| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Center for Neurobiology and Behavior, New York City, New York 10032 2 New York State Psychiatric Institute, New York City, New York 10032
Submitted 18 February 2003; accepted in final form 31 March 2003
 |
ABSTRACT |
|---|
|
|
|
INTRODUCTION |
|---|
|
;
Flament et al. 1992;
Li and Martin 2001;
Olivier et al. 1997; for
review, see Porter and Lemon
1993). In several species, connectional specificity is achieved
partly by eliminating spinal terminations because CS axons terminate within a
more extensive region of the gray matter early in development than in maturity
(Alisky et al. 1992;
Martin et al. 1999;
Theriault and Tatton 1989). In
addition to—or in place of—axon terminal elimination, there also
is extensive local CS axon terminal growth
(Armand et al. 1997; Li and
Martin 2001,
2002). Together, the combined
elimination and addition of terminals could lead to development of the
capability of the CS system to drive spinal motor circuits for controlling
skilled movements.
Early postnatal life is thus a dynamic time for development of the CS
system and ascendancy to its important role in controlling skilled movements.
Not surprisingly, this is also a time of transition, from when CS functions
are not outwardly expressed, either physiologically or behaviorally, to when
the same experimental manipulations point to a clear role for the CS system in
activating spinal motor circuits for producing muscle contraction. For
example, motor cortex stimulation in both cat and monkey—either
electrical or transcranial magnetic stimulation (TMS)—fails to produce
motor effects early in development before CS termination refinement
(Alisky et al. 1992;
Armand et al. 1997;
Martin et al. 1999;
Theriault and Tatton 1989).
But immediately thereafter, stimulation of motor cortex produces muscle
contraction (Bruce and Tatton
1980; Chakrabarty and Martin
2000; Flament et al.
1992; Olivier et al.
1997). In the monkey, the expression of relatively independent
finger movements—a measure of CS function—also coincides with
achieving the mature topographic pattern of terminations
(Armand et al. 1997). The
effects of CS system stimulation in humans during early postnatal development
are similar to those in cat and monkey. Generally there are heightened
thresholds for evoking movements using TMS, and motor effects are inconsistent
during the first year or two of postnatal development. This is before
acquisition of relatively independent finger movements
(Eyre et al. 2000;
Porter and Lemon 1993). This
is followed by a later postnatal drop in TMS threshold
(Eyre et al. 2000), more
consistent responding to TMS stimulation
(Nezu et al. 1997), and a
correlated improvement in hand skills
(Eyre et al. 2000). In the
cat, lesions or inactivation of the motor cortex before CS termination
refinement produces minimal motor control disturbances but robust effects
immediately thereafter (Armand and Kably
1992; Martin et al.
1999,
2000). What could account for
these differences in expression of CS function that occur just before and
after the period of topographic refinement of CS terminals? While the presence
of compensatory mechanisms or maturation of motivational and cognitive systems
may play a role, we have been struck by the possibility that the CS system is
incapable of exciting spinal motor circuits to a sufficient degree for
controlling limb movements until their axon terminals are refined
topographically.
While development of the capacity to drive spinal neurons is certainly
important for the CS system's essential role in skilled movement control, it
is also critical for refining the topography of its spinal terminations. We
showed that CS activity blockade, by intracortical infusion of the GABA
agonist muscimol, impairs development of CS terminations
(Martin et al. 1999). The
silenced cortex fails to develop dense spinal terminations, whereas the active
side develops both dense contralateral and ipsilateral terminations. The
ipsilateral terminations of the active cortex are produced both by maintenance
of transient terminals and new growth beyond the normal territory of transient
terminations. Given this important role for activity in shaping development of
CS terminations, it is all the more paradoxical that cortical motor functions
are not apparent at early postnatal ages.
The principal purpose of this study was to determine if the delay in
expression of cortical motor system function is due to the inability of the CS
synapse to activate its postsynaptic spinal targets. We studied this question
in the cat so that the results could be interpreted in relation to the time
course of activity-dependent refinement, development of CS terminal
morphology, and development of the cortical motor map
(Chakrabarty and Martin 2000;
Li and Martin 2000,
2001,
2002;
Martin and Lee 1999;
Martin et al. 1999). To probe
development of the capacity of CS terminations to evoke spinal postsynaptic
responses, we bypassed the influence of motor cortical circuit development by
stimulating CS axons directly in the medullary pyramid (PT). We recorded
surface-evoked potentials from the spinal cord and local field potentials
within the gray matter in response to PT stimulation. We chose these
extracellular recording approaches for two reasons. First, these are classical
population recording techniques that are effective in sampling the actions of
ensembles of CS axon terminals and their postsynaptic targets
(Baldissera et al. 1981;
Maier et al. 1998). Second,
because it is difficult to maintain acute neonatal spinal preparations, these
extracellular techniques are efficient substitutes for the more
physiologically invasive single-cell recordings requiring paralysis and
artificial ventilation. We show that CS terminals are capable of driving
spinal circuits well before anatomical specificity is achieved and before
primary motor cortex stimulation is effective in producing motor responses.
Given the importance of activity-dependent refinement of CS axon terminals,
this capacity for activating spinal neurons during early postnatal life could
play an important role in development of CS circuit connectivity. Some of the
findings in this paper were presented in an abstract
(Meng et al. 2002).
|
|
METHODS |
|---|
|
2 yr old. Adults were housed individually.
All experiments were conducted with the approval of the New York State
Psychiatric Institute and Columbia University institutional animal care and
use committees. Surgical procedures
Anesthesia was induced with ketamine (30 mg/kg im) and xylazine
(0.6–0.8 mg/kg im) and maintained using intravenous ketamine infusion
(10–30 mg · kg–1 ·
h–1). Wound margins were infiltrated with
lidocaine (2%). This is the same anesthesia protocol used in our earlier study
on development of the cortical motor map
(Chakrabarty and Martin 2000).
Using this anesthesia protocol, animals remained areflexive during surgery and
testing. Animals were placed in a conventional stereotaxic frame. Body
temperature was maintained at 39° by a heating pad. We performed a
cervical laminectomy, exposing the dorsal surface of the cord from
C4 to C6. After exposing the spinal cord, it was covered
with warm mineral oil. The C8 spinous process was clamped to
stabilize the vertebral column. The combination of C8 spinal
fixation and placement of the head in the stereotaxic frame produced adequate
stabilization of the cervical cord for recording surface and depth field
potentials.
Pyramidal tract stimulation
A craniotomy was made over the posterior portion of the occipital bone to access the pyramid. In adults, an electrode was placed at standard stereotaxic anteroposterior (AP) and mediolateral (ML) coordinates (AP 13.5 ML 0.75 mm). In kittens, we adjusted the AP location so that the electrode was adjacent to the caudal edge of the occipital bone, at ML 0.5 mm. The electrode was inserted parallel to the vertical plane. This resulted in consistent PT stimulating electrode placement within the caudal medulla (see Fig. 1A).
|
We inserted a pair of concentric electrodes (separated by 2 mm, oriented in the sagittal plane; Microprobe) into the medulla to stimulate CS axons in the pyramid. We found that we were able to activate CS axons more effectively (i.e., produce larger spinal potentials) when we used a pair of electrodes (caudal electrode as anode) than a single electrode with the anode at the center and the cathode as the surround electrode. We lowered the electrodes to a depth at which contact occurred with the ventral surface of the occipital bone, just beneath the pyramid. The point of contact was noted under a dissecting scope as a small jarring motion of the electrodes. We then raised the electrodes between 0.5 and 1.5 mm, to a depth that evoked the largest surface potential at C4 at the lowest stimulus amplitude. We used a constant current stimulator (AM System, Model No. 2100; biphasic stimulus 20-µS duration); current varied in relation to the animal's age (see RESULTS). In all experiments, we determined the position of the electrodes histologically. In six experiments, we unilaterally sectioned the dorsolateral column of the spinal cord in the horizontal plane at C5 while recording at C6 to verify the conduction path for PT activation.
Spinal recordings and data acquisition
Surface potentials were recorded above the dorsal root entry zone in the
center of the target spinal segment (C4, C5, and/or
C6) using a silver ball electrode. Depth recordings were made using
a glass micropipette (tip diameter between 10 and 20 µm, filled with 4 M
NaCl; 0.1–0.4 M
). The micropipette was mounted to a Kopf
electrode carrier. The pipette was connected to a unity gain preamplifier,
which in turn was connected to an AC-coupled amplifier (AM Systems, Model No.
1800). Filters cut signals <100 Hz and >10 kHz. In each experiment in
which depth recordings were made, penetrations were made at 10 sites, equally
spaced within the C6 segment. We recorded at 500-µm-depth
intervals from just beneath the pial surface to the a point within the ventral
gray or white matter. The output of the amplifier was digitized on-line
(Axograph, Digidata; Axon Instruments) and responses to individual stimuli or
a set of 10 stimuli (as an ensemble average) were saved on disk for subsequent
analysis (see RESULTS for particular measurements.) The records all
show positive up. Data were collected at a sampling rate of 20 kHz, in 20-ms
trials.
Data analysis
We used the program Axograph for the Apple Macintosh computer to measure the onset and amplitude of individual responses. The results of the analyses were used to create two databases. One contained all data from the surface recordings and the other from the depth recordings. Most depth recordings were accompanied by a concurrent surface record close to the penetration. We determined the significance of age-related and site-related differences using an unpaired t-test, ANOVAs, and linear regressions using the program Statview. Specific analyses are described in RESULTS.
Histology
At the end of each experiment a lesion was made through the caudal PT stimulating electrode. The animal was killed with an overdose of anesthesia and either perfused transcardially with saline, followed by 10% formal-saline or the brain and cervical spinal cord were removed in situ and placed in fixative. After a suitable postfixation time, transverse sections were cut through the caudal medulla and either transverse or longitudinal sections were cut through the region of the spinal cord examined. Tissue was Nissl-stained and examined for the PT electrode lesion site and track, spinal recording sites, and spinal lesion site. Depths in the spinal cord were measured in relation to landmarks (i.e., gray matter borders) or dye spots after tissue was mounted onto slides but before staining.
|
|
RESULTS |
|---|
|
Experiments were conducted on 19 cats between postnatal week (PW) 4 and adult (PW4, n = 2; PW5, n = 2; PW6, n = 1; PW7, n = 1; PW8, n = 1; PW9, n = 2; PW10, n = 1; PW11, n = 1; PW13, n = 2; adult, n = 6) to determine age-dependent changes in spinal cord potentials produced by PT stimulation. Surface potentials were recorded in all cats; a laminar series throughout the gray matter was recorded in two PW4 animals, two adults, and in one of each of the cats at the following ages: 6, 8, 10, and 13 wk. Key developmental parameters are summarized in Table 1. We combined animals into three groups: weeks 4–5, which correspond to ages before CS topographic refinement; weeks 6–13, which correspond to kittens at ages after topographic refinement; and adults.
|
Stimulation over approximately a 1-mm dorsoventral distance (slightly less in the kitten) in the ventral medulla evoked short latency spinal potentials. Stimulation 1 mm ventral or dorsal to the pyramidal stimulation site either evoked no response or substantially smaller responses, often with a different configuration. Figure 1A is a Nissl-stained section through the ventral medulla in a PW8 animal showing the lesion produced at the tip of one of the stimulating electrodes within the PT. PT stimulation at all ages evoked two distinct potentials recorded from the surface of the cervical cord, at consistent latency (Fig. 1, B and C; PW5). While the onset of the first potential was typically obscured by the stimulus artifact, especially in older animals where the latency was shortest, the peak was discernable. Figure 1B shows ensemble averages of surface potentials (n = 10), aligned with the stimulus artifact and recorded from the fourth, fifth, and sixth cervical segments. The time to the peak of the first potential (arrows) and the onset of the second potential (arrows and vertical tic marks) increased from C4 (thick trace) to C6 (thin trace). This is consistent with an increasing conduction distance from C4 to C6. When the traces were aligned with the peak of the initial response (Fig. 1C), the onset times of the late response did not increase systematically with more distal segments (single arrow; vertical tic marks). This is consistent with a relatively constant conduction distance from the CS tract branch points in the lateral white matter at the different segments to the segmental gray matter terminations. We occasionally recorded small potentials near the end of the volley. These could be mediated by slowly conducting corticospinal axons, with a latency about twice that of the initial potential.
Figure 2A plots the
time to peak of the first response, the measure used to monitor its latency,
for all animals in which recordings were made at C4, C5,
and C6. In all cases, the latency of the first response increased
with increasing conduction distance (F = 67.24; P <
0.0001). The interval between the peak of the first and onset of the second
potential is plotted in Fig.
2B for recordings at the three cervical levels. We used
this interval as a measure of the intrasegmental conduction time, although it
is likely also to include pre- and postsynaptic processes. Overall, there were
no systematic changes in these values from C4 to C6
(F = 0.196; P = 0.6632), presumably because conduction
distances (and times) remained relatively stable across segments. However, we
did note a large decrease in latency in one PW5 animal and a small increase in
the adult. These differences could reflect variations in the strength of
activation resulting in different summation times, differences in conduction
distance, and in conduction velocity of CS axon branches
(Shinoda et al. 1976).
|
These large potentials were eliminated with lesion of the dorsolateral spinal white matter, the principal location of descending CS projection in the cat. After conducting the experiment in six cats, we acutely lesioned the dorsolateral column at C5 while continuing to record at C6. Ensemble averages of the C6 surface potentials recorded immediately before and after making the lesion are shown in Fig. 3, A and C. A Nissl-stained longitudinal section through the lesion is shown in B. The volley was substantially reduced and its configuration changed. The postsynaptic field response was eliminated.
|
Our findings are consistent with the first potential corresponding to the
conducted CS tract volley and the second potential corresponding to the
initial postsynaptic response in the gray matter. The two-potential pattern
that we recorded is identical to what Elger and colleagues
(Elger et al. 1977) recorded
in the rat in response to motor cortex stimulation. In that study, the second
potential occurred at the onset of monosynaptic excitation of spinal neurons.
In the following discussion, we refer to the initial surface response (and
corresponding depth records; see following text) as the volley and the later
response, as the postsynaptic field potential.
Postsynaptic field potentials were evoked by PT stimulation in both immature and mature cats
Ensemble averages of surface potentials recorded at C6 at different postnatal ages are shown in Fig. 4. Potentials recorded at PW4 had substantially longer latencies than in older animals. We show data from two animals because the delay times (horizontal arrows) were noticeably different. For the volley, this is likely due to reduced myelination of the CS tract. We were unable to measure volley rise time in all animals because a portion of the beginning of the volley was obscured by the stimulus artifact. To monitor changes in volley duration, we instead measured volley fall-time (i.e., interval between volley peak and the nadir between the volley and the second potential). This value was reduced significantly during development (F = 5.46; P = 0.0165). The delay in the postsynaptic field response could be due to reduced conduction along very fine preterminal axons or a delayed postsynaptic response due to ineffective early summation.
|
At 5 wk, the two potentials merged as a positive-negative-positive
sequence. This sequence is similar to that in the rat
(Elger et al. 1977) and the
cat (Maier et al. 1998). We
adjusted the scales in Fig. 4
to show the potentials at approximately the same amplitude. While the
potentials could be roughly the same size in young and old animals, implying
similarity in the CS termination density and strength, this is misleading
because spinal cord diameter increased substantially with age. The distance
between the surface electrode and the CS tract in the lateral white matter was
400 µm at 4 wk and 1,700 µm in the adult. The amplitude of
extracellular potentials decrease with distance. The tracings in
Fig. 4 also show that there was
late activity.
The graphs in Fig. 5
summarize the age-related changes in several stimulus and response parameters.
A plots the current thresholds for evoking the spinal responses.
There was a larger difference between the mean thresholds of animals 
5 wk
and adults (251 mA) than between older kittens (6–13 wk) and adults (86
mA). Pair-wise comparison of data from animals in these three groups revealed
significant differences between each (5 wk and 6–13 wk: t =
5.30; P = 0.0003; 5 wk and adults: t = 39.76; P
< 0.0001; 6–13 wk and adults: t = 3.1; P = 0.008).
We found that conduction velocity (Fig.
5B) increased continuously (for the 3 groups in
Table 1, F = 45.001;
P < 0.0001). While in the oldest kitten conduction velocity was
not as fast as in adults, by other measures (see following text), the oldest
kitten appeared to have a mature CS system. The reduction in stimulation
threshold (in mA) for evoking spinal potentials
(Fig. 5A) and the
increase in CS tract conduction velocity are both likely to be due to
increased myelination. These changes are similar to those reported, using TMS,
in monkey (Eyre et al. 2000;
Flament et al. 1992;
Olivier et al. 1997) and human
(Eyre et al. 2000;
Koh and Eyre 1988).
|
Although difficult to substantiate in this small sample, volley latency (Fig. 5C) seemed to remain relatively constant after week 4. For example, latencies in the two oldest kittens were comparable to those of the adults while conduction velocity was only 60–80%. Moreover, several of the kittens between 5 and 13 wk had volleys of similar latency despite considerable intervening growth (i.e., conduction distance; E). Between weeks 5 and 13 there was only a 22% reduction in latency despite a 41% increase in conduction path. This suggests relative constancy of central conduction time. On the other hand, the latency to the onset of the postsynaptic surface field potential (Fig. 5D) decreased gradually. Our findings indicate, for all postnatal ages and adults, the presence of a robust postsynaptic field potential and either a stepwise or gradual maturation of many CS electrophysiological characteristics.
Laminar analysis of local field potentials
We next examined local field potentials from the surface of the cord to the ventral gray matter in response to PT stimulation in animals of different ages. There were two purposes for these depth recordings. First, we determined the change in the amplitude of the field potentials at the various depths as a measure of the local gray matter territories that could be influenced by CS activation at different postnatal ages. Second, we determined if there was a consistent relationship between the surface and depth recordings, to obtain insight into the generators of the postsynaptic response. For each animal, we made 10 penetrations, equally spaced within C6. Within each penetration, we recorded at 500-µm intervals, from the surface into the ventral horn. In some cases, the deepest recordings were in the anterolateral white matter.
Figure 6 (left) shows a transverse section through C6 in a PW8 kitten. During the experiment, we marked two depths with dye (1,000 and 3,000 µm). Each tracing in the panel on the right is the average of recordings at the 10 different rostrocaudal sites within C6 at the stated depth. For example, the trace marked "surface" is the ensemble average of surface recordings at the 10 sites. The purpose of using averages from sites distributed along the segment was to estimate more accurately the pattern of CS activation within the segment. As we show in the following text, a slightly different pattern of potentials was recorded in a minority of penetrations. These averaged recordings, as in the other kittens and adults, revealed a multiphasic potential. And as discussed in the following text, the dorsoventral distribution varied with age.
|
The first two phases (positive-negative;
Fig. 6; light gray shading)
occurred during the surface volley, and the second two phases
(positive-negative; Fig. 6;
dark gray shading), during the surface-recorded postsynaptic response. Based
on classical field potential analysis
(Lorente de Nó 1947),
the initial positive-negative potential is consistent with recording local
outward conductances (i.e., current source) as impulses are distant to the
pipette tip, followed by local inward conductances (i.e., current sink) as
impulses are conducted toward and recorded at the tip of the pipette. These
local inward conductances would predominantly correspond, in sequence, to the
impulses conducted along local preterminal axons, their subsequent
depolarization, and early postsynaptic depolarization. The subsequent
potentials (dark gray shading; positive-negative), based on their longer
latency, reflect postsynaptic actions (presumably both excitation and
inhibition). In this, and other kittens (10 wk), there was a conspicuous
late phasic negative potential (blackened panel) that occurred during the
surface recorded postsynaptic field response. This potential inverts dorsally
and ventrally, suggesting that it is a depolarizing potential with a
distribution shown by the blackened rectangle. This negative potential had an
age-dependent dorsoventral distribution (see following text).
Figure 7 shows depth recordings from two kittens (4 and 6 wk old) and an adult. One major difference between the form of the evoked responses in kittens (including the one shown in Fig. 6) and adults is that the late surface and deep potentials were roughly inverted (i.e., out of phase) in the kittens. This corresponded partly to the prominent negative response in the kittens (Fig. 6; black rectangle). In addition, in adults the duration of the surface volley was shorter so that the late depth potentials roughly paralleled the late surface potential (i.e., in phase). The significance of this latter difference is not clear.
|
Age-dependent reduction in the dorsoventral distribution of depth potentials
Across all experiments, there were two penetration classes. 84% showed
substantial changes in the form of the field potentials across depths.
Figure 7 shows data from this
penetration type. In the 4 wk old animal
(Fig. 7, left), large
field potentials—negative, positive, and late negative— during the
surface volley and postsynaptic field response were recorded at all but the
deepest depth, or >90% of the dorsoventral extent of the gray matter. The
ventral border of the ventral horn was at 3,510 µm; thus the deepest
recording—where the potential changed configuration and size—was
at the ventral border of the gray and white matter. The dorsoventral
distribution of the large initial negative and subsequent positive and
negative potentials was more restricted by week 6 (middle), being
eliminated both superficially and deeper. In the adult
(Fig. 7, right),
potentials recorded from the most superficial and deep sites were small and
invariant. However, within the intermediate zone layers the potential
increased substantially. This increase occurred between 2,000 and 3,500 µm,
which was 35% of the dorsoventral extent of the gray matter in this
animal. The ventral horn border was 5,270 µm. Similar results were obtained
in the second adult subjected to a laminar analysis, where there was a
qualitative increase in the amplitude of the local potential also between
2,000 and 3,500 µm.
Figure 8 presents two examples of data from a minority of penetrations (16%) in kittens where the potentials at the various depths, when present, did not change substantially from the form of the surface response. These sites were present in 6-, 8-, 10-, and 14-wk-old kittens. Similar to the typical sites, field potential amplitude varied, being largest within the middle layers, and the dorsoventral distribution was more restricted in the older than younger kittens (e.g., compare 7, left and right). However, for a given age, these sites had a more restricted dorsoventral distribution than the more typical sites. These sites may simply have sparse CS terminations, resulting in a limited focus of activation, or have matured earlier, resulting in a more restricted distribution. Adjacent penetrations recorded the typical kind of potentials, suggesting local anisotropy in the distribution of CS terminations.
|
The results of the depth analysis show a dorsoventral restriction in both the volley and postsynaptic field potentials. Because the occurrence of the volley and postsynaptic field potentials were tightly correlated for both typical (i.e., Figs. 6 and 7) and atypical (i.e., Fig. 8) sites, the volley was not merely recording impulses in the CS tract at a distance. Rather the potential was due to local conductances associated with CS axon terminals.
As animals grew older, there was a striking restriction in the dorsoventral distribution over which large potentials could be recorded. We quantified this by measuring the amplitude of the second positive potential (i.e., positive peak within the darkly shaded region in Figs. 6, 7, 8; measured between the baseline and peak). We chose this potential to measure because it was clearly present at all ages. However, other phases of the potential also showed the same dorso-ventral changes, as can be seen in Figs. 6, 7, 8. Figure 9 plots the average values (across the 10 penetrations in each animal) of the potential, measured at 500-µm intervals. Because there was over a 50% increase in the size of the gray matter during development (e.g., at PW4, the gray matter was 2.8 mm but 4.3 mm in the adult), we normalized the depth measurements to the dimensions of the gray matter for each animal. The first recording (i.e., 0%) was at or close to the dorsal border of the gray matter, whereas the deepest recordings were at or close to the ventral border (100%). For a given animal, the largest potentials at any age were within the deeper laminae of the dorsal horn and the intermediate zone. This corresponds approximately to depths between 40 and 60%. Only in the oldest kitten (13 wk) and the adult was the focus of activation sharply restricted. This redistribution could reflect both the elimination of branches and local growth. This redistribution could not, however, be explained by the age-dependent increase in the size of the spinal cord. Although the gray matter was only 2% larger at week 13 than week 10, there was a qualitative change in the distribution of field potential amplitude, from broad to focused. In contrast, the area of the gray matter was 18% larger in the adult than the week 13 animal, yet the distributions of field potential amplitudes had a similar focus in the intermediate zone. These results show that there is a developmental restriction in the distribution of physiological effects of CS axon terminals.
|
|
|
DISCUSSION |
|---|
|
,
2001,
2002). We propose that CS
synapses are not transmitting information for specifying muscle control at
young postnatal ages, before the time when cortical stimulation in the cat
produces motor effects (Bruce and Tatton
1980; Chakrabarty and Martin
2000). Although the lack of stimulation efficacy could be due to
several factors, such as poor synchronization of descending signals or a
reduced capacity of spinal neurons to integrate CS inputs, this nonetheless
shows that the CS system cannot sufficiently drive spinal motor circuits to
produce muscle contraction. We reason that if cortical burst stimulation
(Chakrabarty and Martin 2000)
or TMS (Flament et al. 1992;
Olivier et al. 1997) fails to
produce muscle contraction at young ages, then the naturally occurring
discharges of CS neurons would be less able to do so because they are apt to
be desynchronized by comparison. What then could the role be for CS synaptic
activity? We propose that the capability of developing CS terminals to
activate spinal neurons at this age is important for shaping early development
of the topography of CS terminations. We have previously shown that CS neural
activity during early postnatal life is important for the normal development
of the topography of CS terminations
(Martin et al. 1999) through
an activity-dependent competition mechanism
(Martin and Lee 1999).
Sensory-motor cortex activity blockade disrupts development of the normal
topography of CS terminations (Martin et
al. 1999; Martin and Lee
1999) and prehension skills
(Martin et al. 2000). With CS
synapses capable of exciting spinal neurons during early postnatal life,
activity during a brief postnatal period can have long-term consequences for
constructing spinal motor circuits. A role in activity-dependent development
does not necessarily preclude additional roles in other spinal functions.
Indeed, as in the visual system where activity is presumably both shaping
development and providing visual inputs to the brain
(Goodman and Shatz 1993), CS
activity may have important spinal functions in the neonate, such as providing
excitability to segmental circuits to promote reflex and locomotor functions.
Developing CS terminals may also play a role in refining the connections of
other spinal systems. Recently Clowry and coworkers
(Gibson et al. 2000) showed
that early postnatal motor cortex lesion in the rat produced an increase in
muscle afferent terminations in the ventral horn, suggesting interactions
between developing cortical and spinal afferents. Relationship of changes in CS evoked responses to anatomical refinement
All of our depth records showed a consistent association between the amplitude of the initial positive-negative phase, which is associated with presynaptic activation and the local postsynaptic potentials (both early positivity and late negativity). As the first negative potential grew in amplitude with depth, so too did the subsequent positive and second negative potentials. In a minority of sites in young and old animals, there were depths, both superficial and deep, where neither an early negative potential nor the late postsynaptic potential were recorded. This demonstrates that the extracellular potentials that we recorded were very local. If not, then we would have recorded the CS tract volley at all depths.
For the most common sites (i.e., where early biphasic potentials were
recorded; Figs. 6 and
7), we propose that the initial
negative phase measures primarily action potentials and depolarization in
preterminal and terminal axons. This is because the potential was brief—
consistent with summation of synchronous action potential currents—and
negative— consistent with a local inward current (i.e., sink). The
distribution of the amplitude of the potential along the dorsoventral axis
correlated well with refinement of CS terminations examined anatomically. At 4
wk, CS terminals have an extensive dorso-ventral termination pattern, being
present from the superficial dorsal horn to the base of the ventral horn,
including the motor nuclei (Li and Martin
2000,
2002). This is similar to the
distribution of large-amplitude potentials shown in Figs.
6 and
7. By 6 wk, there is
substantial anatomic refinement, so that terminations within the superficial
dorsal horn and the motor nuclei are eliminated (Li and Martin
2000,
2001). In the present study,
many sites in older kittens showed a restricted distribution of pre- and
postsynaptic field potentials along the dorsoventral axis (e.g.,
Fig. 7). However, overall
physiological refinement appeared to be incomplete at this age (e.g.,
Fig. 9) because there was
further refinement into week 10. The distribution of the large potentials
along the dorsoventral axis at week 13 and the adult was highly restricted
(e.g., Figs. 7,
8,
9) because the large increase
in the relative size of the potentials was present only in the intermediate
zone. This increase is likely to be partly due to the increase in local
branching and presynaptic site development that occurs later in development
(Li and Martin 2001,
2002). It is interesting that
large potentials were recorded deep into the ventral horn only in the 4- and
5-wk-old animals. We reported that there are CS terminations within the motor
nuclei in immature animals (weeks 4–5)
(Li and Martin 2000), although
these terminations are not maintained into maturity.
Role of myelination in maturation of CS function
Traditionally, the importance of myelination of corticospinal axons has
been stressed as key to development of cortical motor function. Myelination is
essential for rapid and synchronous activation of spinal neuronal populations,
which are certainly important for proper CS function. While adequate
myelination is a requirement for functional development, else spatial and
temporal summation cannot take place, it is not sufficient. Development of the
CS terminal—axon branching and presynaptic density—is obviously
key. In the cat, we have observed a good correlation between development of
limb motor control and CS axon terminal morphology (e.g.,
(Li and Martin 2001). Prior to
PW8, which is before topographic specificity is achieved, kittens have
impoverished forelimb motor skills. After week 8, skill is noticeably greater;
this is that age that we can begin to train kittens to reach (unpublished
observations; Martin et al.
2000). Clearly this improvement is not due to CS myelination,
which is well developed by week 4 (Oka et
al. 1985). Presynaptic corticospinal site development has not yet
been examined in the primate.
In the present study, we were surprised that the CS volley remained well
synchronized throughout the cervical cord at all ages. As shown in the
averages in Fig. 1B,
for a 5-wk-old animal, there was no noticeable prolongation in the duration of
the volley with conduction distance. Similarly, surface and depth recordings
(e.g., Figs. 4 and
7) revealed little shortening
with age. We were also struck by the difference between the age dependence of
volley and postsynaptic latency (Fig.
5). While the postsynaptic latency showed a progressive decrease,
volley latency was more abrupt. Between PW5 and -13 there was little change in
volley latency despite a large increase in conduction distance. This is
reminiscent of the findings of Eyre and colleagues
(Eyre et al. 1991) using TMS
in humans, showing constancy of central conduction delay during development.
This phenomenon is also present in the monkey, using TMS
(Olivier et al. 1997).
Constancy of central conduction delay is thought to be a way for the
supraspinal motor systems to better predict the time at which their control
signals act at the cord while the body is growing
(Eyre et al. 1991).
Role of development of intrinsic spinal circuits in maturation of CS postsynaptic responses
Development of the spinal postsynaptic responses reflect both the efficacy
of CS synaptic transmission and the response of spinal target neurons to this
input. Unfortunately little relevant information is available about
development of the spinal interneuronal circuits that developing CS terminals
engage. However, the spinal cord is quite mature at 4 wk because young kittens
are capable of a variety of limb and postural reflexes and can locomote
(Levine et al. 1980), albeit
clumsily. Motoneurons have developed their mature electrophysiological
characteristics and limb muscle has mature properties (Bagust et al.
1973,
1974;
Buller et al. 1960). This
mature state of the spinal cord (and skeletal muscle) reflects the fact that
the brain stem motor pathways and primary afferent fibers have developed their
spinal terminations before this age. Despite relative maturity of the spinal
cord during the time that the CS system is refining its spinal terminations,
local circuits may continue to be capable of plastic changes, especially
synaptic strength (Iriki et al.
1990). However, more needs to be learned about spinal motor
circuit plasticity before we can partition changes due to the CS synaptic
action and intrinsic spinal neurons.
Distinct early CS developmental periods for refining axon terminations and for steering motor output
Our findings point to a nonmotor function for CS activity during early
postnatal development. Interestingly, data in the human also suggest that CS
terminals can activate spinal neurons before development of the sorts of
distal skills that are generally attributed to CS control. Eyre and colleagues
(2000,
2001) have recently reported
that TMS evoked motor responses in preterm and term infants before the usual
1–2 yr of age (e.g., Koh and Eyre
1988; Nezu et al.
1997). Surprisingly, they found an early time window, <3 mo of
age, during which TMS thresholds were slightly lower than in older neonates
(Eyre et al. 2001).
Afterwards, threshold decreased with age. Using an immunological marker
(GAP-43) on autopsy material, they proposed that CS terminals were present on
motoneurons at preterm ages (Eyre et al.
2000). However, both our present findings and those of Eyre and
colleagues, showing spinal activation at early CS developmental stages, should
not be taken to mean that the CS system is poised to deliver control signals
to the spinal cord for producing muscle contraction at very young ages. On the
contrary, PT electrical stimulation and TMS produce highly synchronous motor
system activation. Natural output from motor cortex is much less synchronous
and therefore much less effective in exciting spinal circuits and motor
output. Nevertheless the ability to weakly excite spinal neurons (but without
muscle contraction) could lead to increased synaptic strength, through Hebbian
or other forms of associative plasticity
(Bailey et al. 2000).
During early postnatal development, CS terminals—like those of some
developing sensory systems and the neuromuscular junction
(Goodman and Shatz
1993)— undergo activity-dependent refinement
(Martin et al. 1999). Through
pruning and accretion, the optimal topography and density of CS presynaptic
sites becomes gradually specified (Li and Martin
2001,
2002). This process could lead
to enhancing synaptic strength sufficiently for the CS system to effectively
control particular spinal circuits.
|
|
DISCLOSURE |
|---|
|
|
|
ACKNOWLEDGMENTS |
|---|
|
|
|
FOOTNOTES |
|---|
Address for reprint requests: J. H. Martin, Center for Neurobiology and Behavior, Columbia University, 1051 Riverside Dr., New York, NY 10032 (E-mail: jm17{at}columbia.edu).
|
|
REFERENCES |
|---|
|
Armand J and Kably B. Recovery of distal skills after neonatal lesion of the sensorimotor cortex in the cat. Neurosci Lett 138: 45–48, 1992.[Web of Science][Medline]
Armand J,
Olivier E, Edgley SA, and Lemon RN. Postnatal development of corticospinal
projections from motor cortex to the cervical enlargement in the macaque
monkey. J Neurosci 17:
251–266, 1997.
Bagust J, Lewis
DM, and Westerman RA. Polyneuronal innervation of kitten skeletal muscle.
J Physiol 229:
241–255, 1973.
Bagust J, Lewis
DM, and Westerman RA. The properties of motor units in a fast and a slow
twitch muscle during post-natal development in the kitten. J
Physiol 237:
75–90, 1974.
Bailey CH, Giustetto M, Huang YY, Hawkins RD, and Kandel ER. Is heterosynaptic modulation essential for stabilizing Hebbian plasticity and memory? Nat Rev Neurosci 1: 11–20, 2000.[Web of Science][Medline]
Baldissera F, Hultborn H, and Illert M. Integration in spinal neuronal systems. In: Handbook of Physiology. The Nervous System. Motor Control. Bethesda, MD: Am. Physiol. Soc., 1981, sect. 1, vol. II, p. 509–596.
Bruce IC and Tatton WG. Synchronous development of motor cortical output to different muscles in the kitten. Exp Brain Res 40: 349–353, 1980.[Web of Science][Medline]
Buller AJ,
Eccles JC, and Eccles RM. Differentiation of fast and slow muscles in the
cat hind limb. J Physiol 150:
399–416, 1960.
Chakrabarty S and Martin JH. Postnatal development of the motor representation in
primary motor cortex. J Neurophysiol
84: 2582–2594,
2000.
Elger CE, Speckmann EJ, Caspers H, and Janzen RW. Cortico-spinal connections in the rat. I. Monosynaptic and polysynaptic responses of cervical motoneurons to epicortical stimulation. Exp Brain Res 28: 385–404, 1977.[Web of Science][Medline]
Eyre JA, Miller
S, Clowry GJ, Conway EA, and Watts C. Functional corticospinal projections
are established prenatally in the human foetus permitting involvement in the
development of spinal motor centers. Brain
123: 51–64,
2000.
Eyre JA, Miller
S, and Ramesh V. Constancy of central conduction delays during development
in man: investigation of motor and somatosensory pathways. J
Physiol 434:
441–452, 1991.
Eyre JA, Taylor
JP, Villagra F, Smith M, and Miller S. Evidence of activity-dependent
withdrawal of corticospinal projections during human development.
Neurology 57:
1543–1554, 2001.
Flament D, Hall
EJ, and Lemon RN. The development of cortico-motoneuronal projections
investigated using magnetic brain stimulation in the infant macaque.
J Physiol 447:
755–768, 1992.
Gibson CL, Arnott GA, and Clowry GJ. Plasticity in the rat spinal cord seen in response to lesions to the motor cortex during development but not to lesions in maturity. Exp Neurol 166: 422–434, 2000.[Web of Science][Medline]
Goodman CS and Shatz CJ. Developmental mechanisms that generate precise patterns of neuronal connectivity. Cell 10 Suppl: 77–98, 1993.
Iriki A, Keller A, Pavlides C, and Asanuma H. Long-lasting facilitation of pyramidal tract input to spinal interneurons. Neuroreport 1: 49–52, 1990.[Medline]
Koh TH and Eyre
JA. Maturation of corticospinal tracts assessed by electro-magnetic
stimulation of the motor cortex. Arch Dis Child
63: 1347–1352,
1988.
Levine MS, Hull CD, and Buchwald NA. Development of motor activity in kittens. Dev Psychobiol 13: 357–371, 1980.[Web of Science][Medline]
Li Q and Martin JH. Postnatal development of differential projections from the caudal and rostral motor cortex subregions. Exp Brain Res 134: 187–198, 2000.[Web of Science][Medline]
Li Q and Martin JH. Postnatal development of corticospinal axon terminal morphology in the cat. J Comp Neurol 435: 127–141, 2001.[Web of Science][Medline]
Li Q and Martin JH. Postnatal development of connectional specificity of corticospinal terminals in the cat. J Comp Neurol 447: 57–71, 2002.[Web of Science][Medline]
Lorente de Nó RA. A Study of Nerve Physiology. Studies From Rockefeller Inst. Med. Res. New York: Rockefeller Institute for Medical Research, 1947.
Maier MA,
Illert M, Kirkwood PA, Nielsen J, and Lemon RN. Does a C3–C4
propriospinal system transmit corticospinal excitation in the primate? An
investigation in the macaque monkey. J Physiol
511: 191–212,
1998.
Martin JH,
Hacking A, and Donarummo L. Impairments in prehension produced by early
postnatal sensorimotor cortex activity blockade. J
Neurophysiol 83:
895–906, 2000.
Martin JH, Kably B, and Hacking A. Activity-dependent development of cortical axon terminations in the spinal cord and brain stem. Exp Brain Res 125: 184–199, 1999.[Web of Science][Medline]
Martin JH and Lee S. Activity-dependent competition between developing corticospinal terminations. Neuroreport 10: 2277–2282, 1999.[Web of Science][Medline]
Meng Z, Li Q, and Martin JH. Postnatal development of corticospinal synapses in cats. Soc Neurosci Abstr 2002.
Nezu A, Kimura S, Uehara S, Kobayashi T, Tanaka M, and Saito K. Magnetic stimulation of motor cortex in children: maturity of corticospinal pathway and problem of clinical application. Brain Dev 19: 176–180, 1997.[Web of Science][Medline]
Oka H, Samejima
A, and Yamamoto T. Post-natal development of pyramidal tract neurons in
kittens. J Physiol 363:
481–499, 1985.
Olivier E,
Edgley SA, Armand J, and Lemon RN. An electrophysiological study of the
postnatal development of the corticospinal system in the macaque monkey.
J Neurosci 17:
267–276, 1997.
Porter R and Lemon R. Corticospinal Function and Voluntary Movement. Oxford, UK: Oxford Science, 1993.
Shinoda Y, Arnold AP, and Asanuma H. Spinal branching of corticospinal axons in the cat. Exp Brain Res 26: 215–234, 1976.[Web of Science][Medline]
Theriault E and Tatton WG. Postnatal redistribution of pericruciate motor cortical projections within the kitten spinal cord. Dev Brain Res 45: 219–237, 1989.[Medline]
This article has been cited by other articles:
|
|
 |
|
  I. Salimi, K. M. Friel, and J. H. Martin Pyramidal Tract Stimulation Restores Normal Corticospinal Tract Connections and Visuomotor Skill after Early Postnatal Motor Cortex Activity Blockade J. Neurosci., July 16, 2008; 28(29): 7426 - 7434. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. M. Friel and J. H. Martin Bilateral Activity-Dependent Interactions in the Developing Corticospinal System J. Neurosci., October 10, 2007; 27(41): 11083 - 11090. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Kamiyama, N. Yoshioka, and M. Sakurai Synapse Elimination in the Corticospinal Projection During the Early Postnatal Period J Neurophysiol, April 1, 2006; 95(4): 2304 - 2313. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. H. Martin The Corticospinal System: From Development to Motor Control Neuroscientist, April 1, 2005; 11(2): 161 - 173. [Abstract] [PDF]
|
|
|
|
|
|
I. Salimi and J. H. Martin Rescuing Transient Corticospinal Terminations and Promoting Growth with Corticospinal Stimulation in Kittens J. Neurosci., May 26, 2004; 24(21): 4952 - 4961. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Z. Meng, Q. Li, and J. H. Martin The Transition from Development to Motor Control Function in the Corticospinal System J. Neurosci., January 21, 2004; 24(3): 605 - 614. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
TOP
ABSTRACT
INTRODUCTION
