| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Section of Neurophysiology, Department of Physiology, University of Copenhagen, 2200 Copenhagen N, Denmark; and 2 Section of Neurobiology and Behavior, Cornell University, Ithaca, New York 14853
 |
ABSTRACT |
|---|
 Abstract
 Introduction
Methods
Results
Discussion
References
|
|---|
Raastad, Morten, Bruce R. Johnson, and Ole Kiehn. Analysis of EPSCs and IPSCs carrying rhythmic, locomotor-related information in the isolated spinal cord of the neonatal rat. J. Neurophysiol. 78: 1851-1859, 1997. To understand better the synaptic language used by neurons in active networks, we have analyzed postsynaptic currents (PSCs) received by interneurons in the isolated spinal cord from neonatal rats during 5-hydroxytryptamine- and N-methyl-D-aspartate-induced fictive locomotion. Using a computer algorithm, we identified PSCs in rhythmically active interneurons in laminae VII and X. To test whether the PSCs actually participated in the transmission of the cyclic, locomotor-related signal, we constructed an analytic current trace based on only the identified events. Each identified PSC was fitted by a mathematical function, and the shape of this function was added to a baseline with time delays given by the time positions of the identified PSCs. By averaging the resulting analytic current trace over several cycles, we showed that the identified PSCs built a cyclic signal locked to the rhythmic activity recorded from the ventral roots. Furthermore, subtraction of the analytic from the original current trace reduced the amplitude of the cyclic signal received by these cells. Thus the identified PSCs contributed to the cyclic information, allowing us to analyze how they built the compound cyclic signal. Most often there was an inverse relationship between the contribution from excitatory and inhibitory PSCs during the cyclic modulation, indicating that there was a reciprocal regulation of the presynaptic inhibitory and excitatory cells. Comparing the most inhibitory and most excitatory halves of the locomotor related cycle, there was a considerably larger modulation of the frequency of PSCs than of their amplitude. The small and sometimes insignificant modulation of PSC amplitude suggests that facilitation and depression had little importance for the information transfer. The modest amplitude modification also suggests that the large range of available PSC amplitudes seen in these neurons was not used very efficiently to code the cyclic information.
Understanding the functional organization of a central neural network requires an analysis of the relevant synaptic input that modulates neuronal activity during network function. It has been difficult to study in detail the synaptic information transfer during activity in central vertebrate networks because the necessary experimental conditions and normal network activity are often unobtainable together. We are interested particularly in the functional organization of the central pattern generator (CPG) networks of the mammalian spinal cord because they have the ability, even in reduced preparations, to produce locomotor rhythms that closely resemble those seen in intact animals (Grillner 1981 The methods for preparing the isolated neonatal rat spinal cord, inducing rhythmic locomotor-like activity from the cord, and recording the activity extracellularly from the ventral roots and from interneurons intracellularly, using tight-seal whole cell recordings have been described previously in detail (Kiehn and Kjaerulff 1996 Preparation
Neonatal rats (0-3 days old) were anesthetized deeply with ether, decapitated, and the spinal cord extending from C1 to L6 including the ventral and dorsal roots was removed. For better access to spinal interneurons, the cord was split midsagittally in most experiments from T13 to L6, and the dorsal side of the cord was discarded. The preparation was transferred to a recording chamber, pinned down, and superfused with oxygenated (95% O2-5% CO2) Ringer solution of the following composition (in mM): 128 NaCl, 4.7 KCl, 25 NaHCO3, 1.2 KH2PO2, 1.25 MgSO4, 2.5 CaCl2, and 20 glucose (pH 7.4) at room temperature. Locomotor activity was induced by bath application of N-methyl-D-aspartic acid (NMDA, 6-7 µM) in combination with 5-hydroxytryptamine (5-HT, 4-20 µM). The drugs were obtained from Sigma (St. Louis, MO) or RBI (Natick, MA).
Recording
We combined ventral root recordings with intracellular tight-seal whole cell recordings (Blanton et al. 1989 Synaptic detection algorithm
As previously described (Raastad et al. 1996 Interneurons receiving cyclic locomotor-related information
Concomitant recordings extracellularly from spinal cord motor roots and intracellularly from cells located close to the central canal or in the intermediate gray matter often showed interneurons with rhythmic membrane potential modulation that correlated with rhythmic ventral root activity (Kiehn et al. 1996
Identification of PSCs giving locomotor-related information
If the detected events carried cycle-related information, one would expect to reduce the magnitude of the total cyclic information a cell received by subtracting these events from the original current trace. An alpha function was fitted to putative PSCs, like those in Fig. 1C, and the area (charge) of the function was subtracted for each identified PSC over several locomotor cycles. This procedure enabled us to test whether a significant proportion of the cyclic information could be accounted for by the detected PSCs. Figure 2A shows 10 s of rhythmic root activity (top) and the simultaneously recorded intracellular current (marked I) when the soma was clamped at
Decomposition of the synaptic information
We next examined the relative contributions of the excitatory and inhibitory synaptic events to the generation of cyclic information. For the rest of the analysis, we arranged the intracellular locomotor-related cycles so that the most inhibitory (positive) phase was always in the first half of the cycle (Fig. 3A, left). We can distinguish three ways the PSCs can build a cyclic signal. The simplest possibility is to transfer the cyclic signal by EPSCs or IPSCs alone. Either a relative increase in the amount of excitatory charge or a relative decrease in the amount of inhibitory charge could make the second half of the cycle more excitatory than the first half. A second possibility is that both the inhibitory and the excitatory charges together form the cyclic signal. This would mean that the excitatory charge increased and the inhibitory charge decreased in the second compared with the first cycle half. A third possibility is that the amount of both inhibitory and excitatory charge increases simultaneously in one cycle half. In this case, the main cyclic signal would be dominated by either the EPSCs or the IPSCs, whichever contributed the largest charge difference between the two halves of the cycle. The less dominating charge in that case would counteract the main signal. The following analysis will demonstrate that all three possibilities are realized in the cyclic information-driving rhythmically active interneurons.
PSC frequency and amplitude contributions to the cyclic signal
The relative changes in synaptic charge comparing the first and the second cycle halves must be due to changes either in the frequency or the average size of the PSCs or both these factors. The total charge contribution can be calculated as the average charge of the PSCs multiplied by their frequency, illustrated in Fig. 4A for the EPSCs and IPSCs from Fig. 3A. In this experiment, the average size (charge) and the number of PSCs occurring within 10 equal time bins during an average cycle show that the size changes very little, whereas there is an obvious modulation in the frequency of PSCs during the cycle.
In this study, we have recorded from rhythmically active interneurons located in the intermediate gray matter and around the central canal in the isolated neonatal rat spinal cord during transmitter-induced locomotion. Neurons in these areas probably are involved in locomotor rhythm generation (Carr et al. 1995 Detection of PSCs
The majority of detected events are probably the result of transmitter release from individual release sites or release from several release sites synchronized by individual presynaptic neurons (unitary PSCs) because of the following observations: 1) the shape was characterized by an abrupt onset, fast and smooth rise time, and a slower decay, similar to unitary synaptic currents in other central neurons; 2) the outward currents were reduced greatly in amplitude, and therefore not detected, at Contribution from IPSCs and EPSCs to the rhythmic signal
In most of the interneurons from which we recorded, the cyclic locomotor related signal was mediated by an inversely related contribution from both EPSCs and IPSCs (Fig. 3). This alternating dominance between EPSC and IPSC activity, or push-pull drive (Russell and Wallen 1983 PSC frequency and amplitude modification
Although significant modulation in both average amplitude and frequency were seen in the cyclic information transfer, frequency modification of the PSCs was contributing more than amplitude modification to the creation of the cyclic locomotor-related signal. The interpretation of the observed changes in PSC frequency depends on several unknown presynaptic factors. If the synapses in this network typically have few release sites combined with low release probability at the individual sites (as seen, for example, in the hippocampus: Arancio et al. 1994
INTRODUCTION
Abstract
Introduction
Methods
Results
Discussion
References
; Rossignol 1996). Such preparations may allow the type of detailed analysis necessary to examine relevant synaptic information transfer in an operating central vertebrate network. We are presently far from understanding these CPG networks because the creation, flow, and cellular integration of the network signals are understood poorly. An important step to understand the mammalian CPG function is to analyze how synaptic signals are used to transmit the cyclic locomotor-related signal. Fast synaptic input, i.e., synapses with ionotropic receptors, could drive cyclic locomotor activity by modulation of the sign (excitatory or inhibitory) and by modulation of the frequency and/or the amplitude of the individual synaptic events. Earlier studies have examined the composition of synaptic inputs driving locomotor activity in vertebrate motoneurons by inverting the inhibitory synaptic potentials, either by filling motoneurons with chloride or by hyperpolarizing them. The locomotor CPG network was found to control motoneuron activity by alternating fast excitatory and inhibitory synaptic events during fictive locomotion (Cazalets et al. 1995; Jordan 1983; Perret 1983; Russell and Wallen 1983). Thus there is a reciprocal activity in presynaptic inhibitory and excitatory neurons to drive the motoneurons to fire rhythmically. In this study, we analyzed the fast synaptic signals received by interneurons in the rat spinal cord that were rhythmically active in phase with ongoing fictive locomotion. We asked if the collaboration between excitatory and inhibitory inputs that drives motoneurons was responsible for cyclic activity in a higher level of the motor pathway, interneurons possibly participating in the spinal locomotor CPG. Alternatively, other forms of synaptic signaling could be dominant here. For example, mutual reciprocal inhibition dominates the synaptic signaling in many invertebrate motor CPGs (Marder and Calabrese 1996; Mulloney and Perkel 1988), and a similar organization has been suggested for the rhythm-generating kernel in the mammalian spinal CPGs (Brown 1911; Jankowska et al. 1967; Lundberg 1979; Pearson and Collins 1993; see also Kiehn et al. 1997 for a review).
; Clamann et al. 1985; Mendell and Henneman 1970). On the other hand, if frequency modification of synaptic input is the dominating mechanism in the control of postsynaptic activity, then the factors controlling presynaptic firing rates are important to study.
; Cowley and Schmidt 1994; Kiehn and Kjaerulff 1996; Kudo and Yamada 1987; Smith and Feldman 1987). This preparation is well suited for studies of network activity at a cellular level because spinal interneurons in the lumbar region can be recorded from with whole cell patch electrodes (Kiehn et al. 1996).
) and neonatal rat (Kjaerulff et al. 1994) and spinal cord lesion studies (Kjaerulff and Kiehn 1996), these areas are thought to be important for locomotion. The majority of cells in these areas in the neonatal rat show rhythmic locomotor activity apparently driven mainly by synaptic input rather than by intrinsic bursting/plateau properties (Kiehn et al. 1996). A detailed analysis of how the network uses the synapses for information transfer has been hampered previously by the fact that the individual synaptic events have been difficult to distinguish from each other during ongoing network activity. By voltage clamping the soma, we were able to gather recordings of clearly distinguishable fast synaptic currents that the cells receive and also reduce any contribution of active membrane properties to the rhythmic locomotor related activity.
METHODS
Abstract
Introduction
Methods
Results
Discussion
References
; Kiehn et al. 1996). Below follows a brief description of these procedures.
; Edwards et al. 1989) from rhythmically active spinal interneurons located around the central canal and in the intermediate gray matter. These are the areas that activity-dependent labeling and lesion studies suggest are involved in rhythm-generation in mammals (Carr et al. 1995; Kjaerulff and Kiehn 1996; Kjaerulff et al. 1994). Activity in the L2 and L5 ventral roots, corresponding to leg flexor and extensor activity, respectively (Cowley and Schmidt 1994; Kiehn and Kjaerulff 1996), was recorded with suction electrodes. Activity in interneurons was recorded with patch electrodes (5-10 M
) pulled from 1.5 mm borosilicate glass without filaments. Whole cell voltage or current recordings were made with either Axopatch 1-D or Axoclamp 2B amplifiers (Axon Instruments; Foster City, CA). The signals were filtered at 2-5 kHz and digitized at 5 kHz. Series resistance was followed throughout the experiments and was usually <25-30 M. The pipette solution contained (in mM) 130 potassium gluconate, 10 N-2-hydroxyethylpiperazine-N
-2-ethanesulfonic acid, 0 or 4 NaCl, 10 ethylene glycol-bis(
-aminoethyl ether)-N,N,N,N-tetraacetic acid, 1 CaCl2, 4 ATP-Mg, and 0 or 0.3 GTP-Li. The pH was adjusted to 7.3 with KOH. Intracellular potentials were not corrected for liquid junction potentials (Neher 1995).
), we used a computer algorithm to detect putative PSCs. In short, an event was considered a PSC candidate if the current deviated from baseline noise more than a specified value within 2 ms. This detection value was set initially so that many events were considered PSC candidates, relative to the final number accepted (see below). An alpha function was fit to the initial 10 ms of the current, to evaluate if the shape resembled a PSC (with an abrupt start, fast rise time and slower decay). The function used was f(t) = amp*
*exp(
*t), where amp is the peak amplitude in picoamps, t is time, and the factor characterizing the time course, and the squared deviation between the function and the original data were minimized. The function was defined only for positive values of t, and the current started at t = 2 ms. The function was fit only to the initial 10 ms of the single PSC because other PSCs and noise often contaminated the PSC tail.
40 to 50 mV) were detected at hyperpolarized potentials (60 to 75 mV). This suggested that the outward events were indeed inhibitory PSCs (IPSCs) (Raastad et al. 1996). This was one of our main criteria for accepting the detected events as PSCs and not noise. Other criteria included visually confirming that the detected events had a shape expected of PSCs and noting that the time courses were different between the outward and inward events (Raastad et al. 1996), which is often the case for inhibitory and excitatory PSCs (Jones and Westbrook 1996). All these features should characterize synaptic events and not noise.
RESULTS
Abstract
Introduction
Methods
Results
Discussion
References
). For example, Fig. 1A (bottom) shows 10 s of rhythmic ventral root activity (L2) during pharmacological induction of a locomotor rhythm with NMDA and 5-HT, whereas Fig. 1A (top) shows an interneuron firing out of phase with the ventral root activity. We cannot conclude that this neuron is a component of the spinal network producing locomotor activity, but it obviously received timing information from the locomotor network. The cyclic, locomotor-related information that the interneuron received appears to be mediated, at least partly, by fast synaptic events. These synaptic events can be better seen under voltage clamp, which reduces their duration and makes them stand out as fast outward and inward transients, probably representing IPSCs and excitatory PSCs (EPSCs), respectively (Fig. 1B, top).
View larger version (58K):
[in a new window]
FIG. 1.
A: 10 s of an intracellularly recorded membrane potential from an interneuron in segment L2 (top) and the concomitant recording of spikes in the L2 ventral root (bottom). Membrane potential is modulated in correlation with the spike activity in the root. B: same cell recorded in voltage-clamp mode shows postsynaptic currents (PSCs) as fast transients in positive and negative directions from a modestly fluctuating baseline. There are more positive transients during high root activity, and more negative transients during low root activity. C: 20 PSCs collected by a computer algorithm (see METHODS) from periods with high root activity (left collection) and 20 PSCs collected during low root activity (right).
). For example, Fig. 1C shows PSCs detected from the cell in Fig. 1 during low and high root activity. Twenty PSCs were detected during high root activity (first half), and 20 were detected during low root activity (second half). It is obvious that both IPSCs and EPSCs occurred during both activity phases of this interneuron. Most synaptic events detected had a time course expected from unitary PSCs, with an abrupt start, fast and smooth rise time, and a slower decay. This suggests that they were due to synchronized transmitter release from one or more release sites in individual presynaptic neurons.
45 mV. All accepted fits obtained from this trace were added at their appropriate time positions to form what we call the analytic current trace (Fig. 2A, II). When the analytic current trace was subtracted from the original trace, the result (Fig. 2A, I II) was a trace with less markedly transient events, suggesting that we identified PSCs that contributed to the rhythm related information. We call the last trace in Fig. 2A (I II) the residual current trace. It contains what was left in the original trace after subtraction of the detected PSCs and contains undetected and rejected PSCs, the residual of poor PSC fits, and perhaps rhythmic modulation of the whole cell current that is not mediated by fast synaptic transmission.
View larger version (34K):
[in a new window]
FIG. 2.
A, top: rhythmic activity in ventral root L2 while the voltage clamp recording (marked I) from an interneuron in the same segment shows inward and outward PSCs as fast transients. Computer algorithm detected fast events with time courses as expected from PSCs (see METHODS). Each event with a peak amplitude above a certain threshold was fitted with an alpha-function. Best fit was added to a line at the time position where it was detected, giving line II, which we call the analytic current trace. When the analytic current trace (II) was subtracted from the original current trace (I), a residual current trace was obtained (I II). This trace may contain undetected PSCs and current modulations that were slower than the fast PSCs. B: cyclic locomotor-related signal in each of the 3 traces above was estimated by calculating the average signal over many cycles (see METHODS for details). Each cycle was divided in 10 equal time bins, and the average current in each of these bins is shown as a bar with standard error as a vertical line. Original current trace contained an obvious cyclic signal when averaged over 20 cycles (left). Analytic trace, composed of identified PSCs that are analyzed in this article, also contained a cyclic signal (middle). Residual current trace also contained a cyclic signal, showing that we detected only a part of the synaptic events that the cell received.
View larger version (33K):
[in a new window]
FIG. 3.
A: detected PSCs that together give a cyclic signal (left) can be separated into a signal from excitatory (EPSCs; middle) and inhibitory (IPSCs; right). Magnitude of the signal was estimated by subtracting the total charge during the second half of the cycle (marked b) from the charge in the first half (marked a). B: calculation of the contribution from excitatory and inhibitory charge is illustrated with an idealized cycle where the most excitatory half (b) is subtracted from the most inhibitory half (a). This difference is equal to the charge difference given by IPSCs plus the charge difference given by EPSCs [IPSC(a-b) + EPSCs(a-b)]. This sum is normalized to 1.0 to compare different experiments. C: normalized cyclic signal is divided into a contribution from EPSCs ( 
) and IPSCs (
). Most cells received a contribution from both IPSCs and EPSCs.
and ). The second possibility, that EPSCs and IPSCs both contributed to the cyclic signal was seen in most interneurons (cells 5-14) and is indicated in the figure by a combination of white and black bars. The third possibility, a simultaneous increase in EPSCs and IPSCs, also was present. This is seen as a negative signal contribution from IPSCs (cells 1 and 2) or EPSCs (cell 16). For cells 1 and 2, this means that there was more inhibitory charge in the second compared with the first half of the cycle. Because inhibitory charge is positive, a-b becomes negative when there is more inhibitory charge in b compared with a. There was, however, a much larger excitatory charge difference in cells 1 and 2, with more excitatory charge in b compared with a, making the sum of the two signals positive (and equal to 1.0 because of the normalization). For cell 16, the negative EPSC contribution means that there was less excitatory charge in the second compared with the first half of the cycle. Because excitatory charge is negative, a-b becomes negative when there is less excitatory charge in b compared with a. The main signal, however, was dominated by the inhibitory charge difference, and the EPSCs resulted only in a reduction of the IPSC-mediated signal. We can conclude from the above analysis that most rhythmically active interneurons received cyclic information through a reciprocal regulation of the activity of both excitatory and inhibitory inputs.
View larger version (30K):
[in a new window]
FIG. 4.
A: cyclic signal given by EPSCs (top histograms) can be divided into a contribution from modulation of the average PSC size (charge) multiplied by their frequency. Same is done for the IPSCs (bottom histograms). During the average cycle, there is no detectable modulation of the average size but an obvious modulation of the frequency of PSCs. B: relative change in average charge and frequency from the first halves to the second halves of the cycles was estimated for all experiments. The results are given as a circle around the mean with a vertical line giving the standard error. 
, differences from 0 with P < 0.05. For the frequency estimates, the vertical line gives the standard deviation in a binomial population with N equal to the sample size. In those cases, where either EPSCs or IPSCs did not give a significant contribution to the signal, the cell is marked only with a vertical line on the unit line. We can see a more obvious modulation of the frequency of PSCs than of their average size.
DISCUSSION
Abstract
Introduction
Methods
Results
Discussion
References
; Kjaerulff and Kiehn 1996; Kjaerulff et al. 1994). The cyclic modulations of interneuron membrane potential correlated with the locomotor associated ventral root activity. The neurons, therefore, received synaptic information from the network generating the rhythm and therefore may have been a part of a locomotor CPG.
65 mV compared with 45 mV (see METHODS and Raastad et al. 1996); and 3) the detected events formed a cyclic signal that was locked to the root activity, which would not be expected from noise or other electrical processes in nerve cells. This last argument strongly suggests that most PSCs appeared in response to presynaptic spikes and not because of spontaneous vesicular release, which would not be expected to build a regular signal.
), is similar to how the rhythmic network drives the motoneurons in lamprey (Russel and Wallen 1983) and mammals (Cazalets et al. 1995; Jordan 1983; Perret 1983).
; Kjaerulff et al. 1994) and also seems to be necessary for the generation of the rhythm (Kjaerulff and Kiehn 1996), they actually may be a part of the mammalian locomotor CPG. It is therefore interesting to compare our findings to investigations of and theories for CPG function. Many motor CPGs in invertebrates are dominated by rhythmic synaptic influence from IPSCs (Marder and Calabrese 1996). In these animals, the basic building blocks of the CPGs are half-center oscillators based on reciprocal inhibitory connections. A similar mutually inhibitory half-center organization has been found in mammalian CPGs generating respiration (Bianchi et al. 1995) and has been proposed for the kernel of the spinal network generating mammalian locomotion (Brown 1911; Jankowska et al. 1967; Pearson and Collins 1993; see Kiehn et al. 1997 for a review). Theoretical studies also show that, in a half-center model, reciprocally inhibitory connections in combination with tonic excitation and/or intrinsic membrane properties are sufficient to generate both alternating and synchronous rhythmic activity (Perkel and Mulloney 1974; Sharp et al. 1996; Wang and Rinzel 1992).
; Gulyàs et al. 1993; Hessler et al. 1993; Rosenmund et al. 1993), the increased frequencies of PSCs simply could be due to an increase in release probability. If the increased frequencies were the result of increased presynaptic spike activity, the additional PSCs could come from the same cells that caused PSCs during low frequencies, or they may come from cells that were active only during the high-activity half of the cycles. This is an important distinction for the identification of cells that contribute to the rhythmic behavior, and this could be illuminated by studying spike frequency modification in extracellular single-unit recordings during locomotion.
; Raastad et al. 1992). A large stochastic variability in the signal from the individual cells, as seen, would reduce the opportunity to use the different amplitudes to code a signal. A stochastic amplitude variability at the individual synapses therefore could be one reason why the large amplitude range is not used efficiently to code the cyclic signal.
) were needed to account for the cyclic signal. With the opportunity to break down the cyclic locomotor-related signal into some of its synaptic elements, as we have described in this article, the preparation is well suited to study how the synaptic code is translated into a spike code.
| |
ACKNOWLEDGEMENTS |
|---|
This work was supported by a European Union Training and Mobility Grant (M. Raastad) and the Danish Research Academy (B. R. Johnson). The work in Ole Kiehn's lab was supported by the Danish Medical Research Council and the NOVO Foundation.
| |
FOOTNOTES |
|---|
Address for reprint requests: M. Raastad, Dept. of Physiology, Section of Neurophysiology, University of Copenhagen, Blegdamsvej 3, 2200 Copenhagen N, Denmark.
Received 7 May 1997; accepted in final form 25 June 1997.
| |
REFERENCES |
|---|
|
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
This article has been cited by other articles:
|
|
 |
|
  L. Ziskind-Conhaim, L. Wu, and E. P. Wiesner Persistent Sodium Current Contributes to Induced Voltage Oscillations in Locomotor-Related Hb9 Interneurons in the Mouse Spinal Cord J Neurophysiol, October 1, 2008; 100(4): 2254 - 2264. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. A. Hinckley, R. Hartley, L. Wu, A. Todd, and L. Ziskind-Conhaim Locomotor-Like Rhythms in a Genetically Distinct Cluster of Interneurons in the Mammalian Spinal Cord J Neurophysiol, March 1, 2005; 93(3): 1439 - 1449. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. J Whelan Developmental aspects of spinal locomotor function: insights from using the in vitro mouse spinal cord preparation J. Physiol., December 15, 2003; 553(3): 695 - 706. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. J. B. Butt, R. M. Harris-Warrick, and O. Kiehn Firing Properties of Identified Interneuron Populations in the Mammalian Hindlimb Central Pattern Generator J. Neurosci., November 15, 2002; 22(22): 9961 - 9971. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M.-C. Perreault Motoneurons Have Different Membrane Resistance during Fictive Scratching and Weight Support J. Neurosci., September 15, 2002; 22(18): 8259 - 8265. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Demir, B.-X. Gao, M. B. Jackson, and L. Ziskind-Conhaim Interactions Between Multiple Rhythm Generators Produce Complex Patterns of Oscillation in the Developing Rat Spinal Cord J Neurophysiol, February 1, 2002; 87(2): 1094 - 1105. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
U. Beierholm, C. D. Nielsen, J. Ryge, P. Alstrom, and O. Kiehn Characterization of Reliability of Spike Timing in Spinal Interneurons During Oscillating Inputs J Neurophysiol, October 1, 2001; 86(4): 1858 - 1868. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Raastad and O. Kiehn Spike Coding During Locomotor Network Activity in Ventrally Located Neurons in the Isolated Spinal Cord From Neonatal Rat J Neurophysiol, May 1, 2000; 83(5): 2825 - 2834. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Ballerini, M. Galante, M. Grandolfo, and A. Nistri Generation of rhythmic patterns of activity by ventral interneurones in rat organotypic spinal slice culture J. Physiol., June 1, 1999; 517(2): 459 - 475. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Raastad, M. Enriquez-Denton, and O. Kiehn Synaptic signaling in an active central network only moderately changes passive membrane properties PNAS, August 18, 1998; 95(17): 10251 - 10256. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
