| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
The Journal of Neurophysiology Vol. 88 No. 1 July 2002, pp. 86-97
Copyright ©2002 by the American Physiological Society
Department of Physiology, The University of Arizona, College of Medicine, Tucson, Arizona 85724-5051
 |
ABSTRACT |
|---|
 TOP
 ABSTRACT
 INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Hornby, T. George,
Jennifer C. McDonagh,
Robert M. Reinking, and
Douglas G. Stuart.
Effects of Excitatory Modulation on Intrinsic Properties of
Turtle Motoneurons.
J. Neurophysiol. 88: 86-97, 2002.
The purpose of this study was to quantify the
effects of excitatory modulation on the intrinsic properties of
motoneurons (MNs) in slices of spinal cord taken from the adult turtle.
Responses were noted following application of an excitatory modulator:
serotonin (5-HT), muscarine, trans-1-amino-1,3-cyclopentane
dicarboxylic acid (tACPD), or all three combined. A sample of 44 MNs
was divided into 2 groups, on the basis of whether MNs did (28/44) or
did not (16/44) demonstrate a nifedipine-sensitive acceleration of discharge during a 2-s, intracellularly injected stimulus pulse. Such
acceleration indicates the development of a plateau potential (PP).
Excitatory modulation lowered the MNs' resting potential, increased
input resistance, decreased rheobase, reduced several afterhyperpolarization values, and shifted the conventional, one-phase stimulus current-spike frequency (I-f) relation to
the left. For both MN groups, the relative efficacy of excitatory
modulation on both non-PP and PP MNs was generally in the following
order: combined application > 5-HT 
muscarine > tACPD. In many instances, the effects of modulation differed
significantly for non-PP versus PP MNs, the most pronounced being in
their I-f relation. To describe this difference, it was
necessary to measure a two-phase relation. In PP MNs, excitatory
modulation considerably increased the slope of the first (initial)
phase and flattened the second (later) phase of this relation. The
latter result bore similarities to that obtained in a previous study,
which addressed MN firing behavior during fictive locomotion of the
high-decerebrate cat.
| |
INTRODUCTION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
It is well-known that various
endogenous neurotransmitters, modulators, and their agonists can
substantially alter the sub- and/or suprathreshold ionic conductances
responsible for the repetitive discharge of motoneurons (MNs) (for
review, Binder et al. 1996
; Powers and Binder
2001; Russo and Hounsgaard 1999). Previous
studies have suggested that the extrinsic modulation of MN conductances active below the action potential (AP) threshold can also alter the
threshold for repetitive firing, thereby causing a shift in the bias of
the stimulus current (I)-spike frequency
(f) relation. Conversely, modulators that
alter conductances activated near or above the AP threshold may change
the slope (gain) of this relationship (Binder et al.
1993). In both cases, modulation of the I-f relation
changes the net synaptic current necessary to produce a given muscle force.
Many MNs possess nonlinear membrane properties that are revealed only
following modulation by neuroactive agents. For example, the plateau
potential (PP; due to a persistent inward current), which is defined as
a sustained depolarization or discharge following a brief excitatory
stimulus (for historical review, see Hornby et al.
2002a) becomes manifest in MNs only following the extrinsic application of a variety of excitatory modulator agonists, and/or following the blockade of outward currents. As a result, excitatory modulation may result in the MN responding nonlinearly to
intracellulary injected and/or synaptic current.
Some of the above effects have been shown previously in a variety of in vivo and in vitro preparations but without provision of quantitative measurements of repetitive firing (see the above reviews). For this reason, little is known about the robustness of these effects, and their functional significance during natural, near-natural, and fictive movements.
In this study, intracellular recording and stimulation of MNs was
undertaken in slices of turtle spinal cord, using the in vitro
techniques of Hounsgaard et al. (1988b). Intrinsic MN
properties, with an emphasis on the I-f relation, were
measured in the control versus modulated condition. The latter was
shown to have a powerful effect on the I-f relation,
particularly in MNs that were generating PPs. It is argued that the
present results are relevant to a previous study in which repetitive MN
firing was measured during fictive locomotion in the high-decerebrate
cat preparation (Brownstone 1989; Brownstone et
al. 1992). Preliminary accounts of some of our results have
appeared previously in abstract form (Hornby et al.
1997, 1998, 2000; Stuart
et al. 1999).
This work was part of the Ph.D. dissertation research of T. G. Hornby.
| |
METHODS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Most of the present techniques were recently described in detail
by our laboratory (Hornby et al. 2002b; McDonagh
et al. 1998a). All procedures were in conformity with
university, state, and federal regulations for the care and use of
laboratory animals.
Dissection and slice preparation
Spinal cord slices (2 mm thick) were obtained from the adult,
North American pond turtle, Pseudemys (Trachemys) scripta
elegans, while deeply anesthetized with pentobarbital sodium, and
perfused intracardially. The slices were maintained in oxygenated
turtle physiological saline at room temperature (25-26°C) for 2-3 h
prior to the recording session. Throughout recording, the test slice was at the same temperature, and perfused continuously with the same
solution at 1 ml/min. At the end of each day's recording session (5-8
h), the slices were refrigerated (4°C) overnight in a sealed
container of turtle physiological saline that had been previously
saturated with 98% O2-2%
CO2. For recording on subsequent days (i.e., 
48
h following surgery), the slices were warmed to room temperature.
Recording and measurement procedures
Intracellular potentials were recorded in spinal MNs at depths
of ~100-300 µm from the cut surface of the slice, using sharp microelectrodes of thin-wall borosilicate glass with filament (1.5 mm
OD). Electrodes were filled with 1 M K+ acetate.
Successful penetration of a spinal MN was characterized by a rapid
negative shift in membrane potential, and maintenance of a resting
potential more negative than 
60 mV, thereby precluding excessive
damage to the cell. The cell sample was restricted to those requiring
an injected current >0.4 nA for the generation of APs. This precluded
the testing of ventral-horn interneurons (see following text for
further MN criteria). The membrane potential of each test MN was
amplified, filtered (0-3 kHz), and recorded on both a tape recorder
and a computer.
MEMBRANE PROPERTIES.
Spinal MNs were studied for an average of 51 min (range, 19-218 min).
Passive (cell-at-rest) properties measured included resting membrane
potential (in mV), input resistance (in M
), and membrane time
constant (in ms). The "steady-state" (not peak) input resistance
was measured as the slope of the linear regression of a voltage
response to four, 2-s constant-current pulses (2 depolarizing, 2 hyperpolarizing; each an average of 4 successive sweeps) of varying
magnitude, which were applied to the cell at its resting potential. The
time constant was averaged from the decay of the voltage response at
the termination of four current steps used during input resistance determination.
-c), because turtle MNs in a SC slice do
not exhibit Kernell's secondary range: i.e., an abrupt increase in
f/I slope for firing rates at higher stimulus strengths
(Hornby et al. 2002a; McDonagh et al.
1998a; cf., however, Hounsgaard et al. 1988b). Note further that the profile of the steady state I-f
relation across vertebrate MNs is somewhat sigmoidal, with the
f/I slope becoming progressively lower at higher current
intensities (e.g., in the lamprey, Buchanan 1993;
turtle, Hounsgaard et al. 1988b; and cat, Granit
et al. 1963; Kernell 1965b). For the issues
addressed since Kernell's seminal (Kernell 1965a-c)
reports on cat MNs, it has been considered sufficient to approximate
the relation as a linear one (e.g., Binder et al. 1993;
Kernell 1999; Powers and Binder 2001).
For the present study, too, a linear approximate was appropriate. In
addition, however, it was necessary to divide the f/I slope
into two phases, as described below.
TWO-PHASE MEASUREMENT OF THE F/I SLOPE. Following application of excitatory modulators, many of our turtle MNs exhibited PP behavior (see following text). The result was a transition point in their I-f relation, with the pretransition phase markedly steeper (greater) than the posttransition one. This behavior necessitated a two-phase measurement of the f/I slope as shown in Fig. 1.
|
). In the test (modulated)
condition when PP behavior was present,
Itrans-ftrans
measurements were made by visual inspection, because there was usually
a clear-cut transition point in the I-f relation. This point
was obscure for all cells in the control state, and in those cells that
did not display modulator-induced PP type behavior. In such case,
Itrans was denoted as the current
midway between Imin and
Imax, and
ftrans was denoted as the
spike-frequency at that current. For the control versus test
(modulated) condition, this latter
Itrans value relative to the
Imin value (162%) of non-PP
generating cells was quite close to the corresponding value of the
PP-generating cells (167%). Our latter
Itrans procedure had the advantage of
allowing an objective (nonarbitrary) statistical comparison between the
initial, steeper f/I slope of the non-PP generating cells
(both in the control and modulated state) versus those with PP behavior
in which the shift of the transition point was dramatic (see following
text). Note further that the term "transition point" refers only to
a change in the slope of the I-f relation. It gives no
direct indication of the onset and/or time course of accelerated
(PP-influenced) MN discharge both within and across the various
stimulus current steps that were used in this study.
The pretransition f/I slope was calculated as the regression
line of all the data before and including the transition point. Similarly, the posttransition f/I slope was calculated as
the regression line of all the data after and including the transition point. The f/I slope ratio was calculated as the latter
slope divided by the former one. This parameter provided an indication of the percentage change in the gain of the I-f relation.
IDENTIFICATION OF MNS.
For largely anatomical reasons, it is not practical to identify MNs by
the antidromic activation of the short, highly dispersed ventral-root
filaments in a slice of turtle spinal cord (McDonagh et al.
1998a; see their Fig. 1). Further, some MNs may have their axons cut when making the test slices. All the tested cells of the
present study had an input resistance value >2.6 M, a rheobase current >0.4 nA, and an f/I slope <50 Hz/nA. Previously,
our laboratory has shown that these values clearly separate MNs from
interneurons using both electrophysiological criteria (McDonagh
et al. 1998a) and morphological evidence from reconstructions
of stained MNs and interneurons (McDonagh et al. 1998b,
1999a). It is conceivable, but not likely, that these
cells could have been segmental or ascending-tract interneurons. For
example, in our previous morphological study of the soma diameter of
>350 µm turtle MNs, we saw virtually no other cells of comparable
diameter in the ventral horn (Callister et al. 1996; see
also Ruigrok et al. 1984).
Pharmacological procedures
Excitatory modulators were applied individually and in
combination at concentrations consistent with previously published results and/or at doses that (at least in our own work) produced maximal alterations in MN behavior without damage to the cell or the
onset of spontaneous discharge. We used the term "excitatory" to
denote the net facilitative effect (increased input resistance, decreased rheobase, increased gain of the I-f relation) of
these agents on neuronal excitability. The agents were first made up as
high-concentration stock solutions in standard turtle physiological saline, and later diluted to the final concentration. They included serotonin (5-HT; 100 µM) (Hounsgaard and Kiehn 1989),
muscarine (20 µM) (Svirskis and Hounsgaard 1998), and
trans-1-amino-1,3-cyclopentane, dicarboxylic acid (t-ACPD;
20 µM). While both t-ACPD (Svirskis and Hounsgaard
1995) and cis-ACPD (Svirskis and Hounsgaard
1998) have been reported to modulate MN behavior, our
preliminary experiments showed that application of t-ACPD was
appropriate for our present purposes; i.e., it substantially modulated
MN properties, including the production of PP-like behavior.
Intrinsic MN properties were first determined in control bathing medium, then remeasured following addition of a modulator, or combination of modulators, to the bathing medium. In preliminary experiments, measurement of MN properties at 6-8 versus 30 min following modulator application revealed only small differences in properties. Therefore statistical comparisons were made between the initial control MN properties versus those measured after 6-8 min of modulator application. Note further that measurements made in a subsequent control bathing medium 1-2 h following the test measurements were almost identical to the initial control ones. Some residual modulatory effects were sometimes noted, however. For this reason, the present procedure was to use separate SC slices for the measurements made on each single MN.
To investigate PP behavior further, the ionic channel blockers TTX (1 µM; a blocker of the fast Na+ channel) and the
dihydropyridine, nifedipine (15 µM; a partial blocker of the L-type
Ca2+ channel) (Hounsgaard and Mintz
1988) were applied for 30 min before the properties of MNs
were reinvestigated. Although higher nifedipine concentrations (50
µM) have been used to block the L-type conductance in turtle spinal
neurons (Russo and Hounsgaard 1994), we used lower
concentrations to reduce the possibility of an N-type
Ca2+ blockade, as has been observed in other
vertebrate neurons (Jones and Jacobs 1990;
Wilkinson and Barnes 1996).
Both blocking agents were dissolved in DMSO and diluted with
physiological saline to appropriate concentrations. Preliminary experiments revealed no changes in intrinsic MN properties following application of DMSO alone. This finding was consistent with previous studies showing that DMSO alone has no significant effect on intrinsic vertebrate neuron conductances, but rather altered glutamatergic (Lu and Mattson 2001) and cholinergic (Kubota et
al. 1998) transmission.
Identification of PPs
BACKGROUND.
Previously, Russell and Hartline (1982) have provided 12 tests, which, in various combinations, have been used to identify PPs
(see also Hartline and Graubard 1992). For turtle MNs,
Hounsgaard et al. (1986) utilized a subset of these
tests for verification of the presence of a PP. Importantly,
Hounsgaard and Kiehn (1989) have emphasized the value of
a single criterion, wherein the MN exhibits an
acceleration, rather than adaptation, of its discharge in
response to a constant-current stimulus. This acceleration was shown to
be due primarily to an increase in an L-type Ca2+
conductance sensitive to the dihydropyridines (Hounsgaard and Kiehn 1989; Svirskis and Hounsgaard 1998). The
significance of this criterion is reinforced by the previous literature
across vertebrates on MN spike-frequency adaptation, because
it has shown consistently that in non-PP generating
vertebrate MNs, firing rate progressively decreases throughout the
constant-stimulus period (e.g., Kernell and Monster
1982; Powers et al. 1999).
PRESENT APPROACH. In 64% of the cells tested, discharge following excitatory modulation was characterized by the appearance of accelerated and maintained discharge (as discussed above) before and/or during the 2nd s of the 2-s stimulus pulses. We interpreted this change as commensurate with PP-like behavior (hereafter termed PP behavior). Similarly, we observed a marked alteration in the I-f relation from near-linearity (1-phase) to a two-phase relation with a steep initial slope followed by a shallower one at higher current intensities. Although this behavior was concomitant with PP behavior, the criterion for PP manifestation was the acceleration of frequency during a 2-s current pulse. This interpretation required some control experiments using previously established criteria (see above). For example, Fig. 2 demonstrated a modulator-induced, nifedipine-sensitive accelerated MN discharge during a 2-s, constant-current intracellular stimulus.
|
, and it was obtained in
several MNs following application of excitatory agents. We further
obtained qualitatively similar results for some TTX-applied MNs, in
which a gradually increasing depolarization of membrane potential was observed during a constant-current, 2-s stimulus (for details, see
Hornby 2000). Based on the above findings, and
consistent with the previous description of PPs in the in vitro turtle
SC slice (Hounsgaard and Kiehn 1989), PP MNs were
identified as those cells demonstrating accelerated firing responses
during the 2-s, constant-current intracellular stimulus pulses used for
determination of the I-f relation.
MEMBRANE POTENTIAL OSCILLATIONS.
For 3/46 MNs, oscillatory membrane behavior was observed following
excitatory modulation, as has been observed previously following
application of muscarine to turtle MNs in a spinal cord slice
preparation (Guertin and Hounsgaard 1999;
Svirskis and Hounsgaard 1998). In two of three cells,
which were excluded from this study, such activity was manifested
subthreshold (i.e., without depolarizing current injection), and stable
recording was not maintained for determination of the full array of
electrophysiological properties. The third cell was included because
its oscillatory behavior was observed only during long-duration (30-s)
intracellular stimulation. Since the protocol for determination of
intrinsic MN properties involved the use of current stimuli 2 s in
duration, this cell's measured properties were not affected by the
oscillatory behavior.
Statistical analysis
Due to the size-related variability of most intrinsic MN
properties (Zengel et al. 1985), changes in measured
parameters are expressed below as a percentage of the control values.
The significance of these differences for the control versus test
condition was evaluated by use of standard paired t-test.
The significance of differences in the relative extent of modulation by
the various modulators was tested by the use of a standard unpaired
t-test. In both cases, significance was noted at
P < 0.05 and <0.01.
| |
RESULTS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
The database for this study consisted of 44 MNs whose properties
were within the range of MNs values reported previously (Hornby et al. 2002b; McDonagh et al. 1998a). AHP
measurements could not be made accurately in 2/44 MNs, a problem
encountered in previous reports from this and other laboratories (e.g.,
Kernell 1966; McDonagh et al. 1998a). For
two other MNs, PP behavior began at Imin, as has been seen previously
(Bennett et al. 1998b) for PPs activated at or below the
threshold for AP generation. For these two latter cells, values are
reported for only the one-phase I-f relation (i.e., the data
point for
Itrans-ftrans
was considered to coincide with that for
Imin-fmin).
The results are reported below in the order of the effects of
excitatory modulation on the following: 1) the 44-cell
sample to a single modulator or combination thereof, 2)
16/44 non-PP versus 28/44 PP MNs, and 3) selected low-
versus high-threshold PP MNs. The analysis featured systematic, tabular
statistical data, with one table presented below and another five
available on request (i.e., Tables 4.1B to 4.3A
in Hornby 2000).
Effects of excitatory modulation on MNs
PASSIVE AND TRANSITIONAL PROPERTIES.
Excitatory modulation reduced the resting membrane potential (range of
the 4 mean modulator effects, 1.2-6.8%), increased the input
resistance (17-32%), decreased rheobase current (23 to 38%), and
decreased the values for the four AHP parameters (0.1 to 23%). The
combined relative extents of change and their statistical significance
were in the parameter order: rheobase current (3/4 modulator-induced
changes significant at P < 0.05) > input
resistance (2/4 significant) > slow AHP amplitude (3/4 significant) > slow AHP half-time decay (2/4 significant) > AHP duration (1/4 significant) > resting potential (2/4
significant) > membrane time constant and fast AHP amplitude
(consistently slight and insignificant effects). The relative efficacy
of the four modulator applications was in the order as follows: 5-HT (significant changes in 6/8 parameters) > combined 3 modulators (5/8 parameters) > muscarine (2/8) > t-ACPD (consistently
modest and relatively insignificant effects).
ACTIVE PROPERTIES: ONE-PHASE I-F RELATION.
The effect of excitatory modulation was relatively stronger on the five
parameters of the one-phase I-f relation than on the cells'
passive and transitional properties. The values of
Imin and
fmax were particularly affected, with
Imin reduced significantly for three
of four modulator applications (range of the 4 mean modulator effects,
21 to 42%), and fmax increased
significantly for all four applications (23-59%). The overall effect
of modulation was to shift the I-f relation to the left and
upward, with a lesser (but consistent) increase in the slope of the
relation. These changes were similar for 5-HT, muscarine, and the
combined modulators, with all such effects much greater than those
evoked by t-ACPD.
ACTIVE PROPERTIES: TWO-PHASE I-f
RELATION.
For the five additional parameters of the two-phase I-f
relation, Fig. 3 shows that only the
posttransition f/I slope was not altered significantly by
the four-modulator applications (mean changes, 0.8-19%). The changes
were significant for Itrans (14 to
31%, 3/4 applications), ftrans
(30-90%, 3/4 applications), pretransition f/I slope
(55-238%, 4/4 applications), and f/I slope ratio (35 to
55%, 4/4 applications). The result of these changes was to not only
shift the I-f relation to the left and upward, but to also
elicit a more pronounced two-phase profile of the I-f
relation. The relative extent of these changes was in the order as
follows: muscarine = combined modulators > 5-HT > t-ACPD.
|
D, for
the effects of 5-HT and muscarine on non-PP versus PP MNs (with similar
results obtained for t-ACPD and the combined 3 modulators).
|
D,
comparison of non-PP versus PP cells. 1) The modulators had
a greater effect on the I-f relation of PP versus non-PP
cells. This was attributable largely to the greater modulator-induced
reduction in Itrans values and
increase in ftrans of PP cells. The
result was to increase the pretransition f/I slope, flatten
the posttransition f/I slope, thereby reducing the
f/I slope ratio of the PP cells to a greater degree than
that observed in non-PP cells. 2) The modulators increased the fmax value to the same relative
extent for PP versus non-PP cells, thereby indicating that the maximum
discharge capability of the MNs was not dependent on the presence of a
PP. 3) The relative efficacy of the modulators was generally
in the order as follows: combined modulators > muscarine = 5-HT > t-ACPD.
EFFECTS OF NIFEDIPINE ON PP MNS.
The effects of nifedipine were two-fold on PP-generating MNs. First, it
consistently reduced but did not abolish the acceleration of MN
discharge during the 2-s depolarizing current (Fig. 2C
above). Second, nifedipine consistently modified the modulator-induced two-phase I-f relation profile back toward but not
reaching the one-phase profile (i.e., similar to the control
condition). This change (Fig. 2D) involved a migration of
the transition point downward (decreased
ftrans) and to the right (increased
Itrans), as compared with the
PP-generating state. Previous studies have shown similar decreases in
the acceleration of discharge following the application of nifedipine
(Hounsgaard and Kiehn 1989; Hounsgaard and Mintz
1988). Such demonstrations of changes in the I-f
relations are quite rare, however (e.g., as in Hounsgaard and
Mintz 1988). Our results suggest that the alteration in the
I-f relation in PP MNs is not solely dependent on
manifestation of PP behavior. Rather, other conductances must also
contribute, at least in part, as occurs in the non-PP cells following
extrinsic modulation (see following text).
GROUPED EFFECTS OF THE FOUR EXCITATORY MODULATORS.
The two samples of non-PP versus PP cells in Fig. 4, A-D,
together with those for t-ACPD and the combined modulators, were often
too small (range, 2-9 cells) for a statistical verification of trends.
Previous reports have emphasized, however, that the mechanisms by which
the three presently used modulators induce PP behavior are relatively
similar. In particular, previous results support the often-stated
argument that changes in PPs are dependent, at least in part, on an
alteration in outward K+ currents
(Hounsgaard and Kiehn 1989; Svirskis and
Hounsgaard 1998). While the possibility of direct modulation of
the L-type Ca2+ conductance by excitatory
agonists is still open to debate (Russo and Hounsgaard
1999), the following analysis grouped the responses of the 16 non-PP versus the 28 PP MNs based on the likelihood of similar
mechanisms being responsible for PP generation.
22 ± 22%) and PP (33 ± 22%) MNs (P < 0.01), there was a
significant difference in the extent of alteration between the groups
(P < 0.05). Strikingly, the changes induced in AHP
parameters were quite different for non-PP versus PP MNs (details in
Hornby et al. 2001a). For PP MNs, only the value of slow
AHP amplitude was changed significantly (11 ± 12%,
P < 0.01). In contrast, the values for all four AHP parameters in non-PP MNs changed significantly (range, 9 to 22%; P < 0.01). Accordingly, differences in the extent of
change in AHP properties (other than slow AHP amplitude) were
significantly greater for non-PP MNs (P < 0.05; see
also Hornby et al. 2001a).
Active properties. Analysis of the one-phase I-f relation.
Table 1 shows that for three of five
parameters of the one-phase I-f relation
(Imin,
fmax, 1-phase f/I slope),
the grouped modulator effect was significant (20-30%) but similar for
both non-PP and PP cells. For the remaining two parameters,
fmin and Imax, this effect was significant for
PP (60 and 10%, respectively) but not non-PP cells. As a result, the
overall modulator effect on the one-phase I-f relation was
to shift the relationship upward and leftward. These effects occurred
to a greater degree in PP versus non-PP cells. There was virtually no
difference in the extent of change in the one-phase f/I
slope between the two cell groups, however.
|
27% vs.
non-PP, 14%). For the latter, note that the negative PP change was
significant, whereas the positive non-PP change was not. Accordingly,
the change in the f/I slope ratio was also far more
pronounced for PP versus non-PP MNs. To further emphasize the
flattening of the f/I slope beyond the transition point, for
7/28 of the PP cells, the posttransition f/I slopes were
<2.0 Hz/nA (vs. 10-15 Hz/nA in the control state). In addition, in
10/28 PP MNs, excitatory modulation induced a negative
posttransition f/I slope for greater than or equal to three
consecutive increasing stimulus current steps, which were applied well
before that producing the Imax value.
None of the entire population of cells exhibited this behavior in the
control condition.
Summary.
There are five features to the above grouped non-PP versus PP MN
comparisons. 1) The passive properties of non-PP versus PP cells were essentially similar, and they did not predict the
differences observed in their I-f relation. 2)
There were significant differences in the effects of modulation on the
transitional properties of non-PP versus PP MNs, with greater effects
on the AHP of non-PP MNs, and the rheobase current of PP MNs.
3) Differences in rheobase were correlated with the extent
of the leftward shift in the I-f relation (i.e., thereby
lowering the Imin value) of the two
cell groups. 4) There was a significantly greater
modulator-induced alteration in the transition point of PP versus
non-PP MNs. It included a greater increase in
ftrans values (albeit not
significant), and a significant leftward shift (i.e., greater decrease)
in Itrans values. This difference in
the relative modulation of the transition point underscored the change
in the fundamental I-f relation from a more-traditional
linear one-phase I-f relation to the newly presented
two-phase one. 5) A substantial proportion (10/28) of PP MNs
exhibited a negative input-output transformation of stimulus current to spike frequency during a component of the
posttransition-phase of the I-f relation.
Effects of excitatory modulation on low- versus high-threshold PP MNs
In recent reports on PP behavior in hindlimb MNs in the
decerebrate cat, Lee and Heckman (1998a,b) reported a
greater propensity and extent of PP behavior in low- versus
high-threshold MNs. The present experimental protocol was not designed
to address this issue. Nonetheless, one analysis demonstrated an
interesting difference, which is shown in Fig.
5. It compares the effects of modulation of the I-f relation of the 5/28 lowest-threshold PP MNs to
that of the 5/28 highest-threshold PP MNs.
|
Three relevant properties had widely different control values for the
low- versus high-threshold PP cells of Fig. 5. These mean values were,
respectively, input resistance (M), 23 versus 4.4; rheobase current
(nA), 0.6 versus 5.2; and, one-phase f/I slope (Hz/nA), 16 versus 4.0. Clearly, the comparison was of the near extremes of the
low- versus high-threshold cells for turtle hindlimb MNs (Hornby
et al. 2002b; McDonagh et al. 1998a).
Figure 5 shows that the low-threshold cells were modulated to a greater extent in their spike-frequency responses, particularly ftrans (84% increase vs. 9% for high-threshold cells) and fmax (27 vs. 6% increase). In contrast, the high-threshold cells exhibited relatively greater changes (by 3- to 8-fold) in their Imin, Itrans, and Imax (values in Fig. 5 legend) and in the steepening of their pretransition slope (409% increase vs. 206% for low-threshold cells), flattening of their posttransition f/I slope (29 vs. 9% reduction), and lowering of the slope ratio (66 vs. 54% reduction). These comparisons suggest that the responses of the two sets of cells to excitatory modulation were qualitatively different.
| |
DISCUSSION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
The key finding of this study was the powerful effect of
excitatory modulation on the I-f relation of spinal MNs in
the adult turtle, particularly in cells exhibiting PPs. This finding is the main point of the discussion, with emphasis on its relation to the
previous results of Brownstone et al. (1992) on the
high-decerebrate cat, and relevant subsequent results on the PP and MN
firing regulation. First, however, it is appropriate to discuss the
relationship between the present results on modulation of the MN
passive and transitional properties and previous studies in vertebrate
preparations. The emphasis is on the responses of non-PP versus PP MNs,
a comparison that has not appeared in previous literature on this
topic. This comparison also raises the issue of why some but not all
MNs displayed PPs in the present study.
Discussion of the present work is restricted to the grouped
effects of the three excitatory agents. The sample sizes were too small
to draw quantitative inferences regarding the relative efficacy of
5-HT, muscarine, and tACPD on non-PP versus PP MN discharge. In most
cases, however, a combination of all three agents was more effective
than any single agent applied alone. This finding was advantageous for
the present intent, however, because the summed effects of the three
agents lead to clear-cut changes in the I-f relation, particularly when
comparing the grouped responses of non-PP versus PP MNs. Our emphasis
below on these grouped responses is also supported by several previous
reports on both turtle and cat preparations (Hounsgaard and
Kiehn 1989; Schwindt and Crill 1980a-c;
Svirskis and Hounsgaard 1998). These reports emphasized
that different modulators have the same general qualitative effect on
at least one of the mechanisms involved in PP generation; i.e., a
reduction of outward K+ conductances. It is
conceded, however, that much further work is required on this issue.
Effects of excitatory modulation on passive and transitional MN properties
The present work supported previous findings demonstrating a
modulator-induced reduction in resting potential, increase in input
resistance, reduction in rheobase current, and reduction in various AHP
parameters (for review: Binder et al. 1996;
Powers and Binder 2001; Russo and Hounsgaard
1999). These changes could result primarily from alterations in
two key potassium conductances. First, and most consistently across
studies and vertebrate species, is a reduction of the resting
K+ conductance, noted primarily by a decrease in
resting potential and increase in input resistance. Application of 5-HT
has generated such effects in rat spinal (Elliott and Wallis
1992; Wang and Dun 1990), phrenic
(Lindsay and Feldman 1993), and facial (Larkman and Kelly 1992) MNs, and in cat (White and Fung
1989) MNs. Similar results have also been noted in mouse spinal
neurons (Nowak and MacDonald 1983), and turtle
(Svirskis and Hounsgaard 1998) and cat
(Zieglgansberger and Reiter 1974) MNs following
modulation by muscarine. Second, and perhaps less consistently, is a
reduction in the KCa-s current, which is
primarily responsible for the slow component of the AHP, and which may
also be active in a MN at rest. Modulation of AHP amplitude and
duration following application of 5-HT has been observed in spinal MNs
of the lamprey (Wallén et al. 1989), turtle
(Hounsgaard and Kiehn 1989), and cat (White and
Fung 1989), and in rat hypoglossal MNs (Berger et al.
1992). Depression of the AHP following modulation by muscarine
has been observed in hippocampal (Fraser and MacVicar
1996; Madison et al. 1987) and neocortical
(Schwindt et al. 1988) cells and has been
thought to indicate a reduction in K+
conductances, as well.
Effects of excitatory modulation on the I-f relation
The quantitative differences between the modulator-induced changes
in the I-f responses of non-PP versus PP MNs have not been reported previously. The major differences were due largely to the
greater effect of excitatory modulation on the transition values
(Itrans-ftrans)
associated with PP behavior. The changes in these parameters were much
more substantial in PP versus non-PP MNs, and they were reflected in
the greater modulator-induced changes in the pre- and posttransition
f/I slope values between cell types. Note also that the
effect of excitatory modulation on non-PP cells was to
increase their posttransition f/I slope, whereas
the corresponding effect on PP cells was a significant reduction in this slope. Indeed, for 7/28 of the PP cells,
the modulated posttransition f/I slope value was <2.0
Hz/nA, and 10/28 PP MNs demonstrated a negative
posttransition slope value for greater than or equal to three
consecutive steps of increasing stimulus strength. This latter finding
is particularly relevant to the previous high-decerebrate cat results
of Brownstone et al. (1992) (see following text).
Mechanisms underlying generation of the two-phase I-f relation in PP MNs
In the present study of the I-f relation, we used a
current-step paradigm like the classical one of Granit et al.
(1966) and Kernell (1965a-c,
1966, 1979, 1999); i.e.,
step-wise increasing, 2-s current pulses at 0.1 Hz. Our procedure may
have brought on the "warm-up" phenomenon (Russo and
Hounsgaard 1994), which can include a decreasing threshold
for PP activation near the AP threshold (Bennett et al.
1998b). The resultant I-f relation should then be
characterized by a pronounced increase in the f/I slope and a shift of the transition point to a lower level of stimulus intensity. These changes were both demonstrated in the present work. Two recent
studies on decerebrate cat MNs provided additional, compelling evidence
for pronounced modulator-induced alterations in the I-f relation following PP activation (Bennett et al. 1998a;
Lee and Heckman 1998a). Their results were qualitatively
similar to the present ones, but a further comparison must await a
study that evaluates the contribution of the stimulation paradigm to
the I-f relation of PP MNs; e.g., by comparing the effects
of current-step versus triangular-wave paradigms in the same test cells.
The pronounced modulator-induced flattening of the f/I slope following the transition point was a key feature of our present results. While all our present PP MNs demonstrated profound reductions in the f/I slope after the transition point, negative f/I slopes were observed for only a few data points, and posttransition f/I slopes were always >0 Hz/nA.
Three mechanisms may contribute to the flattening of the f/I
slope observed following generation of PP behavior. 1) The
mechanism underlying PPs is predominantly a persistent inward current
resulting in a sustained depolarization ~20-30 mV less negative than
the resting membrane potential (Hounsgaard and Mintz
1988). In dorsal horn cells of the turtle spinal cord slice,
Russo and Hounsgaard (1994) have also demonstrated a
two- to fourfold reduction in input resistance concurrent with PP
activation. Although not tested directly in our present sample, such an
alteration in input resistance would render additional synaptic or
intracellular current impressed on the MN less effective in altering
discharge frequency (Lee and Heckman 1998a; cf.,
however, Bennett et al. 1998a,b). 2)
Regulation of intracellular Ca2+ levels may
control PP activation and alter MN discharge rate during intracellular
stimulation. Previous studies have demonstrated reductions in
Ca2+ currents via elevation of intracellular
Ca2+ levels (Tillotson 1979), or
second messengers generated by a ryanodine receptor (Chavis et
al. 1996; Nakai et al. 1996). The regulation of
intracellular Ca2+ may suppress PP activity
following its generation, thereby decreasing discharge frequency and
flattening of the f/I slope. Such behavior could also
account for the modulator-induced negative f/I slope that
was observed in 10/28 of our PP MNs. 3) Saturation of MN discharge capability has been observed previously in cat MNs during strong intracellular current injection (Kernell 1965c),
sensory afferent stimulation (Burke 1968; Cordo
and Rymer 1982), and electrical stimulation of brain stem
centers (Tansey and Botterman 1996; Zajac and
Young 1980). Such "rate-limiting" behavior of MNs
has also been observed during voluntary activation in both cats
(Hoffer et al. 1987) and humans (De Luca et al.
1982; Monster and Chan 1977). Saturation of MN
discharge capability is a less likely possibility for the present
results, however, because fmax was always greater than ftrans, and the
overall f/I slope never reached 0 Hz/nA. The mechanisms for
the observed rate-limiting in other studies may be linked to PP
generation, however (for review, see Hornby et al.
2002a).
In summary, at least three mechanisms, and likely further
as-yet-undetermined ones, can account for the pronounced rate-limiting behavior observed in the PP MNs of the present study. These same mechanisms may also explain the Brownstone et al. (1992)
results, as discussed in the following text.
Ubiquity of PP behavior
It is not clear whether all vertebrate MNs can generate PPs. Early
in the cellular/systems-level PP literature, it was argued that PPs
should assist MNs involved in postural tasks because their proportion
was found to be greater in MNs innervating extensor musculature
(Hounsgaard et al. 1988a). Similarly, this value was emphasized in the prediction that they should be more prevalent in
proximal versus distal limb muscles, and utilized more in tasks requiring static versus dynamic force production (Heckman and Lee 1999b). Alternatively, PPs have been implicated in
segmental interneuronal and MN behavior during rhythmically changing
pattern-generating activity (Hartline et al. 1988;
Kiehn 1991; Selverston 1999). To this
point, no quantitative evidence is available on the relative occurrence
of PPs in MNs supplying different muscles, and when MNs are involved in
different tasks (for further review, see Hornby et al.
2001b).
For the adult cat, early research on a persistent inward current
evident in anesthetized preparations indicated that such behavior might
be present in ~50% of MNs studied (Schwindt and Crill
1980a-c). More recently, there is evidence that PPs are more
fully developed and longer lasting, and hence more efficacious, in
spinal, low-threshold MNs (Lee and Heckman 1998a,b).
Other evidence has shown, however, that high-threshold cat MNs can also generate substantial PPs (Bennett et al. 1998a). The
present work shows that very-high threshold MNs are just as
prone to exhibit PPs, as are low-threshold ones (Fig. 5). Note further
that Lee and Heckman (2000) have recently demonstrated
that most MNs studied under modulatory influences did not generate
self-sustained PPs, although such input allowed strong amplification of
excitatory (Ia) synaptic input in all the tested cells.
The purpose of the various studies cited above was not to directly address the generality of PPs among selected MN populations, and whether PP-generation is task specific. Similarly, there is no evidence at the cellular-molecular level as to why the MNs of a given motor nucleus may or may not generate PPs. Therefore these important functional issues remain open.
Relation between current and previous results on MN behavior in reduced preparations
In a study on the paralyzed high-decerebrate cat
(Brownstone 1989; Brownstone et al.
1992), depolarizing current was intracellularly injected into
spinal extensor MNs during the active phase of their fictive locomotor
cycle (i.e., as brought on by repetitive brain stem stimulation). In
five of six MNs there was no further increase in spike frequency with
increasing stimulus strength; i.e., the f/I slope became
zero. This controversial finding, and the surprisingly limited directly
relevant subsequent studies (i.e., see ad seriatim: Bennett et
al. 1998a,b; Brownstone et al. 1994;
Edwards et al. 1997; Fedirchuk et al.
1998; Heckman and Lee 1999a; Lee and
Heckman 1998a,b; Schmidt 1994) have been
reviewed elsewhere (Hornby 2000). It is sufficient here
to point out that the rate-limiting behavior observed by
Brownstone et al. (1992) has been confirmed recently by
Lee and Heckman (1998a), as described above. Notably,
such rate-limiting behavior is not readily observed in anesthetized, spinalized cat preparations, probably because general anesthetics reduce PPs (Guertin and Hounsgaard 1999). The present
Figs. 4 and 5 add further evidence in support of the original
Brownstone et al. (1992) observations on this
rate-limiting phenomenon, particularly if their results are explained
by their argument that their MNs exhibited PPs during controlled
fictive locomotion (see also Brownstone et al. 1994). It
seems likely that the several mechanisms proposed above for the
flattening of the posttransition f/I slope could all come
into play during elaboration of fictive locomotion in the
high-decerebrate cat.
In conclusion, the present study showed that modulation of turtle MNs
dramatically altered their fundamental input-output relation,
particularly during PP generation. Still open are the issues of the MN
type that exhibited rate-limiting behavior in the present study and the
previous one of Brownstone et al. (1992). Was it the
low-threshold group, as favored by Heckman and Lee (1999b) and suggested indirectly by the human motor unit
results of Monster and Chan (1977)? Conversely, was it
the high-threshold group, as suggested by our Fig. 5 results (see also
Bennett et al. 1998a,b; Hultborn 1999)?
For technical reasons (McDonagh et al. 1999b), the
latter were the most likely ones studied by Brownstone et al.
(1992).
| |
ACKNOWLEDGMENTS |
|---|
We thank Drs. R. F. Fregosi, A. J. Fuglevand, T. M. Hamm, and R. B. Levine for reviewing this dissertation, Drs. C. J. Heckman and R. H. Lee for reviewing earlier drafts of this article, and P. Pierce for technical and editorial help.
This project was supported in part by National Institutes of Health Grants NS-20577 and NS-07309 to D. G. Stuart, NS-01686 to J. C. McDonagh, and GM-08400 to Dr. W. H. Dantzler; The University of Arizona Small Grants Program and College of Medicine Dean's Research Council to J. C. McDonagh; and a Flinn Foundation Fellowship and an Award from the American Psychological Association Minority Program in Neuroscience to T. G. Hornby. This paper's contents are solely the responsibility of the authors and do not necessarily represent the views of the awarding agencies.
| |
FOOTNOTES |
|---|
Address for reprint requests: D. G. Stuart, Dept. of Physiology, The Univ. of Arizona College of Medicine, Tucson, AZ 85724-5051 (E-mail: dgstuart{at}u.arizona.edu).
Present addresses: T. G. Hornby, Dept. of Physical Medicine and Rehabilitation, Northwestern University, 345 East Superior, Chicago, IL 60611; J. C. McDonagh, Arizona School of Health Sciences, 5850 E. Still Circle, Mesa, AZ 85206; R. M. Reinking, Program in Applied Mathematics, The University of Arizona, Tucson, AZ 85721-0089.
Received 5 July 2001; accepted in final form 20 February 2002.
| |
REFERENCES |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
), and following application
of 1 or more excitatory modulators (test, 
) that produced
plateau potential (PP) behavior in many of the cells. Criteria for
selection of the data points
Imin-fmin,
Itrans-ftrans,
and Imax-fmax are
described in the text. In this and several subsequent figures, the
measurement (and visualization) of the 1-phase I-f
relation involved determining the slope of a straight line connecting
Imin-fmin to
Imax-fmax.
[Note: for simplicity, the mean I-f relations in this
and all subsequent figures are drawn with lines connecting the relevant
pairs of points. Therefore as in our previous work (McDonagh et
al. 1998a
) conditions. In each plot, the extreme
data points are for
Imin-fmin and
Imax-fmax, as
provided above in Fig. 
