| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Mammalian Locomotor Laboratory, Department of Neuroscience, Karolinska Institutet, Stockholm, Sweden
Submitted 2 July 2008; accepted in final form 23 September 2008
 |
ABSTRACT |
|---|
|
|
|
INTRODUCTION |
|---|
|
; Robertson and Stein 1988) and walking (Burke et al. 2001; Cazalets et al. 1996; Hochman and Schmidt 1998; Jordan 1983; Orsal et al. 1986; Schmidt 1994; Schmidt et al. 1988; Shefchyk and Jordan 1985a,b), similar to what has been described for nonlimbed swimming CPGs (Kahn 1982; Russell and Wallen 1983; Wallen et al. 1993). This push–pull organization of synaptic inputs to MNs during rhythmic activity has led to the notion that spinal premotor CPG networks are organized in reciprocally coupled flexor and extensor modules/units that, when active in their appropriate phase, provide excitation to agonist MNs and inhibition to antagonist MNs in a symmetric manner (see Grillner 1981; Jordan 1991; Kiehn 2006; McCrea and Rybak 2008). However, some studies have demonstrated a predominance of inhibitory drive to certain groups of MNs during locomotion (Hochman and Schmidt 1998; Schmidt et al. 1988; Shefchyk and Jordan 1985a) and a possible difference in the relative importance of excitation and inhibition in flexor and extensor MNs (Godderz et al. 1990; see their Figs. 9 and 10), suggesting a more asymmetrical CPG output. Quantitative information of the relative contribution of excitatory and inhibitory synaptic inputs in different MN pools, on the other hand, is lacking and needed to evaluate the possible scale and extent of an asymmetrical output and the consequences for the CPG organization. Moreover, a recent study in in vitro turtle spinal cord preparations reported that excitatory and inhibitory conductances in hindlimb MNs during rhythmic scratch-like network activity varied in-phase with the peak at the active phase of the scratch episodes (Berg et al. 2007). Detection of the balanced excitation and inhibition was conditional on quantifying the phase-related inhibitory and excitatory synaptic conductances and has not been described before in any other locomotor studies. This mode of operation is incompatible with the general concept of a reciprocal organization of flexor-related and extensor-related CPG networks.
In the present study we provide a quantitative assessment of the weighting and phase tuning of synaptic inputs to different flexor and extensor MNs during locomotor-like activity in the isolated neonatal mice lumbar spinal cord preparation. This preparation has become a model system for mammalian locomotor experiments (Bonnot et al. 2002; Cazalets and Bertrand 2000; Kiehn 2006; Kiehn et al. 2000; Nishimaru et al. 2000; Schmidt and Jordan 2000; Whelan 2003). We applied the same conductance measurement and decomposition methods (Anderson et al. 2000; Borg-Graham et al. 1998; Marino et al. 2005; Shu et al. 2003) as used in the turtle preparation (Berg et al. 2007) during drug-induced locomotor-like activity. Our results reveal a highly asymmetrical operation of reciprocally organized flexor and extensor CPG modules, with dominance of inhibitory conductances. The asymmetrical operation remained under different modes of operation of the locomotor network. Our data also showed a difference in the operation of distinct CPG modules in the rostral and caudal lumbar segments and provided evidence for a layered architecture of the rodent locomotor CPG.
Parts of this study were previously presented in abstract form (Endo and Kiehn 2007).
|
|
METHODS |
|---|
|
All procedures followed Swedish federal guidelines for animal care and were approved by the local Animal Care and Use Committee. Newborn C57BL6 mice (postnatal days 0–4) were anesthetized with isoflurane and the spinal cords were removed as described previously (Butt et al. 2005; Kiehn and Kjaerulff 1996). The preparations were placed in a recording chamber mounted in an upright microscope and continuously perfused with Ringer solution containing (in mM): 111.14 NaCl, 3.09 KCl, 1.1 KH2PO4, 1.26 MgSO4, 2.52 CaCl2, 25 NaHCO3, and 11.1 glucose (pH 7.4, bubbled with 95% O2-5% CO2). Locomotor-like activities were induced by bath application of N-methyl-D-aspartate (NMDA, 5–10 µM), in combination with serotonin (5-HT, 5–10 µM), and recorded with glass suction electrodes as ventral root (VR) activity in lumbar segments (L) 2, 3, or 5. Dopamine (25–50 µM) was also applied in addition to NMDA and 5-HT in some experiments. The main bursts in the L2 and L3 VRs reflect flexor activity, whereas the L5 VR activity reflects extensor activity (Whelan et al. 2000). The recorded signals were amplified x20,000 and bandwidth-filtered at 100–1,000 Hz using a custom-made amplifier. Whole cell patch-clamp recordings were obtained from visually patched motor neurons (Nishimaru et al. 2005) located in the same segments as the recorded VRs. Patch pipettes were filled with the internal solution, containing (in mM): 138 K-gluconate, 10 HEPES, 5 ATP-Mg, 0.3 GTP-Li, 0.0001 CaCl2, adjusted to pH 7.3 with KOH. Cesium chloride (5 mM) and N-(2,6-dimethylphenyl carbamoylmethyl)triethylammonium bromide (QX-314, 5–10 mM) were added to the solution to reduce intrinsic K+ and Na+ conductances, respectively. MNs were identified by antidromic activation from the ventral roots before QX-314 diffused enough to block action potentials. The resistance of the pipettes was in the range of 3–7 M
. The access resistance was in the range of 9–20 M during recordings and continuously corrected by balancing the bridge in current-clamp recordings. In voltage-clamp recordings, the series resistance was compensated by about 70%. The signals were recorded with a Multiclamp 700A patch-clamp amplifier (Molecular Devices, Palo Alto, CA). All recordings were performed at room temperature (20–23°C). Only recordings where the MN membrane potential, the access resistance, and the locomotor-like activity remained stable throughout the recording period were included in the analysis.
Data analysis
The methods for conductance measurement and decomposition were basically the same as recently used in turtle scratching preparations by Berg et al. (2007) (Anderson et al. 2000; Borg-Graham et al. 1998; Marino et al. 2005; Shu et al. 2003).
To perform the conductance measurements, MNs were recorded during locomotor-like activity (Fig. 1 A). Individual locomotor cycles were divided into flexor phase and extensor phase by using the half-amplitude in rectified and smoothed VR signals as onset and offset reference points. Both phases were then subdivided into five bins of equal length (Fig. 1B). Phases 1–5 therefore correspond to the flexor phase (active phase of L2 and L3; inactive phase of L5) and phases 6–10 correspond to the extensor phase (inactive phase of L2 and L3; active phase of L5). Only locomotor periods where the cycle period varied <10% throughout the measurements were used for analysis. The mean membrane potential in each bin for the recorded MN was then obtained by averaging values from 5 to 10 locomotor cycles. This procedure was performed for recordings obtained at two or three different injection current levels and the total conductance was calculated for each bin as the slope of the current–voltage relationship obtained by linear regression (Fig. 1C). The decomposition of the conductances was based on the following equations
 | (1) |
 | (2) |
; Hochman and Schmidt 1998) due to the electrode solution used in the present experiments. Ee was set to the commonly accepted value of 0 mV. Assuming that Gleak and Eleak were time- and voltage independent, these parameters were set to arbitrary constant values (see following text). Although we could easily determine the actual values for Gleak and Eleak in the nonlocomoting preparations, we have no way to determine these values during locomotor-like activity. Consequently, we could not determine the absolute values of Ge and Gi. We instead calculated the relative changes of Ge (
Ge) and Gi (Gi) as a percentage of the mean value of Gtotal (mean Gtotal) averaged over the locomotor cycle. In this calculation, the shape and amplitude of Ge and Gi profiles are insensitive to the precise value of Eleak and Gleak. The membrane potential was corrected by a liquid junction potential of –10 mV.
|
G divided by maximum amplitude of G, G/max G) was plotted against the locomotor phase in polar coordinates for individual motor neurons. To evaluate mean phase and sharpness of the phase tuning of the conductance profiles, all vectors were summed and normalized by the sum of the length of all vectors. The direction of the resultant vector indicates the mean active phase angle of the conductance. The length of the resultant vector represents the sharpness of the phase tuning. The resultant length of a conductance that increases only in one phase bin will be 1, whereas that of a conductance that increases equally in symmetrically distributed bins will be 0. Summary statistic values are given as means ± SE. Paired and unpaired t-tests, one-way ANOVA, and Games–Howell test were used to assess statistical significance. Differences were considered to be significant at P < 0.05.
Drugs
NMDA, 5-HT, 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX), strychnine hydrochloride, picrotoxin, and dopamine were purchased from Sigma (St. Louis, MO). Tetrodotoxin (TTX) was purchased from Tocris Cookson (Bristol, UK). QX-314 was purchased from RBI (Natick, MA).
|
|
RESULTS |
|---|
|
Whole cell patch-clamp recordings were obtained from a total of 40 lumbar motor neurons (MNs). Locomotor-like activity was induced by bath application of a combination of NMDA and 5-HT (5–10 µM) with (n = 6 MNs) or without (n = 34) dopamine (25–50 µM), and monitored by recording the ventral root activity in lumbar ventral root L2/L3, representing flexor activity, or L5, representing extensor activity (Whelan et al. 2000). MNs were classified as flexor or extensor MNs based on the phase during which the membrane potential (Vm) was most depolarized. Twenty-two MNs were classified as flexors and 18 as extensors. All flexor MNs (n = 19 without dopamine; n = 3 with dopamine) and 6 extensor MNs (all without dopamine) were recorded in the L2/L3 segments, whereas 12 extensor MNs (n = 9 without dopamine; n = 3 with dopamine) were recorded in the L5 segment. None of the recorded MNs in L5 showed out-of-phase oscillations with the L5 root bursts.
The mean locomotor cycle length in all experiments was 2.4 ± 0.1 s, with no significant difference between the data groups (one-way ANOVA, P = 0.17).
Measurements of conductance changes
To derive the total conductance (Gtotal) profile over the locomotor cycle the average Vm was calculated for recordings at two or three different injection current (Iinj) levels in normalized cycles where bins 1–5 define the flexor phase and bins 6–10 the extensor phase (Fig. 1, A and B). The Gtotal-phase trace was derived from the slope of the liner regression line for each phase bin. We calculated and derived the variation of total, excitatory, and inhibitory conductances (Gtotal, Ge, and Gi, respectively) from the Gtotal trace as a percentage of the mean value of the Gtotal (mean Gtotal) averaged over the locomotor cycle, assuming that Gtotal is composed of Ge, Gi, and a constant leak conductance, Gleak (Fig. 1C; see METHODS for detail). To ensure the assumption of the constant Gleak, we used a pipette solution containing Cs+ (5 mM) and QX-314 (5–10 mM) to reduce intrinsic K+ and Na+ conductances, and kept Vm below approximately –60 mV to avoid activations of voltage-dependent conductances. Since locomotor-inducing drugs—NMDA, 5-HT, and dopamine—alone can evoke nonlinear conductances, we examined current responses to voltage ramps (30 mV/s) after the locomotor-like activity was blocked by the additional application of TTX (0.5 µM) or a cocktail of CNQX (20 µM), picrotoxin (10 µM), and strychnine (1 µM). We confirmed that the net current responses could be safely regarded as linear in the voltage range recorded (n = 6 without dopamine; n = 1 with dopamine; Fig. 2).
|
We first describe the general features of conductance changes in flexor and extensor MNs during locomotor activity induced by 5-HT and NMDA without dopamine.
Figure 1 shows an example of a recording from a flexor MN in L2, with rhythmic Vm depolarizations in-phase with the L2 VR burst (Fig. 1, A and B). Gtotal varied moderately throughout the locomotor cycle, whereas the calculated Ge and Gi traces showed clear rhythmic variations (Fig. 1C). Ge increased during the flexor phase when the MN was depolarized, whereas Gi increased in the extensor phase when the MN was hyperpolarized, indicating that this flexor MN received increased excitatory synaptic drive during the flexor phase and increased inhibitory synaptic inputs during the extensor phase. As a consequence of the alternating excitation and inhibition, the total conductance was only moderately varied over the locomotor cycle.
Figure 3 shows an example of an extensor MN recorded in the L5 segment displaying rhythmic depolarizations in-phase with the main burst activity in the L5 ventral root (Fig. 3, A and B). In this extensor MN the peak-to-trough amplitude of the Vm oscillations increased with depolarization in the voltage window investigated (compare –70 and –50 mV; Fig. 3A). Moreover, depolarizing voltage deflections became obvious in the flexor phase when the Vm was hyperpolarized to a level below the equilibrium potential for Cl– (arrows in Fig. 3B, bottom trace). Together these observations suggest that the Vm oscillations were dominated by rhythmic inhibitory synaptic inputs arriving in the inactive flexor phase. The extracted Ge and Gi traces clearly showed that this was the case (Fig. 3C). Although Ge and Gi showed an out-of-phase relationship with a peak of Ge in the extensor phase and a peak of Gi in the flexor phase (reciprocal to the pattern of activity seen for flexor MNs), the amplitude of Ge was only less than one fifth of that of Gi. In contrast to the flexor MN described in Fig. 1, the total conductance in this extensor MN showed a marked rhythmic variation, which peaked during the inactive flexor phase.
|
G-phase traces of all of the recorded MNs are shown in Fig. 4, A and B (top row of panels) and their normalized conductance profiles, G divided by the amplitude of the G (max G), are shown in Fig. 4, A and B (bottom row of panels). Data are divided into flexor (Fig. 4A) and extensor (Fig. 4B) MNs and data obtained in the absence and the presence of dopamine are shown in blue and red, respectively (data in the presence of dopamine are described in a following section). Before drug application the resting conductance measured from responses to small-amplitude current pulses (<200 pA, 1 s) was 22.4 ± 2.7 nS for the L2/L3 flexor MNs, 21.0 ± 3.2 nS for the L2/L3 extensor MNs, and 24.9 ± 1.9 nS for L5 extensor MNs. During locomotor activity the mean Gtotal over the locomotor cycle was 17.9 ± 1.3 nS for the L2/L3 flexor MNs, 22.7 ± 3.2 nS for the L2/L3 extensor MNs, and 24.2 ± 2.5 for L5 extensor MNs. The mean Gtotal during locomotion was significantly smaller than the resting conductance in L2/L3 flexor MNs (86.4 ± 4.5% of the resting conductance, paired t-test, P < 0.05), whereas the extensor MNs did not show significant differences between resting and locomotion, either for L2/L3 MNs nor for L5 MNs (113.2 ± 12.5% for L2/L3 MNs and 97.6 ± 8.7% for L5 MNs, P > 0.05). As can be seen from the simple average trace of normalized Gtotal (dashed line in Fig. 4A, bottom left) the flexor MNs showed little modulation of the total conductance over the locomotor cycle. However, there was a moderately lower Gtotal during their active phase (the flexor phase) than that during the inactive extensor phase. In extensor MNs, the lower conductance state during the active phase was more pronounced than that in the flexor MNs (dashed line in Fig. 4B, bottom left). The amplitude of the Gtotal was 18.4 ± 2.4% of the mean Gtotal value for the L2/L3 flexor MNs, 19.5 ± 4.3% for the L2/L3 extensor MNs, and 34.2 ± 5.1% for the L5 extensor MNs. The amplitude of the Gtotal in L5 extensors appeared to be larger than that in the other MNs (the difference was statistically significant between the L2/L3 flexors and the L5 extensors, Games–Howell test, P < 0.05). These results shows that rodent MNs during 5-HT/NMDA–evoked locomotor-like activity have a lower conductance state during their active phase compared with that during their inactive phase.
|
Ge of individual MNs was plotted against the amplitude of Gi (blue symbols). The amplitude of Gi was in all cases larger than that of Ge. For L2/L3 flexor MNs the mean value of Ge amplitude was 11.4 ± 1.2% of the mean Gtotal and the mean value of Gi amplitude was 24.8 ± 3.2% of the mean Gtotal. For L2/L3 extensor MNs, the corresponding values were 2.6 ± 0.6% (Ge) and 20.7 ± 4.4% (Gi) and for L5 extensor MNs values were 5.4 ± 0.7% (Ge) and 35.9 ± 4.9% (Gi). The differences between Ge and Gi were statistically significant in all groups (paired t-test, P < 0.0002 for the L2/L3 flexor MNs and the L5 extensor MNs, P < 0.01 for the L2/L3 extensors). The Ge to Gi ratios were 0.49 ± 0.04 for the L2/L3 flexor MNs, 0.17 ± 0.05 for the L2/L3 extensor MNs, and 0.18 ± 0.06 for the L5 extensor MNs. Thus the dominance of inhibition is more pronounced in the extensor than in the flexor MNs. These calculations clearly demonstrate an asymmetrical conductance change during the locomotor cycle with a dominance of inhibition in both flexors and extensor MNs that was extended to include both rostral and caudal MNs.
|
Ge and Gi (R = 0.76), suggesting that the two conductances at the MN level are regulated in parallel over a wide range of strength of motor outputs. For extensor MNs this correlation is not significant because Ge values remain relatively constant for a large range of Gi. Moreover, we did not find a significant correlation between the G amplitudes or the Ge to Gi ratio and the locomotor cycle period (|R| <0.521; Fig. 6). These findings suggest that the asymmetrical conductance changes are robust phenomena during transmitter-induced locomotion and found over a large range of network outputs.
|
Inspection of the G profiles shown in Fig. 4 also revealed that there seemed to be a difference in the phase tuning of G curves: the profile of Gi in flexor MNs and Ge in extensor MNs appeared to be broader and more erratic than the Ge in flexor MNs and Gi in extensor MNs. To evaluate these characteristics in more detail, the preferred locomotor phase and the sharpness of the conductance-phase profile of excitation and inhibition were analyzed.
The normalized Ge and Gi were plotted against the locomotor phase in polar coordinates (Fig. 7, A and B, top rows, blue lines); we then calculated the mean vector for each MN (Fig. 7, A and B, bottom rows, blue symbols). The angle of the mean vector represents the preferred active phase and the length represents the sharpness of G-locomotor-phase profile, for each conductance. The preferred locomotor phases of Ge and Gi are plotted for flexor MNs and extensor MNs in Fig. 7C. This plot highlights the timing of the conductance changes as described earlier. In all of the flexor MNs and most (13/15) of the extensor MNs, the preferred phase of conductance change clearly alternated with increased excitation in the active phase and increased inhibition in the inactive phase. Only two L5 extensor MNs showed a concurrent increase of excitation and inhibition with the peaks in the inactive flexor phase. Thus the general feature of the conductance change is excitation alternating with inhibition.
|
Ge was significantly longer than that for Gi in L2/L3 flexor MNs (0.53 ± 0.01 vs. 0.40 ± 0.03, paired t-test, P < 0.0001; Fig. 7D). In contrast, the vector length for Ge in L2/L3 extensor MNs was significantly shorter than that for Gi (0.29 ± 0.05 vs. 0.49 ± 0.04, paired t-test, P < 0.05; Fig. 7D). These results indicate that in L2/L3 the excitation in flexor MNs and the inhibition in extensor MNs, which increase timewise simultaneously during the flexor phase, had a sharper conductance-phase profile than that of conductances that increased during the extensor phase. On the other hand, Ge and Gi in L5 extensors had similar average vector lengths (0.48 ± 0.04 vs. 0.56 ± 0.03, paired t-test, P > 0.1; Fig. 7D). There was also a significant difference between Ge values in extensors in L2/L3 and L5 (unpaired t-test, P < 0.05). These findings suggest that conductances in flexor and extensor MNs in L2/L3 are modulated in a modular fashion and different from those in L5. Asymmetrical conductance profiles are preserved in the presence of dopamine
One possibility for the observed asymmetrical conductance changes is that 5-HT and NMDA bias the network operation in a skewed way, in particular because 5-HT has been reported to have stronger excitatory effects on extensor- than on flexor-related motor circuits in the mammalian spinal cord (see Conway et al. 1988; Schmidt and Jordan 2000; Vult von Steyern and Lømo 2005). To test whether similar profiles are observed when locomotor-like activity is produced by drug combinations that may bias the CPG circuits different from 5-HT we used a cocktail of 5-HT/NMDA in combination with dopamine (25–50 µM) in a small number of experiments (n = 3 flexor MNs in L2; n = 3 extensor MNs in L5) to evoke rhythmic activity. This cocktail is commonly used to induce locomotor-like activity in the rodent (Jiang et al. 1999) and 5-HT and adrenergic modulation is also known to have opposite effects on inhibition in vertebrate CPGs (McDearmid et al. 1997; McLean et al. 2000).
The MNs showed profiles of excitation and inhibition parallel with those in the absence of dopamine. The excitation and inhibition varied out of phase with each other with increased excitation during the active phase and increased inhibition during the inactive phase in all MNs (Figs. 4 and 7C, red lines). The amplitude of Gi (39.4 ± 10.4% of mean Gtotal for L2 flexors MNs; 32.2 ± 2.8% for L5 extensor MNs) was larger than that of Ge (15.1 ± 3.1% for L2 flexor MNs; Ge = 6.2 ± 1.0% for L5 extensor MNs) in all MNs and the dominance of inhibition was most pronounced in extensor MNs (Ge to Gi ratios: 0.40 ± 0.04 for L2 flexor MNs, 0.20 ± 0.5 for L5 extensor MNs; Fig. 5, red symbols). Similarly, there were no obvious differences in the phase tunings [mean vector length: 0.55 ± 0.01 (Ge) and 0.41 ± 0.05 (Ge) for L2 flexor MNs; 0.50 ± 0.03 (Ge) and 0.56 ± 0.07 (Ge) for L5 extensor MNs] from those in the absence of dopamine (Fig. 7, A and B, red). These results suggest that the asymmetrical conductance changes reflect a robust network property for drug-induced locomotion.
|
|
DISCUSSION |
|---|
|
Methodological considerations
The present conductance measurement and decomposition method has previously been used to evaluate the network activity converging onto identified neuron classes in the brain and spinal cord (Anderson et al. 2000; Berg et al. 2007; Borg-Graham et al. 1998; Marino et al. 2005; Shu et al. 2003). The general assumption is that the total conductance consists of excitatory and inhibitory synaptic conductances and a leak conductance, which includes all other conductances. The basic requirement is that no active conductances are activated during the voltage oscillations. In the present study, we have attempted to fulfill this requirement in two ways. First, voltage-dependent potassium and sodium conductances, including the persistent sodium conductance that is known to be expressed in rodent MNs (Kuo et al. 2006; Li and Bennett 2003), were reduced by intracellular Cs+ and QX-314, respectively. Second, recordings were performed at relatively hyperpolarized levels to avoid influences from voltage-dependent conductances including NMDA receptors. Under these conditions, we confirmed that the current–voltage relationship was linear in the presence of locomotion-inducing drugs in the voltage range where the measurements were performed. This also indicated that subthreshold conductances like the transient calcium conductance or hyperpolarization-activated inward cation conductance that are known to be present in rodent motor neurons and to be facilitated by 5-HT (Berger and Takahashi 1990; Kjaerulff and Kiehn 2001; Larkman and Kelly 1997) did not cause a significant nonlinearity. Moreover, as previously demonstrated Ih will mainly contribute with a leak conductance during locomotion and thus will not be modulated in a time-varying way (Kiehn et al. 2000).
We had to set unknown parameters that appear in Eqs. 1 and 2 (see METHODS). We could determine realistic values for the reversal potentials of excitatory and inhibitory synaptic conductances. However, it was impossible to isolate and measure the leak conductance and its reversal potential during the network activity in our experimental conditions (see also Berg et al. 2007). These limitations thus did not allow us to calculate the actual values of the synaptic conductances. Instead, we calculated only the locomotor-phase–related changes of the excitation and inhibition: Ge and Gi, respectively. Ge and Gi do not depend on the precise estimation of the leak-related conductances as long as these are constant throughout the measurement (cf. Berg et al. 2007). Although some voltage-dependent conductances might be enhanced or depressed during the network activity, we believe that these conductances were not significant factors in the present experimental conditions as discussed earlier.
Another possible source of error is the cable properties of MNs. It is possible that a relatively distant location from soma of the excitatory synaptic inputs compared with inhibitory synaptic inputs caused the smaller relative amplitude of Ge than Gi. Nevertheless, the present results obviously reflect a relative impact of the excitation and inhibition on the somatic membrane potential and the output of the MNs.
Activity of motor neurons is regulated by alternating synaptic inputs in a push–pull manner with dominant inhibitory conductance
The synaptic conductance profiles extracted from the rhythmic signals recorded during drug-induced locomotor activity showed in the large majority of MNs alternation between excitation and inhibition. Excitatory conductance peaked in the depolarizing phase, whereas the inhibitory conductance peaked in the hyperpolarizing phase (push–pull). These results confirm a large number of previous in vivo and in vitro studies in the vertebrate spinal cord on rhythmic MN oscillations during walking (Burke et al. 2001; Cazalets et al. 1996; Hochman and Schmidt 1998; Jordan 1983; Orsal et al. 1986; Schmidt 1994; Schmidt et al. 1988; Shefchyk and Jordan 1985a,b), swimming (Kahn 1982; Russell and Wallen 1983; Wallen et al. 1993), and scratching (Perreault 2002; Robertson and Stein 1988). The exception to this general notion of a push–pull organization is a series of recent studies by Hounsgaard and colleagues, which demonstrated a balanced variation of opposing synaptic conductances during the depolarizing phase in hindlimb MNs recorded during scratching movements (Alaburda et al. 2005; Berg et al. 2007). The balanced excitation and inhibition lead to a remarkably high total conductance state during the active phase of MN firing. In the present study, we find signs of balanced variation of opposing synaptic conductances only in two MNs. Moreover, in accordance with previous reports (Burke et al. 2001; Cazalets et al. 1996; Jordan 1983, 1991; Perreault 2002; Schmidt 1994; Schmidt et al. 1988; Shefchyk and Jordan 1985b), total conductance was generally unchanged or moderately lower in the hyperpolarizing phase in the present experiments. We do not have a clear explanation for the discrepancies between the in vitro turtle scratch experiments and other experiments. One possible explanation is that recordings in the turtle were performed from MNs located near the transversely cut end of the cord and that some parts of the pre-MN network are therefore missing.
The synaptic conductance measurements allowed a direct quantitative analysis of time-varying synaptic conductances and in particular to investigate whether the conductance profiles were equal in flexors and extensors. Previous studies have suggested a predominance of inhibitory drive to some mammalian MNs during locomotion (Hochman and Schmidt 1998; Schmidt et al. 1988; Shefchyk and Jordan 1985a). Here, we find that in all MNs the time-varying conductance changes were asymmetric in amplitude, with inhibition dominating over excitation. Furthermore, this inhibitory dominance was largest in extensors (both L2/L3 and L5) where the average excitatory conductance changes range to only 15–20% of the inhibitory conductance changes. In flexors Gi was about twice the amplitude of Ge. Our findings thus show that last-order rhythmic excitatory inputs from the CPG network to MNs are relatively small compared with inhibition and suggest that rhythmicity of extensor MNs is more susceptible to changes in central excitatory drive than flexor MNs. Some indications of this come from findings in cats showing that spontaneous failures of burst of activity in MN firing—so-called deletions—occur more frequently in extensor than in flexor MN pools (Duysens 1977; Lafreniere-Roula and McCrea 2005). Such deletions are rare when sensory input is intact, possibly because a greater part of the amplitude modulation in extensor MNs comes from extensor group I afferent feedback, whereas flexor-directed feedback is more variable (Hiebert and Pearson 1999; Hultborn 2001; McCrea 2001; Nielsen and Sinkjaer 2002; Pearson 2004). The predominance of inhibition also suggests that mammalian MN rhythmicity is regulated around a tonic excitation that will increase the importance of the rhythmic inhibitory signals (see Jordan 1983; Roberts and Tunstall 1990).
Tuning of synaptic conductances reveal modular organization of CPG with tighter rhythm-generating capabilities in flexor modules
The analysis of the phase tuning of the conductance changes in L2/L3 flexor and extensor MNs showed a similarity in the sharpness of phase tuning between excitatory and inhibitory synaptic drives which increased in phase (e.g., excitation of flexors MNs and inhibition of extensor MNs), and a significant difference in phase tuning between synaptic drives, which increased out of phase. These observations support the notion that pre-MN excitatory and inhibitory interneuronal circuits work as cooperating modules/units (see Jordan 1991; Stein 2005), which convey reciprocal activities to flexor and extensor MNs with a common fingerprint.
In contrast to L2/L3, the difference in the phase tuning in L5 extensor MNs was not obvious and the tuning of excitation L5 extensor MNs was different from that of L2/L3 extensor MNs, indicating that MNs in rostral and caudal lumbar segments are under the control of functionally distinct CPG networks, respectively (see also Tresch and Kiehn 1999). In the mouse, MNs located in L5 may belong to the intrinsic foot muscles, whereas MNs located in L2/L3 innervate the hip and knee (McHanwell and Biscoe 1981). Our results are thus compatible with results from the cat where it has been shown that MNs innervating the intrinsic foot muscles, located in caudal lumbar segments, receive different central drive than rostrally located extensor MNs (Burke et al. 2001). In future studies it will obviously be of great interest to be able to perform the conductance measurements in muscle-identified groups of MNs.
The extensor-related synaptic drives (excitation of extensors MNs and inhibition of flexor MNs) were in all cases loosely tuned to the locomotor phase and/or having weaker rhythmic modulation, compared with flexor-related drives (excitation of flexor MNs and inhibition of extensor MNs) that were more sharply tuned and displayed larger rhythmic modulations. These and previous observations (see e.g., Duysens 1977; Lafreniere-Roula and McCrea 2005) imply that the flexor-related premotor CPG network operates more robustly and stably with tighter rhythm-generating capabilities compared with the extensor-related CPG modules. The flexor-related network modules may therefore have a leading role in the operation of locomotor CPG in the spinal cord. However, our observations do not imply that the extensor parts of the CPG are collaterals to rhythmic flexor modules (Brownstone and Wilson 2008; Pearson and Duysens 1976) and are incapable of generating rhythmic activity. In fact, the push–pull nature of the synaptic inputs observed in extensor MNs exclude the pure flexor-generator model (Pearson and Duysens 1976; see also the discussion in Shefchyk and Jordan 1985) and the observed differences in drive between L2/L3 and L5 extensor MNs also preclude an asymmetric model with a rhythmogenic flexor generator localized in the L1–L2 region, as proposed by Brownstone and Wilson (2008).
The asymmetries in CPG operation were observed when we used the two classical drug cocktails to induce rhythmic locomotor activity in rodents (Jiang et al. 1999): 5-HT in combination with NMDA or 5-HT/NMDA combined with dopamine. 5-HT is known to have a stronger excitatory effect on extensor than on flexor motor circuits in the mammalian spinal cord, whereas dopamine presumably has opposite effects (see Conway et al. 1988 and Schmidt and Jordan 2000 for references). Moreover, 5-HT decreases, whereas adrenergic modulation enhances, rhythmic inhibitions in vertebrate CPGs (McDearmid et al. 1997; McLean et al. 2000). However, in the two conditions, asymmetrical conductance changes and difference in synaptic weighting and tuning between flexors and extensors were preserved over the entire frequency range of locomotor outputs investigated. Therefore the conductance changes observed here appear to provide a robust measure of the rodent CPG output to MNs and to reflect the fundamental nature of CPG operation during drug-induced locomotor activity. We cannot, however, rule out the possibility that the asymmetric features are specific to drug-induced locomotion because we did not perform our analysis after neural-evoked locomotion. Nevertheless, the coordination of L2 and L5 activity and frequency of rhythmic output following activation of locomotor regions in the brain stem in rodents (Crone et al. 2008; Gilmore and Fedirchuk 2004; Liu and Jordan 2005; Zaporozhets et al. 2004) are similar to what was observed during drug-induced locomotion.
Consequences for organization of the walking CPG
The findings that the rhythmic last-order inputs to MNs are asymmetric with a dominance of inhibition are compatible with an organization where the number of last-order excitatory interneurons is considerably smaller than the number of last-order inhibitory interneurons both in the flexor and extensor modules/units. Moreover, we find no correlation between the amplitudes of Ge and Gi in extensor MNs and, although there was a correlation between Ge and Gi in flexor MNs, Ge amplitude seemed to saturate for larger values of Gi in flexor MNs as well. These latter observations indicate that the "gain" in the last-order excitatory and inhibitory interneuron pathways (e.g., those controlled by the flexor module/unit) can be controlled independently. This independent gain control may be explained by an organization where the last-order inhibitory interneurons are not downstream collaterals of excitatory last-order interneurons, as proposed in most models of the mammalian CPG models (Brownstone and Wilson 2008; Jordan 1991; Kriellaars et al. 1994; Lafreniere-Roula and McCrea 2005; Noga et al. 2003; Perret et al. 1988; however, see Burke et al. 2001). The pure reciprocal coupling at the last-order interneuron level proposed in previous models is presumably based on findings showing that in some cases of MN recordings a single synapse segmental delay can be observed between excitatory and inhibitory synaptic potentials following stimulation of the mesenchephalic locomotor region (MLR) (Jordan 1991; Noga et al. 2003; Shefchyk and Jordan 1985a). However, in many cases of MLR stimulation this strict single-synapse excitatory–inhibitory coupling is replaced by overlapping excitatory and inhibitory latencies or by pure excitatory or inhibitory inputs (Degtyarenko et al. 1998; Noga et al. 2003). To encompass the strict single-synapse excitatory–inhibitory coupling with independent gain control at the last-order level, we propose a model with populations of excitatory and inhibitory last-order interneurons driven by common upstream excitatory interneurons, which themselves also directly project onto MNs (Fig. 8). Our findings thus support the notion of a multilayer structure of the mammalian CPG consisting of at least two layers (Fig. 8), with the important addition of a skewed number of last-order inhibitory and excitatory interneuron and possibilities for increasing the inhibitory conductance in the face of a constant excitation.
|
; Kriellaars et al. 1994; Lafreniere-Roula and McCrea 2005). Although not shown in the figure our data also suggest that the locomotor CPG is composed of several flexor/extensor modules/units in the rostral and caudal lumbar spinal cord (see also Lafreniere-Roula and McCrea 2005; McCrea and Rybak 2008). Conclusion
The present quantitative conductance analysis in rodent motor neurons clearly shows that the structure and/or operation of flexor-related and extensor-related CPG modules of locomotion are reciprocally but asymmetrically organized. The data predict a layered structure of the rodent CPG with a skewed distribution of last-order inhibitory and excitatory interneurons placed under common control of excitatory second last-order neurons. The overall structure of the proposed model resembles the multilayered modular architecture proposed for locomotor CPGs in felines. The predictions of the model provide a solid basis for further anatomical and electrophysiological investigations of the cellular organization of the mammalian CPG.
|
|
GRANTS |
|---|
|
|
|
ACKNOWLEDGMENTS |
|---|
|
|
|
FOOTNOTES |
|---|
Address for reprint requests and other correspondence: O. Kiehn, Mammalian Locomotor Laboratory, Department of Neuroscience, Karolinska Institutet, Retzius väg 8, S-171 77 Stockholm, Sweden (E-mail: o.kiehn{at}ki.se)
|
|
REFERENCES |
|---|
|
Anderson JS, Carandini M, Ferster D. Orientation tuning of input conductance, excitation, and inhibition in cat primary visual cortex. J Neurophysiol 84: 909–926, 2000.
Berg RW, Alaburda A, Hounsgaard J. Balanced inhibition and excitation drive spike activity in spinal half-centers. Science 315: 390–393, 2007.
Berger AJ, Takahashi T. Serotonin enhances a low-voltage-activated calcium current in rat spinal motoneurons. J Neurosci 10: 1922–1928, 1990.[Abstract]
Bonnot A, Whelan PJ, Mentis GZ, O'Donovan MJ. Locomotor-like activity generated by the neonatal mouse spinal cord. Brain Res 40: 141–151, 2002.[CrossRef][Web of Science]
Borg-Graham LJ, Monier C, Fregnac Y. Visual input evokes transient and strong shunting inhibition in visual cortical neurons. Nature 393: 369–373, 1998.[CrossRef][Web of Science][Medline]
Brownstone RM, Wilson JM. Strategies for delineating spinal locomotor rhythm-generating networks and the possible role of Hb9 interneurones in rhythmogenesis. Brain Res Rev 57: 64–76, 2008.[CrossRef][Medline]
Burke RE, Degtyarenko AM, Simon ES. Patterns of locomotor drive to motoneurons and last-order interneurons: clues to the structure of the CPG. J Neurophysiol 86: 447–462, 2001.
Butt SJ, Lundfald L, Kiehn O. EphA4 defines a class of excitatory locomotor-related interneurons. Proc Natl Acad Sci USA 102: 14098–14103, 2005.
Cazalets JR, Bertrand S. Ubiquity of motor networks in the spinal cord of vertebrates. Brain Res Bull 53: 627–634, 2000.[CrossRef][Web of Science][Medline]
Cazalets JR, Borde M, Clarac F. The synaptic drive from the spinal locomotor network to motoneurons in the newborn rat. J Neurosci 16: 298–306, 1996.
Conway BA, Hultborn H, Kiehn O, Mintz I. Plateau potentials in alpha-motoneurones induced by intravenous injection of L-dopa and clonidine in the spinal cat. J Physiol 405: 369–384, 1988.
Crone SA, Quinlan KA, Zagoraiou L, Droho S, Restrepo CE, Lundfald L, Endo T, Setlak J, Jessell TM, Kiehn O, Sharma K. Genetic ablation of V2a ipsilateral interneurons disrupts left-right locomotor coordination in mammalian spinal cord. Neuron 60: 70–83, 2008.[CrossRef][Web of Science][Medline]
Degtyarenko AM, Simon ES, Burke RE. Locomotor modulation of disynaptic EPSPs from the mesencephalic locomotor region in cat motoneurons. J Neurophysiol 80: 3284–3296, 1998.
Duysens J. Reflex control of locomotion as revealed by stimulation of cutaneous afferents in spontaneously walking premammillary cats. J Neurophysiol 40: 737–751, 1977.
Endo T, Kiehn O. Conductance measurement of synaptic inputs to mammalian motor neurons during locomotion. Soc Neurosci Abstr 33: 188.8, 2007.
Gilmore J, Fedirchuk B. The excitability of lumbar motoneurones in the neonatal rat is increased by a hyperpolarization of their voltage threshold for activation by descending serotonergic fibres. J Physiol 558: 213–224, 2004.
Godderz W, Illert M, Yamaguchi T. Efferent pattern of fictive locomotion in the cat forelimb: with special reference to radial motor nuclei. Eur J Neurosci 2: 663–671, 1990.[CrossRef][Web of Science][Medline]
Grillner S. Control of locomotion in bipeds, tetrapods, and fish. In: Handbook of Physiology. The Nervous System. Motor Control. Bethesda, MD: Am. Physiol. Soc., 1981, sect. 1, vol. II, pt. 2, p. 1179–1236.
Hiebert GW, Pearson KG. Contribution of sensory feedback to the generation of extensor activity during walking in the decerebrate cat. J Neurophysiol 81: 758–770, 1999.
Hochman S, Schmidt BJ. Whole cell recordings of lumbar motoneurons during locomotor-like activity in the in vitro neonatal rat spinal cord. J Neurophysiol 79: 743–752, 1998.
Hultborn H. State-dependent modulation of sensory feedback. J Physiol 533: 5–13, 2001.
Jiang Z, Carlin KP, Brownstone RM. An in vitro functionally mature mouse spinal cord preparation for the study of spinal motor networks. Brain Res 816: 493–499, 1999.[CrossRef][Web of Science][Medline]
Jordan LM. Factors determining motoneuron rhythmicity during fictive locomotion. Symp Soc Exp Biol 37: 423–444, 1983.[Medline]
Jordan LM. Brainstem and spinal cord mechanisms for the initiation of locomotion: single limb analysis. In: Neurological Basis of Human Locomotion, edited by Shimamura M, Grillner S, Edgerton VR. Tokyo: Japan Scientific Societies, 1991, p. 3–20.
Kahn JA. Patterns of synaptic inhibition in motoneurons and interneurons during fictive swimming in the lamprey, as revealed by Cl– injections J Comp Physiol A Sens Neural Behav Physiol 147: 189–194, 1982.[CrossRef]
Kiehn O. Locomotor circuits in the mammalian spinal cord. Annu Rev Neurosci 29: 279–306, 2006.[CrossRef][Web of Science][Medline]
Kiehn O, Kjaerulff O. Spatiotemporal characteristics of 5-HT and dopamine-induced rhythmic hindlimb activity in the in vitro neonatal rat. J Neurophysiol 75: 1472–1482, 1996.
Kiehn O, Kjaerulff O, Tresch MC, Harris-Warrick RM. Contributions of intrinsic motor neuron properties to the production of rhythmic motor output in the mammalian spinal cord. Brain Res Bull 53: 649–659, 2000.[CrossRef][Web of Science][Medline]
Kjaerulff O, Kiehn O. 5-HT modulation of multiple inward rectifiers in motoneurons in intact preparations of the neonatal rat spinal cord. J Neurophysiol 85: 580–593, 2001.
Kriellaars DJ, Brownstone RM, Noga BR, Jordan LM. Mechanical entrainment of fictive locomotion in the decerebrate cat. J Nneurophysiol 71: 2074–2086, 1994.
Kuo JJ, Lee RH, Zhang L, Heckman CJ. Essential role of the persistent sodium current in spike initiation during slowly rising inputs in mouse spinal neurones. J Physiol 574: 819–834, 2006.
Lafreniere-Roula M, McCrea DA. Deletions of rhythmic motoneuron activity during fictive locomotion and scratch provide clues to the organization of the mammalian central pattern generator. J Neurophysiol 94: 1120–1132, 2005.
Larkman PM, Kelly JS. Modulation of IH by 5-HT in neonatal rat motoneurones in vitro: mediation through a phosphorylation independent action of cAMP. Neuropharmacology 36: 721–733, 1997.[CrossRef][Web of Science][Medline]
Li Y, Bennett DJ. Persistent sodium and calcium currents cause plateau potentials in motoneurons of chronic spinal rats. J Neurophysiol 90: 857–869, 2003.
Liu J, Jordan LM. Stimulation of the parapyramidal region of the neonatal rat brain stem produces locomotor-like activity involving spinal 5-HT7 and 5-HT2A receptors. J Neurophysiol 94: 1392–1404, 2005.
Marino J, Schummers J, Lyon DC, Schwabe L, Beck O, Wiesing P, Obermayer K, Sur M. Invariant computations in local cortical networks with balanced excitation and inhibition. Nat Neurosci 8: 194–201, 2005.[CrossRef][Web of Science][Medline]
McCrea DA. Spinal circuitry of sensorimotor control of locomotion. J Physiol 533: 41–50, 2001.
McCrea DA, Rybak IA. Organization of mammalian locomotor rhythm and pattern generation. Brain Res Rev 57: 134–146, 2008.[CrossRef][Medline]
McDearmid JR, Scrymgeour-Wedderburn JF, Sillar KT. Aminergic modulation of glycine release in a spinal network controlling swimming in Xenopus laevis. J Physiol 503: 111–117, 1997.
McHanwell S, Biscoe TJ. The localization of motoneurons supplying the hindlimb muscles of the mouse. Philos Trans R Soc Lond B Biol Sci 293: 477–508, 1981.
McLean DL, Merrywest SD, Sillar KT. The development of neuromodulatory systems and the maturation of motor patterns in amphibian tadpoles. Brain Res Bull 53: 595–603, 2000.[CrossRef][Web of Science][Medline]
Nielsen JB, Sinkjaer T. Afferent feedback in the control of human gait. J Electromyogr Kinesiol 12: 213–217, 2002.[CrossRef][Medline]
Nishimaru H, Restrepo CE, Ryge J, Yanagawa Y, Kiehn O. Mammalian motor neurons corelease glutamate and acetylcholine at central synapses. Proc Natl Acad Sci USA 102: 5245–5249, 2005.
Nishimaru H, Takizawa H, Kudo N. 5-Hydroxytryptamine-induced locomotor rhythm in the neonatal mouse spinal cord in vitro. Neurosci Lett 280: 187–190, 2000.[CrossRef][Web of Science][Medline]
Noga BR, Kriellaars DJ, Brownstone RM, Jordan LM. Mechanism for activation of locomotor centers in the spinal cord by stimulation of the mesencephalic locomotor region. J Neurophysiol 90: 1464–1478, 2003.
Orsal D, Perret C, Cabelguen JM. Evidence of rhythmic inhibitory synaptic influences in hindlimb motoneurons during fictive locomotion in the thalamic cat. Exp Brain Res 64: 217–224, 1986.[Web of Science][Medline]
Pearson KG. Generating the walking gait: role of sensory feedback. Prog Brain Res 143: 123–129, 2004.[Web of Science][Medline]
Pearson KG, Duysens J. Function of segmental reflexes in the control of stepping in cockroaches and cats. In: Neural Control of Locomotion, edited by Herman RN, Grillner S, Stein PS, Stuart DG. New York: Plenum Press, 1976, p. 519–537.
Perreault MC. Motoneurons have different membrane resistance during fictive scratching and weight support. J Neurosci 22: 8259–8265, 2002.
Perret C, Gabelguen JM, Orsal D. Analysis of the pattern of activity in "knee flexor" motoneurons during locomotion in the cat. In: Stance and Motion: Facts and Concepts, edited by Gurfinkle VS, Ioffe ME, Massion J, Roll JP. New York: Plenum Press, 1988, p. 133–141.
Roberts A, Tunstall MJ. Mutual re-excitation with post-inhibitory rebound: a simulation study on the mechanisms for locomotor rhythm generation in the spinal cord of Xenopus embryos. Eur J Neurosci 2: 11–23, 1990.[CrossRef][Web of Science][Medline]
Robertson GA, Stein PS. Synaptic control of hindlimb motoneurones during three forms of the fictive scratch reflex in the turtle. J Physiol 404: 101–128, 1988.
Russell DF, Wallen P. On the control of myotomal motoneurones during "fictive swimming" in the lamprey spinal cord in vitro. Acta Physiol Scand 117: 161–170, 1983.[Web of Science][Medline]
Schmidt BJ. Afterhyperpolarization modulation in lumbar motoneurons during locomotor-like rhythmic activity in the neonatal rat spinal cord in vitro. Exp Brain Res 99: 214–222, 1994.[Web of Science][Medline]
Schmidt BJ, Jordan LM. The role of serotonin in reflex modulation and locomotor rhythm production in the mammalian spinal cord. Brain Res Bull 53: 689–710, 2000.[CrossRef][Web of Science][Medline]
Schmidt BJ, Meyers DE, Fleshman JW, Tokuriki M, Burke RE. Phasic modulation of short latency cutaneous excitation in flexor digitorum longus motoneurons during fictive locomotion. Exp Brain Res 71: 568–578, 1988.[CrossRef][Web of Science][Medline]
Shefchyk SJ, Jordan LM. Excitatory and inhibitory postsynaptic potentials in alpha-motoneurons produced during fictive locomotion by stimulation of the mesencephalic locomotor region. J Neurophysiol 53: 1345–1355, 1985a.
Shefchyk SJ, Jordan LM. Motoneuron input-resistance changes during fictive locomotion produced by stimulation of the mesencephalic locomotor region. J Neurophysiol 54: 1101–1108, 1985b.
Shu Y, Hasenstaub A, McCormick DA. Turning on and off recurrent balanced cortical activity. Nature 423: 288–293, 2003.[CrossRef][Web of Science][Medline]
Stein PS. Neuronal control of turtle hindlimb motor rhythms. J Comp Physiol A Neuroethol Sens Neural Behav Physiol 191: 213–229, 2005.[CrossRef][Web of Science][Medline]
Tresch MC, Kiehn O. Coding of locomotor phase in populations of neurons in rostral and caudal segments of the neonatal rat lumbar spinal cord. J Neurophysiol 82: 3563–3574, 1999.
Vult von Steyern F, Lømo T. Postnatal appearance of 5-HT2A receptors on fast flexor and slow extensor rat motor neurons. Neuroscience 136: 87–93, 2005.[CrossRef][Web of Science][Medline]
Wallen P, Shupliakov O, Hill RH. Origin of phasic synaptic inhibition in myotomal motoneurons during fictive locomotion in the lamprey. Exp Brain Res 96: 194–202, 1993.[Web of Science][Medline]
Whelan P, Bonnot A, O'Donovan MJ. Properties of rhythmic activity generated by the isolated spinal cord of the neonatal mouse. J Neurophysiol 84: 2821–2833, 2000.
Whelan PJ. Developmental aspects of spinal locomotor function: insights from using the in vitro mouse spinal cord preparation. J Physiol 553: 695–706, 2003.
Zaporozhets E, Cowley KC, Schmidt BJ. A reliable technique for the induction of locomotor-like activity in the in vitro neonatal rat spinal cord using brainstem electrical stimulation. J Neurosci Methods 139: 33–41, 2004.[CrossRef][Web of Science][Medline]
This article has been cited by other articles:
|
|
 |
|
  K. J. Dougherty and O. Kiehn Firing and Cellular Properties of V2a Interneurons in the Rodent Spinal Cord J. Neurosci., January 6, 2010; 30(1): 24 - 37. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Frigon and J.-P. Gossard Asymmetric control of cycle period by the spinal locomotor rhythm generator in the adult cat J. Physiol., October 1, 2009; 587(19): 4617 - 4628. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Perez, J. A. Tapia, C. R. Mirasso, J. Garcia-Ojalvo, J. Quevedo, C. A. Cuellar, and E. Manjarrez An Intersegmental Neuronal Architecture for Spinal Wave Propagation under Deletions J. Neurosci., August 19, 2009; 29(33): 10254 - 10263. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. P. Gabriel, R. Mahmood, A. Kyriakatos, I. Soll, G. Hauptmann, R. L. Calabrese, and A. El Manira Serotonergic Modulation of Locomotion in Zebrafish--Endogenous Release and Synaptic Mechanisms J. Neurosci., August 19, 2009; 29(33): 10387 - 10395. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
TOP
ABSTRACT
INTRODUCTION
