. Le Ray, Laboratoire Neurobiologie et Mouvements, CNRS, 31 chemin Joseph Aiguier, 13402 Marseille cedex 20, France.
Received 14 April 1997; accepted in final form 22 July 1997.
| |
REFERENCES |
|---|
|
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
This article has been cited by other articles:
|
|
 |
|
  A. A. V. Hill and D. Cattaert Recruitment in a heterogeneous population of motor neurons that innervates the depressor muscle of the crayfish walking leg muscle J. Exp. Biol., February 15, 2008; 211(4): 613 - 629. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. A. DiCaprio, C. P. Billimoria, and B. Ch. Ludwar Information Rate and Spike-Timing Precision of Proprioceptive Afferents J Neurophysiol, September 1, 2007; 98(3): 1706 - 1717. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Le Ray, D. Combes, C. Dejean, and D. Cattaert In Vivo Analysis of Proprioceptive Coding and Its Antidromic Modulation in the Freely Behaving Crayfish J Neurophysiol, August 1, 2005; 94(2): 1013 - 1027. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Le Bon-Jego, D. Cattaert, and E. Pearlstein Serotonin Enhances the Resistance Reflex of the Locomotor Network of the Crayfish through Multiple Modulatory Effects that Act Cooperatively J. Neurosci., January 14, 2004; 24(2): 398 - 411. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Le Bon-Jego and D. Cattaert Inhibitory Component of the Resistance Reflex in the Locomotor Network of the Crayfish J Neurophysiol, November 1, 2002; 88(5): 2575 - 2588. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Cattaert and M. Bevengut Effects of Antidromic Discharges in Crayfish Primary Afferents J Neurophysiol, October 1, 2002; 88(4): 1753 - 1765. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. L. Ray, F. Clarac, and D. Cattaert Functional Analysis of the Sensory Motor Pathway of Resistance Reflex in Crayfish. II. Integration of Sensory Inputs in Motor Neurons J Neurophysiol, December 1, 1997; 78(6): 3144 - 3153. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
